Mechanism of Growth

Griffith University, Gold Coast, Australia
Explaining the
Mechanism of Growth
in the Past Two
Million Years
The generally accepted fundamental
postulates are contradicted by data
Prof. Ron W. Nielsen

(aka Jan Nurzynski)
2017
Explaining the Mechanism of
Growth in the Past Two
Million Years
The generally accepted fundamental postulates are
contradicted by data
Prof. Ron W. Nielsen

(aka Jan Nurzynski)
Griffith University, Gold Coast, Qld, 4222. Australia
ronwnielsen@gmail.com
2017
________________________________________________________
How to cite this document: Nielsen, R. W. (2017). Explaining the Mechanism of Growth in the
Past Two Million Years: The generally accepted fundamental postulates are contradicted by
data. [Insert here the Website address to the pdf file.]
About the Author: http://home.iprimus.com.au/nielsens/ronnielsen.html

Abstract: Economic growth and the growth of human population in the past 2,000,000 years
are extensively examined. Data are found to be in a clear contradiction of the currently accepted
explanations of the mechanism of growth, which revolve around two fundamental but incorrect
doctrines: (1) the doctrine of stagnation (inappropriately labelled also as Malthusian stagnation,
because Malthus never claimed that his positive checks would cause a long-lasting and wide-
spread stagnation) and (2) the doctrine of explosion described also as a takeoff, sprint, spike or
by other similar attributes. These doctrines and other related postulates are contradicted even
by precisely the same data, which are used in the economic research and by the research results
published in a prestigious scientific journal as early as in 1960. The generally accepted
explanations are not based on a rigorous analysis of data but on impressions created by the
easily misleading features of hyperbolic distributions. Two leading theories: the Demographic
Transitions Theory (or Model) and the Unified Growth Theory are fundamentally incorrect.
Descriptions of the past socio-economic conditions are not questioned. They might have been
harsh, difficult and primitive but they are not reflected in the growth trajectories. They did not
create stagnation in the economic growth and in the growth of population. Likewise, impacts
of the Industrial Revolution on many aspects of life are not questioned. It is only demonstrated
that this event had absolutely no impact on shaping growth trajectories. A general law of growth
is formulated and used to explain the mechanism of growth of human population and of
economic growth. The growth was predominantly hyperbolic. Such a growth is described by
exceptionally simple mathematical function and the explanation of the mechanism of growth
turns out to be also simple.
RON W. NIELSEN, 2017, Explaining the Mechanism of Growth in the Past Two Million Years
__________________________________________________________________________________
Contents
Preface. ii
1. Scientifically unacceptable established knowledge in demography and in economic

research................................ 1
2. Changing the direction of the economic and demographic research... 25
3. Mathematical analysis of the historical economic growth with a search for takeoffs from
stagnation to growth 45
4. Growth of the world population in the past 12,000 years and its link to the economic
growth.. 65
5. Demographic Transition Theory contradicted by data 75
6. Unified Growth Theory contradicted by the absence of takeoffs in the Gross Domestic
Product..................................................... 89
7. The postulate of the three regimes of economic growth contradicted by

data........................... 101
8. Unified Growth Theory contradicted by the mathematical analysis of the historical

growth of human population................................ 124
9. Puzzling properties of the historical growth rate of income per capita

explained...... 143
10. Mathematical analysis of the historical income per capita distributions. 155
11. The unresolved mystery of the great divergence is solved.. 172
12. The dichotomy of Malthusian positive checks: Devastation and regeneration but not
stagnation. 194
13. Industrial Revolution did not boost the economic growth and the growth of population
even in the United Kingdom 217
14. Mathematical analysis of income per capita in the United Kingdom.. 230
15. Interpretations of hyperbolic growth. 241
16. The law of growth 270
17. Mechanism of hyperbolic growth explained 279
18. Population and Economic Growth in Australia: 8,000 BC AD 1700 and Earlier. 294
19. Puzzling Features of Income per Capita Distributions Explained 308
20. Economic Growth and the Growth of Human Population in the Past 2,000,000
Years.. 323
21. Demographic Catastrophes Did Not Shape the Growth of Human Population or Economic
Growth.. 344
i
__________________________________________________________________________________
Preface
The spontaneous and unconstrained economic growth and the growth of human population is
hyperbolic. Unfortunately, this type of growth, if not rigorously analysed, can easily lead to incorrect
conclusions and even experienced researchers can be easily mislead by its features.
Hyperbolic growth is slow over a long time and fast over a short time but it is still the same,
monotonically increasing, growth. Its interpretations should never be based on dividing it into two or
three distinctly different components governed by distinctly different mechanisms. Unfortunately, this
mistake is repeatedly demonstrated in scientific literature. The slow growth is interpreted as stagnation
and the fast growth as explosion. Different mechanisms of growth are suggested for the two conceived
but non-existing parts. They belong to the same, monotonically increasing, trajectory.
Hyperbolic distributions have to be interpreted as a whole. The same mechanism should be applied to
the slow growth and to the fast growth because they belong to the same distribution. If we apply the
mechanism of stagnation to the slow growth, we should apply precisely the same mechanism to the fast
growth, which would be obviously incorrect. Likewise, if we apply the mechanism of explosion to the
fast growth, we should also apply the same mechanism to the slow growth.
Even though hyperbolic distributions appear to be made of two distinctly different components, it is
impossible to locate on them a transition from the slow growth to the fast growth. There is no transition,
or more precisely, the transition occurs slowly over the whole range of hyperbolic distribution.
Unfortunately, current interpretations of the mechanism of economic growth and of the growth of
population are not based on the rigorous analysis of data by on the incorrect and misleading impressions
created by hyperbolic distributions. These distributions have to be carefully and methodically analysed.
Rigorous analysis of data point to a fundamentally serious problem with the currently accepted
interpretations of the economic growth and of the growth of population. It is not just some minor
corrections or adjustments that have to be made but rather it is the need of turning the research around
and basing it on radically new concepts. The fundamental doctrines have to be abandoned and replaced
by the evidence-based interpretations. They have to be based on accepting hyperbolic growth.
I am presenting here a compilation of my articles aimed at correcting the current misconceptions about
the economic growth and about the growth of population. The presented chapters are arranged
approximately in the order of the publication of the relevant articles but there are some exceptions.
Notably, Chapter 2 represents the last publication in these series. It is it presented at the beginning of
this compilation because it contains a convenient overview of conclusions based on this extensive
research. All these chapters explain why economic and demographic research has been on the wrong
track for a long time and why it is unavoidable to turn it around and to follow new directions as indicated
by data. This is how science works. Even the most cherished and the best-established doctrines have to
be abandoned if they are contradicted by data, and the sooner it is done the better.
A guide to chapters
Chapter 1.
Reasons for this series of investigations.
Hyperbolic illusions.
Examples of scientifically unacceptable conjectures.
Chapter 2.
Method of reciprocal values.
First examples showing how the Unified Growth Theory (Galor, 2005, 2011) is contradicted
by data.
An overview of the results presented in the remaining chapters.
Chapters 3 and 6
Analysis of the global and regional growth of the Gross Domestic Product (GDP) during the
AD era.
ii
__________________________________________________________________________________
Economic growth was hyperbolic.

There was no stagnation and no Malthusian trap in the economic growth.
There were no takeoffs at any time to a distinctly faster growth.
Industrial Revolution did not shape growth trajectories.
Unified Growth Theory (Galor, 2005, 2011) contradicted by the same data, which were used
during its development.
Chapter 4
Growth of the world population in the past 12,000 years was hyperbolic. The early observation
of von Foerster, Mora and Amiot (1960) for the AD era have been now confirmed and extended
to the BC era.
There was no stagnation in the growth of the world population and no transition from stagnation
to growth.
What is commonly interpreted as the population explosion is just the natural continuation of
the hyperbolic growth.
There was only one major transition during that time but it was not the usually claimed
transition from stagnation to growth but from hyperbolic growth to hyperbolic growth.
There was only one example when demographic catastrophes might have had impact on
shaping the growth trajectory. It was a unique event when there was an unusual convergence
of five major demographic catastrophes. However, their combined impact was small and short-
lasting.
Chapter 5
Demographic Transitions Theory is contradicted repeatedly by data, even by the same data,
which are used in its defence.
Fluctuations in birth and death rates, as well as fluctuations in the growth rate, cannot be used
as the evidence of the so-called Malthusian stagnation because they do not shape the population
growth trajectories.
Chapter 6
The concept of differential takeoffs in the economic growth proposed by Galor (2005, 2008,
2011, 2012) is examined. Differential takeoffs never happened.
Industrial Revolution had absolutely no impact on shaping economic growth trajectories, global
or regional
Chapter 7
Three regimes of growth postulated by Galor (2005, 2011) did not exist. His postulate is based
on his habitually distorted presentations of data (Maddison, 2001). Precisely the same data,
when analysed, contradict his fundamental postulates.
Industrial Revolution had no impact on shaping economic growth trajectories.
Chapter 8
Growth of population, global and regional, during the AD era was hyperbolic.
There was no stagnation in the growth of population and no transition from stagnation to
growth.
Unified Growth Theory is contradicted by the same data, which were used but not analysed
during its formulation.
Chapter 9
Puzzling properties of the growth rate of income per capita, described by Galor as a mystery of
the growth process (Galor, 2005), are explained. This mystery was created by his habitually
distorted presentation of data.
The growth rate of income per capita (GDP/cap) was increasing monotonically. There was no
sudden spike claimed by Galor.
iii
__________________________________________________________________________________
Industrial Revolution had no impact on shaping the GDP/cap trajectory.

Chapter 10
Gross Domestic Product per capita (GDP/cap), global and regional, was increasing
monotonically by following the linearly-modulated hyperbolic distributions.
There was no stagnation in the growth of income per capita and no takeoffs from stagnation to
growth.
Industrial Revolution had absolutely no impact on changing economic growth trajectories.
There was no boosting in the economic growth. On the contrary, shortly after the Industrial
Revolution, economic growth was diverted to slower trajectories.
Chapter 11
The following claim is examined: The mind-boggling phenomenon of the Great Divergence
in income per capita across regions of the world in the past two centuries, that accompanied the
take-off from an epoch of stagnation to a state of sustained economic growth, presents
additional unresolved mysteries about the growth process (Galor, 2005, p. 220).
This single sentence is overflowing with misinformation.
The great divergence never happened. This feature was created by the habitually distorted
presentations of data.
The epoch of stagnation did not exist.
There was no takeoff from stagnation to growth but a monotonically increasing hyperbolic
growth.
Various regions are on different levels of economic growth but they never diverged to distinctly
different trajectories.
It is shown that we can take any set of hyperbolic distributions and construct a great divergence
by selecting a few strategically located points on these distributions. This constructed great
divergence has nothing to do with the mechanism of the economic growth but reflects only the
distortions introduced to the purely mathematical distributions.
Chapter 12
Effects of Malthusian positive checks (Malthus, 1798) are investigated using the UN data
(UNDP, 2011).
His observations are confirmed.
Malthusian positive checks are responsible for increasing the death rate but they also trigger a
process of regeneration.
Malthusian positive checks do not produce stagnation but stimulate growth.
Chapter 13
Using the data of Maddison (2010) it is shown that Industrial Revolution had absolutely no
impact on shaping the trajectories describing economic growth and the growth of population in
the United Kingdom, the centre of this revolution.
Chapter 14
The fine structure of growth trajectories is explained.
It is incorrect to claim that fluctuations in the growth rate can be taken as the manifestation of
the so-called Malthusian stagnation (see for instance Artzrouni & Komlos, 1985; Lagerlf,
2003) because these fluctuations have no impact on shaping growth trajectories.
Chapter 15
Various attempts of explaining the mechanism of hyperbolic growth are discussed.
iv
__________________________________________________________________________________
Chapter 16
The general law of growth is formulated. This law links directly the force of growth with the
growth trajectories and thus allows for an easy way of studying the mechanism of growth.
Chapter 17
The mechanism of hyperbolic growth of population and of economic growth is explained.
Chapter 18
The growth of the ancient population in Australia is examined because a sudden transition from
a slow to a fast growth, described as the intensification of growth, was claimed by Johnson and
Brook (2011).
It is demonstrated that the claimed intensification never happened. The growth of population in
Australia was increasing monotonically.
The claimed intensification represents a typical mistake made with the interpretation of
hyperbolic distributions but this time it was made in a discipline, which is outside the
mainstream of the demographic research.
Chapter 19
Puzzling properties of income per capita (GDP/cap) are explained. They are purely
mathematical properties of dividing two hyperbolic distributions.
The GDP/cap data are described by the linearly modulated hyperbolic distributions.
Chapter 20
Growth of population was hyperbolic not just in the past 12,000 years as demonstrated in
Chapter 4, but also in the past 2,000,000 years.
This analysis confirms the earlier observation of Deevey (1960) that the growth of population
over such a long time was in three stages. However, Deevey expected that each stage was
leading to an equilibrium. Data show diametrically different shapes: each stage was represented
by a hyperbolic growth, which if continued would increase to infinity at a fixed time.
The commonly claimed epoch of stagnation never existed. There was no stagnation in the
growth of population in the past 2,000,000 years.
The repeatedly claimed takeoffs from stagnation to growth never happened because there was
no stagnation. There were only transitions from hyperbolic growth to hyperbolic growth.
Hyperbolic growth was remarkably stable. There were only two major demographic transitions
in the past 2,000,000 but they were not transitions from stagnation to growth but from
hyperbolic growth to hyperbolic growth.
Using the characteristic properties of income per capita, mathematical description of the
economic growth is also extended over the past 2,000,000 years. The pattern is the same as for
the growth of population.
Chapter 21
A survey of demographic catastrophes shows that individually they were too weak to have a
significant impact on the growth of human population.
Contrary to the established knowledge, demographic catastrophes did not shape the growth of
population and the economic growth.
New directions
1. There is no need to explain the mechanism of the so-called Malthusian stagnation because there
was no stagnation in the economic growth and in the growth of human population. The concept of
stagnation is inapplicable to the description and to the explanation of the mechanism of the
economic growth and of the growth of human population.
2. The concept of the so-called Malthusian stagnation was not proposed by Malthus and neither was
the concept of the so-called Malthusian trap. Malthus has never claimed any form of a prolonged
v
__________________________________________________________________________________
and widespread stagnation in the growth of population or in the economic growth. These concepts
were erroneously introduced much later and are incorrectly linked with Malthus.
3. Malthus noticed the dual effects of his positive checks: they increase the rate of mortality but they
also stimulate growth (Nielsen, 2016a, Malthus, 1798). He also suggested that this stimulating
effect should be further investigated. The combined impact of Malthusian positive checks is that
they increase the growth rate. Malthusian positive checks do not cause stagnation but stimulate
growth.
4. Discussions of the past socio-economic conditions might be interesting but there is no obvious
connection between them and the growth trajectories describing economic growth and the growth
of population. Growth trajectories are immune to the changing socio-economic conditions. They
cannot explain the mechanism growth.
5. There is no need to explain the escape from the so-called Malthusian trap because there was no
trap in the economic growth and in the growth of population.
6. There is no need to explain the transition from stagnation to growth because there was no such
transition. The growth was hyperbolic in the past and continued to be hyperbolic at the time of the
alleged transitions. The alleged transitions represent the natural continuation of the monotonically
increasing hyperbolic distributions.
7. Hyperbolic growth was remarkably stable. There were only two major transitions in the past
2,000,000 but they were not transitions from stagnation to growth but from hyperbolic growth to
a new hyperbolic growth.
8. There is no need to explain the boosting effects of the Industrial Revolution because Industrial
Revolution had no impact on shaping the economic growth trajectories and the trajectories
describing the growth of population.
9. There is no need to explain the mechanism of takeoffs because they never happened.
10. There is no need to explain the differential timing of takeoffs (Galor, 2005, 2011) because there
were no takeoffs.
11. There is no need to explain the numerous mysteries of growth listed by Galor in his Unified Growth
Theory because there are no such mysteries. They were created by a repeatedly distorted
presentations of data. Analysis of the same data shows that there were no mysteries.
12. There is no need to explain the mind boggling phenomenon of the Great Divergence (Galor,
2005, p. 220) because the great divergence never happened. It was again created by a distorted
presentation of data. Any set of monotonically increasing hyperbolic distributions can be used to
create great divergence when they are suitably distorted as repeatedly done in the Unified Growth
Theory and in other similar publications. Economic growth in various regions is now on different
levels of development but they all follow closely similar trajectories governed by the same
mechanism of growth.
13. There is no need to explain the sudden spike (Galor, 2005, 2011) in the growth rate of income per
capita because there was no spike.
14. There is no need to explain a transition from a nearly horizontal to a nearly vertical growth of
income per capita (GDP/cap) because there was no transition. These perplexing and puzzling
features have now been explained: they represent nothing more than just the mathematical property
of dividing two hyperbolic distributions but they also represent the monotonically increasing
distributions without any sudden transition between the two apparently different patterns of growth.
15. Demographic Transitions Theory and the Unified Growth Theory are contradicted by data and
should not be used in the demographic and economic research.
16. Demographic and economic research has to be based on accepting that the historical growth of
population and the economic growth were following hyperbolic distributions.
vi
__________________________________________________________________________________
17. Growth of population and economic growth were predominantly hyperbolic in the past 2,000,000
years and, as first noticed by Deevey (1060), it was in three major staged. Economic growth
followed a similar pattern.
18. The mechanism of the hyperbolic economic growth and of the growth of population has been now
explained (Nielsen, 2016b). Hyperbolic growth is the simplest unconstrained growth prompted by
the fundamental forces: the indispensable force of procreation in the case of the growth of
population and the simplest market force in the case of the economic growth. These were the
dominating forces in the past.
19. Demographic catastrophes did not shape the growth of population or the economic growth. The
only example when they might have caused a minor and insignificant disturbance in the growth
trajectory was when there was an unusual and unique combination of five major demographic
catastrophes.
20. The currently accepted interpretations of the mechanism of economic growth and of the growth of
population are not only in conflict with data used in the economic research but they are also
dangerously misleading because they create the sense of security by claiming that after a long stage
of stagnation, extending over thousands of years, we have finally escaped the Malthusian trap and
entered into a sustained growth regime. Scientific evidence shows that the opposite is true. The
past growth was stable, secure and sustainable, as demonstrated by the largely stable hyperbolic
distributions. In contrast, it is the current growth, which is potentially unsustainable. For the first
time in human existence, our ecological footprint is larger than ecological capacity and it continues
to increase. For the first time in human history we support our existence on the continually
increasing ecological deficit.
21. Conjectures and inspiration are acceptable in science but they have to be controlled and moderated
by data. When conjectures are supported by conjectures, the created system of doctrines and
explanations becomes quickly scientifically untenable.
References
Artzrouni M., & Komlos J. (1985). Population Growth through History and the Escape from the Malthusian Trap: A
Homeostatic Simulation Model. Genus, 41, 21-39. (PMid:12280407)
Deevey, E. S. Jr (1960). The human population. Scientific American, 203(9), 195-204.
Galor, O. (2005). From stagnation to growth: Unified Growth Theory. In P. Aghion & S. Durlauf (Eds.), Handbook of
Economic Growth (pp. 171-293). Amsterdam: Elsevier.
Galor, O. (2008). Comparative Economic Development: Insight from Unified Growth Theory.
http://www.econ.brown.edu/faculty/Oded_Galor/pdf/Klien%20lecture.pdf
Galor, O. (2011). Unified Growth Theory. Princeton, New Jersey: Princeton University Press.
Galor, O. (2012). Unified Growth Theory and Comparative Economic Development.
http://www.biu.ac.il/soc/ec/students/mini_courses/6_12/data/UGT-Luxembourg.pdf
Johnson, C. N., & Brook, B. W. (2011). Reconstructing the dynamics of ancient human populations from radiocarbon dates:
10 000 years of population growth in Australia. Proceedings of the Royal Society B, 278, 3748-3754.
Lagerlf, N-P. (2003). Mortality and early growth in England, France and Sweden. Scand. J. of Economics, 105(3), 419
439.
Maddison, A. (2001). The World Economy: A Millennial Perspective. Paris: OECD.
Maddison, A. (2010). Historical Statistics of the World Economy: 1-2008 AD. http://www.ggdc.net/maddison/Historical
Statistics/horizontal-file_02-2010.xls.
Malthus, T. R. (1798). An Essay on the Principle of Population. London: J. Johnson.
Nielsen, R. W. (2016a). The dichotomy of Malthusian positive checks: Destruction and even more intensified regeneration.
Journal of Economic Bibliography, 3(3), 409-433. http://www.kspjournals.org/index.php/JEB/article/view/934
Nielsen, R. W. (2016b). Mechanism of Hyperbolic Growth Explained. Journal of Economic Library. 3(4), 411-428.
http://kspjournals.org/index.php/JEL/article/view/1074
UNDP (2011). Human development report 2011 Sustainability and equity: A better future for all. New York: United Nations
Development Program.
von Foerster, H., Mora, P., & Amiot, L. (1960). Doomsday: Friday, 13 November, A.D. 2026. Science, 132, 255-296.
Ron W. Nielsen
Gold Coast, Australia
September, 2017
vii
RON W. NIELSEN, 2017, Major Revision of Fundamental Postulates
__________________________________________________________________________________
1
Scientifically unacceptable established knowledge in
demography and in economic research
By Ron W. NIELSENa
Abstract. The established knowledge in demography and in the economic research is based on the concept of the
so-called Malthusian stagnation and on the associated concept of the escape from the Malthusian trap. These two
fundamental concepts were gradually enforced by numerous other related postulates all aimed at explaining the
mechanism of the historical growth of population and of the historical economic growth. Examples of publications
based on the established knowledge are closely examined. They are used to show why the established knowledge
is scientifically unacceptable. It is also pointed out that the established knowledge is contradicted by data and by
their analyses. Interpretations of the historical economic growth and of the historical growth of population has to
be based on accepting hyperbolic growth. However, the discussed examples point to a more serious problem in
these two fields of research. It is a fundamental systemic problem, the problem associated with the way research
is conducted. Doctrines, interpretations and declarations used by the established knowledge have to be often
accepted by faith. Data are either ignored or manipulated to support preconceived ideas. Contradicting evidence
is methodically ignored. To be recognised as science, demographic and economic research has to adhere to the
scientific rules of investigation.
Keywords. Economic growth, Population growth, Gross Domestic Product, Hyperbolic growth, Malthusian
stagnation, Malthusian trap, Malthusian positive checks, Malthusian oscillations, Fertility rate, Mortality rate,
Famines, Pestilence, Wars
JEL. A10, A12, A23, B22, B41, C12, Y80.
1. Introduction
T wo fields of research, economic growth and the growth of population, which might appear to be
distinctly different, are in fact closely related for at least three reasons. First, there is obviously
no economic growth without humans. Second, there is a close correlation between economic
growth and the growth of human population (Nielsen, 2016a, 2016b). Third, in order to understand the
growth of income per capita, measured by the Gross Domestic Product per capita (GDP/cap), it is
obviously necessary to study not only the economic growth but also the growth of human population.
It is inter alia for these reasons, that the best source of information about the historical economic growth,
compiled by the world-renown economist, includes not only the data describing the growth of the GDP
but also the growth of population (Maddison, 2001, 2010).
2. The established knowledge

The established knowledge in demography and in the economic research revolves around two
fundamental concepts: the so-called Malthusian stagnation and the explosion, which is supposed to have
marked a dramatic escape from the so-called Malthusian trap. Gradually and by accretion, in the process
extending over many years, these two fundamental concepts were adorned by various additional
explanations, speculations and conjectures all adding to the now established knowledge based on the
scientifically unacceptable doctrines and beliefs. These two fundamental regimes of growth, stagnation
and explosion, are described as Stage 1 and Stage 2, respectively, in the Demographic Transition Theory
a
AKA Jan Nurzynski, Griffith University, Environmental Futures Research Institute, Gold Coast Campus, Qld, 4222,
Australia.
. +61407201175
. ronwnielsen@gmail.com
Published as: Nielsen, R. W. (2016). Scientifically unacceptable established knowledge in demography and in economic
research. Journal of Economic Library, 3(3), 429-457.
1
__________________________________________________________________________________
(see Nielsen, 2016c and references therein). The epoch of stagnation was supposed to have lasted for
many thousands of years and was allegedly strongly controlled by the Malthusian positive checks
(Malthus, 1798) generating an unstable stage of growth characterised by irregular Malthusian
oscillations. The mechanism of growth is claimed to have changed dramatically at the time of the
alleged population explosion when the growth was supposed to have changed from slow to fast. The
transition from stagnation to explosion is described as the great escape from the Malthusian trap.
We have already demonstrated that the established knowledge is convincingly contradicted by the
relevant data and by their analyses (Biraben, 1980; Clark,1968; Cook,1960; Durand, 1974; Gallant,
1990; Haub, 1995; Kapitza, 2006; Kremer, 1993; Lehmeyer, 2004; Livi-Bacci, 1997; Maddison, 2001,
2010; Mauritius, 2015; McEvedy & Jones, 1978; Nielsen, 2014, 2015a, 2016a, 2016b, 2016c, 2016d,
2016e, 2016f, 2016g, 2016h, 2016i; Podlazov, 2002; Shklovskii, 1962, 2002; Statistics Mauritius, 2014;
Statistics Sweden, 1999; Taeuber & Taeuber, 1949; Thomlinson, 1975; Trager, 1994, United Nations,
1973, 1999, 2013; von Hoerner, 1975, von Foerster, Mora & Amiot, 1960; Wrigley & Schofield, 1981).
The aim of this publication is (1) to outline briefly the origin of the established knowledge, (2) to explain
why the established knowledge is so strongly established, (3) to explain the deceptive evidence in data,
which can be used in support of the established knowledge, (4) to give a few examples of how strongly
the established knowledge is established and (5) to explain why the established knowledge as illustrated
by these examples is scientifically unacceptable.
3. Evidence in data
Data describing the historical growth of population and the historical economic growth are hardly
ever analysed. Recently, attempts were made to use some of these data (Maddison, 2001) but they were
presented in grossly distorted and misleading diagrams, which appear to be supporting the established
knowledge (Ashraf, 2009; Galor, 2005a, 2005b, 2007, 2008a, 2008b, 2008c, 2010, 2011, 2012a, 2012b,
2012c; Galor & Moav, 2002; Snowdon & Galor, 2008). Data were not analysed to learn from them but
manipulated to support preconceived ideas. Such approach to research is scientifically unacceptable.
Data have to be carefully and methodically analysed to avoid drawing incorrect conclusions. Their
superficial examination creates strong impression of stagnation followed by explosion but when closely
analysed they show that the apparent explosion was just the natural continuation of the past hyperbolic
growth.
Global population in 10,000 BC is estimated at only between 1 and 10 million (McEvedy & Jones,
1978, Thomlinson, 1975). Now the population of this size can be located in just a single city. By AD 1,
global population increased to only a few hundred million. The estimated values vary between 170 and
400 million (Biraben, 1980; Durand, 1974; Haub, 1995; McEvedy & Jones, 1978; Thomlinson, 1975;
United Nations, 1973, 1999). Now, the population of this size or even larger can be found in just a
single country.
The first billion of global population was reached around AD 1800 (Biraben, 1980; Durand, 1974;
McEvedy & Jones, 1978; Thomlinson, 1975; United Nations, 1973, 1999) and from that time on the
growth was progressing exceedingly fast. The origin of Homo Sapiens is usually claimed at around
200,000 years ago, but it might have been even earlier (Weaver, Roseman & Stringer, 2008). Thus, it
took many thousands of years for the world population to increase to one billion but after reaching the
first billion, the second billion was added in just only about 130 years (United Nations, 1999). The
process of many hundreds of thousands of years was suddenly compressed to just over 100 years. The
consumption of natural resources and the stress on the environment started to increase rapidly.
If adding one billion in just 130 years sounds too fast, the next billion was added in just 29 years,
the next in 15 years, the next in 13 years, and the next in 12 years, increasing the size of global
population to 6 billion (US Census Bureau, 2016). The last billion, which increased global population
to 7 billion, was added in 13 years (US Census Bureau, 2016). We call it the slowing-down growth but
obviously the slowing down process is still too slow.
Assuming a medium-intensity growth, the size of the world population is projected to increase to
8.39 billion in 2030 and 9.63 billion in 2050 reaching a maximum of 10.48 billion around 2080 (Nielsen,
2006). These projections are in good agreement with the US Census Bureau (2016) projections of 8.34
billion in 2030 and 9.41 billion in 2050. It is what we hope for, but the high intensity growth could lead
to 12.26 billion by the end of the current century (Nielsen, 2006), assuming that such a growth can be
supported by the availability of natural resources.
2
__________________________________________________________________________________
Similar surprising pattern of a slow growth in the past and a fast growth in recent years is reported
for the growth of the Gross Domestic Product (Maddison, 2001, 2010). The first trillion dollars ($1012)
of the GDP (expressed in the 1990 international Geary-Khamis dollars) was reached in 1870. The next
trillion was added in just 51 years, the next in 19 years and the next in only 10 years, increasing global
GDP to $4 trillion in 1950. By 1998, global GDP increased to $34 trillion. The latest estimate for 2014
is $91 trillion (World Bank, 2016) and the projected value for 2050 is $118 trillion (Nielsen, 2015b).
Using such numbers, it would be easy to conclude that there was a long epoch of stagnation in the
past economic growth and in the growth of human population and that this stagnation was followed by
a sudden explosion. However, such a conclusion, which is the corner stone of the established knowledge
in demography and in the economic research, would be unscientific because impressions can be
misleading. Scientific research has to be conducted scientifically. If economic and demographic
research is supposed to be recognised as science they have to adhere to the scientific rules of
investigation.
In science, data have to be methodically analysed. This fundamental requirement in scientific
research appears to have been ignored in economic and demographic research. Hasty conclusion about
stagnation followed by explosion is also clearly incorrect and scientifically unacceptable because over
50 years ago, von Foerster, Mora and Amiot, (1960) demonstrated that the growth of population during
the AD era was hyperbolic. This crucial contribution to science should not have been ignored. It should
have been further investigated because hyperbolic growth rules out the interpretations based on the
assumption of stagnation followed by explosion.
Postulates of the established knowledge are also unacceptable because hyperbolic growth have been
recognised and confirmed by other independent investigations (Kapitza, 2006; Kremer, 1993; Podlazov,
2002; Shklovskii, 1962, 2002; von Hoerner, 1975). Accepting the fundamental postulates of established
knowledge is scientifically unjustified because for a long time now there was a large body of data
describing the growth of population not only during the AD era but also during the BC era (Biraben,
1980; Clark,1968; Cook,1960; Durand, 1974; Gallant, 1990; Haub, 1995; Livi-Bacci, 1997; McEvedy
& Jones, 1978; Taeuber & Taeuber, 1949; Thomlinson, 1975; Trager, 1994, United Nations, 1973,
1999, 2013). These data should have been analysed to check the earlier claims about the hyperbolic
growth.
Fundamental postulates of the established knowledge are now contradicted by the excellent new data
describing economic growth and the growth of population (Maddison, 2001, 2010). These postulates
are scientifically unacceptable because they are consistently contradicted by the analysis of relevant
data (Nielsen, 2013a, 2013b, 2013c, 2014, 2015a, 2016a, 2016b, 2016c, 2016d, 2016e, 2016f, 2016g,
2016h, 2016i).
Data describing birth and death rates and the associated growth of population are limited (Lehmeyer,
2004; Mauritius, 2015; Statistics Mauritius, 2014; Statistics Sweden, 1999; Wrigley & Schofield, 1981)
but they also show consistently that the established knowledge, as expressed in the Demographic
Transition Theory, is contradicted by their analysis (Nielsen, 2016c). We do not even have to analyse
these data mathematically to see that they are in contradiction of the established knowledge because
even though the birth and death rates and the associated growth rates were fluctuating, their time-
dependence does not fit into the patterns claimed by the Demographic Transition Theory. Furthermore,
the corresponding distributions describing the growth of population do not display any form of
stagnation during the alleged Stage 1 or a transition to the alleged Stage 2, which is supposed to
represent the explosion. Data show no such patterns.
Demographic Transition Theory is based on a persistent and blatant disregard for relevant data. This
theory is supported by largely meaningless presentations of data for birth or death rates. These rates
have to be studied together and they should show the expected behaviour, as claimed by the
Demographic Transition Theory, that the gap between them is approximately zero during the alleged
Stage 1 and that it increases during the alleged Stage 2. Such patterns are not confirmed by the best
available data (Lehmeyer, 2004; Mauritius, 2015; Statistics Mauritius, 2014; Statistics Sweden, 1999;
Wrigley & Schofield, 1981), which show that the Demographic Transition Theory is contradicted by
the data describing birth and death rates and by the associated data describing the growth of population.
Paradoxically, when methodically analysed, data used in support of the Demographic Transition Theory
are in fact in its clear contradiction.
3
__________________________________________________________________________________
A theory contradicted by just a single set of data is scientifically unacceptable and the Demographic
Transition Theory was first contradicted by the results of von Foerster, Mora and Amiot (1960) who
demonstrated that the growth of human population during the AD era was hyperbolic and thus that
Stages 1 and 2 claimed by this theory did not exist. The Demographic Transition Theory should have
been rejected or at least fundamentally modified about 50 years ago. Its continuing use over such a long
time has been scientifically unjustified.
The postulate of the so-called Malthusian stagnation followed by explosion, and all other associated
postulates and explanations of the historical economic growth and of the historical growth of population
followed by a mythical escape from the Malthusian trap have no place in science. They may, however,
have a place in the history of science.
4. Hyperbolic growth
Hyperbolic distributions are strongly deceptive and it is easy to make a mistake with their
interpretation. Fortunately, however, analysis of hyperbolic distributions is also trivially simple
(Nielsen, 2014) and it is easy to avoid making an easy mistake.
Examples of two hyperbolic distributions, a hyperbolic distribution describing the growth of the
world population during the AD era and the distribution describing the world economic growth, are
shown in Figures 1 and 2. Their analysis is based on using the method of reciprocal values (Nielsen,
2014). For a sufficiently wide range of data, hyperbolic distributions can be uniquely identified using
this method because if the reciprocal values are decreasing linearly, then the growth is hyperbolic. There
is no other option. It is something similar to the unique identification of the exponential growth. For a
sufficiently large range of good quality data, exponential growth can be uniquely identified by the linear
distribution of the logarithm of the size of a growing entity.
Figure 1. Data describing the growth of the world population (Maddison, 2010) are compared with hyperbolic
distribution.
Figures 1 and 2 show that the growth of human population and economic growth were indeed slow
over a long time, but it was hyperbolic growth, which is slow over a long time and fast over a short
time. It is still the same, monotonically-increasing, growth. It is impossible to divide such a growth into
distinctly-different components and the best way to see it, is to examine the reciprocal values of the size
of the growing entity, in our case the reciprocal values of the GDP or of the size of the population
(Nielsen, 2014).
4
__________________________________________________________________________________
Hyperbolic distributions have to be analysed and interpreted as a whole. The same mechanism has
to be applied to the slow and fast growth. If we apply the mechanism of the so-called Malthusian
stagnation to the slow growth, we have to apply precisely the same mechanism to the fast growth. If we
apply the mechanism of explosion to the fast growth, then precisely the same mechanism should be
applied to the slow growth, which obviously is incorrect because explosion has to be triggered by
something and there was clearly no explosion along the slow growth.
The usually assumed event that was supposed to have triggered population explosion or a sudden
takeoff in economic growth or in the growth of population is the Industrial Revolution but as we can
see in Figures 1 and 2, there was no sudden explosion during the Industrial Revolution or at any other
time. The growth was increasing monotonically. Transition from slow to fast growth takes place all the
time. We could demonstrate this monotonic growth even more clearly by using reciprocal values of
data or by the semilogarithmic display (Nielsen, 2014, 2016a, 2016b, 2016d, 2016e, 2016f, 2016g,
2016h, 2016i) but the primary aim of presenting these two diagrams is to illustrate the deceptive
character of hyperbolic distributions. They can easily lead to incorrect interpretations particularly when
they are not analysed but only used to quote certain, well-selected numbers or when they are deliberately
manipulated and distorted (Ashraf, 2009; Galor, 2005a, 2005b, 2007, 2008a, 2008b, 2008c, 2010, 2011,
2012a, 2012b, 2012c; Galor & Moav, 2002; Snowdon & Galor, 2008) to support preconceived ideas.
Hyperbolic distributions have to be analysed.
Figure 2. Data describing the growth of the world Gross Domestic Product (Maddison, 2010) are compared
with hyperbolic distribution.
Figure 1 shows also that the growth of population is not yet levelling off. It is still following closely
the fast-increasing historical hyperbolic distribution. Maddisons data end in 2008. The point
representing the size of the population in 2014 is from the US Bureau of Census (2016) while the last
two points are the predicted values (Nielsen, 2006). Not until 2030 or maybe even until 2050 could we
expect a clear departure from the historical hyperbolic trend. The future of the population growth is
uncertain, in much the same way as the future of the world economic growth (Nielsen, 2015b).
5. The origin of the concept of stagnation

Two features make the concept of the so-called epoch of Malthusian stagnation deceptively
attractive: (1) it is strongly believable and (2) it is supposed to have originated over 200 years ago. It is
believable because the growth of human population and the economic growth over thousands of years
were indeed slow, so slow that they appear to have been stagnant. It is also an old concept because its
5
__________________________________________________________________________________
origin is traced, to Malthus (1798), perhaps inaccurately because Malthus never used the word
stagnation in his book.
The Malthusian theory, as was outlined initially by Malthus (1978), captures the main attributes of
the epoch of Malthusian stagnation that had characterized most of human existence (Galor, 2005,
p. 221).
The idea of multiple equilibria, or poverty traps, can be retraced back to Malthus (Wang, 2005, p. 36).
The work of Malthus was the first well-documented attempt to understand and explain the
mechanism of growth of human population but it appears that this is also precisely where it ended.
Considering the time when Malthus was writing his book, it was a remarkable achievement, but his
work should have been not only checked but also extended using a large body of data, which were not
available to Malthus but which are readily available to us.
The history of population theory can be summarized in three words: pre-Malthusian, Malthusian, and
post-Malthusian. Hardly ever in intellectual history does one man so dominate a field as does the
Reverend Thomas Robert Malthus in demographic theory. To paraphrase a quotation attributed to
Newton, Malthus shoulders must be climbed (Thomlinson, 1965, p. 47. Italics in the original text.).
the demographic transition experiences three regimes: the Malthusian Regime, the Post-
Malthusian Regime, and the Modern Growth Regime. Any theory attempts (sic) to describe the
process of demographic transition must include these three periods (Wang, 2005, p. 3. Italics added.).
Claiming, suggesting or assuming that something must be accepted just because it comes from a
certain source is not acceptable in science. Any theory can be questioned and even should be questioned,
and if necessarily corrected or rejected. The sooner it is done, the better it is for science. If Malthuss
shoulders must be climbed it is only for the same reason as climbing the shoulders of any giant of human
intellect: to see better and further ahead. It is not just to have a comfortable ride.
However, we are not even climbing Malthuss shoulders. Attaching his name to the concept of
stagnation and calling it Malthusian stagnation sounds like defamation. It is questionable whether
Malthus would be pleased with such a dubious distinction. We are putting our interpretations into his
work and we are claiming that he did it.
If we read his publication carefully, we can find that he was writing not only about the destructive
or impeding effects of positive checks but also about their stimulating effects (Nielsen, 2013b; Malthus,
1798). He even suggested that these stimulating effects should be further investigated. Given enough
time he would have probably studied this issue further. Discussions of the impeding effects of positive
checks, which we label inaccurately as Malthusian stagnation should be balanced by the discussion of
their stimulating effects, which Malthus mentions in his book. Malthusian positive checks cannot cause
stagnation because their combined effect is to increase the growth rate (Nielsen, 2013b).M
The phrase Malthusian stagnation is a misnomer because Malthus never claimed that positive
checks would produce prolonged and wide-spread stagnations in the growth of population and because
we know now that Malthusian positive checks, even if present, were not producing such effects
(Kapitza, 2006; Kremer, 1993; Nielsen, 2013a, 2014, 2016b, 2016c, 2016d; Podlazov, 2002;
Shklovskii, 1962, 2002; von Hoerner, 1975, von Foerster, Mora & Amiot, 1960). They appear to have
been generally either too weak or their destructive impacts were effectively compensated by the well-
known, natural process of regeneration (Nielsen, 2013a, 2013b, 2013c).
It would be interesting to search for impacts of Malthusian positive checks on the growth of
population by investigating the growth of local populations. Generally, there appears to have been no
impact. The only known example (Nielsen, 2016d) is a minor distortion in the growth of the world
population between AD 1200 and 1400, which appears to be correlated with the convergence of five
major demographic catastrophes: Mongolian Conquest (1260-1295) with the total estimated death toll
of 40 million; Great European Famine (1315-1318), 7.5 million; the 15-year Famine in China (1333-
1348), 9 million; Black Death (1343-1352), 75 million; and the Fall of Yuan Dynasty (1351-1369), 7.5
million. In general, demographic catastrophes were too weak to disturb the growth of global population
(Nielsen, 2013c).
Looking for convincing evidence of impacts of Malthusian positive checks on the growth of
population would not be easy because we would have to demonstrate not only clear discontinuities in
the growth of population but also that these discontinuities are correlated with the records of
demographic catastrophes. We would have to know the intensity of these demographic catastrophes not
just in the number of deaths but in their relative impact. However, even then we would have to be aware
of the possibility of spurious correlations.
6
__________________________________________________________________________________
Malthus never claimed that his concepts must be accepted. On the contrary, he was open to new
ideas. Referring to himself in the third person he wrote:
If he should succeed in drawing the attention of more able men to what he conceives to be the principal
difficulty in the way to the improvement of society and should, in consequence, see this difficulty
removed, even in theory, he will gladly retract his present opinions and rejoice in a conviction of his
error (Malthus, 1798, p. viii. Italics added.)
It is interesting that Malthus used arithmetic and geometric progressions to support his arguments
but it is not certain whether he was familiar with the hyperbolic growth, let alone that he appreciated
the difference between hyperbolic and exponential (geometric) types of growth. Even now, hyperbolic
distributions are repeatedly misinterpreted and exponential growth is used to explain the growth of just
about anything.
Malthus claimed that Population, when unchecked, increases in a geometrical ratio (Malthus,
1978, p. 4). Now we know that this is not true. Population, when unchecked does not increase in a
geometrical ratio (exponentially) but hyperbolically (Kapitza, 2006; Kremer, 1993; Nielsen, 2016b,
2016d; Podlazov, 2002; Shklovskii, 1962, 2002; von Hoerner, 1975, von Foerster, Mora & Amiot,
1960).
Malthus did not base his claims on a rigorous analysis of data. If he lived long enough to have better
data, he would have probably discovered that the growth of population is not characterised by a constant
doubling time and consequently that it could not have been increasing exponentially. If he were familiar
with hyperbolic growth, he would have probably discovered that population increases hyperbolically.
However, Malthus did not live long enough, he did not have access to good data and he was probably
unfamiliar with hyperbolic growth. Those who lived after him and those who live now are more
privileged.
6. Examples of questionable claims

6.1. The alleged Law of Population
During the alleged but non-existent epoch of the so-called Malthusian stagnation, birth rates are
claimed to have been high because new generations were needed to support many tiresome and
mundane activities such as hunting, gathering, cultivating crops, caring for children and generally for
coping with harsh living conditions.
According to Classical economists, and early Neo-Classical economists as well, population size was
determined by the demand for labor. This was the Law of Population which constantly operated behind
the seemingly random variations in fertility and mortality induced by epidemic, famine, and war (Lee,
1997, p. 1063).
Claims:
1. Population size was determined by the demand for labour.
2. This is the Law of Growth.
3. This law has been accepted by Classical and early Neo-Classical economists.
4. There were seemingly random variations in fertility and mortality.
5. Random variations were caused by epidemics, famine and war.
6. This law operated constantly behind these seemingly random variations.
It is interesting how much is claimed in this single paragraph and it does not matter whether Lee
agrees with all these claims or just describes them. This quotation represents a typical set of questionable
claims often encountered in publications related to the concept of the epoch of the so-called Malthusian
stagnation. Can we prove them or do we have to accept them by faith?
To prove this Law of Population we would have to have data about the demand for labour and
about the growth of population extending over thousands of years, and we would have to prove that
there is a correlation between the demand for labour and the size of human population. We would have
to prove that population size was determined by the demand for labour. We cannot prove it because we
do not have such data, but we can show that the population data (Nielsen, 2016b, 2016d) do not display
any features that could be linked with this alleged Law of Population. This law has to be accepted by
faith but this law is also in contradiction with data and with their analysis.
It is easy to imagine and claim, without a proof, that there were random variations in the fertility and
mortality. It would be probably more difficult to expect that there were no variations but we have no
information about these variations. We can only imagine them but we cannot analyse them.
7
__________________________________________________________________________________
We have reliable data about the size of human population (Biraben, 1980; Clark,1968; Cook,1960;
Durand, 1974; Gallant, 1990; Haub, 1995; Livi-Bacci, 1997; Maddison, 2001, 2010; McEvedy & Jones,
1978; Taeuber & Taeuber, 1949; Thomlinson, 1975; Trager, 1994, United Nations, 1973, 1999, 2013)
over thousands of years but we have no matching data for fertility and mortality (birth rates and death
rates). We also have no matching data about epidemics, famines and wars to study how they were
correlated with random variations in fertility and mortality. We have absolutely no way of proving
that the Law of Population constantly operated behind the seemingly random variations in fertility
and mortality induced by epidemic, famine, and war. This claim is unscientific because we can never
expect to verify it by data but also it is scientifically unacceptable because data and their analysis give
no support for such declarations.
It should be also noted that growth of population is not determined directly by birth and death rates
but by the difference between these two quantities. This difference determines the growth rate. More
precisely, it determines the rate of natural increase but generally migrations rates are relatively small
and consequently the difference between birth and death rates can be taken as determining the growth
rate.
A constant (non-zero) difference (constant growth rate) produces exponential growth. A zero
difference produces constant population. However, variable difference between birth and death rates
(i.e. the variable growth rate) does not necessarily produce a variable size of the population. In fact,
even large fluctuations in the growth rate are not readily reflected in the growth of population. They
might be reflected only as small and negligible variations (Nielsen, 2016c).
Fluctuations in birth and death rates have no impact on the mechanism of growth because they do
not change population growth trajectories. We can see it even without analysing data. We can easily
check that even for data characterised by large fluctuations in birth and death rates, and consequently
by large fluctuations in the growth rate, the corresponding data, which describe the growth of population
are not affected by such fluctuations. Fluctuations in birth and death rates do not change the general
character of the distributions describing the growth of population (Lehmeyer, 2004; Mauritius, 2015;
Statistics Mauritius, 2014; Statistics Sweden, 1999; Wrigley & Schofield, 1981). These data are well
known. Some of them are even repeatedly used to defend the erroneous Demographic Transition Theory
but no-one cared to check the population data published in the same sources, which list the fluctuating
birth and death rates. While the fluctuating birth and death rates are taken as the confirmation of the
established knowledge, the data describing the growth of population, data coming from precisely the
same sources as the data for birth and death rates, are methodically ignored. Data describing the growth
of population are in contradiction of the Demographic Transition Theory and in contradiction of the
established knowledge.
6.2. The alleged losing battle
According to the concept of the so-called epoch of Malthusian stagnation, as soon as the population
started to increase, it was significantly reduced by numerous factors associated with severe living
conditions.
During the first [stage of the demographic transition], fertility is assumed to have been sufficiently high
to allow a population to grow slowly even in the face of a rather high level of mortality. However,
periodic epidemics of plague, cholera, typhoid and other infectious diseases would in one or two years
wipe out the gains made over decades. Over long periods of time there would, consequently, be almost
no population growth at all (van de Kaa, 2010, p. 87. Italics added.).
Claims:
1. During the first stage of the demographic transition, fertility and mortality are assumed to have
been high.
2. Population was growing slowly.
3. Population growth was strongly controlled by periodic epidemics of plague, cholera, typhoid and
other infectious diseases.
4. Periodic epidemics of plague, cholera, typhoid and other infectious diseases would in one or two
years wipe out the gains made over decades.
5. Over long periods of time there was no population growth at all.
Van de Kaa describes the first of the four stages of growth claimed by the classical Demographic
Transition Theory, the stage corresponding to the mythical but non-existent epoch of the so-called
Malthusian stagnation (Nielsen 2016b, 2016c, 2016d).
8
__________________________________________________________________________________
Here we have a vivid description of what was happening so long ago and over a long time; not only
a vivid description but also an explanation. In science, one would have to do a lot of solid work in order
to be able to make such a sweeping declaration. We would have to prove that our conclusions are
supported by data. We would have to give frequent examples that the growth of population was indeed
controlled by periodic epidemics of plague, cholera, typhoid and other infectious diseases. We would
have to demonstrate convincingly that there were frequent correlations between periodic epidemics of
plague, cholera, typhoid and other infectious diseases and the growth of population. Ideally, we would
also have to prove that these frequent irregularities were caused by periodic epidemics of plague,
cholera, typhoid and other infectious diseases because even observed correlations could be spurious.
Van de Kaa produces no such proof. He does not even give reference to such research. As far as we
can tell, no-one has ever carried out such systematic and well-documented research.
His claims have to be accepted by faith and even more importantly, by a fixated faith because they
are contradicted by data (Biraben, 1980; Clark,1968; Cook,1960; Durand, 1974; Gallant, 1990; Haub,
1995; Livi-Bacci, 1997; Maddison, 2001, 2010; McEvedy & Jones, 1978; Taeuber & Taeuber, 1949;
Thomlinson, 1975; Trager, 1994, United Nations, 1973, 1999, 2013). With only one exception in the
past 12,000 years, between AD 1200 and 1400 (Nielsen, 2016d), there is no convincing evidence of
generally occurring long periods of time when there was almost no population growth at all and
that the growth was controlled by periodic epidemics of plague, cholera, typhoid and other infectious
diseases. The only way we could hope to give support to his claims would be to find exceptions to the
generally observed regularities in the growth of population but even then, his claims would not have a
general application. The established knowledge may sound plausible and convincing but it has to be
accepted by faith.
It is scientifically incorrect to take an easy way out by assuming that something happened, which
we think could have happened and claim with such absolute certainty that it did happen. We might feel
or think that our descriptions are true; we might wish for them to be true, but we should test them by
following the generally accepted process of scientific investigation.
6.3. The alleged food-controlled homeostatic equilibrium
Harsh living conditions, and in particular the availability of food, are supposed to have a suppressive
influence on the growth of human population but these intuitive expectations are again contradicted by
data (UNDP, 2011) showing that growth rate is not directly proportional to the level of affluence but to
the level of deprivation (Nielsen, 2013b). There is also convincing evidence that harsh living conditions
in the distant past did not shape the growth of population (Nielsen, 2016b, 2016d, von Foerster, Mora
& Amiot, 1960). Again, it is scientifically inexcusable to take an easy way out, ignore data and try to
mould science in the image of our wished-for interpretations.
the food-controlled homeostatic equilibrium had prevailed since time immemorial (Komlos, 2000,
p. 320).
the population tends to oscillate in a homeostatic mechanism resulting from the conflict between the
population's natural tendency to increase and the limitations imposed by the availability of food
(Artzrouni & Komlos, 1985, p. 24).
Claims:
1. There was a food-controlled homeostatic equilibrium.
2. This equilibrium prevailed since time immemorial.
3. Population tends to oscillate in a homeostatic mechanism.
4. Oscillations are caused by the natural tendency of the population to increase and by the
limitations imposed by the availability of food.
It is easy to assume that the food-controlled homeostatic equilibrium had prevailed since time
immemorial but it is more difficult to prove it. It is easy to claim that the population tends to oscillate
in a homeostatic mechanism resulting from the conflict between the population's natural tendency to
increase and the limitations imposed by the availability of food but it is more difficult to prove it.
Authors of these confident declarations do not prove anything nor do they give reference to such a
proof because such a proof does not exist. These declarations are in harmony with the established
knowledge but the established knowledge is in conflict with science (Kapitza, 2006; Kremer, 1993;
Nielsen, 2016b, 2016d; Podlazov, 2002; Shklovskii, 1962, 2002; von Hoerner, 1975, von Foerster,
Mora & Amiot, 1960).
9
__________________________________________________________________________________
In order to have these declarations supported by science we would have to work a little harder. We
would have to design a model with the homeostatic equilibrium. We would have to have data for the
availability of food since time immemorial. We would have to have corresponding data describing
the growth of population. These data would have to be at small time intervals in order to detect the
postulated oscillations. We would have to demonstrate convincingly that there were oscillations in the
growth of population and that there was a correlation between the recorded oscillations in the growth
of population and the oscillations in the availability of food. We would have to prove that the
oscillations in the growth of population were caused by the oscillations in the availability of food.
Acceptable evidence would have to be in demonstrating that our mathematical model reproduces all
these oscillations. This would have been science but what we are offered is just a story, which has to be
accepted by faith.
It is easy to claim many things but it is more difficult to prove them. Our postulates and explanations
might sound plausible but they would have to be verified by the rigorous process of scientific
investigation. Data (Biraben, 1980; Clark,1968; Cook,1960; Durand, 1974; Gallant, 1990; Haub, 1995;
Livi-Bacci, 1997; Maddison, 2001, 2010; McEvedy & Jones, 1978; Taeuber & Taeuber, 1949;
Thomlinson, 1975; Trager, 1994, United Nations, 1973, 1999, 2013) give no support for the existence
of the claimed fluctuations or oscillations.
There is no scientific basis for claiming that food-controlled homeostatic equilibrium had prevailed
since time immemorial. This claim has to be accepted by faith. We have to accept by faith that
population tends to oscillate in a homeostatic mechanism resulting from the conflict between the
population's natural tendency to increase and the limitations imposed by the availability of food. It all
might sound plausible but we cannot prove it. However, even if it sounds plausible it is contradicted by
the rigorous analysis of data (Nielsen, 2016b, 2015d).
Artzrouni and Komlos (1985) carried out model calculations, which incorporated the assumed
mechanism of the so-called Malthusian stagnation. Their contribution is important but for reasons,
which were not even noticed in their publication because their results show that the mechanism of the
so-called Malthusian stagnation does not work. We shall discuss this issue in one of our forthcoming
publications.
6.4. The allegedly characteristic features of the past human history
Stage 1 [of the Demographic Transition Theory] presumably characterizing most of human history,
involves high and relatively equal birth and death rates and little resulting population growth (Guest
& Almgren, 2001; p. 621. Italics added.).
This stage is characterized not by changes in average death rates but by a stagnation of death rates at
extremely high levels for a period of what is believed to be thousands of years (Olshansky & Ault,
1986, p. 357. Italics added.).
Claims:
1. Stage 1 proposed by the Demographic Transition Theory characterised presumably most of
human history.
2. During this stage, there were high and relatively equal birth and death rates.
3. During this stage, there was little resulting population growth.
4. This stage was not characterised by changes in the average death rates.
5. This stage was characterised by stagnation of death rates at extremely high levels.
6. This stagnation is believed to have lasted for thousands of years.
It is amazing how firmly the established knowledge is now established if so much can be so easily
claimed. The declaration that Stage 1 proposed by the Demographic Transition Theory was
characterized not by changes in average death rates but by a stagnation of death rates at extremely
high levels for a period of what is believed to be thousands of years has to be accepted by faith and by
faith alone because we can never expect to have systematic data describing death rates to check its
validity. No-one has yet demonstrated the validity of the Demographic Transition Theory. No-one has
yet demonstrated the existence of the first two stages of growth, let alone the existence of all stages of
growth.
Examples used in support of the Demographic Transition Theory are in fact in its direct contradiction
(Nielsen, 2016c). As pointed out earlier (Nielsen, 2016c), the only way to demonstrate the apparent
empirical features, which seem to be in agreement with the Demographic Transition Theory, is by a
10
__________________________________________________________________________________
suitable manipulation of data consisting in stitching together the birth and death rates data for Mauritius
with the data for Sweden.
It should be also remembered that any scientific theory is acceptable only if it is consistently
confirmed by empirical evidence. A single convincingly contradicting evidence questions the validity
of an accepted theory. For the Demographic Transition Theory, it is the other way round. There is not
a single convincing empirical evidence in support of this theory but there is overwhelming empirical
evidence showing that this theory is incorrect. This theory is contradicted by birth and death rates and
by the corresponding distributions describing the growth of population (Nielsen, 2016c). Furthermore,
within the range of analysable data (Biraben, 1980; Clark,1968; Cook,1960; Durand, 1974; Gallant,
1990; Haub, 1995; Livi-Bacci, 1997; Maddison, 2001, 2010; McEvedy & Jones, 1978; Taeuber &
Taeuber, 1949; Thomlinson, 1975; Trager, 1994, United Nations, 1973, 1999, 2013) growth of
population was hyperbolic (Kapitza, 2006; Kremer, 1993; Nielsen, 2016b, 2016d; Podlazov, 2002;
Shklovskii, 1962, 2002; von Hoerner, 1975, von Foerster, Mora & Amiot, 1960). Stages of growth
proposed by the Demographic Transition Theory did not exist.
Birth and death rates may have been high and strongly fluctuating but high and fluctuating birth and
death rates do not prove the existence of a stagnant state of growth because, as mentioned earlier, growth
is determined by the average difference between these two quantities. Furthermore, these two quantities
have to behave in a very specific way to produce the stagnant state of growth. Studying just death rates
or birth rates, or equivalently studying just the fertility rates (Lehr, 2009) cannot be used as the evidence
in support of the Demographic Transition Theory. Using scraps of favourable information while
ignoring contradicting evidence is strongly misleading and consequently scientifically unacceptable.
6.5. The allegedly well-documented evidence
It is well documented that the fluctuations experienced by the worlds population throughout history
did not have a regular, cyclical pattern, but were, to a large extent, brought about by randomly
determined demographic crises (wars, famines, epidemics, etc.). As McKeown and others have pointed
out, the main cause of these fluctuations of the past were mortality crises. There are four kinds of crises:
subsistence crises, epidemic crises, combined crises (subsistence/epidemic), and finally crises from
other causes, which are mainly exogenous (wars, natural or other catastrophes)
Crises followed by periods of population decline during which the nutritional status of the population
improved gave rise to fluctuations which testify to the continued existence of the Malthusian trap:
population would not grow beyond its carrying capacity for long, and when it did, the resulting
overshoot was followed by a crash (i.e. the positive checks such as diseases, famines, wars, etc.)
(Artzrouni & Komlos 1985, p. 24. Italics added.).
Claims:
1. There were fluctuations in the worlds population throughout history.
2. These fluctuations are well documented.
3. It is well documented that these fluctuations did not have a cyclic pattern.
4. It is well documented that these fluctuations were, to a large extent, brought about by randomly
determined demographic crises (wars, famines, epidemics, etc.).
5. The main cause of these fluctuations were mortality crises.
6. There are four types of crises.
7. Crises were followed by periods of population decline.
8. Population decline improved nutritional status.
9. Fluctuations testify to the continuing existence of the Malthusian trap.
10. Population was repeatedly reaching its carrying capacity.
11. Population would not grow beyond its carrying capacity for long.
12. Population growing beyond its carrying capacity was reflected in overshoots.
13. Overshoots were followed by crashes.
If all this is so well documented, where is the documentation of this well documented research? It
would be interesting to see at least a few references to this important and fundamental research work,
to see the data showing fluctuations throughout history, to see a positive proof that the the
fluctuations experienced by the worlds population throughout history are correlated with
demographic crises (wars, famines, epidemics, etc.), that they were brought about by randomly
determined demographic crises. It would be also interesting to see convincing evidence that population
was reaching its carrying capacity, that population would not grow beyond its carrying capacity for
11
__________________________________________________________________________________
long, the convincing evidence of overshoots and crashes, evidence that crashes were associated with
positive checks such as diseases, famines, wars, etc. It would be interesting to see the compelling
evidence of the existence of the Malthusian trap, the demonstration of frequent periods of population
decline, the compelling proof that periods of population decline caused by demographic crises were
improving nutritional status. All this vital and well documented evidence is missing.
What is well documented is the repeated fiction stories, which have to be accepted by faith. We have
many publications propagating such stories. The repeatedly related stories of fiction are by now
accepted as the undisputable facts. What is well documented is a system of beliefs, doctrines, wished-
for explanations, opinions, views, theories, hypotheses, conjectures and speculations, added gradually
over a long time until they became the established knowledge, the well-documented established
knowledge but the knowledge, which is contradicted by science.
In contrast, it is well documented (Biraben, 1980; Clark,1968; Cook,1960; Durand, 1974; Gallant,
1990; Haub, 1995; Livi-Bacci, 1997; Maddison, 2001, 2010; McEvedy & Jones, 1978; Taeuber &
Taeuber, 1949; Thomlinson, 1975; Trager, 1994, United Nations, 1973, 1999, 2013) that the growth of
human population does not show fluctuations or random behaviour. It is well documented that the data
show no signs of frequent overshoots and crashes, no signs of growth reaching its carrying capacity, no
signs of the continued existence of the Malthusian trap, no evidence that the population would not
grow beyond its carrying capacity for long, and no repeated periods of population decline. All these
colourful and dramatic descriptions associated with the narrative based on the assumption of the
existence of the mythical epoch of the so-called Malthusian stagnation are contradicted by data.
It is obvious, that demographic crises were often causing decline in the size of local populations,
depending on the scale of these crises and depending on what we understand by a local crisis.
Sometimes it might have been just a large death toll in a city, in a part of a country, as for instance in
China (Mallory, 1926), or maybe in the whole country or even extending over a few countries. However,
a large death toll does not necessarily mean a significant impact on the growth of human population. A
large death toll should not be immediately interpreted as a population decline; it could have been just a
slower growth over a limited time followed by a more intensified growth, as it happened after AD 1400
for the world population.
All these issues should be closely investigated by examining records of demographic catastrophes.
To arrive at any reasonably supported conclusion, we would have to do some hard work. However, data
which should be used for such investigations are strongly limited. We have no data showing that local
demographic crises were repeatedly causing fluctuations in the growth of regional or global populations.
In fact, the data show remarkably stable growth of human population, generally unaffected by
demographic crises (Nielsen, 2013a, 2013c, 2016b, 2016d).
The opening paragraph in the above quotation contains two interesting and characteristic elements,
the elements occurring repeatedly in the descriptions of the concept of the epoch of the so-called
Malthusian stagnation: (1) it makes a highly-questionable but confident declaration about the existence
of certain features (in this case about the existence of fluctuations) and (2) it equally confidently
explains them while ignoring empirical evidence. The normal progression in scientific research is first
to observe certain features and then try to explain them. We can also reverse the process: we can first
predict the existence of certain features. However, to accept the prediction and the associated
explanation, we would have to demonstrate the existence of the predicted features. This is how science
works but for doctrines accepted by faith scientific process of investigation is too tedious and
consequently it is readily ignored.
So, in this case, we would have to show first that there were significant fluctuations in the birth and
death rates and in the size of human population extending over thousands of years, and then we would
also have to explain them convincingly by demonstrating that they were correlated with demographic
crises. Alternatively, we would have to predict (using a suitable mathematical model) the existence of
fluctuations in birth and death rates and in the size of human population and then we would have to
show that our predictions are confirmed by relevant data.
We cannot prove that there were fluctuations throughout history in the birth and death rates
because we do not have relevant data, but we can prove that there were no fluctuations throughout
history in the size of human population because we have the relevant data (Biraben, 1980; Clark,1968;
Cook,1960; Durand, 1974; Gallant, 1990; Haub, 1995; Livi-Bacci, 1997; Maddison, 2001, 2010;
McEvedy & Jones, 1978; Taeuber & Taeuber, 1949; Thomlinson, 1975; Trager, 1994, United Nations,
12
__________________________________________________________________________________
1973, 1999, 2013). There is nothing in the data, which calls for the explanations of fluctuations in the
growth of population because there are no fluctuations. What needs to be explained is perhaps the
remarkable absence of fluctuations, the absence of random behaviour, crashes, overshoots or periods
of population decline. What needs to be explained is why the growth of population was so remarkably
stable during the past 12,000 years (Nielsen, 2016d) and why it was hyperbolic. The quoted declarations
are in perfect agreement with the established knowledge but they are in conflict with science.
6.6. The allegedly long-run equilibrium between population size and the food supply
Referring to three sources (Habakkuk, 1953; Kunitz, 1983; McKeown, 1983), Komlos explains:
Malthusian positive checks (mortality crises) maintained a long-run equilibrium between population
size and the food supply. Crises followed by periods when human nutritional status was above the level
of subsistence gave rise to cycles. the cycles testify to the continued existence of the Malthusian
population trap: population could not grow beyond an upper bound imposed by the resource and
capital constraints of the economic structure in which it was imbedded. The escape from this trap
occurred only when the aggregate capital stock was large enough and grew fast enough to provide
additional sustenance for the population, which thereby overcame the effects of the diminishing returns
that had hindered human progress during the previous millennia. After escaping from the Malthusian
trap, population was able to grow unchecked. In historic terms, this escape corresponds to the industrial
and demographic revolutions. Removal of the nutritional constraint, at least for the developed part of
the world, resulted in the population explosion (Komlos, 1989, pp. 194, 195. Italics added.).
Claims:
1. There was a long-term equilibrium between population size and the food supply.
2. This equilibrium was maintained by positive checks (mortality crises).
3. Crises were followed by periods when human nutritional status was above the level of
subsistence.
4. This process gave rise to cycles.
5. The cycles testify to the continued existence of the Malthusian population trap.
6. Population could not grow beyond an upper bound imposed by the resource and capital
constraints of the economic structure in which it was imbedded.
7. Malthusian trap was active for millennia.
8. The escape from the Malthusian trap occurred when the aggregate capital stock was large enough
and grew fast enough to provide additional sustenance for the population.
9. The removal of nutritional constrains caused population explosion.
Massive amount of work would have to be done to support all these impressive declarations. We
would have to study food supply over millennia and determine how they were correlated with the
growth of human population. We would have to prove that there was a long-run equilibrium between
population size and the food supply. We would have to study mortality crises over millennia. We
would have to establish a correlation between the growth of human population, food supply and
mortality crises. We would also have to investigate upper bounds of resource and capital constraints
and prove that over millennia the size of the population was repeatedly reaching the limits of these
upper bounds.
Conducting scientific research is not easy but results have a high degree of reliability. Writing
fictions stories, whose general script is already provided by the established knowledge based largely on
faith is much easier, but this is not science.
It is easy to declare so much so quickly and with such a confidence, but it is harder to prove it. It is
also hard to accept it, but accept we must if we want to accept the concept of the epoch of the so-called
Malthusian stagnation promoted by the established knowledge.
The claimed cycles cannot possibly testify to the continued existence of the Malthusian population
trap because they did not exist in the growth of population (Nielsen, 2016b, 2016d, von Foerster,
Mora & Amiot, 1960). Population growth, global and regional, was remarkably stable and
unconstrained. The claim that population could not grow beyond an upper bound imposed by the
resource and capital constraints is contradicted by the analysis of population data. This claim appears
to be based on pure fantasy and on a wished-for mechanism that did not exist. There was no Malthusian
trap in the growth of population.
We know nothing about any possible cycles in birth and death rates because we have no relevant
data extending over a long time in the past. We do not know how large were these alleged cycles. We
13
__________________________________________________________________________________
do not even know whether they existed. Discussions of these cycles are irrelevant because we know
that cycles in birth and death rates are of little or no consequence for explaining the mechanism of
growth (Nielsen, 2016c). Even if they were present they did not have any significant influence on the
growth of the world population in the past 12,000 years (Nielsen, 2016d). They also had no influence
on the growth of regional populations (Nielsen, 2016b). The absence of cycles in the growth of
population combined with the evidence of the steadily increasing growth testify that the Malthusian trap
did not exist. We cannot also claim that there was escape from this trap because there was no trap
in the growth of population. There was also no trap in the economic growth (Nielsen, 2014, 2015a,
2016a, 2016e, 2016f, 2016g, 2016h, 2016i). Again, the quoted declarations are in perfect agreement
with the established knowledge but they are in conflict with science.
6.7. The alleged fluctuation of fertility and mortality rates around zero
Discussing the first stage of the Demographic Transition Theory, Warf explains:
Because both fertility and mortality rates are high, the difference between them natural population
growth is relatively low, fluctuating around zero (Warf, 2010, p. 708. Italics added.).
Claims:
1. During the first stage of the demographic transition fertility and mortality rates were high.
2. The difference between them (the natural population growth) was fluctuating around zero
because they were high.
Just because fertility and mortality rates were high it does not follow that the difference between
them was zero. The difference between them can fluctuate around zero even if they are low. However,
this is just a minor issue.
In this quotation, the natural population growth is identified as the difference between the fertility
and mortality rates. It is, therefore, the rate of natural increase or the growth rate because, in general,
migration rates are relatively small and can be neglected.
We shall recall that while the growth rate fluctuating around a constant value describes exponential
growth, the growth rate fluctuating around zero describes the constant size of the growing entity, i.e.
in our case, the constant size of the population. The claim made by Warf is contradicted by data, which
show that for thousands of years the size of human population was not constant but steadily increasing
(Nielsen, 2016b, 2016d, von Foerster, Mora & Amiot, 1960). Furthermore, the size of population was
increasing hyperbolically. The natural population growth (growth rate) could not have been
fluctuating around zero but it must have been increasing hyperbolically because for the hyperbolic
growth, the growth rate also increases hyperbolically (Nielsen, 2016h).
6.8. The alleged roughly constant population
In line with the accepted interpretations of the first stage of the Demographic Transition Theory,
Lagerlf writes:
The Malthusian Regime in our model is a stable situation where death and birth rates are both high, and
population roughly constant. Moreover, mortality is highly volatile, increasing dramatically in periods
of big epidemic shocks. In periods with mild shocks population expands. This worsens the impact of
the next epidemic, equilibrating population back to its Malthusian state (Lagerlf, 2003a, p. 756. Italics
added.).
In our model, the world can thus be stuck in a Malthusian equilibrium for centuries and then suddenly
escape, and never contract back. As suggested by a referee, this process could possibly be interpreted
in terms of wars, instead of epidemics (Lagerlf, 2003a, p. 766. Italics added.).
Throughout human history, epidemics, wars and famines have shaped the growth path of population.
Such shocks to mortality are the central theme of the model set up by Lagerlf, which endogenously
generates a long phase of stagnant population and living standards, followed by an industrial revolution
and a demographic transition (Lagerlf, 2003b, pp. 434, 435. Italics added.).
Claims:
1. It is assumed that there was a Malthusian regime.
2. It is assumed that Malthusian regime is characterised by high birth and death rates.
3. During the Malthusian regime population is roughly constant.
4. Mortality is highly volatile.
5. Mortality increases dramatically in periods of big epidemic shocks.
6. Population expands when the mortality shocks are mild.
14
__________________________________________________________________________________
7. Expanding population worsens the impact of the next epidemic and equilibrates population to
the Malthusian state.
8. Malthusian equilibrium lasts for centuries.
9. The process of Malthusian equilibrium can be also explained by wars instead of epidemics.
10. Throughout human history, epidemics, wars and famines have shaped the growth path of
population.
11. Model based on the assumption of shocks to mortality generates a long phase of stagnant
population.
12. The long phase of stagnant population and living standards is followed by an industrial
revolution and a demographic transition.
Here again, and quite typically, we have a series of declarations that have to be accepted by faith.
However, paradoxically if not ironically, Lagerlf was on the verge of discovering that doctrines
accepted by faith were contradicted by his own model.
He has carried out an interesting and important research work but unfortunately, he did not finish it:
he did not compare results of his calculations with data (Maddison, 2001), which were available to him
before publication of his work. He did not take the final and the most essential step. If he did, he would
have discovered that the mechanism of the so-called Malthusian stagnation incorporated in his model
did not produce fluctuations in the model-generated growth of population, that model-generated growth
of population was not stagnant and it did not fit the relevant data. He would have found that contrary to
what he claims in his publication, his model generated population was not roughly constant. If he
cared to consult data (Maddison, 2001) he would have also found the population reported by Maddison
was also not roughly constant. We shall discuss these issues in a separate publication.
Lagerlf presents a plot of the growth rate and calls it erroneously Population growth (Lagerlf,
2003b, p. 436). He fails to take the most essential step in this type of work and to use his model-
generated growth rate to calculate model-generated distribution describing the growth of population.
He ignores data (Maddison, 2001) and yet his unfinished work is accepted for publication maybe
because it proclaims loud and clear the doctrines of the established knowledge. Science appears to be
of no importance.
6.9 Incorrect claims about the growth rate
In our model, this leads to a constant rate of population growth prior to the adoption of the Solow
technology. This result is consistent with population data from Michael Kremer (1993), where the
growth rate of population fluctuates around a small constant throughout most of the Malthusian period
(from 4000 B.C. to A.D. 1650) (Hansen & Prescott (2002, p. 1205. Italics added.).
Claims:
1. Growth rate of population fluctuates around small constant during the Malthusian period (i.e.
prior to the adoption of Solow technology).
2. Small and roughly constant growth rate is consistent with population data from Michael Kremer
(1993).
First, it appears that Hansen and Prescott might be confusing constant growth rate with constant
population. It might be the same mistake as it appears to have been made by Lagerlf (2003b). A
constant (non-zero) growth rate does not produce a constant (non-zero) size of population. A constant
(non-zero) growth rate produces exponential growth.
Second, this declaration appears to contain conflicting information. It is hard to imagine that random
forces characterising the mythical Malthusian period would produce a steadily increasing exponential
growth. Steadily-increasing growth suggests the presence of a dominating constant force, overruling
any random forces.
Third, fluctuations in the growth rate are not readily reflected as fluctuations in the growth of
population (Nielsen, 2016c). We can demonstrate it even without carrying mathematical analysis of the
fluctuating growth rate. Data alone (Lehmeyer, 2004; Mauritius, 2015; Statistics Mauritius, 2014;
Statistics Sweden, 1999; Wrigley & Schofield, 1981) show clearly that fluctuating growth rates do not
produce significant fluctuations in the growth of population and that they have no impact on the
mechanism of growth because they do not alter growth trajectories.
Fourth, we would have to show convincingly that the growth rate was indeed fluctuating around a
small constant value as claimed by Hansen and Prescott (2002). There is no such proof because we do
not have the data for the growth rate extending over thousands of years. However, there is a proof that
15
__________________________________________________________________________________
the growth rate during the AD and BC eras was not fluctuating around a small constant value but that
it was increasing hyperbolically because the growth of the population was hyperbolic (Kapitza, 2006;
Nielsen, 2016b, 2026d; Kremer, 1993; Podlazov, 2002; Shklovskii, 1962, 2002; von Hoerner, 1975,
von Foerster, Mora & Amiot, 1960). For the hyperbolic growth, the growth rate increases hyperbolically
with time or in the direct proportion to the size of population (Nielsen, 2016h), as observed also by
Kremer (1993).
Fifth, Kremer (1963) did not carry out an extensive study of the growth rate. He has presented rough
calculations of this quantity using strongly varying local gradients, which do not represent the real
gradient of growth. His calculations are strongly inaccurate for the BC era when individual data values
are separated by large time intervals. It is scientifically unjustifiable to use such calculations and claim
fluctuations around a constant value.
Sixth, for the hyperbolic growth, growth rate is small over a long time because it is also hyperbolic.
Growth rate might appear to vary around a small constant but such interpretation is incorrect. Growth
rate should be preferably calculated using interpolated gradients to avoid spurious effects of strongly-
varying local gradients between adjacent data values. It is also useful to display growth rate using
various types of displays to help in its interpretation. Using the approximate calculations of Kremer
(1993) and claiming that growth rate was varying around small constant is self-misleading and
scientifically unjustified.
This example illustrates that in science it is essential to carry out methodical analysis of data. In
economic and demographic research, this is particularly important because historical economic growth
and historical growth of population were increasing hyperbolically. Hyperbolic distributions are
strongly misleading and can easily lead to their misinterpretations. Furthermore, for hyperbolic
distributions, the growth rate and the gradient increase in a similar fashion. The growth rate increases
hyperbolically and the gradient follows the second-order hyperbolic distribution, both of them
containing the same confusing features of a slow growth over a long time and a fast growth over a short
time, but both increasing monotonically over the entire range of time. Hyperbolic growth of the GDP
and population as well the monotonically-increasing growth rates and gradients cannot be divided into
two or three distinctly different sections. They all have to be analysed and interpreted as a whole. The
same mechanism has to be applied to the slow and to the fast growth because slow and fast growth
belongs to the same, monotonically-increasing distributions.
6.10. The alleged density-dependent variations in mortality
If population density increases the mortality rate rises, equilibrating population back to the Malthusian
trap (Lagerlf, 2003a, p. 765. Italics added.).
This statement has to be also accepted by faith because there is no convincing research supporting
such declaration. Creative imagination appears to be taking full control in the established knowledge.
Here we have an example of an interesting detail added to the concept of the epoch of the so-called
Malthusian stagnation, illustrating how one fantasy can lead easily to a new fantasy and how such
gradual additions reinforce the established knowledge. This statement claims the dependence of
mortality rate on the density of human population. It offers an explanation how the phantom Malthusian
trap regulates the growth of human population. It describes some kind of a general rule that the
Malthusian trap is activated when the population density, not its size, reaches a certain limiting value.
There is no research confirming the described mechanism; no research showing how the growth of
human population depends on its density. Even if we could show some isolated examples of the density-
dependent growth we would have to demonstrate that such mechanism applies also to regional and
global populations. The best data available to us show the time-dependence of the size of human
population and there is nothing in them to suggest any form of density-dependence, let alone the
existence of the Malthusian trap triggered by the density of population.
This statement is yet another example of the leaps of faith, of confident declarations requiring a huge
amount of work to be accepted as a reliable contribution to science. The descriptions of the epoch of
the so-called Malthusian stagnation are full of such unscientific declarations. Indeed, they are made of
them.
Other terms used to describe the alleged stagnant and fluctuating state of growth during this mythical
epoch of Malthusian stagnation are equilibrium trap or population trap (Leibenstein, 1957; Nelson,
1956), multiple equilibria or poverty trap (Wang, 2005).
16
__________________________________________________________________________________
The belief in the stagnant and fluctuating growth is so strong that mathematical models are deemed
successful if they can generate the desired oscillations during this mythical epoch of Malthusian
stagnation, and no-one seems to care to take the next and the most essential step and to compare model
calculations with population data. As long as oscillations of some kind are generated by a mathematical
model, they are taken as the proof of the existence of the epoch of the so-called Malthusian stagnation.
This line of reasoning shows that the primary, if not the exclusive, aim of such mathematical exercises
is to translate a story into a mathematical language and when the translation is done properly, when
mathematical formulae generate any kind of oscillations, large or small, significant or negligible, these
formulae are then taken as a proof of the existence of the so-called Malthusian stagnation.
6.11. The alleged Age of Pestilence and Famine
The so-called epoch of Malthusian stagnation is also described as the Age of Pestilence and Famine
(Omran 1971, 1983, 1998).
In this stage, the major determinants of death are the Malthusian positive checks, namely epidemics,
famines and wars (Omran, 1983, p. 306; Omran, 2005, p. 737).
Even if fertility approached its biologic maximum, depopulation could and did occur as a result of
epidemics, wars and famines, which repeatedly pushed mortality levels to high peaks (Omran, 2005, p.
733).
The pattern of growth [of human population] until about 1650 is cyclic (Omran, 1971, Table 4, p. 533).
Claims:
1. During the Age of Pestilence and Famine (i.e. during the so-called epoch of Malthusian
stagnation) major determinants of death are the Malthusian positive checks (epidemics, famines
and wars).
2. Depopulation was occurring even when fertility was approaching its biological maximum
because epidemics, wars and famines were repeatedly pushing mortality levels to high peaks.
3. Growth of population before AD 1650 was cyclic.
To justify the first claim, we would have to have reliable records of the causes of death over
thousands of years. We would then have to show convincingly that indeed the major causes of death
were epidemics, famines and wars. We would also have to show that there was a clear change in the
causes of death when the so-called epoch of Malthusian stagnation ceased to exist. We cannot present
such proofs because we do not have the supporting data. In principle, therefore, this claim is not
scientific because we cannot check it by data. It has to be accepted by faith.
To justify the second claim, we would have to have reliable records of fertility and mortality over
thousands of years. We would then have to demonstrate that fertility was approaching biological limits,
that such events were coinciding with high mortality peaks and that these high mortality peaks were
caused by epidemics, wars and famines. We do not have relevant data to check whether these
descriptions are true. They are therefore also unscientific and they have to be accepted by faith.
The growth of population, global and regional, before AD 1650 was not cyclic (Nielsen, 2016b,
2016d). This statement is contradicted by data (Biraben, 1980; Clark,1968; Cook,1960; Durand, 1974;
Gallant, 1990; Haub, 1995; Livi-Bacci, 1997; Maddison, 2001, 2010; McEvedy & Jones, 1978; Taeuber
& Taeuber, 1949; Thomlinson, 1975; Trager, 1994, United Nations, 1973, 1999, 2013).
6.12. The alleged main cause of mortality
During the first stage, mortality vacillated at high levels, with infectious disease as the main cause of
death plus a large proportion due to wars and famines (Robine, 2001, p. 191. Italics added.).
Claims:
1. During the first stage of demographic transitions mortality vacillated at high levels.
2. The main causes of death were infectious diseases.
3. Large proportion of death were caused by wars and famines.
We cannot prove that mortality vacillated at high levels because we have no relevant data for the
so-called first stage to carry out such a study, the stage that is assumed to have lasted for thousands
of years. We cannot prove that these imagined and wished-for vacillations were correlated with
infectious disease, wars and famines. We cannot prove that the main causes of deaths were infectious
diseases. We cannot prove that a large proportion of death was due to wars and famines. We do not
have sufficiently extensive records of causes of death extending over thousands of years. We do not
know how the causes of death were changing over time. We do not have the records to help us to
17
__________________________________________________________________________________
distinguish between the major and minor causes. We do not know whether the main cause of death was
the same over thousands of years. The concept of the epoch of Malthusian Stagnation and all these
claims have to be accepted by faith.
6.13. The alleged unsustained growth of population
The first transition phase, called the Age of Pestilence and Famine, is characterized by high and
fluctuating mortality rates, variable life expectancy with low average life span, and periods of
population growth that are not sustained (McKeown, 2009, p. 20S. Italics added.).
Claims:
1. During the Age of Pestilence and Famine (i.e. during the hypothetical but non-existent epoch of
the so-called Malthusian stagnation) mortality rates were high and fluctuating.
2. Average life span was low.
3. There were periods when the population growth was not sustained.
Mortality rates might have been high and fluctuating but we have no data extending over thousands
of years to prove it. Furthermore, we would yet have to show that these hypothetical high and fluctuating
mortality rates could have been responsible for creating stagnation. What we know is that strongly-
fluctuating mortality rates do not change the growth of population (Lehmeyer, 2004; Mauritius, 2015;
Nielsen, 2016c; Statistics Mauritius, 2014; Statistics Sweden, 1999; Wrigley & Schofield, 1981). There
is also nothing in the data and in their analysis to show that low average life span was affecting the
growth of population. As for the periods of population growth that are not sustained this claim is
contradicted by the analysis of data (Nielsen, 2016b, 2016d).
6.14. Positive forces were allegedly balanced by negative forces
The positive forces of growth had existed all along. However, they had been counterbalanced by the
negative forces of malnutrition and disease (Komlos & Baten, 2003, p. 19).
We have no reliable empirical evidence to support this claim, no study of positive and negative
forces, no study of their balancing, and no study of their influence on the growth of human population.
This is not science but story-writing prompted and approved by the established knowledge.
How do we know that the so-called positive forces were balanced by forces of malnutrition and
disease? They obviously were not because economic growth and the growth of population were
hyperbolic and remarkably stable (Nielsen, 2016a, 2016b, 2016d). Such a strong and stable growth
could have been only generated by a strong and dominating force.
Here again, authors of this declaration take an easy way out. They have made no attempt to consult
data available to them at the time of the publication of their paper (Biraben, 1980; Clark,1968;
Cook,1960; Durand, 1974; Gallant, 1990; Haub, 1995; Livi-Bacci, 1997; Maddison, 2001; McEvedy &
Jones, 1978; Taeuber & Taeuber, 1949; Thomlinson, 1975; Trager, 1994, United Nations, 1973, 1999).
They have made no attempt to reconcile their interpretations with the already documented evidence of
hyperbolic growth (Kapitza, 2006; Kremer, 1993; Podlazov, 2002; Shklovskii, 1962, 2002; von
Hoerner, 1975, von Foerster, Mora & Amiot, 1960). Again, this declaration is in perfect agreement with
the established knowledge but is in conflict with science.
6.15. The continuing misinformation
The established knowledge is by now so strongly established that it will be difficult to change it. It
continues to be supported by the scientifically-unsubstantiated claims and descriptions. It would take
volumes to list and discuss all such examples and to show that these repeatedly propagated doctrines,
explanations and interpretations have to be accepted by faith.
The current established knowledge based on the assumption of the so-called Malthusian stagnation
followed by explosion and reinforced by many complicated explanations is similar to the established
knowledge about the dynamics of celestial bodies, interpretations which were established for about two
millennia before they were eventually abandoned. Describing the work of mathematicians of his time,
Osiander wrote:
18
__________________________________________________________________________________
With them it is as though an artist were to gather the hands, feet, head and other members from his
images from divers models, each part excellently drawn, but not related to a single body, and since they
in no way match each other, the result would be monster rather than man (Copernicus, 1995). *
Historical economic growth and historical growth of population can be expected to be described by
a simple mechanism because hyperbolic growth is simple. This issue will be discussed in a separate
publication, where a simple explanation of the mechanism of hyperbolic growth will be also presented.
Hyperbolic growth prevailed for at least 12,000 years for the growth of population (Nielsen, 2016d)
and for hundreds of years for the economic growth (Nielsen, 2016a). The established knowledge in
demography and in economic research offers complicated explanations, which have to be accepted by
faith. Hopefully we shall not have to wait for two thousand years to abandon these erroneous doctrines
and replace them by science.
7. Summary and conclusions

Established knowledge in demography and in economic research is based on a series of doctrines
and explanations revolving around the concept of the so-called Malthusian stagnation and around the
concept of the escape from the so-called Malthusian trap described as explosion, takeoff, sprint or spurt.
It is a system of interpretations, which have to be accepted by faith.
It is easy to understand why these concepts are so attractive because the growth of population and
economic growth were increasing hyperbolically and hyperbolic growth creates an illusion of stagnation
followed by explosion.
It is essential to understand that hyperbolic distributions should be analysed and interpreted as a
whole. If we take just a few examples along the hyperbolic growth, we can easily make a mistake and
arrive at incorrect conclusions. If hyperbolic distributions are already difficult to understand without
their methodical analysis, linearly-modulated hyperbolic distributions (Nielsen, 2015a) describing
income per capita are even more difficult to understand because they create even stronger illusion of
stagnation followed by a sudden explosion. Here again, just taking a few examples along these
distributions is bound to lead to incorrect conclusions. These distributions have to be also analysed with
care. Careful and methodical mathematical analysis of data describing historical economic growth and
the growth of population is unavoidable.
Distributions describing income per capita are generated by a division of two hyperbolic
distributions. The characteristic feature of this ratio is that for a long time the growth of income per
capita was not just slow, as for hyperbolic distributions, but nearly constant. This feature characterises
the division of any hyperbolic distributions, not just the division of the GDP and population (Nielsen,
2015a). It is a purely mathematical property, which has nothing to do with specific properties of
economic growth,
The nearly constant income per capita should never be interpreted automatically as stagnation. The
only way to claim stagnation for this nearly-constant income per capita is to analyse the GDP and
population data separately and to prove that these distributions are not hyperbolic but stagnant.
It is incorrect to take a few values of income per capita, show that they are nearly constant and claim
stagnation. If the GDP and population increase hyperbolically, then income per capita increases by
following the monotonically-increasing linearly-modulated hyperbolic destitution and it is incorrect to
try to divide such a monotonically-increasing distribution into two different sections, slow and fast.
Mathematically, it is impossible to make such a division. It is impossible to identify a point or a range
of points and claim them as marking the place of transition.
Even though the ratio of two hyperbolic distributions is nearly constant over a long time and nearly
vertical over a short time, the transition from the nearly constant to the nearly vertical patterns occurs
all the time along the entire range of such distributions. Linearly-modulated hyperbolic distributions
representing income per capita should be also interpreted as a whole. The same mechanism should be
*
This quotation comes from a letter written by Andreas Osiander, Lutheran theologian and a friend of
Copernicus, a letter addressed to the chief editor, Pope Paul III. Osiander argues in favour of the mathematically
simple and elegant heliocentric system as opposed to the complicated geocentric descriptions. This letter was
later used as an unsigned introduction to the book De revolutionibus orbium coelestium, and was mistakenly
attributed to Copernicus.
19
__________________________________________________________________________________
applied to the nearly constant and to the nearly vertical growth, unless we can prove that the GDP and
population were not following hyperbolic distributions but were stagnant.
We have presented many examples of claims revolving around the concepts of stagnation followed
by explosion. We have shown why such claims are scientifically unacceptable.
The origin of the fundamental concepts of the established knowledge can be traced, perhaps not
entirely correctly, to Malthus (1798). He has presented an important pioneering work but unfortunately
the ensuing studies of economic growth and of the growth of population have taken a wrong turn at a
certain time in the past, perhaps because relevant data were not available.
By the time the relevant data (Biraben, 1980; Clark,1968; Cook,1960; Durand, 1974; Gallant, 1990;
Haub, 1995; Livi-Bacci, 1997; Maddison, 2001, 2010; McEvedy & Jones, 1978; Taeuber & Taeuber,
1949; Thomlinson, 1975; Trager, 1994, United Nations, 1973, 1999, 2013) became available, they were
ignored. More recently, some of them (Maddison, 2001) were manipulated to support the established
knowledge (Ashraf, 2009; Galor, 2005a, 2005b, 2007, 2008a, 2008b, 2008c, 2010, 2011, 2012a, 2012b,
2012c; Galor & Moav, 2002; Snowdon & Galor, 2008). Earlier analyses of data (Kapitza, 2006; Kremer,
1993; Podlazov, 2002; Shklovskii, 1962, 2002; von Hoerner, 1975, von Foerster, Mora & Amiot, 1960)
showing that the growth of population was hyperbolic were also ignored. By now, the established
knowledge is so well established that anything being in its conflict is methodically ignored, rejected or
suppressed. This is not science.
Recent analyses of data (Nielsen, 2014, 2015a, 2016a, 2016b, 2016d, 2016e, 2016f, 2016g, 2016h,
2016i) confirmed the earlier studies (Kapitza, 2006; Kremer, 1993; Podlazov, 2002; Shklovskii, 1962,
2002; von Hoerner, 1975, von Foerster, Mora & Amiot, 1960) and demonstrated that the historical
growth of population and the historical economic growth were hyperbolic. The established knowledge
based on the scientifically-contradicted concepts of stagnation followed by explosion (takeoff or the
escape from the Malthusian trap) has to be replaced by explanations based on accepting hyperbolic
growth.
It is incorrect to interpret the past harsh living conditions as a proof of the existence of the so-called
Malthusian stagnation. Whatever harsh living conditions might have been present in the past, their
effects are generally not reflected in growth trajectories. The only known example is for the growth of
global population between AD 1200 and 1400 coinciding with the convergence of five major
demographic catastrophes (Nielsen, 2016d). However, even then, the recorded effect is small.
Negative effects of the Malthusian positive checks should be never used robotically to describe the
past growth of population or the economic growth. If we want to claim that positive checks were shaping
the growth of population or economic growth, we have to prove it. If we want to claim that the so-called
Law of Population was shaping growth trajectories, we have to prove it. If we want to claims that
demographic catastrophes were shaping the growth of population, we have to prove it. If we want to
claim that Industrial Revolution was shaping growth trajectories, we have to prove it. We cannot take
shelter in the established knowledge because in this respect established knowledge is repeatedly
contradicted by data. Any data we might have, should to be methodically analysed to prove the negative
effects of Malthusian positive checks but whatever we would prove would be just an exception from
the general and well-demonstrated pattern that the historical growth of population and historical
economic growth were not only hyperbolic but that they also remarkably stable.
Interpretations based on the concepts of the so-called Malthusian stagnation and on the claims of the
escape from the so-called Malthusian trap are not only incorrect but also dangerously misleading. They
suggest that after the endless epoch of stagnation we have now entered the sustained growth regime
(Galor, 2005a, 2011). This hypothesis creates a sense of security. In contrast, analysis of data shows
that the past growth was sustainable but now for the first time in human history it is unsustainable and
insecure (Nielsen, 2015b). While in the past, economic growth and the growth of population, global
and regional, were following the slowly increasing hyperbolic trajectories (Nielsen, 2014, 2015a,
2016a, 2016b, 2016d, 2016e, 2016f, 2016g, 2016h. 2016i) indicating the unconstrained and secure
growth, now the growth is at the stage of the dangerously fast increase (see Figures 1 and 2). The growth
is no longer hyperbolic but the current growth increases close to the historical hyperbolic trajectories.
For the first time in human history, these growth trajectories are clearly unsustainable because such a
fast increase cannot be possibly tolerated for much longer.
The established knowledge is not only in conflict with data describing the past economic growth and
the growth of human population but also in conflict with the general knowledge about the current
20
__________________________________________________________________________________
mounting problems threatening our future. The established knowledge in demography and in economic
research created its own world of fiction divorced from the real world.
We have not escaped the Malthusian trap because there was no trap in the economic growth or in
the growth of population. The past growth was unconstrained and sustainable as demonstrated by the
undisturbed hyperbolic distributions. However, now we are in the trap. For the first time in human
history we are in the trap of numerous critical problems, which threaten our global security and our
survival (Nielsen, 2006). For the first time in human history our combined ecological footprint is larger
than the ecological capacity and it continues to increase (WWF, 2010). For the first time in human
history our growth is supported by the increasing ecological deficit.
In order to understand the past and present economic growth, erroneous interpretations revolving
around the concept of the so-called Malthusian stagnation have to be abandoned and replaced by
scientifically acceptable interpretations. What needs to be explained is why the past economic growth
and the growth of population were hyperbolic. Why was the growth so remarkably stable? Why was it
not influenced by many random forces, which might have been present? Why did the growth start to
divert to slower trajectories? Why does it continue so closely to the dangerously fast hyperbolic
trajectories? And the most important questions of all: How to slow down the current growth? How to
control growth?
Examples presented here suggest that there is a problem not just with certain interpretations adopted
and protected by the established knowledge in the demographic and economic research but with the
way research is carried out in these two fields. It is not just the problem with one or two theories, such
as the Demographic Transition Theory or the Unified Growth Theory, which need to be corrected or
most likely replaced. It is not even just the problem with the accepted paradigm based on the concept
of the so-called Malthusian stagnation, which needs to be abandoned. It is a systemic problem. It is a
problem, with the way research is conducted in these two fields. It is a problem with creating stories
and interpretations, which have to be accepted by faith. It is a problem with a selective use of data. It is
a problem of ignoring contradicting evidence, such as the contradicting evidence published over 50
years ago by von Foerster, Mora and Amiot (1960). It is a problem with manipulating and distorting
data to fit the preconceived ideas, as it has been done repeatedly in the Unified Growth Theory and in
other related publications (Ashraf, 2009; Galor, 2005a, 2005b, 2007, 2008a, 2008b, 2008c, 2010, 2011,
2012a, 2012b, 2012c; Galor & Moav, 2002; Snowdon & Galor, 2008). It is a problem of testing data
by a theory rather than testing theory by data. It is a problem with protecting a system of doctrines,
which are accepted on faith.
As outlined briefly elsewhere (Nielsen, 2016i), there are two ways of conducting research: (1) the
dynamic scientific method, which is used in the self-correcting disciplines of science and (2) the stale
method, which is used routinely in the usually emotional and dishonest defence of doctrines accepted
by faith. It is unfortunate, that as pointed out earlier (Nielsen, 2013a, 2013b, 2013c, 2014, 2015a, 2016a,
2016b, 2016c, 2016d, 2016e, 2016f, 2016g, 2016h, 2016i), demographic and economic research appears
to gravitate strongly towards the unscientific method.
The established knowledge revolving around the concept of the so-called Malthusian stagnation has
to be changed because there was no stagnation in the historical economic growth and in the historical
growth of population. There was also no escape from the Malthusian trap because there was no trap.
This paradigm has to be changed because historical economic growth and the historical growth of
population were hyperbolic. However, in order to make the demographic and economic research
scientifically acceptable, the systemic problem has to be also solved. Scientific research can be based
only on the well-known and generally recognised scientific rules of investigation. Anything else is not
science.
References
Ashraf, Q. H. (2009). Essays on Deep Determinants of Comparative Economic Development. Ph.D. Thesis, Department of
Economics, Brown University, Providence.
Biraben, J-N. (1980). An Essay Concerning Mankind's Evolution. Population, Selected Papers, December.
Clark, C. (1968). Population Growth and Land Use. New York, NY: St Martin's Press.
Cook, R. C. (1960). World Population Growth. Law and Contemporary Problems, 25(3), 379-388.
21
__________________________________________________________________________________
Copernicus, N. (1995). On the revolution of heavenly spheres (C. G. Wallis, Trans.). New York: Prometheus Books.
Original work published in 1543.
Durand, J. D. (1974). Historical Estimates of World Population: An Evaluation. Analytical and Technical Reports, Number
10. University of Pennsylvania, Population Center.
Gallant, R. A. (1990). The Peopling of Planet Earth: Human Growth through the Ages. New York, NY: Macmillan
Publishing Company.
Galor, O. (2005a). From stagnation to growth: Unified Growth Theory. In P. Aghion & S. Durlauf (Eds.), Handbook of
Galor, O. (2005b). The Demographic Transition and the Emergence of Sustained Economic Growth. Journal of the
European Economic Association, 3, 494-504. doi. 10.1162/jeea.2005.3.2-3.494
Galor, O. (2007). Multiple Growth RegimesInsights from Unified Growth Theory. Journal of Macroeconomics, 29, 470-
475.
Galor, O. (2008a). Comparative Economic Development: Insight from Unified Growth Theory.
http://www.econ.brown.edu/faculty/Oded_Galor/pdf/Klien lecture.pdf
Galor, O. (2008b). Economic Growth in the Very Long Run. In: Durlauf, S.N. and Blume, L.E., Eds., The New Palgrave
Dictionary of Economics, Palgrave Macmillan, New York. doi. 10.1057/9780230226203.0434
Galor, O. (2008c). Comparative Economic Development: Insight from Unified Growth Theory.
Galor, O. (2010). The 2008 Lawrence R. Klein LectureComparative Economic Development: Insights from Unified
Growth Theory. International Economic Review, 51, 1-44. doi. 10.1111/j.1468-2354.2009.00569.x
Galor, O. (2012a). Unified Growth Theory and Comparative Economic Development.
Galor, O. (2012b). The Demographic Transition: Causes and Consequences. Cliometrica, 6, 1-28. doi. 10.1007/s11698-011-
0062-7
Galor, O. (2012c). Unified Growth Theory and Comparative Economic Development.
Galor, O. & Moav, O. (2002). Natural Selection and the Origin of Economic Growth. The Quarterly Journal of Economics,
117, 1133-1191. doi. 10.1162/003355302320935007
Grebe, T., Schmid, J., & Stiehler, A. (2008). Do individuals recognize cascade behavior of others? An experimental study.
Journal of Economic Psychology, 29(2), 197209.
Guest, A. M., & Almgren, G. (2001). Demographic Transition. In Encyclopedia of Sociology (2nd ed.), Vol. 1, (pp. 621-
630). Detroit: Macmillan Reference.
Habakkuk, H. J. (1953). English population in the eighteenth century. Economic History Review, 6(2), 117-133
Hansen, G. D., & Prescott, E. C. (2002). Malthus to Solow. The American Economic Review, 92(2), 1205-1217.
Haub, C. (1995). How Many People Have Ever Lived on Earth? Population Today, February.
Kapitza, S. P. (2006). Global population blow-up and after. Hamburg: Global Marshall Plan Initiative.
Komlos, J., & Baten, J. (2003). The Industrial Revolution as the escape from the Malthusian trap. Munich Discussion Paper
No. 2003-13, Volkswirtschaftliche Fakultt, Ludwig-Maximilians-Universitt Mnchen
Komlos, J. H. (1989). Thinking about Industrial Revolution. Journal of European Economic History, 18, 191-206.
Komlos, J. H. (2000). The Industrial Revolution as the Escape from the Malthusian Trap. Journal of European Economic
History, 29, 307-331.
Kremer, M. (1993). Population growth and technological change: One million B.C. to 1990. Quarterly Journal of
Economics, 108(3), 681716.
Kunitz, S. J. (1983). Speculations on the European mortality decline. Economic History Review, 36, 349-364.
Lagerlf, N-P. (2003a). From Malthus to modern growth: Can epidemics explain three regimes? International Economic
Review, 44, 755-777.
Lagerlf, N-P. (2003b). Mortality and early growth in England, France and Sweden. Scand. J. of Economics, 105(3), 419
439.
Lehmeyer, J. (2004). Mauritius: Historical demographical data of the whole country.
http://www.populstat.info/Africa/mauritic.htm
Lee, R. D. (1997). Population dynamics: Equilibrium, disequilibrium, and consequences of fluctuations. In O. Stark & M.
Rosenzweig (Eds.), The Handbook of Population and Family Economics (pp.1063-1115). Amsterdam: Elsevier.
Lehr, C. S. (2009). Evidence on the demographic transition. Review of Economics and Statistics, 91(4), 871-887.
Leibenstein, H. (1957). Economic backwardness and economic growth, New York, Willey.
Livi-Bacci, M. (1997). A Concise History of World Population (2nd ed.). Malden, MA: Blackwell Publishers.
Statistics/horizontal-file_02-2010.xls
Mallory, W. H. (1926). China: Land of famine. New York: American Geographical Society.
Mauritius (2015). Demographics of Mauritius. http://en.wikipedia.org/wiki/Demographics_of_Mauritius
McEvedy, C. & Jones, R. (1978). Atlas of World Population History. Harmondsworth, Middlesex, England: Penguin Books
Ltd.
McKeown, T. (1983). Food, infection, and population. Journal of Interdisciplinary History, 14, 227-247.
McKeown, T. (2009). The epidemiologic transition: Changing patterns of mortality and population dynamics. American
Journal of Lifestyle Medicine, 3(1), 19S-26S
22
__________________________________________________________________________________
Nelson, R. R. (1956). Theory of the low-level equilibrium trap in underdeveloped countries. American Economic Review,
46(5), 894-908.
Nielsen, R. (2006). The Green Handbook: Seven trends shaping the future of our planet. New York: St. Martins Press.
Nielsen, R. W. (2013a). No stagnation in the growth of population. http://arxiv.org/ftp/arxiv/papers/1311/1311.3997.pdf
Nielsen, R. W. (2013b). Malthusian stagnation or Malthusian regeneration?
http://arxiv.org/ftp/arxiv/papers/1310/1310.5390.pdf
Nielsen, R. W. (2013c). Impacts of demographic catastrophes. http://arxiv.org/ftp/arxiv/papers/1311/1311.1850.pdf
Nielsen, R. W. (2014). Changing the Paradigm. Applied Mathematics, 5, 1950-1963. doi. 10.4236/am.2014.513188
Nielsen, R. W. (2015a). Unified Growth Theory contradicted by the GDP/cap data.
Nielsen, R. W. (2015b). The Insecure Future of the World Economic Growth. Journal of Economic and Social Thought,
2(4), 242-255.
Nielsen, R. W. (2016a). Mathematical analysis of the historical economic growth with a search for takeoffs from stagnation
to growth. Journal of Economic Library, 3(1), 1-23.
Nielsen, R. W.(2016b). Unified Growth Theory contradicted by the mathematical analysis of the historical growth of human
population. Journal of Economics and Political Economy, 3(2), 242-263.
Nielsen, R. W. (2016c). Demographic Transition Theory and its link to the historical economic growth. Journal of Economic
and Political Economy, 3(1), 32-49.
Nielsen, R. W. (2016d). Growth of the world population in the past 12,000 years and its link to the economic growth.
Journal of Economics Bibliography, 3(1), 1-12.
Nielsen, R. W. (2016e). Unified Growth Theory contradicted by the absence of takeoffs in the Gross Domestic Product.
Economic Review, 3(1), 16-27.
Nielsen, R. W. (2016f). The postulate of the three regimes of economic growth contradicted by data. Journal of Economic
and Social Thought, 3(1), 1-34.
Nielsen, R. W. (2016g). Mathematical analysis of historical income per capita distributions. Economic Review, 3(2), 300-
319.
Nielsen, R. W. (2016h). Puzzling properties of the historical growth rate of income per capita explained. Journal of
Economics Library, 3(2), 196-219.
Nielsen, R. W. (2016i). The Unresolved Mystery of the Great Divergence is Solved. Journal of Economic and Social
Thought, 2(2), 242-255.
Olshansky, S. J., & Ault, A. B. (1986). The fourth stage of the epidemiologic transition: The Age of Delayed Degenerative
Diseases. The Milbank Memorial Fund Quarterly, 64(3), 355-391.
Omran, A. R. (1971). The epidemiologic transition: A theory of the epidemiology of population change. Milbank Memorial
Fund Quarterly, 49(4), 509-538.
Omran, A. R. (1983). The epidemiologic transition theory. A preliminary update. Journal of Tropical Pediatrics, 29, 305-
316.
Omran, A. R. (1998). The epidemiological transition revisited thirty years later. World Health
Statistics Quarterly, 51, 99-191.
Omran, A. R. (2005). The epidemiologic transition: A theory of the epidemiology of population change. The Milbank
Quarterly, 83(4), 731-757.
Podlazov, A.V. (2002). Theoretical demography: Models of population growth and global demographic transition (in
Russian). In Advances in Synergetics: The Outlook for the Third Millennium (pp. 324345). Moscow: Nauka.
Robine, J-M. (2001). Redefining the stages of the epidemiological transition by a study of the dispersion of life spans: The
case of France. Population, 13(1), 173-193.
Shklovskii, J. S. (1962). The universe, life and mind, (in Russian). Moscow: Academy of Science, USSR.
Shklovskii, J. S. (2002). The universe life and mind (5th edn.). John Wiley and Sons, Ltd, New York, US.
Snowdon, B. & Galor, O. (2008). Towards a Unified Theory of Economic Growth. World Economics, 9, 97-151.
Statistics Mauritius, (2014). Digest of Demographic Statistics. Ministry of Finance and Economic Development,
Demography Unit, Port Louis, Mauritius.
http://statsmauritius.govmu.org/English/StatsbySubj/Documents/Digest/Demographic13.pdf
Statistics Sweden (1999). Population development in Sweden in a 250-year perspective. Stockholm: Statistics Sweden.
http://www.scb.se/Grupp/Hitta_statistik/Historisk_statistik/_Dokument/Befolkningsutvecklingen_under_250_ar_Historis
k_statistik_1700talet_1800talet_1900talet.pdf
Taeuber, C., & Taeuber, I. B. (1949). World Population Trends. Journal of Farm Economics, 31(1), 241.
Thomlinson, R. (1975). Demographic Problems, Controversy Over Population Control (2nd ed.). Encino, Ca.: Dickenson
Pub.
Trager, J. (1994). The People's Chronology: A Year-by-Year Record of Human Events from Prehistory to the Present. New
York, NY: Henry Holt and Company.
United Nations (1973). The Determinants and Consequences of Population Trends, Population Studies, No. 50., p.10.
United Nations (1999). The World at Six Billion, http://www.un.org/esa/population/publications/sixbillion/sixbilpart1.pdf
United Nations (2013). World Population Prospects: The 2012 Revision, DVD Edition. Washington DC: United Nations,
Department of Economic and Social Affairs, Population Division, http://esa.un.org/unpd/wpp/Excel-
Data/Interpolated.htm
van de Kaa, D. J. (2010). Demographic Transition. In Y. Zeng (Ed.), Encyclopedia of Life Support Systems, Vol. I (pp. 65-
103). Oxford, UK: Eoloss Publishers Co. Ltd.
23
__________________________________________________________________________________
von Hoerner, S. J. (1975). Population explosion and interstellar expansion. Journal of the British Interplanetary Society, 28,
691-712.
Wang, C. (2005). Institutions, demographic transition, and Industrial Revolution: A unified theory. SSRN Working Paper
Series, 08/2005.
Warf, B. (2010). Demographic Transition. In B. Warf (Ed.), Encyclopedia of Geography (pp. 708-711). SAGE: Thousand
Oaks, CA.
Weaver, T. D., Roseman, C. C., & Stringer C. B. (2008). Close correspondence between quantitative and molecular-genetic
divergence times for Neanderthals and modern humans. Proceedings of the National Academy of Sciences of the United
States of America, 105, 4645-4649.
World Bank (2016). GDP Statistics from the World Bank, http://knoema.com/mhrzolg/gdp-statistics-from-the-world-bank
Wrigley, E. A., & Schofield, R.S. (1981). The Population History of England 1541-1871. London: Edward Arnold
(Publishers) Ltd.
WWF, (2010). Living Planet Report 2010: Biodiversity, biocapacity and development. Gland, Switzerland: WWF
International.
24
__________________________________________________________________________________
Changing the direction of the economic and

demographic research
By Ron W. NIELSENa
Abstract. A simple but useful method of reciprocal values is introduced, explained and illustrated. This method
simplifies the analysis of hyperbolic distributions, which are causing serious problems in the demographic and
economic research. It allows for a unique identification of hyperbolic distributions and for unravelling
components of more complicated trajectories. This method is illustrated by a few examples: growth of the world
population during the AD era; growth of population in Africa; economic growth in Western Europe; and the
world economic growth. They show that fundamental postulates of the demographic and economic research are
contradicted by data, even by precisely the same data, which are used in this research. The generally accepted
postulates are based on the incorrect understanding of hyperbolic distributions, which characterise the historical
growth of population and the historical economic growth. In particular, data used, but never analysed, during the
formulation of the Unified Growth Theory show that this theory is based on fundamentally incorrect premises
and thus is fundamentally defective. In this theory, distorted representations of data are used to support
preconceived and incorrect ideas. Precisely the same data, when properly analysed, show that the theory is
incorrect. Application of this simple method of analysis points to new directions in the demographic and
economic research. It suggests simpler interpretations of the mechanism of growth. The concept or the evidence
of the past primitive and difficult living conditions, which might be perhaps described as some kind of stagnation,
is not questioned or disputed. It is only demonstrated that trajectories of the past economic growth and of the
growth of population were not reflecting any form of stagnation and thus that they were not shaped by these
primitive and difficult living conditions. The concept or evidence of an explosion in technology, medicine,
education and in the improved living conditions is not questioned or disputed. It is only demonstrated that this
possible explosion is not reflected in the economic growth and in the growth of population. Growth trajectories
were increasing monotonically during the generally claimed epoch of stagnation and during the claimed
explosion.
Keywords. Hyperbolic Distributions; Reciprocal Values; Economic Growth; Growth of Human Population;
Industrial Revolution; Unified Growth Theory; Growth Regimes; Gross Domestic Product; GDP
1. Introduction
hat we are going to see will change the fundamental postulates in the demographic and economic
W research. It will change radically the way the mechanism of economic growth and of the growth
of population is interpreted. Maybe the change will not come immediately because it is usually
difficult to change the well-established interpretations and explanations but the change will come
because this is the way science works. Incorrect interpretations are not tolerated for too long and it does
not matter who are their advocates.
It might be expected that a complicated proof would be required to achieve such a radical change of
direction in the economic and demographic research, that perhaps some new and complicated
description of the mechanism of growth would have to be proposed. However, the proof turns out to be
exceptionally simple. No complicated mathematics is required but only the way we describe data using
the simplest mathematical representation: the straight line.
George Plya, Hungarian mathematician, observed that when a proof is too simple, youngsters
will be unimpressed (Plya, 1981), but mathematics does not have to be complicated to be useful. He
a
Australia.
. +61407201175
Published as: Nielsen, R. W. (2017). Changing the Direction of the Economic and Demographic Research.
https://arxiv.org/pdf/1708.08673
25
__________________________________________________________________________________
also pointed out that solving problems is a quintessential human activity and the aim is always to find
the simplest solutions.
We are going to present here a proof so simple that it might look trivial. We are going to show how
to change the confusing and complicated distributions describing the historical economic growth and
the historical growth of human population into the simplest representations. We are going to show how
the distributions, which suggest complicated explanations of the mechanism of growth are in fact so
simple that they suggest also a simple mechanism.
Analysis of data describing the historical economic growth and the historical growth of population
might look complicated but it is exceptionally simple. Anyone can do it. However, there is more to the
analysis of data then just looking for their mathematical descriptions. We are going to demonstrate that
this simple method of analysis makes a significant contribution to a better understanding of the
mechanism of the historical growth of population and of the economic growth. It also demonstrates that
there is a need to replace the traditionally used postulates based largely on impressions and conjectures
by postulates based on the mathematical analysis of data.
2. The common problem

Hyperbolic processes appear to be causing a serious problem in the economic and demographic
research. They create such a strong illusion that it deceives even the most experienced and respected
researchers. The common mistake is to see them as being made of two distinctly different components,
slow and fast, with a clear transition between them (Ashraf, 2009; Artzrouni & Komlos, 1985; Baldwin,
Martin & Ottaviano, 2001; Becker, Cinnirella & Woessmann, 2010; Clark 2003, 2005; Currais, Rivera
& Rungo 2009; Dalton, Coats & Asrabadi, 2005; Desment & Parente, 2012; Doepke, 2004; Ehrlich,
1998; Elgin, 2012; Galor 2005a, 2005b, 2007, 2008a, 2008b, 2010, 2011, 2012a, 2012b; Galor &
Michalopoulos, 2012; Galor & Moav 2001, 2002; Galor & Mountford, 2003, 2006, 2008; Galor & Weil,
1999, 2000; Goodfriend & McDermott 1995; Hansen & Prescott 2002; Jones, 2001; Johnson & Brook
2011; Kelly, 2001; Khan 2008; Klasen & Nestmann 2006; Kgel & Prskawetz 2001; Komlos 1989,
2000, 2003; Komlos & Artzrouni 1990; Lagerlf 2003a, 2003b, 2006, 2010; Lee, 2003, 2011; Matar,
2009; McFalls, 2007; McKeown, 2009; McNeill 2000; Mller & Sharp, 2013; Mongomery, n.d.;
Nelson, 1956; Omran 1971, 1983, 1986, 1998, 2005; Robine 2001; Smil 1999; Snowdon & Galor, 2008;
Steinmann, Prskawetz & Feichtinger, 1998; Strulik, 1997; Tamura 2002; Thomlinson 1965; van de Kaa
2008; Voigtlnder & Voth, 2005; Vollrath, 2011; Wang 2005, Warf 2010; Weisdorf 2004; Weiss
2007). The next step is then to try to explain these two perceived stages of growth and the associated
but non-existent transition by proposing distinctly different mechanisms for each of these imagined
components rather than seeing them as representing a single, monotonically increasing distribution
governed by a single mechanism of growth.
This step leads progressively further away from the correct understanding of studied processes
because all efforts are now concentrated on explaining the non-existing features. An increasing number
of scholars are being involved. They do not analyse the relevant data but only describe their impressions
created by hyperbolic illusions. The participating researchers do not question the existence of the
distinctly different stages of growth or of the postulated transition they take them for granted and
concentrate their attention only on the explanation of these phantom features, proposing new
mechanisms, theories and mathematical descriptions without realizing that the apparent distinctly
different two stages of growth do not exist and that there is no transition but a monotonically increasing
hyperbolic distribution. Their mathematical descriptions, complicated and elaborate as they might be,
are not the descriptions of the studied processes but rather the descriptions of phantom impressions
created by hyperbolic illusions.
The perceived two stages of growth are commonly described as stagnation and sustained growth,
while the perceived but non-existent transition as an escape, sprint, sudden spurt, intensification,
acceleration, explosion or by some other similar terms all emphasizing a clear and dramatic change in
the pattern of growth at a certain time. Variety of forces and mechanisms are then proposed to explain
the phantom stages of growth and of the associated but non-existent transition. Efforts are also made to
determine the precise time of the non-existent transition, often placing it around the Industrial
Revolution but sometimes around 1950, without realizing that the determination of this time is
impossible because there was no unusual acceleration at any particular time or over a certain range of
time.
26
__________________________________________________________________________________
Hyperbolic processes are prone to misinterpretations and consequently they have to be analysed with
care. Fortunately, their analysis is exceptionally simple. To show how to avoid being guided by
hyperbolic illusions we shall describe the simple method of their analysis and illustrate it by a few
examples.
2. The method of reciprocal values

Hyperbolic processes can be easily analysed using the method of reciprocal values. This method is so
simple that it can be explained by using just two elementary equations, and yet so powerful that it can
turn around and revolutionize such fields of research as the economic growth and the growth of human
population, the important fields of study because for the first time in human existence we have now
reached ecological limits of our planet and the correct understanding of these two processes is essential
to avoid the undesirable unsustainable developments. We have to know how these processes work and
how to control them. Incorrect interpretations are potentially dangerous and cannot be tolerated. Every
effort has to be made to identify and eliminate any incorrect and misleading explanations.
The first-order hyperbolic distribution is described by the following simple equation:
t ) ( a0 + a1t ) 1 ,
S (= (1)
where S (t ) is the size of a growing entity, while a0 and a1 are constants. For the hyperbolic growth,
a1 < 0 .
Example of hyperbolic growth is shown in Figure 1. It represents the growth of the world population
during the AD era. We can see that hyperbolic distribution describes well the growth of population
during the entire range of data.
Figure 1. Example of hyperbolic growth. Population data (Maddison, 2001) taken from
the same source as used by Galor in his Unified Growth Theory (Galor, 2005a, 2011)
are compared with hyperbolic distribution.
Data have to be analysed but in general they are not. Meticulous analysis of data is particularly
important in the study of hyperbolic processes because they may be strongly misleading. They easily
create an illusion of stagnation followed by explosion. Unfortunately, on seldom occasions when data
are used and displayed, they are displayed in a grossely distorted and self-misleading way (Ashraf,
2009; Galor 2005a, 2005b, 2007, 2008a, 2008b, 2010, 2011, 2012a, 2012b; Galor & Moav, 2002;
Snowdon & Galor, 2008) as shown in Figure 2.
Figure 2 was repreduced from Galors publication (Galor, 2005a, p. 181). His figure was based on
precisely the same source of data (Maddison, 2001) as used in Figure 1 but in this distorted way they
show no resemblance to the the original data. Such distortions were used repeatedly during the
development of the Unified Growth Theory (Galor, 2005a, 2011) making it scientifically unacceptable,
incorrect and unreliable. This Figure shows incorretly that there was a long epoch of stagnation followed
by a takeoff to a fast growth.
27
__________________________________________________________________________________
Figure 2. Example of a distorted presentation of data used in academic literature (Ashraf, 2009;
Galor 2005a, 2005b, 2007, 2008a, 2008b, 2010, 2011, 2012a, 2012b; Galor & Moav, 2002;
Snowdon & Galor, 2008). Data presented in this figure come from precisely the same source
(Maddison, 2001) as the data presented in Figure 1 but in this distorted way they bear no
resemblance to the original data and they suggest incorrect interpretation of the mechanism of
growth. Population is in millions and the GDP/cap in the 1990 International Geary-Khamis dollars
In discussions of the growth of population or of the economic growth it is easy to use some selected
numbers and show that the growth was slow over a long time and fast over a short time. The slow
growth is then interpreted as stagnation controlled by random forces of growth usually associated with
Malthusian positive checks. The fast growth is interepreted as explosion controlled by distinctly
different forces. The triggering mechanism of the alleged explosion is usually claimed to have been
associated with the Industrial Revolution and Galor conveniently locates this alleged takeoff time
around the time of the Industrial Revolution. Of course the growth was slow over a long time and fast
over a short time because it was hyperbolic. It was not because there was stagnation followed by a
takeoff or explosion leading to a new type of growth.
Hyperbolic distribution shown in Figure 1 is described by the eqn (1) with the following parameters
a0 = 8.724 and a1 = 4.267 103 . The fit to the data is remarkably good. Details of analysis are
described in a separate publication (Nielsen, 2016a). They show that there was a major transition from
a fast hyperbolic growth to a slow hyperbolic growth around AD 1 and that there was a minor
disturbance around AD 1300. However, these details are of no concern to us in our present discussion.
What is important to notice is that the growth of human population was indeed slow over a long time
and fast over a short time but that these features are described remarkably well by a single hyperbolic
distribution. These features represent nothing more than mathematical properties of hyperbolic
distribution. They represent a single mechanism of growth.
It is important to point out that hyperbolic distribution increases monotonically. It makes no sense to
divide it into two or three components and assign different mechanisms of growth to each perceived
component. Hyperbolic distribution cannot and should not be divided into separate components and the
best way to see it is to plot their reciprocal values [ S (t )]1 because they convert hyperbolic distribution
to a straight line:
[ S (t )]=
1
a0 + a1t (2)
Reciprocal values of hyperbolic distribution shown in Figure 1 are plotted in Figure 3. It is precisely
the same distribution as shown in Figure 1 but it is presented in a different way. The confusing features
such as the apparent stagnation followed by a takeoff to a fast growth increasing to infinity are replace
by a clear straight line, which is easy to understand. It is obvious now that it would make no sense to
divide such a straight line into distinctly different components and to claim distinctly different
mechanisms of growth. It is also clear that it is impossible to identify a transition from a slow to a fast
28
__________________________________________________________________________________
growth for hyperbolic distributions. There is no transition at any time. The transition occurs gradually
over the entire range of growth. It is impossible to identify a takeoff time because there was no takeoff.
Figure 3. Reciprocal values of the hyperbolic distribution presented in Figure 1 together with the
reciprocal values of the size of population. Complicated hyperbolic distribution is now represented
by a simple straight line, which is easy to understand. The takeoff around 1800 shown in Figure 2
did not happen. The straight line cannot be divided into two distinctly different components making
it clear that hyperbolic distribution shown in Figure 1 is also made of a single component. The slow
and the fast growth shown in Figure 1 follow the same, monotonically-increasing distribution.
The display in Figure 3 is from AD 1000 for two reasons. (1) There is a large gap between AD 1 and
1000 so the display from AD 1000 shows better the agreement of the fitted hyperbolic distribution with
data. (2) Detailed analysis of data for the AD and BC eras shows clearly that between around 500 BC
and AD 500 there was a massive transition from a fast hyperbolic growth during the BC era to a
significantly slower hyperbolic growth during the AD era (Nielsen, 2016a). The point at AD 1 is right
in the middle of this transition and belongs to an entirely different distribution, the distribution
describing the process of transition.
It should be also noticed that the point at AD 1000 in Figure 3 appears to be much further away from
the fitted distribution then the point in Figure 1. The distributions are precisely the same but the display
of reciprocal values magnifies the discrepancies between data and the calculated curve for small values
(Nielsen, 2016b). The smaller are the values of the data and of the calculated distribution the larger is
the magnification.
Reciprocal values allow for a unique identification of the first-order hyperbolic distributions because
only these distributions are represented then by straight lines. This representation allows also for an
easy study of departures from hyperbolic growth because deviations from a straight line are easy to
notice.
Properties of growth do not change by changing the display of data but certain features, which are
difficult or even impossible to recognize in one display can be easily identified in another. It is essential
to remember that in the display of reciprocal values effects are reversed. Thus, for instance, a deviation
to a slower trajectory will be indicated by an upward bending and deviation to a faster trajectory by a
downward bending. An increasing growth is represented by a decreasing trajectory of the reciprocal
values.
When hyperbolic growth is represented by a mathematically generated and gradually changing
curve, such as shown in Figure 1, it might be clear that there was no particular time when the growth
changed from being nearly horizontal to nearly vertical, but when data represented by discrete points
are displayed, such a conclusion might be less obvious. The illusion becomes particularly strong when
only a few strategically located points are selected (Ashraf, 2009; Galor 2005a, 2005b, 2007, 2008a,
2008b, 2010, 2011, 2012a, 2012b; Galor & Moav, 2002; Snowdon & Galor, 2008) from a significantly
29
__________________________________________________________________________________
larger set of data as if to make the deception even more pronounced. Even if the enforcement of the
perceived illusion is unintended, such crude displays of data lead readily to grossly incorrect
interpretations.
However, if reciprocal values of data are displayed, their analysis is immediately made significantly
simpler because if the data follow a simple, first-order hyperbolic distribution, their reciprocal values
will be clearly aligned along a decreasing straight line. It is then obvious that dividing such a straight
line into two sections and claiming two distinctly different regimes of growth governed by two distinctly
different mechanisms simply makes no sense. It also makes no sense to try to locate a point on the
decreasing straight line and claim a transition to a new trajectory because there is obviously no transition
to a new trajectory on a decreasing straight line.
It should be stressed that in this representation only the first-order hyperbolic distributions describing
growth will follow the decreasing straight-line trajectories. It is for this reason that this simple method
is so useful in identifying the first-order hyperbolic distributions. It is a simple and yet powerful method,
which can be used successfully in the analysis of data describing the historical economic growth and
the growth of human population, global, regional or local, because in general they follow simple, first-
order hyperbolic trajectories. Any deviations from such trajectories can be easily investigated. Higher-
order hyperbolic distributions describing growth will be represented by gradually decreasing
trajectories, which could be fitted using higher-order polynomial functions intercepting the horizontal
axis, while the exponential growth will be represented by a decreasing exponential function.
This method might have a more general application but its specifically intended application described
in this publication is to help to avoid being guided by hyperbolic illusions, the unfortunate common
mistake, which often leads to seriously incorrect conclusions as we shall demonstrate in the examples
2 and 3.
Going beyond the intended application, the first-order decreasing hyperbolic distributions will be
represented by the increasing straight lines. Again, in this representation, any deviation from the
decreasing hyperbolic distributions can be easily detected and investigated. Pareto distributions, which
resemble the decreasing hyperbolic distributions, will be represented by gradually increasing functions,
which in this representation might be also easier to investigate.
We shall now illustrate the application of the method of reciprocal values by using three additional
examples: the growth of human population in Africa, the economic growth in Western Europe and the
world economic growth.
3. Further examples
3.1. Growth of population in Africa
The method of reciprocal values can be used to study fine details of growth trajectories, the study which
can then be used not only to improve the fit to data but also to understand the mechanism of growth.
Some distributions might be made of different components, which could be difficult or even impossible
to see in the direct display of data but they could be easily revealed by displaying their reciprocal values.
An excellent example is the growth of human population in Africa shown in Figure 4, constructed using
Maddisons data (Maddison, 2010). These Figure illustrates the added advantage of using the reciprocal
values of data.
The top panel in Figure 4 contains the direct display of data for the growth of human population in
Africa. The displayed shape suggests hyperbolic growth because it is slow over a long time and fast
over a short time.
However, the reciprocal values of data presented in the lower panel reveal that the growth trajectory
is in fact made of two major components: a slow hyperbolic distribution until around 1870 and a fast
hyperbolic distribution after that year. Parameters describing the two hyperbolic components are
=a0 5.105 101 , a1 = 2.036 102 for the slow component and = a0 1.705 102 , a1 =8.515 102 for the
fast component.
Figure 4 shows also that at a later stage, the fast hyperbolic growth started to be diverted to a slower
trajectory as indicated by the upward bending of the trajectory representing the reciprocal values.
Furthermore, it is now clear that the growth of population in Africa was never stagnant and that there
was never a transition from stagnation to growth. The first stage of growth was hyperbolic and the
transition around 1870 was a transition from hyperbolic growth to another hyperbolic growth. All these
features, which are unrecognisable in the direct display of data are clearly seen in the display of the
reciprocal values.
30
__________________________________________________________________________________
Figure 4. Growth of human population in Africa (Maddison, 2010) illustrates how the
method of reciprocal values can serve as an excellent tool in revealing hidden features of
studied distributions.
The pattern revealed by data contradicts the traditional interpretations of the mechanism growth of
human population. There was no escape from the Malthusian trap because there was obviously no trap
in the growth of population. The growth was slow but it was increasing monotonically with no signs of
restrictions imposed by a mythical trap.
The transition from a slow to a fast hyperbolic growth in Africa occurred around the time of the
Industrial Revolution but it was not a transition from a stagnant growth to a new, so-called sustained
growth regime (Galor, 2005a, 2011) but from a hyperbolic growth to another but faster hyperbolic
growth. It was the boosting that coincides with the intensified colonisation of Africa (Duignan & Gunn,
1973; McKay, Hill, Buckler, Ebrey, Beck, Crowston, & Wiesner-Hanks, 2012; Pakenham, 1992).
Contrary to the commonly accepted interpretations, the boosting in the growth of population was not
triggered by a dramatically decreased intensity of Malthusian positive checks but by their dramatic
escalation. It is clear that the accepted interpretations of the effects of Malthusian positive checks are
incorrect. Their increased intensity does not lead to stagnation but to a more intensified growth
(Malthus, 1798; Nielsen, 2016c). The increased intensity of Malthusian positive checks increases the
mortality rate but it also increases the fertility rate with the net result of increasing the rate of natural
increase or the growth rate. This correlation is also clearly demonstrated even now by the growth of
population in poor countries. The poorer they are the faster is the growth of their populations. Thus,
this simple analysis of data assisted by using the reciprocal values already questions the commonly
accepted interpretations of the mechanism of growth of human population.
As shown in Figure 4, reciprocal values of data reveal the details of the mechanism of growth, which
were impossible to identify by the direct display of data. Even if we cannot yet fully explain these
details, we can already see that the growth of the populations in Africa was following a slow hyperbolic
trend until around 1870. Around that year, the growth of human population in Africa experienced an
unprecedented 4-fold acceleration, which diverted the growth into a significantly faster hyperbolic
trajectory. The fast-hyperbolic growth continued until around 1975 when it started to be diverted to a
new but slower trend.
It is this pattern of growth that we have to explain. It is for this pattern of growth that we have to
propose the mechanism of growth. It is not the imaginary pattern of stagnation followed by explosion.
It is not the fictitious Malthusian regime followed by the mythical takeoff from stagnation to an
imagined sustained growth regime (Galor, 2005a, 2011). It is an entirely different pattern, the pattern
31
__________________________________________________________________________________
indicated by the close analysis of data rather than by the pure fantasy. The aim of scientific investigation
is not to explain figments of imagination but the evidence presented by data.
Data are essential in scientific investigations. Assisted by data we shall not be guided by the
erroneous concept of stagnation but by the clear evidence of hyperbolic growth. We shall also not be
guided by the erroneous concept of a takeoff from stagnation to a sustained growth regime but by the
clear evidence of a transition from a hyperbolic growth to another hyperbolic growth. We shall also be
guided by an observation that at a certain stage, around 1975, the long-lasting pattern of hyperbolic
growth has been eventually abandoned and the growth was diverted to an entirely different trajectory.
3.2. Economic growth in Western Europe

Economic growth is measured using the Gross Domestic Product (GDP) or the GDP per capita
(GDP/cap). Galor and Moav (2002) studied economic growth in Western Europe using the data of
Maddison (Maddison, 2001). They have selected a few, strategically located points from a larger set of
data, joined them by straight lines and concluded that there were two distinctly different regimes of
growth: the Malthusian regime (also labelled as the epoch of stagnation, Malthusian era,
Malthusian epoch, Malthusian steady-state equilibrium, Malthusian stagnation or Malthusian
trap) and the sustained economic growth (described also as the Modern Growth Regime,
sustained economic growth and sustained growth regime). Their distorted representation of
Maddisons data is shown in Figure 5.
Figure 5. A typically distorted and self-misleading representation (Galor & Moav, 2002, p. 1136) of
Maddisons data (Maddison, 2001). Compare it with exactly the same data, but not distorted, presented
in Figure 7.
Referring to this crude display of data they also concluded that the Industrial Revolution had a strong
impact on the economic growth causing a dramatic takeoff from stagnation to a fast growth. They made
no attempt to analyse mathematically Maddisons data (Maddison, 2001) but presented a series of
mathematical equations describing their imaginations, which were neither related to nor supported by
the source of data they have used.
It is remarkable that data coming from precisely the same source as they have used contradict their
claims and their interpretations of growth. Extensive analysis of the GDP/cap data, global and regional,
is presented in a separate publication (Nielsen, 2016d). It is shown there that GDP/cap data follow the
monotonically increasing trajectories. They are just the linearly modulated hyperbolic trajectories
(Nielsen, 2017a), i.e. hyperbolic trajectories modulated by the linear time-dependence of the reciprocal
values of the size of population. There is no stagnation and no takeoff to a distinctly different regime
32
__________________________________________________________________________________
of growth. Both, the GDP and the population increase hyperbolically (Nielsen, 2016b, 2016e, 2016f)
and thus monotonically. Consequently, their ratios increase also monotonically.
Figure 6 presents the reciprocal values of the Gross Domestic Product (GDP) for Western Europe
(Maddison, 2001) in the vicinity of the alleged takeoff. The data are well aligned along a decreasing
straight line, which means that they were following the simplest, first-order, hyperbolic distribution
given by the eqn (1).
Figure 6. Reciprocal values of data describing the Gross Domestic Product (GDP) in Western Europe
(Maddison, 2001) in the vicinity of the Industrial Revolution. This is precisely the same source of data
as used by Galor and Moav (2002) to construct their distorted representation shown in Figure 5.
Contrary to their claim, Industrial Revolution had no effect on shaping the economic growth trajectory
in Western Europe, the centre of this revolution. The two regimes of growth claimed by them also did
not exist. The GDP is in billions of the 1990 International Geary-Khamis dollars.
Industrial Revolution was between 1760 and 1840 (Floud & McCloskey, 1994), or around 1800 as
shown in Figure 6. This figure demonstrates clearly and convincingly that the claimed takeoff around
the time of the Industrial Revolution did not happen because the reciprocal values of the GDP data
follow an undisturbed straight-line trajectory representing an undisturbed hyperbolic growth. It is now
clear that there was no takeoff and no escape, great or small, from the hypothetical but non-existing
Malthusian trap, at least from the alleged trap in the economic growth. Maybe there were some other
traps but maybe they are just figments of imagination. It is clear, however, that Industrial Revolution
had absolutely no impact on shaping the economic growth trajectory in Western Europe, the centre of
this revolution.
Industrial Revolution had, no doubt, many other impacts but they are not reflected in the economic
growth trajectory. Their study could be important and interesting but they will not explain the growth
of the GDP. The mechanism of growth was immune to the changes introduced by the Industrial
Revolution. Whatever dramatic changes the Industrial Revolution might have introduced to the general
style of living, to technology and even to the economic marked, these changes obviously were not
shaping the economic growth trajectory.
The absence of a takeoff eliminates also the need for assuming the existence of two distinctly
different regimes of growth. It obviously makes no sense to divide the straight line into two arbitrarily
selected sections and claim distinctly different trajectories governed by distinctly different mechanisms
of growth. What might not have been clear in the direct display of data, is now perfectly obvious if we
display the reciprocal values of data. This display abolishes all elaborate theories and untidy
explanations incorporating such concepts as traps, escapes, takeoffs and stagnation and replaces them
by a simple interpretation of the mechanism of growth suggested by the simple equation describing
33
__________________________________________________________________________________
hyperbolic growth. This conclusion is in agreement with the general observation that natural phenomena
can be usually explained by using simple descriptions.
In Figure 7, the hyperbolic trajectory corresponding to the straight line shown in Figure 6 is extended
to AD 1. The economic growth in Western Europe is well described by a simple, first-order, hyperbolic
distribution. The corresponding parameters are:= a0 9.697 102 and a1 = 5.020 105 . The point at
1950 is not fitted by the hyperbolic trend because from the early 1900s the economic growth in Western
Europe started to be diverted to a slower trajectory, which is again contrary to the claimed boosting or
a transition from stagnation to growth. There was a transition but it was a transition from a
monotonically increasing hyperbolic growth to a slower trajectory.
Figure 7. Data for the Gross Domestic Product (GDP) in Western Europe (Maddison, 2001)
compared with the monotonically increasing hyperbolic distribution. The GDP is in billions of the
1990 International Geary-Khamis dollars.
We cannot claim that the growth was sustained only after the Industrial Revolution because it was
sustained equally strongly during the postulated but non-existent epoch of stagnation. Figure 6 and
Figure 7 show clearly that the concept of two stages of growth is unsupported by data. When stripped
of the hyperbolic illusions, the economic growth is revealed as a simple process, which can be described
using just one, simple mathematical trajectory until the early 1900s when it started to be diverted to a
slower, non-hyperbolic, trajectory. There is no compelling need to make this simple description
complicated.
Growth of the GDP was slow in the past because it was hyperbolic. However, while being slow it
was not stagnant. The growth was fast in recent years because it was hyperbolic. It followed the same
undisturbed hyperbolic distribution as in the past.
We now have a completely different understanding of the economic growth in Western Europe, an
important turnaround in the economic research. Rather than wasting the valuable time, energy and
financial resources on trying to explain the phantom features created by hyperbolic illusions and
magnified by the customary crude representation of data (Ashraf, 2009; Galor 2005a, 2005b, 2007,
2008a, 2008b, 2010, 2011, 2012a, 2012b; Galor & Moav, 2002; Snowdon & Galor, 2008) we can now
focus our attention on the relevant task of trying to explain why the economic growth was so stable over
such a long time and why it was hyperbolic. Rather than writing numerous articles based on impressions
and publishing them in peer-reviewed scientific journals and in academic books we can now concentrate
our attention on the understanding of the science of economic growth. In our investigations, we shall
not be guided by impressions, we shall not be guided by the customary crude representations of data
but by their rigorous mathematical analysis
3.3. Global economic growth

Another example of the application of the method of reciprocal values is the global economic growth.
It is an important example because it questions Galors Unified Growth Theory (Galor, 2005a, 2011)
34
__________________________________________________________________________________
representing the culmination of his work extending over 20 years (Baum, 2011). His theory is based on
an uncritical acceptance of the common interpretations, descriptions and explanations used in the
economic and demographic research. In this sense, his theory offers no new insights.
The fundamental postulate of this theory is again the existence of three regimes of growth: the slow
and stagnant Malthusian Regime, the short and intermediary Post-Malthusian Regime and the fast,
Sustained Growth Regime. Galor also accepts that Industrial Revolution played a crucial role in the
alleged dramatic takeoff from a prolonged stagnation into a rapid and sustained growth.
The welcome initiative in his theory is that he makes an attempt of using repeatedly Maddisons data
(Maddison, 2001). However, he makes not even a single attampt to test his theory by the rigorous
analysis of data. This is a serious omission. The usual practice in any scientific theory is to test it by
data or at least to suggest how it can be tested by data. Galor does not follow this accepted practice. He
does not test his mathematical descriptions by data. Data are used repeatedly but they are never
analysed. They are presented in a typically distorted way, as illustrated in Figures 2 and 5, and in this
distorted way they seem to support the preconceived ideas. His work is based on prejudice and no
attempt is made to check its validity.
When data are used but manipulated to confirm preconceived ideas we are not dealing with science.
We also make no progress and we are not learning anything new or useful.
We shall now use exactly the same source of data and show that the Unified Growth Theory is
scientifically unsustainable. For more extensive discussion of these issues see other publications
(Nielsen, 2016a, 2016b, 2016d, 2016e, 2016f, 2016g, 2016h, 2016i, 2016j, 2017a).
It is hard to see how much can be rescued from Galors Unified Growth Theory. It is hard to see
how many of his descriptions and explanations are based on pure and unsubstantiated speculations. His
theory would have to be minutely analysed. However, its major premises are untenable. All his mind
boggling mysteries of the growth process (Galor, 2005a, p 220), for instance, can be easily explained
(Nielsen, 2016a, 2016d, 2016g, 2016i) there are no mysteries. All his mysteries were created by his
repeatedly distorted presentations of data coming from a reputable source (Maddison, 2001), the data
used during the formulation of his theory but never properly analysed.
His theory certainly does not explain the mechanism of growth because it revolves around the
descriptions of hyperbolic illusions. It does not even describe economic growth. His descriptions are
incorrect because again they are based on the distorted presentations of data and on the unsubstantiated
prejudice.
Theories come and go. Scientific integrity is not tarnished by proposing incorrect explanations and
interpretations but by refusing to correct them or to reject them when they are contradicted by reliable
data.
Reciprocal values of data for the world Gross Domestic Product (GDP) (Maddison, 2001) are shown
in Figure 8. They follow closely a decreasing straight line, which means that the economic growth was
increasing hyperbolically. It is clear that there was no takeoff of any kind, large or small, around the
time of the Industrial Revolution and no repeatedly claimed great escape from the postulated but non-
existing Malthusian trap. The data do not support the existence of the three regimes of growth and thus
contradict the fundamental postulates of the Unified Growth Theory.
The last point of the data shown in Figure 8 is not fitted by the straight line, suggesting a possible
diversion to a slower trajectory. This region can be studied more closely using the extended compilation
of the economic growth data (Maddison, 2010). Their reciprocal values between 1700 and 2003 are
shown in Figure 9 demonstrating clearly that while the Unified Growth Theory claims an unusually
accelerated growth after the alleged but non-existent epoch of stagnation, the data show the opposite
behaviour: a diversion to a slower trajectory after the earlier vigorous, well-sustained and secure
economic growth. Rather than being boosted by the Industrial Revolution, the economic growth
continued along the undisturbed hyperbolic trajectory for about one hundred years after this revolution
and then started to be diverted to a slower trajectory.
Figure 9 illustrates again how the method of reciprocal values can unravel useful details about a
studied process. Not only does it help in an unambiguous and easy identification of hyperbolic
distributions but also it helps in an easy detection of deviations from such distributions. The world
economic growth continues to increase but from the early 1900s it started to be diverted away from the
faster accelerating historical hyperbolic trajectory to a slower trend.
The point of intersection of the reciprocal values with the horizontal axis is the point of singularity
when the growth escapes to infinity. No growth can go beyond this point and any growth close to it
may become unstable, unsustainable and catastrophic. Figures 8 and 9 show how close we are now to
the point of the potential global economic instability and unsustainability.
35
__________________________________________________________________________________
Figure 8. Fundamental concepts of the Unified Growth Theory are contradicted by precisely the same
data (Maddison, 2001), which were used (but never analysed) during its development. Reciprocal
values of data follow closely a decreasing linear distribution representing a monotonically increasing
hyperbolic growth. The three regimes of growth claimed by Galor (2005a, 2011) did not exist. There
was no takeoff around the time of the Industrial Revolution or around any other time. The monotonically
increasing hyperbolic growth remained undisturbed until the 1990s. The GDP is in billions of the 1990
International Geary-Khamis dollars.
Figure 9. Maddisons data (Maddison, 2010) show clearly that while the Unified Growth Theory
(Galor, 2005a, 2011) claims a transition from stagnation to a vigorous growth, the data show the
opposite behaviour: a transition from a vigorous hyperbolic growth to a slower trajectory, as indicated
by the upward bending of the growth trajectory of the reciprocal values during the 1990s and 2000s.
There was no stagnation and no boosting in the economic growth at any time. The claimed (Galor,
2005a, 2011) but non-existent three regimes of growth are also shown. Their existence is contradicted
by data. The GDP is in billions of the 1990 International Geary-Khamis dollars.
Unified Growth Theory claims that after a long epoch of stagnation we have now reached an era of
sustained economic growth, the term repeated 82 times in the first detailed formulation of this theory
(Galor, 2005a), the potentially misleading description because while it is true that the current economic
growth is still sustained the past economic growth was not only sustained but also it was increasing
36
__________________________________________________________________________________
along a more secure trajectory, far away from the point of singularity. Even though the growth is now
diverted to a slower trajectory any further increase can be potentially dangerous.
Reciprocal values of data show that for the first time during the AD era, and probably for the first
time in human existence, we are now trapped between the already high level of economic growth and a
point of no return, or equivalently between the very small reciprocal values of the GDP and zero. Any
intrusion into this narrow gap has to be closely monitored. Even if the trend of the reciprocal values of
the GDP data does not cross the horizontal axis any close approach to this axis could be dangerous,
because it could trigger global economic instability and even a possible global economic collapse.
This simple analysis of data shows how dangerous are the generally accepted postulates presented
in the Unified Growth Theory. The concept of a transition from stagnation to the so-called sustained
growth regimes suggests that now for the first time in human history we can enjoy the sustained
economic growth. Data, however, reveal a diametrically different pattern of growth. It was in the past
that the economic growth was sustainable because it was following a stable hyperbolic distribution,
which was still far away from the point of singularity. Now, however, the reciprocal values of the GDP
are so close to zero, i.e. to the point when the GDP escapes to infinity, that the economic growth is no
longer easily sustainable. The possibility of a serious economic instability is real. Now, the economic
growth has to be closely monitored and controlled. The claim that we are now in the regime of sustained
economic growth is dangerously inaccurate and misleading.
The method of reciprocal values can be also used to demonstrate that two other postulates of the
Unified Growth Theory, the postulate of the differential takeoffs and the postulate of the great
divergence, are contradicted by the mathematical analysis of data coming from the same source, which
was used during the formulation of this theory (Nielsen, 2016b, 2016e, 2016i). Takeoffs never happened
and consequently it makes no sense to claim that they happened at different times for developed and
developing regions. The so called great divergence also never happened. Different regions are on
different levels of development but they follow closely similar trajectories. They are like athletes
running along similar tracks. They do not run in distinctly different directions as incorrectly claimed in
the Unified Growth Theory but in the same direction.
If the economic growth continued along the historical hyperbolic trajectory it would have already
reached a point of no return as indicated by the fitted straight line crossing the horizontal axis. To use
the colourful description of von Foerster, Mora and Amiot (1960), we have been saved from
experiencing a doomsday in the global economic growth. However, the danger of an excessive and
unsustainable growth is still not averted.
Under a suitable control, the economic growth can continue for a long time, but this is precisely the
important point: from now on the economic growth has to be closely monitored and controlled because
it can easily become unsustainable.
Data between 1965 and 2003 follow closely exponential trajectory. Exponential growth does not
increase to infinity at a fixed time but this is hardly any consolation because eventually such a growth
also becomes unsustainable.
Any other continually increasing growth can be unsustainable unless it is increasing to a certain
constant asymptotic value. However, it is extremely difficult to control such a growth because the
growth rate would have to finely tuned to decrease slowly to zero. A constant growth rate, even if small,
would represent the undesirable exponential growth. A growth rate fluctuating around zero would be
safe but our general tendency is to try to increase the growth rate or at least to keep it constant, both
options leading to unsustainable economic growth.
Data describing the world economic growth (Maddison, 2001) are compared in Figure 10 with the
hyperbolic trajectory calculated using the straight-line fitted to the reciprocal values shown in Figure 8.
Parameters describing the historical hyperbolic growth of the world GDP are:= a0 1.716 102 and
6
a1 = 8.671 10 .
Now the puzzling features of the economic growth, the features that prompted so many discussions
in numerous peer-reviewed scientific journals culminating in the formulation of the Unified Growth
Theory, are manifestly clear, and their explanation is surprisingly simple. Over hundreds of years, the
world economic growth was slow because it was hyperbolic. Over a short time, until the early 1900s,
the economic growth was fast because it was hyperbolic it followed the same undisturbed hyperbolic
trajectory as in the past. The apparent transition from a slow to a fast growth is just an illusion created
by the hyperbolic distribution. There was no unusually accelerated transition from the slow to the fast
economic growth. The acceleration was gradual over the entire range of time.
The study presented here shows how important it is to have a clear understanding of the economic
growth and how the simple method of reciprocal values can assist in such studies. Application of this
37
__________________________________________________________________________________
method can not only assist in unravelling different components of growth trajectories but also to avoid
being guided by hyperbolic illusions, which are the source of numerous misinterpretations of economic
growth and of the growth of population culminating in the formulation of the fundamentally flawed and
strongly misleading Unified Growth Theory (Galor, 2005a, 2011).
Figure 10. The data for the world Gross Domestic Product (GDP) (Maddison, 2001) follow
closely the first-order hyperbolic distribution. The claimed three regimes of growth (Galor,
2005a, 2011) did not exist. They are replaced by an uninterrupted and monotonically increasing
hyperbolic growth. The GDP is in billions of the 1990 International Geary-Khamis dollars.

We have described a simple but effective method of analysis of hyperbolic distributions and we have
explained its application by using the growth of the world population during the AD era. We have then
demonstrated the flexibility of this method by using an example of the growth of human population in
Africa. This method can be used to identify uniquely the first order hyperbolic distributions, to reveal
hidden components of growth trajectories and to remove hyperbolic illusions, which are the source of
numerous misinterpretations of economic growth and of the growth of population, misinterpretations
prevailing over a long time in academic literature. This simple method redirects the economic and
demographic research from explanations of phantom features created by hyperbolic illusions, to
explanations based on the scientific analysis of data.
We have presented two other examples of analysis of data: the economic growth in Western Europe
and the global economic growth. All these four examples show that the rigorous analysis of data
contradicts the established knowledge in demography and in the economic research, and in particular,
that it contradicts the fundamental postulates of the Unified Growth Theory (Galor, 2005a, 2011).
However, what we have presented here is just a tip of an iceberg. An entirely new world is opened when
we analyse more data (Nielsen, 2016a, 2016b, 2016c, 2016d, 2016e, 2016f, 2016g, 2016h, 2016i, 2016j,
2016k, 2016l, 2016m, 2016n, 2016o, 2016p, 2017a, 2017b, 2017c, 2017d), the world without stagnation
in the economic growth and in the growth of population, without takeoffs from the alleged stagnation
to growth, the world without complicated and untidy explanations of the mechanism of growth but the
elegant world where data can be described by simple mathematical distributions, the world, which opens
up new vistas for the demographic and economic research.
38
__________________________________________________________________________________
Impressions can be strongly deceptive and persuasive. It is clear that the earth does not move, and
that it does not lie elsewhere than at the centre declared Aristotle. Fortunately, however, in science,
incorrect interpretations are sooner or later corrected.
1. Stagnation. Research based on impressions and reinforced by the customary crude and self-
misleading representations of data (Ashraf, 2009; Galor 2005a, 2005b, 2007, 2008a, 2008b, 2010,
2011, 2012a, 2012b; Galor & Moav, 2002; Snowdon & Galor, 2008), such as shown in Figures 2
and 5, seems to confirm the generally accepted belief that there was an epoch of stagnation in the
economic growth and in the growth of population. Scientific analysis of precisely the same (but
undistorted) data demonstrates that there was no stagnation and that the economic growth and the
growth of population followed monotonically increasing hyperbolic distributions.
2. Takeoffs. Research based on impressions seems to indicate that there was a transition from
stagnation to growth described usually as a takeoff or explosion. Scientific analysis of precisely
the same (but undistorted) data demonstrates that there was no takeoff or explosion and that
economic growth and the growth of population continued to follow the monotonically increasing
hyperbolic distributions. What appears as a takeoff or explosion is in fact the natural continuation
of hyperbolic growth.
3. Industrial Revolution. Research based on impressions seems to indicate that Industrial Revolution
played a crucial role in the economic growth and in the growth of population causing a dramatic
acceleration (boosting) in the growth trajectories, described as takeoffs. Scientific analysis of
precisely the same (but undistorted) data demonstrates that Industrial Revolution had absolutely
no impact on shaping growth trajectories. Industrial Revolution can be linked to other impacts but
not to shaping the population or the economic growth trajectories. This might be surprising but the
evidence in data is undisputable and we have to accept it.
4. Regimes of growth. Research based on impressions seems to suggest that there were two or maybe
even three distinctly different regimes of growth governed by distinctly different mechanisms
(Galor, 2005a, 2011). Scientific analysis of precisely the same (but undistorted) data demonstrates
that these two or three distinctly different regimes of growth did not exist. The growth was
hyperbolic until recently when it started to be diverted to slower trajectories.
5. Mysteries. Research based on impressions resulted in claiming a series of mind-boggling and
perplexing mysteries of the growth process (Galor, 2005a, pp. 177, 220). Scientific analysis
of precisely the same data demonstrates that all these mysteries belong to the world of fiction
created by a good dose of fantasy guided by the misleading impressions and reinforced by the
customarily distorted presentations of data (Ashraf, 2009; Galor 2005a, 2005b, 2007, 2008a,
2008b, 2010, 2011, 2012a, 2012b; Galor & Moav, 2002; Snowdon & Galor, 2008) such as shown
in Figures 2 and 5. Science is supported by a methodical analysis of data. There are no mysteries
when precisely the same data are properly analysed.
In particular, the mystery of the great divergence is explained: there was no great divergence
(Nielsen, 2016i). Various regions are on different levels of economic growth but they all follow
closely similar trajectories. Their economic growth did not diverge into distinctly different
trajectories as incorrectly suggested by the crude representations of data.
The mystery of the alleged sudden spike in the growth rate of income per capita has been
explained: there was no sudden spike (Nielsen, 2016g). The growth rate of income per capita
followed a monotonically increasing trajectory, which is readily represented by a mathematical
distribution derived using hyperbolic growth for the growth of the GDP and for the growth of
population.
The mystery of the puzzling features of income per capita has been explained (Nielsen, 2017a).
The distribution representing income per capita is nothing more than just a linearly modulated
hyperbolic distribution. It reflects nothing more than the purely mathematical property of dividing
two hyperbolic distributions.
Other questions listed by Galor as representing the mysteries of the growth process can be easily
answered. They refer to features that do not exist, features based on impressions reinforced by
ineffectual handling of empirical evidence. They are in the same category as the question Why
does the sun revolve around the earth?
6. Mechanism. Research based on impressions leads to proposing numerous complicated mechanisms
of growth. Scientific analysis of data shows that the mechanism of growth is exceptionally simple
39
__________________________________________________________________________________
(Nielsen, 2016p), which is hardly surprising because hyperbolic distributions are described by an
exceptionally simple equation [see eqn (1)].
7. Unified Growth Theory. Research based on impressions prompted the development of a Unified
Growth Theory (Galor, 2005a, 2011). Mathematical analysis shows that the fundamental postulates
of this theory are contradicted by the same data, which were used during its development. Galor
could have saved 20 years of his life and could have directed his academic skills to developing a
useful theory if he did what any scientist is supposed to do: if he based his deductions and
explanations on a scientific analysis of data. He had access to excellent data but he did not analyse
them. He was guided by preconceived ideas and he supported them by distorted presentations of
data.
The analysis data suggests new lines of research. Thus, for instance, the relevant question is not why
the historical economic growth was so unstable in the past or what caused the perceived transition from
alleged stagnation to growth but why the economic growth was so remarkably stable in the past. The
same question applies to the growth of population but it was already answered (Nielsen, 2016c, 2017d).
The growth of population was remarkably stable because of the combination of the generally low
impacts of demographic catastrophes (at least on the global and regional scales) and the high level of
human resilience expressed in the efficient process of regeneration (Malthus, 1798; Nielsen, 2016c). If
we accept that there is a close relationship between the growth of population and the economic growth,
then the question about the stability of the historical economic growth has been also already answered.
However, it is possible that some new insights could be still added to this explanation.
The relevant question is not why the Industrial Revolution and the unprecedented technological
development boosted the economic growth because they did not. The relevant question is why the
Industrial Revolution and the unprecedented technological development did not boost the economic
growth. Why these apparently strong technological and socio-economic forces had no impact on
shaping the economic growth trajectories.
The relevant question is not why the economic growth increased so fast in modern time, because we
have shown that this fast increase was just the natural continuation of the monotonically increasing
hyperbolic growth until in recent years it started to be diverted to a slower but still fast-increasing
trajectory. The relevant question is why the economic growth was diverted to a slower trajectory. What
new force or forces were so strong that they were able to overpower the historically strong force of
growth. Another relevant question is also whether this new trajectory is likely to develop into a
historically preferable and potentially catastrophic, hyperbolic growth. Furthermore, the relevant
question is how to control the current fast economic growth. The same question applies also to the
growth of population but it was at least partly answered in the study of the effects of Malthusian positive
checks (Nielsen, 2016c). The primary if not exclusive way of controlling the growth of human
population is to improve the living conditions in developing countries.
The method of reciprocal values is so simple that it can be used by anyone and it is, therefore,
expected that it will be of interest to many scientists who look for a simple method of analysis of
empirical evidence, a method that does not involve any complicated mathematical formulae, any
intricate mathematical algorisms or the use of powerful computers but a simple display of data and a
remarkably simple fitting procedure. We have demonstrated that even a simple mathematical method
can have a dramatic influence on scientific research.
It is essential to understand that by claiming that there was no stagnation in the economic growth or
in the growth of population we are not claiming that there was no stagnation in the standard of living.
We are only claiming that the two processes were decoupled. We might, if we insist, describe the past
general living conditions as primitive or even stagnant, but there is no evidence that they were shaping
the trajectories describing the growth of population or the economic growth.
It is also essential to understand that by claiming that there was no takeoff in the economic growth
or in the growth of population we are not claiming that there was no takeoff in the technological
development, or generally in the intellectual progress and in the dramatic changes in human experience
and in living conditions. We are only claiming that these possible takeoffs had no impact on changing
the economic growth trajectories or the trajectories describing the growth of population. There were no
takeoffs in any of these two processes. Industrial Revolution can be linked with many changes in human
living experience but all these changes had no impact on changing the economic or demographic growth
trajectories.
There is no reason why scientific evidence presented here and in other related publications (Nielsen,
2016a, 2016b, 2016c, 2016d, 2016e, 2016f, 2016g, 2016h, 2016i, 2016j, 2016k, 2016l, 2016m, 2016n,
40
__________________________________________________________________________________
2016o, 2016p, 2017a, 2017b, 2017c, 2017d) should not be accepted by the scientific community. The
only alternative option is to reject data but this would be no longer science.
Even Galor and his associates accept the same data and use them in their research. Their unfortunate
mistake was only in choosing to support their investigations by the grossly distorted and self-misleading
representations of data (Ashraf, 2009; Galor 2005a, 2005b, 2007, 2008a, 2008b, 2010, 2011, 2012a,
2012b; Galor & Moav, 2002; Snowdon & Galor, 2008). Consequently, the only way to reject scientific
evidence and to accept the doctrines of stagnation and takeoffs, and all other associated erroneous
explanations of the dynamics of the economic growth and of the growth of population, is to accept data
but distort them in such a way as to make them to conform with preconceived ideas, but then again it is
not science.
Evidence in data is overwhelming and leaves no room for accepting incorrect interpretations. In order
to have progress in the demographic and economic research, incorrect interpretations of growth have to
be abandoned and a new paradigm has to be developed. There is no other, scientifically justified, way.
A serious mistake in scientific investigations is not in stumbling and in making mistakes but in refusing
to learn from them and to correct them.
References
Ashraf, Q, H. (2009). Essays on deep determinants of comparative economic development. Providence, Rode Island: Brown
University.
Artzrouni, M., & Komlos, J. (1985). Population growth through history and the escape from the Malthusian Trap: A
homeostatic simulation model. Genus, 41, 21-39. PMid:12280407.
Baldwin, R. E., Martin, P., & Ottaviano, G. I. P. (2001). Global income divergence, trade, and industrialization: The
geography of growth take-offs. Journal of economic growth, 6, 5-37. http://dx.doi.org/10.1023/A:1009876310544.
Baum, D. (2011) Oded Galor: Economic Growth Process over 100,000 Years. http://news.brown.edu/features/2011/06/galor
Becker, S. O., Cinnirella, F., & Woessmann, L. (2010). The trade-off between fertility and education: Evidence from before
the demographic transition. J Econ Growth, 15,177204, DOI 10.1007/s10887-010-9054-x.
Clark, G. (2003). The great escape: The Industrial Revolution in theory and in history. Working Paper. Davis: University of
California.
Clark, G. (2005). The condition of the working class in England, 1209-2004. Journal of Political Economy, 113(6), 1307-
1340. http://dx.doi.org/10.1086/498123.
Currais, L., Rivera, B., & Rungo, P. (2009). Health improvements and the transition out of Malthusian stagnation. Bulletin
of Economic Research, 61(4), 335-352. http://dx.doi.org/10.1111/j.1467-8586.2008.00291.x.
Dalton, T. R., Coats, R. M., & Asrabadi, B. R. (2005). Renewable resources, property-rights regimes and endogenous
growth. Ecological Economics, 52, 31 41. http://dx.doi.org/10.1016/j.ecolecon.2004.03.033
Desment, K., & Parente, S. L. (2012). The evolution of markets and the revolution of industry: a unified theory of growth. J
Econ Growth, 17, 205234. http://dx.doi.org/10.1007/s10887-012-9080-y
Doepke, M. (2004). Accounting for fertility decline during the transition to growth. Journal of Economic Growth, 9(3),
347-383. http://dx.doi.org/10.1023/B:JOEG.0000038935.84627.e4
Duignan, P., & Gunn, L. H. (Eds.) (1973). Colonialism in Africa 1870 1960: A Bibliographic Guide to Colonialism in Sub-
Saharan Africa. Cambridge, UK: Cambridge University Press.
Ehrlich, P. R. (1998). The population bomb. Cutchogue, NY: Buccaneer Books.
Elgin, C. (2012). A theory of economic development with endogenous fertility. Macroeconomic Dynamics, 16, 686-705.
http://dx.doi.org/10.1017/S1365100510000842
Floud, D., & McCloskey, D. N. (1994). The Economic History of Britain since 1700. Cambridge, UK: Cambridge
University Press.
Galor, O. (2005a). From Stagnation to Growth: Unified Growth Theory. In P. Aghion & S. Durlauf (Eds.), Handbook of
Galor, O. (2005b). The demographic transition and the emergence of sustained economic growth. Journal of the European
Economic Association, 3(2/3), 494-504. http://dx.doi.org/10.1162/jeea.2005.3.2-3.494
Galor, O. (2007). Multiple growth regimes Insights from unified growth theory. Journal of Macroeconomics, 29(3), 470-
475. http://dx.doi.org/10.1016/j.jmacro.2007.06.007
Galor, O. (2008a). Economic growth in the very long run. In S. N. Durlauf & L. E. Blume (Eds.), The New Palgrave
Dictionary of Economics. New York: Palgrave Macmillan. The New Palgrave Dictionary of Economics Online,
http://dx.doi.org/10.1057/9780230226203.0434, PMCid:PMC2694025
Galor, O. (2008b). Comparative economic development: Insight from Unified Growth Theory.
Galor, O. (2010). The 2008 Lawrence R. Klein lecture Comparative economic development: Insights from Unified Growth
Theory. International Economic Review, 51(1), 1-44. http://dx.doi.org/10.1111/j.1468-2354.2009.00569.x
Galor, O. (2012a). The demographic transition: causes and consequences. Cliometrica, 6, 1-28.
http://dx.doi.org/10.1007/s11698-011-0062-7
Galor, O. (2012b). Unified Growth Theory and comparative economic development.
41
__________________________________________________________________________________
Galor, O., & Michalopoulos, M. (2012). Evolution and the growth process: Natural selection of entrepreneurial traits.
Journal of Economic Theory, 147, 759-780. http://dx.doi.org/10.1016/j.jet.2011.04.005
Galor, O., & Moav, O. (2001). Evolution and growth. European Economic Review, 45, 718-729.
http://dx.doi.org/10.1016/S0014-2921(01)00112-X
Galor, O., & Moav, O. (2002). Natural Selection and the Origin of Economic Growth. The Quarterly Journal of Economics,
117(4), 1133-1191. http://dx.doi.org/10.1162/003355302320935007
Galor, O., & Mountford, A. (2003). Trading population for productivity. Working paper, Brown University.
Galor, O., & Mountford, A. (2006). Trade and the great divergence: The family connection. The American Economic
Review, 96(2), 299-303. http://dx.doi.org/10.1257/000282806777212378
Galor, O., & Mountford, A. (2008). Treading population for productivity: Theory and evidence. Review of Economic
Studies, 75, 1143-1179. http://dx.doi.org/10.1111/j.1467-937X.2008.00501.x
Galor, O., & Weil, D. N. (1999). From Malthusian stagnation to modern growth. The American Economic Review, 89(2),
150-154. http://dx.doi.org/10.1257/aer.89.2.150
Galor, O., & Weil, D. N. (2000). Population, Technology, and Growth: From Malthusian Stagnation to the Demographic
Transition and beyond. The American Economic Review, 90(4), 806-828. http://dx.doi.org/10.1257/aer.90.4.806
Goodfriend, M., & McDermott, J. (1995). Early development. American Economic Review, 85, 116133.
http://dx.doi.org/10.1257/00028280260344731
Jones, C. I. (2001). Was the industrial revolution inevitable? Economic growth over the very long run. Advances in
Macroeconomics, 1, 143.
http://dx.doi.org/10.1098/rspb.2011.0343
Kelly, M. (2001). Linkages, thresholds, and development. Journal of Economic Growth, 6, 39-53.
http://dx.doi.org/10.1023/A:1009846027382
Khan, A. (2008). The Industrial Revolution and the demographic transition. Business Review, First Quarter, 9-15.
Klasen, S., & Nestmann, T. (2006). Population, population density and technological change. Journal of Population
Economics, 19(3): 611-626. http://dx.doi.org/10.1007/s00148-005-0031-1
Kgel, T., & Prskawetz, A. (2001). Agricultural Productivity Growth and Escape from the Malthusian Trap. Journal of
Economic Growth, 6(4), 337-357. http://dx.doi.org/10.1023/A:1012742531003
Komlos, J. H. (2000). The Industrial Revolution as the escape from the Malthusian trap. Journal of European Economic
History, 29, 307-331.
Komlos, J. (2003). The Industrial Revolution as the escape from the Malthusian trap. Munich Discussion Paper No. 2003-
13. Volkswirtschaftliche Fakultt, Ludwig-Maximilians-Universitt Mnchen
Komlos, J., & Artzrouni, M. (1990). Mathematical investigations of the escape from the Malthusian trap. Mathematical
Population Studies, 2(4), 147-157. http://dx.doi.org/10.1080/08898489009525313
PMid:12283330
Lagerlf, N-P. (2003a). From Malthus to Modern Growth: Can Epidemics Explain Three Regimes? International Economic
Review, 44, 755-777. http://dx.doi.org/10.1111/1468-2354.t01-1-00088
Lagerlf, N-P. (2003b). Mortality and Early Growth in England, France and Sweden. Scand. J. of Economics, 105(3), 419
439. http://dx.doi.org/10.1111/1467-9442.t01-2-00006
Lagerlf, N-P. (2006). The Galor-Weil model revisited: A quantitative exercise. Review of Economic Dynamics, 9(1), 116-
142. http://dx.doi.org/10.1016/j.red.2005.07.002
Lagerlf, N-P. (2010). From Malthusian war to Solovian peace. Review of Economic Dynamics, 13(3), 616-636.
http://dx.doi.org/10.1016/j.red.2009.10.008
Lee, R. (2003). The demographic transition: Three centuries of fundamental change. Journal of Economic Perspectives,
17(4), 167-190. http://dx.doi.org/10.1257/089533003772034943
Lee, R. (2011). The outlook for population growth. Science, 333, 569-573. http://dx.doi.org/10.1126/science.1208859
PMid:21798936 PMCid:PMC3155613
http://dx.doi.org/10.1787/9789264189980-en
Matar, H. F. (2009). The ephemeral existence of humanity. Mankind Quarterly, 49(3/4), 381-392.
McFalls, J.A. Jr, (2007). Population: A Lively Introduction (5th ed.). Washington DC: Population Reference Bureau.
McKay, J. P., Hill, B. D., Buckler, J., Ebrey, P. B., Beck, R. B., Crowston, C. H., & Wiesner-Hanks, M. E. (2012). A History
of World Societies: From 1775 to Present. Volume C From 1775 to the Present. Ninth edition. Boston, MA: Bedford
Books.
Journal of Lifestyle Medicine, 3(1), 19S-26S. http://dx.doi.org/10.1177/1559827609335350
PMid:20161566 PMCid: PMC2805833
McNeill, J. R. (2000). Something new under the sun; An environmental history of the twentieth-century world. New York:
W. W. Norton & Company, Inc.
Mller, N. F., & Sharp, P. (2013). Malthus in cointegration space: Evidence of a post-Malthusian pre-industrial England. J
Econ. Growth, 31(1), 68-93.
Montgomery, K. (n.d.). The demographic transition. http://www.uwmc.uwc.edu/geography/demotrans/demtran.htm
42
__________________________________________________________________________________
Nelson, R. R. (1956). Theory of the low level equilibrium trap in underdeveloped countries. American Economic Review,
46(5), 894-908.
Nielsen, R. W. (2016a). Growth of the world population in the past 12,000 years and its link to the economic growth.
Journal of Economics Bibliography, 3(1), 1-12. http://www.kspjournals.org/index.php/JEB/article/view/607
Nielsen, R. W. (2016b). Mathematical analysis of the historical economic growth with a search for takeoffs from stagnation
to growth. Journal of Economic Library, 3(1), 1-23. http://www.kspjournals.org/index.php/JEL/article/view/606
Nielsen, R. W. (2016c). The dichotomy of Malthusian positive checks: Destruction and even more intensified regeneration.
Nielsen, R. W. (2016d). Mathematical analysis of the historical income per capita distributions. Economic Review, 3(2),
300-319. http://www.kspjournals.org/index.php/TER/article/view/766/889
Economic Review, 3(1), 16-27. http://www.kspjournals.org/index.php/TER/article/view/650;
Nielsen, R. W. (2016f). Unified Growth Theory contradicted by the mathematical analysis of the historical growth of human
http://www.kspjournals.org/index.php/JEST/article/view/777/911
Nielsen, R. W. (2016g). Puzzling properties of the historical growth rate of income per capita explained. Journal of
Economics Library, 3(2), 241-256. http://www.kspjournals.org/index.php/JEL/article/view/756/867
Nielsen, R. W. (2016h). The postulate of the three regimes of economic growth contradicted by data. Journal of Economic
and Social Thought, 3(1), 1-34. http://www.kspjournals.org/index.php/JEST/article/view/669;
http://arxiv.org/ftp/arxiv/papers/1602/1602.05718.pdf.
Nielsen, R. W. (2016i). The unresolved mystery of the great divergence is solved. Journal of Economic and Social Thought,
3(2), 196-219. http://www.kspjournals.org/index.php/JEST/article/view/777/911
Nielsen, R. W. (2016j). Industrial Revolution did not boost the economic growth or the growth of population even in the
United Kingdom. Journal of Economic Bibliography, 3(4), 434-446.
http://kspjournals.org/index.php/JEB/article/view/1053
Nielsen, R. W. (2016k). Scientifically unacceptable established knowledge in demography and in economic research.
Journal of Economic Library, 3(3), 429-457. http://kspjournals.org/index.php/JEL/article/view/918
Nielsen, R. W. (2016l). Demographic Transition Theory and its link to the historical economic growth. Journal of Economic
and Political Economy, 3(1), 32-49. http://www.kspjournals.org/index.php/JEPE/article/view/618/798
Nielsen, R. W. (2016m). Mathematical analysis of income per capita in the United Kingdom. Economic Review, 3(4), 453-
463. http://kspjournals.org/index.php/TER/article/view/1054
Nielsen, R. W. (2016n). Interpretations of Hyperbolic Growth. Journal of Economics and Political Economy, 3(4), 594-626.
http://kspjournals.org/index.php/JEPE/article/download/1063/1099
Nielsen, R. W. (2016o). The Law of Growth. Journal of Economic and Social Thought, 3(4), 349-357.
http://kspjournals.org/index.php/JEPE/article/view/1063
Nielsen, R. W. (2016p). Mechanism of Hyperbolic Growth Explained. Journal of Economic Library. 3(4), 411-428.
Nielsen, R. W. (2017a). Puzzling Features of Income per Capita Explained. Journal of Economics Bibliography. 4(1), 10-24.
http://www.kspjournals.org/index.php/JEB/article/download/1155/1216
Nielsen, R. W. (2017b). Population and Economic Growth in Australia: 8,000 BC AD 1700 and Extended to 60,000 BC.
Journal of Economic and Social Thought, 4(1), 41-54.
http://www.kspjournals.org/index.php/JEST/article/download/1156/1206
Nielsen, R. W. (2017c). Economic Growth and the Growth of Human Population in the Past 2,000,000 Years. Journal of
Economic Bibliography, 4(2). Forthcoming.
Nielsen, R. W. (2017d). Demographic Catastrophes Did Not Shape the Growth of Human Population or Economic Growth.
Journal of Economic and Social Thought, 4(2). Forthcoming.
Fund Quarterly, 49(4), 509-538. http://dx.doi.org/10.2307/3349375 PMid:5155251
316. http://dx.doi.org/10.1093/tropej/29.6.305 PMid:6672237
Omran, A. R. (1986). The fourth stage of the epidemiologic transition: The age of delayed degenerative diseases. The
Milbank Quarterly, 64(3), 355-391. http://dx.doi.org/10.2307/3350025
Omran, A. R. (1998). The epidemiological transition revisited thirty years later. World Health Statistics Quarterly, 51, 99-
191.
Quarterly, 49(4), 731-757. http://dx.doi.org/10.1111/j.1468-0009.2005.00398.x
PMid:16279965 PMCid: PMC2690264
Pakenham, T. (1992). The Scramble for Africa: White Mans Conquest of the Dark Continent from 1876-1912. New York:
Avon Books.
Plya, G. (1981). Mathematical Discovery: On Understanding, Learning, and Teaching Problem Solving. John Wiley &
Sons, Inc., Hoboken, NJ.
43
__________________________________________________________________________________
Robine, J-M. (2001). Redefining the Stages of the Epidemiological Transition by a Study of the Dispersion of Life Spans:
The Case of France. Population, 13(1), 173-193.
Smil, V. (1999). Detonator of the Population Explosion. Nature, 400, 415. http://dx.doi.org/10.1038/22672
Snowdon, B., & Galor, O. (2008). Towards a Unified Theory of Economic Growth. World economics, 9(2), 97-151.
Steinmann, G., Prskawetz, A., & Feichtinger, G. (1998). A model on the escape from the Malthusian trap. J Popul Econ,
11(4), 535-550. http://dx.doi.org/10.1007/s001480050083 PMid:12294786
Strulik, H. (1997). Learning-by-doing: Population pressure, and the theory of demographic transition. Journal of Population
Economics, 10(3), 285-298. http://dx.doi.org/10.1007/s001480050044 PMid:12292961
Tamura, R. F. (2002). Human capital and the switch from agriculture to industry. Journal of Economic Dynamics and
Control, 27, 207-242. http://dx.doi.org/10.1016/S0165-1889(01)00032-X
Thomlinson, R. (1965). Population dynamics: Causes and consequences of world demographic change. New York:
Random House.
van de Kaa, D. J. (2008). Demographic Transition. Encyclopedia of Life Support Systems, Demography, Vol. 2. Paris:
UNESCO
Voigtlnder, N., & Voth, H.-J. (2006). Why England? Demographic factors, structural change and physical capital
accumulation during the industrial revolution. Journal of Economic Growth, 11(4), 319361.
http://dx.doi.org/10.1007/s10887-006-9007-6
Vollrath, C. (2011). The agricultural basis of comparative development. J Econ Growth, 16, 343370.
http://dx.doi.org/10.1007/s10887-011-9074-1
http://dx.doi.org/10.1126/science.132.3436.1291
Wang, C. (2005). Institutions, demographic transition, and Industrial Revolution: A Unified Theory. SSRN Working Paper
Series, 08/2005.
Warf, B. (2010). Demographic transition. In B. Warf (Ed.), Encyclopedia of Geography (pp. 708-711) Thousand Oaks, CA:
SAGE.
Weisdorf, J. L. (2004). From stagnation to growth: Revising three historical regimes. Journal of Population Economics, 17,
455-472. http://dx.doi.org/10.1007/s00148-004-0182-5
Weiss, V. (2007). The population cycle drives human history from a eugenic phase into a dysgenic phase and eventual
collapse. The Journal of Social, Political, and Economic Studies, 32(3): 327-358.
44
__________________________________________________________________________________
Mathematical Analysis of the Historical Economic Growth with a

Search for Takeoffs from Stagnation to Growth
By Ron W. NIELSENa
Abstract. Data describing historical economic growth are analysed. Included in the analysis is the world and
regional economic growth. The analysis demonstrates that historical economic growth had natural tendency to
follow hyperbolic distributions. Parameters describing hyperbolic distributions have been determined. A search
for takeoffs from stagnation to growth produced negative results. This analysis throws a new light on the
interpretation of the mechanism of the historical economic growth and suggests new lines of research.
Keywords. Historical economic growth; regimes of growth; stagnation; takeoffs; Malthusian trap; hyperbolic
growth.
JEL. A10, B15, B16, B22, C29.
1. Introduction
T he latest publication of excellent data by the world-renown economist (Maddison, 2001, 2010)
offers an unprecedented opportunity to study the mechanism of the historical economic growth.
Earlier study (Nielsen, 2014), based on these data, indicated that historical economic growth can
be described using hyperbolic distributions in much the same way as the growth of human population
(von Foerster, Mora & Amiot, 1960). Unlike exponential growth, which is more familiar and which can
be easier to understand, hyperbolic distributions are strongly deceptive because they appear to be made
of two distinctly different components, slow and fast, joined perhaps by a certain transition component.
This illusion is so strong that even the most experienced researchers can be easily deceived particularly
if their research is based on a limited body of data, as it was in the past. Fortunately, Maddisons data
solve this problem, and fortunately also their analysis is trivially simple because, as pointed out earlier
(Nielsen, 2014), hyperbolic distributions can be easily identified and analysed using the reciprocal
values of data.
Hyperbolic distribution describing growth is represented by a reciprocal of a linear function:
1
S (t ) = , (1)
a kt
where S (t ) is the size of the growing entity, in our case the Gross Domestic Product (GDP), while a and
k are positive constants.
The reciprocal of such hyperbolic growth, 1/ S (t ) , is represented by a decreasing linear function:
1
= a kt . (2)
S (t )
a
Australia.
. +61407201175
. r.nielsen@griffith.edu.au; ronwnielsen@gmail.com
Published as: Nielsen, R. W. (2016). Mathematical analysis of the historical economic growth with a search for takeoffs
from stagnation to growth. Journal of Economic Library, 3(1), 1-23.
45
__________________________________________________________________________________
Hyperbolic distributions should not be confused with hyperbolic functions ( sinh(t ) , cosh(t )
etc). Furthermore, reciprocal functions should not be confused with inverse functions. Thus, for
instance, for the expression given by the eqn (1) the objective of finding the inverse function would be
to calculate time t for a given size S (t ) . The roles of the dependent and independent variables would be
reversed. For the reciprocal function, the objective is to convert eqn (1) into eqn (2). The roles of
dependent and independent variables are not changed.
Reciprocal values help in an easy and generally unique identification of hyperbolic growth because
in this representation hyperbolic growth is given by a decreasing straight line. Apart from serving as an
alternative way to analyse data, reciprocal values allow also for the investigation of even small
deviations from hyperbolic distributions because deviations from a straight line can be easily noticed.
Reciprocal values allow also for an easy identification of different components of growth. This
property can be used in comparing empirical information with theoretical interpretations (Galor, 2005,
2011), which are based on the assumption of the existence of different components of growth.
When comparing mathematically-calculated distributions with the reciprocal values of data, we have
to remember that the sensitivity of the reciprocal values to small deviations increases with the
decreasing size S of the growing entity.
Suppose we have two values of S at a given time: S1 and S 2 , representing, for instance, the empirical
and calculated values. It is clear that
1 S
=
, (3)
S S1S 2
where (1/ S ) is the difference between two inverse values and S is the difference between S values.
For a given S , (1 / S ) increases rapidly with the decreasing S1 and S 2 values. The separation
of small values of data from calculated distributions are magnified. Similar magnifications, though less
pronounced, are also shown in the semilogarithmic displays of data. We shall use both displays to
examine the quality of fits to the data.
It should be noted that the decreasing reciprocal values describe growth, while a deviation to larger
reciprocal values describes decline. Consequently, a diversion to a faster trajectory will be indicated by
a downward bending of a trajectory of the reciprocal values, away from an earlier observed trajectory,
while the diversion to a slower trajectory will be indicated by an upward bending.
The data describing the historical economic growth (Maddison, 2001, 2010) do not allow for a
detailed analysis below AD 1500 because there are two large gaps in the data: between AD 1 and 1000
and between AD 1000 and 1500. The best sets of data are from AD 1500. However, the compilation
prepared by Magnuson appears to be the best and the most reliable source of data describing the
historical economic growth.
Throughout the analysis presented here, the values of the Gross Domestic Product (GDP) will be
expressed in billions of the 1990 International Geary-Khamis dollars. All diagrams are presented in the
Appendix
Theories play an important role in scientific research because they crystallise interpretations of
studied phenomena. However, theories have to be always tested by data. In science it is important to
look for data confirming theoretical explanations but it is even more important to discover contradicting
evidence, because data confirming a theory confirm only what we already know but contradicting
evidence may lead to new discoveries.
Currently, the most complete theory describing the mechanism of the historical economic growth
appears to be the Unified Growth Theory (Galor, 2005, 2008, 2011, 2012). One of the fundamental
postulates of this theory is the postulate of the existence of three regimes of growth governed by three
distinctly different mechanisms: (1) the Malthusian regime of stagnation, (2) the post-Malthusian
regime, and (3) the sustained-growth regime.
According to Galor (2005, 2008, 2011, 2012), Malthusian regime of stagnation was between 100,000
BC and AD 1750 for developed regions and between 100,000 BC and AD 1900 for less-developed
regions. The claimed starting time appears to be based entirely on conjecture because Maddisons data
are terminated at AD 1 and even they contain significant gaps below AD 1500. The post-Malthusian
46
__________________________________________________________________________________
regime was allegedly between AD 1750 and 1850 for developed regionsand and from 1900 for less-
developed regions. The sustained-growth regime was supposed to have commenced around 1850 for
developed regions.
Unified Growth Theory (Galor, 2005, 2008, 2011, 2012) can be tested in many ways but the easiest
way to test it is to look for the dramatic takeoffs from stagnation to growth. These takeoffs are described
as a remarkable or stunning escape from the Malthusian trap (Galor, 2005, pp. 177, 220). It is a
signature, which cannot be missed.
This change in the pattern of growth is described as the sudden take-off from stagnation to growth
(Galor, 2005, pp. 177, 220, 277) or as a sudden spurt (Galor, 2005, 177, 220). According to Galor,
for developed regions, the end of the Malthusian regime of stagnation coincides with the Industrial
Revolution. The take-off of developed regions from the Malthusian Regime was associated with the
Industrial Revolution (Galor, 2005, p. 185). Indeed, the Industrial Revolution is considered to have
been the prime engine of economic growth (Galor, 2005, p. 212).
This signature is characterised by three features: (1) it should be a prominent change in the pattern
of growth, (2) it should be a transition from stagnation to growth and (3) it should occur at the time
predicted by the theory. For developed regions, the postulated takeoffs should occur around AD 1750,
or around the time of the Industrial Revolution, 1760-1840 (Floud & McCloskey, 1994). For less-
developed regions, they should occur around 1900. The added advantage of using this simple test is that
there are no significant gaps in the data around the time of the postulated takeoffs and consequently the
stagnation and the expected prominent transitions from stagnation to growth should be easily
identifiable.
A transition from growth to growth is not a signature of the postulated takeoff from stagnation to
growth. Thus, a transition is from hyperbolic growth to another hyperbolic growth or to some other
steadily-increasing trajectory is not a signature of the sudden takeoff from stagnation to growth.
Likewise, a transition at a distinctly different time is not a confirmation of the theoretical expectations.
2. World economic growth

Results of mathematical analysis of the world economic growth are presented in Figures 1-3.
Reciprocal values of historical data can be fitted using a straight line (representing hyperbolic growth)
between AD 1000 and 1955. From around 1955, the world economic growth started to be diverted to a
slower trajectory as indicated by the upward bending of the reciprocal values. This section is magnified
in Figure 2. Global economic growth is now approximately exponential (Nielsen, 2014, 2015a).
Hyperbolic fit to the world GDP data (Maddison, 2010) is shown in Figure 3. The fit is remarkably
good. The point at AD 1 is 77% away from the fitted curve. We would need more data between AD 1
and 1000 to decide whether such a difference is of any significance but it could reflect a pattern similar
to the pattern observed for the growth of human population (Nielsen, 2016). Hyperbolic economic
growth of the historical GDP has been uniquely identified by the straight-line fitting the reciprocal
values of data.
Parameters describing hyperbolic trajectory fitting the data between AD 1000 and 1955 are:
= a 1.684 102 and= k 8.539 106 . Its singularity is at t = 1972 . However, from around 1955, the
world economic growth started to be diverted to a slower trajectory bypassing the singularity by 17
years (see Table 1).
The search for a takeoff in the world economic growth produced negative results. The data reveal a
different pattern of growth than claimed by the Unified Growth Theory (Galor, 2005, 2008, 2011, 2012).
The theory claims a long period of stagnation followed by a sudden takeoff. The data show a stable
hyperbolic growth followed by a diversion to a slower trajectory.
The data also demonstrate that the Industrial Revolution had no impact on changing the economic
growth trajectory. These results might not be surprising because the world economic growth is
represented by the economic growth in developed and less-developed regions. However, even then, it
would be hard to expect that the data would follow such a remarkably stable and specific trajectory. We
would expect some distortions reflecting takeoffs around the time of the Industrial Revolution for
developed regions and takeoffs around 1900 for less-developed regions. We see no signs of such
distortions and no signs of the presence of such takeoffs.
47
__________________________________________________________________________________
The straight-line representing the reciprocal values of the GDP data shown in Figure 1 follows the
data closely until 1955. There was no boosting in the economic growth, no unusual acceleration at any
time between AD 1000 and 1955. The world economic growth was increasing monotonically before
and after the Industrial Revolution as shown by either a steadily increasing hyperbolic distribution in
Figure 3 or by the steadily-decreasing straight line (representing hyperbolic distribution) shown in
Figure 1. Which point on a straight line should be selected to mark a boundary between different
patterns of growth? How can we claim different patterns of growth on a straight line if the straight line
shows clearly only one pattern? There was no takeoff in the world economic growth at any time, let
alone around the time of the Industrial Revolution or around 1900.
Economic growth may have been slow over a long time but it was not stagnant. The growth was
hyperbolic, and the characteristic feature of hyperbolic growth is a slow growth over a long time and a
fast growth over a short time. Hyperbolic growth increases monotonically and it is impossible to locate
a place marking a transition from a slow to fast growth because such a transitions does not exist.
Hyperbolic growth of the world economy is in harmony with the hyperbolic growth of the world
population (Nielsen, 2016; von Foerster, Mora & Amiot, 1960). In both cases, the growth was indeed
slow over a long time and fast over a short time. In both cases the growth creates an illusion of stagnation
followed by a sudden takeoff. However, in both cases the growth was hyperbolic. There was no
stagnation and no sudden takeoff. Furthermore, in both cases the growth started to be diverted, relatively
recently, to slower trajectories.
3. Western Europe
The growth of the GDP in Western Europe is shown in Figures 4-6. Western Europe is represented
by the total of 30 countries: Austria, Belgium, Denmark, Finland, France, Germany, Italy, the
Netherlands, Norway, Sweden, Switzerland, the United Kingdom, Greece, Portugal, Spain and by 14
small, but unspecified countries. Ireland is missing in this list because it was included only from 1921.
The best hyperbolic fit to the data is between AD 1500 and 1900. Parameters for this distribution
=
are a 9.859 102 and= k 5.112 105 . The point of singularity is at t = 1929 . Between 1900 and
1910, economic growth started to be diverted to a slower, but still fast-increasing, trajectory bypassing
the singularity by 29 years (see Table 1).
The most complete set of data for Western Europe is for Denmark, France, the Netherlands and
Sweden. They are analysed separately and results are presented in Figures 7 and 8. According to
Maddison (2010), these four countries accounted for 34% of the total GDP of the 30 countries of
Western Europe in 2008.
Parameters describing the historical hyperbolic growth of the GDP in these four countries are:
= a 3.821 101 and = k 1.986 104 . The point of singularity is at t = 1923 . From around 1875
economic growth in Denmark, France, the Netherlands and Sweden was diverted to a slower trajectory,
bypassing the singularity by 48 years.
The quality of the hyperbolic fit to the data is virtually the same as for the total of the 30 countries
but now the fitted curve passes also through the AD 1 point. However, it still does not reproduce the
point at AD 1000. This point is only 41% below the fitted hyperbolic distribution.
The historical growth of the GDP in Western Europe was definitely hyperbolic from AD 1500 to
1900 but there is also a good indication that it might have been hyperbolic from AD 1 (see Figures 7
and 8). Even if we make allowance for this uncertainty, the search for a sudden takeoff around the
expected time, i.e. around the time of the Industrial Revolution, produced negative results for the 30
countries of Western Europe and for the four (Denmark, France, the Netherlands and Sweden)
characterised by the most complete sets of data.
The claim of a stunning or remarkable takeoff is contradicted by data. There was no takeoff of any
kind and at any time, stunning or less stunning, remarkable or less remarkable, sudden or gradual
none at all. The Industrial Revolution, the alleged prime engine of economic growth (Galor, 2005a,
p. 212), made no impression on changing the economic growth trajectory in regions where this engine
should have been working most efficiently. Industrial Revolution brought many other important
changes but, surprisingly perhaps, did not change the economic growth trajectory in the countries
closest to this monumental development.
48
__________________________________________________________________________________
4. Eastern Europe
Systematic data for Eastern Europe are available only for seven countries: Albania, Bulgaria,
Czechoslovakia, Hungry, Poland, Rumania and Yugoslavia. For other countries there are no data until
1990. The analysis of the historical data for Eastern Europe is summarised in Figures 9-11.
The best hyperbolic fit to the data is between AD 1000 and 1890. Hyperbolic parameters are:
= a 7.749 101 and
= k 4.048 104 . The point of singularity is at t = 1915 . From around 1890,
economic growth in Eastern Europe was diverted to a slower trajectory, bypassing the singularity by 25
years.
There was no stagnation and no takeoff at any time. Industrial Revolution had no impact on changing
the economic growth trajectory in the countries of Eastern Europe.
5. Former USSR
The analysis of the data for the countries of the former USSR is presented in Figures 12-14. The
hyperbolic fit to the data is between AD 1 and 1870. Parameters fitting the data are:= a 6.547 101
=
and k 3.452 104 . The point of singularity is at t = 1897 . From around 1870, or maybe even a little
earlier (shortly after the Industrial Revolution) economic growth in the Former USSR was diverted to
a slower trajectory, bypassing the singularity by at least 27 years.
There was no stagnation and no takeoff at any time. Industrial Revolution had no impact on changing
the economic growth trajectory in the countries of former USSR.
6. Asia
Analysis of the historical economic growth in Asia (including Japan) is summarised in Figures 15-
17. The best hyperbolic fit is between AD 1000 and 1950. Parameters fitting the data are:
= a 2.303 102 and= k 1.129 105 . The point of singularity is at t = 2040 .
Asia is made primarily of less-developed countries (BBC, 2014, Pereira, 2011) and consequently,
according to the Unified Growth Theory (Galor, 2005, 2008, 2011, 2012), economic growth in this
region should have been characterised by stagnation until around 1900, the year marking the alleged
stunning escape from the Malthusian trap, the escape, which was supposed to have been manifested by
the postulated dramatic takeoff. (Until AD 1900, Japans conrtibution to the total economy in Asia was
on average only 5%.) The data and their analysis show that there was no stagnation, at least from AD
1000 and no expected takeoff. The data reveal a steadily increasing hyperbolic growth until around
1950. From around that year economic growth was diverted to a faster trajectory. This boosting can be
seen clearly in Figures 16 and 17 and it occurred close to the time of the postulated takeoff from
stagnation to growth. However, it was not a transition from stagnation to growth but from hyperbolic
growth to a slightly faster trajectory of a different kind. It is, therefore, not the takeoff postulated in the
Unified Growth Theory. Furthermore, it was only a temporary boosting, which is now returning to the
original hyperbolic trajectory and, as indicated by the reciprocal values of the data, this new growth is
likely to be slower than the original trajectory. Thus, it is a boosting of a completely different kind. It
would be interesting to explain it but we cannot be helped by Unified Growth Theory because it
discusses mechanisms, which are repeatedly contradicted by data. This transition is not even recognised
in this theory
Reciprocal values of data presented in Figure 16 show that the economic growth became temporarily
slower at the time overlapping the time of the Industrial Revolution, 1760-1840 (Floud & McCloskey,
1994), because while the point in 1820 is still located on the straight line, representing hyperbolic
growth, the point in 1870 is above this line. The deceleration in the economic growth occurred sometime
between 1820 and 1870.
This brief deceleration was followed by a transient growth between 1870 and 1940, which appears
to have been also hyperbolic but a little faster than the earlier hyperbolic growth. This transition
occurred earlier than the postulated takeoff around 1900 and it was not a transition from stagnation to
growth but a transition from hyperbolic growth to hyperbolic growth. Furthermore, it was also a minor
transition, which could be hardly noticed in the direct display of data shown in Figure 17. In summary,
therefore, the examination of data for the economic growth in Asia demonstrates that the postulated
49
__________________________________________________________________________________
takeoff (Galor, 2005, 2008, 2011, 2012) never happened. There was no stagnation and no sudden
dramatic escape to a new and rapid growth.
7. Africa
Results of the analysis of the economic growth in the 57 African countries are presented in Figures
18-20. Reciprocal values of the GDP data, presented in Figures 18 and 19, show clearly that the
economic growth was following two hyperbolic distributions. At first it was a slow hyperbolic growth
between AD 1 and 1820 characterised by parameters = a 1.244 101 and
= k 5.030 105 and by the
singularity at t = 2473 . Then, around 1820, this slow hyperbolic growth was replaced by a significantly
faster hyperbolic growth characterised by parameters = a 4.192 101 and
= k 2.126 104 and by the
singularity at t = 1972 . Defined by the parameter k, this new growth was 4.2 times faster than the earlier
hyperbolic growth. From around 1950, this fast hyperbolic growth was diverted to a slower, non-
hyperbolic trajectory, bypassing singularity by 22 years.
Africa is also made of less-developed countries (BBC, 2014; Pereira, 2011) so according to the
Unified Growth Theory (Galor, 2005, 2008, 2011, 2012) it should have experienced stagnation in the
economic growth until around 1900 followed by a clear takeoff around that year. These expectations
are contradicted by the economic growth data because (1) economic growth was not stagnant but
hyperbolic until 1950, (2) there was no takeoff from stagnation to growth around 1900 or around any
other time and (3) shortly after the expected time of the takeoff, economic growth in Africa started to
be diverted to a slower trajectory.
Acceleration in the economic growth in Africa occurred around 1820, but it was not a transition
from stagnation to growth but from growth to growth. Even more specifically, it was a transition from
the hyperbolic growth to another hyperbolic growth. It was also acceleration at a wrong time, not around
1900 but around the time of the Industrial Revolution. This acceleration can be explained by noticing
that it appears to coincide with the intensified colonisation of Africa (Duignan & Gunn, 1973; McKay,
et al. 2012; Pakenham, 1992). The fast-increasing GDP after 1820 was not reflecting the rapidly
improving living conditions of African population brought about by the beneficial changes caused by
the Industrial Revolution but the rapidly increasing wealth of new settlers and their countries of origin
at the expense of the deploring living conditions of native populations.
The search for the takeoff from stagnation to growth, claimed by the Unified Growth Theory (Galor,
2005, 2008, 2011, 2012), produced negative results. The data show also that there was no stagnation in
the economic growth over the entire range of time, from AD 1 to the present time.
8. Latin America
Results of the analysis of the economic growth in Latin America are presented in Figures 21 - 23.
Data for Latin America are difficult to analyse because there was a significant decline in the economic
growth between AD 1500 and 1600 but they also appear to follow two distinctly different hyperbolic
trajectories. However, the identification of the first trajectory is not as clear as for Africa. The
identification of the second hyperbolic trajectory is more convincing. Our tentative conclusion is that
the economic growth in Latin America was following a slow hyperbolic distribution between AD 1 and
1500 and a fast hyperbolic distribution between AD 1600 and around 1870.
The tentatively assigned slow hyperbolic growth between AD 1 and 1500 is characterised by
parameters= a 4.421 101 and = k 2.093 104 . Its singularity is at t = 2113 . The better determined
fast hyperbolic growth between AD 1600 and 1870 is characterised by parameters = a 1.570 100 and
= k 8.224 104 . Its singularity is at t = 1910 . Defined by the parameter k, this growth was 3.9 times
faster than the earlier hyperbolic growth. From around 1870, this fast hyperbolic growth started to be
diverted to a slower trajectory bypassing the singularity by 40 years. The transition from the earlier
apparent hyperbolic growth to a new and rapid hyperbolic growth, which occurred between around AD
1500 and 1600 appears to coincide with commencement of the Spanish conquest (Teeple, 2002).
Latin America is also made of less-developed countries (BBC, 2014; Pereira, 2011) so again,
according to the Unified Growth Theory (Galor, 2005, 2008, 2011, 2012), the economic growth in this
regions should have been stagnant until around 1900 and fast-increasing from around that year. This
pattern of growth is not confirmed by data. The data show a diametrically different pattern: (1) there is
50
__________________________________________________________________________________
no convincing evidence of the existence of stagnation over the entire range of time between AD 1 and
1870 but there is a sufficiently convincing indication of the hyperbolic growth particularly between AD
1600 and 1870, (2) there was no takeoff from stagnation to growth at any time, and (3) around the time
of the postulated takeoff in 1900 there was a diversion to a slower trajectory in 1870.
Even if the identification of the hyperbolic growth between AD 1 and 1500 is questioned, the overall
pattern of growth in Latin America is similar to the pattern in Africa: a slow hyperbolic growth is
followed by a fast hyperbolic growth. However, in any case, there is no convincing evidence that the
growth was ever stagnant. On the contrary, there is sufficiently convincing evidence that the growth
was never stagnant. It was clearly not stagnant between AD 1600 and 1870.
There was also no takeoff, dramatic or modest, from stagnation to growth around the expected time
of 1900, first because the growth before that year was not stagnant but hyperbolic and second because
around the time of the expected remarkable takeoff the economic growth started to be diverted to a
slower trajectory. The search for the postulated takeoff produced negative results.

Results of mathematical analysis of the historical economic growth are presented in Table 1. The
listed parameters a and k are for the fitted hyperbolic distributions. The last column shows the results
of the search for the takeoffs from stagnation to growth claimed by the Unified Growth Theory (Galor,
2005, 2008, 2011, 2012).
Table 1. Summary of the mathematical analysis or the historical economic growth

Region/Countries a k Hyperbolic Singularity Proximity Takeoff
Range
World 1.684 102 8.539 106 1000 1955 1972 17 X
Western Europe 9.859 102 5.112 105 1500 1900 1929 29 X
Western Europe (4) 3.821 101 1.986 104 1 1875 1923 48 X
Eastern Europe 7.749 10 1
4.048 104 1000 1890 1915 25 X
Former USSR 6.547 101 3.452 104 1 1870 1897 27 X
Asia 2.303 102 1.129 105 1000 1950 2040 90 X
Africa 1.244 101 5.030 105 1 1820 2473
4.192 10 1
2.126 104 1820 1950 1972 22 X
Latin America 4.421 10 1
2.093 10 4 1 1500 2113
1.570 100 8.224 104 1600 1870 1910 40 X
Notes: a and k Hyperbolic growth parameters [see eqn (1)]. Hyperbolic Range - The empirically-confirmed range of time
when the economic growth can be described using hyperbolic distributions. Singularity - The time of the escape to infinity for
a given hyperbolic distribution. Proximity - Proximity (in years) of the singularity at the time when the economic growth
departed from the hyperbolic growth to a new trajectory. Western Europe (4) - Four countries of Western Europe: Denmark,
France, the Netherlands and Sweden. X - No takeoff. The takeoff from stagnation to growth claimed by the Unified Growth
Theory (Galor, 2005, 2008, 2011, 2012) never happened.
This analysis demonstrates that the natural tendency for the historical economic growth was to
increase hyperbolically. In general, there is a remarkably good agreement between the data and the
calculated hyperbolic distributions.
Unlike the more familiar exponential distributions, which are easier to understand because they show
more readily a gradually increasing growth, hyperbolic distributions appear to be made of two or maybe
even three components: a slow component, a fast component and perhaps even a transition component
located between the apparent slow and fast components. This illusion is so strong that even the most
experienced researchers can be deceived particularly if they have no access to good sets of data, which
was in the past. Now, however, excellent data are available (Maddison, 2001, 2010) and we can use
them not only to check the earlier interpretations of economic growth but also to expand the scope of
the economic research.
The postulate of the existence of the epoch of Malthusian stagnation is suggested by a slow economic
growth over a long time but this slow growth is just a part of the hyperbolic growth, which is
convincingly identified using reciprocal values. Hyperbolic distributions create also the illusion of a
sudden takeoff but this feature is also a part of the hyperbolic growth. Hyperbolic growth is slow over
a long time and fast over a short time but the slow and fast growth are the integral features of the same
monotonically increasing distribution, which is easier to understand by using the reciprocal values of
51
__________________________________________________________________________________
the growing entity (Nielsen, 2014). In such displays, the illusion of distinctly different components
disappears because hyperbolic growth is then represented by a decreasing straight line, which is easy to
understand. It then becomes obvious that hyperbolic distribution cannot be divided into distinctly
different sections governed by different mechanism because it makes no sense to divide a straight line
into arbitrarily chosen sections and claim different mechanism to such arbitrarily-selected sections. It
is also then clear that it is impossible to pinpoint the transition from a slow to a fast growth. Which point
on a straight line should we select to identify such a transition? The transition does not happen at any
specific time but gradually over the whole range of time.
Our search for the postulated takeoffs from stagnation to growth (Galor, 2005, 2008, 2011, 2012)
produced negative results: there were no takeoffs. Galors elaborate discussion revolving around his
postulated three regimes of growth and the postulated takeoffs from stagnation to growth are irrelevant
because there were no takeoffs in the growth of the GDP and in the growth of income per capita
(GDP/cap) (Nielsen, 2015b). In science, just one contradicting evidence in data is sufficient to show
that a theory advocating the contradicted postulate or postulates has to be either rejected or revised to
bring it in the agreement with empirical evidence. In the case of the Unified Growth Theory (Galor,
2005, 2008, 2011, 2012), the postulated takeoffs from stagnation to growth are contradicted repeatedly
by the economic growth in Western Europe, Eastern Europe, former USSR, Asia, Africa, and Latin
America as well as by the world economic growth.
The data and their analysis suggest new lines of research of economic growth. They suggest that our
attention should not be directed towards explaining the mechanism of stagnation and of the sudden
takeoffs from stagnation to growth because these features are contradicted by data. What needs to be
explained is why the historical economic growth was hyperbolic and why relatively recently it was
diverted to a slower trajectory. Maddison published excellent data describing not only economic growth
but also the growth of human population and these data can be used effectively in trying to explain the
historical economic growth.
References
BBC, (2014). The North South Divide.
http://www.bbc.co.uk/bitesize/standard/geography/international_issues/contrasts_development/revision/2/
Floud, D., & McCloskey, D. N. (1994). The Economic History of Britain since 1700. Cambridge: Cambridge University
Press.
Economic Growth, (pp. 171-293). Amsterdam: Elsevier.
Galor, O. (2008). Comparative Economic Development: Insight from Unified Growth Theory.
Galor, O. (2012). Unified Growth Theory and Comparative Economic Development.
Books.
Nielsen, R. W. (2015a). The Insecure Future of the World Economic Growth. Journal of Economic and Social Thought,
2(4), 242-255.
Nielsen, R. W. (2015b). Unified Growth Theory Contradicted by the GDP/cap Data.
Nielsen, R. W. (2016). Growth of the world population in the past 12,000 years and its link to the economic growth. Journal
of Economics Bibliography, 3(1), 1-12
Avon Books.
Pereira, E. (2011). Developing Countries Will Lead Global Growth in 2011, Says World Bank.
http://www.forbes.com/sites/evapereira/2011/01/12/developing-countries-will-lead-global-growth-in-2011-says-world-
bank/
Teeple, J. B. (2002). Timelines of World History. London: Dorling Kindersley Publishing Inc.
von Foerster, H., Mora, P., & Amiot, L. (1960). Doomsday: Friday, 13 November, A.D. 2026. Science, 132,1291-1295.
52
__________________________________________________________________________________
53
__________________________________________________________________________________
Appendix
World Economic Growth
Figure 1. Reciprocal values of the GDP data (Maddison, 2010) are fitted using straight line between AD 1000
and 1955 representing hyperbolic growth. There was no stagnation and no takeoff from stagnation to growth,
claimed by the Unified Growth Theory (Galor, 2005, 2008, 2011, 2012). Industrial Revolution had no impact on
changing the economic growth trajectory. From around 1955, the economic growth started to be diverted to a
slower trajectory.
Figure 2. Reciprocal values of the GDP data (Maddison, 2010) showing the diversion of the economic growth to
a slower trajectory from around 1955, as indicated by the upward bending. The current global economic growth
is approximately exponential (Nielsen, 2014, 2015a).
54
__________________________________________________________________________________
Figure 3. World GDP data (Maddison, 2010) fitted using hyperbolic distribution. The point at AD 1 is 77%
higher than the calculated distribution. There was no stagnation and no takeoff from stagnation to growth. Both
features were incorrectly claimed by the Unified Growth Theory (Galor, 2005, 2008, 2011, 2012). Industrial
Revolution had no impact on changing the economic growth trajectory. From around 1955, the world economic
growth started to be diverted to a slower but still fast-increasing trajectory, which is now approximately
exponential (Nielsen, 2014, 2015a).
Western Europe
The total of 30 countries
Figure 4. Reciprocal values of the GDP data (Maddison, 2010) for Western Europe are compared with the
hyperbolic distribution represented by the decreasing straight line. The growth was hyperbolic from at least AD
1500 to 1900. There was no takeoff from stagnation to growth. Industrial Revolution had no impact on changing
the economic growth trajectory in Western Europe, the centre of this revolution. On the contrary, from around
1900, shortly after the Industrial Revolution, the economic growth in Western Europe started to be diverted to a
slower trajectory as indicated by the upward bending of the trajectory representing the reciprocal values of
data.
55
__________________________________________________________________________________
Figure 5. Reciprocal values of the GDP data (Maddison, 2010) for Western Europe between AD 1500 and 2008
showing a diversion to a slower trajectory from around 1900. There was no takeoff from stagnation to growth,
claimed incorrectly by the Unified Growth Theory (Galor, 2005, 2008, 2011, 2012). Industrial Revolution had
absolutely no impact on changing the economic growth trajectory in Western Europe, the centre of this
revolution.
Figure 6. Economic growth in Western Europe. The GDP data (Maddison, 2010) are compared with hyperbolic
distribution. The growth was hyperbolic from at least AD 1500 to around 1900. The point at AD 1 is 42%
higher than for the calculated distribution and 48% lower at AD 1000. There was no takeoff from stagnation to
growth, claimed incorrectly by the Unified Growth Theory (Galor, 2005, 2008, 2011, 2012). Industrial
Revolution had no impact on changing the economic growth trajectory in Western Europe, the centre of this
revolution. From around 1900, economic growth in Western Europe started to be diverted to a slower
trajectory.
56
__________________________________________________________________________________
Denmark, France, Netherlands and Sweden
Figure 7. Reciprocal values of the GDP data (Maddison, 2010) describing economic growth in four countries of
Western Europe (Denmark, France, the Netherlands and Sweden) compared with the straight line representing
hyperbolic growth fitting the data between AD 1 and 1875. From around 1875, or shortly after the Industrial
Revolution, economic growth in these four countries started to be diverted to a slower trajectory. Industrial
Revolution did not boost economic growth. There was no takeoff from stagnation to growth.
Figure 8. Economic growth in Denmark, France, the Netherlands and Sweden. The data (Maddison, 2010) are
compared with hyperbolic distribution. The point at AD 1000 is 41% lower than for the calculated distribution.
From around 1875, the economic growth started to be diverted to a slower trajectory. There was no takeoff
from stagnation to growth.
57
__________________________________________________________________________________
Eastern Europe
Figure 9. Reciprocal values of the GDP data (Maddison, 2010) for Eastern Europe are compared with the
hyperbolic distribution represented by the decreasing straight line. Economic growth was hyperbolic from at
least AD 1000. The takeoff from stagnation to growth never happened because there was no stagnation.
Industrial Revolution did not boost the economic growth in Eastern Europe.
Figure 10. Reciprocal values of the GDP data (Maddison, 2010) for Eastern Europe showing that from around
1890, shortly after the Industrial Revolution, the economic growth started to be diverted to a slower trajectory.
There was no takeoff from stagnation to growth because there was no stagnation. Industrial Revolution did not
boost the economic growth in Eastern Europe. Hyperbolic growth around that time remained undisturbed.
58
__________________________________________________________________________________
Figure 11. Economic growth in Eastern Europe. GDP data (Maddison, 2010) are compared with the best
hyperbolic fit. The point at AD 1 is 51% higher than for the calculated distribution. From around 1890, shortly
after the Industrial Revolution, economic growth started to be diverted to a slower trajectory. Industrial
Revolution did not boost the economic growth in Eastern Europe. Contrary to the Unified Growth Theory
(Galor, 2005, 2008, 2011, 2012), there was no stagnation and no takeoff from stagnation to growth.
Former USSR
Figure 12. Reciprocal values of the GDP data (Maddison, 2010) for the former USSR compared with the
hyperbolic distribution represented by the decreasing straight line. Data indicate that the economic growth was
hyperbolic from AD 1 to 1870. Industrial Revolution did not boost the economic growth. There was no
stagnation and no takeoff from stagnation to growth. Shortly after the Industrial Revolution, the economic
growth in Eastern Europe started to be diverted to a slower trajectory. Unified Growth Theory (Galor, 2005,
2008, 2011, 2012) is contradicted by the economic growth data.
59
__________________________________________________________________________________
Figure 13. Reciprocal values of the GDP data (Maddison, 2010) for the former USSR showing that from around
1870, shortly after the Industrial Revolution, economic growth started to be diverted to a slower trajectory.
Figure 14. Economic growth in the former USSR. GDP data (Maddison, 2010) are compared with the
hyperbolic fit. The growth was hyperbolic from AD 1 to 1870. From around 1870, shortly after the Industrial
Revolution, economic growth started to be diverted to a slower trajectory. Epoch of stagnation did not exist in
the economic growth. Industrial Revolution did not boost the economic growth. There was no takeoff from
stagnation to growth because there was no stagnation but a steadily-increasing growth. Unified Growth Theory
(Galor, 2005, 2008, 2011, 2012) is contradicted by the economic growth data.
Asia (including Japan)
Figure 15. Reciprocal values of the GDP data (Maddison, 2010) for Asia (including Japan) compared with the
hyperbolic distribution represented by the decreasing straight line. Economic growth was hyperbolic from at
least AD 1000. There was no expected transition from stagnation to growth.
60
__________________________________________________________________________________
Figure 16. Reciprocal values of the GDP data (Maddison, 2010) for Asia (including Japan). The data show a
minor deceleration of growth towards the end of the time of the Industrial Revolution followed by a slightly
faster hyperbolic growth between 1870 and 1940. The expected takeoff from stagnation to growth around 1900
(Galor, 2005, 2008, 2011, 2012) did not happen. The data show a small boosting around 1950 but it was not a
transition from stagnation to growth. The search for the postulated takeoff (Galor, 2005, 2008, 2011, 2012)
produced negative results.
Figure 17. Economic growth in Asia (including Japan). The data (Maddison, 2010) are compared with the
hyperbolic distribution. The point at AD 1 is 76% higher than the calculated value. The data show a minor
boosting around 1950 but it was not a transition from stagnation to growth but from the hyperbolic growth to a
slightly faster trajectory, which is now coming closer to the earlier hyperbolic trajectory. The boosting was not
only small but also it did not last long. The search for the postulated takeoff from stagnation to growth (Galor,
2005, 2008, 2011, 2012) produced negative results.
61
__________________________________________________________________________________
Africa
Figure 18. Reciprocal values of the GDP data (Maddison, 2010) for Africa compared with hyperbolic
distributions represented by the decreasing straight lines. There was no stagnation in the economic growth.
Economic growth was increasing hyperbolically between AD 1 and around 1820 and again from 1820 to around
1950. The expected takeoff from stagnation to growth (Galor, 2005, 2008, 2011, 2012) never happened. The
acceleration around 1820 was not a transition from stagnation to growth but from growth to growth. It also
occurred earlier than expected (in 1820 rather than around 1900). Furthermore, close to the postulated takeoff
in 1900, economic growth started to be diverted to a slower trajectory. The search for the takeoff from
stagnation to growth around 1900 produced negative results. Unified Growth Theory (Galor, 2005, 2008, 2011,
2012) is contradicted by data.
Figure 19. Reciprocal values of the GDP data (Maddison, 2010) for Africa showing that from around 1950
economic growth started to be diverted to a slower trajectory. There was no takeoff around 1900, not even from
growth to growth.
62
__________________________________________________________________________________
Figure 20. Economic growth in Africa. Data (Maddison, 2010) are compared with hyperbolic distributions. The
claimed takeoff from stagnation to growth (Galor, 2005, 2008, 2011, 2012) never happened because there was
no stagnation. Furthermore, the transition from hyperbolic growth to hyperbolic growth occurred earlier
(around 1820) than the postulated takeoff from stagnation to growth (around 1900). From around 1950, close to
the claimed but non-existing takeoff from stagnation to growth, economic growth started to be diverted to a
slower trajectory. Unified Growth Theory (Galor, 2005, 2008, 2011, 2012) is contradicted by data.
Latin America
Figure 21. Reciprocal values of the GDP data (Maddison, 2010) for Latin America are compared with
hyperbolic distributions represented by the decreasing straight lines. The pattern of growth in Latin America is
similar to the pattern of growth in Africa. The expected takeoff from stagnation to growth around 1900 (Galor,
2005, 2008, 2011, 2012) did not happen, because there was no stagnation and because, from around 1870,
economic growth started to be diverted to a slower trajectory.
63
__________________________________________________________________________________
Figure 22. Reciprocal values of the GDP data (Maddison, 2010) for Latin America showing that from around
1870, i.e. close to the time of the expected takeoff (around 1900) from stagnation to growth (Galor, 2005, 2008,
2011, 2012) economic growth started to be diverted to a slower trajectory. The data show also that the takeoff
from stagnation to growth could not have happened because there was no stagnation.
Figure 23. Economic growth in Latin America. Economic growth data (Maddison, 2010) are compared with
hyperbolic distributions. Unified Growth Theory (Galor, 2005, 2008, 2011, 2012) is contradicted by data.
Economic growth was not stagnant before the postulated takeoff from stagnation to growth (around 1900) but
hyperbolic. The growth was also stable and hyperbolic around the time of the Industrial Revolution in the
Western world. The transition from stagnation to growth could not have happened because there was no
stagnation. Furthermore, from around 1870, i.e. from around the time of the postulated takeoff, economic
growth started to be diverted to a slower trajectory. The search for the takeoff from stagnation to growth
produced negative results.
64
__________________________________________________________________________________
4
Growth of the World Population in the Past 12,000 Years and Its
Link to the Economic Growth
By Ron W. NIELSENa
Abstract. Data describing the growth of the world population in the past 12,000 years are analysed. It is shown
that population, if unchecked, does not increase exponentially, as expected by Malthus, but hyperbolically. This
analysis reveals three episodes of hyperbolic growth: 10,000 - 500 BC, AD 500 - 1200 and AD 1400 -1950,
representing the total of about 89% of the past 12,000 years. The remaining 11% were taken by three demographic
transitions: 500 BC - AD 500, AD 1200 - 1400 and AD 1950 - present. The first two transitions were between
sustained hyperbolic trajectories. The current transition is from a sustained hyperbolic growth to a yet unknown
trajectory. The often claimed but never proven existence of Malthusian stagnation in the growth of human
population is convincingly contradicted by the mathematical analysis of data. There was also never any form of a
dramatic transition from stagnation to growth, described often as a takeoff, because there was no stagnation in the
growth of the world population. The perceived takeoff is just the natural continuation of hyperbolic growth.
Correct understanding of the historical growth of human population is essential in the correct interpretation of the
historical growth of income per capita.
Keywords.Growth of human population, Economic growth, Growth of income per capita, Stagnation, Takeoffs,
Hyperbolic growth, Demographic transitions.
JEL.A12, B22, B25, F01, N00, Y80.
1. Introduction
T he study of historical economic growth involves not only the study of the Gross Domestic
Product (GDP) but also the study of the growth of population, because as pointed out by Galor
(2005, 2011), it is important to understand the relationship between the these two process and
particularly to understand income per capita (GDP/cap) distributions. It is for this reason that the latest
and the most extensive compilations of historical GDP data, published by the world-renown economist,
include also data describing the historical growth of human population (Maddison, 2001, 2010).
Maddison lists not only the GDP data but also the calculated GDP/cap values.
About 50 years ago, von Foerster, Mora and Amiot (1960) demonstrated that human population was
increasing hyperbolically during the AD era. We now have far better and more extensive sets of data
compiled not only by Maddison (2001, 2010) but also by Manning (2008) and by the US Census Bureau
(2016). The last two compilations are based on virtually the same primary sources but they are
complimentary.
Maddisons compilation is useful in studying the growth of the population not only global but also
regional and national. However, his data are terminated in AD 1. Furthermore, they also contain
significant gaps below AD 1500. The data compiled by Manning and by the US Census Bureau are
significantly richer but they are limited only to the description of the world population. However, they
extend down to 10,000 BC.
Preliminary examination of Maddisons data indicates that economic growth and the growth of
human population followed similar trajectories. Consequently, by using the rich set of data extending
down to 10,000 BC we might gain better insight not only into the historical growth of human population
but also of the economic growth.
a
Australia.
. +61407201175
Published as: Nielsen, R. W. (2016). Growth of the world population in the past 12,000 years and its link to the economic
growth. Journal of Economics Bibliography, 3(1), 1-12.
65
__________________________________________________________________________________
2. The data
Procedures adopted in estimating historical populations are described by Durand (1977). The data
for the AD era are of exceptionally good quality. Between AD 400 and 1850, independent estimates are
within 10% of their corresponding averaged values. The estimates after 1850 are within 1.5%. The
largest deviations of around 30% are for the AD 1 data. The two estimates for AD 200 differ by 15%
from their average value. The BC data are less accurate and less consistent but when closely analysed
they are also found to follow a well-defined trajectory.
3. Analysis of population data

In order to understand hyperbolic distributions, it is useful to compare them with the more familiar
exponential distributions. The differential equation describing exponential growth is given by the
following simple equation:
1 dS (t )
=k, (1)
S (t ) dt
where S (t ) is the size of a growing entity, in our case the size of population, and k is an arbitrary constant.
The left-hand side of this equation represents the growth rate. For k > 0 the eqn (1) describes growth,
while for k < 0 it describes decay.
The solution of the eqn (1) is
S (t ) = aekt , (2)
where a is a constant related to the constant of integration.

The eqn (2) gives
ln S =
(t ) ln a + kt . (3)
The logarithm of the size of the growing entity increases linearly with time. Exponential growth can
be easily identified by plotting data using semilogarithmic scales of reference because in such
presentation exponential growth is represented by an increasing straight line.
Data for the growth of the population during the BC and AD eras (Manning, 2008; US Census
Bureau, 2015) are shown in Figure 1. They are compared with the best fit obtained by using exponential
function. The world population was not increasing exponentially.
Let us now examine hyperbolic growth. This type of growth is described by the following differential
equation:
1 dS (t )
= kS (t ) , (4)
S (t ) dt
where k > 0 .
It is a slight modification of the eqn (1). Here, the growth rate is not constant but directly proportional
to the size of the growing entity. The solution of this equation, which can be found by substitution
S (t ) = Z 1 (t ) , is given by the following simple formula:
1
S (t ) = . (5)
a kt
It is just a reciprocal of a linearly-decreasing function. Consequently,
66
__________________________________________________________________________________
1
= a kt (6)
S (t )
Figure 1. Data describing the growth of the world population (Manning, 2008; US Census Bureau, 2016) are
compared with the best fit obtained by using exponential function. The world population was not increasing
exponentially. The BC time scale is identified by the negative numbers.
Reciprocal values of the size of a growing entity follow a decreasing straight line. This representation
simplifies the analysis of hyperbolic distributions. We can use this dependence to identify uniquely
hyperbolic growth, in much the same way as the linearly increasing logarithm of a growing entity can
be used to identify exponential growth.
It is useful to understand the difference between the exponential growth and the hyperbolic growth.
For the exponential growth, the growth rate is constant. It does not matter how large is the size of the
growing entity, the growth rate never changes. For this reason, exponential growth can be characterised
and identified by using the growth rate or equivalently by using the doubling time. This approach is
inapplicable to the hyperbolic growth or to any other type of growth. It is incorrect to use the doubling
time to characterise any other type of growth. In particular, it is incorrect to use the so-called rule of
70 for any other type of growth because in all other cases the growth rate and the doubling time are
not constant. In order to characterise any other types of growth by the growth rate or by the doubling
time we cannot just present a single value for any of these two quantities at a certain time but we have
to show how their growth rate or the doubling time depends on time or on the size of the growing entity.
For instance, if we look at the eqn (4) we can see that, for the hyperbolic growth, the growth rate is
directly proportional to the size of the growing entity. This is a useful characteristic feature of hyperbolic
growth. Another characteristic feature of hyperbolic growth is that the growth rate per size of the
growing entity is constant.
As discussed elsewhere (Nielsen, 2014), hyperbolic distributions are confusing because they appear
to be made of two distinctly-different components, slow and fast, leading to countless misconceptions
and misinterpretations of the distributions describing the growth of human population or the economic
growth. However, the analysis of these distributions and their interpretation becomes trivially simple if
reciprocal values of data are used, as shown in Figure 2, because according to the eqns 5 and 6, if the
reciprocal values of data follow a decreasing straight line, then the growth is hyperbolic. We can then
fit the reciprocal values to find the mathematical expression for the hyperbolic growth given by the eqn
(5).
67
__________________________________________________________________________________
Figure 2. Reciprocal values of the world population data (Manning, 2008; US Census Bureau, 2015) reveal two
distinctly different hyperbolic trajectories (represented by the decreasing straight lines). They also show a
dramatic demographic transition between around 500 BC and AD 500. Furthermore, they show that there was
no takeoff around the time of the Industrial Revolution. In fact, there was no transition from stagnation to
growth at any time. The size of the population is in billions.
Furthermore, if reciprocal values of data follow a decreasing straight line, the growth is not stagnant.
However, the concept of stagnation is not supported even if the reciprocal values of data do not decrease
linearly. Any monotonically-decreasing trajectory will show that the postulate of stagnation followed
by a takeoff at a certain time is not supported by data. To prove the existence of the epoch of stagnation
it is necessary to prove the presence of random fluctuations often described as Malthusian oscillations.
Such random fluctuations should be clearly seen not only in the direct display of data but also in the
display of their reciprocal values. It they are absent then there is no support in data for claiming the
existence of the epoch of stagnation. However, if the reciprocal values of data follow a decreasing
straight line, then they show, or at least strongly suggest, that the growth was hyperbolic. Positive
identification of any type of growth depends on the range of available data.
It should be also remembered that for the reciprocal values effects are reversed. A diversion to a
slower trajectory will be indicated by an upward bending away from the earlier trajectory, while
diversion to a faster trajectory will be indicated by a downward bending.
Descriptions of economic growth involve frequent discussions of the so-called takeoffs (Galor,
2005, 2011) representing the assumed sudden and prominent change in the growth trajectory, a
transition from the alleged stagnation to growth. For the economic growth or for the growth of human
population represented by their reciprocal values, such sudden takeoffs should be indicated by a clear
and strong downward bending of the growth trajectory.
If the straight line representing the reciprocal values of data remains unchanged, then obviously,
there is no change in the mechanism of growth. It makes no sense to divide a straight line into two or
three arbitrarily selected sections and claim different regimes of growth controlled by different
mechanisms for these arbitrarily-selected sections.
68
__________________________________________________________________________________
The analysis of data presented in Figure 2 reveals two distinctly different hyperbolic trajectories for
the BC and AD eras. They are represented by two distinctly different straight lines fitting the reciprocal
values of population data. In this representation, the growth during the AD era is dwarfed by the growth
during the BC era but this part can be better examined by looking at the lower section of Figure 2.
The corresponding hyperbolic distributions are shown in Figure 3. Figures 1 and 2 make it clear that
the growth of human population was not exponential, as claimed by Malthus (1798). Contrary to his
expectations, data and their analysis show that population, if unchecked, increases hyperbolically. It
shows that the growth of human population was increasing hyperbolically not only during the AD era,
as observed by von Foerster, Mora and Amiot (1960), but also during the BC era. This analysis shows
also that Industrial Revolution, 1760-1840 (Floud & McCloskey, 1994) did not boost the growth of
human population, the result being in agreement with the analysis of the historical economic growth
(Nielsen, 2016).
Figure 3. Population, if unchecked, increases hyperbolically. This overall view shows that there was only one
major demographic transition (between around 500 BC and AD 500) from a fast to a significantly slower
hyperbolic trajectory. Industrial Revolution had no impact on the growth of human population. The perceived
population explosion is just the natural continuation of hyperbolic growth.
Results presented in Figures 2 and 3 show that from 10,000 BC to around 500 BC the growth of
human population was hyperbolic. This hyperbolic growth was followed by a demographic transition
between 500 BC and AD 500 from a fast BC hyperbolic trajectory to a significantly slower AD
hyperbolic trajectory. It was not a transition from stagnation to growth because there was no stagnation
in the growth of human population (Nielsen, 2013a).
Hyperbolic parameters fitting world population data are: a = 2.282 and = k 2.210 102 for the BC
trajectory between 10,000 BC and 500 BC, and a = 7.061 and= k 3.398 103 for the AD trajectory
between AD 500 and 2015. Characterised by the parameter k, the BC hyperbolic growth was 6.5 times
faster than the AD growth.
Using the data (Manning, 2008; US Census Bureau, 2016), the fitted hyperbolic distributions (shown
in Figure 3) and the eqn (4) we can now estimate the growth rates during the BC and AD eras. During
the BC era, the growth rate was increasing hyperbolically (monotonically) with time or linearly (and
again monotonically) with the size of the population from around 1.010 104 (0.010%) per year in
69
__________________________________________________________________________________
10,000 BC to around 2.520 103 (0.252%) per year in 500 BC. The growth was slow but not stagnant.
During the AD era, the growth was again approximately hyperbolic between AD 500 and 1950, and the
growth rate increased approximately monotonically from 6.337 104 (0.063%, smaller than in 500 BC)
to 7.805 103 (0.781%) in 1950.
There was no stagnation but hyperbolic growth. The transition between 500 BC and AD 500 was
not a transition from stagnation to growth but from growth to growth. It was not a dramatic takeoff but
a transition to a slower hyperbolic trajectory. These features are important in relating the growth of
population to the economic growth because contrary to the repeated claims presented in the Unified
Growth Theory (Galor, 2005, 2011) there was no stagnation and no dramatic takeoff in the growth of
the GDP (Nielsen, 2016) or in the growth of the GDP/cap (Nielsen, 2015a).
4. Detailed analysis of the AD data

Data for the AD era are of exceptionally good quality and they allow for a closer and minute
examination of the pattern of growth. Even though the hyperbolic trajectory shown in Figures 2 and 3
fits the AD data well, the display of the reciprocal values presented in the lower part of Figure 2 shows
that starting from around AD 1400, some data are systematically above the fitted straight line,
suggesting a shift in the hyperbolic growth around that time.
Reciprocal values of data shown in Figure 4 reveal a clear delay in the growth of the population
between around AD 1200 and 1400 followed by a new and slightly faster hyperbolic trajectory.
Hyperbolic trajectory between AD 500 and 1200 is given by a = 6.940 and = k 3.448 103 , and from AD
1400 by a = 9.123 and = k 4.478 103 . For these new and improved fits to the data, the growth rate was
3
6.610 104 (0.066%) in AD 500, 1.230 103 (0.123%) in AD 1200, 1.568 10 (0.157%) in AD 1400
and 1.142 102 (1.142%) in 1950. The growth was hyperbolic (monotonic) between AD 500 and 1200
and again between AD 1400 and 1950. There was no stagnation and no dramatic takeoff from stagnation
to growth at any time.
Figure 4. Reciprocal values of data for the AD era show a clear but small disturbance in the growth of the
population between AD 1200 and 1400. This disturbance caused a shift to a slightly faster hyperbolic trajectory.
The size of the population in billions.
Transition between AD 1200 and 1400 coincides with the unusual convergence of five strong lethal
events. This minor, but noticeable delay in the growth of population appears to have been caused by a
combined impact of five significant demographic catastrophes (Nielsen, 2013b): Mongolian Conquest
(1260-1295) with the total estimated death toll of 40 million; Great European Famine (1315-1318), 7.5
million; the 15-year Famine in China (1333-1348), 9 million; Black Death (1343-1352), 75 million; and
the Fall of Yuan Dynasty (1351-1369), 7.5 million. This is the only evidence in the data that
demographic catastrophes might have had influence on the growth of the world population and if such
is the case, not one but five of them were need to generate a small distortion.
There is no indication that exogenous conditions after AD 1400 were different than before AD 1200
so the slightly faster hyperbolic growth from around AD 1400 could be perhaps explained by the natural
70
__________________________________________________________________________________
human response to crisis manifested in the intensified process of regeneration (Malthus, 1798; Nielsen,
2013c).
Closer view of the new growth trajectory, starting from around AD 1400, is displayed in Figure 5.
The new hyperbolic growth was undisturbed until around 1950 when it experienced a small but
unsustained acceleration, as indicated by a slight downward bending of the trajectory of the reciprocal
values. This minor boosting lasted for only a short time and soon the growth of human population
started to be diverted to a slower trajectory, as indicated by a change to an upward bending of the
trajectory of reciprocal values.
Figure 5. Between AD 1400 and around 1950 the growth of human population was hyperbolic. Data show a
minor boosting around 1950 followed quickly by a diversion to a slower trajectory. There was no takeoff from
stagnation to growth at any time. Industrial Revolution had no impact on boosting the world population. The
size of the population is in billions.
Again, there was no dramatic takeoff and no transition from stagnation to growth as claimed
repeatedly by Galor (2005, 2011). This hypothetical but non-existed takeoff, which was supposed to
characterise not only the economic growth but also the growth of human population, is consistently
contradicted by the analysis of the economic growth (Nielsen, 2015a, 2016) and by the presented here
analysis of the growth of the world population.
It is remarkable that the growth of the world population was so hyperbolically stable over the past
12,000 years. The data show that during this long time there were only three transitions: 500 BC - AD
500, AD 1200 - AD 1400 and 1950 - present. Each of the two earlier transitions was a change-over
between hyperbolic trajectories. The outcome of the current transition is unknown. The dynamics of
growth in the past 12,000 years is summarised in Table 1.
Table 1. Dynamics of growth of the world population in the past 12,000 years. Time intervals are approximate.
Hyperbolic Growth Demographic Transitions
10,000 BC 500 BC 500 BC AD 500

a = 2.282=, k 2.210 102 Transition from a fast to much slower hyperbolic trajectory
AD 500 1200 AD 1200 1400
a = 6.940=, k 3.448 103 Transition from a slow to a slightly faster hyperbolic
trajectory
AD 1400 1950 1950 present
a = 9.123
= , k 4.478 103 Transition from a hyperbolic trajectory to an unknown
trend
Total time of hyperbolic growth: Total rime of transitions
10,750 years 1265 years
(~89% of the total combined time) (~11% of the total combined time)
5. Implications for the economic growth

As mentioned earlier, preliminary analysis of Maddisons data (Maddison, 2001, 2010) shows close
similarities between distributions describing economic growth and the growth of human population.
Galor also commented that there was a positive relationship between income per capita and population
71
__________________________________________________________________________________
that existed throughout most of human history (Galor, 2005, p. 177). The study of the economic growth
goes hand in hand with the study of the growth of population.
Our analysis demonstrated that the growth of the world population was hyperbolic, and consequently
monotonic. The frequently-claimed but never proven Malthusian stagnation in the growth of human
population did not exist. There was also no transition from stagnation to growth, which could be
described as a sudden takeoff. The fast-increasing growth of the world population in recent years is just
the natural continuation of the hyperbolic growth.
Our analysis shows that with the exception of just two demographic transitions (500 BC - AD 500,
and AD 1200 - 1400) the growth of human population was hyperbolic until around 1950, when it started
to be diverted to a yet unknown trajectory. The first demographic transition (500 BC - AD 500) was
from a faster to a slower hyperbolic growth. It was definitely not a takeoff from stagnation to growth.
The second transition (AD 1200 1400) was from a slow to a slightly faster hyperbolic trajectory (only
30% faster, as indicated by the parameter k). It was also not a transition from stagnation to growth. The
current transition, which commenced around 1950 was initially to a slightly faster trajectory, which was
soon becoming progressively slower than the preceding hyperbolic trajectory. Here again, there was no
transition from stagnation to growth. For about 89% of the past 12,000 years the growth of human
population was hyperbolic and monotonic and there was never a transition from stagnation to growth
because there was no stagnation. Our analysis shows that the growth of human population was
remarkably stable and robust over the past 12,000 years.
Galor wonders what is the origin of the sudden spurt in growth rates of output per capita and
population? (Galor, 2005, p. 177). This puzzle has now been solved: there was no sudden spurt.
Trying to explain this sudden spurt is like trying to explain why there is water in the middle of the
desert when the image of water is created by a mirage. Such an exercise would be obviously ridiculous.
Likewise, in the case of the growth of population or of the economic growth we can explain the illusion
of the spurt but not existence of the spurt. The illusion of the spurt is explained by the hyperbolic
properties but the sudden spurt has never happened. What we see as a sudden spurt is the natural
continuation of the monotonically-increasing hyperbolic distribution and the simplest way to dispel the
illusion of stagnation and of a sudden spurt is to use the reciprocal values of data (Nielsen, 2014) but
we can also use other methods (Nielsen, 2015a). Data have to be rigorously analysed; otherwise it is
easy to be distracted by illusions. Any perfunctory and hasty examination of data is likely to lead to
incorrect conclusions and we can find many examples of such unprofessional use of data in the Unified
Growth Theory (Galor, 2005, 2011).
We have demonstrated that there was no sudden spurt in the growth rate of the world population
because the growth was hyperbolic, which means that the growth rate was also increasing
hyperbolically with time or linearly with the size of the population, in both cases monotonically [see
the eqn (4)]. Such an increase has no room for any form of spurts.
There were also no spurts during the past two demographic transitions. During the first transition
(500 BC - AD 500), the growth rate decreased from 0.252% in 500 BC to 0.066% in AD 500. During
the second transition (AD 1200 - 1400) the growth rate increased only slightly from 0.123% in AD
1200 to 0.157% in AD 1400.
So, our analysis eliminates at least one of Galors spurts: the alleged spurt in the growth rate of
human population. What remains to be explained is the alleged spurt in the growth rate of output per
capita (GDP/cap) but the analysis of this ratio shows that the growth rate of the GDP/cap was also
increasing monotonically (Nielsen, 2015a). There was no spurt at all. Furthermore, the analysis of the
GDP data (Nielsen, 2016) shows that there were no spurts (takeoffs) in the growth of the GDP.
When data are properly analysed they show that what Galor saw as spurts in the growth rates
represented just the natural features of monotonically increasing hyperbolic distributions describing the
growth of population, the growth of the GDP, the growth of the GDP/cap and of their respective
monotonically-increasing growth rates. All these distributions were slow over a long time and fast over
a short time. These features are real but they represent nothing mysterious but the natural properties of
monotonically-increasing hyperbolic distributions. They create strong illusions of stagnations followed
by sudden spurts or takeoffs but when properly analysed they show that there was no stagnation and
that the sudden spurts (takeoffs) never happened.
Galor wonders about the relationship between the income per capita (GDP/cap) and the population
growth, but the answer to this apparent riddle is simple. When closely analysed, the growth of the
72
__________________________________________________________________________________
population is found to be hyperbolic. The growth of the GDP is also hyperbolic (Nielsen, 2016) and
hence, the growth of the GDP/cap is described by the ratio of hyperbolic distributions, which is just a
linearly-modulated hyperbolic distribution (Nielsen, 2015a). The mystery is solved.
The only features, which need to be explained, are not the stagnation and sudden spurts (takeoffs)
because they did not exist but why the growth of human population and the growth of the GDP were
hyperbolic. This issue diverts our attention from the phantom problems, which do not need to be solved,
and directs it to the problems, which need to be solved, because if we could explain why the growth of
the population and the growth of the GDP were hyperbolic, we could also explain the mechanism of
the historical income per capita.
Finally, we shall address a minor issue, which might help to understand at least one discrepancy
between the fitted hyperbolic curve and the GDP data (Nielsen, 2016). In that analysis, we have found
that one point, located at AD 1 was 77% higher than the fitted hyperbolic distribution. In Figure 3 we
can see that something similar can be observed for the growth of human population. The size of the
population in AD 1 was 71% higher than the size determined by the fitted hyperbolic distribution to the
AD data, and the explanation of this discrepancy is simple: there was a maximum in the growth of the
population around AD 1 caused by the transition from a fast-hyperbolic trajectory during the BC era to
a significantly slower hyperbolic trajectory during the AD era. Close similarities between the growth of
the GDP and the growth of the population displayed by Maddisons data (Maddison, 2001, 2010)
suggest that the 77% difference between the GDP value and the fitted hyperbolic distribution at AD 1
(Nielsen, 2016) might reflect a similar maximum in the growth of the GDP as observed in the growth
of the population.

We have analysed the world population data (Manning, 2008; US Census Bureau, 2015) between
10,000 BC and AD 2015. We have found that the growth was hyperbolic during the BC and AD eras.
We have also found that there were just three, relatively, brief demographic transitions during that
time: between 500 BC and AD 500, between AD 1200 and 1400 and currently from around 1950. These
transitions were of a different kind than usually discussed in academic publications. Contrary to the
frequently-repeated but never proven claims, there was never a transition from stagnation to growth
because there was no stagnation.
The first transition was from a fast-hyperbolic trajectory to a significantly slower hyperbolic
trajectory; the second from a slow hyperbolic trajectory to a slightly faster hyperbolic trajectory; and
the current transition from the latest hyperbolic trajectory to a yet unknown trend. The total fraction of
time characterising hyperbolic growth was about 89% of the past 12,000 years and the total time taken
by transitions was only about 11%. Thus, the analysis shows that population, if unchecked, does not
increase exponentially as believed by Malthus but hyperbolically. There was also no stagnation in the
growth of the world population (Nielsen, 2013a), not only during the AD era but also during the BC
era.
The correct understanding of the growth of human population is essential for the correct
understanding of economic growth because, as pointed out by Galor (2005, 2011), and as can be easily
checked using Maddisons data (Maddison, 2001, 2010), there is a close relationship between the
growth of the population and the growth of the GDP. We have demonstrated that the growth of the
world population was hyperbolic. Growth of the GDP was also hyperbolic (Nielsen, 2016). The growth
of the world GDP/cap can be, therefore, described using hyperbolic distributions (Nielsen, 2015a). It is
simply a ratio of hyperbolic distribution describing the growth of the GDP and the hyperbolic
distribution describing the growth of human population. The ratio of two hyperbolic distributions is
simply a linearly-modulated hyperbolic distribution (Nielsen, 2015a). There is nothing profoundly
mysterious, perplexing or mind-boggling about such distributions as repeatedly claimed by Galor (2005,
2011).
It would be a waste of time, money and human resources to try to explain the mechanism of the
historical economic growth, by using phantom features of stagnation and takeoffs. These erroneous and
frequently-used postulates in the Unified Growth Theory (Galor, 2005, 2011) and in other publication
are misleading and unhelpful. Explanations of the mechanism of the historical economic growth and of
the growth of human population have to be based on accepting hyperbolic growth.
73
__________________________________________________________________________________
References
Durand, J. D. (1977). Historical estimates of world population: An evaluation. Population and Development Review,3(3),
253-296. doi. 10.2307/1971891
Floud, D., & McCloskey, D.N. (1994). The Economic History of Britain since 1700. Cambridge: Cambridge University
Press.
Maddison, A. (2010). Historical Statistics of the World Economy: 1-2008 AD. http://www.ggdc.net/.
Manning, S. (2008). Year-by-Year World Population Estimates: 10,000 B.C. to 2007 A.D.
http://www.scottmanning.com/content/year-by-year-world-population-estimates/
Nielsen, R.W. (2013b). Impact of demographic catastrophes. http://arxiv.org/ftp/arxiv/papers/1311/1311.1850.pdf
Nielsen, R.W. (2013c). Malthusian stagnation or Malthusian regeneration?
Nielsen, R.W. (2014). Changing the Paradigm. Applied Mathematics, 5, 1950-1963.doi. 10.4236/am.2014.513188
Nielsen, R.W. (2015a). Unified Growth Theory Contradicted by the GDP/cap Data.
Nielsen, R. W. (2016). Mathematical analysis of the historical economic growth with a search for takeoffs from stagnation to
growth, Journal of Economics Library, 3(1), 1-23.
US Census Bureau (2016). International Data Base. http://www.census.gov/ipc/www/idb/worldpopinfo.php
74
__________________________________________________________________________________
Demographic Transition Theory Contradicted by Data
By Ron W. NIELSENa
Abstract. In the absence of convincing evidence, data for Sweden and Mauritius are used in academic publications
to illustrate the Demographic Transition Theory. These data are closely examined and found to be in clear
contradiction of this theory. Demographic Transition Theory is also contradicted by the best available data for
England. Other examples of contradicting evidence are also discussed. Correct understanding of the mechanism
of growth of human population is important in the correct interpretation of the per capita economic growth.
Keywords. Population growth, Economic growth, Malthusian stagnation, Unified Growth Theory, Demographic
Growth Theory, Stages of growth, Gross Domestic Product.
JEL. A12, A20, B10, C50, Y80.
1. Introduction
H istorical economic growth can be studied using the Gross Domestic Product (GDP). However,
to understand the time dependence of income per capita, expressed as GDP/cap, it is necessary
to understand not only the economic growth, expressed in terms of the GDP, but also the growth
of human population.
The latest theory describing economic growth is the Unified Growth Theory (Galor, 2005, 2011).
The theory, or model, describing the growth of human population is the Demographic Transition Theory
(see for instance Caldwell, 1976, 2006; Casterline, 2003; Coale, 1973; Haupt & Kane, 2005; Kirk, 1996;
Landry, 1934; Lee, 2003; Lehr, 2009; McFalls, 2007; Notestein, 1945; Olshansky & Ault, 1986;
Olshansky, Carnes, Rogers, & Smith, 1997, 1998; Omran, 1971, 1983, 1998, 2005; Rogers &
Hackenberg, 1987; Singha & Zacharia, 1984; Thompson, 1929; van de Kaa, 2008; Warf, 2010). Both
of these theories use similar language and interpretations. Both of them divide growth into distinctly
different stages governed by distinctly different mechanisms. In particular, both of them claim the
existence of the ages-long epoch of Malthusian stagnation followed by a sudden transition to a distinctly
different stage, the transition described as a sudden takeoff, spurt, sprint or explosion.
A study published over 50 years ago (von Foerster, Mora & Amiot, 1960) demonstrated that the
growth of the world population was hyperbolic during the AD era, showing implicitly that the epoch of
stagnation did not exist and that there was no sudden transition to a new type of growth. This study has
shown that the growth of human population during the AD era was following a monotonically increasing
trajectory. As explained elsewhere (Nielsen, 2014), such a growth cannot be divided into distinctly
different sections governed by distinctly different mechanisms of growth. A single mechanism has to be
applied to the whole distribution. For reasons, which are hard to understand, this crucial publication
(von Foerster, Mora & Amiot, 1960) appears to have been ignored in the demographic research.
More recently (Nielsen, 2016), it has been demonstrated that the growth of the world population was
hyperbolic not only during the AD era, as pointed out by von Foerster, Mora and Amiot (1960) but also
during the BC era. Furthermore, it has been shown that there was no stagnation and no transition to a
faster growth as claimed by the Demographic Growth Theory. This study identified only two transitions
in the past but they were transitions of entirely different kind than claimed by the Demographic
Transition Theory. They were transitions from hyperbolic growth to hyperbolic growth. The first
transition was from a fast-hyperbolic growth to a significantly slower hyperbolic growth and the second
a
Australia.
. +61407201175
Published as: Nielsen, R. W. (2016). Demographic Transition Theory and its link to the historical economic growth.
Journal of Economic and Political Economy, 3(1), 32-49.
75
__________________________________________________________________________________
transition from a slow hyperbolic growth to only slightly faster hyperbolic growth. Thus, these two
studies (Nielsen, 2016; von Foerster, Mora & Amiot, 1960) demonstrate that the Demographic
Transition Theory is incorrect. Now we shall discuss additional evidence and we shall show that the
Demographic Transition Theory is contradicted not only by the aggregate data describing the growth of
the population but also by data describing birth and death rates.
Demographic Transition Theory has been described as a ghost story (Abernethy, 1995). This theory
should have been abandoned long time ago but it is still in circulation and many a demographer would
passionately defend its concepts. Abernethy wonders why this dead theory is still being resurrected and
her suggested explanation is that it is because of the respect for elders. However, would elders feel happy
to be so protected?
Science is full of discarded theories and explanations. This is how science works. New ideas are tried
and if they do not work they are replaced by better ideas or simply abandoned. To cling to incorrect
ideas just because we cannot think about something better to replace them or because we feel we might
offend our predecessors is not only unreasonable but also scientifically unjustified.
Friedman, Managing Editor of the Population and Development Review, claims that the
Demographic Transition Theory with its formulaic presentation of the four states is largely a straw
man (Friedman, 2015). This classical version of the Demographic Transition Theory is now known as
the first demographic transition to which a second demographic transition has been added (Lesthaeghe,
2010, 2014; Lesthaeghe & van de Kaa, 1986; van de Kaa, 2001, 2002). The classical four stages of
growth are still there even though they have no convincing support in data. It is fair to say, that nearly
all statements of a general kind about the classical - for me now the first - demographic transition, can
be easily contradicted (van de Kaa, 2002, p. 9). The classical Demographic Transition Theory appears
to have been not only acknowledged but also reinforced by adding the international migration
component. Kirk observed that Demography is a science short on theory, but rich in quantification
(Kirk, 1996, p. 361) but it would be perhaps better to have science without a theory than science with
a theory contradicted by data.
There is no science without data. In science, even the best constructed theory can be undermined and
even abolished by just one contradicting evidence. It would be better to accept that it is perhaps
impossible to have a general theory in the demographic research and that each case should be explained
individually.
The curious feature of the Demographic Transition Theory is that there is not a single convincing
confirmation of this theory in data. Try as we may, we shall never find data showing convincingly the
four stages of growth. It is for this reason that Montgomery had to stitch the data for Sweden and
Mauritius to illustrate this theory (Montgomery, n.d.). I used Mauritius and added Sweden to the end
of it. I smoothed the stage 1 of Mauritius a bit. It is composite more than purely conceptual
(Montgomery, 2012). It should be emphasised that his aim was not to prove this theory but only to
illustrate it.
Data for Sweden are repeatedly used in support of the Demographic Transition Theory but we shall
show that these data serve as an excellent illustration that the Demographic Transition Theory is
contradicted by empirical evidence. Data for Mauritius are sometimes used (e.g. Lutz & Qiang, 2002)
but we shall show that they also do not support this theory. The best and the most extensive data are for
England (Wrigley & Schofield, 1981). We shall demonstrate that the Demographic Transition Theory
is also contradicted by these data.
It is taken for granted that the first stage, which is believed to have lasted for thousands of years, was
characterised by strong fluctuations in birth and death rates but we have absolutely no data to prove it.
We do not have data for death and birth rates extending over thousands of years, so in this sense at least
this part of the theory is unscientific. We have no choice but to accept it by faith.
However, much more has to be accepted by faith. No-one has ever proven the existence of the first
stage of growth (the epoch of stagnation) proposed by the Demographic Transition Theory. In fact, this
concept is contradicted by data (Nielsen, 2013, 2016; von Foerster, Mora &Amiot, 1960). There was no
stagnation in the growth of human population. However, for doctrines accepted by faith, contradictions
in data are routinely and promptly ignored. The only way to accept this stage of growth is by faith and
by ignoring population data (Maddison, 2010; Manning, 2008; US Bureau of Census, 2016) and their
contradicting evidence, but then it is no longer science. Countless descriptions of this mythical epoch
and of the mechanism of growth during that long time have to be accepted by faith.
76
__________________________________________________________________________________
No-one has ever proven that there was a transition from the first to the second stage. No-one has ever
proven that there was population explosion at a certain time. Rapid growth of the population, interpreted
as population explosion, is real but it is just the natural continuation of hyperbolic growth (Nielsen,
2014, 2016), the type of growth, which was identified over 50 years ago (von Foerster, Mora & Amiot,
1960) but which was also conveniently ignored. The transition from the alleged first to the second stage
has to be accepted by faith and by ignoring not only the evidence published over 50 years ago but also
the extensive population data (Maddison, 2010; Manning, 2008; US Bureau of Census, 2015).
No-one has ever proven that the mechanisms of growth during the alleged first and second stages
were different. No-one has ever proven that the Industrial Revolution boosted the growth of human
population. All these concepts and more have to be accepted by faith supported perhaps occasionally by
the misinterpretation of selected data.
It is believed that strong fluctuations in birth and death rates are reflected in fluctuations in the size
of the population. These assumed fluctuations, described often as Malthusian oscillations, have been
extensively discussed in peer-reviewed literature but no-one cared to check whether fluctuations in birth
and death rates have any influence on the growth of human population. We shall demonstrate that these
fluctuations have absolutely no impact on the growth of human population.
It is believed that the growth of population was stagnant for thousands of years and that it was
characterised by random fluctuations. According to this belief, there were periods of time when the
population did not grow at all and that any gains in the growth of human population made over decades
were wiped out in one or two years (van de Kaa, 2008). Such confident declarations are inaccurate and
misleading. They might apply to some local populations, sometimes, but they certainly do not apply to
the growth of the world population (Nielsen, 2013, 2016). This claim is also not supported by the
regional population data (Maddison, 2010).
Normally, in any scientific investigation, empirical evidence such as published over 50 years ago
(von Foerster, Mora &Amiot, 1960) would have been further investigated. Why was it ignored in the
demographic research? This early observation is now convincingly confirmed (Nielsen, 2016) by new
data (Maddison, 2010; Manning, 2008; US Bureau of Census, 2015). The growth of the population in
the past was hyperbolic. It was slow but it was not stagnant or random. The first stage proposed by the
Demographic Transition Theory did not exist and there was no transition from stagnation to growth.
2. Demographic Transition Theory

We have already mentioned certain features of the Demographic Transition Theory but in order to
understand the discussed examples for Sweden, Mauritius and England we shall now present its brief
outline. Its general concepts are illustrated in Figure 1. In this figure, 1 represents hypothetical birth
rate, 2 hypothetical death rate, 3 the corresponding rate of natural growth (the difference between 1 and
2, generally the same as the growth rate) and 4 the growth of population.
Figure 1. Fundamental concepts of the Demographic Transition Theory based on the illustrations presented by
Montgomery (n.d.) and by van de Kaa (2001, 2002).
77
__________________________________________________________________________________
Demographic Transition Theory describes changes in birth and death rates, in the size of the
population and in the rate of natural increase. According to this theory, changes in socio-economic
conditions lead to transitions in death and birth rates, which in turn are reflected in the growth of human
population (see for instance Caldwell, 1976, 2006; Casterline, 2003; Coale, 1973; Haupt & Kane, 2005;
Kirk, 1996; Landry, 1934; Lee, 2003; Lehr, 2009; McFalls, 2007; Notestein, 1945; Olshansky & Ault,
1986; Olshansky, Carnes, Rogers, & Smith, 1997, 1998; Omran, 1971, 1983, 1998, 2005; Rogers &
Hackenberg, 1997; Singha & Zacharia, 1984; Thompson, 1929; van de Kaa, 2008; Warf, 2010). These
transitions are supposed to have been taking place in four fundamental stages, to which other stages
could be added.
Stage 1 is supposed to have been the pre-industrial stage of stagnation; Stage 2 is supposed to
represent the post-industrial stage of explosion; Stage 3 is the stage of the slowing-down growth; and
Stage 4 is the stage of a stable size of the population. The number of stages can be extended to five
(Haupt & Kane, 2005; Olshansky, Carnes, Rogers, & Smith, 1998; van de Kaa, 2008) or maybe even
to six (Myrskyla, Kohler & Billari, 2009).
The theory was proposed in its inchoate form in 1929 (Thompson, 1929) but the word transition
was not used until 1934 (Landry, 1934). The first clear outline of this theory is attributed to Notestein
(1945). Its fundamental concepts illustrated in Figure 1 are based on the illustration prepared by
Montgomery (n.d.) and by van de Kaa (2001, 2002).
Stage 1 is claimed to have prevailed since time immemorial (Komlos, 2000, p. 320), i.e. for many
thousands of years. The characteristic feature of this stage is supposed to have been the high birth and
death rates fluctuating around the same constant value and producing a stagnant state of growth. The
size of the population remained approximately constant and the rate of natural increase approximately
zero. This stage is described as the Preindustrial Age, the Preindustrial Society, the Malthusian Regime,
the Epoch of Malthusian Stagnation, the Pre-Demographic Transition Stage and the Age of Pestilence
and Famine. Living conditions during that long time are claimed to have been characterised by poor
health care, poor hygiene, inadequate diets, as well as unsanitary drinking water and bacterial diseases
(Warf, 2010:708). During this stage, there was a continuing struggle for survival and the growth of the
population was fluctuating around zero (Warf, 2010, p. 708).
Stage 2 is supposed to have been dramatically different. It was the stage of population explosion,
usually linked with the Industrial Revolution, the stage of transition from ages-long stagnation to a rapid
growth of the population. The rate of natural increase is supposed to have started to increase rapidly
and the size of the population exploded. This stage is described as the Early Industrial Society, the Early
Industrial Age, the Post-Malthusian Regime, the Early-Demographic Transition and the Age of
Receding Pandemics. The transition from Stage 1 to Stage 2 is described as the escape from the
Malthusian trap, the great escape and as the population explosion.
The characteristic feature of this stage is supposed to have been the rapidly declining death rate
described as the mortality transition, allegedly caused by the generally improving living conditions
reflected in a substantially better health care, better hygiene, better access to clean water, improved
sanitation and increased food production (Chrispeels & Sadava 1994; Galor & Weil, 2000; Thomlinson,
1965). These postulated new growth-promoting forces ignited a population explosion (McFalls,
2007). Another characteristic feature of this stage is the continuing high birth rate over a certain time
followed by its gradual decline.
Stage 3 is the stage of the slowing down growth and is described as the Mature Industrial Age, the
Late Industrial Society, the Modern Growth Regime, the Stage of the Late Demographic Transition, or
the Stage of Degenerative and Man-made Diseases. The difference between the mechanism of growth
in Stages 2 and 3, is explained by a change in personal preferences prompted by such factors as women
joining work force, better education, the availability of contraceptives and by the general tendency to
have smaller number of children in order to improve the standard of living.
Stage 4 is the stage of a stable size of the population and is described as the Post-industrial Society,
the Post-industrial Age, the Age of Delayed Degenerative Diseases, the Post-Demographic Transition
Stage or the Stage of Invincibility. This stage is characterised by a close balance between birth and
death rates, similar to the balance claimed for Stage 1, but now both rates are low. Low birth rate is
explained by personal preferences of replacing quantity by quality. The impact of infectious diseases
during this stage is claimed to be low and to be replaced by harmful changes in the lifestyle. Mortality
78
__________________________________________________________________________________
is now associated with smoking and obesity, as well as, to a lesser extent, car accidents, suicides, and
homicides (Warf, 2010, p.710).
We shall now examine the data for Sweden, Mauritius and England and we shall show that they are
in contradiction with the Demographic Transition Theory, but in perfect harmony with other
contradicting evidence (Nielsen, 2013, 2016; von Foerster, Mora & Amiot, 1960).
In comparing this theory with empirical evidence, it is essential to understand the characteristic
features of the alleged Stage 1 and of the transition to Stage 2.
The alleged Stage 1 should be characterised by strong fluctuations in birth and death rates.
On average, birth and death rates should be high and they should be fluctuating around nearly
constant values.
The gap between the fluctuating birth and death rates during this first stage should be on average
zero.
There should be convincing evidence of stagnation in the growth of the population during the alleged
Stage 1.
There should be a clear and convincing transition from Stage 1 to Stage 2, marked by a clear change
in the pattern of growth of human population, from stagnation to growth, so clear that it could be
described as a takeoff, spurt, or explosion.
The transition should be marked by a clear change in the pattern of birth and death rates. On average,
death rates should start to decrease, while the birth rates should, for a certain limited time, remain
constant and then they should also start to decrease.
The gap between birth and death rates should be progressively getting wider from approximately
zero in Stage 1 to a certain maximum value, which would mark the beginning of Stage 3.
3. Examination of data for Sweden

Data for Sweden (Statistics Sweden, 1999), used repeatedly in support of the Demographic
Transition Theory, are displayed in Figure 2.
Figure 2. Demographic Transition Theory is contradicted by the data for Sweden (Statistics Sweden, 1999).
These data appear to be in support of the four stages of growth (cf Figure 1): Stage 1 with its large
and strongly fluctuating birth and death rates; Stage 2 with its widening gap between the average values
of birth and death rates; Stage 3 with its decreasing difference between the birth and death rates; and
79
__________________________________________________________________________________
Stage 4 with its low and nearly equal birth and death rates. These data show also the gradually
decreasing fluctuations in birth and death rates.
However, what we should notice immediately is that birth and death rates in the alleged Stage 1 do
not fluctuate around the same constant value. On average, there is a large gap between them.
In order to produce a stagnant state of growth, birth and death rates have to vary around the same
constant value. It is essential for the difference between them to be on average zero. This difference
could vary between negative or positive values but on average it should not be larger than zero.
Data for death rates or birth rates are often used in support of the Demographic Transition Theory
but such data are meaningless. They should be used together because Demographic Transition Theory
describes how both of them should behave. If the gap between birth and death rates is not on average
zero, there is no stagnation in the growth of population. If the gap is not on average zero, the growth of
population will reveal the presence of a non-random force, constant for the exponential growth, steadily-
increasing as for instance for the hyperbolic growth or steadily decreasing as for the logistic growth.
In the same source of data (Statistics Sweden, 1999), which are used to defend the Demographic
Transition Theory, there are also aggregate data describing the growth of population in Sweden, shown
in the lower section of Figure 2. These data clearly demonstrate that the four stages of growth did not
exist. Aggregate data should never be ignored in testing the Demographic Transition Theory.
Data for Sweden should have never been used to illustrate the validity of the Demographic Transition
Theory because such illustrations are incorrect and misleading. When used in classrooms or lecture
rooms, they do not teach science. When used in academic publications in support of the Demographic
Transition Theory they propagate unscientific and incorrect information.
Death rate shown in the upper section of Figure 2 is decreasing in the apparent agreement with Stage
2 but it was also decreasing in the apparent Stage 1. There was no clear mortality transition, which could
be claimed as marking the change from Stage 1 to Stage 2. Consequently, the apparent Stage 2 cannot
be identified as Stage 2, which puts in question the other apparent stages.
The widening gap during the apparent Stage 2 is only slightly larger than the gap during the apparent
Stage 1. Such a small change could not have produced a desired transition from a stagnant growth
during the alleged Stage 1 to an explosive growth during the apparent Stage 2. In fact, the wide gap
between birth and death rates during the alleged Stage 1 is obviously so large that there must have been
no stagnation during this stage but a steadily-increasing growth of population, and indeed this
expectation is confirmed by the aggregate data describing the growth of population in Sweden and
shown in the lower part of Figure 2. These data show clearly that the four stages of growth did not exist.
Demographic Transition Theory neither describes nor explains the growth of human population in
Sweden and is in gross disagreement with data
The data also show that even large fluctuations in birth and death rates and the resulting fluctuations
in the rate of natural increase had no impact on the growth of population in Sweden. The fluctuations
in birth and death rates did not produce the normally expected Malthusian oscillations in the growth of
population. Maybe much larger fluctuation could have produced some noticeable ripples in the growth
of population but the fluctuations shown in Figure 2 had no effect.
A study of such fluctuations might be interesting for another reason but it has no bearing on
explaining the mechanism of growth of human population. If we look at Figure 2, we can see that some
points for the rate of natural increase are located far from the prevailing trend and yet even such large
fluctuation had no noticeable effect on the recorded size of the population.
Summary of the contradicting evidence:
Contrary to the Demographic Transition Theory, the gap between birth and death rates during the
alleged Stage 1 is not close to zero.
Such a wide gap cannot produce a stagnant state of growth characterised by a zero rate of natural
increase, and indeed the data show that the rate of natural increase during this alleged Stage 1 was not
zero.
The gap between birth and death rates during the alleged Stage 2 is only slightly larger than during
the alleged Stage 1.
Such a difference in the size of the gap cannot produce the population explosion, and indeed there
was no population explosion in Sweden during the displayed time.
Mortality transition (the decreasing death rate) commenced during the alleged Stage 1 and
consequently, the alleged Stage 1 is not Stage 1.
80
__________________________________________________________________________________
Population data demonstrate that the four stages of growth did not exist. They show that there was
a steadily increasing growth of population.
Demographic Transition Theory is contradicted by the data for Sweden.
4. Examination of data for Mauritius

Data for Mauritius (Lehmeyer, 2004; Mauritius, 2015; Statistics Mauritius, 2014; UN, 2013) are
shown in Figure 3. Please notice that the time scales for the two diagrams are not the same.
Using the data for Mauritius in support of the Demographic Transition Theory (e.g. Lutz & Qiang,
2002) is surprising, because the population in Mauritius represents a minute fraction of the world
population, and thus these data can be hardly considered as representing typical patterns of birth and
death rates. Furthermore, these data are poorly documented and it is uncertain, which areas were
included in the population surveys.
Figure 3. Demographic Transition Theory is contradicted by the data for Mauritius (Lehmeyer, 2004;
Mauritius, 2015; Statistics Mauritius, 2014; UN, 2013).
Data describing birth and death rates, shown in the upper section of Figure 3 appear to be supporting
the Demographic Transition Theory. Birth and death rates are at first high and they appear to be
fluctuating around the same constant value, suggesting Stage 1 of growth.
There is also a clear mortality transition at a certain time marked by the rapidly decreasing death
rate, accompanied by an increasing gap between birth and death rates, in good agreement with the
pattern expected for Stage 2, characterised by a transition from stagnation to an explosive growth of the
population. Gradually, the gap between birth and death rates narrows suggesting Stage 3 with a
possibility of developing into Stage 4.
However, this apparent agreement with the theory becomes questionable when we look at the time
scale. The epoch of stagnation as indicated by the merging birth and death rates lasted for only around
20 years. We could, perhaps, extend it to 40 years but we can see that the gap between birth and death
rates started to increase from around 1920. This epoch is probably nothing more than a temporary
delay in the growth of the population.
One of the fundamental principles of scientific investigation is that no relevant data should be
ignored. Consequently, in order to understand the patterns displayed by birth and death rates we have
to include also data describing the growth of population. These data are shown in the lower part of
81
__________________________________________________________________________________
Figure 3 and they now make it perfectly clear that they do not support the Demographic Transition
Theory, because the population was increasing before the apparent Stage 1. Consequently, the apparent
Stage 1 is not Stage 1, which means that the apparent Stage 2 is not Stage 2. The whole pattern of
growth is incompatible with the Demographic Transition Theory. The growth of population in Mauritius
was increasing, sometimes faster and sometimes slower, in complete disagreement with the
Demographic Transition Theory.
Data show that over the displayed time the growth of population was at first slow, then fast, slowing
down, slow, fast, and slowing down again. Data for Mauritius demonstrate that there were more
demographic transitions than claimed by the Demographic Transition Theory.
While the gap between birth and death rates is close to zero as required by the Demographic
Transition Theory for the Stage 1, the empirical evidence indicates that the stagnant state of growth
lasted for only about 20 years or at best for only 40 years. The required evidence should be for at least
a few hundred years, but in principle it should be for thousands of years.
The apparent Stage 1 is not Stage 1 because it was preceded by a fast growth of population, which
is supposed to characterize Stage 2.
Thus, the apparent Stage 2 is not Stage 2 even though it looks like Stage 2 because Stage 2 (fast
growth of population) was before Stage 1
Population data show that there were three, maybe even four, stages of growth during the displayed
short time but these stages have nothing to do with the Demographic Transition Theory.
Demographic Transition Theory is contradicted by the data for Mauritius.
5. Examination of data for England

Probably the best, the most reliable and the most extensive demographic data we might ever expect
to have are for England (Wrigley & Schofield, 1981) between 1541 and 1871. These data are important
not only because of their high accuracy but also because they extend into the time well before of the
Industrial Revolution, dated between 1760 and 1840 (Floud & McCloskey, 1994) and thus they should
reveal the impacts, if any, of this Revolution on the growth of population. Furthermore, it is important
that these data are for England, where the impacts of the Industrial Revolution on the growth of human
population should be strong and clear.
It is here, in England, that we should expect a clear confirmation of a change from high birth and
death rates fluctuating around the same constant value to a new pattern characterised by a rapidly
widening gap between these two quantities, indicating a clear transition from stagnation to population
explosion. It is here, in England, that we should be able to see a clear correlation between the Industrial
Revolution and the morality transition (the decreasing death rate); the clear confirmation of the
beneficial effects of modern progress; the clear evidence of a dramatic escape from the Malthusian trap;
the dramatic transition from Malthusian stagnation (marked by a stagnant stage of growth characterised
by Malthusian oscillations) to a rapid and sustained growth of human population.
Birth and death rates, together with the corresponding rate of natural increase in England are shown
in Figure 4. The time-dependent patterns are entirely different than claimed by the Demographic
Transition Theory (cf Figure 1).
Birth and death rates were always high before, during and after the Industrial Revolution. They
were also not fluctuating around a common constant value before the Industrial Revolution and there
was no morality transition coinciding with this event. In fact, Industrial Revolution had no impact on
the time-dependent distributions of birth and death rates. It is as if this crucial development, which was
supposed to have had such a dramatic impact on the growth of human population, had never happened.
Data show not just one but two mortality transitions (decreasing mortality rate) both beginning well
before the commencement of the Industrial Revolution. Correspondingly, the data show not just one
but two maxima in the rate of natural increase. The increase in the rate of natural increase leading to
these two maxima began well before the Industrial Revolution. This increase was clearly not caused by
the Industrial Revolution. A delayed response to the benefits of progress associated with the Industrial
Revolution could be easily explained, but it would be hard, if not impossible, to explain the anticipated
response.
82
__________________________________________________________________________________
Figure 4. Demographic Transition Theory is contradicted by the data for England (Wrigley & Schofield, 1981).
Growth of human population in England between 1541 and 1871 (Wrigley & Schofield, 1981) is
shown in Figure 5. The top panel shows all the data at yearly intervals. The lower panel shows data at
larger time intervals to allow for comparing them with the numerical integration of the fluctuating rates
of natural increase.
Figure 5. Growth of human population in England (Wrigley & Schofield, 1981).
Data for birth and death rates and for the growth of population, shown in Figures 4 and in the upper
panel of Figure 5, respectively, are at yearly intervals. However, while the data for birth and death rates
and for the corresponding rate of natural increase show strong fluctuations, the data for the growth of
population do not show even a slightest effect of these fluctuations. The growth of population in
England was immune to the fluctuations in birth and death rates.
83
__________________________________________________________________________________
These models of Malthusian oscillations, although elegant and intriguing, must be viewed as quite
speculative in their application to any actual populations (Lee, 1997). Indeed, their presence is
contradicted by the data for England, Sweden and Mauritius as well as by the analyses of the world
population data (Nielsen, 2013, 2016).
We have demonstrated that fluctuations in birth and death rates are not reflected in the growth of
human population. However, we can now take a different approach to our investigation and ask whether
fluctuations in birth and death rates can generate fluctuations in the calculated growth of human
population. Suppose we use the empirically-determined birth and death rates or the rate of the natural
increase representing the difference between birth and death rates, and suppose that we use these rates
to calculate the size of the population, will they produce fluctuations in the calculated distribution?
In order to answer this question, we have carried out numerical integration of the following
differential equation:
1 dS (t )
= Re (t ) (1)
S (t ) dt
where S (t ) is the calculated size of human population and Re (t ) is the empirically-determined, and
fluctuating, rate of natural increase shown in Figure 4.
Migration rates are relatively small (Wrigley & Schofield, 1981) and can be neglected. However, if
the calculated distribution of the size of the population is not going to agree with population data, they
will have to be included.
Results of these numerical calculations are shown in the lower part of Figure 5. The calculated curve
is displayed in steps of one year but it follows the data so closely that in order to see any possible
fluctuations in its calculated values we had to show the data at 10-year intervals. Fluctuating birth and
death rates or the corresponding fluctuating rate of natural increase do not produce fluctuations in the
calculated distribution describing the growth of population.
Growth of population shown in Figure 5 does not display the expected pattern of stagnation followed
by explosion claimed by the Demographic Transition Theory. There was a steady growth of population
well before the Industrial Revolution. This growth was briefly interrupted but it was resumed again
around 1700. Growth of population in England is not correlated with the Industrial Revolution. There
is no indication of the prolonged Malthusian stagnation, no evidence of Malthusian oscillations, no clear
evidence of the existence of stage one and no transition to a new stage. It is just a growth, which was
increasing, halted for a while and started to increase again.
Demographic Transition Theory is contradicted by data for England, the very data, which could have
been expected to serve as the strongest confirmation of this theory. Fluctuations in birth and death rates
shown in Figure 4 had no impact on the growth of the population. The typical pattern of stagnation
followed by explosion is not confirmed by data. Population was increasing well before the onset of the
Industrial Revolution. After a short delay, the population started to increase again but the onset of this
new growth was also before the Industrial Revolution
Growth of human population can be also studied using the reciprocal values of data, 1 / S (t ) . Such
a study gives a new insight into the interpretation of data. This method has been discussed elsewhere
(Nielsen, 2014).
Reciprocal values of the size of human population in England and their absolute gradient, calculated
directly from data and interpolated, are shown in Figure 6.
The deceasing reciprocal values, 1/ S (t ) , of the size of the population indicate an increasing growth,
and vice versa. The top section of Figure 6 shows that the population in England was steadily increasing
well before the Industrial Revolution, as indicated by the steadily-decreasing reciprocal values. After
only a brief interruption, the size of the population in England continued to increase, confirming the
pattern of growth shown in Figure 5.
There was no stagnation that could be identified as Stage 1 proposed by the Demographic Transition
Theory. This stage did not exist. In its place, there was, in general, a steadily-increasing growth. The
data show a temporary distortion of this trajectory but the general pattern was a continuing increase of
the population. Furthermore, the data show no clear impact of the Industrial Revolution on the growth
of population. There was no transition to a distinctly new stage. This pattern of growth is in
contradiction with the pattern proposed by the Demographic Transition Theory.
84
__________________________________________________________________________________
The gradient of 1/ S (t ) shown in Figure 6 is generally negative. It is, therefore, more convenient to
represent it using its absolute value.
The absolute value of the gradient of 1/ S (t ) is a convenient indicator allowing for detecting
changes in the growth pattern. We shall call it the absolute gradient of 1/ S (t ) . A constant absolute
gradient of 1/ S (t ) represents hyperbolic growth. The decreasing absolute gradient of 1/ S (t ) indicates
gradual diversion to a slower trajectory, while the increasing absolute gradient indicates gradual
diversion to a faster trajectory.
Figure 6. Reciprocal values, 1 / S (t ) , of the size of the population and their absolute gradient calculated
directly from data and by interpolation.
The absolute values of the gradient of 1/ S (t ) data are shown in the lower part of Figure 6. If we
compare the upper and the lower sections of this figure we can see that the growth of population in
England was steadily increasing, as indicated by the decreasing reciprocal values 1/ S (t ) , but it was
gradually getting slower, as indicated by the decreasing absolute gradient of 1/ S (t ) . After a short
period of instability, the growth of human population in England started to increase again from around
1690 and was accelerating, as indicated by the downward bending of the reciprocal trajectory and by
its increasing absolute gradient. The onset of this new growth occurred about 70 years before the onset
of the Industrial Revolution. Contrary to the general beliefs, the Industrial Revolution did not boost the
growth of human population in England where its impacts should be stronger than anywhere else. The
growth of population was boosted briefly before the Industrial Revolution.
After a certain time, the acceleration of the growth of human population started to grow weaker, as
indicated by the gradient approaching its maximum value. The absolute gradient of the reciprocal values
reached its maximum around 1800 or right in the middle of the Industrial Revolution and then started
to decrease. The growth of the population in England started to be gradually diverted to a slower
trajectory during the Industrial Revolution.
Thus, the data for England show that there was no stagnation before the Industrial Revolution and
no boosting of growth by the Industrial Revolution. On the contrary, the Industrial Revolution coincides
with the slowing-down growth as indicated by the maximum in the absolute gradient of 1/ S (t ) .
85
__________________________________________________________________________________
If we wanted to claim a cause-effect link between the Industrial Revolution and the growth of human
population in England we could conclude that the Industrial Revolutions slowed down the growth of
the population and diverted it to a slower trend as indicated by the decreasing absolute gradient of the
reciprocal values. However, a reasonable and more convincing conclusion is that the Industrial
Revolution had no impact on the growth of human population. The two processes were totally
independent and it is incorrect to link them by any cause-effect associations. The growth of human
population in England must have been prompted by different forces than the forces associated with the
Industrial Revolution and with the numerous random forces repeatedly proposed to explain the epoch
of stagnation, which did not exist.
The time dependence of birth and death rates in England between 1541 and 1871 (Wrigley &
Schofield, 1981) is in contradiction with the first two stages of growth claimed by the Demographic
Transition Theory. The expected pattern of the fluctuating birth and death rates around a common high
constant value followed by a clear transition to a new stage around the time of the Industrial Revolution
is contradicted by data.
The first stage of growth proposed by the Demographic Transition Theory did not exist.
The data show not just one mortality transition (decreasing death rate) as claimed by the
Demographic Transition Theory, but two, both of them beginning well before the onset of the Industrial
Revolution.
There is no positive correlation between the Industrial Revolution and the time dependence of birth
and death rates.
Data for England show that there were more demographic transitions than can be accounted for by
the Demographic Transition Theory.
With the exception of a minor delay between around 1656 and 1682, the growth of human population
in England was steadily increasing.
Reciprocal values of data for the size of human population also confirm that Industrial Revolution
had no impact on the growth of population in England, where this impact should have been stronger
than anywhere else
Rather than being boosted by the Industrial Revolution, the growth of the population in England
started to be diverted to a slower trajectory from around 1800, i.e. during the Industrial Revolution.
Demographic Transition Theory is contradicted by the data for England between 1541 and 1871.

Data for Sweden, used repeatedly in support of the Demographic Transition Theory, are shown to
be in its direct contradiction. They show that the four stages of growth claimed by the Demographic
Transition Theory did not exist. There was no stagnation (no Stage 1) and no transition to a new stage
(Stage 2) claimed by the Demographic Transition Theory. There was no population explosion and no
transitions to stages three and four. There was just a steadily-increasing, single-stage, growth of
population. The gap between death and birth rates in the apparent Stage 1 was large and there was no
dramatic change in its size during the usually claimed but non-existent transition from Stage 1 to Stage
2.
The data for Mauritius, used sometimes in support of the Demographic Transition Theory (e.g. Lutz
& Qiang, 2002) also show a clear disagreement with this theory. The apparent Stage 1 suggested by the
birth and death rates, even if accepted, lasted for only a few decades. However, when aggregate data
are included, they show that the apparent Stage 1 was not Stage 1 because it was not preceded by
stagnation but by a steadily increasing growth of population. The pattern of growth of human population
does not fit into the pattern claimed by the Demographic Transition Theory
The exceptionally good data for England, 1541-1871, are also in contradiction with the Demographic
Transition Theory. The expected stages in birth and death rates are not confirmed by data. There were
two mortality transitions during that time, both commencing well before the onset of the Industrial
Revolution. There is no correlation between the Industrial Revolution and the time-dependence of the
birth and death rates in England. There was no stagnation followed by population explosion.
Industrial Revolution did not boost the growth of population in England. On the contrary, the data
show that from around 1800, i.e. during the Industrial Revolution, the growth of population in England
86
__________________________________________________________________________________
started to be diverted to a slower trajectory. Consequently, if we wanted to link the Industrial Revolution
with the growth of population we could claim that Industrial Revolution slowed down the growth of the
population. However, a more convincing conclusion is that the two processes were totally independent.
The data indicate that it is incorrect to use the Industrial Revolution to explain the mechanism of growth
of human population, even in England, the centre of this revolution.
While the data for Sweden show a steady growth of the population without any signs of four stages
claimed by the Demographic Transition Theory, the data for Mauritius and England demonstrate that
there were more stages than one can account for by using the Demographic Transition Theory.
The study presented here also shows that even large fluctuations in birth and death rates and in the
corresponding growth rate have no impact on the growth of population. It is possible that exceptionally
large fluctuations in birth and death rates and in the corresponding growth rate might produce some
noticeable ripples in the distribution describing the growth of population but it would be unconvincing
to claim that such possible small ripples had a strong contribution to the mechanism of growth. It is,
therefore, incorrect to imagine that fluctuations in birth and death rates can produce the often-claimed
Malthusian oscillations in the size of the population.
A study published over 50 years ago (von Foerster, Mora & Amiot, 1960) demonstrated that the
growth of human population was hyperbolic. These results are in the direct contradiction of the concept
of Malthusian stagnation in the growth of population and they have been in the direct contradiction for
over 50 years but they appear to have been ignored in favour of the generally accepted but scientifically
unsupported doctrine of stagnation. Results of this study show that there was no Malthusian stagnation
in the growth of population and consequently that Stage 1 claimed by the Demographic Transition
Theory did not exist. In science, even one contradicting evidence is sufficient to show that a contradicted
theory is incorrect. Now we have extensive data, which can be used successfully to refine and correct
the established knowledge (the knowledge established largely on questionable conjectures) including
the data describing population and economic growth (Maddison, 2010; Manning, 2008; US Bureau of
Census, 2015). All these data show consistently that there was no stagnation in the growth of human
population (Nielsen, 2013, 2016). They show clearly that Stage 1 claimed by the Demographic
Transition Theory did not exist and that there was no transition to the alleged Stage 2. They show that
the Demographic Transition Theory is contradicted by data.
Demographic Transition Theory has a strong link with the Unified Growth Theory (Galor, 2005,
2011), which also claims, incorrectly, the existence of the epoch of stagnation and a dramatic transition
to a new stage of economic growth described repeatedly as a takeoff from stagnation to growth. Both
theories, the Demographic Transition Theory and the Unified Growth Theory, are based on dubious
speculations nourished by fertile imagination and by illusions often reinforced by the incorrect
interpretation of hyperbolic growth.
The growth of human population was slow over a long time and fast over a short time but it was
never stagnant. It was slow because it was hyperbolic and fast because it was hyperbolic. It was a
monotonically-increasing hyperbolic distribution (Nielsen, 2016; von Foerster, Mora & Amiot, 1960).
Economic growth, whether expressed in terms of the GDP or GDP/cap, was slow over a long time and
fast over a short time but it was also not stagnant. It was slow because it was hyperbolic and fast because
it was hyperbolic (Nielsen, 2014, 2015). There was no stagnation and no sudden takeoff from the alleged
but non-existent stagnation to growth. There was no transition to a new and faster pattern of growth, a
transition that could be described as a takeoff. It was a monotonically-increasing growth.
Demographic Transition Theory is incorrect and the only way to accept it is by ignoring the
repeatedly contradicting empirical evidence and by placing full trust in stories based largely on creative
imagination perfected by accretion over many years and by many people, each new imagined
explanation or concept creating new ideas and all of them growing into the established knowledge in
demography.
The Demographic Transition Theory (or Model) has been considered to have been a ghost story for
at least 20 years (Abernethy, 1995) and it is not clear why its concepts have not been abandoned long
time ago. It would be probably better to accept openly and clearly that each case should be studied
individually and that it is not necessary to reconcile them with some kind of a master theory, which at
present does not exist.
Scientific principles of investigation can be used even in the absence of an all-encompassing theory,
and the fundamental principle is to refrain from ignoring any relevant data particularly if they contradict
87
__________________________________________________________________________________
the accepted interpretations. It appears that the continuing use of the Demographic Transition Theory
makes the demographic research unscientific because by now and over many years this field of research
evolved into a strong system of concepts many of which can be accepted only by faith.
There is also another problem with the continuing acceptance of this theory. Demographers might
be aware of the fundamental problems associated with the Demographic Transition Theory. However,
many teachers, lecturers and even university professors might be less informed. They accept this theory
as scientific and they teach it to younger generations, who, in turn, accept it as presented to them
believing that they learn science. However, they are neither taught nor learn science. They learn
scientifically-contradicted narrative cloaked as science.
For instance, quite recently, Thompson and Roberge (2015) published an article in which they
present a diagram showing the four stages of growth proposed by the Demographic Transition Theory.
They show how to help students to unpack this rich display of information (p. 254) without being
aware that they are helping to unpack this rich source of misinformation. It would be more useful to
teach students why the diagram they see is a misleading source of misinformation. It is a fiction story,
a ghost story, presented as science, but teachers might not be aware of the problems permeating the
corridors of science. At best, Demographic Transition Theory could be taught in the history of science
and used as a good illustration of scientifically-unsupported theories.
Correct understanding of the growth of human population is important but the misleading
information presented by the Demographic Transition Theory is seriously harmful because this theory
does not explain the growth of human population but presents concepts and explanations, which when
closely examined are contradicted by empirical evidence. It is better to have no theory than a misleading
theory. This theory is the source of misinformation presented as science but it is not science.
References
Abernethy, V. (1995). The demographic transition model: A ghost story. Population and Environment: A Journal of
Interdisciplinary Studies, 17(1), 3-5. doi. 10.1007/BF02208273
Caldwell, J. C. (1976). Towards a reinstatement of demographic transition theory. Population and Development Review, 2(3-
4), 321-266.
Caldwell, J. C. (2006). Demographic Transition Theory. Dordrecht, Netherlands: Springer.
Casterline, J. B. (2003). Demographic transition. In P. Demeny & G. McNicoll (Eds.), Encyclopedia of Population, Vol. I
(pp. 210-216). Farmington Hills, MI: Macmillan Reference.
Chrispeels, M. J., & Sadava, D. E. (1994). Human population growth: Lessons from demography. In: D. Jones & T. L.
Bartlett (Eds.), Plants, Genes, and Agriculture (pp. 13-34). Boston, MA: Jones and Bartlett Publishers.
Coale, A. (1973). The demographic transition reconsidered. Proceedings of the International Population Conference, 53-72.
Liege, Belgium: IUSSP
Press,
Friedman, R. (2015). Private communication.
Transition and Beyond. The American Economic Review, 90(4), 806-828. doi. 10.1257/aer.90.4.806
Haupt, A., & Kane, T. T. (2005). Population handbook (5th edn.). Washington, DC: Population Reference Bureau.
Kirk, D. (1996). Demographic Transition Theory. Population Studies, 50(3), 361-387. doi. 10.1080/0032472031000149536
History, 29, 307-331.
Landry, A. (1934). La revolution dmographique: tudeetessai sur les problemes de la population. Paris: INED.
Lesthaeghe, R. (2010). The unfolding story of the second demographic transition. Population and Development Review,
36(2), 211-251. doi. 10.1111/j.1728-4457.2010.00328.x
Lesthaeghe, R. (2014). The second demographic transition: A concise overview of its development. Proceedings of the
National Academy of Sciences, 111, 1811218115. doi. 10.1073/pnas.1420441111
Lesthaeghe, R., & van de Kaa, D.J. (1986). Twee demografischetransities? In D.J. van de Kaa and R. Lesthaeghe (Eds.),
Bevolking: GroeienKrimp (pp. 9-24). Deventer: Van LoghumSlaterus
Lee, R. (2003). The demographic transition: Three centuries of fundamental change. Journal of Economic Perspectives,
17(4), 167-190. doi. 10.1257/089533003772034943
Lehr, C. S. (2009). Evidence on the demographic transition. Review of Economics and Statistics, 91(4), 871-887. doi.
10.1162/rest.91.4.871
88
__________________________________________________________________________________
Lutz, W., & Qiang, R. (2002). Determinants of human population growth. Phil. Trans. R. Soc. Lond., 357, 11971210. doi.
10.1098/rstb.2002.1121
Maddison, A. (2010). Historical statistics of the world economy: 1-2008 AD. http://www.ggdc.net/maddison/Historical
http://www.scottmanning.com/content/year-by-year-world-population-estimates/and references therein.
McFalls, J.A. Jr, (2007). Population: A Lively Introduction (5th ed.). Washington DC: Population Reference Bureau.
Montgomery, K. (n.d.). The Demographic Transition. http://www.marathon.uwc.edu/geography/Demotrans/demtran.htm
Montgomery, K. (2012). Private communication
Myrskyla, M., Kohler, H-P., & Billari, F. (2009). Advances in development reverse fertility declines. Nature, 460, 741-743.
doi. 10.1038/nature08230
Notestein, F. (1945). Population: The long view. In T. Shultz (ed.), Food for the World (pp. 36-57). Chicago: University of
Chicago Press.
Nielsen, R. W. (2013). No stagnation in the growth of population. http://arxiv.org/ftp/arxiv/papers/1311/1311.3997.pdf
Nielsen, R. W. (2015). Unified Growth Theory Contradicted by the GDP/cap Data.
Nielsen, R. W. (2016). Growth of the World Population in the Past 12,000 Years and Its Link to the Economic Growth.
Journal of Economic Bibliography, 3(1), 1-12.
Olshansky, S. J., Carnes, B., Rogers, R.G., & Smith, L. (1997). Infectious diseases - New and ancient threats to world health.
Population Bulletin, 52(2), 2-52.
Olshansky, S. J. Carnes, B. A, Rogers, R. G. & Smith, L. (1998). Emerging infectious diseases: The Fifth stage of the
epidemiologic transition? World Health Statistics Quarterly, 51(2-4), 207-217.
Fund Quarterly, 83(4), 509-538. doi. 10.1111/j.1468-0009.2005.00398.x
Omran, A. R. (1983). The epidemiologic transition theory. A preliminary update. Journal of Tropical Pediatrics, 29(6), 305-
316. doi. 10.1093/tropej/29.6.305
191. doi. 10.2307/3349375
Quarterly, 83(4), 731-757. doi. 10.1111/j.1468-0009.2005.00398.x
Rogers, R. G., & Hackenberg, R. (1987). Extending epidemiologic transition theory, A new stage. Social Biology, 34(3-4),
234-243. doi. 10.1080/19485565.1987.9988678
Singha, V. C., & Zacharia, E. (1984). Elements of demography. New Delhi: Allied Publishers Pvt Ltd.
Random House.
Thompson, W. S. (1929). Population. American Journal of Sociology, 34(6), 959-975.
Thompson, V., Roberge, M. C. (2015). An Alternative Visualization of the Demographic Transition Model. Journal of
Geography, 114, 254259. doi. 10.1080/00221341.2014.983144
UN (2013). World Population Prospects: The 2012 Revision, DVD Edition. Washington DC: United Nations, Department of
Economic and Social Affairs, Population Division, http://esa.un.org/unpd/wpp/Excel-Data/Interpolated.htm
US Census Bureau (2015). International Data Base. http://www.census.gov/ipc/www/idb/worldpopinfo.php and references
therein.
van de Kaa, D. J. (2001). Demographic Transition, Second. In N. J. Smelser & B. Baltes (Eds.), International Encyclopedia
of the Social and Behavioral Sciences (pp. 3486-3488). Oxford: Elsevier Science Ltd.
van de Kaa, D. J. (2002). The idea of a second demographic transition in industrialized countries. Paper presented at the
Sixth Welfare Policy Seminar of the National Institute of Population and Social Security, Tokyo, 29 January.
van de Kaa, D. J. (2008). Demographic Transition. In Y. Zeng (Ed.). Demography. Encyclopedia of Life Support Systems.
Vol. 2. Paris: UNESCO.
von Foerster, H., Mora, P., & Amiot, L. (1960). Doomsday: Friday, 13 November, A.D. 2026. Science, 132, 1291-1295. doi.
10.1126/science.132.3436.1291
Warf, B. (2010). Demographic Transition. In Warf B (Ed.). Encyclopedia of Geography (pp. 708-711). Thousand Oaks, CA:
SAGE.
(Publishers) Ltd.
89
__________________________________________________________________________________
Unified Growth Theory Contradicted by the Absence of Takeoffs

in the Gross Domestic Product
By Ron W. NIELSENa
Abstract. Data describing historical economic growth are analysed. They demonstrate convincingly that takeoffs
from stagnation to growth, claimed in the Unified Growth Theory, never happened. This theory is again
contradicted by the same data which were used, but never properly analysed, during its formulation. The claimed
takeoffs were constructed by Galor by the manipulation of data and by their grossly distorted presentations. Now,
the same data are used to show that the takeoffs did not exist. The absence of the claimed takeoffs demonstrates
that the postulate of the differential takeoffs is also contradicted by data. Furthermore, this analysis demonstrates
that the mathematically-analysable data contradict the concept of the prolonged Malthusian stagnation, its effects
on the economic growth as well as the concept of a dramatic escape from the Malthusian trap.
Keywords. Historical economic growth, regimes of growth, Malthusian stagnation, takeoffs, Malthusian trap,
hyperbolic growth.
JEL.A10, B10, B22, F01, N10.
1. Introduction
O ne of the fundamental postulates of the Unified Growth Theory (Galor, 2005a, 2011) is the
postulate of takeoffs from stagnation to growth. This feature is supposed to mark a boundary
between the ages-long epoch of Malthusian stagnation and a new epoch of a rapid economic
growth. An easy way to test the Unified Growth Theory is to look for such postulated takeoffs because
they should be easily identifiable. The added advantage of using this test is that it also checks the validity
of yet another postulate of this theory, the postulate of the differential takeoffs.
In our analysis, we shall use the excellent data published by the world-renown economist (Maddison,
2010). The data presented in this compilation are virtually the same as in his earlier compilation
(Maddison, 2001), which Galor was using during the formulation of his Unified Growth Theory. The
difference between the two compilations is that the new set of data was extended to the 21st century.
These extended data are not essential for testing the Unified Growth Theory but they help in
demonstrating the latest transitions from the historical hyperbolic growth to slower trajectories.
Unfortunately, Galor did not analyse Maddisons data. His interpretations of the mechanism of
economic growth are based on strongly questionable quotations of isolated numbers, on the unfortunate
simplistic and self-misleading examination of data and on the habitual use of grossly distorted diagrams
(Ashraf, 2009; Galor, 2005a, 2005b, 2007, 2008a, 2008b, 2008c, 2010, 2011, 2012a, 2012b, 2012c;
Galor & Moav, 2002; Snowdon & Galor, 2008).
Historical economic growth and the growth of human population can be described using hyperbolic
distributions (Nielsen, 2014, 2015a, 2015b, 2015c, 2015d, 2016a, 2016b, 2016c, 2016d; von Foerster,
Mora & Amiot, 1960) Unlike the better-known exponential growth, which is easier to understand,
hyperbolic distributions are strongly deceptive because they appear to be made of two distinctly
different components, slow and fast, joined perhaps by a certain transition component. This illusion is
so strong that even the most experienced researchers can be easily deceived particularly if their research
is based on a limited body of data, as it was in the past. Fortunately, Maddisons data solve this problem,
and fortunately also their analysis is trivially simple because, as pointed out earlier (Nielsen, 2014),
a
Australia.
. +61407201175
Published as: Nielsen, R. W. (2016). Unified Growth Theory contradicted by the absence of takeoffs in the Gross Domestic
Product. Economic Review, 3(1), 16-27.
90
__________________________________________________________________________________
hyperbolic distributions can be easily identified and analysed using the reciprocal values of data.
Consequently, if in the past, researchers were basing their conclusions on the strongly-limited sets of
data and imagined that there was a prolonged epoch of stagnation followed by sudden takeoffs in
various geographical regions, now there is no excuse to continue with such interpretations because we
have excellent sets of data, which lead to the entirely different conclusions. It is, therefore surprising, if
not disappointing, that Galor, who had access to these excellent data and even used them during the
formulation of his theory, did not analyse them properly but followed the traditional and incorrect
interpretations of the historical economic growth.
Theories play an important role in scientific research because they crystallise interpretations of
studied phenomena. However, theories have to be always tested by data. In science, it is important to
look for data confirming theoretical explanations but it is even more important to discover contradicting
evidence, because data confirming a theory confirm only what we already know but contradicting
evidence may lead to new discoveries.
According to Galor, historical economic growth can be divided into three distinctly-different stages
governed by three distinctly different mechanisms: (1) the Malthusian regime of stagnation, (2) the
post-Malthusian regime, and (3) the sustained-growth regime. We have already demonstrated that this
postulate of the three regimes of economic growth is contradicted by the data for Western Europe,
Eastern Europe, Asia, countries of former USSR, Africa and Latin America (Nielsen, 2015a, 2015b,
2015c, 2015d, 2016b), ironically by the same data which were used but never analysed by Galor.
This fundamental postulate of the three regimes of growth is used repeatedly throughout the narrative
of the Unified Growth Theory and serves as the essential support for the discussed interpretations and
explanations. Without this corner stone, the whole theory becomes unsupported.
According to Galor, The take-off of developed regions from the Malthusian Regime was associated
with the Industrial Revolution and occurred at the beginning of the 19th century, whereas the take-off
of less developed regions occurred towards the beginning of the 20th century and was delayed in some
countries well into the 20th century (Galor, 2005a). Even more precisely (Galor, 2008a, 2012a),
Malthusian regime of stagnation was supposed to have been between 100,000 BC and AD 1750 for
developed regions and between 100,000 BC and AD 1900 for less-developed regions. The post-
Malthusian regime was allegedly between AD 1750 and 1850 for developed regions and from 1900 for
less-developed regions. The sustained-growth regime was supposed to have commenced around 1850
for developed regions.
The claimed starting time of the Malthusian regime appears to be based entirely on conjecture
because Maddisons data are terminated at AD 1 and even they contain significant gaps below AD
1500. The claimed date of 100,000 BC is also hanging in the middle of nowhere because the origin of
Homo sapiens is usually placed around 200,000 BC. However, Weaver, Roseman & Stringer (2008)
have pointed out that the divergence of the lineages of modern humans and Neanderthals might have
occurred around 309,000 BC or even 433,000 BC.
We have no mathematically analysable data over such a long time so any claim of the existence of
Malthusian stagnation in the economic growth in the distant past is based on questionable conjectures.
However, we have mathematically-analysable data describing the growth of the population from 10,000
BC and they show that the growth of the population was not stagnant but hyperbolic not only during
the AD era, as pointed out over 50 years ago by von Foerster, Mora and Amiot (1960) but also during
the BC era (Nielsen, 2016c).
Hyperbolic growth was slow in the past but it was not stagnant. Slow hyperbolic growth should
never be interpreted as stagnant because if we want to interpret the slow hyperbolic growth as stagnant,
and governed by the usually assumed multitude of random forces, we should use precisely the same
mechanism to explain the fast hyperbolic growth. It is impossible to divide the monotonically-
increasing hyperbolic distributions into slow and fast components (Nielsen, 2014). Hyperbolic
distributions have to be interpreted as a whole and the same mechanism has to be applied to the apparent
slow growth and to the apparent fast growth. There is no clearly defined transition between the apparent
slow and the apparent fast growth.
The alleged transition at the end of the postulated regime of Malthusian stagnation for various
regions and countries is described by Galor as the sudden take-off from stagnation to growth (Galor,
2005a, pp. 177, 220, 277), as a sudden spurt (Galor, 2005a, 177, 220) or as remarkable or
91
__________________________________________________________________________________
stunning escape from the Malthusian trap (Galor, 2005a, pp. 177, 220). It is a signature, which cannot
be missed.
For developed regions, this signature is supposed to have coincided with the onset of the Industrial
Revolution, 1760-1840 (Floud & McCloskey, 1994). Indeed, Industrial Revolution is considered to
have been the prime engine of economic growth (Galor, 2005a, p. 212).
The signature of the takeoffs is characterised by three features: (1) it should be a prominent change
in the pattern of growth, (2) it should be a transition from stagnation to growth and (3) it should occur
at the time claimed by the theory. For developed regions, the postulated takeoffs should occur around
AD 1750. For less-developed regions, they should occur around 1900.
growth. Thus, for instance, a transition from hyperbolic growth to another hyperbolic growth is not a
signature of the sudden takeoff from stagnation to growth. Likewise, a transition at a distinctly different
time is not a confirmation of the theoretical expectations.
We shall now demonstrate that the postulated takeoffs never happened and consequently that the
concept of the differential takeoffs is contradicted by data, because in the absence of takeoffs it makes
no sense to claim that they occurred at different times for different regions. In the future, we shall also
demonstrate that The mind-boggling phenomenon of the Great Divergence (Galor, 2005a, p. 220) is
mind-boggling only because it is hard to understand how anyone familiar with mathematics could be
puzzled by such an artificially-created structure. If hyperbolic distributions are not properly analysed
they can be used to generate such phantom and totally meaningless features. Scientific analysis of
Maddisons data opens a new outlook on the interpretation of the historical economic growth.
Throughout the analysis presented here, the values of the Gross Domestic Product (GDP) will be
expressed in billions of the 1990 International Geary-Khamis dollars. Parameters describing the fitted
distributions were determined by the mathematical analysis (Nielsen, 2016b) of Maddisons data
(Maddison, 2010).

Results of mathematical analysis of the world economic growth are presented in Figure 1. If the
Unified Growth Theory (Galor, 2005a, 2008a, 2011, 2012a) is correct, we should see clear signs of two
takeoffs: around 1750 for developed regions and around 1900 for less-developed regions. We see none
of them.
The data and their analysis are in the direct contradiction of this theory. They show that the economic
growth was remarkably stable and that the claimed or wished-for takeoffs never happened. The absence
of the two claimed takeoffs is strikingly conspicuous. Galors claim of the spectacular or stunning
escapes from Malthusian trap (Galor, 2005a, pp. 177, 220) is spectacularly and stunningly contradicted
by the analysis of the economic-growth data, the same data, which he used, but never properly analysed,
during the formulation of his theory.
The absence of the takeoffs has been also demonstrated for the income per capita data (GDP/cap) for
the world economic growth (Nielsen, 2015e). In science, such single demonstration would have been
sufficient to show that the Unified Growth Theory needs to be revised to bring it in agreement with data,
however, when closely analysed this theory is found to be repeatedly contradicted by data (Nielsen,
2015a, 2015b, 2015c, 2015d, 2015e, 2016a, 2016b).
Hyperbolic growth of the world economy is in harmony with the hyperbolic growth of the world
population (Nielsen, 2016c; von Foerster, Mora & Amiot, 1960). In both cases, growth was indeed slow
over a long time and fast over a short time. In both cases the growth creates an illusion of stagnation
followed by a sudden takeoff. However, in both cases the growth was hyperbolic. There was no
stagnation and no sudden takeoff. Furthermore, in both cases the growth started to be diverted, relatively
recently, to slower trajectories.
3. Western Europe
The growth of the GDP in Western Europe is shown in Figure 2. Results of analysis show that there
was no takeoff from stagnation to growth because (1) there was no stagnation and (2) because the
economic growth, which is described well by the hyperbolic trajectory, was stable during the time of
the alleged takeoff. The takeoff simply did not happen.
92
__________________________________________________________________________________
The claim of the stunning or remarkable takeoff is contradicted by data. There was no takeoff of any
kind, stunning or less stunning, remarkable or less remarkable, sudden or gradual; none at all. The
Industrial Revolution, the alleged prime engine of economic growth (Galor, 2005a, p. 212), made no
impression on changing the economic growth trajectory in the region where this engine should have
been working most efficiently. Industrial Revolution brought many other important changes but,
surprisingly perhaps, did not change the economic growth trajectory in the countries closest to this
monumental development.
4. Eastern Europe
The analysis of the historical data for Eastern Europe is summarised in Figure 3. There was no
stagnation and no takeoff at any time. Industrial Revolution had no impact on changing the economic
growth trajectory in the countries of Eastern Europe.
5. Former USSR
The analysis of data for countries of the former USSR is presented in Figure 4. There was no
stagnation and no takeoff at any time. Industrial Revolution had no impact on changing the economic
growth trajectory in countries of the former USSR.
6. Asia
Analysis of the historical economic growth in Asia (including Japan) is summarised in Figure 5.
(Until AD 1900, Japans contribution to the total economic growth in Asia was on average only 5%.)
Asia is made primarily of less-developed countries (BBC, 2014; Pereira, 2011) and consequently,
according to the Unified Growth Theory (Galor, 2005a, 2008a, 2011, 2012a), economic growth in this
region should have been stagnant until around 1900, the year marking the alleged stunning escape from
the Malthusian trap, the escape manifested by the postulated dramatic takeoff.
The data and their analysis show that there was no stagnation and no claimed takeoff from stagnation
to growth. The data reveal a steadily increasing and stable hyperbolic growth until around 1950. From
around that year, economic growth was diverted to a slightly faster trajectory. This boosting occurred
close to the time of the postulated takeoff from stagnation to growth. However, it was not a transition
from stagnation to growth but from growth to growth.
It should be noted that this temporary boosting is now returning to the original hyperbolic trajectory
and is likely to move to the other side. It is already following a slower trajectory, because its gradient
is smaller than the gradient of the historical trajectory. It would be interesting to explore and explain
the mechanism of this boosting but we shall not find its explanation in the Unified Growth Theory. This
theory does not even notice this feature. Closer but still preliminary inspection of data seems to suggest
that the observed boosting around 1950 in the economic growth in Asia was cased primarily, if not
entirely, by the increased economic activity in Japan.
7. Africa
Results of analysis for Africa are presented in Figure 6. Africa is also made of less-developed
countries (BBC, 2014; Pereira, 2011) so according to the Unified Growth Theory (Galor, 2005a, 2008a,
2011, 2012a) it should have experienced stagnation in the economic growth until around 1900 followed
by a clear takeoff from stagnation to growth around that year. These expectations are contradicted by
the economic growth data because (1) economic growth was not stagnant but hyperbolic (Nielsen,
2015d, 2016b), (2) there was no takeoff from stagnation to growth around 1900 or around any other
time and (3) shortly after the expected time of the takeoff, economic growth in Africa started to be
diverted to a slower trajectory.
As discusses elsewhere (Nielsen, 2015d, 2016b), there was an acceleration in the economic growth
in Africa around 1820. However, this acceleration occurred significantly earlier than the expected
takeoff around 1900 and it was not a transition from stagnation to growth but from growth to growth.
Even more specifically, it was a transition from the hyperbolic growth to another hyperbolic growth.
This acceleration can be explained by noticing that it appears to coincide with the intensified
colonisation of Africa (Duignan & Gunn, 1973; McKay, Hill, Buckler, Ebrey, Beck, Crowston, &
Wiesner-Hanks, 2012; Pakenham, 1992). The fast-increasing GDP after 1820 was not reflecting the
93
__________________________________________________________________________________
rapidly improving living conditions of African population brought about by the beneficial changes
caused by the Industrial Revolution but the rapidly increasing wealth of new settlers and their countries
of origin at the expense of the deploring living conditions of the native populations.
The takeoff from stagnation to growth, claimed by the Unified Growth Theory (Galor, 2005a, 2008a,
2011, 2012a), did not happen in the region where stagnation should have been prominently present.
Economic growth was always stable in Africa (Nielsen, 2015d, 2016b) and now it is being diverted to
a slower trajectory. Escape from the Malthusian trap never happened because there was no trap.
Economic growth was never stagnant in Africa but hyperbolic.
8. Latin America
Results of the analysis of the economic growth in Latin America are presented in Figure 7. Latin
America is also made of less-developed countries (BBC, 2014; Pereira, 2011) so again, according to
the Unified Growth Theory (Galor, 2005a, 2008a, 2011, 2012a), economic growth in this region should
have been stagnant until around 1900 and fast-increasing from around that year. This pattern of growth
is stunningly contradicted by data, the same data, which were used, but never properly analysed, during
the formulation of this theory. At the time of the claimed stunning and remarkable escape from the
Malthusian trap (Galor, 2005a, pp. 177, 220) economic growth in Latin America was already diverted
to a slower trajectory.

Results of mathematical analysis of Maddisons data (Maddison, 2010) show convincingly that
takeoffs from stagnation to growth, claimed repeatedly in the Unified Growth Theory (Galor, 2005a,
2008a, 2011, 2012a) never happened. The growth of the GDP was not stagnant but hyperbolic and, in
general, remarkably stable.
It is essential to understand that claims about the existence of the epoch of Malthusian stagnation in
the economic growth or in the growth of human population are not supported by the scientifically-
analysable data. They are based on conjectures and impressions and they introduce the unwelcome and
undesirable ballast in the economic and demographic research, directing them into unproductive
channels, which move the economic and demographic research away from science and develop them
into a fiction, because in the absence of scientifically analysable data the concepts of stagnation and of
the dramatic escape from the mythical Malthusian trap are supported by fertile imagination and creative
writing, which are not moderated by the rigorous process of scientific investigation.
A clear way of demonstrating that the doctrine of Malthusian stagnation and its effects on the
economic growth or on the growth of human population is incorrect is by demonstrating the absence of
takeoffs from the alleged stagnation to growth. As demonstrated here, such takeoffs did not exist in the
economic growth. They also did not exist in the growth of human population (Nielsen, 2016c, 2016d).
Demographic Transition Theory, the only theory used by demographers to explain the historical growth
of human population, also claims the existence of Malthusian stagnation followed by a dramatic takeoff
from stagnation to growth but this theory is repeatedly contradicted by data (Nielsen, 2016e).
Slow economic growth or the growth of human population is routinely interpreted as stagnation but
such interpretations are incorrect because the slow growth is an integral part of the hyperbolic growth,
which cannot be divided into slow and fast components (Nielsen, 2014) and which has to be interpreted
as a whole by using the same mechanism for the whole distribution. We already know that the growth
of human population during the AD and BC eras was not stagnant but hyperbolic from at least 10,000
BC (Nielsen, 2016c; von Foerster, Mora & Amiot, 1960). We do not have mathematically-analysable
data for the economic growth over such a long time, but the data we have (Maddison, 2010) show
conclusively that during the time described by these data, economic growth was also hyperbolic and
consequently that it was not stagnant. Furthermore, we have also proven that Galors concept of the
existence of the three regimes of growth is contradicted by the analysis of the economic growth in
Western Europe, Eastern Europe, Asia, countries of the former USSR, Africa and Latin America
(Nielsen, 2015a, 2015b, 2015c, 2015d, 2016a).
There is no scientific support for the concept of Malthusian stagnation and for the dramatic escape
from the Malthusian trap, which is supposed to have been manifested in the dramatic takeoffs.
Mathematically analysable data describing economic growth and the growth of human population show
94
__________________________________________________________________________________
repeatedly and consistently that takeoffs from stagnation to growth never happened because there was
no stagnation. Mathematically analysable data show repeatedly and consistently that the economic
growth and the growth of human population were hyperbolic. Concepts of prolonged stagnation
followed by a remarkable or stunning escape from Malthusian trap (Galor, 2005a, pp. 177, 220)
are repeatedly and consistently contradicted by data.
In science, such overwhelming evidence would have been more than sufficient to show that the
theory is unacceptable and that it should be either thoroughly revised or rejected and replaced by a more
suitable theory, a theory based on scientific analysis of data, a reliable theory, which could be used in
the economic growth research. In its present form, Unified Growth Theory is neither reliable nor useful.
In fact, it is unscientific and strongly misleading.
Our analysis of Maddisons data (Maddison, 2010) shows not only that the concept of Malthusian
regime of stagnation followed by dramatic escapes from Malthusian trap is incorrect but also that the
concept of the differential takeoffs is incorrect because we cannot have differential takeoffs without
takeoffs.
Unified Growth Theory is riddled with questionable claims and interpretations. In due time, we shall
demonstrate that this theory is contradicted by regional GDP/cap data in much the same way as it is
contradicted by the global data (Nielsen, 2015e). We shall show that this theory is contradicted by the
economic growth in the UK, the centre of the Industrial Revolution where the Unified Growth Theory
should have the strongest support. It can be also shown that this theory is contradicted by the economic
growth in other individual countries.
We shall demonstrate that the postulate of the great divergence is also based on the incorrect
interpretation of the mathematical properties of hyperbolic distributions. Furthermore, we shall
demonstrate that Galors repeated interpretation of growth rates of income per capita is incorrect.
In its present form, Unified Growth Theory is unacceptable. In order to improve it, it would be
necessary to examine it closely to determine not only how much of it is based on the incorrect
interpretation of data but also how much is just pure fantasy. However, the best solution would probably
be to replace it by a new theory.
Close analysis of Maddisons data (Maddison, 2010) opens new and fascinating avenues for the
economic research. Rather than devoting time and financial resources on explaining features based on
impressions and on the already contradicted conjectures, we can focus our attention of explaining the
features confirmed by the scientific analysis of data. In particular, the relevant and still unanswered
questions are why the historical economic growth was hyperbolic, what mechanism should we use to
explain this type of growth and why, relatively recently, the economic growth, global and regional, has
been diverted to generally slower trajectories. Even the temporarily slightly boosted economic growth
in Asia appears to be also a part of the generally-observed diversions to slower trajectories.
References
Ashraf, Q.H. (2009). Essays on Deep Determinants of Comparative Economic Development. Ph.D. Thesis, Department of
Duignan, P., & Gunn, L.H. (Eds.) (1973). Colonialism in Africa 1870 1960: A Bibliographic Guide to Colonialism in Sub-
Floud, D., & McCloskey, D.N. (1994). The Economic History of Britain since 1700. Cambridge: Cambridge University
Press.
475.
95
__________________________________________________________________________________
0062-7
117, 1133-1191. doi.10.1162/003355302320935007
McKay, J. P., Hill, B.D., Buckler, J., Ebrey, P. B., Beck, R. B., Crowston, C. H., & Wiesner-Hanks, M. E. (2012). A History
Books.
Nielsen, R. W. (2015a). Unified Growth Theory Contradicted by the Economic Growth in Europe.
Nielsen, R. W. (2015b). Unified Growth Theory Contradicted by the Economic Growth in Asia.
Nielsen, R. W. (2015c). Unified Growth Theory Contradicted by the Economic Growth in the Former USSR.
Nielsen, R. W. (2015d). Unified Growth Theory Contradicted by the Economic Growth in Africa,
Nielsen, R. W. (2015e). Unified Growth Theory Contradicted by the GDP/cap Data.
Nielsen, R. W. (2016a). Unified Growth Theory Contradicted by the Economic Growth in Latin America.
Nielsen, R. W. (2016b). Mathematical Analysis of the Historical Economic Growth with a Search for Takeoffs from
Stagnation to Growth. Journal of Economic Library, 3(1), 1-23.
Nielsen, R. W. (2016c). Growth of the World Population in the Past 12,000 Years and Its Link to the Economic Growth.
Nielsen, R. W. (2016d). Unified Growth Theory Contradicted by the Mathematical Analysis of the Historical Growth of
Human Population. http://arxiv.org/ftp/arxiv/papers/1601/1601.07291.pdf
Nielsen, R. W. (2016e). Demographic Transition Theory and Its Link to the Historical Economic Growth. Journal of
Economics and Political Economy, 3(1), 32-49.
bank/
Avon Books.
Snowdon, B., & Galor, O. (2008). Towards a Unified Theory of Economic Growth. World Economics, 9, 97-151.
von Foerster, H., Mora, P., & Amiot, L. (1960). Doomsday: Friday, 13 November, A.D. 2026. Science, 132, 1291-1295. doi.
10.1126/science.132.3436.1291
Weaver, T.D., Roseman, C.C., & Stringer C.B. (2008). Close correspondence between quantitative and molecular-genetic
States of America, 105, 4645-4649. doi. 10.1073/pnas.0709079105
96
__________________________________________________________________________________
Appendix
World Economic Growth
Figure 1. No takeoffs from stagnation to growth. Two postulated takeoffs are indicated (Galor, 2005a, 2008a,
2011, 2012a): for developed regions, around 1750 and for less-developed regions around 1900. The world
economic growth was not stagnant but hyperbolic and it was remarkably stable. Industrial Revolution, the
prime engine of economic growth (Galor, 2005a, p. 212), had no impact on changing the economic growth
trajectory. Unified Growth Theory (Galor, 2005a, 2008a, 2011, 2012a) is contradicted by data.
Western Europe
Figure 2. No takeoff from stagnation to growth. Economic growth in Western Europe was not stagnant but
hyperbolic and it was remarkably stable. Industrial Revolution, the prime engine of economic growth (Galor,
2005a, p. 212), had no impact on changing the economic growth trajectory where this engine should have
worked most efficiently. Unified Growth Theory (Galor, 2005a, 2008a, 2011, 2012a) is contradicted by data.
97
__________________________________________________________________________________
Eastern Europe
Figure 3. No takeoff from stagnation to growth. Economic growth in Eastern Europe was not stagnant but
2005a, p. 212), had no impact on changing the economic growth trajectory. Unified Growth Theory (Galor,
2005a, 2008a, 2011, 2012a) is contradicted by data.
Former USSR
Figure 4. No takeoff from stagnation to growth. Economic growth in the former USSR was not stagnant but
2005a, p. 212), had no impact on changing the economic growth trajectory. Unified Growth Theory (Galor,
2005a, 2008a, 2011, 2012a) is contradicted by data.
98
__________________________________________________________________________________
Asia
Figure 5. No takeoff from stagnation to growth. Economic growth in Asia (including Japan) was not stagnant
but hyperbolic before the alleged takeoff and it was remarkably stable. The minor boosting after the alleged
takeoff was not a transition from stagnation to growth but a transition from growth to growth. It was similar to
the commonly-observed transitions to slower trajectories but in this case, it was preceded by a minor and
temporary boosting. Unified Growth Theory (Galor, 2005a, 2008a, 2011, 2012a) is contradicted by data.
Africa
Figure 6. No takeoff from stagnation to growth. Economic growth in Africa was not stagnant but hyperbolic.
Unified Growth Theory (Galor, 2005a, 2008a, 2011, 2012a) is contradicted by data. Shortly after the alleged
dramatic but non-existent escape from the postulated Malthusian trap, economic growth in Africa started to be
99
__________________________________________________________________________________
Latin America
Figure 7. No takeoff from stagnation to growth. Economic growth in Latin America was not stagnant but
hyperbolic. At the time of the alleged takeoff, economic growth in Latin America was already following a
slower trajectory. The alleged takeoff is replaced by a slower growth. The spectacular or stunning escapes
from Malthusian trap (Galor, 2005a, pp. 177, 220) never happened because there was no stagnation and no trap.
Unified Growth Theory (Galor, 2005a, 2008a, 2011, 2012a) is contradicted by data.
100
__________________________________________________________________________________
The Postulate of the Three Regimes of Economic Growth

Contradicted by Data
By Ron W. NIELSENa
Abstract. Economic growth in Western Europe, Eastern Europe, Asia, countries of the former USSR, Africa and
Latin America were analysed. It is demonstrated that the fundamental postulate of the Unified Growth Theory
about the existence of the three regimes of growth (Malthusian regime, post-Malthusian regime and sustained-
growth regime) is contradicted by data. These regimes did not exist. In particular, there was no escape from the
Malthusian trap because there was no trap. Economic growth in all these regions was not stagnant but hyperbolic.
Unified Growth Theory is fundamentally incorrect. However, this theory is also dangerously misleading because
it claims a transition from the endless epoch of stagnation to the new era of sustained economic growth, the
interpretation creating the sense of security and a promise of prosperity. The data show that the opposite is true.
Economic growth in the past was sustained and secure. Now, it is supported by the increasing ecological deficit.
The long-term sustained and secure economic growth has yet to be created. It did not happen automatically, as
suggested incorrectly by the Unified Growth Theory.
Keywords. Regional economic growth, Gross Domestic Product, Unified Growth Theory, Malthusian stagnation,
post-Malthusian regime, sustained-growth regime, Industrial Revolution, hyperbolic growth.
JEL. A10, C12, C20, F00, N00, O10.
1. Introduction
T here is no science without data but there is also no science without scientific analysis of data.
We can have excellent data but if we do not analyse them properly we are likely to draw incorrect
conclusions. A perfect example is the Unified Growth Theory (Galor, 2005a, 2011). Excellent
data (Maddison, 2001) were available and even used during its formulation but they were never properly
analysed. Now, it can be easily demonstrated that the fundamental postulates of this theory are
repeatedly contradicted by data, making it fundamentally incorrect and, consequently, unacceptable.
Many attractive theories and explanations can be formulated but if they are not based firmly on the
rigorous analysis of data they are only, at best, just interesting stories. They may contain elements of
truth but folklores of many cultures are full of such stories and they also contain elements of truth.
Fantasy and leaps of faith might be inspiring and productive even in scientific research but they have
to be soon tested by the scientific process of investigation.
However, if one leap of faith is followed by another, if one fantasy creates another, then we no longer
deal with science but with fiction. It is then easy to lose scientific perspective and defend emotionally
the widely-accepted dogmas, based on faith.
Any theory that cannot be checked by data is unacceptable even if it is based on scientifically
attractive ideas. Such a theory has to be put aside until it can be checked by relevant data. Even if a
theory is confirmed by many sets of data it can be still challenged by a single set of contradicting data.
Any theory contradicted by just one set of good data has to be either revised or rejected. Any research,
any intellectual activity, which ignores these fundamental principles of scientific investigation is
unscientific even if it is intellectually stimulating and attractive.
In science, it is important to look for data confirming theoretical explanations but it is even more
important to discover contradicting evidence, because data confirming a theory confirm only what we
already know but contradicting evidence may lead to new discoveries.
a
Australia.
. +61407201175
Published as: Nielsen, R. W. (2016). The postulate of the three regimes of economic growth contradicted by data. Journal
of Economic and Social Thought, 3(1), 1-34.
101
__________________________________________________________________________________
If scientific analysis of data is found to be in agreement with a proposed theory, this theory may then
be considered to be supported by data and its explanations of studied phenomena may then be accepted.
However, if just one set of data is found to be in contradiction with this theory, then this theory can no
longer be accepted in its original form. It has to be then either modified to bring it in agreement with
data, or rejected if such modification is impossible. There is no scientific gain in accepting such a theory.
On the contrary, its continuing acceptance is detrimental to science.
When an incorrect theory is rejected we can then look for a better explanation of studied phenomena.
There are no sentimental values in scientific research and no emotional attachments, and any scientist
should be prepared to have his or her theories challenged by science.
2. Unified Growth Theory

Currently, the most complete theory of the historical economic growth appears to be the Unified
Growth Theory (Galor, 2005a, 2011). It follows closely the traditional interpretations of economic
growth. One of its fundamental postulates is the existence of the three regimes of growth. It claims that
the historical economic growth in various countries and regions can be divided into three distinctly
different regimes of growth governed by distinctly different mechanisms. We shall show that these three
regimes did not exist.
The alleged regimes are:
The regime of Malthusian stagnation. According to Galor, and indeed according to the currently
accepted interpretations, this regime characterized most of human history (Galor, 2005a, p. 178).
Economic growth was allegedly in the endless state of stagnation described as the Malthusian trap or
the Malthusian steady-state equilibrium (e.g. Galor, 2005a. pp. 236, 237, 244). Galor claims that this
epoch of stagnation commenced in 100,000 BC (Galor 2008a, 2012a) and was terminated in around
AD 1750, or around the time of the Industrial Revolution, 1760-1840 (Floud & McCloskey, 1994) in
developed regions and around AD 1900 in less-developed regions.
The beginning of this regime in 100,000 BC is highly speculative because Maddisons data do not
extend to the BC era. Furthermore, the emergence of Homo Sapiens is usually claimed to have been
around 200,000 BC or maybe even earlier (Weaver, Roseman, & Stringer, 2008). We simply do not
know about the economic growth in such a distant past because we do not have relevant data. Judging
by the available evidence (Nielsen, 2016a, 2016b, 2016c), the growth was probably hyperbolic but
whatever we might want to suggest will be based on speculations. However, we do not have to go so
far back in time to test the Unified Growth Theory because the postulate of the existence of the three
regimes of growth cannot be even tested using the economic growth data for the BC era. Even if such
data were available they would be inapplicable for this purpose. We could only test the concept of the
regime of Malthusian stagnation but not the concept of the other two regimes because the existence of
all three regimes of growth is not claimed for the BC era but only for the AD era. All of them were
supposed to have been manifested around the time of the Industrial Revolution. The data we need to
use are the data of Maddison (2001, 2010) because they cover the time when the alleged three regimes
were supposed to have existed.
The post-Malthusian regime. According to Galor (2008a, 2012a), this regime was between AD 1750
and 1870 for developed regions but it commenced a little later, in around AD 1900, for less-developed
regions. Thus, the alleged escape from the Malthusian trap and the commencement of the fast-economic
growth occurred around the onset of the Industrial Revolution for developed regions and a little later
for less-developed regions.
The sustained-growth regime. According to Galor (2008a, 2012a), this regime commenced around
AD 1870 for developed regions.
The general idea of this interpretation of the historical economic growth is that after the endless
epoch of the Malthusian steady-state equilibrium, humans were finally able to break through the
impenetrable barrier of stagnation, escape the Malthusian trap and enter into a new era of sustained and
rapid economic growth. This concept is not only incorrect but also dangerous because the data
describing the historical economic growth (Maddison, 2001, 2010) present a diametrically opposite
interpretation. The economic growth was sustained and secure in the past (Nielsen, 2016a) but now it
entered a stage of the insecure future (Nielsen, 2015a).
102
__________________________________________________________________________________
We shall now demonstrate that Golors concept of the three regimes of growth is contradicted by
the economic growth data (Maddison, 2010). We shall show that his three regimes of growth have no
correlation with data and no positive connection with the real world. Within the range of the
mathematically-analysable data, there was no stagnation and no transition to a fast-economic growth,
described as the sustained-growth regime or the modern-growth regime. We shall show that during this
alleged new, fast-increasing and sustained-growth regime, economic growth started to be diverted from
the fast-increasing historical hyperbolic trajectories to slower trajectories.
Historical economic growth, global and regional, was so well sustained that it followed stable
hyperbolic trajectories. However, such trajectories escape to infinity at a fixed time and any growth,
which follows them, has to be, at a certain stage, diverted to a slower trajectory. Economic growth,
global and regional, is now diverted to slower trajectories. However, the momentum gained during the
sustained historical growth keeps on propelling the economic growth along trajectories, which are still
increasing too fast to feel comfortable about their future.
Galors Unified Growth Theory is not based on the scientific analysis of data. He had access to the
excellent set of data (Maddison, 2001). He even used them, or more correctly he manipulated them to
support his theory, but he did not analyse them. Now, precisely the same data can be used to show that
his theory is fundamentally incorrect.
Regrettably, Unified Growth Theory is based on impressions created by the customary disfigured
presentation of data (Ashraf, 2009; Galor, 2005a, 2005b, 2008a, 2008b, 2008c, 2010, 2011, 2012a,
2012b, 2012c; Galor & Moav, 2002; Snowdon & Galor, 2008). Example of such distorted presentation
of data is shown in Figure 1. Such a way of handling data is a perfect prescription for drawing incorrect
conclusions. (All figures are presented in the Appendix.)
In science, data are treated with respect because the primary aim of science is to discover the truth,
and for this purpose there is nothing as reliable as good sets of data. Not all data can be accepted but
we have to have good reasons for rejecting data. If reasons for rejecting data are unacceptable, then
reasons for rejecting data have to be rejected.
Many attractive theories and explanations may be formulated but they all have to pass the test of
data. Without such a test, they are just stories, which might or might not be true.
Galors predecessors might be excused for believing in the existence of Malthusian stagnation and
in the dramatic impact of the Industrial Revolution on changing the economic growth trajectories
because they were using strongly limited information. They had no access to the excellent source of
data published by the world-renown economist (Maddison, 2001). Galor not only had access to these
data but he also used them repeatedly during the formulation of his theory but unfortunately, he distorted
them so much that they were creating an impression of being in agreement with his postulates.
In our discussion, we shall use the latest data describing economic growth (Maddison, 2010). This
publication contains some additional information but any of Maddisons compilations, the compilation
used by Galor or this new compilation, can be used to demonstrate that the Unified Growth Theory is
contradicted by data. The advantage of using the new compilation (Maddison, 2010) is that it helps to
understand the recent transitions to slower trajectories because the earlier compilation did not include
the data for the 21st century.
3. Method of analysis and related issues

We shall use two ways of displaying data: (1) semilogarithmic display of the GDP data and (2) the
display of their reciprocal values, 1/GDP. These two types of display are suitable for studying data
varying over a large range of values. The GDP values will be expressed in billions of 1990 International
Geary-Khamis dollars.
Hyperbolic distributions, which describe the historical economic growth (Nielsen, 2016a), are
represented by the simple mathematical formula:
S (t=) (a kt )1 (1)
where, in our case, S (t ) is the GDP while a and k are positive constants.
The reciprocal values of hyperbolic distributions are represented by straight lines:
1
= a kt (2)
S (t )
103
__________________________________________________________________________________
In general, hyperbolic growth can be uniquely identified by the decreasing straight line of the
reciprocal values of the size of the growing entity in much the same way as the exponential growth can
be identified by their logarithm. Reciprocal values of data can also help in identifying easily any
deviations from hyperbolic trend because deviations from a straight line are easy to notice.
In using the reciprocal values, it should be remembered that a deviation to a slower trajectory is
indicated by an upward bending away from the previous linear trend while deviations to faster
trajectories are indicated by downward bending. In particular, any form of boosting or takeoff,
repeatedly claimed by Galor for global and regional economic growth, should be indicted by a clear
change in the downward direction of the reciprocal values.
If the straight line fitting the reciprocal values of data remains undisturbed, it shows that there was
no diversion to a faster or slower trajectory. In particular, if the straight line does not show a change in
the downward direction (if the gradient of the trajectory of the reciprocal values remains constant) then
there was no boosting in the economic growth. We obviously cannot claim a change of direction on an
undisturbed straight line.
If the reciprocal values of data follow a decreasing straight line, the growth is not stagnant but
hyperbolic. However, the concept of stagnation is not supported even if the reciprocal values of data do
not decrease linearly. Any monotonically-decreasing trajectory will show that the postulate of
stagnation followed by a takeoff at a certain time is not supported by data.
To prove the existence of the epoch of stagnation it is necessary to prove the presence of random
fluctuations often described as Malthusian oscillations. Such random fluctuations should be clearly seen
not only in the direct display of data but also in the display of their reciprocal values. It they are absent
then there is no support in data for claiming the existence of the epoch of stagnation. Furthermore, if
data do not show a clear takeoff from stagnation to growth at the postulated time, then there is no
support for Galors repeatedly-claimed takeoffs. However, if the reciprocal values of data follow a
decreasing straight line, then they show, or at least strongly suggest, that the growth was hyperbolic.
If the straight line representing the reciprocal values of data remains unchanged, then obviously,
there is no change in the mechanism of growth. It is impossible to divide a straight line into different
sections and claim different mechanism of growth for each of such arbitrarily selected section. It is
impossible to claim, for instance, a transition from stagnation to growth as repeatedly claimed by Galor
in his Unified Growth Theory if the reciprocal values of data follow an undisturbed straight line. It is
impossible to claim the existence differential takeoffs if there were no takeoffs. It is also impossible to
claim that the Industrial Revolution changed the economic growth trajectory if the reciprocal values of
data demonstrate that there was no change, i.e. that their linear trend remained undisturbed.
No-one has yet demonstrated the existence of Malthusian stagnation in the economic growth or in
the growth of human population. For instance, Lee pointed out that these models of Malthusian
oscillations are speculative when applied to the growth of human population (Lee, 1997, p. 1097).
However, from the descriptions of Malthusian stagnation, its signature and the alleged escape from the
Malthusian trap should be easy to identify. This signature is schematically presented in Figures 2 and
3.
For the direct display of GDP data (Figure 2), the signature of the regime of Malthusian stagnation
can be identified by random fluctuations or oscillations around an approximately horizontal line. Over
much longer sections of time, perhaps extending over thousands of years, fluctuations around the
horizontal line might be replaced by fluctuations around a certain irregular trajectory (increasing,
decreasing or randomly oscillating), which would be probably difficult to describe mathematically
because the general concept of Malthusian stagnation is that it was controlled by random forces. Such
random forces are hardly expected to generate monotonically-increasing distributions (Artzrouni &
Komlos, 1985; Lagerlf, 2006; McKeown, 2009; Komlos, 1989; van de Kaa, 2008). For the
monotonically-increasing distributions, random forces are either too weak or they average out (Kapitza,
2006) and the growth is controlled by a certain dominant force, which could be constant (for the
exponential growth), increasing with time or with the size of the growing entity (as for the hyperbolic
growth) or even decreasing (as for the logistic growth).
The signature of the remarkable or stunning escape from the Malthusian trap (Galor, 2005a, pp.
177, 220) to the sustained economic growth should be easily identified by a clear takeoff from the earlier
stagnant distribution to a fast-increasing growth. The alleged escape should occur around AD 1750 for
developed regions and around AD 1900 for less-developed regions (Galor, 2008a, 2012a).
104
__________________________________________________________________________________
For the reciprocal values of data (Figure 3), the epoch of Malthusian stagnation can be again
identified by random fluctuations around an approximately horizontal line or around an irregular
trajectory but the escape from the Malthusian trap will be identified by a clear downward trend. It should
be noted that in the display of the reciprocal values of GDP data, small fluctuations are magnified,
which means that in this display, epoch of Malthusian stagnation should be easy to identify because it
should be characterised by strong fluctuations.
Maddisons data are indispensable in studying the historical economic growth but they have a
strongly-limited range because they contain a large gap between AD 1 and 1000, and between AD 1000
and 1500. The most useful sets of data are from AD 1500. However, this shortcoming is immaterial
because all the action described by Galors three regimes of growth takes place after AD 1500. Within
the range of the good sets of data, i.e. commencing from AD 1500, we should see clearly all the
hallmarks of Galors postulate of the three regimes of growth. We should see the signature of the regime
of Malthusian stagnation, the effects of the Industrial Revolution, which was supposed to have been
the prime engine of economic growth (Galor, 2005a, p. 212), the signature of the escape from the
alleged Malthusian trap and a clear evidence of the uninterrupted era of the fast-increasing and sustained
economic growth after stagnation. All these features should be clearly displayed. If they are not, then
there is no support in the data for Galors interpretations of the historical economic growth based on
such distorted presentations of data as shown in Figure 1. Such presentations have no place in the
scientific research.
The discussion presented here is the extension of the mathematical analysis of the historical
economic growth (Nielsen, 2016a). We have already demonstrated that the historical economic growth
was hyperbolic and thus that implicitly it gives no support for the doctrine of the three regimes of
growth. Now, we shall show it explicitly.
It is essential to understand the fundamental features of hyperbolic distributions (Nielsen, 2014).
Hyperbolic growth is slow over a long time and fast over a short time, but it is still the same,
monotonically-increasing distribution, which is impossible to divide into two or three different,
mathematically-justified components. The easiest way to see it, is by using the reciprocal values [see
the eqn (2)] because the confusing hyperbolic growth is then represented by a decreasing straight line.
It is then clear that it is impossible to divide such a straight line into distinctly different, mathematically-
justified components and claim distinctly different mechanisms of growth for each of these arbitrarily
selected components.
Even though hyperbolic growth is slow over a long time it is not stagnant. Slow hyperbolic growth
should never be interpreted as stagnant because if we want to interpret the slow perceived part of
hyperbolic growth as stagnant, and governed by the usually assumed multitude of random forces, we
should use precisely the same mechanism to explain the perceived fast growth. The perceived slow and
fast components belong to the same, monotonically-increasing distribution. It is impossible to divide a
monotonically-increasing hyperbolic distribution into the mathematically-justifiable slow and fast
sections because it is obviously impossible to divide a straight line describing the reciprocal values and
representing the hyperbolic distribution into distinctly-different and mathematically-justifiable sections
(Nielsen, 2014). It is scientifically unjustifiable to use different mechanisms of growth for such
arbitrarily selected sections. Hyperbolic distributions have to be interpreted as a whole and the same
mechanism has to be applied to the apparent slow growth and to the apparent fast growth. There is no
clearly defined transition between the apparent slow and the apparent fast growth.
These comments apply also to the income per capita distributions represented by the Gross Domestic
Product per capita (GDP/cap). Such distributions are even more confusing than hyperbolic distributions.
They are linearly-modulated hyperbolic distributions, i.e. the monotonically-increasing hyperbolic
distributions representing the growth of the GDP modulated by the monotonically-decreasing linear
distributions representing the reciprocal values of the size of the population (Nielsen, 2015b). A product
or a ratio of monotonic distributions cannot generate a non-monotonic distribution.
Even though the GDP/cap distributions appear to be made of two or maybe even three different
components, as claimed incorrectly by Galor, they are increasing monotonically and it is impossible to
divide them into distinctly different, mathematically-justifiable components. We can demonstrate it by
calculating gradients or the growth rates of the GDP/cap distributions and by showing that they increase
monotonically (Nielsen, 2015b). Any attempt to divide the GDP/cap distributions into distinctly-
different components is strongly subjective and mathematically unjustified.
105
__________________________________________________________________________________
4. Analysis of data for Western Europe

We shall analyse two sets of data for Western Europe: (1) the data for 12 selected countries and the
data for the total of 30 countries. The 12 selected countries are made of Austria, Belgium, Denmark,
Finland, France, Germany, Italy, the Netherlands, Norway, Sweden, Switzerland and the United
Kingdom. According to Maddison (2010), in 2008, these 12 countries accounted for 85% of the total
GDP of the 30 countries of Western Europe. The total of the 30 countries includes also Ireland, Greece,
Portugal, Spain and 14 other small west European countries.
The reason for analysing these two groups separately is that the listed 12 countries represent the
most advanced economies, where the effects of the Industrial Revolution and the escape from the
Malthusian trap should be most clearly visible. Consequently, for these 12 countries we should expect
the best agreement between the Unified Growth Theory and the data.
Economic growth between AD 1 and 2008 in the 12 countries of Western Europe is shown in Figures
4 and 5. The growth in the total of 30 countries is shown in Figures 6 and 7.
Hyperbolic parameters describing economic growth in the 12 countries of Western Europe are:
= a 1.147 101 and
= k 5.961 105 . The corresponding singularity is in 1923 but the economic growth
was diverted to a slower trajectory around 1900, bypassing the singularity by about 23 years.
Hyperbolic fit to the data is remarkably good between AD 1500 and 1900 and acceptable below AD
1500. The point at AD 1 is only 27% higher than the fitted distribution and the point at AD 1000 is 54%
lower. The critical range of time for testing the Unified Growth Theory is from AD 1500. It is in this
range of time that we should be able to see transition from stagnation to growth and later a transition to
the alleged sustained growth regime.
The data presented in Figures 4 and 5 clearly demonstrate that there is no support for the existence
of the alleged regime of Malthusian stagnation. However, there is a convincing support for the
hyperbolic growth at least between AD 1500 and 1900, the range of time where the signature of
Malthusian stagnation should be still clearly displayed for about 300 years. The data show that during
that time economic growth was following a steadily-increasing hyperbolic trajectory. There is no sign
of the existence of Malthusian stagnation.
Absolutely nothing had happened at the end of the alleged Malthusian regime. There was no
transition from stagnation to growth at any time. Around the time of the beginning of the postulated
regime of sustained-growth, when the economic growth was supposed to have been launched from
stagnation to a fast-increasing trajectory, the growth started to be diverted to a slower trajectory.
It is remarkable also that the Industrial Revolution had absolutely no impact on shaping the economic
growth trajectory in these 12 countries. They should experience the greatest benefits of this revolution
and they probably did but these benefits did not boost the economic growth. Technological innovations
were used in the economic growth but they did not change in the slightest the economic growth
trajectory. In countries, where effects of the Industrial Revolution, the prime engine of economic
growth (Galor, 2005a, p. 212), should have been most clearly reflected in the relevant data, we see no
impacts of this engine. There must have been some other force, which was much stronger than any
forces associated with the Industrial Revolution.
This is an interesting issue, which should be studied and explained but it is futile to look for its
explanation in the Unified Growth Theory. This interesting feature has not been even noticed by Galor,
which is hardly surprising because it is hard or even impossible to carry out scientific research and draw
reliable and scientifically-justified conclusions by repeatedly distorting data in such a way as shown in
Figure 1.
Galors Unified Growth Theory has no relevance to the description, let alone to the explanation of
the mechanism of the economic growth, even in countries where his theory should be best fitted. Here,
in the leading countries of Western Europe, where the effects of the Industrial Revolution should be
most prominently displayed in the data describing economic growth, where the remarkable and
stunning escape from the Malthusian trap (Galor, 2005a, pp. 177, 220) should be remarkably obvious,
there are no signs of the impacts of the Industrial Revolution on the economic growth and no signs of
any escape from the Malthusian trap, remarkable or less-remarkable, because there was no trap.
Economic growth was increasing undisturbed and unconstrained along a hyperbolic trajectory until
around 1900 when it started to be diverted to a slower but still fast-increasing trajectory.
106
__________________________________________________________________________________
Galors three regimes of growth are totally dissociated from reality. They describe events that never
happened.
Stories and explanations presented by Galor in his theory have no relevance to the explanation of
the mechanism of the economic growth even in these 12 leading countries of Western Europe. His
stories might be explaining or describing something else, e.g. social conditions or the style of living,
but even then, one wonders about the degree of reliability of such descriptions. His narrative does not
explain the mechanism of the economic growth.
Results of the analysis of the economic growth in the total of 30 countries of Western Europe are
presented in Figures 6 and 7. Hyperbolic parameters are: = a 9.859 102 and
= k 5.112 105 . The
corresponding singularity is in 1929 but the economic growth was diverted to a slower trajectory around
1900, bypassing the singularity by about 29 years. The point at AD 1 is 42% higher than the calculated
hyperbolic distribution and at AD 1000 it is 48% lower.
The analysis of economic growth in the total of 30 countries of Western Europe leads to the same
conclusions as for the 12 leading countries: Unified Growth Theory is contradicted by the economic
growth data in Western Europe where the effects discussed by Galor should have been most
convincingly confirmed. In contrast, they are convincingly contradicted.
5. Analysis of data for Eastern Europe

Results of the analysis of economic growth in Eastern Europe, based on using Maddisons data
(Maddison, 2010), are presented in Figures 8 and 9. Hyperbolic parameters fitting the data are:
= a 7.749 101 and = k 4.048 104 . The point at AD 1 is 51% higher than the calculated curve. The
singularity is in 1915 but the economic growth was diverted to a slower trajectory around 1890,
bypassing the singularity by 25 years.
Unified Growth Theory is clearly contradicted by the economic growth data for Eastern Europe. The
epoch of Malthusian stagnation did not exist within the range of the mathematically-analysable data.
Outside of this range, any claim about the existence of the regime of Malthusian stagnation and about
its effects on the economic growth has to be based on questionable conjectures. Such a claim would be
also in conflict with the analysable data.
The data show no transition from stagnation to growth at any time because the growth was
hyperbolic. There was no remarkable or stunning escape from the Malthusian trap (Galor, 2005a,
pp. 177, 220) because there was no trap. Industrial Revolution did not boost the economic growth in
Eastern Europe.
There was also no boosting of the economic growth at the time of the transition from the alleged
post-Malthusian regime to the alleged sustained growth regime. Soon after the commencement of this
phantom sustained-growth regime, economic growth in Eastern Europe started to be diverted to a slower
trajectory. Galors regimes of growth are clearly dissociated from reality. They do not describe the real
world but the world of fiction created by preconceived ideas and supported by the habitually-distorted
presentation of data (Ashraf, 2009; Galor, 2005a, 2005b, 2008a, 2008b, 2008c, 2010, 2011, 2012a,
2012b, 2012c; Galor & Moav, 2002; Snowdon & Galor, 2008).
6. Analysis of data for Asia

Asia (excluding Japan) is made primarily, if not exclusively, of less-developed countries (BBC,
2014; Pereira, 2011). According to Galor, this region should have experienced the epoch of stagnation
until around 1900 followed by the post-Malthusian regime commencing around that year. If Galors
claims are correct, we should see clear signs of stagnation in the data until around 1900 and a clear
transition (a dramatic takeoff) from stagnation to growth around that year.
Economic growth in Asia between AD 1 and 2008 is presented in Figure 10. There is absolutely no
correlation between the data and the three key events indicated in this figure: The Industrial Revolution,
the alleged Malthusian regime and the alleged post-Malthusian regime, which were supposed to have
been shaping the economic growth.
During the alleged Malthusian regime of stagnation, economic growth in Asia was increasing
hyperbolically at least from AD 1000 but the point at AD 1 is also not far from the calculated hyperbolic
=
distribution. Parameters fitting the data are a 2.493 103 and
= k 1.238 105 .
107
__________________________________________________________________________________
The data show no signs of stagnation within their mathematically-analysable range, no signs of the
Malthusian steady-state equilibrium and no signs of Malthusian oscillations. Assuming the existence of
all such features is not only unnecessary but also scientifically unjustified because in science
complicated interpretations are rejected in favour of simpler explanations. The data follow a steadily-
increasing hyperbolic distribution, suggesting a simple mechanism of growth because hyperbolic
distributions are described by a simple mathematical formula [see the eqn (1)].
The concept of stagnation is dramatically contradicted by data and so is the transition to the alleged
post-Malthusian regime, which was supposed to have been a transition from stagnation to growth. We
see no such transition but a continuation of the hyperbolic growth. The claimed by Galor takeoff did
not happen. There was a minor and hard-to-notice disturbance in the economic growth around 1950 but
the growth soon returned to its historical hyperbolic trajectory. The overall evidence in the data is that
the prop-up structures (the alleged different regimes of growth) used by Galor are not only totally
redundant but also strongly misleading. They can, and even should, be removed because the data reveal
a totally different pattern of growth.
The data and their analysis show that nothing dramatic occurred during the alleged transition from
the postulated Malthusian regime of stagnation to the alleged post-Malthusian regime, which is
supposed to mark the escape from the postulated Malthusian trap and leading to a sustained growth
regime. There was no escape from the trap because there was no trap. During the postulated Malthusian
trap, the economic growth was steadily increasing and it was obviously unconstrained. It is futile to
claim random fluctuations and oscillations when there are none. Why should we even contemplate to
make it all more complicated when the data show that the growth was much simpler?
If not for Maddison and his data, the established knowledge in the economic research would have
remained established, assuming that the knowledge based on conjunctions and dubious assumptions
could be regarded as established, but now it has to be revaluated and changed. However, new insights
should be welcome, particularly if they suggest a simpler explanation of the historical economic growth.
Reciprocal values of the GDP data, 1/GDP, shown in Figure 11, also demonstrate that the Unified
Growth Theory is contradicted by the same data, which were used during its development, the data
published by Maddison in 2001 (Maddison, 2001) but later extended to include economic growth during
the 21st century (Maddison, 2010).
During the alleged Malthusian regime of stagnation, reciprocal values of data were decreasing along
a straight line indicating an undisturbed, hyperbolic economic growth. The data show also that nothing
dramatic had happened at the end of this alleged epoch of stagnation. There was no transition to a new
regime of growth. In particular, there was no transition from stagnation to growth, as claimed by Galor,
but a continuation of the hyperbolic growth. The concept of different regimes of growth proposed by
Galor is convincingly contradicted by data.
7. Analysis of data for the former USSR

Economic growth in countries of the former USSR between AD 1 and 2008 is presented in Figure
12. Reciprocal values of the GDP data, 1/GDP, are shown in Figures 13 and 14. The growth was
hyperbolic between AD 1 and around 1870. Parameters describing hyperbolic growth are
= a 6.547 101 and= k 3.452 104 .
During the entire range of the mathematically-analysable data the epoch of Malthusian stagnation
did not exist. Galors regimes of growth are hanging there without having any connection with data.
The remarkable or stunning escape from the Malthusian trap did not happen because there was no
trap. Galors Malthusian regime ends in the middle of nowhere. Absolutely nothing (remarkable or less-
remarkable, stunning or less stunning) happened on the border between the alleged Malthusian regime
and the post-Malthusian regime. There was also no stunning or remarkable escape from the alleged
Malthusian trap at the onset of the alleged sustained-growth regime because there was obviously no
trap but the steadily-increasing hyperbolic growth from as early as AD 1. There was no dramatic
increase in the economic growth. On the contrary, around the time of the alleged but non-existent
transition to the so-called regime of sustained growth, economic growth started to be diverted to a
slower trajectory.
What is remarkable about this confrontation of Galors theory with the empirical evidence is that
there is such a consistently repeated and stunning disagreement between his theory and the data. The
108
__________________________________________________________________________________
data also demonstrate that the Industrial Revolution had absolutely no impact on changing the economic
growth trajectory in the countries of former USSR. Here again we see that the prime engine of
economic growth (Galor, 2005a, p. 212) did nothing to change to growth trajectory. Whatever this
engine might have been doing, it certainly did not boost the economic growth. The data and their
analysis give no support for the concept of Malthusian stagnation and for the assumption of the
existence of the steady-state Malthusian equilibrium. Even more strongly, the data are in clear
contradiction of these concepts. Economic growth was increasing along a remarkably-stable hyperbolic
trajectory. There was no escape from the Malthusian trap, let alone a remarkable or stunning escape
as claimed by Galor (2005a, pp. 177, 220), because there was no trap. The growth was unconstrained
because the hyperbolic trajectory remained undisturbed.
The concept of stagnation is dramatically contradicted by data and so is the alleged transition from
stagnation to growth. Such a transition never happened. On the contrary, from around 1870, economic
growth in the countries of former USSR started to be diverted to a slower trajectory, away from its
faster, historical hyperbolic trajectory.
8. Analysis of data for Africa

Africa is a perfect example of a cluster of countries, which belong to the group of less-developed
and least-developed countries. Out of the total of 48 least-developed countries in the world, 34 are in
Africa (Bangla News, 2015; UNCTAD, 2013). With just one minor exception, Africa is made entirely
of less-developed and least-developed countries (BBC, 2014; Pereira, 2011). The exception is Western
Sahara, a small country in transition made of around 586,000 people (UNDATA, 2015).
Maddisons data for Africa serve, therefore, as an excellent source of information to test Galors
hypothesis of the existence of the distinctly different regimes of economic growth in less-developed
regions. We shall demonstrate that this hypothesis is dramatically and clearly contradicted by data.
Reciprocal values of data describing economic-growth in Africa are presented in Figure 15.
Economic growth was clearly hyperbolic between AD 1 and around 1820 because the reciprocal values
follow a straight line. There was definitely no stagnation. The concept of the regime of Malthusian
stagnation is clearly contradicted by data. To prove its existence, one would have to demonstrate a
stagnant state of growth characterised by random Malthusian oscillation around an approximately
horizontal line as shown in Figure 3. The data contain no such signature. On the contrary they show a
steadily-increasing and remarkably-stable hyperbolic growth. There are no signs of any possible
fluctuations, which in this representation of data should be strongly magnified.
Furthermore, Galors concept of Malthusian stagnation extending to 1900 ignores not only the data
between AD 1 and 1820 but also the clear and dramatic transition, which occurred around 1820. It was
not a transition from stagnation to growth but from growth to growth, the transition from a slower but
steadily-increasing hyperbolic growth to a faster and steadily-increasing hyperbolic growth. This
pattern is in clear contradiction of the Unified Growth Theory (Galor, 2005a, 2008a, 2011, 2012a).
The concept of the regime of stagnation ignores the steadily-increasing economic growth before
1820, the dramatic change in the pattern of growth around that year and the new hyperbolic growth
after 1820. The claim of Malthusian stagnation ending in 1900 for less-developed countries ignores also
that absolutely nothing unusual had happened around that year. The economic growth continued
undisturbed. The postulated Malthusian regime ends in the middle of nowhere. There is no justification
for claiming the regime of Malthusian stagnation and no justification for terminating it in AD 1900 or
at any other time because there was no stagnation.
In addition, the data demonstrate the existence of a feature, which is ignored by Galor: the diversion
to a slower trajectory around 1950 indicated by the upward bending of the trajectory of the reciprocal
values. It was impossible for Galor to notice any interesting features in data because he persistently and
consistently was distorting the representation of data in the manner illustrated in Figure 1. Such
unfortunate manipulation of data was doomed to lead to incorrect conclusions. According to Galor, the
economic growth was supposed to have been boosted from stagnation to growth (at the end of his
alleged Malthusian regime) and launched into a fast-increasing growth, but data present an entirely
different interpretation: economic growth was increasing fast along a hyperbolic trajectory during the
alleged regime of Malthusian stagnation but shortly after the time of the postulated transition to a faster
growth the data it started to follow a slower trajectory. Data used by Galor tell one story, while Galor
109
__________________________________________________________________________________
tells another, but in science data have the priority and it does not matter how much desirable it is for the
researcher to have his or her cherished doctrines supported by data.
The disagreement between Galors theory and the data is also clearly demonstrated in Figures 17
and 18. Over the range of the mathematically-analysable data the Malthusian regime did not exist. The
data show no evidence of the features characterising the epoch of Malthusian stagnation. In contrast,
the data show steadily-increasing hyperbolic distributions.
In his description of economic growth, Galor did not even notice that there was a strong transition
around AD 1820, let alone that it was a transition from one hyperbolic distribution to another. He also
did not notice that that the postulated epoch of Malthusian stagnation ends in the middle of nowhere
(see Figure 18).
Many important details are easily lost in the habitually distorted presentations of data (Ashraf, 2009;
Galor, 2005a, 2005b, 2008a, 2008b, 2008c, 2010, 2011, 2012a, 2012b, 2012c; Galor & Moav, 2002;
Snowdon & Galor, 2008) as illustrated in Figure 1. It is hard or even impossible to draw reliable
conclusions by using such distorted diagrams and by making no attempt to analyse data. Conclusions
based on impressions are likely to be incorrect. It is hard or even impossible to do science without
following the principles of scientific investigation.
9. Analysis of data for Latin America

Results of analysis of the economic growth in Latin America based on Maddisons data (Maddison,
2010) are shown in Figures 19 and 20.
Data suggest the existence of two hyperbolic growth trajectories: a slow trajectory between AD 1
and 1500 and a fast trajectory between AD 1600 and 1870. The slow trajectory is characterised by
parameters= a 4.421 101 and = k 2.093 104 . The singularity for this trajectory was at t = 2113 .
The fast trajectory is characterised by parameters = a 1.570 100 and
= k 8.224 104 . The singularity
for this new trajectory was at t = 1910 . However, from around 1870, i.e. from around the time of the
alleged takeoff from stagnation to growth (Galor, 2008a, 2012a), economic growth in Latin America
started to be diverted to a slower trajectory bypassing the singularity by a safe margin of 40 years. The
illusion of a takeoff is replaced by a diversion to a slower growth.
The characteristic features of the economic growth in Latin America are similar to the features in
Africa. In both cases, a slow hyperbolic growth was followed by a much faster hyperbolic trajectory
and this transition can be correlated with the intensified colonisation of Latin America (Bethell, 1984).
The data for Latin America are in clear disagreement with the Unified Growth Theory. The economic
growth was slow before AD 1500 but there is no basis for claiming that it was stagnant. Hyperbolic
trajectory between AD 1 and 1500 could be questioned but it is consistent with the similar, but much
clearer, pattern in Africa and is in perfect agreement with the repeated evidence of hyperbolic growth in
other regions. There is definitely no convincing support for the existence of the epoch of stagnation.
The data show a brief economic decline between AD 1500 and 1600, which appears to be coinciding
with the commencement of the intensified Spanish conquest (Bethell, 1984). However, from around AD
1600, economic growth in Latin America was following a fast-increasing hyperbolic trajectory. The
change from a slow to fast economic growth occurred about 300 years before the alleged takeoff around
1900. Furthermore, as in Africa, it was not a transition from stagnation to growth but from hyperbolic
growth to hyperbolic growth. This feature is ignored in the Unified Growth Theory. It is not even noticed
by Galor because how is it possible to notice any useful patters if data are so eagerly distorted as shown
in Figure 1? Remarkably also, at the time of the alleged but non-existent remarkable escape from the
Malthusian trap (Galor, 2005a, p. 177) in around AD 1900, economic growth in Latin America was
already diverted to a slower trajectory.
Unified Growth Theory presents a story, which is contradicted by data. There is no correlation
between the data and the narrative of this theory. In his habitually crude display of data, Galor could not
have seen all these important features. He appears to have been guided by the inherited ideas, which
unfortunately he did not check by the rigorous analysis of the new and excellent data (Maddison, 2001)
available to him and even used by him at the time of the formulation of his theory. The updated
compilation of the data describing the historical economic growth (Maddison, 2010) was also available
to him even before the publication of his book (Galor, 2011) and certainly during his continuing
disseminations of the same erroneous ideas after its publication. As mentioned earlier, any of these
110
__________________________________________________________________________________
compilations can be used to show that his theory is fundamentally incorrect because during the time
when there were supposed to have been transitions between these alleged regimes of growth each of
these two compilation shows a clear disagreement with Galors theory.

We have analysed economic growth in Western Europe, Eastern Europe, Asia, former USSR, Africa
and Latin America (Maddison, 2010). We have found that the fundamental concepts of the Unified
Growth Theory (Galor, 2005a, 2011) are contradicted by the same data, which were used by Galor but
never analysed during the formulation of his theory.
The wished-for events did not happen. The real world refused to comply with the preconceived ideas
and with the imagined interpretations, which were creating such an attractive story.
It seems to be obvious that the Industrial Revolution should have a strong and decisive effect on the
economic growth but it did not. It seems to be obvious that a slow growth is stagnant but it is not. What
seems to be obvious is not necessarily true. It is obvious that the Sun moves around the Earth but it does
not. It is clear that the earth does not move, and that it does not lie elsewhere than at the centre
(Aristotle). Even the greatest thinkers can make a mistake.
Empirical evidence has to be methodically and carefully analysed; otherwise we shall be creating
our own stories, which might be interesting, exciting and convincing but they will be just stories of
fiction. They will have nothing to do with science. In science, we learn from nature. Any attempt to
mould nature into the image fashioned by our creative imagination is bound to fail and the perfect
example is the Unified Growth Theory. Creative imagination is important in science but it has to be
checked, controlled and moderated by data.
Within the range of the mathematically-analysable data, the three regimes of growth, the Malthusian
regime, the post-Malthusian regime and the sustained-growth regime did not exist. There is no
correlation between the data and these three postulated regimes of growth. In particular, there was no
escape from the Malthusian trap because there was no trap.
During the time described by the mathematically-analysable data, economic growth was hyperbolic
and generally undisturbed. Only most recently, around the time when according to the Unified Growth
Theory it should have been boosted from stagnation to growth, economic growth started to be diverted
from the fast-increasing hyperbolic trajectories to slower trajectories. Unified Growth Theory does not
explain, let alone describe the historical economic growth because it is based on the fundamentally
incorrect premises and on the unscientific methods of investigation.
The concept of the three regimes of growth was supported by the distorted presentation of data
(Ashraf, 2009; Galor, 2005a, 2005b, 2008a, 2008b, 2008c, 2010, 2011, 2012a, 2012b, 2012c; Galor &
Moav, 2002; Snowdon & Galor, 2008). When properly displayed and analysed, the same data show that
the Unified Growth Theory is fundamentally incorrect.
The reliable and correct interpretation of the historical economic growth might appear to have no
practical application because what was in the past is in the past. Why should the distant past have any
influence on our present economic growth? However, the correct understanding of the past economic
growth may well decide about our future.
Galors interpretations of the historical economic growth are not only scientifically unacceptable but
also dangerously incorrect because they create the false sense of security. They present a picture of the
unsustained economic growth in the past and of a transition to a new era of sustained economy after the
remarkable or stunning escape from the Malthusian trap (Galor, 2005a, pp. 177, 220). At last, after
the endless suffering, straggle, and deprivation, humans escaped the tyranny of the Malthusian regime
and now they can enjoy the sustained economic growth with its prosperous future. This is a pleasing
story but the opposite is true.
Rigorous analysis of data shows convincingly that the past economic growth was sustained and
secure because it followed the remarkably stable hyperbolic trajectories (Nielsen, 2016a). This
conclusion is in harmony with the study of ecological footprints, which shows that until the late 1900s
global ecological footprint was lower than the ecological capacity (WWF, 2010). It was in the past that
the economic growth was not only sustained but also sustainable. Now it is not, because it is supported
by the increasing ecological deficit. Indeed, mathematical analysis of the economic growth shows that
its future is insecure (Nielsen, 2015a).
111
__________________________________________________________________________________
Economic growth was not in a trap in the past but now it is in a trap of our continuing drive to
increase not only the GDP but also the GDP/cap. We seem to see no limit to prosperity but the limit is
imposed by the ecological limits and by the fast-increasing trajectories of economic growth. While the
Unified Growth Theory suggests a prosperous future based on the assumption of the existence of the
sustained growth regime after the alleged Malthusian regime, the data indicate that unless we take
decisive steps to control the current economic growth our future is insecure (Nielsen, 2015a).
In its present form, Galors Unified Growth Theory is unacceptable. It has to be either thoroughly
revised or rejected and replaced by a new theory aimed at explaining why the economic growth was
hyperbolic in the past, why it was increasing along such remarkably stable trajectories, why it started
to be diverted to slower, but still fast-increasing, trajectories and, most importantly, how to create a
sustainable economic future.
Propelled by the gained momentum of the historical economic growth, the current growth continues
to increase too fast. It has to be slowed down. The sustainable and secure economic growth has yet to
be created. It has not been created automatically at the end of the alleged but non-existent Malthusian
regime as suggested incorrectly by the Unified Growth Theory.
References
Artzrouni, M., & Komlos, J. (1985). Population Growth through History and the Escape from the Malthusian Trap: A
Homeostatic Simulation Model. Genus, 41(3-4), 21-39.
Bangla News, (2015). Bangladesh Chairmen of Least Developed Countries.
http://www.en.banglanews24.com/fullnews/bn/115940.html
Bethell, L. (1984). The Cambridge History of Latin America: Vol. I and II, Colonial Latin America. Cambridge, UK:
Cambridge University Press.
Floud, D. & McCloskey, D.N. (1994). The Economic History of Britain since 1700. Cambridge: Cambridge University
Press.
European Economic Association, 3(3), 494-504. doi. 10.1162/jeea.2005.3.2-3.494
Galor, O. (2008b). Economic Growth in the Very Long Run. In Durlauf, S.N. and Blume, L.E., Eds., The New Palgrave
Growth Theory. International Economic Review, 51(1), 1-44. doi. 10.1111/j.1468-2354.2009.00569.x
Galor, O. (2012b). The Demographic Transition: Causes and Consequences. Cliometrica, 6(1), 1-28. doi. 10.1007/s11698-
011-0062-7
117, 1133-1191. doi. 10.1162/003355302320935007
Komlos, J. H. (1989). Thinking about Industrial Revolution. Journal of European Economic History,18, 191-206.
Lagerlf, N-P. (2006). The Galor-Weil Model Revisited: A Quantitative Exercise. Review of Economic Dynamics, 9(1), 116-
142.doi. 10.1016/j.red.2005.07.002
Lee, R. D. (1997). Population dynamics: equilibrium, disequilibrium, and consequences of fluctuations. In O. Stark & M.
McKeown, T. (2009). The Epidemiologic Transition: Changing Patterns of Mortality and Population Dynamics. American
Journal of Lifestyle Medicine, 3(1), 19S-26S.doi. 10.1177/1559827609335350
Nielsen, R. W. (2014). Changing the Paradigm. Applied Mathematics, 5, 1950-1963.doi. 10.4236/am.2014.513188
Nielsen, R. W. (2015a). The Insecure Future of the World Economic Growth. Journal of Economic and Social Thought,
2(4), 242-255.
112
__________________________________________________________________________________
Nielsen, R. W. (2015b). Unified Growth Theory Contradicted by the GDP/cap Data.

Nielsen, R. W. (2016a). Mathematical Analysis of the Historical Economic Growth with a Search for Takeoffs from
Stagnation to Growth. Journal of Economic Library, 3(1), 1-23.
Nielsen, R. W. (2016b). Unified Growth Theory Contradicted by the Mathematical Analysis of the Historical Growth of
Human Population. http://arxiv.org/ftp/arxiv/papers/1601/1601.07291.pdf
Nielsen, R. W. (2016c). Growth of the World Population in the Past 12,000 Years and Its Link to the Economic Growth.
UNCTAD, (2013). Map of the LDCs. http://unctad.org/en/Pages/ALDC/Least%20Developed%20Countries/LDC-Map.aspx
UNDATA, (2015), Western Sahara.http://data.un.org/CountryProfile.aspx?crName=Western%20Sahara
van de Kaa, D. J. (2008). Demographic Transition. In Y. Zeng (Ed.). Demography. Encyclopedia of Life Support Systems.
Vol. 2. Paris: UNESCO.
Weaver, T. D., Roseman, C.C., & Stringer C.B. (2008). Close correspondence between quantitative and molecular-genetic
States of America, 105, 4645-4649.
International.
113
__________________________________________________________________________________
Appendix
Galors distorted presentation of data
Figure 1. Two examples of the ubiquitous, grossly-distorted and self-misleading diagrams used to create the Unified Growth
Theory (Galor, 2005a, 2011). Madisons data (Maddison, 2001) were used during the formulation of this theory but they
were never analysed. Such a state-of-the-art was used to construct a system of scientifically-unsupported concepts,
interpretations and explanations. These erroneous concepts are now repeatedly contradicted by the same data, which were
used during the formulation of this theory.
Schematic diagrams of the signatures of Malthusian stagnation and takeoffs
Figure 2. The direct display of hypothetical GDP data serving as the schematic representation of the signature of Malthusian
stagnation (random fluctuations around an approximately horizontal line) followed by the escape from the Malthusian trap
into the sustained economic-growth regime around AD 1750 for developed regions and around AD 1900 for less-developed
regions as claimed by Galor (2005a, 2008a, 2011, 2012a). If these signatures are missing, Unified Growth Theory is
contradicted by data.
Figure 3. The display of the reciprocal values of the same hypothetical data as shown in Figure 2, serving as the schematic
representation of the signature of Malthusian stagnation (random fluctuations around an approximately horizontal line)
followed by the escape from the Malthusian trap into the sustained economic-growth regime around AD 1750 for developed
regions and around AD 1900 for less-developed regions as claimed by Galor (2005a, 2008a, 2011, 2012a. If these signatures
are absent, the Unified Growth Theory is contradicted by data.
114
__________________________________________________________________________________
Western Europe
Figure 4. Economic growth in the 12 selected countries of Western Europe representing the most-advanced economies where
the Unified Growth Theory should have the strongest confirmation. There was no transition from stagnation to growth at any
time because there was no stagnation. The growth was hyperbolic before and after the alleged transition around AD 1750.
Industrial Revolution did not boost the economic growth. The remarkable or stunning escape from the Malthusian trap
(Galor, 2005a, pp. 177, 220) did not happen because there was no trap. Galors three regimes of growth have no relevance to
the description, let alone to the explanation, of the mechanism of the economic growth. During the alleged sustained growth
regime, when the economic growth was supposed to follow a fast-increasing trajectory after the epoch of stagnation,
economic growth was diverted to a slower trajectory.
Figure 5. Reciprocal values of the GDP data, 1/GDP, for the economic growth in the 12 selected countries of Western
Europe. Unified Growth Theory (Galor, 2005a, 2011) is contradicted by Maddisons data (Maddison, 2010). Galors three
regimes of growth have no relevance to the description, let alone to the explanation, of the mechanism of the economic
growth. There was no transition from stagnation to growth at any time because there was no stagnation. There was no
remarkable or stunning escape from the Malthusian trap (Galor, 2005a, pp. 177, 220) because there was no trap.
Industrial Revolution did not boost the economic growth even in the countries where its effects should be most pronounced.
During the alleged sustained growth regime, when the economic growth was supposed to follow a fast-increasing trajectory
after the epoch of stagnation, economic growth was diverted to a slower trajectory, as indicated by the upward bending of
the trajectory of the reciprocal values.
115
__________________________________________________________________________________
Figure 6. Economic growth in the total of 30 countries of Western Europe. The data give no clear support for the existence of
the alleged Malthusian regime of stagnation. Industrial Revolution did not boost the economic growth in Western Europe.
The remarkable or stunning escape from the Malthusian trap (Galor, 2005a, pp. 177, 220) did not happen because there
was no trap in the economic growth. Galors three regimes of growth have no relevance to the description and to the
explanation of the mechanism of the economic growth in Western Europe. During the alleged sustained growth regime, when
the economic growth was supposed to follow a fast-increasing trajectory after the epoch of stagnation, economic growth was
Figure 7. Reciprocal values of the GDP data, 1/GDP, for the economic growth in the total of 30 countries of Western
Europe. Unified Growth Theory (Galor, 2005a, 2011) is contradicted by Maddisons data (Maddison, 2010). Galors three
regimes of growth have no expected correlation with data. There was no transition from stagnation to growth at any time
because there was no stagnation. There was no remarkable or stunning escape from the Malthusian trap (Galor, 2005a,
pp. 177, 220) because there was no trap in the economic growth. Industrial Revolution did not boost the economic growth in
Western Europe. During the alleged sustained growth regime, when the economic growth was supposed to follow a fast-
increasing trajectory after the alleged but non-existent epoch of stagnation, economic growth was diverted to a slower
trajectory
116
__________________________________________________________________________________
Eastern Europe
Figure 8. Economic growth in Eastern Europe. Galors three regimes of growth have no relevance to the description, let
alone to the explanation, of the mechanism of the economic growth. Unified Growth Theory is contradicted by data. The
alleged Malthusian regime of stagnation did not exist. Industrial Revolution did not boost the economic growth in Eastern
Europe. The remarkable or stunning escape from the Malthusian trap (Galor, 2005a, pp. 177, 220) did not happen
because there was no trap in the economic growth. During the alleged sustained growth regime, when the economic growth
was supposed to follow a fast-increasing trajectory after the epoch of stagnation, economic growth was diverted to a slower
trajectory
Figure 9. Reciprocal values of the GDP data, 1/GDP, for the economic growth in Eastern Europe. Unified Growth Theory
(Galor, 2005a, 2011) is contradicted by Maddisons data (Maddison, 2010). Galors three regimes of growth have no
expected connection with data. There was no transition from stagnation to growth at any time because there was no
stagnation. There was no remarkable or stunning escape from the Malthusian trap (Galor, 2005a, pp. 177, 220) because
there was no trap. Industrial Revolution did not boost the economic growth in Eastern Europe. Galors theory has no
relevance to the description, let alone to the explanation, of the mechanism of the economic growth. During the alleged
sustained growth regime, when the economic growth was supposed to follow a fast-increasing trajectory after the epoch of
stagnation, economic growth was diverted to a slower trajectory, as indicated by the upward bending of the trajectory of the
reciprocal values.
117
__________________________________________________________________________________
Asia (excluding Japan)
Figure 10. Economic growth in Asia (excluding Japan) between AD 1 and 2008. Maddisons data (Maddison, 2010) are
compared with the hyperbolic distribution and with their unsubstantiated interpretations promoted by Galor (2005a, 2011).
Economic growth was hyperbolic from at least AD 1000 until 2008. The minor delay after the Industrial Revolution was
followed by the compensating recovery. The concept of Malthusian regime of stagnation is contradicted by data. The escape
from the Malthusian trap never happened because there was no trap in the economic growth. There was no dramatic
transition from stagnation to growth because there was no stagnation.
Figure 11. Reciprocal values of the GDP data, 1/GDP, for Asia (excluding Japan) demonstrate that there is no correlation
between the claimed events (Industrial Revolution, the alleged Malthusian regime of stagnation and the alleged post-
Malthusian regime) and the data (Maddison, 2010). The postulated dramatic and remarkable takeoff around 1900 never
happened. The Malthusian regime of stagnation and the post-Malthusian regime did not exist.
118
__________________________________________________________________________________
Former USSR
Figure 12. Economic growth in countries of the former USSR between AD 1 and 2008, as represented by Maddisons data
(Maddison, 2010), is compared with the hyperbolic distribution and with the unsubstantiated interpretations of the
mechanism of growth proposed by Galor (2005a, 2008a, 2011, 2012a). The alleged Malthusian regime of stagnation did not
exist and neither did the alleged post-Malthusian and sustained-growth regimes. The Industrial Revolution had absolutely no
impact on changing the economic growth trajectory. There was also no dramatic transition to a new and faster economic
growth after the alleged epoch of stagnation, no transition from stagnation to growth at any time because there was no
stagnation. There was no escape from the Malthusian trap because there was no trap. In place of all these imaginary and
wished-for features there was the undisturbed and well-sustained hyperbolic growth. During the alleged sustained growth
regime (Galor, 2005a, 2011), when the economic growth was supposed to follow a fast-increasing trajectory after the epoch
of stagnation, economic growth was diverted to a slower trajectory.
Figure 13. Reciprocal values of the GDP data, 1/GDP, for the former USSR are compared with the hyperbolic distribution
represented by the decreasing straight line. There was no stagnation. Throughout the entire range of the alleged Malthusian
regime during the AD era, economic growth was hyperbolic.
119
__________________________________________________________________________________
Figure 14. The end part of the plot of the reciprocal values of the GDP data, 1/GDP, for the former USSR. Economic growth
was hyperbolic until around AD 1870 when it started to be diverted to a slower trajectory indicated by an upward bending of
the reciprocal values. Industrial Revolution did not boost the economic growth. The alleged Malthusian regime of stagnation
did not exist and there was no transition from stagnation to growth at any time because there was no stagnation. The
stunning or remarkable escape from the Malthusian trap (Galor, 2005a, pp. 177, 220) did not happen because there was
no trap. During the alleged sustained growth regime, when the economic growth was supposed to follow a fast-increasing
trajectory after the epoch of stagnation, economic growth was diverted to a slower trajectory.
Africa
Figure 15. Reciprocal values of the GDP data (Maddison, 2010) for Africa compared with the hyperbolic distributions
represented by the decreasing straight lines. The two distinctly different regimes of growth postulated by Galor (2005a,
2008a, 2011, 2012a) did not exist. There was no transition from stagnation to growth at any time because there was no
stagnation.
120
__________________________________________________________________________________
Figure 16. Reciprocal values of the GDP data (Maddison, 2010) for Africa between AD 1500 and 2008 compared with the
hyperbolic distributions represented by the decreasing straight lines. The two distinctly different regimes of growth
postulated by Galor (2005a, 2008a, 2011, 2012a) did not exist. His postulate ignores the data. There was no transition from
stagnation to growth because there was no stagnation. During the alleged post-Malthusian regime, when the economic growth
was supposed to start to follow a fast-increasing trajectory after the alleged epoch of stagnation, economic growth was
Figure 17. GDP data (Maddison, 2010) for Africa between AD 1 and 2008 compared with hyperbolic distributions. The two
distinctly different regimes of growth postulated by Galor (2005a, 2008a, 2011, 2012a) did not exist. His postulate ignores
the data. Galor did not even notice the transition between hyperbolic trajectories around 1820 because he did not analyse data
but manipulated them to support his preconceived ideas. There was no stagnation and no transition to a faster growth at the
end of the alleged regime of Malthusian stagnation. There was no escape from the Malthusian trap because there was no trap.
During the alleged post-Malthusian regime, when the economic growth was supposed to start to follow a fast-increasing
trajectory after the alleged epoch of stagnation, economic growth was diverted to a slower trajectory
121
__________________________________________________________________________________
Figure 18. GDP data (Maddison, 2010) for Africa between AD 1500 and 2008 compared with hyperbolic distributions. The
two distinctly different regimes of growth postulated by Galor (2005a, 2008a, 2011, 2012a) did not exist. His postulate
ignores the clear evidence in data. The data are in clear contradiction of Galors theory. There was no transition from
stagnation to growth because there was no stagnation. During the alleged post-Malthusian regime, when the economic growth
was supposed to start to follow a fast-increasing trajectory after the alleged epoch of stagnation, economic growth was
diverted to a slower trajectory
Latin America
Figure 19. Economic growth in Latin America between AD 1 and 2008. Maddisons data (Maddison, 2010) are compared
with hyperbolic distributions and with their unsubstantiated interpretations proposed by Galor (2005a, 2008a, 2011, 2012a).
The data suggest two hyperbolic distributions, the pattern similar to the economic growth in Africa. The alleged transition
from stagnation to growth never happened because the economic growth was not stagnant but hyperbolic. Around the time
of the postulated by Galor remarkable escape from the alleged Malthusian trap (Galor, 2005a, p. 177) at the end of the
alleged regime of stagnation, the economic growth started to be diverted to a slower trajectory. There was no escape from
the Malthusian trap because there was no trap.
122
__________________________________________________________________________________
Figure 20. Reciprocal values of the GDP data, 1/GDP, for Latin America between AD 1 and 2008. Maddisons data
(Maddison, 2010) are compared with hyperbolic distributions represented by the decreasing straight lines and with their
unsubstantiated interpretations proposed by Galor (2005a, 2008a, 2011, 2012a). During the alleged regime of stagnation, the
growth was hyperbolic. The data suggest two hyperbolic distributions, the pattern similar to the economic growth in Africa.
The alleged transition from stagnation to growth around AD 1900 did not happened because there was no stagnation. Around
the time of the alleged takeoff from stagnation to growth, economic growth started to be diverted from the fast-increasing
hyperbolic trajectory to a slower trajectory as indicated by the upward bending of the trajectory of the reciprocal values.
There was no escape from the Malthusian trap because there was no trap. The transition from the slow to fast growth
occurred around 300 years before the expected takeoff in AD 1900 and it was not a transition from stagnation to growth but
from growth to growth. This feature, as well as the diversion to a slower trajectory at the time of the claimed takeoff around
AD 1900, is not even noticed in the Unified Growth Theory.
123
__________________________________________________________________________________
Unified Growth Theory Contradicted by the Mathematical

Analysis of the Historical Growth of Human Population
By Ron W. NIELSENa
Abstract. Data describing historical growth of human population, global and regional (Western Europe, Eastern
Europe, Asia, former USSR, Africa and Latin America), are analysed. Results are in harmony with the earlier
analysis of the historical growth of the world population in the past 12,000 years and with other independent
studies. This analysis is also in harmony with the study of the historical economic growth. Within the range of
analysable data, there was no Malthusian stagnation. Takeoffs from stagnation to growth, postulated by the
Unified Growth Theory never happened. There were no escapes from the Malthusian trap because there was no
trap in the growth of population. This analysis and the earlier studies of the Gross Domestic Product lead to the
conclusion that there were also no takeoffs in the income per capita distributions, claimed by the Unified Growth
Theory. Consequently, the claimed differential timing in takeoffs never happened. Unified Growth Theory is
contradicted yet again by the mathematical analysis of the same data, which were used, but never analysed, during
the formulation of this theory. The study presented here, as well as earlier publications on the related topics, shows
also that certain fundamental postulates used in the economic and demographic research are repeatedly
contradicted by the mathematical analysis of data.
Keywords. Population growth, Economic growth, Unified Growth Theory, Regimes of growth, Malthusian trap,
Gross Domestic Product, Income per capita, Hyperbolic growth.
JEL. A10, A12, B41, C50, F00, F01, Y80.
1. Introduction
H istorical economic growth can be studied using the Gross Domestic Product (GDP). However,
to understand the time dependence of the income per capita (GDP/cap) it is necessary to
understand not only the economic growth, expressed in terms of the GDP, but also the growth
of human population. We have already analysed the GDP data (Nielsen, 2016a, 2016b, 2016d). Now,
we shall analyse the growth of human population using the same source of data (Maddison, 2010). The
aim of these studies is to investigate the validity of the Unified Growth Theory (Galor, 2005a, 2011)
and to understand the mechanism of growth as revealed by data because the correct interpretation of the
historical growth can help in the correct interpretation of the current economic growth and of the growth
of population.
Our earlier analysis (Nielsen, 2016a, 2016b, 2016d) demonstrated that the historical economic
growth, regional and global, was hyperbolic. Thus, if unchecked, the natural tendency for the economic
growth is to follow hyperbolic distributions. This type of spontaneous growth is undesirable because
hyperbolic distributions escape to infinity at a fixed time. To avoid such a rapid and potentially
catastrophic increase economic growth has to be closely monitored and controlled.
Analysis published over 50 years ago (von Foerster, Mora &Amiot, 1960) demonstrated that the
growth of the world population was also hyperbolic during the AD era. The follow-up analysis (Nielsen,
2016c) demonstrated that the growth of the world population was hyperbolic not only during the AD
era but also during the BC era, for the total of around 12,000 years. This particular analysis identified
two demographic transitions in the past growth of the population: between 500 BC and AD 500 and
between AD 1200 and 1400. However, these transitions were of a different kind than the transitions
a
Australia.
. +61407201175
Published as: Nielsen, R. W. (2016). Unified Growth Theory contradicted by the mathematical analysis of the historical
growth of human population. Journal of Economics and Political Economy, 3(2), 242-263.
124
__________________________________________________________________________________
assumed routinely in demographic research. They were not transitions from stagnation to growth but
from growth to growth, or more precisely, from hyperbolic growth to hyperbolic growth. The first
transition was from a fast-hyperbolic growth during the BC era to a significantly slower hyperbolic
growth during the AD era. During this transition, the size of human population reached a maximum
around AD 1 and after reaching a minimum between AD 400 and 500 it resumed it slower hyperbolic
growth during the AD era. However, the starting size of global population in AD 500 was significantly
larger than in 10,000 BC and the slower hyperbolic growth increased rapidly to reach a large size of the
population in only about 2000 years. During this first demographic transition, the growth rate decreased
from 0.252% in 500 BC to 0.066% in AD 500. The second transition was hardly noticeable but it
resulted in a change from a slow hyperbolic trajectory to a slightly faster hyperbolic trajectory. During
this transition, after a short delay in the growth of the population, the growth rate increased only
marginally from 0.123% in AD 1200 to 0.157% in AD 1400. Currently the growth of the world
population experiences a third demographic transition to a yet unknown trajectory.

The latest and the most elaborate theory describing economic growth is the Unified Growth Theory
(Galor, 2005a, 2011). It is not a theory, which is widely accepted by economist and used in their
research. In fact, the opposite seems to be true. However, we are using this theory as an example for
two reasons. First, it is a theory, which is firmly based on traditional but erroneous assumptions about
the historical economic growth and about the historical growth of human population. Our primary aim
here, the same aim we had in earlier publications (Nielsen, 2014, 2015a, 2016a, 2016b, 2016c, 2016d,
2016e), is not just to test the validity of this theory or the validity of a similar Demographic Transition
Theory (see Nielsen, 2016e and references therein) but to test the validity of the fundamental postulates
used in the economic and demographic research. Second, Unified Growth Theory appears to be the only
theory where Maddisons data (Maddison, 2001) were used systematically but unfortunately, they were
never analysed. They were manipulated and distorted to support preconceived ideas. Now, precisely the
same data can be used to show that the preconceived ideas used and promoted in this theory are
incorrect (Nielsen, 2014, 2015a, 2016a, 2016b, 2016d). Here we have a difference between a study
based on the manipulation of data and a study based on the rigorous analysis of data.
In the last years of his life, Magnusson, the world-renown economist, published excellent data
describing not only the economic growth as expressed by the Gross Domestic Product (GDP) but also
the growth of human population, global, regional and national (Magnuson, 2001, 2010). These data are
a treasure trove, which can be used in the economic and demographic research. In particular, they can
be used to test the fundamental postulates supporting these two fields of research. Galor used the earlier
compilation of these data (Magnuson, 2001) but any of them can be used to test the fundamental
postulates supporting economic and demographic research, and in particular to test the validity of the
Unified Growth Theory.
Unfortunately, this theory and its fundamental postulates are based on the habitually distorted and
self-misleading presentations of data (Ashraf, 2009; Galor, 2005a, 2005b, 2007, 2008a, 2008b, 2008c,
2010, 2011, 2012a, 2012b, 2012c; Galor & Moav, 2002; Snowdon & Galor, 2008). This
counterproductive approach to research was used to promote such scientifically-unsupported concepts
as the concept of the three regimes of growth (Malthusian regime of stagnation, post-Malthusian regime
and sustained-growth regime), the concept of sudden takeoffs from stagnation to growth, the concept
of differential timing of takeoffs and the concept of the great divergence. An example of the strongly
deceptive and misleading diagrams used in the Unified Growth Theory and in other related publications
is shown in Figure 1. (All diagrams are presented in the Appendix.)
Hyperbolic distributions do not have to be distorted to be confusing. They are already sufficiently
confusing and it is easy to make mistakes with their interpretations. Hyperbolic distributions have to be
carefully and methodically analysed and fortunately their analysis becomes trivial when using the
reciprocal values of data (Nielsen, 2014). Displays, such as presented in Figure 1, which is based on a
figure presented by Galor (2005a, p. 181), are self-misleading and they inevitably lead to incorrect
conclusions.
The correct and accurate display of Maddisons data (Maddison, 2001), precisely the same data as
used but never scientifically analysed during the formulation of the Unified Growth Theory (Galor,
125
__________________________________________________________________________________
2005a, 2011), is presented in Figure 2. Analysis of these data reveals that they follow monotonically-
increasing distributions, which are impossible to divide into distinctly-different regimes of growth
governed by distinctly-different mechanisms (Nielsen, 2014, 2015a). There was no stagnation and no
transition from stagnation to growth. There was no escape from the Malthusian Trap because there was
no trap in the economic growth.
Whether expressed by using the GDP or GDP/cap, economic growth was slow over a long time and
fast over a short time but it was monotonically increasing all the time. What appears as stagnation was
a part of the monotonically-increasing distribution, and what appears as a sudden takeoff was the natural
continuation of the same monotonically-increasing distribution.
Attempts to determine the time of the perceived transition from slow to fast growth are bound to be
unsuccessful because there was no transition (Nielsen, 2014, 2015a). The growth of the GDP is
described by hyperbolic distributions (Nielsen, 2014, 2016a) and the growth of the GDP/cap by the
linearly-modulated hyperbolic distributions (Nielsen, 2015a).
One of the fundamental postulates of the Unified Growth Theory is the postulate of the existence of
three regimes of growth governed by three distinctly different mechanisms: (1) the Malthusian regime
of stagnation, (2) the post-Malthusian regime, and (3) the sustained-growth regime. This postulate
applies not only to the growth of the GDP but also to the growth of human population because Galor
discusses the growth of income per capita, (GDP/cap), which is made of two components: the growth
of the GDP and the growth of population.
According to Galor (2005a, 2008a, 2011, 2012a), Malthusian regime of stagnation was between
100,000 BC and AD 1750 for developed regions and between 100,000 BC and AD 1900 for less-
developed regions. The claimed starting time appears to be based entirely on conjecture because
Maddisons data are terminated at AD 1 and even they contain significant gaps below AD 1500. The
post-Malthusian regime was allegedly between AD 1750 and 1850 for developed regions and from
1900 for less-developed regions. The sustained-growth regime was supposed to have commenced
around 1850 for developed regions.
Unified Growth Theory (Galor, 2005a, 2008a, 2011, 2012a) can be tested in many ways but the
easiest way to test it is to look for the dramatic takeoffs from stagnation to growth. These takeoffs are
described as a remarkable or stunning escape from the Malthusian trap (Galor, 2005a, pp. 177,
220). It is a signature, which cannot be missed.
This change in the pattern of growth is described as the sudden take-off from stagnation to growth
(Galor, 2005a, pp. 177, 220, 277) or as a sudden spurt (Galor, 2005a, 177, 220). According to Galor,
for developed regions, the end of the Malthusian regime of stagnation coincides with the Industrial
Revolution. The take-off of developed regions from the Malthusian Regime was associated with the
Industrial Revolution (Galor, 2005a, p. 185). Indeed, the Industrial Revolution is considered to have
been the prime engine of economic growth (Galor, 2005a, p. 212).
The signature of takeoffs is characterised by three features: (1) it should be a prominent change in
the pattern of growth, (2) it should be a transition from stagnation to growth and (3) it should occur at
the time predicted by the theory. For developed regions, the postulated takeoffs should occur around
AD 1750, or around the time of the Industrial Revolution, 1760-1840 (Floud & McCloskey, 1994). For
less-developed regions, they should occur around 1900. The added advantage of using this simple test
is that there are no significant gaps in the data around the time of the postulated takeoffs and
consequently the stagnation and the expected prominent transitions from stagnation to growth should
be easily identifiable.
growth. Thus, for instance, a transition from hyperbolic growth to another hyperbolic growth or to some
other steadily-increasing trajectory is not a signature of the sudden takeoff from stagnation to growth.
The takeoffs claimed by Galor are in the income per capita (GDP/cap), which means that there
should be takeoffs from stagnation to growth in at least one of these components (in the GDP or in the
population or in both of them) at a specific time (Galor, 2008a, 2012a). We have already demonstrated
that the Unified Growth Theory is contradicted by the GDP data describing the world economic growth
as well as the economic growth in Western Europe, Eastern Europe, countries of the former USSR,
Asia, Africa and Latin America (Nielsen, 2016a, 2016b, 2016d). We have also demonstrated that the
Unified Growth Theory is contradicted by the data describing the growth of the world income per capita
126
__________________________________________________________________________________
(Nielsen, 2015a). Our next step now is to extend our analysis to the growth of population and thus to
extend our study of income per capita, not only global but also regional.
We have already demonstrated that there were no takeoffs in the growth of the GDP. Consequently,
to confirm the Unified Growth Theory we would have to show not only that there were takeoffs from
stagnation to growth in the growth of the population but also that these takeoffs occurred at the specific
time claimed by Galor (2008a, 2012a), around AD 1750 for developed regions (Western Europe,
Eastern Europe and the former USSR) and at around AD 1900 for less developed regions (Asia, Africa
and Latin America). We shall now demonstrate that there were no such takeoffs. Thus, we shall
demonstrate implicitly that there were no takeoffs in the income per capita, which means that Galors
postulate of the differential timing in takeoffs is also contradicted by data, because we cannot have
differential timing in takeoffs without takeoffs.
3. Essentials of the mathematical analysis

1 ,
S (t ) = (1)
a kt
where S (t ) is the size of the growing entity, in our case the population, while a and k are positive
constants.
As pointed out earlier (Nielsen, 2014), hyperbolic distributions are confusing because they create an
illusion of being made of two components, slow and fast, with perhaps even a third component in the
middle. It is easy to make a mistake with their interpretations. Fortunately, these distributions are easy
to analyse by using the reciprocal values of data, 1/ S (t ) :
1
= a kt . (2)
S (t )
In this representation, data follow a decreasing straight line, which obviously cannot be divided into
two or three distinctly different components.
Reciprocal values help in an easy and generally unique identification of hyperbolic growth. Apart
from serving as an alternative way to analyse data, reciprocal values allow also for the investigation of
even small deviations from hyperbolic distributions because deviations from a straight line can be easily
noticed.
The illusion of different components also disappears when using semilogarithmic scales of
reference. Both types of displays help in an easy identification of disagreements between data and fitted
curves for small values of data and we shall use both of these displays.
4. Growth of the world population

Results of mathematical analysis of the world population are presented in Figures 3 and 4. Reciprocal
values of historical data identify uniquely hyperbolic distribution between AD 1000 and around 1950
because they follow a decreasing straight line. From around 1950, the growth of the world population
started to be diverted to a slower trajectory but first it was slightly boosted. The boosting was small (it
is hardly noticeable in the displayed diagrams) and it did not last long.
Hyperbolic fit to the world population data (Maddison, 2010) is shown in Figure 4. The fit is
remarkably good. The point at AD 1 is 75% higher than the fitted curve. This discrepancy is in perfect
agreement with the analysis of the growth of the world population over the past 12,000 years (Nielsen,
2016c), which demonstrated a maximum around that year.
Parameters describing hyperbolic trajectory fitting the data between AD 1000 and 1950 are:
= a 7.739 100 and= k 3.765 103 . Its singularity is at t = 2056 . However, from around 1950, the
growth of the world population started to be diverted to a slower trajectory bypassing the singularity by
a safe margin of 106 years. This diversion was first manifested in a minor and short-lasting boosting of
the growth of the world population.
127
__________________________________________________________________________________
The data are in disagreement with the Unified Growth Theory (2005a, 2011), which erroneously
claims stagnation and takeoffs from stagnation to growth. There was no stagnation but monotonically-
increasing hyperbolic distribution. There were also no takeoffs from stagnation to growth around AD
1750 for developed regions and around AD 1900 for less-developed regions because there was no
stagnation and because hyperbolic growth continued undisturbed. If there were such takeoffs in the
respective regions they must have been too weak to change the growth of global population because the
growth trajectory was remarkably stable during these alleged takeoffs. Furthermore, our analysis shows
that the Industrial Revolution had no impact on the growth trajectory. Unified Growth Theory is yet
again demonstrably contradicted by data.
With the absence of takeoffs in the growth of the population and with the earlier demonstrated
absence of takeoffs in the growth of the GDP (Nielsen, 2016a), this analysis shows that there were no
takeoffs in the income per capita (GDP/cap) distribution confirming our previous analysis based on the
earlier compilation of Maddisons data (Maddison, 2001). Unified Growth Theory (Galor, 2005a, 2011)
is contradicted by data, which were used but never analysed during the formulation of this theory.
5. Western Europe
Growth of population in Western Europe is shown in Figures 5 and 6. Western Europe is represented
by the total of 30 countries: Austria, Belgium, Denmark, Finland, France, Germany, Italy, The
Netherlands, Norway, Sweden, Switzerland, United Kingdom, Greece, Portugal, Spain and by 14 small
but unspecified countries. Ireland is missing in this list because it was included only from 1921.
The straight line fitting the reciprocal values of data, shown in Figure 5, identifies uniquely
hyperbolic distribution between AD 1000 and around 1915. Parameters describing hyperbolic growth
in Western Europe are:= =
a 7.542 101 and k 3.749 102 . The point of singularity is at t = 2012 .
From around 1915, the growth of population in Western Europe started to be diverted to a slower, but
still fast-increasing, trajectory bypassing the singularity by a safe margin of 97 years. The size of the
population in AD 1 is 89% larger than for the fitted hyperbolic distribution. This discrepancy is probably
reflecting the maximum in the growth of the world population around that year (Nielsen, 2016c).
Figures 5 and 6 show that hyperbolic growth between AD 1000 and 1915 remained undisturbed.
Industrial Revolution had absolutely no impact on changing the hyperbolic growth trajectory in the
region where the effects of this revolution should be most prominent. There was no takeoff from
stagnation to growth at the postulated time (Galor, 2008a, 2012a) because there was no stagnation in
the growth of population. There was even no transition to a faster hyperbolic trajectory.
With the absence of the takeoff in the growth of the population in Western Europe and with the
earlier demonstrated absence of the takeoff in the growth of the GDP (Nielsen, 2016a), this analysis
shows that there was no takeoff in the income per capita (GDP/cap) distribution. Unified Growth Theory
(Galor, 2005a, 2011) is contradicted by data, which were used but never analysed during the formulation
of this theory.
6. Eastern Europe
Results of analysis of the growth of population in Eastern Europe are summarized in Figures 7 and
8. Reciprocal values of data shown in Figure 7 identify uniquely hyperbolic distribution between AD
1000 and around 1935. From that year, the growth of population started to be diverted to a slower
trajectory.
Hyperbolic parameters are: = a 3.055 102 and
= k 1.525 101 . The point of singularity is at
t = 2003 . Figures 7 and 8 demonstrate that the Industrial Revolution had no impact on the trajectory
of the growth of population in Eastern Europe and that there was no takeoff from stagnation to growth
at the postulated time (Galor, 2008a, 2012a) because there was no stagnation but hyperbolic growth.
There was even no takeoff to a faster hyperbolic growth. The size of the population in AD 1 was 45%
higher than the calculated curve reflecting probably the maximum in the growth of the world population
around that year (Nielsen, 2016c).
With the absence of the takeoff in the growth of population in Eastern Europe and with the earlier
demonstrated absence of the takeoff in the growth of the GDP (Nielsen, 2016a), this analysis shows
that there was no takeoff in the income per capita (GDP/cap) distribution. Unified Growth Theory
128
__________________________________________________________________________________
of this theory.
7. Former USSR
The analysis of data for countries of the former USSR is presented in Figures 9 and 10. Reciprocal
values shown in Figure 9 identify uniquely hyperbolic distribution between AD 1 and around 1920.
Hyperbolic fit to the data is between AD 1 and 1870. Parameters fitting the data are: = a 2.618 102
=
and k 1.333 10 . The singularity is at t = 1965 . From around 1920, the growth of population in the
1
former USSR started to be diverted to a slower trajectory, bypassing the singularity by around 45 years.
Figures 9 and 10 show that the Industrial Revolution had no impact on shaping the growth of human
population in countries of the former USSR. There was also no takeoff from stagnation to growth around
the postulated time Galor (2008a, 2012a) or around any other time because the growth was not stagnant
but hyperbolic. There was even no transition to a faster hyperbolic trajectory but there was a transition
to a slower, non-hyperbolic growth around 1920.
With the absence of the takeoff in the growth of the population in countries of the former USSR and
with the earlier demonstrated absence of the takeoff in the growth of the GDP (Nielsen, 2016a), this
analysis shows that there was no takeoff in the income per capita (GDP/cap) distribution. Unified
Growth Theory (Galor, 2005a, 2011) is contradicted by data, which were used but never analysed during
the formulation of this theory.
8. Asia
Analysis of the growth of human population in Asia (including Japan) is summarised in Figures 11
and 12. Reciprocal values presented in Figure 11 identify uniquely hyperbolic distribution between AD
1000 and around 1920. Parameters describing this distribution are: = a 1.068 101 and
= k 4.999 103
. The point of singularity is at t = 2135 .
Asia is made primarily of less-developed countries (BBC, 2014; Pereira, 2011) and consequently,
according to Galor (2008a, 2012a), the growth of human population in Asia should have been
characterised by stagnation until around 1900, the year marking the alleged stunning escape from the
Malthusian trap, the supposed escape manifested by the postulated dramatic takeoff. (The population
of Japan before AD 1900 was on average less than 4% of the total population of Asia.) The data and
their analysis show that there was no stagnation, at least from AD 1000 and there was also no expected
takeoff.
The data reveal a steadily increasing hyperbolic growth until around 1920. From around that year
the growth of population was diverted to a faster trajectory. This boosting can be seen clearly in Figures
11 and 12 and it occurred close to the time of the postulated takeoff from stagnation to growth. However,
it was not a transition from stagnation to growth but from hyperbolic growth to a slightly faster
trajectory of a different kind. It is, therefore, not the takeoff postulated by Galor. Furthermore, it was
only a temporary boosting, which is now returning to the original hyperbolic trajectory and, as indicated
by the reciprocal values of data, this new trend is likely to be slower than the original trajectory.
With the absence of the postulated takeoff in the growth of population in Asia and with the earlier
of this theory.
9. Africa
Results of analysis of the growth of human population in 57 African countries are presented in
Figures 13 and 14. Reciprocal values identify uniquely two hyperbolic trajectories: AD 1-1840 and AD
1840-1980. At first it was a slow hyperbolic growth characterised by parameters = a 5.794 101 and
=k 2.473 102 and by the singularity at t = 2343 . Then, around 1840, this slow hyperbolic growth
was replaced by a significantly faster hyperbolic growth characterised by parameters=a 1.571 102
=
and k 7.834 10 and by singularity at t = 2006 . Defined by the parameter k, this new growth was
2
129
__________________________________________________________________________________
3.2 times faster than the earlier hyperbolic growth. From around 1980, this fast-hyperbolic growth was
diverted to a slower, non-hyperbolic trajectory, bypassing singularity by 26 years.
Africa is also made of less-developed countries (BBC, 2014; Pereira, 2011) so according to Galor
(2008a, 2012a) it should have experienced stagnation until around 1900 followed by a clear takeoff
around that year. These expectations are contradicted by data because (1) the growth of population was
not stagnant but hyperbolic until around 1980 and (2) because there was no takeoff from stagnation to
growth around 1900 or around any other time. In fact, around that time hyperbolic growth continued
unaffected in contradiction of the wished-for interpretations.
The acceleration in the growth of human population in Africa occurred around 1840, but it was not
a transition from stagnation to growth but from growth to growth. Even more precisely, it was a
transition from hyperbolic growth to another hyperbolic growth.
Africa is the only region where the commencement of the rapid growth of population coincides with
the Industrial Revolution but it also the region, which demonstrates that the usually-claimed effects of
Industrial Revolution are contradicted by empirical evidence. According to the generally accepted
interpretation, Industrial Revolution improved medical care and introduced many other beneficial
effects, which were supposed to have caused population explosion. Data for Africa demonstrate that
this wished-for mechanism does not work because the growth of population on this continent coincides
with the rapidly-deteriorating living conditions of native populations brought about by the intensified
colonisation (Duignan & Gunn, 1973; McKay, Hill, Buckler, Ebrey, Beck, Crowston, & Wiesner-
Hanks, 2012; Pakenham, 1992).
This fast growth of population in Africa, which commenced around the time when living conditions
started to deteriorate rapidly, is easy to explain by noticing that the growth rate of population is directly
proportional to the level of deprivation (Nielsen, 2013), the process, which is diametrically opposite to
the usually claimed influence of the Industrial Revolution. It appears, therefore, that it is not the
improved living conditions but the increased level of deprivation that have a stimulating effect on the
growth of population.
With the absence of the takeoff around AD 1900 in Africa and with the earlier demonstrated absence
of the takeoff in the growth of the GDP (Nielsen, 2016a), this analysis shows that there was no takeoff
in the income per capita (GDP/cap) distribution. Unified Growth Theory (Galor, 2005a, 2011) is
contradicted by data, which were used but never analysed during the formulation of this theory.
10. Latin America

Results of analysis of population growth in Latin America are presented in Figures 15 and 16. Data
for Latin America are difficult to analyse because there was a significant decline in the growth of
population between AD 1500 and 1600 but they also appear to follow two distinctly different hyperbolic
trajectories, which can be easily identified using the reciprocal values of data (see Figure 15). However,
the identification of the first trajectory is not as clear as for Africa. The identification of the second
hyperbolic trajectory is more convincing. Tentative conclusion is that the growth of population in Latin
America was following a slow hyperbolic distribution between AD 1 and 1500 and a fast-hyperbolic
distribution between AD 1600 and around 1900.
The tentatively assigned slow hyperbolic growth between AD 1 and 1500 is characterised by
parameters= a 1.765 102 and
= k 8.242 102 . Its singularity is at t = 2142 . The better determined
fast hyperbolic growth between AD 1600 and 1900 is characterised by parameters = a 6.561 102 and
= k 3.371 101 . Its singularity is at t = 1947 . Defined by the parameter k, this growth was 4.1 times
faster than the earlier hyperbolic growth. From around 1900, this fast-hyperbolic growth started to be
diverted to a slower trajectory bypassing the singularity by 47 years. The transition from the earlier
apparent hyperbolic growth to a new and rapid hyperbolic growth, which occurred between around AD
1500 and 1600 appears to coincide with the commencement of the Spanish conquest (Bethell, 1984)
and with the rapidly-deteriorating living conditions. The mechanism of this fast growth of population
in Latin America appears to be the same as the mechanism of the fast growth in Africa, which
commenced around AD 1840.
according to Galor (2008a, 2012a), the growth of human population in this region should have been
stagnant until around 1900 and fast-increasing from around that year. This pattern of growth is
130
__________________________________________________________________________________
contradicted by data. The data show a diametrically different pattern: (1) there is no convincing
evidence of the existence of stagnation over the entire range of time between AD 1 and 1900 (there are
no signs of Malthusian oscillations) but there is a sufficiently convincing evidence of hyperbolic growth
particularly between AD 1600 and 1900; (2) there was no takeoff from stagnation to growth at any time;
and (3) at the time of the postulated takeoff in 1900 the growth of population started to be diverted to a
slower trajectory. The wished-for takeoff is replaced by a slower growth. However, even if we had a
takeoff around that time it would have been a takeoff of a different kind, not a takeoff from stagnation
to growth as required by the Unified Growth Theory (Galor, 2005a, 2011) but a takeoff from growth to
growth.
With the absence of the takeoff in the growth of population in Latin America and with the earlier
of this theory.

Results of mathematical analysis of the historical growth of human population are summarised in
Table 1. The listed parameters, a and k, are for the fitted hyperbolic distributions.
This analysis demonstrates that the natural tendency for the historical growth of human population
was to increase hyperbolically. In general, there is a remarkably good agreement between the data and
the calculated hyperbolic distributions.
Table 1. Summary of the mathematical analysis or the historical growth of population

Region/Countries a k Hyperbolic Range Singularity Proximity Takeoff
World 3 1000 1950 2056 106 X
7.739 10 0
3.765 10
3.749 102
Western Europe 1000 1915 2012 97 X
7.542 10 1
1.525 101
Eastern Europe 1000 1935 2003 68 X
3.055 102
1.333 101
Former USSR 1 1920 1965 45 X
2.618 102
4.999 103
Asia 1000 1920 2135 215 X
1.068 101
2.473 102
Africa 1 1840 2343
5.794 101 1840 1980 2006 26 X
1.571 102 7.834 102
8.242 102
Latin America 1 1500 2142
1.765 102 1600 1900 1947 47 X
6.561 102 3.371 101
Notes: a and k Hyperbolic growth parameters [see eqn (1)]. Hyperbolic Range - The empirically-confirmed range of time
when the growth of population can be described using hyperbolic distributions. Singularity - The time of the escape to infinity
for a given hyperbolic distribution. Proximity - Proximity (in years) of the singularity at the time when the growth of population
departed from the hyperbolic growth to a new trajectory. X - No takeoff from stagnation to growth. Takeoffs claimed by the
Unified Growth Theory (Galor, 2005a, 2008a, 2011, 2012a) did not happen because there was no stagnation in the growth of
population. The growth was monotonically hyperbolic.
Unlike the more familiar exponential distributions, which are easier to understand because they show
more readily a gradually increasing growth, hyperbolic distributions appear to be made of two or maybe
even three components: a slow component, a fast component and perhaps even a transition component
located between the apparent slow and fast components. The illusion is so strong that even the most
experienced researchers can be deceived particularly if they have no access to good sets of data, which
was in the past. Now, however, excellent data are available (Maddison, 2001, 2010) and we can use
them to check the earlier interpretations of economic growth and of the growth of human population.
The postulate of the existence of the epoch of Malthusian stagnation is suggested by a slow growth
over a long time but this slow growth is just a part of the hyperbolic growth, which can be convincingly
identified using reciprocal values. Hyperbolic distributions create also the illusion of a sudden takeoff
but this feature is also a part of hyperbolic growth.
Hyperbolic growth is slow over a long time and fast over a short time but the slow and fast growth
are the integral features of the same monotonically increasing distribution, which is easier to understand
131
__________________________________________________________________________________
by using the reciprocal values of the growing entity (Nielsen, 2014). In such displays, the illusion of
distinctly different components disappears because hyperbolic growth is then represented by a
decreasing straight line, which is easy to understand. It then becomes obvious that hyperbolic
distribution cannot be divided into distinctly different sections governed by different mechanism
because it makes no sense to divide a straight line into arbitrarily chosen sections and claim different
mechanism for such arbitrarily-selected section. It is then also clear that it is impossible to determine
the transition from a slow to fast growth. Which point on a straight line should we select to identify
such a transition? The transition does not happen at any specific time but gradually over the whole range
of time.
Our analysis shows that the Industrial Revolution had generally no impact on the growth of human
population. The only boosting of growth, which coincided with the Industrial Revolution was in Africa
but this boosting appears to have not been caused by the usually assumed beneficial effects of the
Industrial Revolution but by the rapidly deteriorating living conditions associated with the colonisation
of Africa (Duignan & Gunn, 1973; McKay, Hill, Buckler, Ebrey, Beck, Crowston, & Wiesner-Hanks,
2012; Pakenham, 1992). Our analysis also shows that the postulated takeoffs from stagnation to growth
(Galor, 2005, 2008a, 2011, 2012a) never happened because there was no stagnation in the growth of
population. We have shown earlier (Nielsen, 2016a) that there were no takeoffs in the growth of the
GDP, global or regional. The demonstrated now absence of takeoffs in the growth of population shows
that the claimed by Galor takeoffs in the income per capita (GDP/cap) did not exist.
Galor describes the imaginary and non-existing features, which have nothing to do with the
economic growth or with the growth of human population, features which were conjured from such
habitually distorted displays as shown in Figure 1, interpretations based on impressions, which were
never checked by the scientific analysis of data. They describe a world of fiction. All his explanations
of the mechanism of economic growth based on these and other imaginary features are not only
irrelevant but also misleading.
Galors Unified Growth Theory is fundamentally incorrect and is repeatedly contradicted by data
(Nielsen, 2014, 2015a, 2016a, 2016b, 2016d), ironically by the same data, which were used but never
analysed during the formulation of this theory. The evidence contradicting the fundamental postulates
of the Unfired Growth Theory is overwhelming and further evidence will be presented in forthcoming
publications. This evidence questions not only the fundamental postulates of the Unified Growth Theory
but also many similar postulates used traditionally in economic and demographic research, postulates
which are based largely on impressions and conjectures but postulates, which are repeatedly
contradicted by the analysis of Maddisons data (Maddison, 2001, 2010) as well as by other related
research (Kapitza, 2006; Kremer, 1993; Podlazov, 2002; Shklovskii, 1962, 2002; von Foerster, Mora
& Amiot, 1960; von Hoerner, 1975; see also Nielsen, 2016c, 2016e and references therein).
In science, just one contradicting evidence is sufficient to show that contradicted postulates need to
be closely examined and revised. Unified Growth Theory is scientifically unacceptable and so are also
many traditional interpretations of the historical economic growth and of the growth of human
population.
Data and their analysis suggest new lines of research. There is no need to waste time to discuss and
explain the mechanism of stagnation and takeoffs from stagnation to growth because these features are
contradicted by data. What needs to be explained is why the historical economic growth and the growth
of human population were hyperbolic and why relatively recently they were diverted to slower but still
fast-increasing trajectories. There is also a need to find out how to control these fast-increasing
trajectories.
Unified Growth Theory is not only spurious but also dangerously misleading. It claims erroneously
that after a long epoch of stagnation in the economic growth we have now entered a sustained-growth
regime. This concept suggests a prosperous and secure future. However, mathematical analysis of data
shows that the past economic growth was stable and sustainable but now it increases alarmingly fast
(Nielsen, 2015b, 2016a). The false sense of security is replaced by the realisation of the urgent need to
control and regulate economic growth and by the generally know need to control the growth of
population.
132
__________________________________________________________________________________
References
Bethell, L. (Ed.). (1984). The Cambridge History of Latin America: Vol. I and II, Colonial Latin America. Cambridge, UK:
Floud, D. & McCloskey, D. N. (1994). The Economic History of Britain since 1700. Cambridge: Cambridge University
Press.
Galor, O. (2005a). From stagnation to growth: Unified Growth Theory. In P. Aghion& S. Durlauf (Eds.), Handbook of
Galor, O. (2007). Multiple Growth RegimesInsights from Unified Growth Theory. Journal of Macroeconomics, 29(3),
470-475. doi. 10.1016/j.jmacro.2007.06.007
Dictionary of Economics. Palgrave Macmillan, New York. doi. 10.1057/9780230226203.0434
0062-7
117, 1133-1191. doi. 10.1162/003355302320935007
Kremer, M. (1993). Population Growth and Technological Change: One Million B.C. to 1990. Quarterly Journal of
Economics, 108(3), 681716.
Books.
Nielsen, R. W. (2013). Malthusian stagnation or Malthusian regeneration?
Nielsen, R. W. (2015a). Unified Growth Theory Contradicted by the GDP/cap Data.
http://arxiv.org/ftp/arxiv/papers/1511/1511.09323.pdf; https:/www.yumpu.com/user/notrelluf
Nielsen, R. W. (2015b). The insecure future of the world economic growth. Journal of Economic and Social Thought, 2(4),
242-255.
Nielsen, R. W. (2016b). The postulate of the three regimes of economic growth contradicted by data. Journal of Economic
Nielsen, R. W. (2016c). Growth of the world population in the past 12,000 years and its link to the economic growth.
Nielsen, R. W. (2016d). Unified Growth Theory contradicted by the absence of takeoffs in the Gross Domestic Product.
Nielsen, R. W. (2016e). Demographic Transition Theory and its link to the historical economic growth. Journal of Economic
Avon Books.
bank/
133
__________________________________________________________________________________
Podlazov, A. V. (2002). Theoretical demography: Models of population growth and global demographic transition (in
691-712.
134
__________________________________________________________________________________
Appendix
Figure 1. Example of the ubiquitous, grossly-distorted and self-misleading diagrams used to create the Unified Growth
Theory (Galor, 2005a, 2011). Madisons data (Maddison, 2001) were used during the formulation of this theory but they
were never analysed. Such state-of-the-art was used to construct a system of scientifically-unsupported concepts,
interpretations and explanations.
Figure 2. The same data (Maddison, 2001) as used in Figure 1 but now displayed accurately and analysed. They follow
monotonically-increasing distributions, which cannot be divided into distinctively-different components (Nielsen, 2014,
2015a).
135
__________________________________________________________________________________
World Population
Figure 3. Reciprocal values of the world population data (Maddison, 2010) identify uniquely hyperbolic distribution between
AD 1000 and around 1950 because they follow a decreasing straight line. From around 1950, the growth of population
started to be diverted to a new trajectory. Industrial Revolution had no impact on changing the growth trajectory. There were
also no takeoffs from stagnation to growth around the postulated times for developed and less-developed regions (Galor,
2008a, 2012a). This analysis and the absence of takeoffs in the corresponding GDP distribution (Nielsen, 2016a) show that
there were no takeoffs in the income per capita (GDP/cap) distribution. Unified Growth Theory is contradicted yet again by
the same data which were used but not analysed during the formulation of this theory.
Figure 4. Growth of the world population. Data of Maddison (2010) are compared with hyperbolic distribution. The point at
AD 1 is 75% higher than the fitted curve because there was a maximum in the growth of the world population around that
time (Nielsen, 2016c). Industrial Revolution had no impact on the growth of population. There were no takeoffs from
stagnation to growth around the postulated time (Galor, 2008a, 2012a) for developed and less-developed regions. This
analysis and the absence of takeoffs in the corresponding GDP distribution (Nielsen, 2016a) show that there were no takeoffs
in the income per capita (GDP/cap) distribution. Unified Growth Theory is contradicted yet again by the same data which
were used but not analysed during the formulation of this theory.
136
__________________________________________________________________________________
Western Europe
Figure 5. Reciprocal values of population data for Western Europe (Maddison, 2010) identify uniquely hyperbolic
distribution between AD 1000 and around 1915 because they follow a decreasing straight line. From around 1915, the
growth of population started to be diverted to a slower trajectory. Industrial Revolution had no impact on changing the
growth trajectory in the region where its influence should have been most pronounced. There was also no takeoff from
stagnation to growth around the postulated time (Galor, 2008a, 2012a). This analysis and the absence of the takeoff in the
corresponding GDP distribution (Nielsen, 2016a) show that there was no takeoff in the income per capita (GDP/cap)
distribution. Unified Growth Theory is contradicted yet again by the same data which were used but not analysed during the
formulation of this theory.
Figure 6. Growth of human population in Western Europe. Data of Maddison (2010) are compared with hyperbolic
distribution. The point at AD 1 is 89% higher than the fitted curve. This discrepancy might be reflecting the maximum in the
growth of the world population (Nielsen, 2016c). Industrial Revolution had no impact on the growth of population in Western
Europe where the effects of this revolution should have been most prominent. There was no takeoff from stagnation to growth
around the postulated time (Galor, 2008a, 2012a). This analysis and the absence of the takeoff in the corresponding GDP
distribution (Nielsen, 2016a) show that there was no takeoff in the income per capita (GDP/cap) distribution. Unified
Growth Theory is contradicted yet again by the same data which were used but not analysed during the formulation of this
theory.
137
__________________________________________________________________________________
Eastern Europe
Figure 7. Reciprocal values of population data for Eastern Europe (Maddison, 2010) identify uniquely hyperbolic
distribution between AD 1000 and around 1935 because they follow a decreasing straight line. From around 1935,
hyperbolic growth started to be diverted to a slower trajectory. Industrial Revolution had no impact on changing the growth
trajectory in Eastern Europe. There was also no takeoff from stagnation to growth around the postulated time (Galor,
2008a, 2012a). This analysis and the absence of the takeoff in the corresponding GDP distribution (Nielsen, 2016a) show
that there was no takeoff in the income per capita (GDP/cap) distribution. Unified Growth Theory is contradicted yet again
by the same data which were used but not analysed during the formulation of this theory.
Figure 8. Growth of human population in Eastern Europe. Data of Maddison (2010) are compared with hyperbolic
distribution. The point at AD 1 is 45% higher than the fitted curve. This discrepancy might be reflecting the maximum in the
growth of the world population (Nielsen, 2016c) around that time. Industrial Revolution had no impact on the growth of
population in Eastern Europe. There was no takeoff from stagnation to growth around the postulated time (Galor, 2008a,
2012a). This analysis and the absence of the takeoff in the corresponding GDP distribution (Nielsen, 2016a) show that there
was no takeoff in the income per capita (GDP/cap) distribution. Unified Growth Theory is contradicted yet again by the
same data which were used but not analysed during the formulation of this theory.
138
__________________________________________________________________________________
Former USSR
Figure 9. Reciprocal values of population data for the former USSR (Maddison, 2010) identify uniquely hyperbolic
distribution between AD 1 and 1920 because they follow closely the decreasing straight line. From around 1920 the growth
started to be diverted to a slower trajectory. Industrial Revolution had no impact on changing the growth trajectory. There
was also no takeoff from stagnation to growth around the postulated time (Galor, 2008a, 2012a) or around any other time
because there was no stagnation. There was even no transition to a faster hyperbolic growth. This analysis and the absence
of the takeoff in the corresponding GDP distribution (Nielsen, 2016a) show that there was no takeoff in the income per capita
(GDP/cap) distribution. Unified Growth Theory is contradicted yet again by the same data which were used but not analysed
during the formulation of this theory.
Figure 10. Growth of human population in countries of the former USSR. Data of Maddison (2010) are compared with
hyperbolic distribution. Industrial Revolution had no impact on the growth of population. There was no takeoff from
stagnation to growth around the postulated time (Galor, 2008a, 2012a) or around any other time because there was no
stagnation. This analysis and the absence of the takeoff in the corresponding GDP distribution (Nielsen, 2016a) show that
there was no takeoff in the income per capita (GDP/cap) distribution. Unified Growth Theory is contradicted yet again by the
same data which were used but not analysed during the formulation of this theory.
139
__________________________________________________________________________________
Asia (including Japan)
Figure 11. Reciprocal values of population data for Asia (Maddison, 2010) identify uniquely hyperbolic distribution between
AD 1 and 1920 because they follow closely the decreasing straight line. From around 1920, the growth started to be diverted
to a temporary faster trajectory. There was no takeoff from stagnation to growth around the postulated time (Galor, 2008a,
2012a) because there was no stagnation. The temporary boosting around 1920 appears to be a part of the commonly
observed transition from the historical hyperbolic growth to a slower trajectory. This analysis and the absence of the takeoff
in the corresponding GDP distribution (Nielsen, 2016a) show that there was no takeoff in the income per capita (GDP/cap)
distribution. Unified Growth Theory is contradicted yet again by the same data which were used but not analysed during the
formulation of this theory.
Figure 12. Growth of human population in Asia. Data of Maddison (2010) are compared with the hyperbolic distribution.
There was no stagnation but a hyperbolic growth between at least AD 1000 and 1920. The size of the population at AD 1 is
80% higher than the fitted hyperbolic distribution, reflecting probably the maximum in the growth of the world population
around that year (Nielsen, 2016c). There was no takeoff from stagnation to growth around the postulated time (Galor,
2008a, 2012a) because there was no stagnation before the temporary boosting from around 1920. This analysis and the
absence of the takeoff in the corresponding GDP distribution (Nielsen, 2016a) show that there was no takeoff in the income
per capita (GDP/cap) distribution. Unified Growth Theory is contradicted yet again by the same data which were used but
not analysed during the formulation of this theory.
140
__________________________________________________________________________________
Africa
Figure 13. Reciprocal values of population data for Africa (Maddison, 2010) identify uniquely two hyperbolic distributions:
AD 1-1840 and AD 1840-1980 because they follow closely the decreasing straight lines. From around 1980 the growth
started to be diverted to a slower trajectory. There was no takeoff from stagnation to growth around the postulated time
(Galor, 2008a, 2012a) because there was no stagnation. However, there was a transition around AD 1840 from a slow to a
fast-hyperbolic trajectory. This analysis and the absence of the takeoff in the corresponding GDP distribution (Nielsen,
2016a) show that there was no takeoff in the income per capita (GDP/cap) distribution. Unified Growth Theory is
contradicted yet again by the same data which were used but not analysed during the formulation of this theory.
Figure 14. Growth of human population in Africa. Data of Maddison (2010) are compared with two hyperbolic distributions,
AD 1-1840 and AD 1840-1980. There was no stagnation but a hyperbolic growth. There was no takeoff from stagnation to
growth around the postulated time (Galor, 2008a, 2012a) because there was no stagnation. The fast-hyperbolic growth,
continued undisturbed until 1980 when it started to be diverted to a slower trajectory. Around 1840, there was a transition
from a slow to a fast-hyperbolic trajectory. This analysis and the absence of the takeoff in the corresponding GDP
theory.
141
__________________________________________________________________________________
Latin America
Figure 15. Reciprocal values of population data for Latin America (Maddison, 2010) identify two hyperbolic distributions:
AD 1-1500 and AD 1600-1900 because they follow closely the decreasing straight lines. From around 1900 the growth
started to be diverted to a slower trajectory. There was no takeoff from stagnation to growth around the postulated time
(Galor, 2008a, 2012a) but there was a transition around the postulated takeoff to a slower trajectory. Data replace Galors
takeoff by a transition to a slower trajectory. This analysis and the absence of the takeoff in the corresponding GDP
theory.
Figure 16. Growth of human population in Latin America. Data of Maddison (2010) are compared with two hyperbolic
distributions, AD 1-1500 and AD 1600-1900. There was no stagnation but a hyperbolic growth. There was no takeoff from
stagnation to growth around the postulated time (Galor, 2008a, 2012a) because there was no stagnation. The fast-hyperbolic
growth continued undisturbed until 1900 when it started to be diverted to a slower trajectory. Data replace Galors takeoff
by a transition to a slower trajectory. This analysis and the absence of the takeoff in the corresponding GDP distribution
(Nielsen, 2016a) show that there was no takeoff in the income per capita (GDP/cap) distribution. Unified Growth Theory is
contradicted yet again by the same data which were used but not analysed during the formulation of this theory.
142
__________________________________________________________________________________
Puzzling properties of the historical growth rate of income per

capita explained
By Ron W. NIELSEN a
Abstract. Galor discovered many mysteries of the growth process. He lists them in his Unified Growth Theory
and wonders how they can be explained. Close inspection of his mysteries reveals that they are of his own creation.
They do not exist. One of his claimed mysteries is the mystery of the alleged sudden spurt in the growth rate of
income per capita and in the growth of population. This sudden spurt never happened. Precisely the same data,
which were used in support of the Unified Growth Theory are in fact in its direct contradiction. They show that
the created mysteries of growth do not exist. The difference between the diametrically opposite conclusions is that
in order to support the Unified Growth Theory and to create the mysteries of growth data were appropriately
manipulated and distorted but the contradicting evidence is based on their rigorous analysis. The mechanism of
the historical economic growth and of the growth of human population is yet to be explained but it would be
unproductive to try to look for explanations in the Unified Growth Theory. However, the problem is much deeper
than just the examination of this theory. Demographic Growth Theory is based on the incorrect but deeply
entrenched postulates developed by accretion over many years and now generally accepted in the economic and
demographic research, postulates revolving around the concept of Malthusian stagnation and around a transition
from stagnation to growth. The study presented here and earlier similar publications show that these postulates
need to be replaced by interpretations based on the mathematical analysis of data and on the correct understanding
of hyperbolic distributions.
Keywords. Economic growth, Population growth, Growth rates, Gross Domestic Product, Income per capita,
Unified Growth Theory, Hyperbolic growth
JEL. A10, A12, C02, C12, C50, F01, Y80.
1. Introduction
I n the subsection entitled Mysteries of the growth process (Galor, 2005a, p. 220) presented in his
Unified Growth Theory (Galor, 2005a, 2011), Galor asks a series of questions about the mysteries
of economic growth. We can take his questions one by one and show that all these mysteries were
of his own creation.
His theory is not based on the scientific analysis of data but on impressions supported by the
habitually distorted presentation of data (Ashraf, 2009; Galor, 2005a, 2005b, 2007, 2008a, 2008b,
2008c, 2010, 2011, 2012a, 2012b, 2012c; Galor & Moav, 2002; Snowdon & Galor, 2008). Such
approach to research can easily create many mysteries that simply do not exist.
One of Galors questions about the alleged mysteries of growth process is What is the origin of the
sudden spurt in growth rates of output per capita and population that occurred in the course of the take-
off from stagnation to growth? (Galor, 2005a, p. 220). In just one sentence, Galor presents two
incorrect doctrines: the doctrine of the presence of the sudden spurts and the doctrine of transition from
stagnation to growth, both created by the failure to follow scientific principles of investigation, which
require that data should not be manipulated to support preconceived ideas but that they should be
methodically analysed.
We shall show that this question makes as much sense as the question, Why does the sun rotate
around the earth, and the answer to both of them is similar: The sun does not rotate around the earth
and there was no sudden spurt in the growth rates of output per capita and population. There was also
a
Australia.
. +61407201175
Published as: Nielsen, R. W. (2016). Puzzling properties of the historical growth rate of income per capita explained.
Journal of Economics Library, 3(2), 241-256.
143
__________________________________________________________________________________
no takeoff from stagnation to growth because the growth was hyperbolic (Kapitza, 2006; Kremer, 1993;
Nielsen, 2014, 2015, 2016a, 2016b, 2016c, 2016d, 2016e; Podlazov, 2002; Shklovskii, 1962, 2002; von
Foerster, Mora & Amiot, 1960; von Hoerner, 1975). Hyperbolic growth is monotonic, and consequently
it is also characterised by the monotonically-increasing growth rate. There is no sudden spurt in this
type of distributions. These two doctrines expressed so confidently in just one sentence in Galors
publication are incorrect.
Output per capita (also described as income per capita and measured using the GDP/cap) is
represented by the ratio of two, monotonically-increasing, hyperbolic distributions (Nielsen, 2015). The
growth rate of this ratio is monotonic. It cannot contain the sudden spurt claimed erroneously by
Galor.
Galors questions about the mysteries of growth are strongly misleading because they describe
features created by the repeatedly distorted presentations of data. The created features and the associated
questions divert attention from the correct understanding of the mechanism of economic growth and of
the growth of population. Galors theory does not explain the mechanism of economic growth but
describes phantom features he created.
We have already discussed (Nielsen, 2014, 2015, 2016a, 2016b, 2016c, 5016d) various aspects of
Galors theory (Galor, 2005a, 2011). We shall now focus our attention on the discussion of his
unsubstantiated claims about the growth rates. We shall use precisely the same data (Maddison, 2001)
as used by Galor (2005a, 2011), who unfortunately did not analyse them.
Unified Growth Theory is fundamentally incorrect but it is just an embodiment of the incorrect
concepts used traditionally in the economic and demographic research, concepts developed by accretion
over many years and now so strongly entrenched that it will be probably difficult to uproot them and
replace them by correct interpretations. However, it is expected that it is in the interest of every
economist and demographer to have scientific basis for their research.
These erroneous interpretations revolve around the concept of Malthusian stagnation and around the
concept of transition from stagnation to growth. The study presented here and in earlier publications
(Kapitza, 2006; Kremer, 1993; Nielsen, 2014, 2015, 2016a, 2016b, 2016c, 2016d, 2016e; Podlazov,
2002; Shklovskii, 1962, 2002; von Foerster, Mora & Amiot, 1960; von Hoerner, 1975) demonstrate that
these traditional interpretations need to be replaced by interpretations based on the mathematical
analysis of data and on the correct understanding of hyperbolic distributions.
The latest data of Maddison (2001, 2010) serve as a rich source of information. They are in perfect
agreement with other similar data (e.g. Biraben, 1980; Clark,1968; Cook,1960; Durand, 1967, 1974,
1977; Gallant, 1990; Haub, 1995; Livi-Bacci, 1997; McEvedy & Jones, 1978; Taeuber & Taeuber,
1949; Thomlinson, 1975; Trager, 1994). When mathematically analysed, conclusions based on these
data are also in harmony with earlier similar studies (Kapitza, 2006; Kremer, 1993; Podlazov, 2002;
Shklovskii, 1962, 2002; von Foerster, Mora & Amiot, 1960; von Hoerner, 1975). Their combined
message is that the demographic and economic research has to be based on accepting the unambiguous
and consistent evidence in data that the historical growth of human population and of economic growth
were hyperbolic and that such a growth cannot be divided into two or three different regimes of growth
governed by distinctly different mechanisms of growth. Hyperbolic growth is slow over a long time and
fast over a short time but it is still the same growth governed by the same mechanism of growth.
Hyperbolic distributions have to be interpreted as a whole and not in parts. What appears as stagnation
is hyperbolic growth and what appears as takeoff or explosion is the natural continuation of the same
hyperbolic growth.
2. Fundamental mathematics
Growth rate R( S ) of a growing entity S can be defined as:
1 dS
R( S ) , (1)
S dt
where S can represent the GDP, the size of the population or any other growing entity.
144
__________________________________________________________________________________
Let us assume that we have two growing entities S1 and S2 , and that we want to calculate the growth
rate of the ratio of these two entities, i.e. the growth rate R( S1 S2 ) . It is easy to see that
1 d ( S1 S2 )
R ( S1 S 2 ) =
R ( S1 ) R( S2 ) . (2)
S1 S2 dt
We have obtained an interesting equation: the growth rate of the ratio of two distributions is the
difference between the growth rates of its two components. Thus, for instance, the growth rate of the
GDP/cap is given by the difference between the growth rate of the GDP and the growth rate of
population.
If two growing entities increase monotonically (as it is in the case of the historical economic growth
and of the historical growth of population) their growth rates also increase monotonically and
consequently the growth rate of their ratio, which is represented by the difference between the
monotonically-increasing growth rates of each of the two components, is also increasing monotonically.
It does not contain a sudden spurt.
Hyperbolic growth is described by the following simple differential equation:
1 dS
= kS , (3)
S dt
where S can represent the GDP or the size of the population, or indeed any other hyperbolically-
increasing entity, while k is a positive constant.
If we compere this differential equation with the general definition of the growth rate given by the
eqn (1) we can see that the characteristic feature of hyperbolic growth is that its growth rate is directly
proportional to the size of the growing entity:
R ( S ) = kS . (4)
The growth rate of hyperbolic distributions increases hyperbolically. The time dependence of the
growth rate of hyperbolic distributions creates precisely the same illusions as the time dependence of
hyperbolic growth (Nielsen, 2014). The growth rate of hyperbolic distribution is slow over a long time
and fast over a short time but it is a monotonically-increasing distribution, which cannot be divided into
mathematically-justifiable slow and fast components because the transition from slow to fast growth
occurs all the time along the entire time-range of such a distribution. The growth rate of hyperbolic
growth does not contain any sudden spurt at any time.
The equation (3) can be solved easily by substitution S = Z 1 . Its solution is:
1
S= , (5)
( a kt )
where a is a constant, which can be determined empirically by comparing the calculated curve with
data.
So, now, if we use the eqn (2) and if we assume that S1 and S2 are hyperbolic, then
1
S1 = (6)
(a1 k1t )
and
1
S2 = . (7)
( a2 k 2 t )
Consequently, by using the eqn (4) we have
145
__________________________________________________________________________________
k1
R ( S1 ) = (8)
a1 k1t
and
k2
R( S2 ) = . (9)
a2 k2t
If we now use these expressions in the eqn (2) we shall get
k k2
R ( S1 S 2 ) = 1 = , (10)
a1 k1t a2 k2t (a1 k1t )(a2 k2t )
where
=
k1a2 k2 a1 . (11)
The eqn (10) can be also presented as
1
R ( S1 / S 2 ) = , (12)
A0 + A1t + A2t 2
where
a1a2
A0 = , (13)

k a +k a
A1 = 1 2 2 1 , (14)

kk
A2 = 1 2 . (15)

So, while the growth rate of hyperbolic distributions is described by the reciprocal of a linear
function [see the eqns (8) and (9)] the growth rate of the ratio of hyperbolic distributions is described
by the reciprocal of the second-order polynomial [see the eqn (12)]. We could call it the second-order
hyperbolic distribution. It is a distribution, which resembles closely the first-order hyperbolic
distribution (the reciprocal of the linear function) because it also increases slowly over a long time and
escapes to infinity at a fixed time. However, it is a monotonically-increasing distribution, which does
not contain a sudden spurt.
It obviously makes no sense to claim a sudden spurt in the monotonically changing second-order
polynomial distribution and it obviously makes no sense to claim a sudden spurt in the reciprocal of the
second-order polynomial. The sudden spurt can be created by distorting data, as Galor did, but then it
is no longer science. Whether created on purpose to support preconceived ideas or carelessly, a similar
distorted presentation of evidence would be unacceptable even outside science. However, distortion of
evidence is not uncommon in defending doctrines accepted by faith. The distorted presentation of
empirical evidence makes the Unified Growth Theory (Galor, 2005a, 2011) scientifically unacceptable.
Another way to understand that the growth rate of the ratio of two hyperbolic distributions (e.g. the
growth rate of the GDP/cap distribution) cannot contain a sudden spurt is by looking at the denominator
of the eqn (10), which is given by a product of two linearly decreasing functions. Multiplication of two
linear distributions produces a monotonic distribution, which does not contain a sudden spurt.
Had Galor analysed the data (Maddison, 2001) he was using, he would have perhaps found that the
GDP and the size of the population were increasing hyperbolically. Maybe, then, it would have been
clear to him that monotonic distributions cannot be characterised by the non-monotonic sudden spurts.
Such an analysis should have been prompted by the discovery made over 50 years ago that the growth
of human population during the AD era was hyperbolic (von Foerster, Mora & Amiot, 1960). This vital
discovery, published in a well-known journal, is not even mentioned in Galors theory, maybe because
it was an inconvenient discovery. Such an analysis should have been also prompted by other related
studies (Kapitza, 2006; Kremer, 1993; Podlazov, 2002; Shklovskii, 1962, 2002; von Hoerner, 1975).
Finally, it should have been prompted by the data describing the historical growth of human population
(e.g. Biraben, 1980; Clark,1968; Cook,1960; Durand, 1967, 1974, 1977; Gallant, 1990; Haub, 1995;
146
__________________________________________________________________________________
Livi-Bacci, 1997; McEvedy & Jones, 1978; Taeuber & Taeuber, 1949; Thomlinson, 1975; Trager,
1994). All these works have been ignored and a scientifically unreliable theory has been created with
many mysteries of the growth process, mysteries that do not exist.
As in the case of hyperbolic distributions which can be studied easily by investigating the reciprocal
values of the size of the growing entity, 1/ S (Nielsen, 2014), an easy way to study the growth rate of
the ratio of hyperbolic distributions is by using the reciprocal values of the growth rate, 1/ R ( S1 S 2 ) .
As 1/ S converts the confusing hyperbolic distribution to a linear function, which is then easy to
understand, so also the reciprocal values, 1/ R ( S1 S 2 ) , convert the second-order hyperbolic distribution
into an easy-to-interpret second-order polynomial. In both cases, the confusing features, which create
the illusion of a slow growth over a long time followed by a sudden spurt disappear and are replaced
by much simpler distributions.
Confusing features of hyperbolic distributions can be also clarified by using semilogarithmic
displays. Such displays are routinely used for distributions, which vary over a large range of values. We
shall use them in our present discussion.

3.1. Growth of the GDP, population and income per capita (GDP/cap)
According to Galor, historical economic growth is characterised by takeoffs from stagnation to
growth, which occurred around AD 1750 for developed regions and around AD 1900 for less-developed
regions (Galor, 2008a, 2012a). He describes them as stunning or remarkable escapes from the
Malthusian trap (Galor, 2005a, pp. 177, 220). Such remarkable escapes should be readily identifiable
in data describing economic growth and the growth of human population. In particular, for data
describing the world economic growth and the growth of population, we should see clearly two takeoffs,
around AD 1750 and around AD 1900.
Results of our analysis of precisely the same data (Maddison, 2001) as used, but never analysed, by
Galor during the formulation of the Unified Growth Theory (Galor, 2005a, 2011), are presented in
Figures 1-3.
Figure 1. Data for the Gross Domestic Product (Maddison, 2001), precisely the same data as used but never
analysed by Galor during the formulation of the Unified Growth Theory (2005a, 2011), are compared with the
first-order hyperbolic distribution [eqn (5)]. The GDP is expressed in billions of 1990 International Geary-
=
Khamis dollars. Parameters describing the fitted hyperbolic distribution are: a 1.716 102 and
=k 8.671 106
147
__________________________________________________________________________________
Figure 2. Data describing the growth of the world population during the AD era (Maddison, 2001) are
compared with hyperbolic distribution. The large discrepancy at AD 1 is because of the maximum in the growth
of the world population around that year associated with the transition from a fast-increasing hyperbolic
distribution during the BC era to a slower hyperbolic distribution during the AD era (Nielsen, 2016b).
=
Parameters describing the fitted hyperbolic distribution are a 8.724 100 and
= k 4.267 103 .
The alleged takeoff from the assumed stagnation to growth for developed countries coincides with
the onset of the Industrial Revolution, AD 1760-1840 (Floud & McCloskey, 1994), which according to
Galor was the prime engine of economic growth (Galor, 2005a, p. 212). It is, therefore, yet another
reason why the takeoff for developed countries and the associated sudden spurt in the growth rates
should be easy to identify because the alleged prime engine should have been working most effectively
in these countries.
Figure 3. World income per capita (GDP/cap). Data (Maddison, 2001) are compared with the linearly-
modulated hyperbolic distribution (Nielsen, 2015) representing the ratio of hyperbolic distributions of the GDP
and of the size of population. Income per capita was increasing monotonically. Such monotonic increase cannot
produce a non-monotonic growth rate claimed by Galor (2005a, 2011). It cannot produce the sudden spurt in
the growth rates of output per capita (Galor, 2005a, p. 220). His stunning or remarkable takeoffs from
stagnation to growth (Galor, 2005a, pp. 177, 220) did not happen. Industrial Revolution, the prime engine of
economic growth (Galor, 2005a, p. 212) had no impact on changing the economic-growth trajectory. All these
stories are contradicted by data (Maddison, 2001), precisely the same data as used but never analysed during
the formulation of the Unified Growth Theory.
Economic growth, as described by the Gross Domestic Product (GDP) shown in Figure 1, was
hyperbolic. The alleged prime engine of economic growth (Galor, 2005a, p. 212) did nothing to
change the growth trajectory. This is an interesting issue, which requires further investigation because
148
__________________________________________________________________________________
technological discoveries were used to support economic growth but surprisingly perhaps they had
absolutely no impact on changing the growth trajectory. It is as if economic growth was controlled by
some other unknown and much stronger force, which was active before the Industrial Revolution and
remained undisturbed during and after the Industrial Revolution. The alleged takeoffs from stagnation
to growth did not happen because there was no stagnation. Economic growth was hyperbolic before and
after the alleged takeoffs. The takeoffs claimed by Galor simply did not exist.
The growth of population during the AD era, shown in Figure 2, was also hyperbolic, at least from
around AD 1000, in perfect agreement with the discovery made over 50 years ago by von Foerster,
Mora and Amiot (1960). The discrepancy at AD 1 is explained by the analysis of the growth of human
population in the past 12,000 years (Nielsen, 2016b), which revealed a maximum around that year
associated with the transition from a fast-increasing hyperbolic growth during the BC era to a slower
hyperbolic growth during the AD era. This extended analysis demonstrated that there was an
uninterrupted hyperbolic growth between 10,000 BC and around 500 BC, followed by a transition to a
new hyperbolic growth commencing around AD 500. It also revealed a small disturbance of the
hyperbolic growth between AD 1200 and 1400. The data show that during the past 12,000 years there
was no stagnation and no sudden takeoff at any time, both in the growth of the population and in the
growth rate.
It is remarkable that so many independent studies are in such perfect agreement: Maddisons data
(Maddison, 2001, 2010) and their analysis (Nielsen, 2014, 2015, 2016a, 2016c, 2016d); the estimates
of the size of human population not only during the AD era but also during the BC era (e.g. Biraben,
1980; Clark,1968; Cook,1960; Durand, 1967, 1974, 1977; Gallant, 1990; Haub, 1995; Livi-Bacci, 1997;
McEvedy & Jones, 1978; Taeuber & Taeuber, 1949; Thomlinson, 1975; Trager, 1994) and their analysis
(e.g. Kremer, 1993; Nielsen, 2016b; Kapitza, 2006); the discovery made by von Foerster, Mora and
Amiot (1960) and similar identifications of hyperbolic growth by Podlazov (2002), Shklovskii (1962,
2002) and von Hoerner (1975).
In contrast, Unified Growth Theory and the generally accepted but questionable postulates used in
the economic and demographic research describe events and processes occurring in the fictitious world
characterised by Malthusian stagnation, takeoffs, sudden spurts and by the remarkable or stunning
escapes from the Malthusian traps (Galor, 2005a, pp. 177, 220), the world which is entirely different
than the world revealed by data and by their mathematical analysis.
There appears to be no formal definition of Malthusian stagnation. Maybe some kind of stagnation
can be used to describe social conditions but this concept is totally irrelevant in the study of the
mechanism of economic growth and of the growth of human population because they were hyperbolic.
There was no stagnation and no transition from the imagined stagnation to growth. Using such
descriptions to explain the mechanism of the historical economic growth or the historical growth of
population is unscientific because these postulates are consistently contradicted by data.
Results of analysis of income per capita (GDP/cap) presented in Figure 3 also demonstrate a
monotonically-increasing distribution at least from AD 1500, i.e. during the time when Galors
remarkable and stunning effects should be clearly visible. What is remarkable about this
distribution is that nothing remarkable or stunning ever happened. The growth of the GDP/cap was
remarkably stable. Industrial Revolution did not accelerate the growth of income per capita and there
were no sudden takeoffs at any time.
Such monotonically-increasing distributions, as presented in Figures 1-3, cannot generate the
sudden spurt (Galor, 2005a, p. 220) in the corresponding growth rates and we shall now see that they
indeed did not.
3.2. Growth rates of the GDP, population and income per capita (GDP/cap)
Results of analysis of growth rates are shown in Figures 4-6. Empirical growth rates were calculated
using Maddisons data (Maddison, 2001) and interpolated gradients. The predicted growth rates were
calculated using the fitted distributions shown in Figures 1-3.
As expected, the growth rate of the world GDP and population were increasing monotonically.
Industrial Revolution did not accelerate their growth. The remarkable or stunning escapes from the
Malthusian trap (Galor, 2005a, pp. 177, 220), which were supposed to have been reflected in takeoffs
from stagnation to growth, never happened because there was no stagnation and the trap did not exist
in the economic growth and in the growth of population. Growth rates were increasing along remarkably
robust trajectories.
149
__________________________________________________________________________________
Analysis of growth rates shows a remarkable contradiction of Galors claims by precisely the same
data, which he used, but never analysed, during the formulation of his Unified Growth Theory. His
wished-for and claimed features never happened. Growth rates were increasing monotonically. There
was absolutely no sudden spurt at any time.
In order to support his preconceived ideas, Galor ignored not only the analysis carried out over 50
years ago (von Foerster, Mora & Amiot, 1960) but also research contributions of his many other
predecessors (Biraben, 1980; Clark,1968; Cook,1960; Durand, 1967, 1974, 1977; Gallant, 1990; Haub,
1995; Kapitza, 2006; Kremer, 1993; Livi-Bacci, 1997; McEvedy & Jones, 1978; Podlazov, 2002;
Shklovskii, 1962, 2002; Taeuber & Taeuber, 1949; Thomlinson, 1975; Trager, 1994; von Hoerner,
1975). Galors claims are in conflict with science. They are not just unsupported by science they are
repeatedly contradicted by the scientific analysis of data and most notably by the analysis of precisely
the same data, which he used during the formulation of his theory.
Figure 4. Growth rate of the world GDP was increasing monotonically. There was no sudden spurt. The
claimed takeoffs did not happen. Industrial Revolution did not accelerate economic growth.
Figure 5. Growth rate of the world population. Empirical growth rate calculated using Maddisons data
(Maddison, 2001) and interpolated gradients is compared with the predicted growth rate calculated using
parameters of the fitted hyperbolic distribution displayed in Figure 2. Galors claims (Galor, 2005a, 2011) are
remarkably contradicted by the analysis of Maddisons data (Maddison, 2001), precisely the same data, which
he used but never properly analysed. Galors mystery of the sudden spurt in the growth rate of population
(Galor, 2005a, p. 220) is solved there was no sudden spurt.
150
__________________________________________________________________________________
Figure 6. Growth rate of the world income per capita (GDP/cap). Empirical growth rate calculated using
Maddisons data (Maddison, 2001) and interpolated gradients is compared with the predicted growth rate
calculated using parameters of the fitted hyperbolic distributions displayed in Figures 1 and 2. Galors claims
(Galor, 2005a, 2011) are remarkably contradicted by the analysis of Maddisons data (Maddison, 2001),
precisely the same data, which he used but never properly analysed. Galors mystery of the sudden spurt in
the growth rate of income per capita (Galor, 2005a, p. 220) is solved there was no sudden spurt.
Mathematical analysis of Maddisons data (Maddison, 2001), precisely the same data as used by
Galor, solves also his mystery of the sudden spurt in growth rates of output per capita (Galor, 2005a,
p. 220) there was no spurt. Results of analysis are presented in Figure 6. Growth rate of income per
capita (GDP/cap) was increasing monotonically. Industrial Revolution did not accelerate the growth of
income per capita. The postulated takeoffs from stagnation to growth (yet another mystery of growth
claimed by Galor) did not happen because there was no stagnation and because the growth rate was
increasing steadily without any major interruption. The only real mystery is why the growth rate of
income per capita was so remarkably stable over such a long time.
Hyperbolic distributions, which increase monotonically, are characterised by monotonically-
increasing growth rates, as shown in Figures 5-6. Claiming the existence of sudden spurts in such
distributions is scientifically unjustifiable. Going a step further and claiming that such an imaginary
spurt is a mystery, which needs to be explained encourages other researchers to carry out pointless and
unproductive research.
It is useful to compare the mathematical analysis of Maddisons data presented in Figure 6 with the
distorted presentation used by Galor reproduced in Figure 7. Both figures are based on precisely the
same set of data (Maddison, 2001). The contrast is striking. Now we can better appreciate the lack of
scientific support for his theory. It is surprising that his theory was ever published. It is also surprising
that similar distorted diagrams and the associated unscientific claims and explanation were published
in peer-reviewed literature (Galor, 2005b, 2007, 2010, 2011, 2012b; Galor & Moav, 2002; Snowdon &
Galor, 2008)
While data and their analysis, displayed in Figure 6, present monotonically-increasing growth rate
of income per capita, Galors distorted presentation of precisely the same data show a clear sudden
spurt. Maybe Galor was so strongly guided by the traditional interpretations of economic growth that
he could not accept the clear contradicting evidence or maybe he simply did not know how to analyse
data. In any case, intentional or unintentional, such ubiquitous distorted diagrams used repeatedly in his
theory can be hardly expected to lead to reliable conclusions. On the contrary, they can be expected to
lead only to incorrect conclusions.
151
__________________________________________________________________________________
Figure 7. Galors distorted, strongly suggestive and misleading presentation of Maddisons data (Maddison,
2001) describing the growth rate of output (income) per capita (Galor, 2005a, p. 179). Precisely the same data,
when correctly displayed and analysed (see Figure 6), show that the sudden spurt in the growth rate of output
per capita claimed by Galor (2005a, p. 220) did not exist.
Galor gives also many isolated examples of small growth rates in the past and significantly larger
values at a later stage of growth but all these examples are not only meaningless but also strongly
misleading. They reflect nothing more than just the natural properties of hyperbolic distributions. Using
them to prove stagnation and transitions from stagnation to growth is scientifically irresponsible.
Of course, growth rates of income per capita (GDP/cap) were small over a long time and significantly
larger at a certain later stage of growth because they were following monotonically-increasing second-
order hyperbolic distributions [see eqn (12)]. Hyperbolic distributions (second-order or first-order) are
slow over a long time and fast over a short time but they are still the same, monotonically-increasing
distributions. They are not characterised by sudden spurts. There is no profound mystery about them
that needs to be explained by some elaborate research or mathematical formulations. It is just a simple
and straightforward hyperbolic growth. The mystery is solved. Picking up some isolated numbers from
such hyperbolic distributions and drawing some profound conclusions based on such examples is
unscientific. The only mystery that needs to be explained is why the economic growth and the growth
of population were hyperbolic and why they were so remarkably stable (undisturbed) over such a long
time.

Galor discovered many mysteries of the growth process (Galor, 2005a, p. 220). One of his
mysteries is the sudden spurt in growth rates of output per capita and population that occurred in the
course of the take-off from stagnation to growth (Galor, 2005a, p. 220).
His discoveries are based on the crude and distorted presentations of data (Ashraf, 2009; Galor,
2005a, 2005b, 2007, 2008a, 2008b, 2008c, 2010, 2011, 2012a, 2012b, 2012c; Galor & Moav, 2002;
Snowdon & Galor, 2008). His mysteries are of his own creation. They do not need to be explained
because they do not exist. They describe the world of fiction.
Historical economic growth and the growth of human population were hyperbolic (Kapitza, 2006;
Kremer, 1993; Nielsen, 2014, 2016a, 2016b, 2016c, 2016d; Podlazov, 2002; Shklovskii, 1962, 2002;
von Foerster, Mora & Amiot, 1960; von Hoerner, 1975). Hyperbolic distributions are monotonic and
they are characterised by the monotonically-increasing growth rates. Sudden spurts do not exist in such
distributions.
It is essential to understand that it is incorrect to take a slowly-increasing distribution and
automatically claim the evidence of Malthusian stagnation. The state of stagnation might occur when
there is a strong interference between a primary force propelling growth and some random opposing
forces. Effects of stagnation should be reflected in the growth trajectory, which should be clearly
152
__________________________________________________________________________________
unstable. If the growth follows a steadily-increasing trajectory without any clear signs of random
behaviour, then there is no need to complicate the description of the mechanism of growth by
introducing random forces whose presence is undetectable. The fundamental principle of scientific
investigation is to look for the simplest descriptions and explanations. Introducing unnecessary
complications is simply unscientific.
It appears that the established knowledge in demography and in economic research is strongly based
on a series of postulates revolving around the concept of Malthusian stagnation and around the alleged
transition from the imagined stagnation to growth. Complicated mechanisms and interactions are then
used (and gradually made even more complicated) to explain the mechanism of growth. Galor went one
step further and reinforced these incorrect interpretations by his persistently distorted presentations of
data (such as shown in Figure 7) and by his repeated quotations of certain well-selected figures to
support his preconceived ideas, figures which were supposed to illustrate the concepts of stagnation and
takeoffs from stagnation to growth but when closely analysed illustrate nothing else than the simple
hyperbolic growth. All such complicated explanations are contradicted by data. Close examination of
data shows that there was no stagnation and no transition from stagnation to growth. Data show
consistently that the mechanism of the economic growth and of the growth of human population must
have been simple because hyperbolic growth is exceptionally simple [see eqn (5)].
Some types of growth might be slow and stagnant but hyperbolic growth is not stagnant even when
it is slow. It is prompted by the same mechanism during the time when it is slow and when it is fast. If
the mechanism of Malthusian stagnation is used to explain the slow hyperbolic growth, precisely the
same mechanism should be used to explain the fast growth, which is commonly described as explosion.
It is incorrect to divide hyperbolic distributions into two or three sections and assign different
mechanisms of growth to each of such arbitrarily selected sections. Hyperbolic distributions have to be
explained as a whole and the same mechanism should be applied to the apparent slow and to the
apparent fast sections.
It is incorrect to take a hyperbolic distribution and look for a sudden takeoff from the imagined
stagnation to growth, as Galor did repeatedly. It is impossible to divide hyperbolic distribution into such
distinctly different sections and the best way to see it, is by using the reciprocal values (Nielsen, 2014)
because hyperbolic distribution is then represented by a decreasing straight line and it is obvious that it
is impossible to claim a change of direction on the straight, which shows no change of direction.
Hyperbolic growth is not the only type of growth that can be slow over a certain time but not
stagnant. Exponential growth is initially slow but it gradually becomes faster. At a certain stage, as if
suddenly, it becomes overwhelmingly fast, the effect described as the second half of the chessboard
(Kurzweil, 1999). Logistic growth is also initially slow but it is not stagnant.
The difference between hyperbolic and exponential distributions is that for hyperbolic distributions
the apparent (but non-existent) transition from a slow to fast growth is more clearly articulated. That is
why hyperbolic distributions are so often misinterpreted, particularly if they are distorted as it is done
repeatedly and persistently in Galors publications. However, this apparent transition from slow to fast
growth does not happen at any given time or even over a certain specific range of time. It happens all
the time. The acceleration is gradual along the entire range of hyperbolic distribution.
Growth of income per capita (GDP/cap) is represented by the ratio of two hyperbolic distributions
(Nielsen, 2015). The ratio of monotonically increasing- distributions is characterised by the
monotonically-increasing growth rate. We have shown that while the growth rate of the GDP and
population increases hyperbolically, the growth rate of income per capita (GDP/cap) increases by
following a second-order hyperbolic distribution (the reciprocal of the second order polynomial). There
is no sudden spurt in any of these distributions and we have demonstrated this point by using the world
economic growth and the growth of human population.
When doctrines accepted by faith are defended, scientific rules of engagement are readily violated.
Contradicting data are then either ignored or manipulated to support preconceived ideas. Economic and
demographic research has no place for this type of activities. However, intentionally or unintentionally,
such unscientific approach to research appears to have been adopted during the formulation of the
Unified Growth Theory (Galor, 2005a, 2011). Numerous preceding research works (e.g. Biraben, 1980;
Clark,1968; Cook,1960; Durand,1967, 1974, 1977; Gallant, 1990; Haub, 1995; Kapitza, 2006; Kremer,
1993; Livi-Bacci, 1997; McEvedy & Jones, 1978; Podlazov, 2002; Shklovskii, 1962, 2002; Taeuber &
Taeuber, 1949; Thomlinson, 1975; Trager, 1994; von Foerster, Mora and Amiot, 1960; von Hoerner,
153
__________________________________________________________________________________
1975) were ignored and the excellent data of Maddison (2001) were manipulated and distorted to
support a series of preconceived ideas.
Hyperbolic distributions may be confusing but there is no excuse for distorting them to make them
even more confusing. There is also no excuse for failing to analyse hyperbolic distributions because
their analysis is trivially simple (Nielsen, 2014). The analysis of the growth rates is in the same category.
Graphically, all these distributions become abundantly clear by using either the semilogarithmic scales
of reference of by displaying the reciprocal values of growing entities or of their corresponding growth
rates.
Galors Unified Growth Theory is scientifically unacceptable and so are also many traditional
interpretations of economic growth and of the growth of human population, interpretations based on the
incorrect understanding of the mathematical properties of hyperbolic distributions. The recent and
readily-accessible Maddisons data (Maddison, 2001, 2010) make it now easy to study the mechanism
of the historical economic growth and of the growth of human population. They demonstrate that certain
fundamental postulates revolving around the concept of Malthusian stagnation and around the assumed
transition from the non-existent stagnation to growth, still used routinely in the established knowledge
in demography and in economic research, are no longer acceptable.
References
Durand, J. (1967). A Long-range View of World Population Growth. The Annals of the American Academy of Political and
Social Science: World Population, 369, 2.
Durand, J. (1977). Historical Estimates of World Population: An Evaluation," Population and Development Review, 3(3),
256.
Floud, D. & McCloskey, D. N. (1994). The Economic History of Britain since 1700. Cambridge: Cambridge University
Press.
Publishing Company.
475.
0062-7
117, 1133-1191. doi. 10.1162/003355302320935007
Economics, 108(3), 681716.
Kurzweil, R. (1999). The Age of Spiritual Machines: When Computers Exceed Human Intelligence. New York: Penguin.
154
__________________________________________________________________________________

McEvedy, C. & Jones, R. (1978). Atlas of World Population History. Middlesex, England: Penguin.
Nielsen, R. W. (2014). Changing the paradigm. Applied Mathematics, 5, 1950-1963. doi. 10.4236/am.2014.513188
Nielsen, R. W. (2016b). Growth of the world population in the past 12,000 years and its link to the economic growth.
Nielsen, R. W. (2016c). The postulate of the three regimes of economic growth contradicted by data. Journal of Economic
Nielsen, R. W. (2016e). Demographic Transition Theory and its link to the historical economic growth. Journal of Economic
Taeuber, C. & Taeuber, I. B. (1949). World Population Trends. Journal of Farm Economics, 31(1), 241.
Pub.
691-712.
155
__________________________________________________________________________________
10
Mathematical analysis of historical income per capita

distributions
By Ron W. NIELSENa
Abstract. Data describing historical growth of income per capita [Gross Domestic Product per capita (GDP/cap)]
for the world economic growth and for the growth in Western Europe, Eastern Europe, Asia, former USSR, Africa
and Latin America are analysed. They follow closely the linearly-modulated hyperbolic distributions represented
by the ratios of hyperbolic distributions obtained by fitting the GDP and population data. Results of this analysis
demonstrate that income per capita was increasing monotonically. There was no stagnation and there were no
transitions from stagnation to growth. The usually postulated dramatic escapes from the Malthusian trap never
happened because there was no trap in the economic growth. Unified Growth Theory is fundamentally incorrect
because its central postulates are contradicted repeatedly by data, which were used but never analysed during the
formulation of this theory. The large body of readily-available data opens new avenues for the economic and
demographic research. They show that certain fundamental postulates revolving around the concept of Malthusian
stagnation need to be replaced by the evidence-based interpretations. Within the range of analysable data, which
for the growth of population extends down to 10,000 BC, growth of human population and economic growth were
hyperbolic. There was no Malthusian stagnation and there were no transitions to distinctly faster trajectories.
Industrial Revolution had no impact on changing growth trajectories.
Keywords. Economic growth, Income per capita, Unified Growth Theory, Hyperbolic growth, Industrial
Revolution, Takeoffs, Malthusian stagnation
JEL. A10, C12, C20, C50, F00.
1. Introduction
T he aim of this publication is to present the direct proof that contrary to the fundamental postulate
of the Unified Growth Theory (Galor, 2005a, 2011) distributions describing historical growth of
income per capita cannot be divided into three, distinctly-different regimes of growth governed
by distinctly different mechanisms. The indirect proof was presented earlier (Nielsen, 2016a, 2016c,
2016d,) by showing that the historical growth of the Gross Domestic Product (GDP) and of human
population were hyperbolic and that postulated by Galor three regimes of growth did not exist.
Mathematical analysis of the latest data (Maddison, 2001, 2010) brings a new insight into the
interpretation of the historical economic growth. Within the range of analysable data there was no
Malthusian stagnation, no alleged takeoffs from stagnation to growth and no escapes from the
hypothetical Malthusian trap because there was no trap (Nielsen, 2014, 2015a, 2016a, 2016b, 2016c,
2016d, 2016e, 2016f).
Unified Growth Theory serves as a good example of traditional interpretations of economic growth,
interpretations revolving around the concept of Malthusian stagnation. It is also a theory, which appears
to be based on Maddisons data (Maddison, 2001) but it is not. Ironically, even though these excellent
data were used during the formulation of this theory, they were never mathematically analysed. Unified
Growth Theory is not based on the scientific analysis of data but on impressions supported by the
habitually distorted presentation of data (Ashraf, 2009; Galor, 2005a, 2005b, 2007, 2008a, 2008b,
2008c, 2010, 2011, 2012a, 2012b, 2012c; Galor & Moav, 2002; Snowdon & Galor, 2008). Data were
used unprofessionally and they were presented in such a way, intentionally or unintentionally, that they
a
Australia.
. +61407201175
Published as: Nielsen, R. W. (2016). Mathematical analysis of the historical income per capita distributions. Economic
Review, 3(2), 300-319.
156
__________________________________________________________________________________
appear to support preconceived ideas. However, when closely analysed they are found to be in the direct
contradiction of the promoted concepts, explanation and interpretations.
Historical economic growth and historical growth of human population were hyperbolic (Kapitza,
2006; Kremer, 1993; Nielsen, 2016a, 2016b, 2016d; Nielsen, 2014, 2015a, 2016a, 2016b, 2016c, 2016d,
2016e, 2016f; Podlazov, 2002; Shklovskii, 1962, 2002; von Foerster, Mora and Amiot, 1960; von
Hoerner, 1975)). Hyperbolic distributions are confusing and they are often misinterpreted in studies of
economic growth and of the growth of human population. They present an image of a slow growth over
a long time followed by a fast growth over a short time. These distributions are, therefore usually
divided into two distinctly-different segments, slow and fast. The selected slow segment is then claimed
to represent the epoch of Malthusian stagnation while the selected fast segment is assumed to represent
an entirely new type of growth. The alleged transition between these two arbitrarily-selected segments
is then described as explosion, takeoff, sudden spurt, sprint or the dramatic escape from the Malthusian
trap. Distinctly-different mechanisms are also assigned for the two perceived segments of growth.
Often, however, interpretations of historical growth are not even based on any attempt to examine
relevant data. Isolated examples are used to support the concept of stagnation followed by explosion.
Even worse, more often than not, interpretations and explanations are just based on impressions and
suppositions. Claims of the existence of Malthusian stagnation and transitions to different stages of
growth are supported by a good dose of creative imagination.
There is no mathematically justifiable reason for dividing hyperbolic distributions into two or three
distinctly-different components (Nielsen, 2014). It is mathematically impossible to divide hyperbolic
distributions into slow and fast components. Hyperbolic distributions are slow over a long time and fast
over a short time but they increase monotonically. Growth rate also increases monotonically without
any unusual acceleration at any time. It increases hyperbolically with time or linearly with the size
growing entity (Nielsen, 2016f). Concepts of stagnation and takeoffs from stagnation to growth are
scientifically unjustified. They are contradicted by the analysis of data describing economic growth and
the growth of population (Kapitza, 2006; Kremer, 1993; Nielsen, 2014, 2015a, 2016a, 2016b, 2016c,
2016d, 2016e, 2016f; Podlazov, 2002; Shklovskii, 1962, 2002; von Foerster, Mora and Amiot, 1960;
von Hoerner, 1975).
Hyperbolic distributions have to be interpreted as a whole and the same mechanism has to be used
for the apparent slow and for the apparent fast segments. These segments do not exist even though they
appear to exist. The best way to demonstrate that these segments do not exist is by using reciprocal
values of hyperbolic distributions (Nielsen, 2014).
Historical economic growth is even more confusing than the historical growth of human population
because economic growth is often described using income per capita represented by the Gross Domestic
Product per capita (GDP/cap). It is a ratio of hyperbolic distributions and it creates an even stronger
illusion of different stages of growth than the illusion created by hyperbolic distributions. It has been
demonstrated (Nielsen, 2015a) that the characteristic features of the GDP/cap distributions, which are
interpreted as the epoch of stagnation followed by a sudden takeoff, are nothing more than mathematical
properties of dividing two hyperbolic distributions. It is incorrect to claim that these features
characterise uniquely economic growth.
The ratio of two hyperbolic distributions, which includes the GDP/cap ratio, increases monotonically
and there is no mathematically-justifiable reason for dividing them into distinctly different regimes of
growth. There is no mathematical justification for assigning different mechanisms of growth to the two
perceived but non-existing segments of income per capita distributions.
Growth of income per capita was slow over a long time and fast over a short time but it was
increasing monotonically. The ratio of monotonically-increasing hyperbolic distributions can only
produce a monotonically-changing distribution, increasing or decreasing, depending on the location of
singularities of hyperbolic distributions (Nielsen, 2015a). Such a ratio cannot produce a distribution
with a sudden discontinuity, which could be described as a takeoff.
The growth of income per capita has to be explained by using the same mechanism for the whole
distribution, slow and fast. We shall now demonstrate that the empirical distributions describing income
per capita were indeed increasing monotonically and that there were no sudden takeoffs from stagnation
to growth as claimed incorrectly in the Unified Growth Theory (2005a, 2011).
157
__________________________________________________________________________________

Maddisons data (Maddison, 2001, 2010) offer an unprecedented opportunity to test the past and
present explanations of economic growth and of the growth of human population, explanations based
on strongly-limited sources of empirical information and on creative imagination. Now, this rich body
of data brings new and refreshing insights into the interpretation of the historical economic growth and
of the growth of population, interpretations confirmed by other sources of relevant data (Biraben, 1980;
Clark,1968; Cook,1960; Durand, 1967, 1974, 1977; Gallant, 1990; Haub, 1995; Livi-Bacci, 1997;
McEvedy & Jones, 1978; Taeuber & Taeuber, 1949; Thomlinson, 1975; Trager, 1994) and by the earlier
research (Kapitza, 2006; Kremer, 1993; Podlazov, 2002; Shklovskii, 1962, 2002; von Foerster, Mora
and Amiot, 1960; von Hoerner, 1975)
It is both unfortunate and ironic that Galor had access to all this information but failed to use it to
make useful contribution to science. He repeatedly distorted empirical distributions to support his
preconceived ideas. An example of such distorted and self-misleading presentations of data is shown in
Figure 1 but remarkably, precisely the same data, when properly analysed, demonstrate that the Unified
Growth Theory is fundamentally incorrect.
Hyperbolic distributions do not have to be distorted to be confusing. They are already sufficiently
confusing and it is easy to make mistakes with their interpretations. Distorted presentations, such as
repeatedly used by Galor, make the interpretation of these distributions even more difficult. The
example presented in Figure 1 is based on a figure presented by Galor (2005a, p. 181). Such self-
misleading presentations of data can be expected to lead inevitably to incorrect conclusions. It is hard
to understand why such distorted diagrams were repeatedly used by Galor because the analysis of
hyperbolic distributions is trivially simple (Nielsen, 2014).
The fundamental postulates of the Unified Growth Theory are based on the assumption of the
existence of three, distinctly-different regimes of economic growth: Malthusian regime of stagnation,
post-Malthusian regime and the sustained-growth regime. According to Galor (2005a, 2008a, 2011,
2012a), Malthusian regime of stagnation was between 100,000 BC and AD 1750 for developed regions
and between 100,000 BC and AD 1900 for less-developed regions. The post-Malthusian regime was
allegedly between AD 1750 and 1850 for developed regions and from 1900 for less-developed regions.
The sustained-growth regime was supposed to have commenced around 1850 for developed regions.
Figure 1. Example of the ubiquitous, grossly-distorted and self-misleading diagrams used to create the Unified
Growth Theory (Galor, 2005a, 2011). Maddisons data (Maddison, 2001) were used during the formulation of
this theory but they were never analysed. Such state-of-the-art was used to construct a system of scientifically-
unsupported interpretations, explanations and mysteries of the growth process (Galor, 2005a, p. 220).
The end of the regime of Malthusian stagnation was supposed to have been characterised by dramatic
takeoffs from stagnation to growth, described as a remarkable or stunning escape from the
Malthusian trap (Galor, 2005a, pp. 177, 220). It is a signature, which cannot be missed. This change in
the pattern of growth is described as the sudden take-off from stagnation to growth (Galor, 2005a,
pp. 177, 220, 277) or as a sudden spurt (Galor, 2005a, 177, 220). According to Galor, the end of the
Malthusian regime of stagnation for developed regions coincides with the Industrial Revolution. The
158
__________________________________________________________________________________
take-off of developed regions from the Malthusian Regime was associated with the Industrial
Revolution (Galor, 2005a, p. 185). Indeed, the Industrial Revolution is considered to have been the
prime engine of economic growth (Galor, 2005a, p. 212).
For developed regions, the postulated sudden takeoffs from stagnation to growth is supposed to have
occurred around AD 1750, or around the time of the Industrial Revolution, 1760-1840 (Floud &
McCloskey, 1994). For less-developed regions, the takeoff was allegedly around 1900. A transition
from growth to growth is not a signature of the postulated sudden takeoff from stagnation to growth.
Thus, for instance, a transition from hyperbolic growth to another hyperbolic growth or to some other
steadily-increasing trajectory is not a signature of the sudden takeoff from stagnation to growth.
In the diagrams presented below, income per capita (GDP/cap) is in 1990 International Geary-
Khamis dollars.
3. Hyperbolic growth
It has been shown earlier that over the range of analysable data historical growth of population and
historical economic growth were hyperbolic (Nielsen, 2014, 2015a, 2016a, 2016b, 2016c, 2016d,
2016e, 2016f). For the economic growth, the range of analysable data extends down to AD 1 but for the
growth of the world population it extends to 10,000 BC. These results are consistent with the analysis
carried out over 50 years ago for the growth of the world population during the AD era (von Foerster,
Mora, & Amiot, 1960) and with other similar studies (Kapitza, 2006; Kremer, 1993; Podlazov, 2002;
Shklovskii, 1962, 2002; von Hoerner, 1975)
Demographic and economic research has to be based on the acceptance of hyperbolic descriptions
of the historical growth of population and of the historical economic growth. Hyperbolic growth,
confirmed repeatedly and consistently by data (Biraben, 1980; Clark,1968; Cook,1960; Durand, 1967,
1974, 1977; Gallant, 1990; Haub, 1995; Livi-Bacci, 1997; Maddison, 2001, 2010; McEvedy & Jones,
1978; Taeuber & Taeuber, 1949; Thomlinson, 1975; Trager, 1994) leaves no room for the outdated
interpretations revolving around the concept of Malthusian stagnation followed by sudden takeoffs to a
distinctly faster growth. Mathematical analysis of data contradicts consistently and repeatedly this
hypothetical but unsupported concepts, including the concept that the Industrial Revolution changed
the relevant growth trajectories. It did not.
1
S (t ) = , (1)
a kt
where S (t ) is the size of the hyperbolically growing entity (e.g. the GDP or the size of the population),
while a and k are positive constants.
Distribution describing the time-dependence of income per capita (GDP/cap) is the ratio of two
hyperbolic distributions: the hyperbolic distribution describing the growth of the GDP (Nielsen, 2016a)
and the hyperbolic distribution describing the growth of population (Nielsen, 2016d). A GDP/cap
distribution can be also interpreted as a ratio of two linearly decreasing distributions describing the
respective reciprocal values or as a product of a hyperbolic distribution representing the GDP and a
linear function representing the reciprocal values of the size of the population. Consequently, the
GDP/cap ratio can be simply described as the linearly-modulated hyperbolic distribution, where the
linear modulation of the GDP distribution is done by the reciprocal values of the size of population
(Nielsen, 2015a).
4. Growth of the world GDP/cap

Results of mathematical analysis of the world GDP/cap are presented in Figure 2. The fitted
distribution represents the linearly-modulated GDP distribution (Nielsen, 2015a). Parameters
describing the GDP data are=a 1 1.684 102 and=k 1 8.539 106 while parameters describing the
=
world population data are k 2 3.765 103 .
a 2 7.739 100 and=
159
__________________________________________________________________________________
For the growth of the world GDP/cap we should see the signature of two takeoffs: around AD 1750
for developed regions and around AD 1900 for less-developed regions, yet we see none of them. There
was no stagnation before the Industrial Revolution and no transition from stagnation to growth around
AD 1750 for developed regions or around AD 1900 for less-developed regions, as claimed by Galor
(2008a, 2012a).
Figure 2. Maddisons data (Maddison, 2010) describing growth of the world income per capita (GDP/cap) are
compared with the linearly-modulated hyperbolic distribution (Nielsen, 2015a) obtained by fitting the GDP and
population data (Nielsen, 2016a, 2016d). The alleged takeoffs from stagnation to growth around AD 1750 for
developed regions and around 1900 for less-developed regions, as claimed by Galor (2008a, 2012a), did not
happen. Industrial Revolution had no impact on changing the trajectory describing the growth of income per
capita (GDP/cap).
The data show a minor disturbance around AD 1900 which looks like a minor boosting. However,
it is definitely not a remarkable or stunning escape from the Malthusian trap (Galor, 2005a, pp.
177, 220) because (1) the growth deviated only slightly from the historical trajectory, (2) this minor
deviation was not preceded by stagnation and (3) because it was only temporary disturbance and the
growth soon returned to the original trajectory. Furthermore, rather than being permanently and
spectacularly propelled along a distinctly new trajectory, as implied by Galors claims of remarkable
and stunning takeoffs (Galor, 2005a, pp. 177, 220), economic growth as described by data, started to
be diverted to a slower trajectory. There was definitely no transition from stagnation to growth. There
was no dramatic escape from the Malthusian trap because there was no trap in the growth of income
per capita.
5. Western Europe
Growth of the GDP/cap in Western Europe is shown in Figure 3. Maddisons data (Maddison, 2010)
are compared with the linearly-modulated hyperbolic distribution obtained by dividing two hyperbolic
distributions: the distribution describing the growth of the GDP (Nielsen, 2016a) and the distribution
describing the growth of population (Nielsen, 2016d). Parameters describing the displayed curve are:
=a1 9.859 102 and = k1 5.122 105 for the GDP and= k2 3.749 102 for the
a2 7.542 101 and=
growth of the population.
Results presented in Figure 3 are particularly important because they show that contrary to the
generally accepted interpretations, Industrial Revolution had absolutely no impact on changing the
growth trajectory of income per capita in the region where its impact should have been most
pronounced. Galors claim that the Industrial Revolution was the prime engine of economic growth
(Galor, 2005a, p. 212) is remarkably contradicted by the same data, which he used during the
formulation of his theory. This and other examples show how important Maddisons data are in
correcting the outdated interpretations of the historical economic growth.
160
__________________________________________________________________________________
Figure 3. Maddisons data (Maddison, 2010) describing the growth of income per capita (GDP/cap) in Western
Europe are compared with the linearly-modulated hyperbolic distribution (Nielsen, 2015a) obtained by fitting
the GDP and population data (Nielsen, 2016a, 2016d). The alleged takeoff from stagnation to growth around
AD 1750 (Galor, 2008a, 2012a) did not happen. Industrial Revolution had no impact on the trajectory
describing the growth of income per capita (GDP/cap). The analysis of data used by Galor shows that the
prime engine of economic growth (Galor, 2005a, p. 212) had absolutely no impact on changing the economic
growth trajectory in the region where this engine should have been most effective and where its impacts
should have been most pronounced. From around AD 1900, the growth of the GDP/cap started to be diverted to
a slower trajectory.
It has been shown earlier (Nielsen, 2016a) that economic growth was hyperbolic not only for the
total of 30 countries of Western Europe but also for the four countries, Denmark, France, the
Netherlands and Sweden, described by the most complete sets of data and representing the most
advanced economies. For these countries, hyperbolic growth was between AD 1 and 1875 when it
started to be diverted to a slower trajectory. There was no Malthusian stagnation, no takeoff and no
escape from the Malthusian trap, because there was no trap. Industrial Revolution had absolutely no
impact on changing economic growth trajectories in these four progressive countries where the impact
of this revolution should be clearly demonstrated in the economic growth data.
Analysis of Maddisons data (Maddison, 2010) demonstrates that the remarkable or stunning
escape from the Malthusian trap (Galor, 2005a, pp. 177, 220) never happened because there was no trap
in the growth of income per capita. Whether expressed in terms of the GDP or GDP/cap, economic
growth was remarkably undisturbed during the time of the Industrial Revolution and continued
undisturbed until around 1900, when it started to be diverted to a slower trajectory.
6. Eastern Europe
Results of analysis of the growth of income per capita in Eastern Europe are summarized in Figure
4. Maddisons data (Maddison, 2010) are compared with the linearly-modulated hyperbolic distribution
obtained by dividing two hyperbolic distributions: the distribution describing the growth of the GDP
(Nielsen, 2016a) and the distribution describing the growth of population (Nielsen, 2016d). Parameters
describing the fitted GDP/cap distribution are:= a1 7.749 101 and= k1 4.048 10 4 for the GDP
and= a2 3.055 102 and= k2 1.525 101 for the growth of the population.
Growth of income per capita was slow but it was not stagnant. It was following the linearly-
modulated hyperbolic distribution. Industrial Revolution had absolutely no impact on shaping the
growth trajectory. The stunning takeoff postulated by Galor did not happen. His theory is repeatedly
and consistently contradicted by the data he used during the formulation of his theory. Rather than being
boosted by the Industrial Revolution, the growth of the GDP/cap started to be diverted to a slower
trajectory from as early as around AD 1850.
161
__________________________________________________________________________________
Figure 4. Maddisons data (Maddison, 2010) describing the growth of income per capita (GDP/cap) in Eastern
Europe are compared with the linearly-modulated hyperbolic distribution (Nielsen, 2015a) obtained by fitting
the GDP and population data (Nielsen, 2016a, 2016d). The alleged takeoff from stagnation to growth around
AD 1750 (Galor, 2008a, 2012a) did not happen. Industrial Revolution had no impact on the trajectory
describing the growth of income per capita (GDP/cap). From around AD 1850, rather than being boosted by
the Industrial Revolution, the growth of the GDP/cap started to be diverted to a slower trajectory.
7. Former USSR
Results of analysis of the growth of income per capita in the former USSR are summarized in Figure
5. Maddisons data (Maddison, 2010) are compared with the linearly-modulated hyperbolic distribution
obtained by dividing two hyperbolic distributions: the distribution describing the growth of the GDP
(Nielsen, 2016a) and the distribution describing the growth of population (Nielsen, 2016d). Parameters
describing the fitted GDP/cap distribution are:= a1 6.547 10 1 and= k1 3.452 104 for the GDP
and= a2 2.618 102 and = k2 1.333 101 for the growth of the population.
Figure 5. Maddisons data (Maddison, 2010) describing the growth of income per capita (GDP/cap) in the
former USSR are compared with the linearly-modulated hyperbolic distribution (Nielsen, 2015a) obtained by
fitting the GDP and population data (Nielsen, 2016a, 2016d). The alleged takeoff from stagnation to growth
around AD 1750 (Galor, 2008a, 2012a) did not happen. Industrial Revolution had no impact on the trajectory
describing the growth of income per capita (GDP/cap). From around AD 1870, rather than being boosted by
the Industrial Revolution, the growth of the GDP/cap started to be diverted to a slower trajectory.
Growth of income per capita in the countries of the former USSR was following closely the linearly-
modulated hyperbolic distribution from AD 1. The growth was slow but not stagnant. Growth of the
GDP and population were monotonic (Nielsen, 2016a, 2016d) and consequently the growth of income
per capita (GDP/cap) was also monotonic. The remarkable or stunning takeoff (Galor, 2005a, pp.
177, 220) claimed by Galor never happened. This wished-for feature is repeatedly contradicted by the
162
__________________________________________________________________________________
analysis of economic and population data (Nielsen, 2014, 2016a, 2016b, 2016c, 2016d, 2016e, 2016f)
and by the analysis of the GDP/cap distributions. Soon after the alleged, but non-existent sudden takeoff
from the non-existent stagnation to growth, the growth of income per capita in countries of the former
USSR started to be diverted to a new and slower trajectory.
8. Asia
Analysis of the growth of income per capita (GDP/cap) in Asia (including Japan) is summarised in
Figure 6. Maddisons data (Maddison, 2010) are compared with the linearly-modulated hyperbolic
distribution obtained by dividing two hyperbolic distributions: the distribution describing the growth of
the GDP (Nielsen, 2016a) and the distribution describing the growth of the population (Nielsen, 2016d).
Parameters describing the fitted GDP/cap distribution are:= a1 2.303 10 2 and= k1 1.129 105 for
the GDP and= a2 1.068 101 and = k2 4.999 103 for the growth of population.
Asia is made primarily of less-developed countries (BBC, 2014; Pereira, 2011) so the alleged
stunning takeoff from the alleged stagnation to growth should have occurred around AD 1900 (Galor,
2008a, 2012a). The data show a certain degree of boosting shortly after the time of the claimed
stunning takeoff from stagnation to growth. However, this boosting is not a transition from stagnation
to growth because the preceding trajectory was not stagnant and because the boosted trajectory follows
closely the historical trend. It was obviously only a temporary boosting because the boosted trajectory
is progressively coming closer to the historical trajectory and judging by its decreasing gradient it is
likely to move to the other side.
The growth of income per capita in the past was slow but it is the mathematically-expected
characteristic of dividing two hyperbolic trajectories (Nielsen, 2015a). To claim Malthusian stagnation,
we would have to demonstrate Malthusian oscillation during that time. Obviously, they are missing in
this display and in the display of the GDP and population distributions (Nielsen, 2016a, 2016d).
Figure 6. Maddisons data (Maddison, 2010) describing the growth of income per capita (GDP/cap) in Asia are
compared with the linearly-modulated hyperbolic distribution (Nielsen, 2015a) obtained by fitting the GDP and
population data (Nielsen, 2016a,2016d). The boosting, which commenced around AD 1950 follows closely the
original historical trajectory and is likely to cross it and move to the other side. This temporary disturbance is a
part of the commonly-observed recent transitions to slower trajectories. The only difference in this case is that
the transition to a slower trajectory was preceded by a relatively minor boosting.
The observed boosting could be probably explained by Japans contribution to the total GDP/cap.
Until 1900, Japans contribution was less than 5% but by 1950 it gradually increased to 12% and by
2000 it climbed to 20%. Japan belongs to the more-developed countries so according to Galor (2008a,
2012a) it should have experienced remarkable and stunning takeoff (Galor, 2005a, pp. 177, 220) in
its GDP/cap around 1750 but it did not. On the other hand, Asia should have experienced a sudden
explosion in the GDP/cap growth around 1900 but it did not. It did experience a minor boosting close
163
__________________________________________________________________________________
to that time but as already pointed out this minor boosting is probably associated with the increased
economic activity in Japan. There was no dramatic transition from stagnation to growth as claimed by
Galor but only a transition from the non-stagnant, linearly-modulated hyperbolic trajectory to a
temporarily faster growth, which appears to have been caused primarily, if not entirely, by the
increasing contribution of Japans economy, the contribution, which should have commenced
explosively around 1750 but it did not. Impressions prompted by wished-for features and reinforced by
distorted presentations of data such as shown in Figure 1 could be persuasive but they can be also
strongly misleading. Data have to be rigorously analysed.
9. Africa
Results of analysis of the growth of income per capita in Africa are presented in Figure 7. As
demonstrated earlier (Nielsen, 2016a, 2016d), the GDP and population data for Africa can be fitted
using two hyperbolic distributions, a slow distribution followed by a fast distribution. The transition
from the slow to fast distribution occurred around 1820 for the growth of the GDP and around 1840 for
Parameters describing the fitted GDP/cap distribution between AD 1 and 1820 are:
=a1 1.244 101 and = k1 5.030 105 for the GDP and= a2 5.794 101 and = k2 2.473 102 for the
growth of the population. For the GDP/cap distribution from AD 1840, parameters are:
=a1 4.192 10 1 and = k1 2.126 10 4 for the GDP and= a2 1.571 102 and = k2 7.834 102 . The
fit to the transient region between AD 1820 and 1840 was obtained by polynomial interpolation.
Figure 7. Maddisons data (Maddison, 2010) describing the growth of income per capita (GDP/cap) in Africa
are compared with the linearly-modulated hyperbolic distributions (Nielsen, 2015a) obtained by fitting the GDP
and population data (Nielsen, 2016a, 2016d). The alleged takeoff from stagnation to growth around AD 1900
for less-developed regions (Galor, 2008a, 2012a) did not happen because the GDP/cap trajectory was not
stagnant before that year and because it continued undisturbed after this year until around 1950 when it started
to be diverted to a slower trajectory.
Africa presents an interesting and unique case when the singularity for the hyperbolic trajectory
describing the growth of population between AD 1 and 1840 is earlier than the singularity for the
hyperbolic trajectory describing the growth of the GDP between AD 1 and 1820. For the growth of the
population, the point of singularity is at t = 2343 while for the growth of the GDP it is at t = 2473
(Nielsen, 2016a, 2016d).
For the linearly-modulated hyperbolic distribution the escape to infinity at a fixed time occurs when
the singularity for the growth of the GDP is earlier than the singularity for the growth of population
(Nielsen, 2015a). If the singularity for the growth of population occurs earlier, as in Africa, then the
164
__________________________________________________________________________________
GDP/cap ratio decreases slowly with time and then escapes rapidly to zero at the time of the singularity
for the growth of population. The decreasing GDP/cap distribution between AD 1 and the early 1800s
in Africa does not represent an unusual and distinctly new mechanism of economic growth but simply
the mathematical property of dividing two hyperbolic distributions describing the growth of the GDP
and population. In particular, it does not represent Malthusian stagnation because both the GDP and the
population were increasing hyperbolically (Nielsen, 2016a, 2016d). It simply shows that the growth of
population was just slightly faster than the economic growth. In all other regions, and globally, the past
economic growth was faster than the growth of population, the unwelcome situation because if both
trajectories are hyperbolic, such a growth leads eventually to excessively fast-increasing trajectories of
income per capita, which, if unchecked, increase to infinity at a fixed time.
The best but still not the ideal solution would be to keep the economic growth and the growth of
population increasing at approximately the same rate. In such a case, we could avoid approaching a
critical growth in the income per capita. However, the problem of the growth of population and of the
economic growth would still remain unsolved. Both cannot growth indefinitely and it does not matter
whether they follow hyperbolic trajectories or some other increasing trajectories.
The past economic growth in Africa was close to a perfect balance between the economic growth
and the growth of population. However, between AD 1820 and 1840, this nearly perfect balance
changed dramatically to resemble the generally experienced trend of reckless and irresponsible
economic growth. It does not mean that a secure economic growth can be only achieved by adopting
primitive living conditions. Progress can be made but there is a limit to how much such a progress can
be supported by the availability of natural resources. The current global economic growth is insecure
(Nielsen, 2015b)
Africa is made of less-developed countries (BBC, 2014; Pereira, 2011) so according to Galor (2008a,
2012a) it should have experienced stagnation until around AD 1900 followed by a clear takeoff around
that year. These expectations are contradicted by data.
In contradiction of Galors interpretations of economic growth (Galor, 2005a, 2008a, 2011, 2012a),
the Malthusian regime of stagnation did not exist. The GDP and population were increasing
hyperbolically during the entire time of the alleged but non-existent regime of Malthusian stagnation,
from AD 1 to 1900 and even after that year. Unrecognised by Galor (because he did not analyse data
but preferred to use distorted diagrams) there was a transition between two hyperbolic trajectories
during his assumed but non-existent regime of Malthusian stagnation, transition from a slowly
decreasing income per capita described by the linearly-modulated hyperbolic trajectory to a fast-
increasing trajectory. When studied separately, it was a transition from slow hyperbolic growth of the
GDP or population to a fast-increasing hyperbolic growth. It was not a transition from stagnation to
growth, claimed erroneously by Galor, but transition from growth to growth.
Africa is the only region where the economic growth was boosted at the time of the Industrial
Revolution but it is also the poorest region, where the claimed Malthusian stagnation should have been
most clearly demonstrated. According to Galor, Malthusian stagnation should have prevailed in Africa
until around 1900 (Galor, 2008a, 2012a). This hypothesis, which appears to have been confirmed by
his manipulation of data, is clearly and convincingly contradicted by their mathematical analysis.
Analysis of data describing the GDP and population in Africa shows that there was no stagnation
over the entire range of the AD era (Nielsen, 2016a, 2016d). Economic growth (as described by the
GDP) and the growth of population were following the steadily-increasing and undisturbed hyperbolic
trajectories but around the time of the Industrial Revolution they were diverted to faster hyperbolic
trajectories. There are no signs of Malthusian stagnation before and after the Industrial Revolution and
before AD 1900, which was supposed to mark the end of the epoch of Malthusian stagnation.
Hyperbolic growth, even if slow, does not represent Malthusian stagnation. Convincing signature of
Malthusian stagnation is in the form of random fluctuations often described as Malthusian oscillations.
This signature is missing in the data for Africa and for other regions. Data for Africa and for other
regions, show steadily-increasing hyperbolic distributions describing economic growth and the growth
of population (Nielsen, 2016a, 2016d).
Analysis of the GDP/cap data for Africa shows that after a transition from a slowly-decreasing
trajectory before around 1840 (which as we have pointed out does not represent Malthusian stagnation
but the mathematical property of dividing monotonically-increasing hyperbolic trajectories) the growth
of the GDP/cap in Africa was following a vigorously-increasing trajectory during the alleged
165
__________________________________________________________________________________
Malthusian stagnation. So, while the suitable manipulation of data (Galor, 2005a, 2011) appears to be
confirming preconceived ideas, mathematical analysis of precisely the same data shows that the
preconceived ideas are clearly incorrect.
Here it might be a good place to point out that harsh and primitive living conditions in the past should
not be immediately interpreted as a proof of the existence of Malthusians stagnation. Africa is at present
the poorest region. It is also the only region where the growth of income per capita was negative over
hundreds of years, between AD 1 and 1840 but there was no Malthusian stagnation in the growth of
population and in the economic growth in Africa (Nielsen, 2016, 2016d). There was also no Malthusian
stagnation in other regions. Living conditions in the past in all regions were primitive and often harsh
by modern standards and yet in all of them there was no Malthusian stagnation in the economic growth
and in the growth of population. Worldwide, living conditions in the distant past were harsh and
primitive and yet for as long as we can trace it by using analysable data, for a time extending down to
10,000 BC, there was no Malthusian stagnation in the growth of human population. With the exception
of just two past and relatively brief transitions, the world growth of population was hyperbolic for nearly
12,000 years (Nielsen, 2016b).
Mathematical analysis of data finds no confirmation of the existence of the hypothetical epoch of
Malthusian stagnation. It is a vague concept, which has no application in the explanation of the
dynamics of economic and demographic growth. Its continuing presence in academic discussions as a
tool to explain the dynamics of growth is not only irrelevant but also counterproductive because it
diverts attention form finding correct explanations of the mechanism of the growth process.
It is also useful to compare results of the analysis for Africa with the results for Western Europe.
Industrial Revolution, the prime engine of economic growth (Galor, 2005a, p. 212) should have
worked most efficiently in Western Europe and its effects should have been most convincingly
confirmed by data in this region, but these effects are convincingly contradicted by data: the alleged
engine did not change the economic growth or the growth of population trajectories in Western Europe.
Likewise, Malthusian stagnation should have been most prominently confirmed in Africa but it is not.
There was never any form of stagnation in the economic growth in Africa.
Furthermore, while in Western Europe, Industrial Revolution had absolutely no impact on changing
the economic growth trajectory, in Africa there was a spectacular acceleration of growth during the time
of the Industrial Revolution but it was not the acceleration from stagnation to growth but from growth
to growth. The wished-for features are contradicted by data showing that even plausible stories and
explanations should not be accepted in science unless they can be confirmed by relevant data; otherwise
they are at best just interesting stories with no scientific merit.
The alleged sudden acceleration (takeoff) in income per capita is supposed to have been associated
with the benefits of progress such as better health care, better housing, better education, higher standard
of living and generally better living conditions. However, data show that in Europe there was no takeoff
in the income per capita at the time of the Industrial Revolution, while in Africa there was a dramatic
acceleration without a dramatic improvement in the style of living. On the contrary, this dramatic
boosting in income per capita in Africa at the time of the Industrial Revolution appears to coincide with
the dramatic deterioration of living conditions of native populations. It occurred around the time of the
intensified colonisation of Africa (Duignan & Gunn, 1973; McKay, Hill, Buckler, Ebrey, Beck,
Crowston, & Wiesner-Hanks, 2012; Pakenham, 1992).
If a sudden takeoff is supposed to mark the remarkable or stunning escape from the Malthusian
trap (Galor, 2005a, pp. 177, 220), then the only such takeoff occurred in Africa. However, this dramatic
takeoff did not mark the transition from stagnation to growth, because there was no stagnation. It also
did not mark the dramatic escape from the Malthusian trap because there was no Malthusian trap in the
economic growth or in the growth of population in Africa: both were increasing hyperbolically before
the claimed but non-existent takeoff around AD 1900. The sudden and steep change in the economic
growth and in the growth of population in Africa between AD 1820 and 1840 did not mark the dramatic
escape from the hypothetical Malthusian trap but the transition from freedom and independence to the
trap of misery, deprivation and suffering of the native population. Poetic explanations could be
interesting but they are not necessarily scientifically sound. Creative imagination can play an important
role in science as long as it is tested by the scientific process of investigation.
166
__________________________________________________________________________________
10. Latin America

Results of analysis of economic growth in Latin America are presented in Figure 8. Data for Latin
America contain a discontinuity in the growth of the GDP and population between AD 1500 and 1600
(Nielsen, 2016a, 2016d). This discontinuity is reflected in the discontinuity of the growth of income per
capita (GDP/cap).
Parameters describing the slowly-increasing, linearly-modulated hyperbolic trajectory are
=a1 4.421 101 and= k1 2.093 104 for the GDP and = a2 1.765 102 and = k2 8.242 10 2 for the
population. The fast-increasing trajectory from AD 1600 is described by the following parameters:
=
a1 1.570 100 and
= k1 8.224 104 for the GDP and= k2 3.371101
a2 6.561102 and= for the
population. The discontinuity in the economic growth and in the growth of population coincides with
the onset of Spanish conquest (Bethell, 1984). However, after this relatively brief delay, economic
growth and the growth of human population were following fast-increasing hyperbolic trajectories.
according to Galor (2008a, 2012a), the growth of income per capita (GDP/cap) in this region should
have been stagnant until around AD 1900 and fast from around that year. This pattern of growth is
contradicted by data. Data show a diametrically different pattern: (1) there is no convincing evidence
of the existence of stagnation over the entire range of time between AD 1 and 1900 (convincing
evidence of Malthusian stagnation requires the presence of random fluctuations) but there is a
sufficiently convincing evidence of the linearly-modulated hyperbolic growth particularly between AD
1600 and 1900; (2) there was no takeoff from stagnation to growth at any time; and (3) at the time of
the postulated takeoff in 1900 the growth of income per capita started to be diverted to a slower
trajectory. The wished-for takeoff is replaced by a slower growth. However, even if we had a takeoff
around that time it would have been a takeoff of a different kind, not a takeoff from stagnation to growth
as required by the Unified Growth Theory (Galor, 2005a, 2011) but a takeoff from growth to growth.
Figure 8. Maddisons data (Maddison, 2010) describing growth of income per capita (GDP/cap) in Latin
America are compared with the linearly-modulated hyperbolic distributions (Nielsen, 2015a) obtained by fitting
the GDP and population data (Nielsen, 2016a, 2016d). There was a discontinuity in the economic growth and in
the growth of population between AD 1500 and 1600 reflected in the discontinuity in the GDP/cap distribution.
The alleged takeoff from stagnation to growth around AD 1900 (Galor, 2008a, 2012a) did not happen because
the GDP/cap trajectory was not stagnant before that year and because there was no sudden acceleration in
growth around that year. On the contrary, around the alleged takeoff there was a transition to a slower
trajectory.

Results of mathematical analysis of the historical income per capita (GDP/cap) distributions are
summarised in Table 1. Listed parameters ( a1 , k1 , a2 , k2 ) describe the fitted, linearly-modulated
167
__________________________________________________________________________________
hyperbolic trajectories (Nielsen, 2015a) represented by the ratios of hyperbolic distributions describing
the growth of the GDP and population. Parameters a1 and k1 describe hyperbolic distributions fitting the
GDP data (Nielsen, 2016a), while parameters a2 and k2 describe hyperbolic distributions fitting
population data (Nielsen, 2016d).
Results of this analysis demonstrate explicitly that the postulated by Galor (2005a, 2008a, 2011,
2012a) takeoffs in the income per capita (GDP/cap) did not happen. There were no transitions from
stagnation to growth because within the mathematically-analysable data Galors regime of Malthusian
stagnation did not exist in the growth of income per capita. Growth of income per capita was following
the linearly-modulated hyperbolic distributions until recently when it started to be diverted to slower
trajectories.
Galors Unified Growth Theory (Galor, 2005a, 2011) is contradicted yet again by data. The
remarkable and stunning escape from the Malthusian trap (Galor, 2005a, pp. 177, 220) never
happened because there was no trap in the economic growth and in the growth of population. His claim
of the existence of the differential timing of takeoffs is also contradicted by data because we cannot
have differential timing of takeoffs without takeoffs. Galor describes phenomena that did not exist. His
explanations of economic growth are based on phantom features created by hyperbolic illusions and
magnified by his habitually distorted presentations of data such as illustrated in Figure 1. His theory is
irrelevant and misleading.
Galor had access to the excellent data of Maddison (2001). He even used them during the
formulation of his theory but he did not attempt to analyse them, which is surprising because their
analysis is trivially simple (Nielsen, 2014). Now, precisely the same data can be used to demonstrate
that his Unified Growth Theory (Galor, 2005a, 2011) is repeatedly contradicted by data (Nielsen, 2014,
2015a, 2016a, 2016c, 2016d, 2016e, 2016f).
Table 1. Summary of the mathematical analysis of the historical income per capita (GDP/cap)
distributions
Region a1 k1 a2 k2 Stagnation Takeoff
1.684 102 8.539 106 3.765 103

World X X
7.739 100
9.859 102 5.112 105 3.749 102
Western Europe X X
7.542 101
7.749 101 4.048 104 1.525 101
Eastern Europe X X
3.055 102
6.547 101 3.452 104 1.333 101
Former USSR X X
2.618 102
2.303 102 1.129 105 4.999 103
Asia X X
1.068 101
1.244 101 5.030 105 2.473 102
Africa X X
5.794 101
4.192 101 2.126 104 1.571 102 7.834 102
4.421 101 2.093 104 8.242 102
Latin America X X
1.765 102
1.570 100 8.224 104 6.561 102 3.371 101
Notes: a1 , k1 , a2 , k2 Parameters describing linearly-modulated hyperbolic distributions (ratios of
hyperbolic distributions). Parameters a1 , k1 describe hyperbolic growth of the GDP, while a2 , k2
describe hyperbolic growth of population [see eqn (1)]. X No stagnation/takeoff. Within the range of
the mathematically-analysable data the claimed by Galor (2005a, 2008a, 2011, 2012a) Malthusian
regime of stagnation did not exist. The claimed takeoffs from stagnation to growth never happened.
Unified Growth Theory is fundamentally incorrect and scientifically unacceptable. It is a theory

based on scientifically unsupported concepts created by impressions and reinforced by the manipulation
of data. Excellent data of Maddison (2001) were not analysed but presented repeatedly using distorted
and misleading diagrams such as shown in Figure 1. Such distorted presentation of data appears not
only ubiquitously in the Unified Growth Theory but also in other related publications (Ashraf, 2009;
Galor, 2005a, 2005b, 2007, 2008a, 2008b, 2008c, 2010, 2011, 2012a, 2012b, 2012c; Galor & Moav,
2002; Snowdon & Galor, 2008). Selected but meaningless and misleading values of data were also
repeatedly quoted to support the concept of stagnation followed by takeoffs from stagnation to growth.
This is an unscientific approach to research.
168
__________________________________________________________________________________
Galors handling of data is puzzling. Maybe he does not know how to analyse data, but this
conclusion is hard to accept because he appears to be familiar with mathematics and the analysis of
hyperbolic distributions is trivially simple (Nielsen, 2014). It looks as if, for whatever reason, he
purposefully manipulated data to support his preconceived ideas.
Assisted by the excellent data of Maddison (2001) available to him at the time of the formulation of
his theory, Galor was on the verge of making an important contribution to science by showing that
economic growth was hyperbolic, in agreement with what has been already known about the growth of
population (Kapitza, 2006; Kremer, 1993; Podlazov, 2002; Shklovskii, 1962, 2002; von Foerster, Mora
and Amiot, 1960; von Hoerner, 1975). He would have discovered that there was no Malthusian
stagnation and no takeoffs from stagnation to growth the economic growth, the features which were not
confirmed in the published results describing the growth of population. However, he missed this
excellent opportunity because he failed to follow the fundamental principles of scientific investigation,
which require that theories should be tested by data and that research should be guided and moderated
by data.
The most plausible explanation why Galor appears to have been reluctant to be guided by data and
why he apparently manipulated data to support his preconceived ideas is that he was blinded by
prejudice. It is a kind of fear or reluctance to accept evidence contradicting the established knowledge.
Psychologists describe it as cascade behaviour, information cascade, informational avalanche, illusion
of truth, illusory truth, illusion of familiarity, running with the pack, following the crowd, herding
behaviour, bandwagons and path depending choice (Anderson & Holt, 1997; Begg, Anas & Farinacci,
1992; Bikhchandani, Hirshleifer & Welch, 1992, 1998; De Vany & Lee, 1999; De Vany & Walls, 1999;
Easley & Kleinberg, 2010; Grebe, Schmid & Stiehler, 2008; Ondrias, 1999; Parks & Tooth, 2006;
Ramsey, Raafat, Chater & Frith, 2009; Walden & Browne, 2003).
In the demographic and economic research, this phenomenon is demonstrated by the reluctance to
accept the compelling contradicting evidence in data simply because many demographers or economists
would not agree with the contradicting evidence. It is safer to follow the crowd and run with the pack.
Tradition is stronger than science and only an outsider who has not been blinded by prejudice and who
is not afraid of being rejected by the crowd might dare to show that the accepted doctrines are incorrect.
He or she is then likely to be ridiculed and rejected but science is a self-correcting discipline so sooner
or later such resistance to accept the overwhelming empirical evidence will have to be broken, but it
would be better for science and scientists if the required change in the paradigm is accepted sooner
rather than later.
The evidence contradicting currently-accepted interpretations is overwhelming: historical economic
growth and historical growth of population were hyperbolic (Kapitza, 2006; Kremer, 1993; Nielsen,
2014, 2015a, 2016a, 2016b, 2016c, 2016d, 2016e, 2016f; Podlazov, 2002; Shklovskii, 1962, 2002; von
Foerster, Mora and Amiot, 1960; von Hoerner, 1975). Hyperbolic growth should be the basis for
explaining the mechanism of the historical growth of population and of the historical economic growth.
Interpretations revolving around the concept of Malthusian stagnation and around assumed
transitions from stagnation to growth are repeatedly and consistently contradicted by data and by their
mathematical analyses. Historical economic growth and historical growth of population cannot be
divided into distinctly different regimes governed by distinctly different mechanisms of growth.
Hyperbolic growth has to be explained as a whole. The same mechanism has to be applied to the slow
and fast hyperbolic growth because it is mathematically impossible to divide hyperbolic distributions
into distinctly different sections (Nielsen, 2014). Once we can explain properly the mechanism of the
past growth we might be able to understand better the current growth and how it should be controlled.
References
Anderson, L. R., & Holt, C. A. (1997). Information cascades in the laboratory. American Economic Review, 87(5), 847-862.
Begg, I. M., Anas, A., & Farinacci, S. (1992). Dissociation of process in belief: Source recollection, statement familiarity
and the illusion of truth. Journal of Experimental Psychology. 121(4), 446 458.
169
__________________________________________________________________________________
Bethell, L. (Ed.). (1984). The Cambridge History of Latin America: Vol. I and II, Colonial Latin America. Cambridge, UK:
Bikhchandani, S., Hirshleifer, D., & Welch, I., A. (1992). Theory of fads, fashion, custom, and cultural change as
informational cascades. Journal of Political Economy, 100(5), 992-1026.
De Vany, A., & Lee, C. (1999). Information cascades in multi-agent models.
https://www.researchgate.net/publication/5061066_Information_Cascades_in_Multi-Agent_Models
De Vany, A., & Walls, R. (1999). Uncertainty in the movie industry: Does star power reduce the terror of the box office?
Journal of Cultural Economics, 23, 285318.
Social Science: World Population, 369, 1-8.
Durand, J. (1977). Historical Estimates of World Population: An Evaluation. Population and Development Review, 3(3),
256-293.
Press.
Easley, D., & Kleinberg, J. (2010). Networks, crowds, and markets. Cambridge, UK: Cambridge University Press.
Publishing Company.
475.
Dictionary of Economics. Palgrave Macmillan, New York. doi. 10.1057/9780230226203.0434
0062-7
117, 1133-1191. doi. 10.1162/003355302320935007
Economics, 108(3), 681716.
McEvedy, C., & Jones, R. (1978). Atlas of World Population History. Middlesex, England: Penguin.
242-255.
170
__________________________________________________________________________________
Nielsen, R. W. (2016d). Unified Growth Theory contradicted by the mathematical analysis of the historical growth of human
Nielsen, R. W. (2016f). Puzzling properties of the historical growth rate of income per capita explained. Journal of
Books.
Avon Books.
Ondrias, K. (1999). The brain, consciousness & illusion of truth (E. Nazinska, Trans. & Ed.). Boca Raton, FL: Universal
Publishers.
Parks, C. M., & Tooth, J. P. (2006). Fluency, familiarity, aging, and the illusion of truth. Aging, neuropsychology, and
cognition, 13, 225253
bank/
Ramsey M., Raafat, R. M., Chater, N., & Frith, C. (2009). Herding in humans. Trends in Cognitive Sciences, 13(10), 420-
428.
Pub.
Von Hoerner, S. J. (1975). Population explosion and interstellar expansion. Journal of the British Interplanetary Society, 28,
691-712.
Walden, E. A., & Browne, G. J. (2003). Running with the pack: An empirical examination of information cascades in the
adoption of novel technologies. Lubbock: Texas Tech. University.
171
__________________________________________________________________________________
11
The unresolved mystery of the great divergence is solved
By Ron W. NIELSENa
Abstract. The so-called great divergence in the income per capita is described in the Unified Growth Theory as
the mind-boggling and unresolved mystery about the growth process. This mystery has now been solved: the great
divergence never happened. It was created by the manipulation of data. Economic growth in various regions is at
different levels of development but it follows similar, non-divergent trajectories. Unified Growth Theory is shown
yet again to be incorrect and scientifically unacceptable. It promotes incorrect and even potentially dangerous
concepts. The distorted presentation of data supporting the concept of the great divergence shows that economic
growth is now developing along moderately-increasing trajectories but mathematical analysis of the same data
and even their undistorted presentation shows that these trajectories are now increasing approximately vertically
with time. So, while the distorted presentation of data used in the Unified Growth Theory and the spuriously-
created great divergence suggest the generally sustainable and secure economic growth, the undistorted
presentation of data demonstrates that the growth is unsustainable and insecure. Similar dangerously incorrect
concept promoted by the Unified Growth Theory is the repeated doctrine of takeoffs from the hypothetical but
non-existent stagnation to growth. They also suggest prosperous and secure future. Such takeoffs never happened
but even without them the current economic growth is insecure.
Keywords. Economic growth, Unified Growth Theory, Regional economic growth, Great Divergence, Income
per capita, Hyperbolic growth
JEL. A10, A12, C12, C20, C50, F00, N00, Y80.
1. Introduction
T hose who are less familiar with the scientific process of investigation might not be aware that
there is also unscientific approach, which unfortunately appears to be used sometimes even in
academic circles. It is important to have a clear understanding of these two different ways of
investigation in order to be able to distinguish between acceptable and unacceptable claims and
conclusions.
In science, theories are tested by data. In unscientific discussions, data are tested by theories. In
unscientific presentations, selective use of data is common. Data are manipulated, distorted or rejected
if they do not agree with preconceived ideas.
In the scientific research, contradicting evidence is not only accepted but looked for because it
usually leads to new discoveries. In unscientific discussions, contradicting evidence is studiously
rejected because it threatens the established knowledge.
In science, data are rigorously analysed. In non-scientific discussions, rigorous analysis is avoided
and interpretations of data are based on impressions, but impressions can be misleading and even great
thinkers can make a mistake. It is clear that the earth does not move, and that it does not lie elsewhere
than at the centre (Aristotle). Appearances and logical explanations are not necessarily reliable. Plato
is my friend, Aristotle is my friend, but my greatest friend is truth (Sir Isaak Newton).
It is also important to understand the limitations of mathematics. Elaborate stories and explanations
can be translated into mathematical language but such translations are meaningless unless they can be
tested by data.
We should never be mesmerised by complicated mathematical formulae and presentations. The
essential question is whether the presented mathematics can be tested by relevant data. If stories
a
Australia.
. +61407201175
Published as: Nielsen, R. W. (2016). The unresolved mystery of the great divergence is solved, Journal of Economic and
Social Thought, 3(2), 196-219.
172
__________________________________________________________________________________
translated into mathematics cannot be tested by data, if they have to be accepted by faith, then obviously,
they have no scientific value and they can be ignored or even rejected. Mathematical formulations
should be making testable predictions. A story dressed up in a mathematical gown will be just a story
unless it makes a testable prediction.
A good example of the unscientific approach to research is the Unified Growth Theory (Galor,
2005a, 2011). Data are manipulated and distorted. Selected data, which appear to support preconceived
concepts, are repeatedly quoted. Excellent data of Maddison (2001) are used during the formulation of
this theory but they are never analysed. They are presented in distorted ways to support preconceived
ideas. Galor translates his assumed and scientifically-unsupported interpretations of economic growth
into many complicated but rather primitive mathematical formulae. However, he does not make even a
single mathematical prediction, which can be tested directly by data. His mathematical expressions do
not describe growth trajectories that could be compared with data, even with data he uses during the
formulation of his theory. Ironically, precisely the same data, when analysed, are in direct contradiction
of his theory.
His concepts can be only tested indirectly by showing that within the range of the mathematically-
analysable data there was no stagnation, no sudden takeoffs, no remarkable or stunning escapes
from the Malthusian trap (Galor, 2005a, pp. 177, 220) and no transition from stagnation to the so called
sustained growth regime (Nielsen, 2014, 2015a, 2016a, 2016b, 2016c, 2016d, 2016e, 2016f, 2016g,
2015h). Economic growth in the past was sustainable and secure, as indicated by the steadily-increasing
hyperbolic trajectories, but now it is unsustainable and insecure (Nielsen, 2015b). The numerous
mathematical formulae used in the Unified Growth Theory do not describe or explain the historical
economic growth because they incorporate concepts, which are either contradicted repeatedly by data
or have to be accepted by faith.
We have already demonstrated (Nielsen, 2014, 2015a, 2016a, 2016d, 2016e, 2016f, 2016g, 2016h)
that Galors Unified Growth Theory is fundamentally incorrect because it is based on fundamentally-
incorrect ideas. We have shown that within the range of the mathematically-analysable data, historical
economic growth and the growth of population were hyperbolic. For the economic growth, the range of
evidence is limited but for the growth of human population it can be extended to 10,000 BC (Nielsen,
2016b). We have demonstrated that within the range of analysable data, there was no Malthusian
stagnation and no Malthusian trap in economic growth and in the growth of population. The growth
was slow over a long time but it was steadily increasing and there was no transition at any time in the
past that could be described as a sudden takeoff, spurt, sprint or explosion. We have demonstrated that
Galors claim of sudden takeoffs is repeatedly contradicted by data. There were no takeoffs and
consequently there was also no differential timing of takeoffs. During the time of the claimed takeoffs,
economic growth and the growth of population were either continuing to increase along undisturbed
and remarkably stable hyperbolic trajectories or they were diverted to slower trajectories. This
conclusion applies not only to the growth of the Gross Domestic Product (GDP) and population but also
to the growth of income per capita (GDP/cap). We do not have to try to explain the mechanism of the
epoch of Malthusian stagnation and of the escape from the Malthusian trap because there was no
stagnation and no trap in the economic growth and in the growth of population. What we have to explain
is why the growth in the past was hyperbolic, why it was so remarkably stable and why it started to be
diverted recently to new, non-hyperbolic trajectories.
2. The concept of the great divergence

The concept of the great divergence belongs to a set of other phantom mysteries about the growth
process (Galor, 2005a, p. 220) invented by Galor and reinforced by the habitually distorted
presentations of data (Ashraf, 2009; Galor, 2005a, 2005b, 2007, 2008a, 2008b, 2008c, 2010, 2011,
2012a, 2012b, 2012c; Galor & Moav, 2002; Snowdon & Galor, 2008). One example of such distorted
presentation of data used routinely by Galor is shown in Figure 1. In contrast, the accurate presentation
of precisely the same data, together with their mathematical analysis, is shown in Figure 2.
In the distorted and appropriately manipulated presentation of data shown in Figure 1 we can see
clearly the non-monotonic growth of population and of the GDP/cap. After the apparent long stagnation,
we see a sudden takeoff to a new regime of growth. Galor made no attempt to analyse data, which is
173
__________________________________________________________________________________
surprising because their analysis is trivially simple (Nielsen, 2014). The manipulated data appear to
support the concept of stagnation and takeoffs described usually as the escape from the Malthusian trap.
Figure 1. Example of the ubiquitous, grossly-distorted and self-misleading diagrams used to create the Unified
Growth Theory (Galor, 2005a, 2011). Maddisons data (Maddison, 2001) were used during the formulation of
this theory but they were never analysed. Such state-of-the-art was used to construct a system of scientifically-
unsupported interpretations, explanations and mysteries of the growth process (Galor, 2005a, p. 220).
In contrast, the accurate display of precisely the same data suggests entirely different interpretation.
General features presented in Figure 1 are still maintained but now mathematical analysis of these data
shows that the GDP/cap and the size of the population were increasing monotonically (Nielsen, 2014,
2015a, 2016a, 2016d, 2016e, 2016f, 2016g, 2016h). There were no sudden takeoffs from stagnation to
growth because there was no stagnation and because the acceleration was gradual along the entire range
of these distributions. The gradient and the growth rate of the GDP/cap distribution were changing
monotonically without any discontinuity, which could be claimed as a takeoff (Nielsen, 2015a, 2016g,
2016h).
Figure 2. Precisely the same data (Maddison, 2001) as used in Figure 1but now displayed accurately and
analysed. They follow monotonically-increasing distributions, which cannot be divided into distinctively-
different components (Nielsen, 2014, 2015a, 2016a, 2016d, 2016e, 2016f, 2016g, 2016h).
Even though the GDP/cap distribution seems to suggest a sudden increase, this feature is just an
illusion, which is dispelled by the mathematical analysis of data or even by using semilogarithmic scales
of reference (Nielsen, 2015a, 2016g). The GDP/cap is the ratio of two distributions, the distribution
describing the growth of the GDP and the distribution describing the growth of population. Both were
increasing hyperbolically and monotonically (Nielsen, 2015a, 2016a, 2016d). The displayed features
174
__________________________________________________________________________________
(slow growth over a long time and fast growth over a short time) represent nothing more than
mathematical properties of monotonically-increasing hyperbolic distributions. They are not the unique
properties of economic growth but economic growth happens to be hyperbolic.
It is impossible to locate a transition from the slow to fast growth for hyperbolic distributions
(Nielsen, 2014) because such a transition does not exist. The GDP/cap distributions are simply the
linearly-modulated and monotonically-increasing hyperbolic distributions (Nielsen, 2015a).
The distorted diagram used by Galor to support his erroneous concept of the great divergence is
presented in Figure 3. This distorted presentation of Maddisons data was reproduced from Galors
publication (Galor, 2005a, p. 175). It shows that over a long time there was hardly any difference in the
economic growth for various regions. However, from around the time of the Industrial Revolution,
1760-1840 (Floud & McCloskey, 1994), there was a sudden takeoff and the economic growth diverged
into distinctly different trajectories.
We have already demonstrated that there were no takeoffs in the growth of the GDP and GDP/cap
(Nielsen, 2015a, 2016a, 2016e, 2016g) and consequently there was also no differential timing of
takeoffs claimed by Galor in his Unified Growth Theory (Galor, 2005a, 2011). We have also
demonstrated that there were no takeoffs in the growth of the world population in the past 12,000 years
(Nielsen, 2016b) and in the growth of regional populations (Nielsen, 2016d). The incorrectly-claimed
takeoffs by Galor represent just the natural continuations of hyperbolic growth. Analysis of data shows
that at the time of the alleged takeoffs, and in clear contradiction of the Unified Growth Theory,
economic growth in various regions was either continuing to increase along undisturbed hyperbolic
trajectories or started to be diverted to slower trajectories.
Figure 3. A typical distorted presentation of Maddisons data (Maddison, 2001) used by Galor to support his
concepts of takeoffs and of the great divergence (Galor, 2005a, p. 175).
Now we shall show that there was no great divergence in the income per capita. We shall show again
that the Unified Growth Theory is scientifically unacceptable. It does not describe the mechanism of
economic growth. It describes phantom features constructed by the manipulation of data.
We shall show that the great divergence never happened. However, we shall also explain how Galor
constructed his great divergence. We shall show how the great divergence can be constructed by a
distorted presentation of any distributions, which increase slowly over a long time and fast over a short
time. They do not have to be distributions describing economic growth.
3. Analysis of the early data of Maddison

We shall first investigate precisely the same data (Maddison, 2001) as used by Galor (2005a, 2011)
during the formulation of his Unified Growth Theory and we shall show that they do not support the
concept of the great divergence. Results of mathematical analysis of these data are shown in Figures 4-
9. Parameters of the fitted distributions have been listed earlier (Nielsen, 2016g). The fitted curves are
the linearly-modulated hyperbolic distributions (Nielsen, 2015a) obtained by dividing hyperbolic
distributions fitting the corresponding GDP and population data (Nielsen, 2016a, 2016d). All GDP/cap
values are in 1990 International Geary-Khamis dollars.
175
__________________________________________________________________________________
Figure 4. Growth of income per capita, i.e. Gross Domestic Product per capita (GDP/cap), in Western Europe
(Maddison, 2001; Nielsen, 2016g). From around 1913, economic growth in Western Europe started to depart
from the historical linearly-modulated hyperbolic distribution. However, it continued to increase close to the
historically-predicted trajectory.
Figure 5. Growth of income per capita in Eastern Europe (Maddison, 2001; Nielsen, 2016g). From around
1870, economic growth in Eastern Europe started to depart from the historical linearly-modulated hyperbolic
distribution. However, it continued to increase close to the historically-predicted trajectory. The growth was not
diverted to a distinctly different and gently-increasing trajectory as claimed by Galor (2005a, 2011; cf Figure
3).
176
__________________________________________________________________________________
Figure 6. Growth of income per capita in countries of the former USSR (Maddison, 2001; Nielsen, 2016g).
From around 1870, economic growth in the former USSR started to depart from the historical linearly-
modulated hyperbolic distribution. However, it continued to increase close to the historically-predicted
trajectory.
Figure 7. Growth of income per capita in Asia (Maddison, 2001; Nielsen, 2016g). The data follow closely the
linearly-modulated hyperbolic distribution. There was no divergence to a distinctly different and gently-
increasing trajectory as claimed by Galor (2005a, 2011; cf Figure 3).
177
__________________________________________________________________________________
Figure 8. Growth of income per capita in Africa (Maddison, 2001; Nielsen, 2016g). The data follow closely the
linearly-modulated hyperbolic distributions. There was no divergence to a distinctly different and gently-
increasing trajectory as claimed by Galor (2005a, 2011; cf Figure 3).
Figure 9. Growth of income per capita in Latin America (Maddison, 2001; Nielsen, 2016g). From around 1913,
economic growth in Latin America started to depart from the historical linearly-modulated hyperbolic
distribution. However, it continued to increase close to the historically-predicted trajectory. The growth was not
diverted to a distinctly different and gently-increasing trajectory as claimed by Galor (2005a, 2011, cf Figure 3).
178
__________________________________________________________________________________
Figure 10. Growth of income per capita in all regions, including Western offshoots (Maddison, 2001). They are
all increasing in approximately the same direction. There is no divergence to distinctly different trajectories.
The data for Western Offshoots were not analysed because of their poor quality, but they are
displayed in Figure 10. Their economic growth is similar to the growth in Western Europe in the sense
that they are clearly ahead of other regions. However, distributions presented in Figures 4-9 show that
economic growth in all regions follows similar trajectories. The difference between regions is not in
their divergence to distinctly different trajectories as claimed incorrectly by Galor but in their levels of
economic development.
Distributions presented in Figures 4-9 are clearly different than the distorted distributions
constructed by Galor and presented in Figure 3. In Galors distorted presentation of data there is a
cluster of regions (Eastern Europe, Asia, Africa and Latin America) whose economic growth follows
distinctly different trajectories than the growth in Western Europe. This information is incorrect because
the analysis of precisely the same data shows clearly that all distributions are similar, including the
distribution representing the economic growth in Africa. They are all following similar trajectories with
a common tendency to increase nearly vertically and close to the historical linearly-modulated
hyperbolic trajectories.
The common characteristic feature of all these empirical distributions shown in Figures 4-9 is that
they have changed gradually from being nearly horizontal to nearly vertical. We shall show later that
when such distributions become nearly vertical it is easy to distort them and construct the great
divergence, and it does not matter whether they follow the fitted linearly-modulated hyperbolic
distributions or not.
The contrast between Maddisons data and their distorted image constructed by Galor is particularly
clear if we compare Figure 3 with Figure 8. In Figure 3, the data for Africa follow a gently-increasing
trajectory after around 1800, i.e. a trajectory characterised by a small gradient. The correct display of
the same data presented in Figure 8 shows diametrically opposite features: the data for Africa follow a
steep trajectory, i.e. the trajectory characterised by a large gradient. This trajectory is approximately
vertical.
In Galors distorted presentation of data the trajectory for Africa after around 1800 is distinctly
different than the trajectory for Western Europe. However, precisely the same data displayed in Figures
4 and 8 demonstrate that the trajectories for Africa and Western Europe are similar. The only difference
is that Africa is further behind in its level of development.
In Galors distorted presentation of data, economic growth in Eastern Europe, Asia and Latin
America follows also gently increasing trajectories after around 1800, similar to the trajectory for
179
__________________________________________________________________________________
Africa. However, precisely the same data displayed properly in Figures 5, 7, 8 and 9 show that they all
follow approximately vertical trajectories in much the same way as the data for Western Europe. The
only difference is again that Western Europe is further ahead but it is further ahead on the virtually the
same trajectory.
With such distorted presentation of data, it is not surprising that Galor discovered so many mind-
boggling and perplexing mysteries of the growth process (Galor, 2005a, pp. 177, 220), mysteries
representing phantom features created by the manipulation of data.
In contrast with his distorted presentation of data, the gradient of all empirical trajectories in this
section of time is large. They all increase approximately vertically. Such a growth cannot be explained
by claiming that larger size of population demands larger GDP. What we have here is the increasing
GDP per person. It is a growth that reflects our surprisingly fast-increasing demands.
Galors theory conveys dangerously incorrect information. According to his distorted presentation
of data shown in Figure 3, income per capita in certain regions (Eastern Europe, Asia, Africa and Latin
America) is following gently increasing trajectories after around 1800. Such trajectories are relatively
safe. However, the correct presentation of precisely the same data shows that in all regions income per
capita is following the dangerously fast-increasing trajectories. The data show that there is now a critical
urgency to regulate economic growth but Galors theory suggests that there is no danger.
According to his erroneous theory, after a long epoch of stagnation we have escaped the tyranny of
the Malthusian trap and now we can enjoy the sustained growth regime. Furthermore, according to his
erroneous concept of the great divergence, economic growth in most regions diverged to the generally
safe trajectories. However, according to the precisely the same data, all regions are now following
dangerously fast-increasing trajectories and for all of them, without exception, economic crisis seems
to be strongly probable.
4. Analysis of the latest data of Maddison

Data published by Maddison in 2010 show even more clearly that there was no divergence in the
economic growth. These data were available to Galor before the publications of his book (Galor, 2011)
but unfortunately, they were not analysed. Had Galor analysed these data he would have soon
discovered many interesting features characterising economic growth, features, which are repeatedly in
contradiction with his Unified Growth Theory (Galor, 2005a, 2011).
Results of analysis of these new data (Maddison, 2010) are shown in Figures 11-16. Their combined
display is presented in Figure 17.
The mystery of Galors mind-boggling and perplexing phenomenon of the Great Divergence
(Galor, 2005a, pp. 177, 220) has now been solved there is no mystery. This mystery and all other of
his mysteries were created by the manipulation of data. In Galors publications (Galor, 2005b, 2007,
2008a, 2008b, 2008c, 2010, 2012a, 2012b, 2012c; Galor & Moav, 2002; Snowdon & Galor, 2008) data
are repeatedly manipulated and presented using distorted diagrams.
The common characteristic feature of Maddisons data describing the growth of income per capita
(Maddison, 2001, 2010) in various regions is again that their nearly horizontal trajectories changed
gradually into nearly vertical trajectories. They have never diverged into distinctly different trajectories
as claimed by Galor (see Figure 3).
Economic growth in all regions is now following new trajectories but all of them continue to increase
close to the historical, linearly-modulated hyperbolic trajectories, which escape to infinity at a fixed
time. In contrast with Galors interpretation based on his erroneous concept of the great divergence, all
new trajectories are critically fast. They do not increase to infinity at a fixed time but they pose virtually
the same danger as the historical, linearly-modulated hyperbolic trajectories because they are close to
the trajectories, which increase to infinity at a fixed time.
180
__________________________________________________________________________________
Figure 11. Growth of income per capita in Western Europe (Maddison, 2010; Nielsen, 2016g). Between 1900
and 1913, economic growth in Western Europe started to depart from the historical linearly-modulated
hyperbolic distribution. However, it continues to increase close to the historically-predicted trajectory.
Figure 12. Growth of income per capita in Eastern Europe (Maddison, 2010; Nielsen, 2016g). From around
1850 economic growth in Eastern Europe started to depart from the historical linearly-modulated hyperbolic
distribution. However, it continues to increase close to the historically-predicted trajectory. The growth was not
diverted to a distinctly different and gently-increasing trajectory as claimed by Galor (2005a, 2011; cf Figure
3).
181
__________________________________________________________________________________
Figure 13. Growth of income per capita in countries of the former USSR (Maddison, 2010; Nielsen, 2016g).
Close to around 1870 economic growth in countries of the former USSR started to depart from the historical
linearly-modulated hyperbolic distribution. However, it continues to increase close to the historically-predicted
trajectory.
Figure 14. Growth of income per capita in Asia (Maddison, 2010; Nielsen, 2016g). After a brief decline between
1940 and 1950, the growth of income per capita in Asia was diverted to a slightly faster trajectory. However, it
continues to increase close to the historically-predicted linearly-modulated hyperbolic distribution. The growth
was not diverted to a distinctly different and gently-increasing trajectory as claimed by Galor (2005a, 2011; cf
Figure 3).
182
__________________________________________________________________________________
Figure 15. Growth of income per capita in Africa (Maddison, 2010; Nielsen, 2016g). In clear contradiction of
Galors claim supported by his distorted presentation of Maddisons data, the growth of income per capita did
not diverge to a slowly-increasing trajectory but continues to increase along a nearly vertical trend close to the
historical linearly-modulated hyperbolic distribution (cf Figure 3).
Figure 16. Growth of income per capita in Latin America (Maddison, 2010; Nielsen, 2016g). In clear
contradiction of Galors claim supported by his distorted presentation of Maddisons data, growth of income per
capita continued to increase along a nearly vertical trajectory close to the historical linearly-modulated
hyperbolic distribution. The growth was not diverted to a distinctly different and gently-increasing trajectory as
claimed by Galor (2005a, 2011; cf Figure 3).
183
__________________________________________________________________________________
Figure 17. Growth of income per capita in all regions, including Western offshoots (Maddison, 2010). Even
without carrying out mathematical analysis of data it is clear that they all follow similar, nearly-vertical
trajectories. The mystery of the mind-boggling and perplexing phenomenon of the Great Divergence
(Galor, 2005a, pp. 177, 220) has now been explained there is no mystery. The great divergence never
happened. This mystery, as well as all his other unresolved mysteries about the growth process (Galor,
2005a, p. 220) have been created by the mind-boggling, perplexing and self-misleading manipulation of data.
In Galors distorted presentation of data shown in Figure 3, economic growth in various regions
follows similar trajectories for a long time and then diverges to distinctly different trajectories. In the
correct and undistorted presentation of data shown in Figure 17, economic growth in various regions
follows similar trajectories all the time. Some regions are slower in their economic development but
they all race in the same direction and along virtually the same trajectory. They do not fan out into
distinctly different directions as claimed by Galor.
We do not have to explain the mechanism of the great divergence because the great divergence never
happened. It is a feature created by the distorted presentation of data. If we want to explain the currently
observed differences in the economic growth we should not be misguided by the Unified Growth Theory
and we should not attempt to explain why different regions follow distinctly different trajectories,
because they do not follow distinctly different trajectories. We should rather try to explain why different
regions follow similar trajectories and why for some regions economic growth is faster while for other
regions it is slower.
5. Geometric distortions
We shall now explain how Galor constructed his unresolved mysteries about the growth process
(Galor, 2005a, p. 220): (1) his mind-boggling and perplexing phenomenon of the Great Divergence
(Galor, 2005a, pp. 177, 220) and (2) his equally mind-boggling but fictitious takeoffs from the alleged
but non-existent stagnation to growth. To demonstrate how such mysteries are created, we can take any
close family of distributions, which change slowly over a large range of independent variable and fast
over its short range. We can use hyperbolic distributions, linearly-modulated hyperbolic distributions,
a set of empirical distributions such as shown in Figures 10 and 17, or any other hyperbolic-like
distributions. By their simple manipulation, we can easily create Galors mysteries about the growth
process (Galor, 2005a, p. 220) but they will not be unresolved mysteries. They will not even be
mysteries because we shall demonstrate and explain their origin. We shall demonstrate that these
184
__________________________________________________________________________________
alleged mysteries do not represent unique properties of economic growth but the introduced by us
disfigurations of hyperbolic-like distributions.
For our demonstration, we have chosen three, closely-related linearly-modulated hyperbolic
distributions shown in Figure 18. Like the historical income per capita distributions, each of these
arbitrary distributions is represented by a ratio of two hyperbolic distributions. However, they have
absolutely nothing to do with economic growth. They are purely mathematical functions f ( x) , g ( x)
and h( x) where x is an arbitrary independent variable. This variable could be time but it could be also
anything else. The common feature of these distributions is that they start from approximately the same
value at x = 0 , they increase monotonically (they are not characterised by sudden takeoffs at any time)
and they increase to infinity within a small range of x values. They do not diverge.
Figure 18. Three arbitrarily-chosen, linearly-modulated, hyperbolic distributions, f ( x) , g ( x) and

h( x) . They increase monotonically from approximately the same value at x = 0 to infinity within
approximately the same time. They do not diverge.
However, if we follow Galors example we can use these non-divergent and monotonically-
increasing distributions and construct a new set of diverging distributions, which will be also
characterised by clear takeoffs. All we have to do is to select a few strategically-located points at certain
constant x-values and join them by straight lines. This is precisely what Galor was doing repeatedly
during the formulation of his Unified Growth Theory (Galor, 2005a, 2011) and in his other publications
(Galor, 2005b, 2007, 2008a, 2008b, 2008c, 2010, 2012a, 2012b, 2012c; Galor & Moav, 2002; Snowdon
& Galor, 2008).
We have selected three values of independent variable, x = 0 , x = 150 and x = 179.6 , and by
following Galors example, we have connected the corresponding values of f ( x) , g ( x) and h( x) by
straight lines. We have now constructed typical distributions used by Galor to formulate his Unified
Growth Theory. We have also constructed the great divergence and the takeoffs. Results are shown in
Figure 19.
185
__________________________________________________________________________________
Figure 19. This figure explains how the mind-boggling and perplexing phenomenon of the Great
Divergence (Galor, 2005a, pp. 177, 220) was invented by Galor and how he created his otherwise non-existing
takeoffs from the non-existing stagnation to growth. By following his approach to research, the corresponding
values of the purely-mathematical functions f ( x) , g ( x) and h( x) at x = 0 , x = 150 and x = 179.6 were
joined by straight lines. The monotonically-increasing distributions are now replaced by distorted diagrams in
much the same way as Maddisons data were replaced by Galor by his distorted diagrams. We have constructed
the meaningless mind-boggling and perplexing phenomenon of the Great Divergence (Galor, 2005a, pp.
177, 220) preceded by the equally meaningless takeoffs at x = 150 .
It would be incorrect to claim that our constructed distributions shown in Figure 19 represent the
original distributions, which were shown in Figure 18, but Galor repeatedly and incorrectly uses his
distorted diagrams as representing Maddisons data. His repeatedly used diagrams are the
misrepresentations of data and his conclusions based on such diagrams or on quoting some isolated
numbers selected from hyperbolic distributions are scientifically unacceptable and strongly misleading.
By using the constructed great divergence and the takeoffs shown in Figure 19 and by constructing
more of such diagrams we could now create a unified growth theory describing properties of the
distorted diagrams and insist that they represent mathematical properties of f ( x) , g ( x) and h( x)
functions or the properties of other similar distributions. However, it would be nave for us to expect
that people familiar with mathematics would be impressed by our scholarly performance and by the
mysteries we have created. It would be nave to expect that they would accept our explanations of the
claimed mathematical properties of hyperbolic-like distributions, and yet Galor expects that economists
will accept his distorted representations of Maddisons data and his explanations of economic growth
based on such repeatedly distorted presentations of data reinforced by the numerous quotations of well-
selected and isolated numbers, which are supposed to represent a reliable empirical confirmation of his
theory.
Like Galor, we could claim the existence of takeoffs from stagnation to growth for our mathematical,
monotonically-increasing functions f ( x) , g ( x) and h( x) . We could try to explain these phantom
takeoffs by some fanciful mechanisms, but such explanations would be unacceptable because the
original functions increase monotonically. They are not characterised by sudden takeoffs. These
takeoffs do not exist. We have created them by distorting the original functions in much the same way
as Galor created them by distorting Maddisons data.
186
__________________________________________________________________________________
Like Galor, we could also claim the existence of the great divergence and try to explain it by some
complicated mechanisms but again our claim and our explanations would be unacceptable because the
original functions do not diverge. We have created the great divergence, which does not characterise
the original functions but only their distorted representations. Like Galor, we could claim the existence
of the unresolved mysteries (Galor, 2005a, p. 220) about mathematical functions but the only
audience we could hope to impress would be people who are not familiar with mathematics but it is
also possible that even people unfamiliar with mathematics would soon notice that what we are doing
is just clever or maybe even not so clever sophism.
Conclusions based on the distorted representations of mathematical distributions f ( x) , g ( x) and
h( x) can be obviously rejected. Likewise, conclusions based on Galors distorted representations of
Maddisons data can be and even should be rejected. Galor presents many curious and seemingly logical
stories about economic growth but his stories are either repeatedly contradicted by data (Nielsen, 2014,
2015a, 2016a, 2016b, 2016d, 2016e, 2016f, 2016g, 2016h) or have no convincing confirmation in data.
They have to be accepted largely by faith. Stories of fiction can be also attractive, logical and convincing
but they will remain stories of fiction.
It would be incorrect to claim that the distorted diagrams presented in Figure 19 represent the
mathematical distributions f ( x) , g ( x) and h( x) . Likewise, it would be incorrect to claim that the
distorted diagrams used repeatedly by Galor in his Unified Growth Theory and in his other publications
represent Maddisons data.
It would be incorrect to claim that the distorted diagrams presented in Figure 19 describe the
mathematical functions f ( x) , g ( x) and h( x) . Likewise, it would be incorrect to claim that the distorted
diagrams presented by Galor in his Unified Growth Theory and in his many other publications describe
economic growth. They describe the world of fiction.
It could be hardly expected that explanations of the properties of mathematical functions f ( x) , g ( x)
and h( x) based on their distorted representations shown in Figure 19 could be ever accepted by people
familiar with mathematics. Likewise, it can be hardly expected that explanations of economic growth
based on such distorted presentations of data as used by Galor in his Unified Growth Theory and in his
other publications can be accepted by the scientific community.
6. Discussion and conclusions

We have analysed Maddisons data (Maddison, 2001, 2010) and we have demonstrated that the great
divergence claimed by Galor (2005a, 2011) and shown in Figure 3 never happened. Various regions are
now on different levels of development but their economic growth did not diverge into distinctly
different trajectories as claimed by Galor (see Figure 3). Their income per capita increases along similar,
approximately vertical trajectories.
The disagreement between Galors claim and the data can be demonstrated using the early
Maddisons data (Maddison, 2001), which Galor used in their habitually distorted presentations during
the formulation of his Unified Growth Theory (Galor, 2005a, 2010). However, the disagreement
between his claims and the data becomes even more pronounced if we display the latest data of
Maddison (2010), which were available to Galor before the publication of his book (Galor, 2011).
The data do not even have to be analysed mathematically to show that they contradict Galors claim
of the existence of the great divergence but their mathematical analysis is helpful. Galors claims
expressed in his Unified Growth Theory and in his other similar publications are based on his failure to
adhere to the fundamental and indispensable principles of scientific investigation, which require that
data should be rigorously analysed, that conclusions should not be based on impressions and that data
should not be manipulated to support preconceived ideas. His theory, his claims and his interpretations
are scientifically unacceptable.
The mind-boggling phenomenon of the Great Divergence in income per capita across regions of
the world in the past two centuries, that accompanied the take-off from an epoch of stagnation to a state
of sustained economic growth, presents additional unresolved mysteries about the growth process
(Galor, 2005a, p. 220). It is interesting how a single sentence can contain so much misinformation.
His mysteries have now been solved: he has created them by the manipulation of data.
187
__________________________________________________________________________________
The great divergence never happened and neither did the takeoffs from Malthusian stagnation to
growth (Nielsen, 2014, 2015a, 2016a, 2016b, 2016c, 2016d, 2016e, 2016f, 2016g, 2016h). Within the
range of analysable data there was no stagnation and no transition from stagnation to growth. Features
described by Galor as takeoffs are not takeoffs but the natural continuations of monotonically-
increasing hyperbolic distributions describing the growth of the GDP or population, or the natural
continuations of monotonically-increasing linearly-modulated distributions describing the growth of
the GDP/cap. Hyperbolic distributions or linearly-modulated hyperbolic distributions are slow over a
long time and fast over a short time but they do not change suddenly from slow to fast at any time. They
increase monotonically all the time.
Hyperbolic growth excludes the interpretations revolving around the concept of Malthusian
stagnation and around takeoffs from stagnation to growth described usually as the escape from the
Malthusian trap. The evidence contradicting such interpretations is overwhelming. It is remarkable that
so many independent studies are in such perfect agreement: Maddisons data (Maddison, 2001, 2010)
and their analysis (Nielsen, 2014, 2015a, 2016a, 2016d, 2016e, 2016f, 2016g, 2016h); the estimates of
the size of human population not only during the AD era but also during the BC era (e.g. Biraben, 1980;
Clark,1968; Cook,1960; Durand, 1967, 1974, 1977; Gallant, 1990; Haub, 1995; Livi-Bacci, 1997;
McEvedy & Jones, 1978; Taeuber & Taeuber, 1949; Thomlinson, 1975; Trager, 1994) and their analysis
(e.g. Kremer, 1993; Nielsen, 2016b; Kapitza, 2006); the discovery made by von Foerster, Mora and
Amiot (1960) and similar identifications of hyperbolic growth by Podlazov (2002), Shklovskii (1962,
2002) and von Hoerner (1975).
According to Galor, the differential timing of the take-off from stagnation to growth across
countries, and the corresponding variations in the timing of the demographic transition, led to a great
divergence in income per capita as well as population growth (Galor, 2005a, p. 218). This is a good
example how fiction can be created even in science. Non-existent takeoffs have been constructed by
distorted presentations of data. These non-existent takeoffs were then used to explain the non-existent
differential timing of takeoffs, and now the same phantom takeoffs are used to explain the origin of the
non-existent great divergence constructed by the manipulation of data.
Galor wonders about the underlying driving forces that triggered the recent transition between these
regimes and the associated phenomenon of the Great Divergence in income per capita across countries
(Galor, 2005a, pp. 174, 219). While it is interesting to study reasons for differences in the level of
economic growth of various regions and countries, there is no need to wonder about the underlying
driving forces of the great divergence because the great divergence never happened.
Galor claims that In the course of the Great Divergence the ratio of GDP per capita between the
richest region and the poorest region has widened considerably from a modest 3 : 1 ratio in 1820, to a
5 : 1 ratio in 1870, a 9 : 1 ratio in 1913, a 15 : 1 ratio in 1950, and 18 : 1 ratio in 2001. (Galor, 2005a,
p. 174). All these ratios are probably correct but the conclusion is incorrect because there was no great
divergence. Economic growth in various regions is on different levels of development but it follows the
virtually the same, non-diverging, trajectories.
This is a good example of being guided by impressions and of using them to draw hasty conclusions.
Data have to be rigorously analysed. Using isolated numbers, as done repeatedly by Galor, is likely to
lead to incorrect interpretations particularly if such a use of isolated numbers is combined with the
repeatedly distorted presentation of data, such as shown in Figures1 and 3. Taking shortcuts and using
them to draw hasty conclusions based usually on preconceived ideas and on wished-for interpretations
does not represent scientific process of investigation. The ratios listed by Galor do not prove the
existence of the great divergence. We have already demonstrated that the great divergence never
happened. The listed ratios represent nothing more than hyperbolic growth and different levels of
development along virtually identical trajectories.
Current economic growth in various regions and countries is at different levels of development. For
countries characterised by high human development, income per capita can be as high as tens of
thousands of dollars but for countries characterised by low human development it can be about one
hundred times lower (Nielsen, 2006). However, Maddisons data show that economic growth in all
regions, without exception, is developing along virtually the same trajectories.
Galors interpretation of economic growth is potentially dangerous because it creates a false sense
of security. He shows that gradients of the current economic-growth trajectories are in general small
and consequently the imminent economic crisis is unlikely (see Figure 3).
188
__________________________________________________________________________________
However, data convey totally different information. Economic growth in all regions is now
increasing rapidly along virtually vertical trajectories (see Figures 4-17). They resemble the historical
linearly-modulated trajectories, which increase to infinity at a fixed time. For such trajectories,
economic crisis can be expected because the growth has to be supported by excessively large per annum
increase in the GDP per capita. The created stress can be too high to be manageable over a long time.
There is also a danger of reaching quickly natural limits to growth.
Warning signs can be already seen in Eastern Europe, in countries of the former USSR and in Africa
(Figures 12, 13, 15). Their growth of income per capita suffered reversals but after a certain time it
managed to recover and follow again the nearly vertical trajectories. Certain degree of instability can
be also observed in Latin America (Figure 16).
The preferred option would be to follow now gently-increasing trajectories but all regions, without
exception, appear to be caught up in the general frenzy to increase rapidly their per capita economic
growth. When they are temporarily left behind they soon resume their hazardous race. Current
trajectories do not increase to infinity at a fixed time but they increase to infinity in a short time, which
is hardly a consolation.
All these important warning signs are not even noticed in the Unified Growth Theory. Unified
Growth Theory appears to suggest a prosperous future after an ages-long epoch of a hypothetical
stagnation but the data show that the future of economic growth is approaching rapidly levels of
unsustainability. It has been shown that the world economic growth follows unsustainable trajectory
(Nielsen, 2015b). However, the analysis of Maddisons data presented here suggests that this is a
common danger shared by all regions. There is not a single region, whose economic growth diverged
to a safer trajectory.
The two opposite interpretations of economic growth have also essential impact on research
activities. In order to explain Galors great divergence we would have to explain why there was a
transition to distinctly different trajectories of economic growth. Such attempts would be a waste
financial and human resources and a waste of time because the great divergence never happened. What
we have to explain is why different regions follow virtually the same trajectories and why they follow
such potentially-hazardous, fast-increasing trajectories. Why there is such a strong desire to increase
the GDP per capita so quickly everywhere and how to control these dangerous tendencies.
Galor claims that the transitions from a Malthusian epoch to a state of sustained economic growth
and the emergence of the Great Divergence have shaped the current growth process in the world
economy (Galor, 2005a, p. 221). They did not because there was no emergence of the Great
Divergence. Galor describes phantom features he created by his manipulation of data. These phantom
features could not have shaped the past growth and they do not shape the current growth because they
did not and do not exist. Galor describes the world of fiction and events that never happened. He then
uses these non-existing phenomena to weave his theory around them.
Transitions from the Malthusian epoch of stagnation to a state of sustained economic growth never
happened because there was no stagnation. Economic growth was sustained in the past because it
followed steadily-increasing hyperbolic trajectories. Takeoffs, which are supposed to represent the
claimed transitions from stagnation to growth, never happened (Nielsen, 2014, 2015a, 2016a, 2016b,
2016c, 2016d, 2016e, 2016f, 2016g, 2016h).
Galors claims are based on distorted presentations of data and generally on repeatedly violating the
fundamental principles of scientific investigation. They are based on impressions rather than on the
rigorous scientific analysis of empirical evidence.
Galor claims that the unified growth theory sheds light on the perplexing phenomenon of the Great
Divergence in income per capita across regions of the world in the past two centuries (Galor, 2005a,
p. 177). If it does, then his theory is a fiction because the perplexing phenomenon of the great divergence
never happened.
Why did we devote so much time on the discussion of Galors Unified Growth Theory (Nielsen,
2014, 2015a, 2016a, 2016b, 2016c, 2016d, 2016e, 2016f, 2016g, 2016h)? The answer is simple. As for
Isaak Newton, the aim of any scientific investigation is to discover the truth. It is not a person dislike.
Science looks for correct interpretations but Galors theory is so obviously incorrect that it attracted
immediate attention.
However, there is also another important reason: Galors Unified Growth Theory is not only
incorrect but also dangerously incorrect because it diverts attention from the urgent need to monitor,
189
__________________________________________________________________________________
control and regulate the current economic growth. It would be unwise to accept his theory and his
explanations because his incorrect explanations of the historical economic growth are linked strongly
with the current economic growth, which affects our future.
Galor claims that after a long epoch of stagnation we are now in the regime of sustained economic
growth. His theory also strongly suggests that the current economic growth is not only sustained but
also sustainable because in general it follows slowly-increasing trajectories (see Figure 3). The future
appears to be safe and secure.
However, precisely the same data, which he used during the formulation of his theory, show that the
opposite is true. It was in the past that the economic growth was safe and secure but now it follows
strongly hazardous trajectories. Recent analysis of the world economic growth also indicates that its
future is insecure (Nielsen, 2015b), which is hardly surprising because our current combined ecological
footprint is already significantly higher than the ecological capacity (WWF, 2010).
Why did Galor manipulate data? Why did he repeatedly present distorted diagrams to support his
preconceived ideas? Why did he quote isolated and well-chosen but otherwise meaningless numbers to
support his arguments? Why did he create such an elaborate work of fiction?
If we assume that he did not do it all on purpose, then a possible explanation is that he did not know
how to analyse data. However, this explanation is unconvincing because he appears to be familiar with
mathematics. Anyone familiar with mathematics can see quickly that plots of Maddisons data display
characteristic features of hyperbolic distribution. Anyone familiar with mathematics knows also that
the analysis of hyperbolic distributions is trivially simple (Nielsen, 2014). However, equally surprising
is why his publications escaped the scrutiny of the peer-review system.
The most plausible explanation is probably that he was blinded by prejudice. It is what psychologists
describe as the cascade behaviour, information cascade, informational avalanche, illusion of truth,
illusory truth, illusion of familiarity, running with the pack, following the crowd, herding behaviour,
bandwagons and path depending choice (Anderson & Holt, 1997; Begg, Anas & Farinacci, 1992;
Bikhchandani, Hirshleifer & Welch, 1992, 1998; De Vany & Lee, 1999; De Vany & Walls, 1999;
Easley & Kleinberg, 2010; Grebe, Schmid & Stiehler, 2008; Ondrias, 1999; Parks & Tooth, 2006;
Ramsey, Raafat, Chater & Frith, 2009; Walden & Browne, 2003). It is the fear of being different, of
taking risks, of sticking the neck out, of claiming something, which is not commonly accepted.
In certain areas of intellectual activities, this problem creates nearly insurmountable obstacles. In the
demographic and economic research, this phenomenon is demonstrated by the reluctance to accept the
compelling contradicting evidence simply because many demographers or economists would not agree
with the contradicting evidence. It is safer to follow the crowd and run with the pack. Tradition is
stronger than science and only an outsider who has not been blinded by prejudice and who is not afraid
of being rejected by the crowd might dare to show that the accepted doctrines are incorrect. He or she
is then risking to be ridiculed and rejected but science is a self-correcting discipline so sooner or later
such resistance to accept the overwhelming empirical evidence will have to be overcome, but it would
be better for science and scientists if the required change in the paradigm is accepted sooner rather than
later.
We now have a large body of data (Biraben, 1980; Clark,1968; Cook,1960; Durand, 1967, 1974,
1977; Gallant, 1990; Haub, 1995; Livi-Bacci, 1997; Maddison, 2001, 2010; McEvedy & Jones, 1978;
Taeuber & Taeuber, 1949; Thomlinson, 1975; Trager, 1994), which can be used to improve our
understanding of the economic growth and of the growth of human population. Correct understanding
of these two processes might have essential impact on our future.
The recent mathematical analysis of data (Nielsen, 2014, 2015a, 2016a, 2016b, 2016c, 2016d, 2016e,
2016f, 2016g, 2016h) reveals many interesting features, which call for further investigation. The past
economic growth and the growth of human population were hyperbolic. Within the range of analysable
data, which for the growth of human population extends down to 10,000 BC, there was no Malthusian
stagnation. Hyperbolic growth was slow but remarkably steady. There were no transitions from
stagnation to growth because there was no stagnation. There was no escape from the Malthusian trap in
the economic growth or in the growth of population because there was no trap. There were no takeoffs
from stagnation to growth claimed by Galor (2005a, 2011). There was no differential timing of takeoffs,
claimed also by Galor, because there were no takeoffs.
We have demonstrated (Nielsen, 2016h) that there was no sudden spurt in growth rates of output
per capita (Galor, 2005a, p. 220). Contrary to the similar claim made by Galor, there was also no
190
__________________________________________________________________________________
sudden spurt in the growth rate of human population in the past 12,000 years (Nielsen, 2016b). The
unresolved mysteries about the growth process listed by Galor (2005a, p. 220) have now been solved.
They do not exist. They are phantom mysteries created by Galor through the manipulation of data.
Industrial Revolution had no impact on changing the trajectories of economic growth and of the
growth of population. There was no population explosion. What is perceived as takeoffs or explosions
are just the natural continuations of hyperbolic growth (Nielsen, 2014). There was also no mind-
boggling and perplexing phenomenon of the Great Divergence in income per capita across regions of
the world in the past two centuries (Galor, 2005a, pp. 177, 220).
Recently, economic growth and the growth of human population started to be diverted to slower
trajectories but these new trajectories continue to increase close to the historical hyperbolic trajectories.
Analysis of data shows that not only the Unified Growth Theory but also the Demographic Transition
Theory, which is based on similar assumptions, is repeatedly contradicted by empirical evidence
(Nielsen, 2016c).
All these features suggest new lines of investigation aimed at answering many important questions
about economic growth and about the growth of human population. Why the economic growth and the
growth of human population were hyperbolic. Why the hyperbolic growth was so remarkably stable
over such a long time in the past. Why was it not affected by many random forces that were no doubt
present? Why the economic growth and population growth trajectories were not affected by the
Industrial Revolution. The only exception where there is a correlation between the Industrial Revolution
and the economic growth and the growth of population is Africa, the poorest region. This boosting can
be explained by the colonisation of Africa (Duignan & Gunn, 1973; McKay, Hill, Buckler, Ebrey, Beck,
Crowston, & Wiesner-Hanks, 2012; Pakenham, 1992) rather than by the beneficial effects of the
Industrial Revolution. What models should be used to explain the historical hyperbolic economic
growth and the growth of human population? What are the common features that link these two
processes? Why was the economic growth and the growth of human population diverted relatively
recently to new, non-hyperbolic trajectories? Are these new trajectories likely to change again into the
apparently preferred hyperbolic growth? How to prevent such an undesirable event? What should be
done to make the growth of population and economic growth sustainable? Much work needs to be done
but it would unwise and potentially dangerous to be guided by the Unified Growth Theory.
References
Anderson, L. R., & Holt, C. A. (1997). Information cascades in the laboratory. American Economic Review, 87(5), 847-862.
Begg, I. M., Anas, A., & Farinacci, S. (1992). Dissociation of process in belief: Source recollection, statement familiarity
and the illusion of truth. Journal of Experimental Psychology. 121(4), 446 458.
Bikhchandani, S., Hirshleifer, D., & Welch, I., A. (1992). Theory of fads, fashion, custom, and cultural change as
informational cascades. Journal of Political Economy, 100(5), 992-1026.
De Vany, A., & Lee, C. (1999). Information cascades in multi-agent models.
https://www.researchgate.net/publication/5061066_Information_Cascades_in_Multi-Agent_Models
De Vany, A., & Walls, R. (1999). Uncertainty in the movie industry: Does star power reduce the terror of the box office?
Journal of Cultural Economics, 23, 285318.
Social Science: World Population, 369, 1-8.
Durand, J. (1977). Historical Estimates of World Population: An Evaluation. Population and Development Review, 3(3),
256-293.
Easley, D., & Kleinberg, J. (2010). Networks, crowds, and markets. Cambridge, UK: Cambridge University Press.
Press.
Publishing Company.
191
__________________________________________________________________________________
475.
http://www.econ.brown.edu/Faculty/Oded_Galor/pdf/Klien%20lecture.pdf
0062-7
117, 1133-1191. doi. 10.1162/003355302320935007
Economics, 108(3), 681716.
Books.
Nielsen, R. W. (2015b). The Insecure Future of the World Economic Growth. Journal of Economic and Social Thought,
2(4), 242-255.
Nielsen, R. W. (2016g). Mathematical analysis of historical income per capita distributions.
Ondrias, K. (1999). The brain, consciousness & illusion of truth (E. Nazinska, Trans. & Ed.). Boca Raton, FL: Universal
Publishers.
Avon Books.
Parks, C. M., & Tooth, J. P. (2006). Fluency, familiarity, aging, and the illusion of truth. Aging, neuropsychology, and
cognition, 13, 225253
Ramsey M., Raafat, R. M., Chater, N., & Frith, C. (2009). Herding in humans. Trends in Cognitive Sciences, 13(10), 420-
428.
192
__________________________________________________________________________________
Pub.
691-712.
Walden, E. A., & Browne, G. J. (2003). Running with the pack: An empirical examination of information cascades in the
adoption of novel technologies. Lubbock: Texas Tech. University.
International.
193
__________________________________________________________________________________
12
The dichotomy of Malthusian positive checks:
Destruction and even more intensified regeneration
By Ron W. NIELSEN
Abstract. Impacts of Malthusian positive checks are investigated using data compiled by the United Nations
Development Program. We show how the intensity of Malthusian positive checks are correlated with indicators
describing the standard of living such as ecological footprint, income per capita, severe poverty, access to pure
water and access to sanitation facilities. We then show that mortality increases exponentially with the intensity
of Malthusian positive checks (adult mortality, under-five mortality, maternal mortality and deaths by polluted
water). However, total fertility rate also increases exponentially. One of the important results of our analysis is
that the growth rate and the rate of natural increase are directly proportional to the intensity of Malthusian
positive checks. The regenerating impacts triggered by Malthusian positive checks do not just keep the growth
undisturbed they stimulate growth and make it even faster. Using data for Africa, we also show that while the
death rate increases with the intensity of hunger, total fertility rate, birth rate and the rate of natural increase also
increase. Records from China also show that Malthusian positive checks of wars and famines triggered the
intensified growth of population. Thus, contrary to the generally promoted interpretations, hunger and famines
do not necessarily suppress the growth of population. Similar patterns of stress-induced growth are also observed
in nature. This study is closely related to the problem of controlling the growth of human population.
Keywords. Malthusian positive checks, Population growth, Income per capita, Birth and death rates, Total
fertility rate, Growth rate, Rate of natural increase
JEL. A12, C12, Y80.
1. Introduction
M altus (1798) is well known for a dubious reason of having his name associated with
the erroneous concept of stagnation expressed in such phrases as Malthusian
stagnation, Malthusian regime, epoch of Malthusian stagnation, Malthusian trap and
escape from Malthusian trap, the concept he never proposed or advocated, the concept based
on impressions, on a good dose of fantasy and on the suitable manipulation of data (Ashraf,
2009; Galor, 2005a, 2005b, 2007, 2008a, 2008b, 2008c, 2010, 2011, 2012a, 2012b, 2012c;
Galor & Moav, 2002; Snowdon & Galor, 2008). The concept of stagnation and all other related
concepts have been repeatedly and convincingly contradicted by data and by their
mathematical analyses (Biraben, 1980; Clark,1968; Cook,1960; Durand, 1974; Gallant, 1990;
Haub, 1995; Kapitza, 2006; Kremer, 1993; Lehmeyer, 2004; Livi-Bacci, 1997; Maddison,
2001, 2010; Mauritius, 2015; McEvedy & Jones, 1978; Nielsen, 2013a, 2013b, 2013c, 2014,
2015, 2016a, 2016b, 2016c, 2016d, 2016e, 2016f, 2016g, 2016h, 2016i; Podlazov, 2002;
Shklovskii, 1962, 2002; Statistics Mauritius, 2014; Statistics Sweden, 1999; Taeuber &
Taeuber, 1949; Thomlinson, 1975; Trager, 1994, United Nations, 1973, 1999, 2013; von
Hoerner, 1975, von Foerster, Mora & Amiot, 1960; Wrigley & Schofield, 1981).
Australia.
. +61407201175
Published as: Nielsen, R. W. (2016). The dichotomy of Malthusian positive checks: Destruction and even more intensified
regeneration. Journal of Economic Bibliography, 3(3), 409-433.
194
__________________________________________________________________________________
Malthus (1798) proposed three fundamental mechanisms contributing to the growth of

human population: (1) the devastating effects of positive checks, (2) the regenerating effects of
positive checks and (3) the effects of preventive checks. He did not explain the mechanism of
growth of population. He only claimed that Population, when unchecked, increases in a
geometrical ratio (Malthus, 1978, p. 4). Now we know that this is not true. Population, when
unchecked does not increase in a geometrical ratio (exponentially) but hyperbolically (Kapitza,
2006; Kremer, 1993; Nielsen, 2016b, 2016d; Podlazov, 2002; Shklovskii, 1962, 2002; von
Hoerner, 1975, von Foerster, Mora & Amiot, 1960).
Malthus carried out an important pioneering work but it was for the future generations of
researchers to explore these three proposed contributing mechanisms and to understand their
impacts, if any, on the growth of population. Malthus did not have sufficient data to carry out
such research. He described lethal effects of demographic catastrophes but he never claimed
explicitly that they would create a lasting stagnation in the growth of human population, let
alone that they would produce the epoch of stagnation as it is now erroneously claimed (Galor,
2005a, 2011). Indeed, his claim that population if unchecked increases exponentially seems to
suggest that he did not imagine prolonged or lasting effects of such positive checks.
Now we know that the growth of the world population, for instance, might have been
checked only once in the past 12,000 years by the unusual convergence of no fewer than five
major demographic catastrophes, which introduced only a minor disturbance between AD
1200 and 1400 (Nielsen, 2016d). There is also no convincing evidence of the frequently
occurring devastating effects of Malthusian positive checks in the growth of regional
populations (Nielsen, 2016b). These surprising results could be perhaps explained in two ways:
(1) that the relative impacts of demographic catastrophes were generally too small and (2) that
the devastating effects of positive checks were to a certain degree compensated by their
regenerating effects, which Malthus mentions in his book.
The aim of the current publication is to continue the work of Malthus and to investigate
impacts of positive checks, the work Malthus could not do because he did not have relevant
data. He described the devastating effects of positive checks but he also did not fail to notice
and record their positive effects in stimulating growth.
We shall start from where Malthus was forced to stop and we shall investigate how the
effects of positive checks are reflected in the growth of human population. We shall assume
that the intensity of Malthusian positive checks can be measured by the level of deprivation.
We shall first define this indicator. We shall then see how this indicator is reflected in the
standard of living. Using this indicator, we shall then see how Malthusian positive checks are
reflected in the destructive effects such as the increased death rates. We shall then investigate
the other side of these positive checks and demonstrate how they are reflected in the process
of regeneration, such as in the increased rate of natural increase, the increased growth rate and
the increased total fertility rate. This study will allow us to extend the work of Malthus, which
he published around 200 years ago, and to understand better the effects of his positive checks,
the effects outlined only briefly in his book.
2. Measuring the intensity of Malthusian positive checks

Effects of Malthusian positive checks can be studied conveniently using the data compiled
by the United Nations Development Program (UNDP, 2011). These data are linked with the
three-dimensional Human Development Index (HDI) (UNDP, 2010), which is defined using
the levels of health, education and income. Human Development Index varies between 0 and
1 and measures the level of human development or the level of prosperity. The HDI close to 1
is for prosperous countries.
We could use this index to describe indirectly the intensity of Malthusian positive checks
but then in order to understand the studied correlations, which we are going to present, and to
195
__________________________________________________________________________________
study the effects Malthusian positive checks, we would have to translate mentally the HDI into
the levels of deprivation. In order to link the UNDP data directly to the level of deprivation and
thus to the intensity of Malthusian positive checks, it is better to introduce just a slight
modification of the HDI and define the Level of Deprivation Index (LDI) as:
LDI 1 HDI (1)
This index varies from around 0 (for the low level of deprivation) to around 1 (for the high
level of deprivation). At the high end of the spectrum, this index is linked directly or indirectly
with such conditions as poor health care, low income, severe poverty, inadequate access to
sanitation facilities, inadequate access to pure water, hunger, inadequate housing, poor
education, devastating effects of wars and military conflicts, high incidents of infectious
diseases and all other conditions, which are usually identified as representing the intensity of
Malthusian positive checks. As the level of deprivation decreases, the intensity of Malthusian
positive checks also decreases.
It should be noted that for Malthus positive checks were not necessarily represented by the
usually claimed great calamities such as wars, famines and pestilence. Furthermore, for him,
positive checks did not have to apply to the whole country but to certain groups of people
within a given country or even to individual families.
Notwithstanding, then, the institution of the poor laws in England, I think it will be allowed that
considering the state of the lower classes altogether, both in the towns and in the country, the
distresses which they suffer from the want of proper and sufficient food, from hard labour and
unwholesome habitations, must operate as a constant check to incipient population. (Malthus, 1798,
p.31. Italics added.).
Labour would be ill paid. Men would offer to work for a bare subsistence, and the rearing of
families would be checked by sickness and misery (Malthus, 1798, p.64. Italics added.).
In all mathematical formulae and diagrams presented in this study, x is reserved exclusively
for the LDI, which will be always used as an independent variable, while y will be used for any
relevant dependent variable. It should be also noted that while the derived formulae can have
a general application, the exact values of the constants apply only to the data published by
UNDP (2011).
We shall first examine how the level of deprivation, i.e. how the intensity of Malthusian
positive checks, is reflected in the standard of living represented by such indicators as the
ecological footprint (EF), income per capita, the intensity of severe poverty, access to clean
water and access to sanitation facilities. We shall then examine the devastating effects of
Malthusian positive checks as reflected in the increased morality. Finally, we shall examine
the regenerating effects by showing how growth-promoting indicators depend on the intensity
of Malthusian positive checks.
3. Malthusian positive checks reflected in the standard of living

We shall now present three examples showing how the intensity of Malthusian positive
checks as described by the Level Deprivation Index are correlated with the standard of living.
3.1. Malthusian positive checks reflected in the ecological footprint (EF)
Ecological footprint (EF) measures the level of consumption of natural resources and the
level of the associated damage to the environment (Ewing, Moore, Goldfinger, Oursler, Reed
& Wackernagel, 2010). The footprint is expressed in global hectares per person [gha/cap] of
the biologically productive surface area: crops, grazing, fishing, forests for timber and
firewood, forests for carbon dioxide absorption and land for human habitat.
The dependence of the ecological footprint (EF) on the Level of Deprivation Index (LDI),
based on the UDDP data (UNDP, 2011), is shown in Table 1. Small ecological footprint is
associated with a small consumption of natural resources and with a high intensity of
196
__________________________________________________________________________________
Malthusian positive checks as reflected in the Level of Development Index (LDI). These data
show that only a small fraction of human population is responsible for the excessively large
ecological footprint.
Table 1. The average Level of Deprivation Index (LDI) and the average ecological footprint (EF),
expressed in gha/cap, for various levels of human development, based on the UNDP data (UNDP,
2011).
Level of development LDI EF Population

[gha/cap] [Million]
Very high human development 0.111 5.8 1,130
High human development 0.259 2.5 973
Medium human development 0.370 1.7 3,546
Low human development 0.544 1.2 1,260
The average global ecological footprint in 2011 calculated using the UNDP data (UNDP,
2011) was 3 gha/cap. According to the data shown in Table 1, ecological footprint was higher
than this average value for 16.5% of the world population. For this small groups of people, the
average intensity of Malthusian positive checks was low (LDI = 0.111). In contrast, for 18.4%
of global population, the intensity of Malthusian positive checks was approximately five times
higher (LDI = 0.544). The dependence of the ecological footprint on the level of deprivation is
shown in Figure 1. Each dot in this diagram (and in all other diagrams presented in this
publication) represents one of the 187 countries listed in the UNDP compilation (UNDP, 2011).
Figure 1. Correlation between the intensity of Malthusian positive checks as measured by the LDI
and the ecological footprint (EF). The data are from the United Nations Development Program
(UNDP, 2011). They are compared with two best fits using hyperbolic distributions.
The best fit to the data presented in Figure 1 is obtained by using the second-order
hyperbolic distribution:
y = (a0 + a1 x + a2 x 2 ) 1 , (2)
where x is the LDI and y is the ecological footprint. For this set of data (UNDP. 2011),
a0 = 0.126 , a1 = 4.406 and a2 = 1.139 .
197
__________________________________________________________________________________
However, a satisfactory fit can be also obtained using a much simpler, first-order, hyperbolic
distribution
y = ax 1 , (3)
where a = 0.646 .
For the large values of the LDI, ecological footprint increases slowly with the decreasing
level of deprivation. Thus, in the extreme case, for countries characterised by the high values
of the LDI, i.e. by the high intensity of Malthusian positive checks, a large reduction in the
level of deprivation, and thus in the intensity of Malthusian positive checks can be achieved by
only a relatively small increase in the ecological footprint.
We shall show later how the intensity of growth of population increases with the intensity
of the LDI, i.e. with the intensity of Malthusian positive checks. Figure 1 suggests that large
reductions in the intensity of growth of human population could be achieved by improving
living conditions of poor countries through a relatively small increase in their ecological
footprint.
In contrast, as we can see from the correlation presented in Figure 1, a large increase in the
ecological footprint of rich countries (characterised by the low LDI values) results in only
marginal improvement in their standard of living. Their standard of living is already so high
that to improve it by only a small degree requires enormous increase in their consumption of
natural resources, which is not only unfair for poor countries but also imprudent because a
better distribution of wealth could contribute significantly to reducing the growth of human
population and to the global security.
3.2. Malthusian positive checks reflected in the income per capita
The dependence of the Gross Domestic Product per person (GDP/cap) on the level of
deprivation is shown in Figure 2.
Figure 2. Exponential dependence of the Gross Domestic Product per person (GDP/cap) on the Level
of Deprivation Index (LDI). The GDP is in the purchasing power parity of 2009 international dollars
The GDP/cap decreases exponentially with the increasing level of deprivation, i.e. with the
intensity of Malthusian positive checks. The range of the GDP/cap is between $319 for the
Demographic Republic of Congo and $91,379 for Qatar, with the US ($45,989) and
Switzerland ($45,224) being located in the middle.
The best fit to the data is obtained using exponential function,
198
__________________________________________________________________________________
y = be rx , (4)
where x is the LDI and y is the GDP/cap. For these particular set of data (UNDP, 2011),
b = $71,144 and r = 6.97 .
Results presented in Figure 2 lead to the same conclusions as results shown in Figure 1. At
the far end of the LDI scale, a small increase in the GDP/cap by only around $4,000, on average,
would advance countries from the low to medium level of human development. In contrast, an
increase by around $20,000 would be needed to advance countries from high to very high level
of human development. We can also look at it by comparing the increase in the GDP/cap
needed to decrease the LDI by the same interval. For instance, to decrease the LDI from 0.500
to 0.400 one would need to increase, on average, the GDP/cap by only $2,000. In contrast, to
decrease the LDI from 0.300 to 0.200 one would need to increase the GDP/cap by around
$9,000.
3.3. Malthusian positive checks reflected in the severe poverty and in other related
indicators
The dependence of the fraction of the population living in severe poverty on the intensity of
Malthusian positive checks is shown in Figure 3.
Figure 3. The fraction of population living in severe poverty (y) represented as a function of the LDI
(x), i.e. as a function of the intensity of the Malthusian positive checks.
It is essential to notice two important features of the correlation presented in Figure 3. First,
the correlation is linear. Second, the correlation is characterised by a certain threshold below
which the fraction of the population living in severe poverty is on average zero. The intensity
of Malthusian positive checks decreases linearly with the level of poverty. However, when the
fraction of the population living in severe poverty reaches its zero value, the intensity of
Malthusian positive checks reaches a certain threshold level. Any further decrease in the
intensity of Malthusian positive checks is no longer correlated with the level of severe poverty,
but it will continue to be correlated with the ecological footprint and with income per capita.
This linear correlation indicates that severe poverty can be reduced even to zero without
trying to reduce the intensity of Malthusian positive checks to zero. The reduction in the level
of severe poverty will come first. After that, there could be other improvements, which would
be reflected in other parameters describing the standard of living.
Similar step-wise linear correlations apply also to the fraction of the population living below
the poverty line and to the Multidimensional Poverty Index (MPI). The same type of
199
__________________________________________________________________________________
correlations applies also to the size of the population with no access to clean water and to the
size of the population with no access to sanitation facilities. All these data can be fitted using
a simple mathematical expression:
=y m( x n) , y 0, (5)
where n is a threshold below which y 0 .

For the data presented in the UNDP report (UNDP, 2011) and for the fraction of the
population living in severe poverty, m = 190 and n = 0.358. In terms of the ecological footprint
and of the GDP/cap, this threshold corresponds to 1.7 gha/cap and $5,858/cap, respectively.
Above these thresholds, the fraction of the population living in severe poverty could be
expected to be on average negligibly small.
For all indicators, mentioned earlier and characterised by such linear correlations, and for
the data listed in the UNDP report (UNDP, 2011) the respective thresholds in the LDI vary
between 0.293 and 0.371. In terms of the EF and the GDP/cap, they vary between 1.6 and 2.1
gha/cap for the EF, and between $5,350 and $9,218 for the GDP/cap. These figures suggest
that a moderate improvement in the living conditions of poor countries could have an enormous
impact on reducing the level of poverty and on improving access to clean water and to
sanitation facilities. We shall see later that the added benefit of improving the standard of living
in poor countries could be a significant reduction in the growth of population.
4. The lethal impacts of Malthusian positive checks

We can now take the next step and try to understand the destructive effects of Malthusian
positive checks. An example of the dependence of mortality on the intensity of Malthusian
positive checks is shown in Figure 4 for adult mortality.
Figure 4. Exponential dependence of adult mortality (per 1000 adult population) on the level of
deprivation (LDI), i.e. on the intensity of Malthusian positive checks.
The effect is quite remarkable: mortality increases exponentially. We can analyse other
relevant data (UNDP, 2011) and we shall get consistently similar results. Such exponential
increase applies to deaths due to polluted water, maternal mortality and under-five mortality.
200
__________________________________________________________________________________
It would appear that the exponential dependence of mortality on the intensity of Malthusian
positive checks could be expected to apply also to other forms of mortality.
Malthus did not study how mortality depends on the intensity of positive checks. He only
pointed out that positive checks can be linked with the increased mortality. Now we know not
only that mortality increases with the intensity of positive checks but also how it increases it
increases exponentially.
The exponential distribution shown in Figure 4 is described by the eqn (4) but now with the
positive parameter r. For the set of data listed by UNDP (2011), b = 124.83 and r = 2.925.
Adult mortality is on average 66% higher in countries characterised by low human
development than in countries characterised by medium human development, nearly 130%
higher than in countries characterised by high human development and 255% higher than in
countries characterised by very high human development.
5. The regenerating impacts of Malthusian positive checks

Intuitively, one might expect that high mortality should be reducing the size of population
and thus that it should be suppressing growth. This is what Malthus expected. These facts
seem to shew that population increases exactly in the proportion that the two great checks to it,
misery and vice, are removed, and that there is not a truer criterion of the happiness and
innocence of a people than the rapidity of their increase (Malthus, 1798, p. 34).
Thus, according to Malthus, the smaller is the intensity of misery and vice, the faster should
be the growth of the population. Furthermore, his comment suggests that it should be a linear
correlation.
If his interpretation of growth is correct, we should expect that the growth rate of population
should be decreasing with the increasing intensity of Malthusian positive checks, i.e. with the
increasing level of deprivation. Furthermore, the high growth rate could be used as an indicator
of the happiness and innocence because there is not a truer criterion of the happiness and
innocence of a people than the rapidity of their increase.
We are now going to show that Malthusian positive checks stimulate growth, which is hardly
surprising because it is well known that poor countries are characterised by a rapid growth of
population. Malthus observed this phenomenon of stress-induced growth but he did not follow
his observation by a closer investigation perhaps because his access to relevant data was
strongly limited. It is also obvious that rapid growth of population in poor countries does not
contribute to their happiness.
5.1. Total fertility rate increases exponentially with the intensity of Malthusian positive
checks
Total fertility rate is defined as the number of children that would be born to each woman
if she were to live to the end of her child-bearing years and bear children at each age in
accordance with prevailing age-specific fertility rates (UNDP, 2011, p. 142). The dependence
of total fertility rate on the level of deprivation, i.e. on the intensity of Malthusian positive
checks is shown in Figure 5.
Analysis of the UNDP data (UNDP, 2011) leads to remarkable results. It shows that while
morality increases exponentially with the intensity of Malthusian positive checks, total fertility
rate also increases exponentially. This is the first and important indication that the growth of
human population is not slowed down by the increased mortality.
201
__________________________________________________________________________________
Figure 5. Total fertility rate increases exponentially with the intensity of Malthusian positive checks,
i.e. with the increasing level of deprivation (LDI).
5.2. Growth rate is directly proportional to the intensity of Malthusian positive checks
The correlation between the level of deprivation and the growth rate is shown in Figure 6.
The data show that, on average, the annual growth rate is directly proportional to the level of
deprivation, i.e. to the intensity of Malthusian positive checks. The larger is the intensity of
Malthusian positive checks the larger is the growth rate.
Figure 6. The dependence of the annual growth rate on the level of deprivation, i.e. on the intensity of
Malthusian positive checks.
Contrary to the intuitive expectations and contrary to the repeated claims of the existence of
the mythical epoch of Malthusian stagnation, the growth of human population is not decreased
by the Malthusian positive checks but increased. However, when the intensity of Malthusian
positive checks is exceptionally high and when they continue over a long time, the growth of
202
__________________________________________________________________________________
population might be temporarily slowed down. This effect was observed in the growth of the
world population but even then, the temporary disturbance in the growth of population was
followed by their more accelerated growth (Nielsen, 2016d).
The straight line fitting the empirical growth rate data shown in Figure 6 is given by
=y m( x n) , (6)
where for this particular set of data (UNDP, 2011) m = 4.3 and n = 0.
In countries characterised by the low level of human development, and consequently
experiencing high intensity of deprivation and of the associated mortality, the growth of
population as given by the UNDP data (UNDP, 2011) was on average about 5 times faster than
in countries characterised by the very high human development and experiencing the low level
of deprivation and mortality.
5.3. Rate of natural increase is directly proportional to the intensity of Malthusian positive
checks
The rate of natural increase is defined as the difference between the death and birth rates
and thus excludes the immigration and emigration rates. The correlation between the rates of
natural increase and the levels of deprivation can be studied by using the 2002 data for the rates
of natural increase (US Census Bureau, 2002) and the 2002 data for the HDI extrapolated from
the tabulated data for 2000 and 2005 (UNDP, 2011). Results are presented in Figure 7.
Figure 7. The dependence of the rate of natural increase on the level of deprivation, i.e. on the
intensity of Malthusian positive checks.
The fitted straight line, represented by the eqn (6), corresponds to m = 3.9 and n = 0 . These
data show that the rate of natural increase is also directly proportional to the level of
deprivation.
In principle, the rate of natural increase gives a better representation of the impacts of
Malthusian positive checks because it is not obscured by contributions from immigrations and
emigrations. However, by comparing Figures 6 and 7, we can see that the linear dependence
applies to the growth rate and to the rate of natural increase suggesting that contributions from
immigration and emigration are in general negligibly small. Individual points might be shifted
but the general trend is the same. The gradients of the straight lines fitting the data are also
similar, 4.3 for the growth rate and 3.9 for the rate of natural increase. It does not matter whether
203
__________________________________________________________________________________
we are using growth rate or the rate of natural increase, results are the same: contrary to the
widely accepted but erroneous doctrine of Malthusian stagnation, the intensified presence of
Malthusian positive checks is correlated with the intensified growth of population.
Results presented in Figures 6 and 7 are most surprising. The regenerating impacts of
Malthusian positive checks do not just keep the growth rate constant they stimulate growth
and make it even faster.
5.4. Devastating impacts of Malthusian positive checks are correlated with their
regenerating impacts
The dichotomy of Malthusian positive checks can be also illustrated by the correlations
between their opposite impacts, the correlations between destructive and regenerating effects.
We could study how various forms of mortality (adult mortality, under-five mortality, maternal
mortality or deaths by polluted water) are correlated with various forms of regenerating effects
(total fertility rate, growth rate and the rate of natural increase). For the listed here effects we
would have 12 such correlations. However, this study of multiple correlations can be reduced
to a study of two types of correlations: (1) correlations between exponential distributions
describing total fertility rate and various forms of mortality, and (2) correlations between
exponential and linear distributions, with exponential distributions representing various forms
of mortality while linear distributions representing growth rate or the rate of natural increase.
As an example, we show two such correlations: (1) the correlation between adult mortality
(exponential) and total fertility rate (exponential), displayed in Figure 8 and (2) the correlation
between adult mortality (exponential) and the growth rate (linear), displayed in Figure 9.
5.4.1. Correlations between exponential impacts
Figure 8. The dichotomy of Malthusian positive checks: the increasing destructive impacts (mortality)
are correlated with the increasing total fertility rate. Similar correlations exist also for under-five
mortality, maternal mortality and deaths by polluted water. Parameters are given by the correlations
of the LDI with adult mortality and with total fertility rate.
As we have already seen, adult mortality and total fertility increase exponentially. If we
represent mortality by y:
y = berx (7)
204
__________________________________________________________________________________
and fertility by z:
z = be r x , (8)
then the correlation between y and z is given by
z = Ae B ln y , (9)
where
r
B= (10)
r
and
A = be B ln b . (11)
The eqn (9) can be also expressed as
=
ln z ln A + B ln y (12)
5.4.2. Correlations between linear and exponential impacts
Figure 9. The dichotomy of Malthusian positive checks: the increasing destructive impacts (mortality)
are correlated with the increasing growth rate. Similar correlations exist also for under-five
mortality, maternal mortality and deaths by polluted water as well as between the rate of natural
increase and all these forms of mortality. Parameters are given by the correlations of the LDI with
adult mortality and with the annual growth rate.
Correlations between the growth rate or the rate of natural increase and various forms of
mortality also illustrate the dichotomy of Malthusian positive checks. As we have seen, the rate
of natural increase and the growth rate increase linearly with the intensity of Malthusian
positive checks while mortality increases exponentially. If we follow similar procedure as
outlined earlier for the correlations between mortality and fertility, we shall find that
205
__________________________________________________________________________________
ln y ln b
=z m n+
r
(13)
r
where z is the annual growth rate or the rate of natural increase and y is the mortality, which
could be adult mortality, under-five mortality, maternal mortality or deaths by polluted water.
For the UNDP data (UNDP, 2011), n = 0 and this formula is reduced to
m
=z [ln y ln b] (14)
r
An example of these correlations is shown in Figure 9. As the intensity of the destructive

effects (mortality) of Malthusian positive checks increases, the intensity of the regenerating
effects also increases and is reflected in the increased growth rate or in the increased rate of
natural increase.
It is generally believed that Malthusian positive checks cause stagnation in the growth of
population and in the associated economic growth (see for instance Artzrouni & Komlos, 1985;
Desment & Parente, 2012; Galor, 2005a, 2011; Galor & Weil,1999, 2000; Guest & Almgren,
2001; Hansen & Prescott, 2002; Komlos,1989, 2000; Komlos & Baten, 2003; Lagerlf, 2003a,
2001b; Lee,1997; Leibenstein, 1957; McKeown, 1983, 2009; Nelson, 1956; Olshansky & Ault,
1986; Omran, 1971, 1983, 1998, 2005; Robine, 2001; van de Kaa, 2010; Vollrath, 2011; Wang,
2005; Warf, 2010; Weisdorf, 2004). The concept of Malthusian stagnation is at the root of the
established knowledge in demography and in economic research, the knowledge, which is
largely based on conjectures, impressions and even on distorted presentations of data (Ashraf,
2009; Galor, 2005a, 2005b, 2007, 2008a, 2008b, 2008c, 2010, 2011, 2012a, 2012b, 2012c;
Galor & Moav, 2002; Snowdon & Galor, 2008).
As discussed elsewhere (Nielsen, 2016j), the currently established knowledge in
demography and in economic research, based on the concept of Malthusian stagnation, is
scientifically unacceptable. There is now an overwhelming evidence that the so-called
Malthusian stagnation never existed in the growth of population and in the economic growth
(Biraben, 1980; Clark,1968; Cook,1960; Durand, 1974; Gallant, 1990; Haub, 1995; Kapitza,
2006; Kremer, 1993; Lehmeyer, 2004; Livi-Bacci, 1997; Maddison, 2001, 2010; Mauritius,
2015; McEvedy & Jones, 1978; Nielsen, 2013a, 2013b, 2013c, 2014, 2015, 2016a, 2016b,
2016c, 2016d, 2016e, 2016f, 2016g, 2016h, 2016i; Podlazov, 2002; Shklovskii, 1962, 2002;
Statistics Mauritius, 2014; Statistics Sweden, 1999; Taeuber & Taeuber, 1949; Thomlinson,
1975; Trager, 1994, United Nations, 1973, 1999, 2013; von Hoerner, 1975, von Foerster, Mora
& Amiot, 1960; Wrigley & Schofield, 1981). The investigation of the UNDP data (2011)
contributes to the explanation why there was no stagnation. One of the contributing factors was
the dichotomy of Malthusian positive checks. As originally noticed by Malthus and as now
confirmed by the study of the UNDP data (UNDP, 2011) the destructive action of Malthusian
positive checks is accompanied by their regenerating impacts. Destruction induces
regeneration.
5.5. Summary of the observed correlations
Summary of the observed correlations between the Level of Deprivation Index (LDI)
representing the intensity of Malthusian positive checks and a series of indicators illustrating
the standard of living, the destructive impact of Malthusian positive checks and their
regenerating impacts is presented in Table A1 (in the Appendix). This table includes also
correlations between the destructive and regenerating impacts of Malthusian positive checks.
5.6. Hunger and famines are correlated with the intensified growth of population
206
__________________________________________________________________________________
The dichotomy of Malthusian positive checks can be also studied by investigating impacts
of hunger. As with other, specifically mentioned indicators describing levels of deprivation,
hunger is just the reflection of the whole spectrum of Malthusian positive checks. Natural
disasters, climatic shocks, conflict, and insecurity are major causes of hunger. But hungers
root causes are tied to a lack of access by individuals to the resources they need to produce,
sell, and buy food (Sheeran, 2008, p. 180). The tragic fact is that, although our planet
produces enough food for everyone, one person in seven still goes to bed hungry each night.
25,000 people die every day including one child every 5 seconds from hunger-related
causes (Sheeran, 2008, p. 180). The overall finding is that 3.1 million children younger than
5 years die every year from undernutrition; that is a staggering 45% of total child deaths in
2011 (Horton & Lo, 2013, p. 371).
Hunger appears to be one of the leading causes of death in the world. Every year over 10
million people die of hunger and hunger-related diseases. Nearly six million of these are
children under the age of five; that is one childs death approximately every six seconds.
(Gibson, 2012, p. 18). This should be compared with other leading causes of death in the world
in 2012: ischaemic heart disease, 7.4 million deaths per year; stroke, 6.7 million; COPD, 3.1
million; lower inspiratory infections, 3.1 million; trachea bronchus lung cancers, 1.6 million;
HIV/AIDS, 1.5 million; diarrhoeal diseases, 1.5 million; diabetes mellitus, 1.5 million; road
injury, 1.3 million; hypertensive heart disease, 1.1 million (WHO, 2014).
Again, it is repeatedly but erroneously claimed that lethal effects of hunger and famines
suppress the growth of human population and create a stagnant state of growth. We shall
demonstrate that such is not the case. These popular and widely-accepted interpretations are
incorrect. They are based on scientifically unsupported dogmas (Nielsen, 2016j).
5.6.1. Evidence from Africa
Table A2 (in the Appendix) and Figure 10 present a series of growth-related indicators for
two groups of African countries, one group where hunger stress is 35% and another where
hunger stress is less than 5%.
Data presented in Table A2 and Figure 10 are based on the examination of three sources of
reference (PRB, 2010; UNDP, 2011; WFP, 2010). In Figure 10, birth and death rates are
expressed in per cent while infant mortality rate in per cent of live births.
Table A2 and Figure 10 show that on average, and for these set of data, countries exposed
to high level of hunger stress experience 71% higher intensity of Malthusian positive checks
as expressed by the LDI and have strongly reduced access to natural resources, as reflected in
their ecological footprint, when compared with countries experiencing low hunger stress.
Countries with high hunger stress experience 39% higher death rate and a massive 120% higher
infant mortality rate. However, for these countries, total fertility rate is 47% higher, birth rate
is 35% higher, the rate of natural increase is also 35% higher and the population increase factor
is 26% higher, all these indicators showing that the natural response to the lethal Malthusian
checks is the increased rate of procreation and the intensified process of regeneration. Thus,
contrary to the generally accepted interpretations, hunger does not reduce the growth of human
population but is associated with a faster growth.
It is, of course, impossible to isolate hunger as a single stress factor. Malthusian positive
checks are interconnected. However, if for instance, hunger stress increases the susceptibility
to infectious diseases, then the primary stress factor is still hunger.
Data make it clear that in countries suffering high hunger stress population growth is faster
than in countries experiencing significantly lower hunger stress. Clearly, Malthusian positive
checks bring not only the destruction as reflected in the increased death tolls but also the
regeneration. Furthermore, the regeneration process is more powerful than the process of
destruction because the growth of population is not just at the same level as in countries
experiencing low hunger stress but faster.
207
__________________________________________________________________________________
Figure 10. The dichotomy of Malthusian positive checks as reflected in the intensity of hunger stress.
Populations suffering high hunger stress experience a higher death rate and a higher rate of infant
mortality than populations experiencing a small hunger stress. However, populations suffering high
hunger stress are also characterised by a higher total fertility rate, higher birth rate, higher rate of
natural increase and higher population increase factors.
5.6.2. Evidence from China

A prominent example of devastating impacts of Malthusian positive checks associated with
famines is China. Between the years 108 B.C. and 1911 AD, there were 1828 famines or one
nearly every year in some of the provinces. Untold millions have died of starvation. In fact, the
normal death rate may be said to contain a constant famine factor (Mallory, 1926, p. 1).
However, China is also an excellent example of regenerating impacts of Malthusian positive
checks. In spite of the tremendously high death rate, particularly of infants, due to lack of
modern medical knowledge, in spite of the depopulating effect of terrible famines, and in spite
of the immense loss of life caused by civil wars we find today a denser population on the plains
than ever before; and since there has been no appreciable influx from other countries we must
ascribe the present conditions to the excessive birth rate (Mallory, 1926, p. 87).
6. The dichotomy observed by Malthus and in nature

While Malthus is well known for suggesting lethal effects of positive checks, he is not so
well known for being aware of the existence of a competing mechanism, the mechanism of
spontaneous regeneration and preservation of identity.
208
__________________________________________________________________________________
The absolute population at any one period, in proportion to the extent of territory, could
never be great, on account of the unproductive nature of some of the regions occupied; but
there appears to have been a most rapid succession of human beings, and as fast as some were
mowed down by the scythe of war or of famine, others rose in increased numbers to supply
their place. Among these bold and improvident Barbarians, population was probably but little
checked, as in modern states, from a fear of future difficulties (Malthus, 1798, p. 15. Italics
added).
If Malthus had access to the data available to us he would have probably presented a more
appropriate description of people exposed to lethal effects of positive checks. We would be
reluctant to describe people living in poor countries as bold and improvident Barbarians. We
also would not like to use the same description to parents of the post-war baby boomers.
Apart from this general observation, Malthus presents also data, which suggest that high-
intensity positive checks are linked with the intensified process of regeneration (Malthus, 1798,
pp. 36-40). He has noticed, for instance that the greatest proportion of births to burials, was
in the five years after the great pestilence (Malthus, 1798, p. 37). He concludes that Great
and astonishing as this difference is, we ought not to be so wonder-struck at it as to attribute it
to the miraculous interposition of heaven. The causes of it are not remote, latent and
mysterious; but near us, round about us, and open to the investigation of every inquiring mind
(Malthus, 1798, p. 40). He was convinced that these incidents of intensified growth, after the
episodes of epidemics, were not only the manifestation of the natural law of growth but also
that they should be closely examined. It is, therefore, disappointing that numerous scholars
who refer to the work of Malthus overlooked his suggestion that the effects of regeneration
should be further examined. Malthus also lists many examples of successful regeneration in
such places as Flanders, Palestine, London, Turkey, Egypt, China, Naples and Lisbon (Malthus,
1798, p. 35).
While emphasising the importance of food in supporting the growth of human population,
he did not fail to notice that people can live almost upon the smallest possible quantity of
food (Malthus, 1798, p. 41), which implies that hunger or famines should not be immediately
identified as factors controlling the growth of population. It is incorrect to suggest that
Malthusian positive checks (mortality crises) maintained a long-run equilibrium between
population size and the food supply (Komlos, 1989, p. 194). It is incorrect to claim that the
food-controlled homeostatic equilibrium had prevailed since time immemorial (Komlos,
2000, p. 320). It is not immediately obvious the Throughout human history, epidemics, wars
and famines have shaped the growth path of population (Lagerlf, 2003b, p. 435).
Malthus uses China as an example where the lower classes of people are in the habit of
living almost upon the smallest possible quantity of food and are glad to get any putrid offals
that European labourers would rather starve than eat (Malthus, 1798, p. 41). He also cautions
against using food as a factor controlling the growth of population. He points out twice in his
book that in some cases population may permanently increase without a proportional increase
in the means of subsistence (Malthus, 1798, pp. 41, 43). Furthermore, he points out that there
could be some variations in the proportion between the number of inhabitants and the quantity
of food consumed, arising from the different habits of living that prevail in each state
(Malthus, 1798, p. 42). Food consumption is not proportional to the size of population. The
relation between the growth of population and food consumption is not immediately obvious.
Malthus placed a significant emphasis on the role of positive checks but he also made an
attempt to present a balanced interpretation of growth, a balanced view which is conspicuously
missing in the numerous publications referring to Malthus and describing erroneously the
effects of Malthusian positive checks as Malthusian stagnation. While making attempts to
praise Malthus for his work such publications are in fact diminishing the importance of his
contribution.
209
__________________________________________________________________________________
His early observations, combined with the vast body of data available now to us, help to
understand the mechanism of growth of human population. It is a process, which might be
influenced by the devastating impacts of Malthusian positive checks but also a process, which
is influenced by their regenerating impacts.
The phenomenon of regeneration noticed and recorded by Malthus is similar to the well-
known process observed in nature. It is the natural and spontaneous process of self-preservation
of living organisms triggered by stressful conditions. It is the resilience of ecological systems
(Holling, 1973). There are numerous examples and definitions of this omnipresent
phenomenon.
According to Cumming et al. (2005, p. 976), resilience is the ability of the system to
maintain its identity in the face of internal change and external shocks and disturbances. The
definition proposed by the National Research Council (NRC) is the continued ability of a
person, group, or system to adapt to stress such, as any sort of disturbance so that it may
continue to function, or quickly recover its ability to function, during and after stress (NRC,
2011, pp. 13, 14). Resilience is the ability to handle stresses or recover from disturbances or
shocks (Bapna, McGray, Mock & Withey, 2009, p. 3). In general, resilience refers to a
systems capacity to deal with change and to continue to develop (Boyd, et al., 2008, p. 391).
There are also other definitions of resilience, all describing either the ability of a quick and
efficient recovery or the ability to cope with stress.
Malthus noticed the existence of this mechanism of regeneration. This process is also well
known in science but for reasons, which are hard to understand, it is overlooked in publications
based on the erroneous assumption of the existence of the epoch of Malthusian stagnation.
While repeatedly describing the lethal effects of Malthusian positive checks, a balanced
interpretation suggested originally by Malthus is missing.

Using the UNDP data (UNDP, 2011), we have investigated impacts of Malthusian positive
checks. We have assumed that a convenient way of measuring the intensity of Malthusian
positive checks is to use the Level of Deprivation Index (LDI), which we defined using the
well-known Human Development Index (HDI). This approach allows not only for studying
impacts of Malthusian positive checks but also for describing them mathematically. Our
empirical formulae are simple but mathematical formulae do not have to be complicated to be
useful.
First, we have investigated how the intensity of Malthusian positive checks is reflected in
the standard of living as represented by the ecological footprint (EF), income per capita,
(GDP/cap), levels of severe poverty, access to clean water and access to sanitation facilities.
We have found that the ecological footprint (EF) decreases hyperbolically with the intensity of
Malthusian positive checks while the GDP/cap decreases exponentially. We have also found
that severe poverty, inadequate access to clean water and to sanitation facilities depend linearly
on the intensity of Malthusian positive checks. However, we have also found that these linear
correlations are characterised by certain thresholds. Thus, our analysis indicates that severe
poverty can be eliminated without the necessity of reducing the intensity of Malthusian positive
checks or equivalently the levels of deprivation to zero or close to zero. The level of the severe
poverty reaches its zero value at a certain threshold of the intensity of Malthusian positive
checks. A significant reduction in the level of severe poverty can be achieved by only a
relatively small increase in the average income per capita in poor countries.
We have investigated the lethal effects of Malthusian positive checks and we have found
that mortality increases exponentially with the increasing level of deprivation, i.e. with the
increasing intensity of Malthusian positive checks. However, we have found that total fertility
rate also increases exponentially with the increasing intensity of Malthusian positive checks.
210
__________________________________________________________________________________
One of the important results of our analysis is that growth rate increases with the intensity
of Malthusian positive checks. The rate of natural increase also increases. The larger is the
intensity of the destructive impacts of Malthusian positive checks, the faster is the growth of
population. The destructive impacts of Malthusian positive checks are not just balanced by
their regenerating process they stimulate an even faster growth. These results suggest that the
essential step in controlling the growth of population is to reduce the levels of severe poverty.
Helping poor countries to help themselves is not an option.
We have also investigated correlations between destructive and regenerating impacts of
Malthusian positive checks and again we have derived simple mathematical formulae
describing these correlations. We have demonstrated that the intensity of the regenerating
process increases with the increasing intensity of the destructive process of Malthusian positive
checks. We have derived a general formula showing how the total fertility rate increases with
the increasing mortality. We have also derived a simple mathematical formula showing how
the rate of natural increase and the growth rate increase with the increasing mortality. We have
presented diagrams for the adult mortality but the same formulae apply also to other forms of
mortality such as maternal mortality, under-five mortality and the mortality caused by polluted
water.
Our investigation shows that contrary to the interpretations based largely on intuition and
impressions, growth of population is not controlled by the increased mortality. On the contrary,
the increased mortality stimulates growth. Our study suggests that in order to have better
control of the growth of human population, levels of deprivation experienced by poor countries
should be significantly reduced. The first and the essential step is to improve the economic
status of these countries. However, helping poor countries to increase their income per capita
is only a partial solution. This step should be accompanied by making a wider range of
accessible options such as options for education and employment available to people living in
poor countries. The improvement of economic status should also go hand in hand with the
improvement in gender equality, which will facilitate better family planning. Only by
improving living conditions of poor countries we can hope to have a better, long-term, control
on the growth of human population and on its stabilisation. Successful control of the growth of
human population is essential for controlling our ever-increasing ecological footprint (Ewing,
Moore, Goldfinger, Oursler, Reed & Wackernagel, 2010; WWF, 2010) and for finding at least
some solutions to the current critical trends shaping the future of our planet (Nielsen, 2006).
References
Artzrouni M, & Komlos J. (1985). Population Growth through History and the Escape from the Malthusian Trap: A
Bapna, M., McGray, H., Mock, G., & Withey, L. (2009). Enabling Adaptation: Priorities for Supporting the Rural Poor in a
Changing Climate. World Resources Institute Issue Brief. http://pdf.wri.org/issue_brief_enabling_adaptation.pdf
Boyd, E., Osbahr, H., Ericksen, P. J., Tompkins, E. L., Lemos, M. C., & Miller, F. (2008). Resilience and Climatizing
Development: Examples and policy implications. Development, 51(3), 390396. http://dx.doi.org/10.1057/dev.2008.32
Cumming, G. S., Barnes, G., Perz, S., Schmink, M., Sieving, K. E., Southworth, J., Binford, M., Holt, R. D., Stickler, C., &
Van Holt, T. (2005). An exploratory framework for the empirical measurement of resilience. Ecosystems 8:975-987.
Desment, K., & Parente, S. L. (2012). The Evolution of Markets and the Revolution of Industry: A Unified Theory of
Growth. Journal of Economic Growth, 17, 205-234. (http://dx.doi.org/10.1007/s10887-012-9080-y)
Ewing, B., Moore, D., Goldfinger, S., Oursler, A, Reed, A., & Wackernagel, M. (2010). The Ecological Footprint Atlas
2010. Oakland: Global Footprint Network.
Publishing Company.
211
__________________________________________________________________________________
475.
0062-7
117, 1133-1191. doi. 10.1162/003355302320935007
Galor, O., & Weil, D. N. (1999). From Malthusian Stagnation to Modern Growth. The American Economic Review, 89, 150-
154. (http://dx.doi.org/10.1257/aer.89.2.150)
Transition and Beyond. The American Economic Review, 90, 806-828. (http://dx.doi.org/10.1257/aer.90.4.806 )
Guest, A. M., & Almgren, G. (2001). Demographic Transition. In Encyclopedia of Sociology (2nd ed.), Vol. 1, (pp. 621-
630). Detroit: Macmillan Reference.
Gibson, M. (2012). Food Security A Commentary: What Is It and Why Is It So Complicated? Food, 1, 18-27.
Holling, C.S. (1973). Resilience and stability of ecological systems. Annual Review of Ecology and Systematics 4, 123.
doi:10.1146/annurev.es.04.110173.000245
Horton, R., & Lo, S. (2013). Nutrition: A quintessential sustainable development goal. The Lancet, 382, 371-372.
History, 29, 307-331.
Komlos, J., & Baten, J. (2003). The Industrial Revolution as the escape from the Malthusian trap. Munich Discussion Paper
No. 2003-13, Volkswirtschaftliche Fakultt, Ludwig-Maximilians-Universitt Mnchen
Economics, 108(3), 681716.
Review, 44, 755-777.
439.
Lee, R. D. (1997). Population dynamics: Equilibrium, disequilibrium, and consequences of fluctuations. In O. Stark & M.
Leibenstein, H. (1957). Economic backwardness and economic growth, New York, Willey.
Ltd.
McKeown, T. (1983). Food, infection, and population. Journal of Interdisciplinary History, 14, 227-247.
Journal of Lifestyle Medicine, 3(1), 19S-26S
Nelson, R. R. (1956). Theory of the low level equilibrium trap in underdeveloped countries. American Economic Review,
46(5), 894-908.
212
__________________________________________________________________________________

Nielsen, R. W. (2013b). Malthusian stagnation or Malthusian regeneration?
Nielsen, R. W. (2013c). Impacts of demographic catastrophes. http://arxiv.org/ftp/arxiv/papers/1311/1311.1850.pdf
Nielsen, R. W. (2015). Unified Growth Theory contradicted by the GDP/cap data.
Nielsen, R. W.(2016b). Unified Growth Theory contradicted by the mathematical analysis of the historical growth of human
Nielsen, R. W. (2016d). Growth of the world population in the past 12,000 years and its link to the economic growth.
Nielsen, R. W. (2016g). Mathematical analysis of historical income per capita distributions. Economic Review, 3(2), 300-
319.
Nielsen, R. W. (2016i). The Unresolved Mystery of the Great Divergence is Solved. Journal of Economic and Social
Thought, 3(2), 196-219.
Nielsen, R. W. (2016j). Scientifically unacceptable established knowledge in demography and in economic research. Journal
of Economic Library, 3(3), 429-457.
NRC (2011). Building Community Disaster Resilience Through PrivatePublic Collaboration. Washington, DC: The
National Academic Press.
316.
191.
Quarterly, 83(4), 731-757.
PRB (2010). World Population Data Sheet. Washington, DC: Population Reference Bureau.
Robine, J-M. (2001). Redefining the stages of the epidemiological transition by a study of the dispersion of life spans: The
case of France. Population, 13(1), 173-193.
Sheeran, J. (2008). The challenge of hunger. The Lancet, 371, 180-181.
Pub.
UNDP (2010). Human development report 2010. The real wealth of nations: Pathways to human development (20th
anniversary ed.) New York: United Nations Development Program.
United Nations (1973). The Determinants and Consequences of
Population Trends, Population Studies, No. 50., p.10.
213
__________________________________________________________________________________
United Nations (2013). World Population Prospects: The 2012 Revision, DVD Edition. Washigton DC: United Nations,
US Census Bureau, (2004). Global Population Profile: 2002. Washington, DC: US Census Bureau.
Vollrath, C. (2011). The Agricultural Basis of Comparative Development. Journal of Economic Growth, 16, 343370.
(http://dx.doi.org/10.1007/s10887-011-9074-1)
691-712.
Wang, C. (2005). Institutions, demographic transition, and Industrial Revolution: A unified theory. SSRN Working Paper
Series, 08/2005.
Warf, B. (2010). Demographic Transition. In B. Warf (Ed.), Encyclopedia of Geography (pp. 708-711). SAGE: Thousand
Oaks, CA.
Weisdorf, J. L. (2004). From Stagnation to Growth: Revising Three Historical Regimes. Journal of Population Economics,
17, 455-472. (http://dx.doi.org/10.1007/s00148-004-0182-5)
WFP (2011). Hunger Map 2011.
http://documents.wfp.org/stellent/groups/public/documents/communications/wfp229328.pdf
WHO (2014). The top 10 causes of death: Fact sheet No 310. http://www.who.int/mediacentre/factsheets/fs310/en/
(Publishers) Ltd.
International.
214
__________________________________________________________________________________
Appendix
Table A1. Mathematical dependence of listed indicators (y) on x = LDI (the intensity of
Malthusian positive checks) with the corresponding parameters describing the UNDP data
(UNDP, 2011). The table includes also two types of correlations between destructive and
regenerating impacts.
Indicator Formula Parameters

Standard of living
Ecological footprint (EF) y = ax 1 a = 0.646
GDP/cap y = be rx b = $71,144, r = -6.97
Population in severe poverty =y m( x n) m = 190, n = 0.358
Population below the poverty line =y m( x n) m = 195, n = 0.308
Multidimensional Poverty Index =y m( x n) m = 1.23, n = 0.293
Pop. with no access to clean water =y m( x n) m = 175, n = 0.371
Pop. with no access to san. facilities =y m( x n) m = 220, n = 0.327
Lethal impacts of positive checks
Deaths due to polluted water y = be rx b = 5, r = 9
Maternal mortality y = be rx
b = 3.58, r = 8.59
Under-five mortality y = be rx b = 2.4, r = 6.7
Adult mortality y = be rx b = 62.42, r = 2.93
Regenerating impacts
Total fertility rate y = be rx b = 1, r = 2.6
Growth rate =y m( x n) m = 4.3, n = 0
Rate of natural increase (RNI) =y m( x n) m = 3.9 , n = 0
Regenerating v lethal impacts
Total fertility rate (z) v mortality (y) =
ln z ln A + B ln y eqns (7) (12)
m
=
Growth rate or RNI (z) v mortality (y) z [ln y ln b] eqn (13)
r
215
__________________________________________________________________________________
Table A2. The dichotomy of Malthusian positive checks reflected in the contrasting levels of
hunger stress
Destruction Regeneration
Country LDI DR IMR TFR BR RNI PIF
Hunger Stress: 35%
Angola 0.514 15 102 2.7 43 2.8 2.4
Burundi 0.684 10 66 6.4 42 3.2 2.7
Chad 0.672 16 125 6.0 45 2.9 2.4
CAR 0.657 15 102 4.7 37 2.1 2.1
DRC 0.714 17 111 6.1 45 2.8 2.2
Eritrea 0.651 8 43 4.7 34 2.6 1.9
Ethiopia 0.637 10 77 5.3 37 2.7 2.0
Malawi 0.600 15 84 5.7 42 2.7 2.4
Mozambique 0.678 14 86 5.6 41 2.8 2.6
Sierra Leone 0.674 15 89 5.0 37 2.2 2.5
Zambia 0.570 15 84 6.3 46 3.1 3.3
Average Values 0.641 13.6 88.1 5.3 40.8 2.7 2.4
Hunger Stress: < 5%
Algeria 0.602 5 22 2.3 19 1.5 1.3
Egypt 0.356 5 23 2.9 25 2.0 1.5
Gabon 0.326 9 45 3.4 27 1.8 1.8
Libya 0.240 4 14 2.5 22 1.8 1.4
Morocco 0.418 6 30 2.2 19 1.3 1.2
South Africa 0.381 14 48 2.4 21 0.6 1.1
Tunisia 0.302 6 18 2.1 18 1.2 1.2
Average Values 0.375 9.8 40.0 3.6 30.2 2.0 1.9
Ratio 1.71 1.39 2.20 1.47 1.35 1.35 1.26
LDI Level of Deprivation Index; DR Death Rate; IMR Infant Mortality Rate; TFR Total Fertility Rate; BR
Birth Rate; RNI Rate of Natural Increase; PIF Population Increase Factor; CAR Central African Republic; DRC
Demographic Republic of Congo; Ratio High stress/Low stress.
Death and birth rates are per 1000 of the population. Infant Mortality Rate is defined as The annual number of deaths of
infants under age 1 per 1,000 live births (PRB, 2010). The Rate of Natural Increase is in per cent. Population Increase
Factor gives the projected population in 2050 as a multiple of the population in 2011.
216
__________________________________________________________________________________
13
Industrial Revolution did not boost economic growth

and the growth of population even in the United
Kingdom
By Ron W. NIELSEN
Abstract. Data describing economic growth and the growth of human population in the United Kingdom are
analysed. Contrary to the widely-accepted interpretations, Industrial Revolution had no impact on shaping
trajectories of economic growth and of the growth of population. Within the range of analysable data, there was
also no Malthusian stagnation. Consequently, there was no escape from Malthusian trap because there was no
trap in the economic growth and in the growth of human population. The United Kingdom was the centre of the
Industrial Revolution and yet its data are in the direct contradiction of the currently accepted interpretations. It
is fortunate that natural processes did not comply with our fanciful and wished-for explanations of the
mechanism of growth. If they did, if the generally claimed takeoffs did occur, it would have been a disaster
because economic growth and the growth of population would have been already unmanageable everywhere.
Keywords. United Kingdom, Economic growth, Population growth, Income per capita, Malthusian stagnation,
Industrial Revolution
JEL. A10, A12, C12, Y80.
1. Introduction
t is generally believed that at a certain stage of human history, economic growth and the
I growth of population experienced a dramatic transition from the endless stagnation to a

sustained growth (see for instance, Artzrouni & Komlos, 1985; Clark, 2003; Galor, 2005a,
2007; Galor & Weil, 2000; Hansen & Prescott, 2002; Klasen & Nestmann, 2006; Kgel &
Prskawetz, 2001; Komlos, 2003; Manfredi & Fanti, 2003; Weiss, 2007). This dramatic event
is described as the great escape from the Malthusian trap, as a take-off from stagnation to
growth (Galor, 2005a, 2011) or as a transition to a new stage of self-sustained growth
(Kgel & Prskawetz, 2001, p. 338). It was supposed to have been a break from Malthusian
equilibrium (Clark, 2005, p. 1314), the escape characterised by the unprecedented increase
in population growth (Galor, 2005b, p. 494), the population sprint (Thomlinson, 1965, p.
312) and the time when the population growth accelerated (Kgel & Prskawetz, 2001, p.
338). It was allegedly a massive simultaneous take-off in economic growth and population
growth (Kgel & Prskawetz, 2001, p. 338).
This alleged dramatic escape from the imagined Malthusian trap is claimed to have been
strongly prompted and assisted by the Industrial Revolution. The rapid and far-reaching
technological and sociological changes associated with this event are claimed to have been the
Australia.
. +61407201175
Published as: Nielsen, R. W. (2016). Industrial Revolution did not boost the economic growth or the growth of population
even in the United Kingdom. Journal of Economic Bibliography, 3(4), 434-446.
217
__________________________________________________________________________________
driving force of the economic and demographic transition (Bar & Leukhina, 2005; Clark, 2005;
Galor, 2005a; Galor & Mountford, 2006; Goodfriend and McDermott, 1995; Khan, 2008;
Komlos, 1989, 2003; Lucas, 2002; Manfredi & Fanti 2003; Matar, 2009; imurina & Tica,
2006; Tamura, 2002; Weiss, 2007). Clark claims that the Industrial Revolution represented a
break from the Malthusian equilibrium (Clark, 2005, p. 1314). According to Weiss, Industrial
Revolution facilitated an endogenous take-off from the Malthusian trap (Weiss, 2007, p.
327). Likewise, Komlos claimed that The industrial revolution can therefore be
conceptualized as a break out of the Malthusian demographic regime. It was a period of both
economic and demographic expansion (Komlos, 1989, p. 203). He wrote that Industrial
Revolution was also accompanied by an acceleration in population growth (Komlos, 2003, p.
18). The Industrial Revolution drove the demographic transition (Khan, 2008, p. 9). It
brought in its wake an accelerated growth in the size of human populations (Matar, 2009,
p. 381). According to Galor, In the first phase of the Industrial Revolution, prior to the
implementation of significant education reforms, physical capital accumulation was the prime
engine of economic growth (Galor, 2005a, p. 212).
All such claims, descriptions and explanations of the past growth are not based on a
scientific analysis of relevant data but on a good dose of fantasy. They may sound plausible
but they have to be accepted by faith. Science has no room for such dubious speculations.
Inevitably, when faith is defended, contradicting data are either ignored or suitably manipulated
to support the preconceived ideas (Ashraf, 2009; Galor, 2005a, 2005c, 2007, 2008a, 2008b,
2008c, 2010, 2011, 2012a, 2012b, 2012c; Galor and Moav, 2002; Snowdon & Galor, 2008).
Remarkably, however, precisely the same data, when closely analysed, demonstrate that the
preconceived ideas are incorrect (Nielsen, 2014, 2015a, 2016a, 2016b, 2016c, 2016d, 2016e,
2016f, 2016g). The established knowledge in demography and economic research, the
knowledge revolving around the concept of Malthusian stagnation and around the concept of
the escape from the alleged Malthusian trap is not based on the scientific process of
investigation (Nielsen, 2016h).
In science, even one contradicting evidence in data is sufficient to question contradicted
interpretations and then to try to revise them or even reject them, but in the case of the historical
economic growth and of the growth of population we now have more than one contradicting
evidence (e.g. Biraben, 1980; Clark, 1968; Cook, 1960; Durand, 1967, 1974, 1977; Gallant,
1990; Haub, 1995; Kapitza, 2006; Kremer, 1993; Livi-Bacci, 2007; Maddison, 2001, 2010;
McEvedy & Jones, 1978; Nielsen, 2014, 2015a, 2016a, 2016b, 2016c, 2016d, 2016e, 2016f,
2016g, 2016i, 2016j; Podlazov, 2002; Shklovskii, 1962, 2002; Taeuber & Taeuber, 1949;
Thomlinson, 1975; Trager, 1994; von Foerster, Mora & Amiot, 1960; von Hoerner, 1975).
We have carried out extensive investigation of the leading postulates used to explain the
historical growth of human population and the historical economic growth (Nielsen, 2014,
2015a, 2016a, 2016b, 2016c, 2016d, 2016e, 2016f, 2016g, 2016i, 2016j). In particular, we have
demonstrated that within the range of analysable data, economic growth and the growth of
population were hyperbolic. There was no Malthusian stagnation and there was never a takeoff
from stagnation to growth, which could be described as the escape from the Malthusian trap
because there was no trap in the economic growth and in the growth of population.
The range of analysable data describing economic growth, global, regional and national,
extends down to AD 1 (Maddison, 2001, 2010) but for the world economic growth it was
extended down to 1,000,000 BC (De Long, 1998). Maddisons data and the extended estimates
show clearly that there was never stagnation in the economic growth.
For the growth of human population, regional and national estimates also extend only down
to AD 1 (Maddison, 2001, 2010) but for the global growth, many estimates are available
extending down to 10,000 BC (Biraben, 1980; Clark,1968; Cook,1960; Durand, 1974; Gallant,
1990; Haub, 1995; Livi-Bacci, 2007; McEvedy & Jones, 1978; Taeuber & Taeuber, 1949;
218
__________________________________________________________________________________
Thomlinson, 1975; Trager, 1994, United Nations, 1973, 1999, 2013). They were also extended
down to 1,000,000 million years ago (Deevey, 1960; cited by Kapitza, 2006, Kremer, 1993 and
Livi-Bacci, 2007). From the distance of one million years it does not really matter whether it
is a million years ago or million years BC. The evidence again is clear and consistent: there
was no Malthusian stagnation but a steadily-increasing growth, interrupted only twice in the
past million years, or maybe three times if we count the minor disturbance between AD 1200
and 1400 (Nielsen, 2016j). Each time, population growth was diverted from one hyperbolic
growth to another. Our analysis, which is in harmony with earlier research (Kapitza, 2006;
Kremer, 1993; Podlazov, 2002; Shklovskii, 1962, 2002; von Foerster, Mora & Amiot, 1960;
von Hoerner, 1975), shows that fundamental postulates accepted by the established knowledge
in demography and in economic research are repeatedly and clearly contradicted by data.
We have also demonstrated that the Industrial Revolution, 1760-1840 (Floud & McCloskey,
1994) had no impact on shaping the trajectories of economic growth and of the growth of
population. Now we are going to demonstrate that Industrial Revolution had absolutely no
impact on shaping the growth of population and the economic growth in the United Kingdom,
the very centre of this revolution where its effects should be most convincingly demonstrated.
In our diagrams, population data will be expressed in billions while the data for the Gross
Domestic Product (GDP) in billions of 1990 International Geary-Khamis dollars. The GDP per
capita (GDP/cap) values will be expressed in 1990 International Geary-Khamis dollars.
2. Analysis of the growth of population in the UK

Hyperbolic growth can be uniquely identified by studying the reciprocal values of data
(Nielsen, 2014) because hyperbolic growth is then represented by a decreasing straight line.
For a sufficiently large range of data, if they follow a decreasing straight line, the growth is
hyperbolic. In such displays, it is also easy to identify even small deviations from hyperbolic
distributions because deviations from a straight line are easy to notice. In particular, any
boosting in the economic growth or in the growth of population, such as the expected boosting
caused by the Industrial Revolution should be readily identified.
For the reciprocal values, effects are reversed. A boosting of growth is indicted by a clear
change of the trend in the downward direction while a diversion to a slower trajectory is
indicated by an upward bending in the growth trajectory. Results of our analysis of population
data (Maddison, 2010) in the United Kingdom are presented in Figures 1-3.
Figure 1. Hyperbolic growth of population in the UK between AD 1 and 1850 as demonstrated by the
decreasing straight line fitting the reciprocal values of the population data. Industrial Revolution did
not boost the growth of the population in the UK. On the contrary, it coincided with the
commencement of the gradually slowing down growth.
219
__________________________________________________________________________________
Figure 2. The end part of the plot presented in Figure 1 showing that from around 1850, just at the
end of the Industrial Revolution, the growth of the population in the UK started to be diverted to a
slower trajectory. Industrial Revolution did not boost the growth of population in the UK.
Figure 3. Growth of population in the UK between AD 1 and 2008. Growth was hyperbolic between
AD 1 and 1850. From around 1850, towards the end of the Industrial Revolution, the growth of
population started to be diverted to a slower trajectory. Industrial Revolution had no impact on
shaping the growth trajectory.
In Figure 1 we show the reciprocal values of population data. Between AD 1 and 1850 they
follow closely a decreasing straight line, showing that the growth of population was hyperbolic.
Within the range of analysable data, which extends down to AD 1, the mythical epoch of
Malthusian stagnation did not exist in the UK. The proof of the existence of Malthusian
stagnation would have to be based on the demonstration of the existence of Malthusian
oscillations. The data displayed in Figure 1 follow closely an undisturbed linear distribution
representing an undisturbed and stable hyperbolic trajectory, indicating that even if random
220
__________________________________________________________________________________
Malthusian forces were present they had no effect on changing the growth trajectory. Any
assumption of the presence of such forces is irrelevant.
It is also clear that the Industrial Revolution, 1760-1840 (Floud & McCloskey, 1994) had
absolutely no impact on changing the growth trajectory. Data displayed in Figures 1-3 show
clearly that there was no often-claimed boosting in the growth of population, no sprinting,
explosion or any form of strong acceleration. On the contrary, from around 1850, shortly after
the Industrial Revolution, the growth of population started to be diverted to a slower trajectory
as indicated by the upper bending of the trajectory of reciprocal values shown clearly in Figure
2 and by a clear deviation from the hyperbolic trajectory shown in Figure 3.
These are remarkable results because the UK was in the centre of the Industrial Revolution.
It is here that the effects of this revolution should be most strongly and most convincingly
demonstrated but the data are in the direct contradiction of such expectations. It seems obvious
that Industrial Revolution brought about many changes in the style of living and in social
interactions, beneficial or detrimental, but all these changes had no effect on the growth of
human population. It is as if this monumental event never happened.
Hyperbolic growth is described by a simple formula:
1
S (t ) = (1)
a kt
where S (t ) is the size of the growing entity (in our case either population or the GDP), while
a and k are positive constants.
The increasing hyperbolic distribution, which could be also called the first-order hyperbolic
distributions, is just the reciprocal of a decreasing straight line. That is why, a decreasing
straight line of the reciprocal values identifies uniquely the first-order hyperbolic distribution.
Parameters of the hyperbolic distribution shown in Figure 3 are: = a 1.210 103 and
= k 6.366 101 . Its singularity is at t = 1901 . If continued along its historical trajectory, the
growth of the population in the UK would have escaped to infinity in 1901. Fortunately, from
around 1850 is started to be diverted to a slower trajectory bypassing the singularity by a safe
margin of 51 years.
3. Analysis of the economic growth in the UK

Results of mathematical analysis of the historical GDP data for the UK are presented in
Figures 4-6, while for the historical income per capita (GDP/cap) they are shown in Figure 7.
Reciprocal values displayed in Figure 4 show that the growth of the GDP was at first
increasing along a fast, hyperbolic trajectory, as shown by a steep straight line fitting the
reciprocal values of data. However, from around AD 1600, i.e. about 160 years before the
commencement of the Industrial Revolution, the growth of the GDP was diverted to a slower
hyperbolic trajectory as indicated by a less-steep straight line. This slower trajectory remained
totally unaffected by the Industrial Revolution. This event did not even manage to revert the
economic growth to the state experienced before AD 1600, when the hyperbolic trajectory was
significantly faster. This slower hyperbolic growth continued until around AD 1850 when it
started to be diverted to even slower trajectory indicated by an upward bending shown in Figure
5.
There was definitely no boosting in the economic growth caused by or associated with the
Industrial Revolution. There was even no visible delay in the diversion to a slower trajectory.
Industrial Revolution had no effect on the economic growth trajectory.
Again, these results are remarkable because the UK was right at the centre of the Industrial
Revolution and it should have experienced its strong effect on the economic growth and on the
growth of population. Technological and sociological changes brought about by the Industrial
Revolution were present but, surprisingly perhaps, they did not accelerate the economic
221
__________________________________________________________________________________
growth. It was as if economic growth were prompted by some other, much stronger force,
which overruled any possible impacts of the Industrial Revolution. It would be interesting to
identify this force but it is clear that the usual explanations based on the hypothetical forces of
Malthusian stagnation and on the equally hypothetical forces of the Industrial Revolution,
including the forces of technological development, are irrelevant for explaining the mechanism
of the historical economic growth.
Figure 4. Reciprocal values of the GDP data for the UK (Maddison, 2010) are compared with the
decreasing linear distributions representing hyperbolic growth. The growth of the GDP was following
a fast-increasing hyperbolic distribution (represented by a fast-decreasing straight line) until AD
1600. From around that year and until around AD 1850, the economic growth was following a slower
hyperbolic trajectory. Within the range of analysable data, i.e. from AD 1, the mythical epoch of
Malthusian stagnation did not exist.
Figure 5. The end part of the display shown in Figure 4. The slower hyperbolic growth, which
commenced around AD 1600 (as indicated by the gently-decreasing straight line), continued
undisturbed until AD 1850, i.e. throughout the entire time of the Industrial Revolution, which had
absolutely no impact on the economic growth trajectory. There was no escape from the Malthusian
trap because there was no trap. From around AD 1850, the growth of the GDP started to be diverted
to a slower trajectory, as indicated by the upward bending of the reciprocal values trajectory.
222
__________________________________________________________________________________
Figure 6. Economic growth (as described by the GDP) in the UK. The growth was hyperbolic
between AD 1 and 1600 and again (but a little slower) between AD 1600 and 1850. From around
1850, the growth started to be diverted to a slower but non-hyperbolic trajectory. Within the range of
analysable data, i.e. from AD 1, the mythical epoch of stagnation did not exist. Economic growth was
steadily increasing. Industrial Revolution did not boost the economic growth. There was no escape
from the Malthusian trap because there was no trap.
Figure 7. Growth of income per capita (GDP/cap) in the UK between AD 1 and 2008. The GDP data
follow closely the empirically-determined linearly-modulated hyperbolic distributions (defined in
Nielsen, 2015a). Industrial Revolution did not change the growth trajectory. From around 1850, the
growth of the GDP/cap started to be diverted to a slower trajectory.
Hyperbolic fits to the GDP data are shown in Figure 6. The fast hyperbolic growth between
AD 1 and 1600 is described by = a 3.120 100 and = k 1.849 103 Its singularity is at
t = 1687 . Contrary to the doctrine of Malthusian stagnation, economic growth was remarkably
223
__________________________________________________________________________________
fast. If continued, it would escape to infinity about 73 years before the commencement of the
Industrial Revolution. The slower hyperbolic growth of the GDP, which commenced in around
AD 1600 is described by = a 1.106 100 and
= k 5.909 104 . Its singularity is at t = 1872 .
This was also a steadily-increasing economic growth at the time when it was supposed to have
been stagnant.
Population and economic growth data for the UK, and in particular the relatively fast
economic growth before AD 1600, show how absurd is the concept of Malthusian stagnation.
This concept is consistently contradicted by the analysis of other data describing economic
growth and the growth of population (Kapitza, 2006; Kremer, 1993; Nielsen, 2014, 2015a,
2016a, 2016b, 2016c, 2016d, 2016e, 2016f, 2016g, 2016i, 2016j; Podlazov, 2002; Shklovskii,
1962, 2002; von Foerster, Mora & Amiot, 1960; von Hoerner, 1975).
The data for the UK show also how absurd is the doctrine of the boosting effects of the
Industrial Revolution. The GDP and GDP/cap were already following fast-increasing
trajectories before the Industrial Revolution. If continued, these trajectories would increase to
infinity in AD 1872. Any boosting by the Industrial Revolution would have been disastrous.
Fortunately, natural processes did not comply with this ludicrous concept. Economic growth
in the UK was not boosted by the Industrial Revolution but it was soon diverted into a slower
pathway.
The same argument applies to the global and regional economic growth and to the global
and regional growth of population. Propelled by the historical hyperbolic growth, they are now
increasing too fast. Any boosting by the Industrial Revolution, any differential timing of the
alleged takeoffs claimed by Galor (2005a, 2011), would be disastrous because it would propel
economic growth and the growth of population along even faster trajectories and would render
them unmanageable. Even now, we are approaching a serious global crisis but with the
mechanism of growth approved by the established knowledge, this crisis would have occurred
much earlier.
Nature or naturally occurring process take no notice of what we think is logical. Imagination
is important in science but imagination has to be checked by meticulous analysis of data. We
can propose convincing explanations but what we think as convincing is not necessarily what
is reflected in the real world. Scientific research has to be conducted scientifically; otherwise
it is not scientific.
We can write as many fiction stories as we can possibly imagine them. They can be
interesting and attractive but they have no place in science. Any theory that cannot be checked
by data is regarded as unscientific and any theory that is contradicted by data has to be modified
or even rejected and replaced by a new theory. Deliberately distorting the presentation of data
(Ashraf, 2009; Galor, 2005a, 2005c, 2007, 2008a, 2008b, 2008c, 2010, 2011, 2012a, 2012b,
2012c; Galor and Moav, 2002; Snowdon & Galor, 2008) to make them comply with
preconceived ideas is not only unscientific but also self-defeating we learn nothing from such
mutilations of scientific evidence.
Doctrines of Malthusian stagnation and of the dramatic impacts of the Industrial Revolution
on the growth of population and on the economic growth are repeatedly and consistently
contradicted. These two doctrines and all the associated explanations and elaborate descriptions
have no place in the economic and demographic research and the sooner they are abandoned
the better. The continuing use of these doctrines to explain the historical economic growth and
the historical growth of population is scientifically unjustified.
Defined by the ratio of the parameter k, hyperbolic growth between AD 1 and 1600 was
about three times faster than the hyperbolic growth between AD 1600 and 1850. The mythical
epoch of stagnation did not exist. The transition around AD 1600 was not the usually-imagined
transition from stagnation to growth but from growth to growth. It was not boosting but a
transition from a fast to a slower hyperbolic growth. There is absolutely no expected correlation
224
__________________________________________________________________________________
between the economic growth in the UK and the Industrial Revolution. No expected boosting
and no transition from stagnation to growth because there was no stagnation. The wished-for
takeoff is replaced by a transition to a slower trajectory. The established knowledge in the
economic research is spectacularly contradicted by data, which were expected to give the most
convincing support for the generally accepted doctrines.
The data refuse to comply with the desired and wished-for interpretations of the mechanism
of economic growth. There was no wished-for escape from the Malthusian trap because there
was no trap in the economic growth. There was also no trap in the growth of population in the
UK. The only way to defend the established knowledge is to reject the data for the UK but then
we would have to reject also other data and their analysed (Biraben, 1980; Clark, 1968; Cook,
1960; Durand, 1967, 1974, 1977; Gallant, 1990; Haub, 1995; Kapitza, 2006; Kremer, 1993;
Livi-Bacci, 2007; Maddison, 2001, 2010; McEvedy & Jones, 1978; Nielsen, 2014, 2015a,
2016a, 2016b, 2016c, 2016d, 2016e, 2016f, 2016g, 2016i, 2016j; Podlazov, 2002; Shklovskii,
1962, 2002; Taeuber & Taeuber, 1949; Thomlinson, 1975; Trager, 1994; von Foerster, Mora
& Amiot, 1960; von Hoerner, 1975)
Results of mathematical analysis presented Figure 7 show that the growth of income per
capita in the UK can described by two linearly-modulated hyperbolic distributions. The
trajectory was calculated by dividing two hyperbolic distributions fitting the GDP data between
AD 1 and 1850 (see Figures 4-6) by the hyperbolic distribution fitting the population data
between AD 1 and 1850 (see Figures 1-3). For the discussion of the linearly-modulated
hyperbolic distributions see Nielsen (2015a).
The growth of income per capita follows closely the empirically-determined growth
trajectory. Industrial Revolution had no impact on changing the linearly-modulated hyperbolic
growth. From around 1850, shortly after this industrial event, the growth of income per capita
started to be diverted to a slower trajectory.

The United Kingdom was in the centre of the Industrial Revolution. It is, therefore, the
perfect place to test the currently accepted concept that the Industrial Revolution boosted
economic growth and the growth of population. This concept is closely linked with the concept
of Malthusian stagnation and the concept of the escape from the Malthusian trap. All these
props are used to explain the mechanism of the economic growth and of the growth of human
population. We have already demonstrated that all these accepted interpretations are
contradicted by data (Nielsen, 2014, 2015a, 2016a, 2016b, 2016c, 2016d, 2016e, 2016f, 2016g,
2016i, 2016j). Now, we have focused our attention of the centre of the Industrial Revolution.
We have analysed the data (Maddison, 2010) describing the growth of population, the
growth of the GDP and the growth of the GDP/cap in the UK. We have demonstrated that the
historical growth of population and of the GDP were hyperbolic. Consequently, the historical
growth of income per capita (GDP/cap) was linearly-modulated hyperbolic (Nielsen, 2015a).
We have demonstrated that over the entire range of the mathematically-analysable data,
which in this case extends down to AD 1, the epoch of Malthusian stagnation did not exist. The
growth of the population and the economic growth were increasing steadily without any signs
of Malthusian stagnation.
We have demonstrated that the Industrial Revolution had absolutely no impact on shaping
the growth of population and the economic growth in the UK, the very centre of this revolution
where its effects should have been most clearly demonstrated. Thus, we have demonstrated yet
again that the often-claimed effects of the Industrial Revolution on shaping the growth of
population or on shaping the economic growth are contradicted by data.
The established knowledge in demography and in economic research is scientifically
unacceptable (Nielsen, 2016h). It is contradicted by data and it flies in the face of everything
225
__________________________________________________________________________________
we know about the current economic and demographic problems, which need to be urgently
solved.
There was no transition from stagnation to a sustained growth regime (Galor, 2005a, 2011).
The past growth was stable and secure as demonstrated by the largely undisturbed hyperbolic
distributions but now, even though it became diverted from the fast increasing hyperbolic
distributions to slower trajectories, it is still too fast and consequently insecure (Nielsen,
2015b). We might still have a sustained economic growth and sustained growth of population
but it is generally acknowledged that in the long run our sustained growth is unsustainable
because for the first time in human history we have already reached and crossed the ecological
capacity of our planet (WWF, 2010).
The currently accepted paradigm based on the concept of Malthusian stagnation, on the
concept of the escape from the Malthusian trap and on the concept of the boosting effects of
innovations and technological development as represented by the Industrial Revolution, by the
progress in medicine and by the dramatic changes in the style of living, is not only scientifically
untenable but it is also potentially dangerous because it propagates the idea that after the
endless epoch of stagnation we have now entered at last the sustained growth regime (Galor,
2005a, 2011). The real world in different. We have not escaped a Malthusian trap because there
was no trap in the growth of population and in the economic growth. However, after the ages-
long stable and secure growth, our current growth is no longer sustainable. For the first time in
human history we have found ourselves in the trap of the fast-increasing economic growth and
in the fast-increasing growth of population.
The erroneous traditional interpretations of economic growth and of the growth of human
population are well illustrated in the Unified Growth Theory (Galor, 2005a, 2011) based firmly
on these incorrect concepts. In conformity with the traditional interpretations, Galor divided
economic growth and the growth of population into three regimes: the Malthusian Epoch, the
Post-Malthusian Regime and the Sustained Growth Regime. Economic growth and the growth
of population is then explained using various complicated mechanisms, different for each of
the imagined regimes. These erroneous concepts are supported by suitably distorted
presentations of data (Ashraf, 2009; Galor, 2005a, 2005c, 2007, 2008a, 2008b, 2008c, 2010,
2011, 2012a, 2012b, 2012c; Galor and Moav, 2002; Snowdon & Galor, 2008).
Data are never analysed. In their distorted presentations, they appear to support the
erroneous concepts based on impressions and on a good dose of fantasy. However, when
analysed, precisely the same data show that the traditionally accepted doctrines have no support
in science. Furthermore, they suggest that the mechanism of the historical economic growth
and of the growth of population must be simple because they are described by the exceptionally
simple mathematical distributions.
While the paradigm based on the concept of the endless epoch of stagnation followed by a
sustained growth regime, creates a sense of security and prosperity, the data show that the
opposite is true. It was the past growth that was safe and secure because it is described by the
generally steadily increasing trajectories. However, now, for the first time in human history,
our economic growth and the growth of human population is uncertain and insecure. We might
reach a certain maximum in the growth of human population during the current century but we
might not (Nielsen, 2006, 2015b, 2016h). The future is far from certain.
Interpretations of the mechanism of the historical economic growth and of the historical
growth of population have to be based on data, and data are in the direct contradiction of the
currently accepted paradigm (Nielsen, 2014, 2015a, 2016a, 2016b, 2016c, 2016d, 2016e,
2016f, 2016g, 2016h, 2016i, 2016j). These interpretations have to be based on accepting
hyperbolic growth. There is no choice: the traditional paradigm based on the concept of
Malthusian stagnation followed by a distinctly new regime of sustained growth has to be
226
__________________________________________________________________________________
replaced by the evidence presented by data that the past growth was hyperbolic but that,
relatively recently, it was diverted to new trajectories.
Hyperbolic distributions may be confusing. They may create an illusion of stagnation
followed by an explosion but this illusion is not a valid excuse for creating the whole system
of scientifically unsupported doctrines and interpretations because the analysis of hyperbolic
distributions is trivially simple (Nielsen, 2014). Anyone can do it and see that the currently
accepted paradigm based on the assumption of the existence of Malthusian stagnation followed
by the alleged escape from the Malthusian trap has no scientific support.
References
Artzrouni, M. & Komlos, J. (1985). Population Growth through History and the Escape from the Malthusian Trap: A
Homeostatic Simulation Model. Genus, 41, 21-39.
Bar, M., & Leukhina, O. (2005). A model of historical evolution of output and population. Federal Reserve Bank of
Minneapolis.
Clark, G. (2003). The Great Escape: The Industrial Revolution in Theory and in History. Working Paper. Davis: University
of California.
Clark, G. (2005). The condition of the working class in England, 1209-2004. Journal of Political Economy, 113(6) ,1307-
1340.
De Long, J. B. (1998). Estimates of World GDP One Million B.C Present.
http://delong.typepad.com/print/20061012_LRWGDP.pdf
Social Science: World Population, 369, 2.
Durand, J. (1977). Historical Estimates of World Population: An Evaluation," Population and Development Review, 3(3),
256.
Press.
Publishing Company.
Economic Association, 3(2-3), 494-504.
Galor, O. (2005c). The Demographic Transition and the Emergence of Sustained Economic Growth. Journal of the
European Economic Association, 3, 494-504. http://dx.doi.org/10.1162/jeea.2005.3.2-3.494
475.
Dictionary of Economics, Palgrave Macmillan, New York. http://dx.doi.org/10.1057/9780230226203.0434
Growth Theory. International Economic Review, 51, 1-44. http://dx.doi.org/10.1111/j.1468-2354.2009.00569.x
Galor, O. (2012b). The Demographic Transition: Causes and Consequences. Cliometrica, 6, 1-28.
http://dx.doi.org/10.1007/s11698-011-0062-7
Galor, O. and Moav, O. (2002). Natural Selection and the Origin of Economic Growth. The Quarterly Journal of
Economics, 117, 1133-1191. http://dx.doi.org/10.1162/003355302320935007
Galor, O., & Mountford, A. (2006). Trade and the Great Divergence: The family connection. The American Economic
Review, 96(2): 299 303.
227
__________________________________________________________________________________
Galor, O. & Weil, D. N. (2000). Population, technology, and growth: From Malthusian stagnation to the demographic
transition and beyond. The American Economic Review, 90(4), 806-828.
Goodfriend, M. & McDermott, J. (1995). Early development. American Economic Review, 85, 116133.
Khan, A. (2008). The Industrial Revolution and the demographic transition. Business Review, First Quarter, 9-15.
Klasen, S. and Nestmann, T. 2006, Population, population density and technological change, Journal of Population
Economics, 19(3): 611-626.
Kgel, T. & Prskawetz, A. (2001). Agricultural Productivity Growth and Escape from the Malthusian Trap. Journal of
Economic Growth, 6(4), 337-357.
Komlos, J. (2003). The Industrial Revolution as the escape from the Malthusian trap. Munich Discussion Paper No. 2003-
13. Volkswirtschaftliche Fakultt, Ludwig-Maximilians-Universitt Mnchen.
Economics, 108(3), 681716.
Lucas, R. E. (2002). The Industrial Revolution: Past and Future. Cambridge, MA: Harvard University Press.
Manfredi, P. & Fanti, L. (2003). The Demographic Transition and Neoclassical Models of Balanced Growth. In N. Salvadori
(Ed.), The Theory of Growth: A Classical Perspective (pp. 161-185). Cheltenham: Edward Elgar.
Matar, H. F. (2009). The ephemeral existence of humanity. Mankind Quarterly, 49(3/4), 381-392.
Nielsen, R. W. (2014). Changing the Paradigm. Applied Mathematics, 5, 1950-1963.
http://dx.doi.org/10.4236/am.2014.513188
242-255.
Nielsen, R. W. (2016b). Mathematical analysis of historical income per capita distributions. Economic Review, 3(2), 300-
319.
Nielsen, R. W. (2016c). The unresolved mystery of the great divergence is solved. Journal of Economic and Social Thought,
3(2), 196-219.
Economic Review, 3(1), 16-27. http://www.kspjournals.org/index.php/TER/article/view/650
Nielsen, R. W. (2016e). Puzzling properties of the historical growth rate of income per capita explained. Journal of
and Social Thought, 3(1), 1-34. http://www.kspjournals.org/index.php/JEST/article/view/669
Nielsen, R. W. (2016g). Unified Growth Theory contradicted by the mathematical analysis of the historical growth of human
Nielsen, R. W. (2016h). Scientifically unacceptable established knowledge in demography and in economic research.
Journal of Economic Library, 3(3), 429-457.
Nielsen, R. W. (2016i). Demographic Transition Theory and its link to the historical economic growth. Journal of Economic
Nielsen, R. W. (2016j). Growth of the world population in the past 12,000 years and its link to the economic growth.
imurina, J. and Tica, J. 2006, Historical perspective of the role of technology in economic development, EFZG Working
Papers Series, No 610, Faculty of Economics and Business, University of Zagreb.
Tamura, R. F. (2002). Human capital and the switch from agriculture to industry. Journal of Economic Dynamics and
Control, 27, 207-242.
Random House.
Pub.
228
__________________________________________________________________________________

691-712.
Weiss, V. (2007). The population cycle drives human history from a eugenic phase into a dysgenic phase and eventual
collapse. The Journal of Social, Political, and Economic Studies, 32(3): 327-358.
International.
229
__________________________________________________________________________________
14
Mathematical analysis of income per capita in the

United Kingdom
By Ron W. NIELSEN
Abstract. Industrial Revolution did not boost economic growth and the growth of population even in the United
Kingdom, the centre of this revolution. On the contrary, from around 1850, economic growth and the growth of
population started to be diverted to slower trajectories. The new trend describing the growth of income per capita
(Gross Domestic Product per capita, i.e. GDP/cap) is now analysed using Maddisons data. Two numerical
solutions and two analytical solutions of the differential equation describing the growth rate of income per capita
are presented. It is demonstrated yet again that even strong fluctuations in the growth rate do not change the
shape of growth trajectories. Thus, contrary to the common misconception, even strong fluctuations in the growth
rate cannot be used as the evidence of the existence of Malthusian stagnation because they do not change the
mechanism of growth. Such strong fluctuations can at best produce only small and negligible ripples in growth
trajectories. Our analysis shows also that the current growth of income per capita in the UK follows an
unsustainable trajectory.
Keywords. United Kingdom, Income per capita, Growth rate, Future growth
JEL. A10, B41, C02, C12, C20, C50, O10.
Australia.
. +61407201175
1. Introduction
ccording to the generally accepted interpretations, the alleged long epoch of the so-
A called Malthusian stagnation in the economic growth and in the growth of human
population was followed by a rapid increase, which is claimed to have been caused by
modern progress reflected in and coinciding with the Industrial Revolution. The United
Kingdom was at the centre of the Industrial Revolution and consequently, this postulated
transition from stagnation to growth should be most clearly demonstrated in this country.
We have already shown that the currently accepted interpretations are incorrect (Nielsen,
2014, 2015a, 2016a, 2016b, 2016c, 2016d, 2016e, 2016f, 2016g, 2016h, 2016i, 2016j, 2016k).
Within the range of analysable data, epoch of Malthusian stagnation did not exist in the
economic growth and in the growth of population, global or regional. Likewise, the Industrial
Revolution, 1760-1840 (Floud & McCloskey, 1994) did not boost the growth trajectories.
In particular, we have demonstrated (Nielsen, 2016k) that even in the United Kingdom,
where the effects of the Industrial Revolution should be most clearly demonstrated, there was
no boosting in the economic growth and in the growth of population. On the contrary, shortly
Australia.
. +61407201175
Published as: Nielsen, R. W. (2016). Industrial Revolution did not boost the economic growth or the growth of population
even in the United Kingdom. Journal of Economic Bibliography, 3(4), 434-446.
230
__________________________________________________________________________________
after the Industrial Revolution, economic growth and the growth of population started to be
diverted to slower trajectories. We have also demonstrated that within the range of the
mathematically analysable data, the mythical epoch of Malthusian stagnation did not exist in
the United Kingdom.
Our analysis and conclusions are supported by data and by earlier investigations (Biraben,
1980; Clark,1968; Cook,1960; Durand, 1974; Gallant, 1990; Haub, 1995; Kapitza, 2006;
Kremer, 1993; Lehmeyer, 2004; Livi-Bacci, 1997; Maddison, 2001, 2010; Mauritius, 2015;
McEvedy & Jones, 1978; Podlazov, 2002; Shklovskii, 1962, 2002; Statistics Mauritius, 2014;
Statistics Sweden, 1999; Taeuber & Taeuber, 1949; Thomlinson, 1975; Trager, 1994, United
Nations, 1973, 1999, 2013; von Hoerner, 1975, von Foerster, Mora & Amiot, 1960; Wrigley
& Schofield, 1981). The only way to accept the doctrine of Malthusian stagnation and the
concept of the boosting effects of the Industrial Revolution is to ignore data or to manipulate
them in such a way as to make them appear to support the erroneous ideas (Ashraf, 2009; Galor,
2005a, 2005c, 2007, 2008a, 2008b, 2008c, 2010, 2011, 2012a, 2012b, 2012c; Galor and Moav,
2002; Snowdon & Galor, 2008).
Our aim now is to investigate the new trend of income per capita in the UK, the trend which
commenced relatively recently when the economic growth started to be diverted from its
historical, linearly modulated hyperbolic trajectory. (For its definition see Nielsen, 2015a.) We
shall first show how to describe growth by using the differential equation defining the growth
rate. This part of the analysis will show that contrary to the generally expected outcomes,
fluctuations and long-term variations in the growth rate have negligible effect on the growth
trajectories. We shall then study the future growth of income per capita.
2. Overview
Distributions describing the growth of population and economic growth in the UK based on
using Maddisons data (Maddison, 2010) are shown in Figure 1-3.
Figure 1. Growth of population in the UK between AD 1 and 2008. Growth was hyperbolic between
AD 1 and 1850. From around 1850, towards the end of the Industrial Revolution, the growth of
population started to be diverted to a slower trajectory. Industrial Revolution had no impact on
shaping the growth trajectory.
231
__________________________________________________________________________________
Figure 2. Economic growth (as described by the GDP) in the UK. The growth was hyperbolic
between AD 1 and 1600 and again (but a little slower) between AD 1600 and 1850. From around
1850, the growth started to be diverted to a slower trajectory. Within the range of analysable data,
i.e. from AD 1, the mythical epoch of stagnation did not exist. Economic growth was steadily
increasing. Industrial Revolution did not boost the economic growth. There was no escape from the
Malthusian trap because there was no trap.
Figure 3. Growth of income per capita (GDP/cap) in the UK between AD 1 and 2008. The GDP data
follow closely the empirically-determined linearly-modulated hyperbolic distributions (defined in
Nielsen, 2015a). Industrial Revolution did not change the growth trajectory. From around 1850, the
growth of the GDP/cap started to be diverted to a slower trajectory.
232
__________________________________________________________________________________
Population and the Gross Domestic Product (GDP) were increasing hyperbolically.
Contrary to the currently accepted interpretations, there was no Malthusian stagnation and the
Industrial Revolution had no impact on the economic growth and on the growth of population
even in the United Kingdom, the very centre of this revolution.
We should notice that at the time of the Industrial Revolution, economic growth and the
growth of population in the United Kingdom were close to escaping to infinity. It was most
fortunate that natural processes did not comply with the imagined interpretations of the growth
mechanism. Any boosting by the Industrial Revolution would have been catastrophic.
While the growth of population was following a single hyperbolic trajectory, the growth of
the GDP experienced a transition around AD 1600 from a fast to a slower hyperbolic trajectory.
This transition is reflected in the income per capita (GDP/cap) shown in Figure 3. Industrial
Revolution did not boost the growth of population or the economic growth. From around 1850,
economic growth and the growth of population started to be diverted to a slower, non-
hyperbolic, trajectory. This simultaneous transition in the growth of population and in the
growth of the GDP is reflected in a clear transition in the income per capita (GDP/cap). It is
the purpose of this publication to investigate this new trajectory.
3. Mathematical method
Our analysis is based on the examination of the growth rate (Nielsen, 2015b):
1 dS (t )
= Re (t ) (1)
S (t ) dt
where S (t ) is the size of the growing entity and Re (t ) is the empirically-determined growth rate.
In the case discussed here, the size of the growing entity is the GDP/cap.
There are two ways of solving this equation: numerical or analytical. If the empirically-
determined growth rate Re (t ) can be described analytically, then
S (t ) = exp f (t )dt (2)

where f (t ) is the analytical representation of Re (t ) .
If Re (t ) cannot be represented by a simple mathematical function, as in the case of randomly-
fluctuating growth rate, then the eqn (1) has to be solved numerically. We also have to solve
the eqn (2) numerically if the integration of the function f (t ) leads to computational problems,
such as when S (t ) has to be expressed by an infinite series. We shall now use both of these
methods, analytical and numerical to describe the growth of income per capita in the UK and
to predict growth.
4. Mathematical analysis
Four representations of the growth rate of income per capita (GDP/cap) in the UK between
AD 1830 and 2008 are shown in Figure 4. They are (1) R(Direct) calculated directly from the
GDP/cap data; (2) R(Refined) calculated using the GDP/cap data and interpolated gradients;
(3) calculated using the best polynomial fit to R(Refined) represented in this case by a sixth-
order polynomial; and (4) calculated by using linear fit to R(Direct). Virtually the same linear
distribution was obtained by fitting R(Refined).
233
__________________________________________________________________________________
Figure 4. Four representations of the growth rate of income per capita (GDP/cap) in the United
Kingdom between AD 1830 and 2008. R(Direct) is the growth rate calculated directly from the
GDP/cap data. R(Refined) is the growth rate calculated using the GDP/cap data and interpolated
gradients. f (t ) - Polynomial is the sixth-order polynomial fitted to R(Refined) and f (t ) - Linear is
the linear fit to R(Direct).
We shall now use all these four representations of the growth rate to describe the GDP/cap
distribution. We shall present two numerical solutions of the eqn (1) by using Re (t ) = R( Direct )
and Re (t ) = R ( Refined ) . We shall also present two analytical solutions f (t ) represented by a
six-order polynomial or by the linear function. Finally, we shall use the linear representation
of the growth rate to predict growth of income per capita.
4.1. Describing the growth trajectory
Two numerical solutions of the eqn (1) are presented in Figure 5. They are so close to the
data that in order to see the difference between them we have to look at a magnified section
(Figure 6) in the region of large fluctuations of R(Direct) (see Figure 4).
Results presented in Figures 5 and 6 show that the two numerical integrations of the eqn (1)
give excellent description of data. However, while the numerical integration using
Re (t ) = R ( Refined ) reproduces the general trend of the GDP/cap distribution, the calculation
based on using Re (t ) = R( Direct ) reproduces the fine structure.
We can now understand the origin of the fine structure, which can be seen in Figure 5, and
even more clearly in Figure 6. These small ripples are caused by strong fluctuations in the
growth rate. However, it is important to notice that even strong fluctuation in the growth rate
do not change the growth trajectory.
It is incorrect to claim that fluctuations in the growth rate represent evidence of the existence
of Malthusian stagnation. They do not. Whatever might be their origin, they have no tangible
effect on the growth trajectories and consequently on the mechanism of growth. Fluctuations
in the growth rate can be neglected when trying to understand the mechanism of growth.
234
__________________________________________________________________________________
Figure 5. The growth of income per capita (GDP/cap) in the UK between 1830 and 2008. Data of
Maddison (2010), are reproduced by carrying out numerical integration of the eqn (1) using
Re (t ) = R ( Direct ) or Re (t ) = R ( Refined ) , both growth rates displayed in Figure 4. Both numerical
calculations give good representation of data.
Figure 6. Magnified section of the GDP/cap distributions showing the difference between the results
of two numerical integrations of the eqn (1). These calculations explain the origin of the fine structure
of GDP/cap distribution.
We shall now turn our attention to the analytical solutions of the eqn (1) given by the eqn
(2). In Figure 7 we show two such solutions, using f (t ) representing either the best, sixth-
order polynomial, fit to R(Refined) or the best linear fit to R(Direct). In order to examine the
differences between these two solutions, we are displaying data every 10 years.
235
__________________________________________________________________________________
Figure 7. Two analytical solutions of the differential eqn (1) compared with the GDP/cap data
(Maddison, 2010). While the solution obtained using f (t ) represented by linear function reproduces
the general trend of the GDP/cap distribution, the solution corresponding to f (t ) represented by the
sixth-order polynomial reproduces the gentle oscillations around the general trend.
Results presented in Figure 7 show that the solution based on using the f (t ) represented by
a linear function reproduces the general trend of the GDP/cap distribution but the calculations
based on using the sixth-order polynomial, which reproduces the oscillating behaviour of
R(Refined), reproduces also the gentle oscillations of the GDP/cap around the general trend.
We can now understand the origin of these gentle oscillations in the GDP/cap distribution:
they are generated by the long-term oscillations of the growth rate. These oscillations are
present in R(Direct) but they are obscured by strong fluctuations. However, they are revealed
in R(Refined), which is calculated using interpolated gradient.
Thus, in summary, combining results presented in Figures 5-7, we can see that small ripples
in the time-dependent distributions describing growth, if present, reflect strong fluctuations in
the growth rate, while gentle oscillations around the prevailing trend reflect the long-term
oscillations in the growth rate.
We have now shown how the interpretations of the mechanism of growth can be simplified.
We do not have to worry about the strong fluctuations or about the long-term oscillations in
the growth rate. We can concentrate our attention of the general trend of the time-dependent
distributions and on simple mathematical representations of the growth rate.
Of course, if we want to go a step further and to try to explain the origin of minor forces,
which have no impact on the general trend, we would have to study the oscillations or minor
ripples in the time-dependent distributions. Maybe such studies could lead to some interesting
discoveries but they would have no impact on explaining the prevailing mechanism of growth.
4.2. Predicting growth
We can now use the GDP/cap data between AD 1830 and 2008 to predict economic growth.
We can see that the growth is not exponential because the best linear fit to the growth rate is
not constant. The linear fit,
f (t=
) a0 + a1t , (3)
236
__________________________________________________________________________________
shown in Figure 4 is described by parameters a0 = 8.964 102 and

= a1 5.549 105 . The
gradient is small but positive, which means that the growth rate is steadily increasing. The
growth rate around AD 1830 was about 1% but by 2000 it increased to around 2%. By 2050, it
is projected to increase to 2.2% and by 2100, to 2.5%.
The predicted growth is faster than the corresponding exponential growth fitting the same
data. Any exponential growth becomes unsustainable after a certain time but the growth of
income per capita in the UK is going to become unsustainable even faster than the
corresponding exponential growth. The predicted growth is shown in Figure 8 and in Table 1.
Figure 8. The projected growth of income per capita in the United Kingdom.
Sustainability of economic growth is defined not only by the availability of natural resources
but also by the associated stress to maintain a given growth. We have defined the relative stress
factor for the growth of the GDP (Nielsen, 2015c). We can use the same definition for the
growth of income per capita. Thus, the relative stress factor for the growth of income per
capita can be defined as
(GDP / cap )t
(4)
(GDP / cap )t0
where (GDP / cap )t is the income per capita at a certain time t and (GDP / cap )t0 is the income
per capita at a certain, fixed time t0 .
The relative stress factor in 2015 was only 35% higher than in 2000. A 35% greater effort
was required to keep the economy growing along this new trajectory. By 2050, the stress factor
is projected to increase to 2.84. Economic output per year will have to be almost three times as
high as in the year 2000 to keep the economy growing along the same trajectory. Such a large
stress might be already hard to tolerate. By the end of the current century, the annual economic
output per year will have to be about 9 times as high as in the year 2000 and by 2150 it would
have to be about 35 times as high. Even with unlimited natural resources, there will come a
time when such a large economic output will be physically impossible to achieve and the
economic growth will either have to be diverted to a new trajectory or it will collapse.
237
__________________________________________________________________________________
Table 1. Projected growth of income per capita in the United Kingdom and the associated
relative stress factor.
Year GDP/cap R Year GDP/cap R

2000 22,031 1.95 1.00 2080 125,268 2.39 5.69
2015 29,717 2.04 1.35 2090 159,536 2.45 7.24
2020 32,924 2.06 1.49 2100 204,291 2.50 9.27
2030 40,579 2.12 1.84 2110 263,033 2.55 11.94
2040 50,289 2.17 2.28 2120 340,520 2.61 15.46
2050 62,662 2.23 2.84 2130 443,246 2.66 20.12
2060 78,508 2.28 3.56 2140 580,121 2.72 26.33
2070 98,899 2.34 4.49 2150 763,419 2.77 34.65
GDP/cap in the 1990 International Dollars; R the growth rate of the GDP/cap, in per cent; - the
relative stress factor, in per cent.
The general drive everywhere, not only in the UK but also in other countries, is to keep the
economic growth rate increasing or at least constant. This is a serious mistake. Even with a
constant growth rate, which describes exponential growth, such economic growth will become,
at a certain stage, impossible to maintain, even if we had unlimited natural resources. To make
the economic growth safe and secure, the growth rate should be now slowly decreasing, not
only in the UK but also globally (Nielsen, 2015c).

We have carried out the analysis of Maddisons data (Maddison, 2010) describing income
per capita (GDP/cap) in the United Kingdom between 1830 and 2008. Our analysis is based on
solving differential equation describing the growth rate. We have presented two numerical and
two analytical solutions of this equation. We have explained the origins of various features of
the time-dependent GDP/cap distribution.
We have demonstrated that strong fluctuations in the growth rate do not change the growth
trajectory. They can, at best, be reflected only as just small ripples along the prevailing trend.
It is incorrect to interpret even strong fluctuations in the growth rate as the evidence of the
existence of Malthusian stagnation because these fluctuations have no impact on shaping
growth trajectories.
Long-term oscillations in the growth rate can be reflected as small oscillations of the growth
trajectory. They are also unlikely to affect the general trend of growth. The mechanism of
growth is determined by the prevailing trend of the growth trajectory.
In order to study the mechanism of growth or to predict its future there is no need to worry
about reproducing mathematically the details of the corresponding growth rate. Random
fluctuations and long-term oscillations in the growth rate can be neglected and the growth rate
can be reproduced by a simplest function. Often, it is possible to do it by using linear functions.
Indeed, using the simplest descriptions of the growth rate is most acceptable.
Our analysis demonstrated that the current economic growth in the UK is unsustainable even
if supported by unlimited natural resources, because after a certain time it will be impossible
to maintain the ever-increasing output. At a certain time in the future, economic growth will
have to start to be diverted to a slower trajectory or it will be likely to collapse.
The same problem applies globally (Nielsen, 2016c). Global economic growth should now,
or soon, be characterised by a slowly decreasing growth rate. The example of the economic
growth in Greece shows that rapid decrease or increase in the growth rate can lead to
catastrophic results (Nielsen, 2015d).
238
__________________________________________________________________________________
References
Press.
Publishing Company.
475.
http://dx.doi.org/10.1007/s11698-011-0062-7
Economics, 108(3), 681716.
http://dx.doi.org/10.4236/am.2014.513188
Nielsen, R. W. (2015b). Mathematics of predicting growth. Economic Review, 2(1), 228-238.
Nielsen, R. W. (2015c). The insecure future of the world economic growth. Journal of Economic and Social Thought, 2(4),
242-255.
Nielsen, R. W. (2015d). Early warning signs of the economic crisis in Greece: A warning for other countries and regions.
Journal of Economics and Political Economy, 2(4), 460-466.
Nielsen, R. W. (2016b). Mathematical analysis of historical income per capita distributions. Economic Review, 3(2), 300-
319.
Nielsen, R. W. (2016c). The unresolved mystery of the great divergence is solved. Journal of Economic and Social Thought,
3(2), 196-219.
Economic Review, 3(1), 16-27. http://www.kspjournals.org/index.php/TER/article/view/650
239
__________________________________________________________________________________
Economics Library, 3(2),241-256.
and Social Thought, 3(1), 1-34. http://www.kspjournals.org/index.php/JEST/article/view/669
Nielsen, R. W. (2016g). Unified Growth Theory contradicted by the mathematical analysis of the historical growth of human
Nielsen, R. W. (2016h). Scientifically unacceptable established knowledge in demography and in economic research.
Nielsen, R. W. (2016j). Growth of the world population in the past 12,000 years and its link to the economic growth.
Nielsen, R. W. (2016k). Industrial Revolution did not boost economic growth and the growth of population even in the
Pub.
691-712.
(Publishers) Ltd.
240
__________________________________________________________________________________
15
Interpretations of Hyperbolic Growth

By Ron W. NIELSEN
Abstract. Hyperbolic growth describes the historical economic growth and historical growth of population, but
their mechanism remains unexplained. Presented here is a brief survey of attempts to understand hyperbolic
growth. Mathematical formulations are in general complicated and there is no clear advantage in using them
because they do not give better description of data than the simple, two-parameter hyperbolic formula. They also
do not explain the mechanism of growth. The well-known simple formula suggests a simple explanation. Two
examples show how two independent investigations were on a brink of making an important and breakthrough
discovery and how their potential discovery was thwarted by the established knowledge in demography and in
economic research. Researchers who could have used their expertise to suggests new research directions and to
advance science were constrained by doctrines, which are widely accepted by faith.
Keywords. Hyperbolic growth, Mechanism of growth, Population growth, Economic growth, Growth models,
Growth theory, Malthusian stagnation
JEL. A10, A12, C02, C12, C20, C50, Y80
1. Introduction
istorical economic growth and the growth of population were hyperbolic (Nielsen, 2014, 2016a,
H 2016b, 2016c). Hyperbolic growth is described by an exceptionally simple mathematical

formula. It is just the reciprocal of a linear function. Many attempts were made to understand
hyperbolic growth or to give an alternative mathematical description. These descriptions or
interpretations tend to be complicated, maybe because hyperbolic distributions appear to be
complicated. Furthermore, they do not explain the mechanism of growth. They also do not give better
description of data than the description furnished by the simple mathematical equation. We shall present
here a few examples of earlier attempts to explain or to describe hyperbolic distributions.
2. Technology and the growth of population

Using correlations between two processes might be tempting to explain the mechanism of growth
but correlations could be spurious and misleading. Just because there is a correlation between two
processes it does not mean that one process influences another. It does not also mean that there is a
cause-effect relation between two observed processes. One has to be on guard when using such
correlations because they can lead easily to loops and to the incorrect interpretation of the mechanism
of growth.
The correlation between technology and the growth of human population is deceptively misleading
and it leads quickly to a dubious loop (Korotayev, Malkov, & Khaltourina, 2006a): technology increases
the carrying capacity, the increased carrying capacity promotes population growth, population growth
promotes the growth of technology, technology increases the carrying capacity, and so on. It is
explaining one unknown mechanism by another unknown mechanism. It is going in circles and
explaining nothing.
Australia.
. +61407201175
Published as: Nielsen, R. W. (2016). Interpretations of Hyperbolic Growth. Journal of Economics and Political Economy,
3(4), 594-626.
241
__________________________________________________________________________________
Technology might be helpful in supporting the existence of people but is it essential? When trying
to explain the mechanism of growth it is necessary to consider first the most obvious and most essential
force or forces. Other forces may be added if the essential force is insufficient to explain growth.
It is obvious that the essential force controlling the growth of population is the force of procreation.
Technology does not produce children and it is even not essential to support growth, as demonstrated
by the fast growth of population in poor countries.
Even if we briefly agree that technology supports or limits the growth of human population, such an
explanation ignores the obvious and indispensable force of growth of human population, the force of
procreation. It ignores the abundant evidence that even without advanced technology people can still
produce children and support them.
This close-loop explanation is supported by the assertion that throughout most of human history
the world population was limited by the technologically determined ceiling of the carrying capacity of
land (Korotayev, Malkov, & Khaltourina, 2006a, p. 18. Italics added.). It is a typical claim based of
pure imagination, a statement that has to be accepted by faith. How can we possibly prove that over
thousands of years and all over the world, the growth of human population was so finely tuned to the
the technologically determined ceiling of the carrying capacity of land?
When this statement was published and when the associated closed-loop explanation was proposed
it was already well known that the growth of human population was hyperbolic, at least during the AD
era (Kapitza,1992, 1996, 2006; Kremer, 1993; Podlazov, 2002; Shklovskii, 1962, 2002; von Foerster,
Mora, & Amiot, 1960; von Hoerner, 1975). Evidence-based indication is that hyperbolic growth was in
general unconstrained and surprisingly robust over a long time. This type of growth contradicts the
concept of the limiting effects of the ceiling of the carrying capacity. This ceiling appears to have been
always much higher than required for supporting growth, the conclusion being in agreement with the
study of the ecological capacity and ecological footprints showing that only recently we have crossed
the ecological limit of our planet (Ewing, Moore, Goldfinger, Oursler, & and Wackernagel, 2010).
To accept this closed-loop explanation we would have to accept, without a proof, that each
component in this loop was not only finely tuned but also that they were all for some mysterious and
unexplained reason increasing hyperbolically: the population was increasing hyperbolically, the
technology was increasing hyperbolically, the carrying capacity was increasing hyperbolically and all
of them were so finely tuned as to increase in unison, in such perfect harmony and so close to each
other. The size of the population would have to be all the time close to the limiting ceiling of the
carrying capacity, which would be so mysteriously increasing.
The proposed closed-loop explanation breaks also down already in the first step. What if the carrying
capacity was already so large that the assumed contribution from technology was inconsequential? The
size of the population in the past was small over a long time. It is hard to accept that our planet was
incapable to support the increasing population.
With the exception of just two demographic transitions in the past 12,000 years (Nielsen, 2016a),
the growth of human population was increasing without any major disturbance. With the small number
of people and with the huge resources of our planet we can reasonably expect that the carrying capacity
was much higher than the size of human population.
It would be unrealistic and unconvincing to assume that the growth of human population over such
a long time was so precisely adjusted to the carrying capacity. It would be unrealistic to expect that this
fine tuning was done so precisely by technological development. To make such a claim we would first
have to prove that the growth of human population was always limited by the carrying capacity of our
planet but we have no such proof and probably we shall never have. Any theory, which attempts to
explain the mechanism of growth of human population by fine tuning of the size of population to the
carrying capacity by technology, economic growth or by any other means is either unscientific (because
it is based on untestable assumptions) or at least strongly questionable.
We would have to have some incredibly advanced technology to measure the carrying capacity and
to adjust the growth of human population so precisely to its ceiling. But even then, we could hardly
expect such a regular hyperbolic growth. By using this advanced technology, we would also have to
control precisely three interacting processes: technological development, the increase in the carrying
capacity and the growth of human population. We would have to make sure that these three processes
are perfectly synchronised and that they follow the closely coupled hyperbolic trajectories.
242
__________________________________________________________________________________
To justify the closed-loop process we would have to explain it without assuming that it was
controlled by any advanced technology. Without such explanation, the mechanism of the proposed
closed-loop remains unexplained and consequently, it does not explain the mechanism of the growth of
human population.
We can also have other questions about this first step in the postulated closed loop. What is the
carrying capacity of our planet? What was the carrying capacity of our planet over the past 12,000 years
or longer? What was the contribution of technology to the carrying capacity? Even if we assume that
technology increases the carrying capacity, is this assumed increase so essential to support the growth
of human population? It is well known that people can survive on very little and that even then they can
still procreate and support children. All they need is basic food, body cover and shelter.
How much damage is caused by technology? How is the technology reducing the carrying capacity?
Can we ignore, for instance, that carbon footprint accounts for about 50% of our total ecological
footprint? (Ewing, Moore, Goldfinger, Oursler, & and Wackernagel, 2010). Can we ignore the pollution
of not only the atmosphere but also of the land and water? Can we ignore climate change, the ever-
increasing weather-related economic losses, the decreasing carrying capacity of people living on
islands, the increasing deforestation, the continuing human-induced extinction of species, the
continuing loss of arable land, the overuse of pesticides, herbicides, artificial fertilisers and other
agricultural chemicals? Can we ignore the ever-increasing urban population and their increasing
dependence on food supply, which comes from the decreasing land resources? Can we ignore how the
huge and the well-stocked arsenal of weapons is relentlessly used to destroy the carrying capacity? Can
we ignore the never-decreasing stream of displaced population?
If we want to claim that technology increases the carrying capacity, we should also consider how
this carrying capacity is decreased by technology. But the essential point is to show that technology was
indeed playing the crucial role in shaping the growth of population, that this assumed force of growth
has to be added to the essential and indispensable force of procreation, that without technology
population would not have been increasing or that it would not have been increasing hyperbolically.
Another problem with linking technological development with the growth of human population is
the misinterpretation of the fundamental mechanism of technological development. Technological
growth is not prompted by the sheer number of people but by concepts, ideas and solutions. This is the
driving force of technological development. People are just carriers of these concepts, ideas and
solutions, or more precisely, carriers of the genetic ability to generate concepts, ideas and solutions.
Is the technological development dependent on the number of people? While it is true that with a
larger number of people we can expect a greater number of ideas and solution, it is also true that the
growth of human population is now slowing down. Does it mean that technological development is also
slowing down because of the slowing down growth of human population? If the growth of human
population is going to reach a maximum and stop growing, will the technology also reach a certain
maximum and stop growing? Will people stop thinking and inventing?
The growth of technology is not determined by the number of people but by the number of creative
ideas, inventions and solutions, which do not appear to be directly proportional to the number of people.
Consequently, even if the size of population is going to be constant, people will not stop being
intellectually active.
The correlation between technology and the growth of human population was investigated by
Kremer (1963). He claims that there is a close correlation between the growth of population and
technological development, which is hardly surprising. However, by observing a correlation between
two processes we can only tell that there is a correlation. The correlation alone does not explain the
mechanism of growth of any of the correlated processes. Correlations can be strongly misleading and
they have to be handled with care.
Kremer claims that the growth rate of human population during the AD era was approximately
proportional to the size of human population indicating that the growth was hyperbolic but he did not
explain why it was hyperbolic. He suggests the correlation between the growth of human population
and the growth of technology but this correlation does not explain the mechanism of growth of any of
them. It does not explain why these two correlated processes are hyperbolic. It is like with the finely-
tuned closed-loop mechanism proposed by Korotayev, Malkov and Khaltourina, (2006a): one process
is explained by another without explaining any of them. The growth of human population is hyperbolic
because the growth of technology is hyperbolic, and the growth of technology is hyperbolic because
243
__________________________________________________________________________________
the growth of population is hyperbolic. It is also explaining one unknown mechanism by another
unknown mechanism and going in circles.
The primary, if not the only, force driving the growth of population is the force of procreation,
which in its simplest representation is the biologically controlled sex drive and biologically controlled
mortality. Until recently, children were not produced by technology. Mortality was also not controlled
by technology. Maybe technology could be claimed to give a better chance of survival but it is definitely
not the primary force of growth. Likewise, the primary force driving technological development can
be identified as concepts and ideas created by people, combined with the efficiency of sharing
information.
The primary force of the growth of population is represented by the biological processes controlling
birth and death. The primary force controlling the growth of technology is represented by concepts,
ideas and generally by creative activities of human population. Biological process controlling birth and
death apply not only to humans but also to other species. The force of creative thinking applies
specifically only to humans. There might be some examples of creative thinking in other species,
particularly in primates, but they are on such a low level that they do not initiate some new lines of
technological development.
If the force responsible for the growth of technology were determined by the sheer number of
people, i.e. by the number of members of this particular species, there would be no reason for excluding
other species from this process. They should be also expected to develop technology but they do not.
The growth of technology and the growth of population are controlled by different forces of growth.
Their fundamental mechanisms of growth are distinctly different.
There is a close correlation between the growth of technology and the growth of human population
only because creative concepts come from humans. Explaining the growth of population by technology
and technology by the growth of population is going in circles and explaining nothing.
3. Convoluted construction
Hyperbolic growth is described by a simple formula:
1
S (t ) = , (1)
C kt
where S (t ) is the size of the growing entity, such as population or the Gross Domestic Product (GDP),
C is the constant of integration, k is a positive constant and t is time.
This expression is a solution of a very simple differential equation:
1 dS (t )
= kS (t ) . (2)
S (t ) dt
Normally, when we already know that the historical growth of population and the historical
economic growth can be described in such a simple way, the next step would be to explain why the
growth was hyperbolic. To this end, we would have to start with some simple and easily acceptable
assumptions and derive the hyperbolic formula based on these assumptions. Maybe we could also start
with acceptable assumptions and derive and alternative formula, which would give a better description
of data. However, if we derived a more complicated formula, which would not give a better description
of data we could then decide that we were on the wrong track and we would have to try another
approach.
In contrast, in the demographic and economic research there appears to be a tendency to construct
mathematical formulae, for no good reason, and to try to make them as complicated as possible. Here
is one such example (Johansen & Sornette, 2001) illustrating an interesting, step by step construction.
Start with the logistic equation of growth
244
__________________________________________________________________________________
= b[K S (t )] .
1 dS (t )
(3)
S (t ) dt
This is already a questionable starting point because we know that population and the GDP do not
grow logistically but hyperbolically. Even now, they do not yet level off (Nielsen, 2016d) to suggest a
conversion to a logistic-type of growth.
Assume that the limit to growth K depends on time.
= b[K (t ) S (t )] .
1 dS (t )
(4)
S (t ) dt
For no apparent reason, delete S (t ) from the right-hand side of the eqn (4).
= b[K (t )] .
1 dS (t )
(5)
S (t ) dt
Again, for no apparently good reason, assume that
K (t ) = [S (t )] ,

(6)
where > 1.
Under this assumption, eqn (6) is now changed to
= b[S (t )] .
1 dS (t )
(7)
S (t ) dt
This equation can be presented as
= b[S (t )] .
dS (t ) +1
(8)
dt
1
We can solve it by substitution S = Z . The solution is
1
S = c bt , (9)

where c is the constant of integration. So now we have
S (t ) = (b ) (tc t ) ,
z z
(10)
where z = 1 / and tc = c / b is the time of singularity when S (t ) escapes to infinity.

Replace (b ) z by an arbitrary and adjustable parameter B and add another arbitrary and adjustable
parameter A to construct
S (t ) = A + B(t c t ) .
z
(11)
Assume that the parameter z is a complex number
245
__________________________________________________________________________________
z = ( + i ) (12)
So now we have
S (t ) = A + B(tc t )
+ i
. (13)
Find the real component of (tc t ) + i .

This is an easy exercise that can be completed using two well-known formulae:
x y = e y ln x (14)
and
e i = cos + i sin . (15)
The answer is
Re(tc t ) + i = (tc t ) cos[ ln(tc t )]. (16)
Assuming that both A and B are real, the formula for S (t ) can be now expressed as
Re S (t ) = A + B Re(tc t ) + i , (17)
which with the help of the eqn (16) gives
Re S (t ) = A + B (tc t ) cos[ ln(tc t )] . (18)
Use the eqn (13) again but now delete i .
S (t ) = A + B(tc t ) .

(19)
Return to the eqn (16) and multiply the right-hand side of this equation by a constant D.
Re(tc t ) + i = D(tc t ) cos[ ln(tc t )] . (20)
Add a phase shift in the eqn (20).
Re(tc t ) + i = D(tc t ) cos[ ln(tc t ) + ] . (21)
Return to the equation (19) and add to it the right-hand side of the eqn (21). We have now constructed
the equation published by Johansen and Sornette (2001).
S (t ) = A + B(tc t ) + D(tc t ) cos[ ln(tc t ) + ]

(22)
This equation contains seven adjustable parameters but we do not know how they are supposed to
be linked with the mechanism of growth. We know how we constructed (not derived) this complicated
and impressive formula but we do not know why we have it and indeed why we should be interested in
using it except perhaps to draw a line through data points, which we could do equally successfully using
pen and paper and obtain equally meaningless result.
246
__________________________________________________________________________________
It is always good to look for mathematical description of data because it could help in understanding
the nature of the observed phenomenon. However, if complicated description is not better than
description given by a simple mathematical formula there is obviously no advantage in using the
complicated description.
Hyperbolic growth described by the eqn (1) gives a satisfactory description of the growth of
population and of the economic growth (Nielsen, 2014, 2016a, 2016b, 2016c). This is a simple formula,
which could be expected to have a simple explanation. But now, we have a significantly more
complicated formula. So, rather than making our task of explaining the mechanism of growth easier we
have made it even more complicated.
In Figure 1, the distribution generated by the complicated eqn (22) is compared with the first-order
hyperbolic distribution described by the eqn (1) and with data. As explained elsewhere (Nielsen, 2016a),
fitting data around AD 1 by using hyperbolic distribution is pointless because around that time there
was a transition from a fast to a slow hyperbolic trajectory. However, if we replace the complicated
formula of Johansen and Sornette by a significantly simpler reciprocal of the second order polynomial
S (t ) = (a0 + a1t + a2t 2 ) 1 (23)
we can generate a virtually identical distribution. There is no clear advantage in using the
complicated formula of Johansen and Sornette. Simple description using the first-order hyperbolic
distribution given by the eqn (1) gives acceptable representation of data but we can also replicate the
complicated seven-parameter calculations but using just three parameters.
Figure 1. Growth of the world population calculated using the Johansen and Sornettes (2001) constructed
formula (22) is compared with the calculations based on significantly simpler formulae given by the eqns (1)
and (24). Population data come from numerous sources compiled by Manning (2008) and by the US Census
Bureau (2016). The parameters for the distribution of Johansen and Sornette given by the eqn (22) are: A 0,
B 1624, D 127, z 1:4, tc 2056; 6:3 and 5.1. Parameters for the hyperbolic distribution
=
given by the eqn (1) are: =
C 7.875 100 and k 3.834 103 . Parameters for the reciprocal second-order
polynomial distribution given by the eqn (24) are:=
a0 3.367 100= , a1 1.172 103 and
1.382 106 .
a2 =
247
__________________________________________________________________________________
The aim of constructing this complicated formula appears to be misplaced because even Johansen
and Sornette used a significantly simpler formula in their analysis of a wide range of data presented in
their Figs 9-32. The formula they used was
S (t ) = a (tc t ) z . (24)
However, even in this simplified form it is already unnecessarily more complicated than the eqn (1)
because S (t ) is no longer represented by the reciprocal of a linear function but by the time difference
taken to the power of z. This expression is linear only if z = 1 . For integer values of z > 1 it describes
higher-order polynomials. For integers z < 1 it describes reciprocals of higher order polynomials.
However, z can be also any other number greater or smaller than zero.
4. The homeostatic simulation model

In conformity with the generally accepted established knowledge in demography and in economic
research (Nielsen, 2016d), Artzrouni and Komlos (1985) imagined that the growth of population can be
divided into two distinctly different regimes: Malthusian stagnation and explosion. These regimes of
growth are assumed by be controlled by two distinctly different mechanisms of growth. They assumed
incorrectly that the growth before the Industrial Revolution was controlled by random forces such as
wars, famines and diseases, the mechanism causing allegedly stagnation in the growth of population.
They also assume, incorrectly, that the growth after around the Industrial Revolution was exponential.
They should have known that their assumptions were unrealistic and incorrect because many years
earlier it has been shown that the growth of population was hyperbolic (von Foerster, Mora, & Amiot,
1960; von Hoerner, 1975). Hyperbolic growth cannot be divided into two regimes of growth, slow and
fast. For this type of growth, Malthusian regime does not exist and the apparent explosion is just the
natural continuation of hyperbolic growth. There was no stagnation in the growth of human population
and in the economic growth and there were no takeoffs leading to distinctly different explosive growth
(Nielsen, 2014, 2015, 2016a, 2016b, 2016c, 2016e, 2016f, 2016g, 2016h).
Their work is important because, unknown to them, they have demonstrated that the established
knowledge is contradicted by science. They did not realise that they made an important discovery
because typically for the research carried out within the constraints of the established knowledge they
did not compare results of their research with data.
To generate the growth of population before the Industrial Revolution, Artzrouni and Komlos carried
out Monte Carlo simulations of the alleged Malthusian regime of stagnation. To describe the alleged
population explosion, they simply assumed exponential growth after the Industrial Revolution. In their
model, the growth of population is given by
S (t )
= rS (t ) . (25)
t
For the constant r, this equation would describe exponential growth. However, in their calculations,
the growth rate r is either constant (after the Industrial Revolution) or time-dependent (before the
Industrial Revolution).
So, more explicitly, they consider two stages of growth. Before the Industrial Revolution the growth
is given by:
S (t )
= r (t ) S (t ) , (26)
t
whereas after the Industrial Revolution it is given by
248
__________________________________________________________________________________
S (t )
= re S (t ) , (27)
t
where re is a certain constant escape rate (Artzrouni & Komlos, 1985, p 27), escape from nowhere
because there was no escape, or more accurately there was nothing to escape from, because the mythical
Malthusian trap did not exist. The growth of population was monotonically hyperbolic, and the
Industrial Revolution had no impact on changing the growth trajectory. However, according to the
established but erroneous knowledge, there was an escape.
Fluctuations in the growth rate r (t ) before the Industrial Revolution are determined by e(t )
described as a nonnegative random variable generated by a Monte Carlo type of simulation (Artzrouni
& Komlos, 1985, p. 27). For no apparent reason, this variable is defined by the following equation:
{ }
e(t ) = 0.1v(t )U (t ) 1 + e 0.15[ y ( t ) 5 ] , (28)
where v(t ) is a random number drawn from a normal distribution with the mean 0 and variance 1, y (t )
is the number of decades the population was in the assumed Malthusian crisis and U (t ) is defined
(again for no clear reason) as
1
U (t ) = 400[ PT P ( t )]
. (29)
1 + 4e
The population is divided into two sectors: the subsistence sector (rural) and the capital producing
sector (urban). In the eqn (29), P (t ) represents the per capita output (production) of the subsistence
sector. If the per capita output is below a certain threshold defined by PT , i.e. if P (t ) < PT , the
population is assumed to be in the Malthusian crisis and cannot grow. If P(t ) PT , the population is
assumed to be out of crisis and can increase.
The per capita output in the subsistence sector is defined as
[ S R (t )]2
P(t ) = C2 [ K (t )]1 , (30)
S (t )
where C2 , 1 and 2 are positive constants with 1 + 2 = 1 , K (t ) is the capital stock and S R (t )
is the population in the subsistence (rural) sector.
The total output in the subsistence sector is given by
QR (t ) = C2 [ K (t )]1 [ S R (t )] 2 . (31)
Likewise, the total production in the capital producing sector (urban) is given by
QU (t ) = C1[ K (t )]1 [ SU (t )] 2 , (32)
where C1 , 1 and 2 are positive constants with 1 + 2 = 1 .

The total population is then given by
S (t ) = S R (t ) + SU (t ) . (33)
Returning to the eqn (28) we should notice that the function U (t ) defining the time-dependent
parameter e(t ) , which plays the essential role in the Monte Carlo simulations, depends on P (t ) , which
in turn depends on the capital stock K (t ) . The growth of the capital stock is described as
249
__________________________________________________________________________________
K (t )
= (t )QU (t ) , (34)
t
where (t ) is defined as
(t ) = 0.01 + 1.778 10 26 e 0.05756t . (35)
The process of Monte Carlo simulations is well described by Artzrouni and Komlos (1985). These
simulations produced most interesting results. Designed to demonstrate the existence of Malthusian
regime of stagnation, the model shows that the regime of Malthusian stagnation did not exist. The
assumed mechanism of stagnation does not produce stagnation but a steadily-increasing growth.
Furthermore, the growth generated by the model during this imaginary regime of stagnation does not
fit the data (see Figure 2).
Figure 2. The established knowledge in demography is contradicted by science. Simulations of the

mechanism of Malthusian stagnation carried out by Artzrouni and Komlos (1985) do not produce stagnation but
a steadily increasing exponential growth. They also do not fit date. These calculations are compared with
hyperbolic distributions (Nielsen, 2016a). The data represent the average values of the size of population
calculated using the compilations of Manning (2008) and of the US Census Bureau (2016).
Parameters describing hyperbolic distributions shown in Figure 2 and defined in the eqn (1) are:
C=
2.282 100 and
= k 2.210 102 for the BC era and
= C 7.061100 and
= k 3.398 103 . The
data come from a variety of sources compiled by Manning (2008) and the US Census Bureau (2016).
The model of Artzrouni and Komlos (1985), designed to reproduce the well documented
fluctuations experienced by the worlds population throughout history (Artzrouni & Komlos, 1985, p.
24), produced instead a steadily increasing growth along exponential trajectory. (In the semilogarithmic
display, exponential growth is represented by an increasing straight line.) Furthermore, their model
calculations do not fit the data. Their results show clearly that the model of Malthusian stagnation does
not work. The mechanism of Malthusian stagnation does not describe the growth of population. The
model based on the assumption of the mechanism of Malthusian stagnation did not generate the required
fluctuations in the growth of population let alone fluctuations that were, to a large extend, brought
250
__________________________________________________________________________________
about by randomly determined demographic crises (wars, famines, epidemics, etc) (Artzrouni &
Komlos, 1885, p. 24).
Thus, if Artzrouni and Komlos took the final step normally expected in scientific investigations, if
they compared theory with data, even with the data used by von Foerster, Mora and Amiot (1960), they
would have made an important discovery that the fundamental concepts of the established knowledge
in demography are incorrect. They would then be able to suggest new lines of research.
It is essential to notice that even though Monte Carlo simulations based on the assumption of the
mechanism of Malthusian stagnation produced exponential growth it would be incorrect to claim that
the mechanism of Malthusian stagnation generates exponential growth. Equation (26) makes it clear
that Artzrouni and Komlos (1985) assumed exponential growth. They assumed that Monte Carlo
calculations were fluctuating around the growth rate describing exponential growth because eqn (26)
describes modulated exponential growth. If we assume exponential growth it is hardly surprising that
we get exponential growth. Fluctuations in the growth rate are not readily reflected as fluctuations of
the growth of population or the GDP (Nielsen, 2016i, 2016k).
5. Lagerlfs model of growth

Lagerlfs model of growth (2003a, 2003b) belongs to the so-called OLG (overlapping generations)
models (Aliprantis, Brown & Burkinshaw, 1990) used for instance by Becker, Murphy and Tamura
(1990) and by Galor (2005a, 2011) to look at the growth of the population from the economic
perspective. The central idea of this approach is to try to explain the growth of population by considering
human capital defined as embodied knowledge and skills (Becker, Murphy & Tamura, 1990, p. S13).
The growth is on the favourable rates of return.
When human capital is abundant, rates of return on human capital investments are high relative to return
on children, whereas when human capital is scarce, rates of return on human capital are low relative to
those on children. As a result, societies with limited human capital choose large families and invest little
in each member; those with abundant human capital do the opposite (Becker, Murphy & Tamura, 1990, p.
S35).
It is a strong assumption, which is hard to accept. One would have to have a strong proof that this
assumption is correct but we do not have such a proof.
It is interesting that neither Becker, Murphy and Tamura (1990), nor Galor (2005a, 2011), nor
Lagerlf (2003b) tried to compare their model predictions with population data. Lagerlf (2003b) came
close to testing his model against data when he generated growth rates in his Monte Carlo simulations
but we shall show that his model is in disagreement with data he was referring to in his publication.
Lagerlfs model is an excellent example of convoluted models characterised by the abundance of
parameters but models, which neither describe data nor explain the mechanism of growth. This model
was also designed to reproduce the epoch of stagnation and the alleged transition from stagnation to
growth at the time of the Industrial Revolution, all as specified by the prescribed instructions of the
established knowledge in demography and in the economic research. Like Artzrouni and Komlos
(1985), Lagerlf was also on the verge of making an important discovery that the established knowledge
in demography and in economic research is contradicted by science. Like Artzrouni and Komlos (1985),
he was on the verge of proving that the epoch of Malthusian stagnation did not exist and that there was
no transition from stagnation to growth. Like Artzrouni and Komlos (1985), he was on the verge of
showing that simulations of Malthusian stagnation do not produce stagnation, that they do not fit data
and that they do not explain the mechanism of growth. He missed making this important discovery
because he did not take the final step normally expected in scientific investigations he did not compare
theory with data. Parameters and definitions used in Lagerlfs model are listed in Table 1.
Assuming that each person (agent) is endowed with a unit of time, the time budget for each agent is
given by
1 = lt + (v + ht ) Bt (36)
At any given time, each person (agent) is assumed either to work or to spend time with children.
Assuming a single economy (or non-interacting economies) and that children consume (produce)
nothing, the output of the consumption (production) of goods is given by
251
__________________________________________________________________________________
Yt = lt ( L + H t ) = Ct (37)
It is simply the productivity per unit of time multiplied by the time spent at work.
The survival rate is given by
Ht 1
Tt = (38)
St t + H t / St
In the absence of mortality shocks ( t = 0 ), the survival rate Tt = 1 .
Table 1. Parameters and definitions used in Lagerlfs theory (Lagerlf, 2003a, 2003b)
Parameter Description
t Time interval or period t assumed to be 25 years, i.e. one generation
Ht Human capital or a component resulting from parental investment (Lagerlf,
2003b, p. 426) called also human capital stock (Lagerlf, 2003a, p. 760)
L The units of skills (Lagerlf, 2003b, p. 426) endowed by nature to every
agent (person)
L + Ht The productivity of a unit of time (Lagerlf, 2003b, p. 426)
v a fixed time cost of rearing one child i.e. the time required to nurse the child
just enough to keep her alive (Lagerlf, 2003a, p. 759)
v Assuming 0 < < 1 , this product measures the direct inheritance of human
capital from one generation to the next reflecting the assumption that less than
100% of the time invested in rearing (nursing) a child is converted into human
capital.
ht The time invested in the education of each child
lt The time input in the consumption good sector (Lagerlf, 2003a, p. 759) i.e.
time spent on production or work
t The mortality shocks (Lagerlf, 2003b, p. 426) which can be interpreted as
epidemics (Lagerlf, 2003a, p. 760), the function assumed to be described by
the probability density function of a log-normal distribution.
Pt the (adult) population size called also population density in the generation t
(Lagerlf, 2003a, p. 760). The fundamental assumption of OLG models is that
people live only for two generations. All adults in the generation t are replaced
by the children born during the generation t. This new generation will be
completely replaced by the next generation.
A( Pt ) The productivity parameter, which enters into the equation of the time-
dependence of human capital
Bt the number of born children (or births) (Lagerlf, 2003a, p. 759). It is the
average number of children per capita of adult population born in the generation
t, i.e. over the entire 25 years.
Tt The survival rate (Lagerlf, 2003a, p. 760). It is the average fraction of the
number of individuals born during the 25 years of the generation t, who survive
to the next generation t + 1 .
(v + ht ) Bt The total time invested in children per capita of the adult population calculated
over the entire time of one generation, i.e. over the total time of 25 years
Yt The output of the consumption (production) of goods
Ct The adult consumption (production) (Lagerlf, 2003b, p. 426)

A parameter ( > 0 ) used in the utility function
252
__________________________________________________________________________________
The production of human capital is given by
H t + 1 = A( Pt )[L + H t ]( v + ht ) (39)
Human capital increases in proportion to the productivity per unit of time multiplied by the time
spent with each child, with the part of this time corrected for the unproductive fraction of time when
nursing a child. By including the parameter 0 < < 1 it is assumed that education is more profitable
for the increasing of human capital than nursing.
Each agent is assumed to maximise a utility function describing personal preferences and is given
by
U t = ln(Ct ) + ln( BtTt ) + ln( L + H t + 1 ) (40)
The first term of the utility function measures the utility (the preference) of consumption
(production), the second measures the utility of surviving children given by BtTt and the third the utility
of human capital of the offspring.
By maximising the utility function, we get the following expression for the optimal (preferred)
number of births
1
Bt = (41)
1 + v + ht
The number of born children depends entirely on the time invested in each child corrected by a factor
dependent on the parameter used in the utility function. The larger the invested time, the smaller is the
number of children, or vice versa.
The annual crude birth and death rates ( Br , t and Dr , t , respectively) are calculated using the following
expressions:
( )
Br , t = 1000 Bt1 / 25 1 (42)
Dr , t = 1000(1 T ) t
1 / 25
(43)
Calculations become significantly more complicated if interacting countries are included. Thus, for
instance, assuming that that a demographic shock in one country is also reflected in other countries, the
survival rate can be expressed as
N 1
H i, t + k H
j i
ij j, t
Ti , t = (44)
N 1 N 1
i , t Pi , t + k ij j, t Pj , t + H i, t + kij H j , t
j i j i
If we look back at eqns (38), (42) and (43) we can see that when t = 0 , then Tt = 1 and the
death rate Dr , t = 0 , which also means that if mortality shocks are low, i.e. if t 0 , the death rate
is also approximately zero. If we assume that the time spent with each child remains approximately the
253
__________________________________________________________________________________
same over time, or equivalently that the number of born children remains approximately the same, a
dramatic decrease in mortality shocks should generate a prominent population explosion.
This mechanism is the essence of the Demographic Transition Theory, which claims that towards
the end of the assumed first stage of human history, interpreted as the epoch of Malthusian stagnation,
the death rate started to fall while the birth rate remained approximately the same, the process creating
allegedly population explosion, the explosion which in fact never happened. This is also the essence of
the three regimes of growth postulated by Galor and Weil (1999, 2000) but contradicted by the analysis
of data (Nielsen, 2016f).
The mechanism of Malthusian stagnation followed by explosion is carefully incorporated in the
Lagerlfs model of growth. In particular, regarding the mortality shock function t , Lagerlf
explains:
To understand the mechanisms driving the results in the calibration later, it is useful to first think of
economies where t is constant over time: Either high or low. To replicate the Three Regimes of Galor
and Weil (1999, 2000), discussed in the introduction, we shall rig the model so that a high- economy
converges toward a locally stable (Malthusian) steady state, whereas a low- economy converges to a
balanced growth path (Lagerlf, 2003a, 763).
The three regimes Lagerlf is writing about are the assumed Malthusian regime of stagnation, which
was supposed to last for thousands of years but which never existed; the post-Malthusian regime marked
allegedly by the rapid increase of population and economy; and the modern growth regime or sustained
growth regime, which allegedly follows a little later but which also represents an imaginary stage of
growth. We have already demonstrated that these three regimes of growth did not exist (Nielsen,
2016f).
Lagerlfs theory is based on the scientifically contradicted fantasy and if he carried his research
properly, if he compared his theory with empirical evidence, he would have soon discovered that he
was guided by fiction. His hard and convoluted work was unnecessary because it has been known for a
long time that the growth of population was hyperbolic (Kapitza,1992, 1996, 2006; Kremer, 1993;
Podlazov, 2002; Shklovskii, 1962, 2002; von Foerster, Mora, & Amiot, 1960; von Hoerner, 1975).
Hyperbolic growth can be described by an exceptionally simple mathematical formula, which is just
the reciprocal of a linear distribution. This type of growth is in contradiction of the concepts of
stagnation and explosion.
Using his model and Monte Carlo simulations, Lagerlf generated growth rate for the growth of
population in England, France and Sweden (Lagerlf, 2003b). His model produced minor fluctuations
in the growth rate, which were interpreted by Lagerlf as the proof of the existence of the regime of
Malthusian stagnation. That was a serious mistake because even large fluctuations in the growth rate
are not readily reflected in the growth of population (Nielsen, 2016i, 2016j), and we do not even have
to carry out laborious calculations to see that fluctuations in the growth rate are not reflected as similar
fluctuations in the growth of population. Data for Sweden are well known (Statistics Sweden, 1999).
They are often used in defence of the Demographic Transition Theory without even realising that the
they are in its contradiction. There, in the same document, for everyone to see, we have graphs showing
not only the fluctuating birth and death rates and fluctuating annual population increase but also a graph
of population growth with no signs of fluctuations. The usual practice of showing fluctuations in birth
and death rates or in the growth rate and claiming that we have a proof of the existence of Malthusian
stagnation is unjustified. These fluctuations are not reflected in the growth of population and
consequently they have no impact on the mechanism of growth. They are, in this respect, irrelevant.
Figure 3 shows an example Lagerlfs results for France. His model-generated growth of population
was calculated using the numerical integration of the following differential equation:
1 dS (t )
= RL (t ) , (45)
S (t ) dt
where RL (t ) is the Lagerlfs, model-generated and fluctuating growth rate, precisely as published in
his paper (Lagerlf, 2003b).
254
__________________________________________________________________________________
Figure 3. The established knowledge in demography is contradicted by science. Simulations of Malthusian

stagnation carried out by Lagerlf (2003b) do not produce stagnation but a steadily increasing growth of
population. Furthermore, his model calculations do not fit data (Maddison, 2010). Model-calculated
distribution follows exponential trajectory because the growth rate was assumed to oscillate around a constant
value. The claimed population explosion is just a small deviation from the exponential trajectory at its end. The
growth of population in France was hyperbolic.
In Figure 3, Lagerlfs model-generated distribution is compared with the exponential distribution

and with data. We also show the hyperbolic distribution fitting the data (Maddison, 2010). These data
were not available to Lagerlf but he had access to similar data (Maddison, 2001) published before the
publication of his work.
Parameters describing hyperbolic distribution are:= C 2.085 101 and= k 9.635 105 [see eqn
(1)]. The exponential distribution, which is so closely followed by Lagerlfs model-generated results,
is described by the following equation:
S (t ) = C 'e rt . (46)
4
Its parameters are= C 3.100 10 and
' 0
= r 9.780 10 .
The tiny, model-generated fluctuations in the growth rate presented in Lagerlfs publication
(Lagerlf, 2003b) could not have possibly generated oscillations in the growth of population. Even large
fluctuations are not readily reflected in distributions describing growth, such as growth of population
or the GDP (Nielsen, 2016i, 2016j). Lagerlf could have seen it clearly if he looked at the data for
Sweden (Statistics Sweden, 1999). He could have also known it if he studied the excellent data for
England (Wrigley & Schofield, 1981). These results show clearly, even without carrying out any
calculations, that even large fluctuations in birth and death rates and in the corresponding growth rate
have no tangible effect on the growth of population and consequently that they have no effect on shaping
the mechanism of growth. These data are clearly contradicting the established knowledge in
demography but they are systematically ignored. The established knowledge in demography is also
contradicted by results published over 50 years ago (von Foerster, Mora, & Amiot, 1960) but they are
also systematically ignored.
255
__________________________________________________________________________________
The important contribution of Lagerlfs Monte Carlo simulations is to show that the mechanism of
Malthusian stagnation and population explosion does not work. Such a mechanism fails to produce the
desired effects of stagnation and explosion and it fails to fit the data.
Results of Lagerlf show that his model-generated distribution was exponential and that his claimed
population explosion is just a minor deviation from the exponential trajectory. However, one might
wonder why model-generated results follow exponential trajectory. Does it mean that the process of
Malthusian stagnation generates exponential growth? No, it does not. Results depend on our
assumptions about the way birth and death rates are fluctuating.
Lagerlf assumed that crude birth rate was a non-zero constant and that crude death rate fluctuated
around a non-zero constant value. Naturally, therefore, his growth rate also fluctuated around a non-
zero constant, which in turn generates exponential growth. If Lagerlf assumed that birth rate was zero
and that the death rate fluctuated also around zero, he would have produced growth rate fluctuating
around zero and thus he would have produced a constant size of population in his Monte Carlo
calculations but he would still have not produced the required fluctuations in the size of population and
his results would have been in a clear disagreement with data. The same applies to the calculations of
Artzrouni and Komlos (1985). If they did not assume the modulated exponential growth during the
postulated epoch of Malthusian stagnation [see the eqn (26)], they would have also produced a constant
population without the so-called Malthusian oscillations.
If Lagerlf took the final step and compared his model-generated distributions with data (Maddison,
2001), if he consulted the available to him literature (Statistics Sweden, 1999; Wrigley & Schofield,
1981) he would have made an important discovery that the concept of Malthusian stagnation followed
by explosion is incorrect, that it is contradicted by data and even by his own model. He could have then
used his expertise to suggest new directions for the demographic and economic research.
The same applies to Galor. He uses Maddisons data but surprisingly he never attempts to analyse
them. He prefers to distort them (Galor, 2005a, 2005b, 2007, 2008a, 2008b, 2008c, 2010, 2011, 2012a,
2012b, 2012c; Galor & Moav, 2002) to support the preconceived but erroneous ideas. He knows
mathematics and he should be familiar with hyperbolic distributions. If he analysed data, the same data
that he used in his publications, he would have soon discovered that the established knowledge in
demography and in economic research is scientifically unsupported. He could have then also used his
expertise to suggest new lines of research. These two examples show how strongly the established
knowledge is established and that even prominent researchers can be easily misled by its system of its
doctrines.
5. Camouflaging the hyperbolic equation

Here is an example how the well-known differential equation describing hyperbolic growth was
disguised as something new, which was supposed to explain the mechanism of growth based on the
assumption that the growth of population is finely-tuned to the technological development. In its
undisguised form, the differential equation (2) describes hyperbolic growth but does not explain its
mechanism. It is just a mathematical equation, which when solved produces hyperbolic distribution.
However, in its disguised form it seems to contain an explanation of the mechanism of growth. It seems
to show that the growth of population is determined by the level of technology or knowledge.
This is a good example, which demonstrates that one should never be mesmerised by complicated
mathematics. Mathematical formulations can be complicated and useful but just because they are
complicated it does not mean that they are useful. Unified Growth Theory (Galor, 2005a, 2011), which
is supposed to explain the mechanism of economic growth, is full of such complicated mathematical
formulations. However, these complicated formulae do not explain anything. They just translate
erroneous concepts into mathematical language. Data describing economic growth and the growth of
population (Maddison, 2001) were used but they were never analysed to check the proposed theory.
They were presented in a distorted way to make the impression that theory is confirmed by data. In the
example presented here, the discussed mathematical equations are relatively simple and even attractive
but they give a corrupted and mathematically unacceptable representation of the well-known differential
equation [eqn (2)] describing hyperbolic growth.
Korotayev (2005) used the following differential equations to describe and explain the growth of
population:
256
__________________________________________________________________________________
dS (t )
= a[bK (t ) S (t )]S (t ) , (47)
dt
dK (t )
= cS (t ) K (t ) . (48)
dt
According to his interpretation K is the level of technology/knowledge, bK corresponds to the

number of people (N) [ S (t ) in our notation], which the earth can support with the given level of
technology (K) (Korotayev, 2005, p. 81). Thus, bK is interpreted as the carrying capacity of the planet.
To fit the population data, Korotayev carried out step-by-step calculations based on the eqns (47)
and (48) but presented in a different form:
K i +1 = K i + cSi K i , (49)
Si +1 = Si + a (bK i +1 Si ) Si . (50)
There is absolutely no reason why K (t ) should represent technology or knowledge. We can call it
whatever we want but just because we call it technology, knowledge or the carrying capacity it does not
mean that it represents these imposed by us concepts. In the logistic model, which is similar to the eqn
(47), it is a constant describing the limit to growth, which may or may not represent the carrying
capacity. However, we shall show that in eqns (47) and (48), K (t ) has nothing to do even with the limit
to growth. It is a variable that does not restrict growth in any way because K (t ) is in fact S (t ) . It is
simply the size of population or the size of any, hyperbolically-increasing quantity. Consequently, even
if we use this set of differential equations and even if we fit data, we cannot claim that we have explained
the growth of human population.
To show that K (t ) is in fact just S (t ) , let us start with the differential equation for the hyperbolic
growth [see eqn (2)]:
dS (t )
= kS 2 (t ) . (51)
dt
It is the same equation as eqn (2) but it is now presented in s slightly different form. Let us now
replace k by
k c a (b 1) . (52)
where c, a and b are constants. Mathematically, this modification is acceptable because k is a constant
and we can always replace a constant by any combination of constants. Normally, we would not do it.
We do it here to show that the eqns (47) and (48) represent a complicated representation of the eqn (51).
However, we shall show that these equations represent also a corrupted form of the eqn (51).
Equation (51) can now be expressed as
dS (t )
= a[bS (t ) S (t )]S (t ) . (53)
dt
This equation is already almost the same as the eqn (47). But now let us corrupt this equation. Let
us replace one S (t ) in the eqn (53) by K (t ) , while keeping the other S (t ) unchanged. So now we have
two equations:
257
__________________________________________________________________________________
dS (t )
= a[bK (t ) S (t )]S (t ) , (54)
dt
K (t ) = S (t ) . (55)
If K (t ) = S (t ) then of course:
dK (t ) dS (t )
= . (56)
dt dt
However, according to the eqns (51), (52) and (55), and supported by the selective treatment of S (t )
, we have
dS (t )
= kS =
2
=
(t ) cS 2
(t ) cS (t ) K (t ) . (57)
dt
So finally, we now have
dS (t )
= a[bK (t ) S (t )]S (t ) , (58)
dt
dK (t )
= cS (t ) K (t ) . (59)
dt
These two equations are precisely the same as the eqns (47) and (48), and functionally the same as
the eqn (51). However, now we have three constants, a, b and c, rather than just one constant k. We also
have one S (t ) disguised as K (t ) , while the other S (t ) retains its identity. The variable K (t ) is just the
size of the population. It has nothing to do with technology, knowledge or carrying capacity.
Korotayevs differential equations do not explain the mechanism of growth. They only describe the
growth of human population using the well-known mathematical differential equation for the hyperbolic
growth. They do not explain why the growth of population was hyperbolic.
We have repeated the calculations of Korotayev (2005) using his eqns (47) and (48) and his step-
by-step procedure defined by eqns (49) and (50). Results are presented in Figure 4. They show that
K (t ) is precisely the same as S (t ) , K (t ) S (t ) . The two distributions are indistinguishable. K (t ) is
not technology, knowledge or carrying capacity but the size of the hyperbolically increasing quantity,
such as population or the GDP.
Korotayev accepts now that he made a mistake: I agree with what you wrote. (Korotayev, 2015).
However, his model and his calculations have been published in a peer-reviewed journal and as far as
we can tell they were never corrected.
This earlier attempt by Korotayev (2005) was followed by a new approach designed to link the
growth of population with economic growth (Korotayev & Malkov, 2012; Korotayev, Malkov &
Khaltourina, 2006a):
dS (t )
= aq (t ) S (t ) , (60)
dt
dq (t )
= bq (t ) S (t ) , (61)
dt
where S (t ) is again the size of human population, a and b are adjustable constant and q (t ) is claimed
to be, again for no convincing reason, the surplus of the GDP per capita.
If we compare eqns (51) and (60), we can see that if the eqn (60) is supposed to describe hyperbolic
growth of population or the GDP, then q (t ) cannot be anything else but S (t ) , the size of the population
258
__________________________________________________________________________________
or the GDP. The eqn (60) is the same as the eqn (51) except that, for no good reason, one S (t ) is now
replaced by q (t ) . However, this also means that a = b and indeed the authors of these two equations
have determined that a = 1.04b , which is as good as a = b . The two equations are identical. They are
not two different equations but the same equation repeated twice, the same equations as eqn (51) but
now one S (t ) is again disguised, this time as q (t ) , which for absolutely no convincing reason is called
the surplus of the GDP per capita.
Figure 4. Results of calculations carried out using eqns (47) and (48) and the step-by-step procedure
defined by eqns (49) and (50). They confirm that K (t ) S (t ) . Eqns (47) and (48) represent a camouflaged
eqn (51), which describes hyperbolic growth. The data represent the average values of the size of the world
population calculated using the compilations of Manning (2008) and of the US Census Bureau (2016).
We can replace S by any letter in the alphabet. We can call the replaced letter anything we want but
in this context, it is nothing else but the size of population or the GDP or the size of any other
hyperbolically increasing quantity. We are back to the original habit of corrupting the perfectly good
and legitimate hyperbolic equation, but now we are not representing one of the S (t ) as K (t ) , which
for no good reason was called technology or knowledge. We are now representing one of the S (t ) as
q (t ) , which again for no convincing reason is called the surplus of the GDP per capita.
In the earlier mistake, the hyperbolic differential eqn (51) or (2) was disguised as two distinctly
different equations. Now it is disguised as two similar equations, which are in fact identical. Previously,
the growth of population was supposed to have been explained by technology, knowledge or the
carrying capacity, which was incorrect and misleading, because the so-called technology or knowledge
or the carrying capacity was nothing else but the size of the hyperbolically increasing quantity S (t ) .
Now, the growth of population is supposed to be explained by the surplus of the GDP per capita, which
is again incorrect and misleading because the claimed surplus of the GDP per capita is just S (t ) , which
represents the size of the hyperbolically increasing quantity. They are making the same mistake as
before. They have not introduced any new idea but present the same mistake in a different mathematical
form.
The next step is to make it all even more mysteriously complicated. For obscure reasons, the growth
of human population is now supposed to be described by a set of three differential equations
(Khaltourina & Korotayev, 2007; Korotayev, Malkov & Khaltourina, 2006a, 2006b):
dS
= aqS (1 L ) , (63)
dt
259
__________________________________________________________________________________
dq
= bqS , (64)
dt
dL
= cqL (1 L ) , (65)
dt
where a, b and c are adjustable constants and L is claimed, without any convincing justification, to
represent the fraction of literate population, which implies that 1 L is the fraction of the illiterate
population (Korotayev, 2015). In these equations, the time dependence is not explicitly displayed. So,
S S (t ) , q q(t ) and L L(t )
Again, if the eqn (63) is supposed to describe hyperbolically increasing distribution, such as
population or the GDP, then q (1 L) S . We can replace one S in the hyperbolic differential equation
(51) by whatever we want but functionally it will be still S.
A modified version of the three equations (63)-(65) are equations containing even more, spurious
and meaningless parameters (Korotayev, Malkov & Khaltourina, 2006b):
dS
= aq1 S 2 (1 L ) 3 ,

(66)
dt
dq
= bq4 S 5 , (67)
dt
dL
= cq6 L7 (1 L ) 8 ,

(68)
dt
where i with i = 1 8 are arbitrary adjustable positive constants not necessarily equal to one
(Korotayev, Malkov & Khaltourina, 2006b, p. 73). The interpretation of these additional parameters is
also obscure.
Korotayev and his associates claim that they can generate hyperbolic growth with a transition to a
new type of growth. However, they did not introduce any new concepts, which could justify this claim.
They have just replaced two equations by three and one spurious variable by two. They follow the same
idea as expressed in the eqns (47) and (48). In the original equations, a spurious variable K (t ) , was
introduced which for no good reason was called technology or knowledge or the limit to growth and
which turned out to be just the size of population or some other hyperbolically increasing quantity.
Now, the original two equations are replaced by three because two spurious variables are introduced,
q (t ) and L(t ) , which for no convincing reason are called the surplus of the GDP per capita and the
fraction of literate population, respectively. The method of calculations is also the same, i.e. as outlined
in the eqns (49) and (50).
Whatever is done is hidden in the obscure calculation procedure. As before, one would have to repeat
their calculations to understand better the source of error or maybe to become convinced that whatever
they are doing is correct. However, from a start, there is no convincing justification for claiming that q
represents the surplus of the GDP per capita and that L represents the fraction of literate population,
described also as potential teachers (Korotayev, Malkov & Khaltourina, 2006a, p. 26, 2006b, p. 73).
There is also no convincing justification for claiming that the growth of population should be so vitally
dependent on the surplus of the GDP per capita and on the number of potential teachers.
We could probably invent many other complicated formulae to replace the simple and working eqn
(2) or (51). We could also label the new introduced variables or constants in whatever way we want but
could we claim that we have contributed to a better understanding of the mechanism of hyperbolic
growth?
6. Microscopic growth theory

The concept of Karev (2005a, 2005b, 2010) and Karev and Kareva (2014) is to see human population
(or other biological systems) as being made of individuals, each characterised by a certain, unique
parameter a. In a more general formulation of this theory, this uniquely defining parameter is a multi-
260
__________________________________________________________________________________

dimensional vector a = (a1 , a2 ...an ) made of many characteristic components. In the extreme case,

we could think that the components of the vector a are made of genes or even of the components of
the whole genome. In such a case, the multidimensional vector would be made of 106-109 components
(Karev, 2005b).
This theory is based on the advanced and aesthetically appealing mathematics. We shall explain the
fundamental concepts of this theory. Once the fundamental ideas are understood, it will be easier for
anyone to read the more advanced description presented by Karev (2005a, 2005b, 2010) and Karev and

Kareva (2014). In our discussion, we shall replace vector a by constant a.
Rather than dealing with individuals characterised by the parameter a, it is assumed that the entire
population is made of a certain number of groups of a-clones, each group characterised by the same
parameter a and each group made of n(t , a ) number of members at a given time t. In order to calculate
the growth of population we first calculate the growth of each group of a-clones. The differential
equation describing the growth of a-clones is given by
1 dn(t , a )
= F (t , a ) . (69)
n(t , a ) dt
The function F (t , a ) is called the per capita reproductive rate (Karev & Kareva, 2014, p. 73) but
the well-known and accepted definition of the net reproductive rate is the number of daughters born per
woman in her lifetime. In the same publication, F (t , a ) appears also as ag ( N ) , where in our notation
N represents S (t ) , and g ( N ) is interpreted as some function, chosen depending on the specifics of
the model (Karev & Kareva, 2014., p. 69). Karev agrees (Karev, 2015) that it would be better to call
F (t , a ) simply as a growth factor, which will depend on the model used in the calculations. However,
if we use the concept of the general law of growth (Nielsen, 2016k), then this factor can be identified
simply as the force of growth, which in the microscopic theory can have a variety of representations.
The factor F (t , a ) contains all the information about the mechanism of growth of each group of a-
clones. The microscopic theory does not describe any single mechanism but gives a complete freedom
to explore a variety of options. Each specifically chosen mathematical representation of the factor
F (t , a ) will describe a certain mechanism of growth of each group of a-clones, but the mechanism will
remain unknown until the chosen mathematical description of F (t , a ) is not only convincingly
explained but also justified.
The additional complication in this theory is that the calculated size S (t ) of the population made of
numerous groups of clones will depend on how their growth is combined. To understand the mechanism
of growth of population it is necessary to explain not only the factor F (t , a ) but also to justify a specific
mathematical way of combining the growth of all clones.
The growth rate of population is given by:
1 dS (t )
= E (t ) F (t , a ) , (70)
S (t ) dt
where E (t ) is a function describing the mathematical way of combining the growth of population in all
groups of clones. So now, the description of the mechanism of growth depends not only on F (t , a ) but
also on E (t ) . In order to explain this mechanism, it is not enough to explain and justify the factor
F (t , a ) but also E (t ) . The force of growth is given by the product of F (t , a ) and E (t ) .
The calculation of E (t ) is based on the assumption that the populations of various groups of clones
are distributed along a certain probability density function p (t , a ) defined as:
261
__________________________________________________________________________________
n(t , a )
p (t , a ) = . (71)
S (t )
The function p (t , a ) describes the probability of having n(t , a ) number of individuals characterised
by the unique parameter a, i.e. the probability distribution of the parameter a.
The definition of E (t ) , based on the publication of Hofbauer and Sigmund (1998), is:

E (t ) = ap (t , a )da . (72)
0
To illustrate the application of this theory to the description of the growth of human population we
shall use three models of growth presented by Karev (2005a) leading to three solutions.
6.1. Solution 1
This solution is based on the assumption that F (t , a ) = a . Consequently,
1 dn(t , a )
=a. (73)
n(t , a ) dt
In this model, it is assumed that each group of a-clones increases exponentially. The growth is
prompted by a constant force generating a constant growth rate. This is the force of unknown nature.
We do not know why this force should be constant. We just assume that it is. Thus, from the very
beginning we cannot explain the mechanism of growth. Whatever we shall calculate will not help us to
understand the growth of population. Maybe we shall be able to fit the data but we already know that
the data can be fitted well (Nielsen, 2016a, 2016c) using the simple expression describing hyperbolic
distribution [see eqn (1)]. The approach proposed by the microscopic theory will offer an alternative
description but it is more complicated and there is no clear reason for preferring this approach.
The growth of human population as a whole is given now by:
1 dS (t )
= E (t )a . (74)
S (t ) dt
Karev (2005a) gives the following expression for E (t )a , determined by his choice to describe
mathematically the probability density function p (t , a ) :
k
E (t )a= + , (75)
s t
where , k and s are adjustable constants (s, k >0, < < ). For t = s , this function escapes
to infinity.
The differential equation for the growth of human population is now given by:
1 dS (t ) k
= + . (76)
S (t ) dt s t
The right-hand side of this equation is again the force of growth of unknown origin and it is even
less acceptable than the constant force because it is more complicated. If we had reservation about using
a constant force of unknown origin to describe the growth of a group of clones [eqn (73)] our reservation
is now even stronger because the force describing the growth of population is significantly more
262
__________________________________________________________________________________
complicated and also unexplained. We can see that explaining the mechanism of growth is becoming
progressively more difficult. We might only hope that perhaps our formula will give a better description
of data but we shall soon see that it does not.
The solution of the eqn (76) presented by Karev (2005a) is
et
S (t ) = S0 , for t < s , (77)
(1 t / s ) k
which is the exponentially modulated hyperbolic-like growth because it increases to infinity when
t = s . It is not clear why we should want to use this distribution when we already have a simpler
distribution given by the eqn (1) fitting the population data.
If = 0 , then
1
S (t ) = S 0 for t < s . (78)
(1 t / s ) k
The size of population approaches singularity when time t approaches the parameter s. For k = 1 it
is the first order hyperbolic growth given by the eqn (1).
We can explain this formula but we cannot explain the mechanism of growth. We cannot explain
why the growth should be expected to behave in this particular way
6.2. Solution 2
Solution 2 is also based on the assumption of an exponential growth of each group of a-clones but
now a different mathematical description is used for the probability density function p (t , a ) , which
gives different expression for E (t )a used in the eqn (74):
1 c
E (=
t )a + . (79)
s t 1 e c ( s t )
E (t )a escapes to infinity when t = s .

The differential equation for the growth of human population is now given by:
1 dS (t ) 1 c
= + , (80)
S (t ) dt s t 1 e c ( s t )
and its solution by:
1 ec ( t s )
S (t ) = S0 . (81)
(1 t s )(1 e sc )
Parameters used by Karev (2005a) are c 0.114 and s = 2026. The corresponding product cs is
large and the second term in the denominator can be neglected. The formula (81) can now be presented
as
1 ec (t s )
S (t ) = S 0 . (82)
(1 t s )
This solution resembles the first-order hyperbolic growth because the denominator is a linear
function of t, and if not for the function appearing in the numerator, the growth of the population would
escape to infinity at t = s . However, when t = s , the numerator is also zero. Close examination of
263
__________________________________________________________________________________
the eqn (82) shows that when t approaches s, this fraction approaches a constant value, which depends
on parameters s and c. Furthermore, calculations show that for t < s , S (t ) increases approximately
hyperbolically but for t > s , it increases approximately exponentially. Thus, the Solution 2 can be
seen as being made of two parts: a hyperbolic growth to t s and an exponential growth from t s
with an instantaneous discontinuity at precisely t = s .
Mathematically, this formula is interesting because it shows that by assuming a certain force of
growth it might be possible to generate a trajectory, which would, at a certain stage, change from
hyperbolic to a different type of growth. If we could explain the nature of this peculiar force and if we
could reproduce data, we would make a huge progress in the understanding of the mechanism of growth.
However, in this particular case we have no clue about the nature of this peculiar force and, as we shall
soon see, the formula given by the eqn (82) does not fit the data.
6.3. Solution 3
Solution 3 is based on the assumption that F (t , a ) , which in Solutions 1 and 2 was constant, is now
represented by the modified logistic growth rate (Gilpin & Ayala, 1973).
S (t ) k
F (t ,=
a ) a 1 . (83)
K
Again, we do not know the nature of this force.

Under this assumption, the growth of each group of a-clones is given by
1 dn(t , a ) S (t ) k
= a 1 , (84)
n(t , a ) dt K
=
where k const > 0 and K is the limit to growth.
Unless k = 1 , the driving force of growth for each group of clones decreases non-linearly with the
size S (t ) of the whole population. The growth of each group of clones is no longer defined by the
parameter a alone, which represents exclusive characteristics of any particular group of clones, but it
also depends on the size of the whole population. The growth of each group of clones is somehow
coupled to the growth of other clones.
The differential equation for the whole population is now given by:
1 dS (t ) S (t ) k
= E (t )a 1 . (85)
S (t ) dt K
Karev (2005a) uses the following expression for E (t )a :
1 1
=
E (t )a + c[ s p ( t ) ]
, (86)
s p (t ) 1 e
where p (t ) is a solution of Cauchy problem:
k
dp(t ) s 1 exp{c[ p(t ) s ]}
= 1 S 0 . (87)
dt s p(t ) 1 Ke sc
So, now, the differential equation describing the growth of human population is given by:
264
__________________________________________________________________________________
1 dS (t ) 1 S (t )
k
1
= s p (t ) + c s p ( t )]
1 , (88)
S (t ) dt 1 e [ K
and the size of population by
s 1 exp{c[ p (t ) s ]}
S (t ) = S 0 . (89)
s p (t ) 1 e sc
The description of the growth of human population is now significantly more complicated. Solutions
given by eqns (78) and (82) were relatively simple because they were based on the assumption of the
simplest type of growth of the individual groups of clones, growth of each clone prompted by a constant
force. Even though we were not able to explain the mechanism of growth of the entire population made
of various groups of clones we could at least explain the mathematical formulae describing growth.
However, in the case of the growth described by the eqn (89) we cannot even understand this formula
let alone to understand the mechanism of growth of the entire population. We do not understand why
the growth of human population should follow this particular trajectory. Even if we could fit the theory
to data precisely and over the entire range of time we would be still unable to explain the mechanism
of growth.
6.4. Comparing theory with data
Solutions 1-3 are shown in Figures 5 - 7. They are compared with data coming from a wide range of
sources compiled by Manning (2008) and by the US Census Bureau (2016).
In Figure 5 we show the reciprocal values of data and the reciprocal values of Solutions 1-3. The
advantage of using this display is that the decreasing linear trends identify uniquely hyperbolic
distributions (Nielsen, 2014).
The common feature of all these solutions is that over the nearly entire range of time during the AD
era they all follow hyperbolic trajectories. However, they reproduce data over a strongly limited range
of time. Consequently, there is no advantage in using these solutions. The microscopic theory does not
give a better description of data than the simple hyperbolic formula, which can reproduce data over the
past 12,000 years (Nielsen, 2016a). Solutions 2 and 3 are indistinguishable in this display. Solution 1 is
only slightly different. Differences between these three solutions can be observed only towards the end
of the time scale, as shown in Figures 6 and 7.
Figure 5. The decreasing straight lines of reciprocal values identify uniquely hyperbolic growth. Reciprocal
values of solutions 1, 2 and 3 [eqns (78), (82) and (89)] are compared with the reciprocal values of the world
population data as compiled by Manning (2008) and by the US Census Bureau (2016).
265
__________________________________________________________________________________
Figure 6. Solutions 1, 2 and 3 [eqns (78), (82) and (89)] are compared with the world population data
compiled by Manning (2008) and the by US Census Bureau (2016).
Figure 7. Trajectories generated by Karevs Solutions 1, 2 and 3 (Karev, 2005a) in the region where they
start to divert to different trajectories are shown together with population data using the compilations of
Manning (2008) and of the US Census Bureau (2016). Solution 1 escapes to infinity. Solution 2 converts to an
exponential growth, while Solution 3 converts into a logistic growth.
While the concept of the microscopic approach to the study of the growth of population is interesting,
it is not only extremely complicated but also it creates serious problems for the explanation of the
266
__________________________________________________________________________________
mechanism of growth. Examples used by Karev (2005a) show that the complicated mathematical
solutions generated by this theory imitate hyperbolic distributions, which can be represented by a much
simpler equation [see eqn (1)]. Furthermore, these solutions reproduce only a very small range of data.
The problem with using this theory to explain the mechanism of growth is well illustrated by
Solutions 1, 2 and 3 given by the eqns (78), (82) and (89) and by the accompanying expressions for
aE (t ) given in the eqns (75), (79) and (86). While we can explain some of these expressions, we cannot
use them to explain the mechanism of growth.
An interesting feature of this exercise is that a single force of growth can describe a trajectory, which
at a certain stage can change from hyperbolic to some other type. If we could find a force that could
reproduce data over the whole range of time and if we could explain the nature of this force, we would
have made a huge progress in explaining the mechanism of growth. However, examples presented by
Karev indicate that finding such a force of growth and explaining its origin is close to impossible.

We have presented here a brief survey of attempts to understand hyperbolic distributions. The
common characteristic of all these attempts is that they are not only complicated but that they are also
unnecessarily complicated because a simple expression given by eqn (1) describes data exceptionally
well (Nielsen, 2016a, 2016c). This simple formula describes not only the growth of population but also
economic growth as expressed by the Gross Domestic Product (Nielsen, 2016b). Furthermore, by using
this simple formula we can also easily describe income per capita and explain its puzzling features
(Nielsen, 2015, 2016g).
Complicated methods used in the interpretations of hyperbolic growth did not yet result in explaining
its mechanism. They also did not produce a better description of data than the descriptions given by the
simple expression represented by the eqn (1).
When mathematical formulations become increasingly complicated it is usually a warning sign that
we are on the wrong track, that we should stop, regroup and look for simpler descriptions and solutions.
A simple formula [eqn (1)] describing population and economic growth suggests that there must be also
a simple explanation of their mechanisms. Such a simpler explanation will be proposed in the next
publication.
References
Aliprantis, C. D., Brown, D. J., & Burkinshaw, O. (1990). Existence and optimality of competitive equilibria. Heidelberg:
Springer-Verlag.
Becker, G. S., Murphy, K. M., & Tamura, R. (1990). Human capital, fertility, and economic growth. Journal of Political
Economy, 98(5), pt. 2, S12-S37.
Ewing, B., Moore, D., Goldfinger, S., Oursler, A., Reed, A., & and Wackernagel, M. (2010). The Ecological Footprint Atlas
2010. Oakland: Global Footprint Network.
475.
0062-7
267
__________________________________________________________________________________

117, 1133-1191. doi. 10.1162/003355302320935007
Galor, O., & Weil, D. N. (1999). From Malthusian stagnation to modern growth. The American Economic Review, 89(2),
150-154.
Galor, O., & Weil, D. N. (2000). Population, technology, and growth: From Malthusian stagnation to the demographic
transition and beyond. The American Economic Review, 90(4), 806-828.
Gilpin, M. E., & Ayala E. J. (1973). Global models of growth and competition. Proc. Nat. Acad. Sci. USA, 70(12), 3590-
3593.
Hofbauer, J. & Sigmund. K. (1998). Evolutionary games and population dynamics. Cambridge, UK: Cambridge University
Press.
Johansen, A., & Sornette, D. (2001). Finite-time singularity in the dynamics of the world population, economic and &
financial indices. Physica A, 294, 465502.
Kapitza, S. P. (1992). A Mathematical Model of World Population Growth (in Russian). Matematicheskoe Modelirovanie, 4,
65-79.
Kapitza, S. P. (1996). The phenomenological theory of world population growth. Physics-Uspekhi, 39(1), 57-71.
Kapitza, S. P. (2006). Global Population Blow-up and After. Hamburg: Global Marshall Plan Initiative.
Karev, G. P. (2005a). Dynamics of inhomogeneous populations and global demography models. http://arxiv.org/ftp/q-
bio/papers/0505/0505039.pdf
Karev, G. P. (2005b). Dynamics of Heterogeneous Populations and Communities and Evolution of Distributions.
http://arxiv.org/abs/q-bio/0507028.pdf
Karev, G. P. (2010). On mathematical theory of selection: Continuous time population dynamics. J. Math. Biol. 60, 107
129. http://arxiv.org/ftp/arxiv/papers/0812/0812.4280.pdf
Karev, G. P. (2015). Private communication
Karev, G. P. & Kareva, I. G. (2014). Replicator Equations and Models of Biological Populations and Communities. Math.
Model. Nat. Phenom., 9(3), 68-95. DOI: 10.1051/mmnp/20149305
Khaltourina D. A. & Korotayev A. V. (2007). A Modified Version of a Compact Mathematical Model of the World System
Economic, Demographic, and Cultural Development. In M. G. Dmitriev, A. P. Petrov, & N. P. Tretyakov (Eds.),
Mathematical Modelling of Social and Economic Dynamics (pp. 274277). Moscow: RUDN.
Korotayev, A. V. & Malkov, A. S. (2012). A Compact Mathematical Model of the World System Economic and
Demographic Growth, 1 CE 1973 CE. http://arxiv.org/ftp/arxiv/papers/1206/1206.0496.pdf
Korotayev, A. (2005). A Compact Macromodel of World System Evolution. Journal of World-Systems Research, 11(11),
379-393.
Korotayev, A. V. (2015). Private communication.
Korotayev, A., Malkov, A. & Khaltourina, D. (2006a). Introduction to Social Macrodynamics: Secular Cycles and
Millennial Trends. Moscow: KomKniga/URSS
Korotayev, A., Malkov, A. & Khaltourina, D. (2006b). Introduction to Social Macrodynamics. Compact Macromodels of
the World System Growth. Moscow: URSS Publishers.
Kremer, M. (1993). Population Growth and Technological Change: One Million B.C. to 1990. The Quarterly Journal of
Economics, 108: 681-716.
Review, 44, 755-777.
439.
F, A. (2010). Historical statistics of the world economy: 1-2008 AD. http://www.ggdc.net/maddison/Historical
http://www.scottmanning.com/content/year-by-year-world-population-estimates/ and references therein.
http://dx.doi.org/10.4236/am.2014.513188
Nielsen, R. W. (2015). Unified Growth Theory contradicted by the GDP/cap data.
Nielsen, R. W. (2016c). Unified Growth Theory contradicted by the mathematical analysis of the historical growth of human
Nielsen, R. W. (2016d). Scientifically unacceptable established knowledge in demography and in economic research.
Turkish Economic Review, 3(1), 16-27.
268
__________________________________________________________________________________
Nielsen, R. W. (2016g). Mathematical analysis of the historical income per capita distributions. Economic Review, 3(2),
300-319.
Nielsen, R. W. (2016h). Industrial Revolution did not boost the economic growth or the growth of population even in the
United Kingdom. Journal of Economic Bibliography,3(4), 434-446.
Nielsen, R. W. (2016j). Mathematical analysis of income per capita in the United Kingdom. Economic Review, 3(4), 453-
463.
Nielsen, R. W. (2016k). The Law of Growth. Journal of Economic and Social Thought, 3(4), 349-357.
therein.
von Hoerner, S. J. (1975). Population Explosion and Interstellar Expansion. Journal of the British Interplanetary Society, 28,
691-712.
(Publishers) Ltd.
269
__________________________________________________________________________________
16
The Law of Growth

By Ron W. NIELSEN
Abstract. A simple law of growth is formulated. It links growth trajectories with the driving force and thus
furnished an easy way to study and interpret the mechanisms of growth. The application of this law is illustrated
by examples.
Keywords. Growth rate, Mechanism of growth, Models of growth
JEL. C02, C20, C50, Y80
1. Introduction
T he aim of this publication is to formulate a simple law of growth, which could be used
to study, interpret and understand any type of growth. This law will link directly the
trajectory of growth with the driving force. The aim here is to facility an easy and
transparent way for studying the mechanism of growth because the mechanism of growth is
defined by the associated driving force.
If we can link the driving force with the growth trajectory we can then easily check our
interpretation of the mechanism of growth. We can use various types of forces to test whether
proposed mechanism is in agreement with the empirical evidence. We can extend our study to
predict growth assuming that the mechanism of growth is going to be unchanged, but we can
also predict growth by assuming a different mechanism of growth. In such a case, we can also
use the general law of growth but with a new, suitably defined driving force.
2. The law of growth

2.1. The definition
The well-known principle in scientific investigations is: Entia non sunt multiplicanda
praeter necessitatem. Before looking for complicated explanations or formulations, it is always
advisable to adopt the simplest possible approach. Complicated explanations might be
impressive and in some cases even unavoidable but the simplest solutions are always more
attractive.
It is well-known and generally accepted that any growth can, and usually is, described by
the growth rate. Once we know the growth rate we can immediately understand whether the
growth is fast or slow. We can even have a certain degree of understanding of its possible
future, whether it can be sustained or not, whether it is too slow and should be speeded up, if
possible, or maybe that it is too fast and should be slowed down.
Thus, for instance, economic growth is routinely described using the percentage of the
annual increase or decline. The growth of human population is also characterised in the same
way.
Australia.
. +61407201175
Published as: Nielsen, R. W. (2016). The law of growth. Journal of Economic and Social Thought, 3(4), 349-357.
270
__________________________________________________________________________________
Often, in order to understand growth, the growth rate is converted into the doubling time.
The equivalent quantity for radioactive isotopes is the half-lifetime calculated using the decay
rate. If we have a radioactive material, which decays within seconds, we do not have to worry
too much about its harmful effects. However, if we have radioactive contamination containing
substantial amounts of radioactive isotopes with half-lifetime of millions of billions of years
we can be sure that we have a serious problem. Likewise, if the doubling time for a spread of
certain infectious diseases is millions of years we do not need to be worried but if the doubling
time is measured in days, then again, we have a serious problem. We do not have to carry out
any laborious calculation. A simple calculation assuming a constant doubling time or a constant
growth rate can lead easily to an approximately correct answer, which in many cased is quite
acceptable.
Even though the simple formula for calculating the doubling time by dividing 70 or 69.3 by
the growth rate expressed in per cent is applicable only to the exponential growth, and should
never be applied to any other type of growth, such calculations are still carried out for other
types of growth because we know that if we calculate the growth rate or the corresponding
doubling time we can have a little better understanding of a given process.
It seems to be obvious that the growth rate reflects the mechanism of growth and that there
must be a close connection between the growth rate and the driving force of growth. The
simplest way of describing this close connection is to assume that the growth rate is directly
proportional to the driving force:
G =F , (1)
where G is the growth rate, F is the driving force, and is a constant, which we could call
the growth promoting factor, or the compliance, because the larger is the parameter , the
faster is the growth.
Growth rate is defined as:
1 dS (t )
G (t ) , (2)
S (t ) dt
where S (t ) is the size of the growing entity, and t is time.

This quantity is sometimes labelled unnecessarily and confusingly as the relative growth
rate to distinguish it from another redundant and confusing term the absolute growth rate,
which describes just the change in the size of the growing entity per unit of time, i.e. dS / dt .
To make it even more confusing, the term absolute growth rate is sometimes replaced by the
growth rate (e.g. Karev & Kareva, 2014) or by the exponential growth rate.
This needless confusion could be easily avoided by leaving the well-known growth rate, as
defined by the eqn (2), alone. It is a widely-used quantity applicable not only to the exponential
growth but also to any other type of growth. All descriptions of growth in terms of per cent of
the increase or in terms of the doubling time use the growth rate defined by the eqn (2), so the
use of the term: the absolute growth rate, for this well-known growth rate represents an
unnecessary and confusing aberration. We should always use the term growth rate only for the
quantity defined by the eqn (2).
If we insist on using dS / dt to describe growth, we should never create confusion by
associating it with the term growth rate but we should simply call it the absolute change. We
do not create science by introducing complicated and confusing terms.
The eqn (1) represents the simplest, general law of growth. There could be many other ways
of linking the growth rate with the driving force but we have assumed the simplest relation.
271
__________________________________________________________________________________
We call the eqn (1) the law of growth rather than the model of growth because this equation
can be used to formulate a variety of models of growth, some of them already well known, but
many of them yet unknown. Rather than using the existing models, such as exponential or
logistic, even if their application could be questionable, we can tailor the models of growth to
the studied processes and by doing so we can then try to explain the mechanisms of growth
described by the relevant driving force.
The eqn (1) can be rewritten as
F = rG , (3)
where r = 1 . This is another representation of the general law of growth.

2.2. Analogies
In the form given by the eqn (3), the law of growth is similar to the Newtons second law
of motion:
F = ma , (4)
where m is the mass of a physical object and a is the acceleration.

Newtons law describes the dynamics of physical objects. If the driving force is zero, the
acceleration is zero, which means that the physical object is either stationary or that it moves
along a straight line with a constant velocity.
The law of growth describes the dynamics of growing entities. If the driving force is zero,
the growth rate is zero and the size of the growing entity remains constant.
In the Newtons law, m is the mass of the physical object. The larger is m the larger force
has to be used to have the same acceleration. The equivalent parameter in the law of growth is
r, which can be interpreted as the resistance to growth. The larger is r the larger must be the
driving force to have the same intensity of growth.
Acceleration is a well-known quantity and because of it, Newtons law can be used easily
to understand the dynamics of physical objects. Growth rate is also a well-known quantity and
because of it, the general law of growth can be also used easily to understand the dynamics of
growing entities.
In its explicit form, Newtons second law of motion can be expressed as
d 2 s (t )
F (t ) = m , (5)
dt 2
where s (t ) is the trajectory of the moving object. The dynamics of the moving object is
explained by linking the trajectory s (t ) with the driving force F (t ) .
Likewise, in its explicit form, the law of growth can be expressed as
1 dS (t )
F (t ) = r . (6)
S (t ) dt
The dynamics of the growing entity (the mechanism of growth) is explained by linking the
size S (t ) of the growing entity with the driving force F (t ) .
For physical objects, the driving force, i.e. the mechanism of motion, is reflected in the
acceleration a (t ) and in the corresponding trajectory s (t ) . If the driving force (the mechanism)
is known, we can use it to calculate the corresponding trajectory of a moving object. However,
272
__________________________________________________________________________________
if the trajectory is known but the driving force (the mechanism) is unknown, we can assume a
driving force (a mechanism) and calculate the corresponding trajectory s (t ) . If our calculations
agree with relevant data, we can then claim that we have explained the mechanism of the
moving object.
For growing entities, the driving force, i.e. the mechanism of growth, is reflected in the
growth rate G (t ) and in the corresponding trajectory describing the size S (t ) of the growing
entity. If the driving force (the mechanism) is known, we can use it to calculate the
corresponding trajectory of a growing entity. However, if the trajectory is known but the
driving force (the mechanism of growth) is unknown, we can assume a driving force (a
mechanism of growth) and calculate the corresponding trajectory S (t ) . If our calculations agree
with relevant data, we can then claim that we have explained the mechanism of growth.
We can calculate the trajectory s (t ) of a physical object directly from the acceleration
without using the Newtons law. However, to understand why a moving object follows a certain
trajectory we have to understand the driving force, and the link between the driving force and
the trajectory is given conveniently by the Newtons law of motion.
Likewise, we can calculate the trajectory S (t ) of the growing entity directly from the growth
rate without using the law of growth. However, to understand why the growth follows a certain
trajectory we have to understand the driving force, and the link between the driving force and
the trajectory is given by the law of growth.
The difference between the Newtons law of motion and the law of growth is that while
Newtons law is a three-dimensional vector, the law of growth is a scalar, which makes the
description of growth much simpler than the description of the dynamics of physical objects.
In Newtons law, mass m represents an intrinsic property of a physical object. For the law
of growth, resistance to growth, r, might have a broader interpretation. It might represent an
intrinsic property of a growing entity but it might also depend on exogenous conditions. In this
respect, there is a close similarity between the law of growth and other similar simple and well-
known laws listed in Table 1.
Table 1. Summary of the analogous laws
Name Law Explanation

Newtons law F = ma F driving force; m mass; a acceleration
Ohms law U = RI U potential; R resistance; I current
HagenPoiseuille law P =rV P pressure difference; r resistance; V
volume velocity
Darcys law P =rVF P pressure gradient; r resistance to flow;
VF volumetric flux
Fouriers law T =rH T temperature gradient; r thermal
resistivity; H heat flux
Law of growth F = rG F driving force; r resistance to growth; G
growth rate
For instance, the law of growth is similar to Ohms law, U = RI , describing the flow of
electricity. The electrical potential, U, plays here the role of the driving force [cf eqn. (3)] and
R is the resistance to flow. The parameter in the law of growth given by the eqn (1) plays
similar role as the conductance, 1/ R , in the Ohms law. Resistance, R, is determined by the
intrinsic property of the conducting material (electrical resistivity) but it also depends on the
geometrical dimensions of the conducting medium (its length and the cross-section area).
273
__________________________________________________________________________________
Furthermore, while resistivity characterises an intrinsic property of the conducting medium, it

also depends on the temperature.
The law of growth is also similar to the HagenPoiseuille law describing the flow of fluids
through cylindrical conduits. In this law, pressure difference plays the role of the driving force.
Resistance to flow depends not only on the intrinsic property of a given liquid (viscosity) but
also on the geometrical dimensions of the cylindrical conduit (its length and its radius).
However, viscosity depends also on the temperature.
The law of HagenPoiseuille is usually expressed using pressure difference and volume
velocity but it can be also presented using pressure gradient and volumetric flux (volume
velocity per unit area). In this form, it resembles the Darcys law describing the flow of fluids
through porous medium where the resistance to flow is given by the ratio of viscosity and
permeability both depending on the temperature.
The law of growth is also similar to the Fouriers law describing the conductive heat transfer,
where heat flux (energy transferred per units of time and area) is given by the product of
conductivity and the temperature gradient, in the same way as the growth rate is given by the
product of and the driving force in the eqn (1). Temperature gradient plays the role of the
driving force while thermal conductivity is equivalent to the parameter . The inverse value
of thermal conductivity is thermal resistivity. This quantity characterises the intrinsic property
of the heat transferring medium but it also depends on the temperature.
3. Examples of applications of the law of growth

We shall now give a few simple examples how the law of growth can be used in the study
of the mechanism of growth. We shall show how we can tailor our interpretations of growth to
understand better its mechanism. We do not have to be restricted to using just a certain, limited
range of models of growth. We can design and use our own models. We can explore a wide
range of mechanisms of growth and check, which of them gives the best description of data. In
general, we might have to solve the relevant differential equations numerically but in many
cases, we might have a convenient analytical solution.
3.1. Exponential growth
We might assume, for instance, that the driving force is constant,
F (t ) = c . (7)
It is the simplest force of growth. By being constant it, obviously, does not depend on time
or on the size of the growing entity. This comment might sound trivial but it is important to
understand that for other types of growth the driving force can depend not only on time but
also on the size of the growing entity, or on the combination of time and size, and that all such
options will describe the multitude of possible models of growth.
If we use this force in the eqn (1) we shall get
1 dS (t )
=k, (8)
S (t ) dt
c .
where k c / r =
The solution of this differential equation is the exponential function,
S (t ) = Cekt (9)
274
__________________________________________________________________________________
where C is related to the constant of integration.

Now, we can understand this growth a little better because we know where it belongs. It
belongs to a specific class or the type of growth, for which the driving force is constant. For
the same intensity c of the driving force, the smaller is the resistance r or the larger is the
compliance (or growth promoting factor) , the larger is the growth rate k and the faster is the
exponential growth.
3.2. The extension of the exponential growth
Let us now use a more general example when the driving force is not constant but depends
on time,
F (t ) = f (t ) . (10)
If we use this force in the eqn (1) we shall have
1 dS (t )
= f (t ) . (11)
S (t ) dt
The solution to this equation is similar to the solution for the eqn (8):
S (t ) = C exp f (t )dt . (12)

Here we have a large variety of models of growth with one of them being the exponential
model of growth characterised by f (t ) = c . We might represent f (t ) by a polynomial function
or by any other function of our choice.
3.3. The logistic model
We might assume that the driving force of growth decreases with the size of the growing
entity. An example could be the growth of a tree. A tree does not grow indefinitely. It might be
growing fast at the beginning but then it reaches a certain average height and does not grow.
Growth of an individual person can be also a good example. Initially the growth is fast but
eventually a given person reaches a certain height and stops growing. In the simplest case, we
might assume that the driving force decreases linearly with the size of a growing entity:
F (t )= a bS (t ) , (13)
where a and b are positive constants.
Using the eqn (1) we then have
1 dS (t )
= [ a bS (t ) ] . (14)
S (t ) dt
This equation can be expressed as
dS (t )
= dt , (15)
[ A BS (t )] S (t )
where A a and B b .
275
__________________________________________________________________________________
The left-hand side of this equation can be easily integrated if we split it into a sum of two
fractions:
B dS (t ) 1 dS (t )
+ =
dt . (16)
A [ A BS (t )] A S (t )
From now on, the integration is easy.
Alternatively, we can solve the eqn (15) by using the general integration formula we have
derived earlier (Nielsen, 2015):
dx 1 v
u v = ln u , (17)
where u = a + bx , v = c + dx and = ad bc .
The solution of the eqn (14) is represented by the sigmoid function:
1
b 1 b
S (t ) = + e at , (18)
a S0 a
where S 0 = S (t = 0) .
We can see that
a
S (t ) K, (19)
b
where K defines the limit of growth.

3.4. The extension of the logistic growth
In the logistic model, the driving force decreases linearly to a certain limit K. However, we
might have many other possibilities. One of them is the modified logistic model introduced by
Gilpin and Ayala (1973). Some of the variations to the logistic growth are shown in Figure 1.
276
__________________________________________________________________________________
Figure 1. Examples of the extensions to the conventional logistic growth.
In Figure 1, the driving force represented by a decreasing straight line to a certain limit K
represents the well-known, conventional logistic model of growth. However, we might have a
force that is initially approximately constant but then changing gradually to a linearly
decreasing. Such a force would describe the initially approximate exponential growth changing
seamlessly into a logistic growth, which would be approaching asymptotically the limit K.
We could also have a force, which could be initially decreasing rapidly with the size of the
growing entity but after a certain time it would start to follow a gently decreasing trajectory. It
would be an approximately fast logistic growth changing seamlessly into a slow logistic
growth.
Another alternative shown in Figure 1 is a force, which is initially increasing with the size
of the growing entity, reaches a certain maximum and then starts to decrease to a certain limit
K. This type of growth could, for instance, follow an approximately pseudo-hyperbolic
trajectory (Nielsen, 2015). However, it would not increase to infinity but it would change
seamlessly to an approximately logistic growth, approaching asymptotically the limit K.
Possibilities are endless and each of them could be tried to fit data and find their best
mathematical representation. However, if we introduce complicated descriptions of the driving
force we might have a problem with explaining why we use a complicated description. For
instance, if we can see that the growth is indeed initially exponential but then gradually levels
off and approaches a certain limit K, we could easily describe such a growth mathematically
by using a constant driving force changing gradually into the linearly decreasing force shown
in Figure 1 but we would still have to explain why the force changed in such a way and why
the growth changed from exponential to logistic.
3.5. Further extensions
Even though the general principle in scientific investigations is to use the simplest
interpretations, in certain cases it might be necessary to try more complicated solutions and the
law of growth offers an easy definition of such more complicated models. The ultimate
extension is to assume that the force of growth depends not only on time but also on the size
of the growing entity. Such an assumption will probably never be used but it shows that we can
have a practically unlimited number of the models of growth.
3.6. The general principle of investigation
Even though the described here general law of growth opens virtually unlimited possibilities
for defining and using a wide variety of models of growth, the general principle of scientific
investigation is to use the simplest descriptions. In the study of the mechanism of growth the
general principle is to use the simplest mechanism of growth as represented by the simplest
driving force.
Using complicated mathematical expressions without understanding why we use them and
without convincingly justifying their use makes absolutely no sense. Even if complicated
expressions lead to a good description of data we have learned nothing about the mechanism
of growth unless we can explain why such complicated mathematical descriptions are
necessary.
The initial and important step in the study of growth is to identify the type of growth. For
instance, if we can show that the growth is not exponential but hyperbolic we can then focus
our attention on a limited range of forces or maybe even on using just an obvious single force
to explain the mechanism of growth. Complicated mathematical descriptions might look
impressive, they might create an aura of science, but simple descriptions are always preferable.
4. Summary and conclusion
277
__________________________________________________________________________________
Using the simplest possible assumption, we have formulated a simple general law of growth.
We have shown that this law is analogous to many other simple but useful laws, one of them
being the Newtons law of motion. Using a few examples, we have shown how this simple law
of growth can be used to define a multiplicity of models of growth, which in turn can be used
to study the mechanism of growth. Even though this general law of growth allows for the
introduction of a wide variety of models of growth, the general recommendation is to use the
simplest descriptions of driving forces to describe and explain the observed phenomena.
References
Gilpin, M. E., & Ayala E. J. (1973). Global models of growth and competition. Proc. Nat. Acad. Sci. USA, 70(12), 3590-
3593.
Nielsen, R. W. (2015). Mathematics of Predicting Growth. Turkish Economic Review, 2(4), 222-238.
278
__________________________________________________________________________________
17
Mechanism of Hyperbolic Growth Explained
By Ron W. NIELSEN
Abstract. Fundamental law of growth is used to explain the mechanism of hyperbolic growth of human population
and of the Gross Domestic Product (GDP). Hyperbolic growth is described by a simple mathematical formula and
the explanation of its mechanism turns out to be also simple. Historical economic growth was prompted by the
familiar net market force, which was on average directly proportional to the existing wealth expressed usually as
the GDP. The larger was the GDP, the stronger was the driving force and the faster was the economic growth. It
is shown that this simple force generates hyperbolic growth. No other force is required. Hyperbolic growth is not
assumed but derived when using this force. Historical growth of population was prompted by the biologically
driven force of procreation, which was on average approximately constant per person. This force includes the
natural, familiar, biologically controlled process of births, aging and dying. Here again, hyperbolic growth is not
assumed but derived when using this force. Explanation of two demographic transitions in the past 12,000 years
in the growth of population and of the currently experienced transition is also proposed. Currently, economic
growth and the growth of population are no longer unconstrained. Other additional forces contribute significantly
to the growth process and the growth is no longer hyperbolic.
Keywords. Hyperbolic growth, Mechanism of growth, Population growth, Economic growth
JEL. A10, A12, A20, B41, C02, C20, C50, Y80
1. Introduction
yperbolic growth gives a remarkably good description of population and economic data (Nielsen,
H 2014, 2016a, 2016b, 2016c). It describes historical growth of the Gross Domestic Product (GDP)
and of population, global and regional and even in individual countries. This conclusion is based
on the analysis of the extensive data published by Maddison (2010). They describe the economic growth
and the growth of population during the AD era, starting from AD 1 and extending to 2008. Hyperbolic
growth describes also remarkably well the growth of global population in the past 12,000 years (Nielsen,
2016a). This analysis is supported by population data coming from a wide range of sources (Biraben,
1980; Clark,1968; Cook,1960; Durand, 1974; Gallant, 1990; Haub, 1995; Livi-Bacci, 1997; Maddison,
2010; McEvedy & Jones, 1978; Taeuber & Taeuber, 1949; Thomlinson, 1975; Trager, 1994, United
Nations, 1973, 1999, 2013).
Hyperbolic growth of population was first noticed by von Foerster, Mora and & Amiot (1960) close
to 60 years ago and it was soon confirmed and accepted by other authors (Kapitza, 2006; Kremer, 1993;
Podlazov, 2002; Shklovskii, 1962, 2002; von Hoerner, 1975). Hyperbolic growth turns out to be
exceptionally stable and generally undisturbed. Many driving forces might be considered as influencing
growth. For the growth of human population, and as pointed out by Kapitza (2006), all these forces can
be arranged in such categories as industrial, economic, cultural, social and biological. However, he also
pointed out that the simple formula describing the growth of the world population suggests that many
of these forces must have been suppressed by the process of averaging (Kapitza 2006, p. 77).
Economic growth is also described by the simple hyperbolic formula and in general it has been also
stable over a long time in the past suggesting a simple explanation of the mechanism of growth and
Australia.
. +61407201175
Published as: Nielsen, R. W. (2016). Mechanism of hyperbolic growth explained. Journal of Economic Library, 3(4), 411-
428.
279
__________________________________________________________________________________
indicating that the growth must have been also controlled by single net force. It is the aim of this
publication to identify these dominating forces of growth and to explain the mechanism of the historical
hyperbolic growth of population and of the GDP.
2. Mechanism of growth
2.1. Mechanism of the historical economic growth
Gross profit may depend on many factors but it obviously depends on the size of investment. Money
makes money. And the money that makes money makes more money (Benjamin Franklin). Economic
growth is directly related to the size of our investments. With the sufficiently high investment, we can
build more retails stores, or larger retail outlets, we can buy more goods for sale, employ more people
in our business, buy more tools and machinery, invest in better equipment to increase production, build
more houses either for sale or for rent, build more factories, improve agriculture, improve our services,
pay for advertising, pay for the transportation and distribution of goods and support all other necessary
activities aimed at generating profit. According to the well-known theory of Cobb and Douglas (1928),
production yield can be described by the following simple equation:
Y = aL K . (1)
where Y is the production yield, a is the so-called total factor productivity, L is labour expressed as
person-hours during a given time, e.g. during one year, K is capital input (the money invested in the
equipment, buildings or anything else to support production), and are constants, + = 1,
0 < < 1 and 0 < < 1 .
In this equation, wealth generates wealth or money makes money not only through the investment
K, which could be passed from one year to another, but also through the ongoing costs of labour.
In essence, therefore, the right-hand side of the eqn (1) represents the investment of a certain amount
of money to produce profit. The left-hand side does not represent the total wealth but the increase in
wealth, which could be the annual increase. This increase is proportional to the money locked as K and
to the annual investment of money expressed as L. We need money to make money. We need wealth to
generate wealth.
In order to explain the mechanism of economic growth we shall look at it from the point of view of
a driving force, because driving force represents the mechanism of growth. For the economic growth,
it is the net market force. We can have many market forces but in order to explain the mechanism of
growth it is best to start with the simplest assumption and make it complicated only if necessary. This
the fundamental principle in scientific research, known as the Occamss razor or the law of parsimony:
Entia non sunt multiplicanda praeter necessitate.
The simplest way to describe mathematically the driving force of economic growth is to assume that
it is directly proportional to the invested wealth. The larger is the circulated wealth, the greater wealth
can be produced.
F = cW , (2)
where W is the total existing wealth and c is a constant.

It is essential to understand that we are dealing here with average quantities. In explaining economic
growth of a country or region or of the world we are not dealing with individual economic units but
with the whole assembly of these units. The eqn (2) describes the average force of economic growth.
The quantity W represents the total wealth of a country, a group of countries or of the whole world,
expressed usually as the GDP and c could represent the average fraction of this wealth used to drive
economic growth. The larger is the already generated wealth, the larger is the driving force of economic
growth when this wealth is invested to produce more wealth. Wealth generates wealth. This principle
and this process appears to be well known and universally accepted. However, this principle has been
never expressed in mathematical form, which could be compared directly with data. It was never used
to describe economic growth trajectories. It was never used to describe and explain the mechanism of
the historical growth of the GDP.
280
__________________________________________________________________________________
In our earlier publication (Nielsen, 2016d), we have formulated a general law of growth:
F = rG , (3)
where G is the growth rate and r is the resistance to growth.

The advantage of using this simple law of growth is that it links the force of growth with trajectories
of a growing quantity. The force of growth represents the mechanism of growth and the law of growth
allows for defining this force, i.e. for defining the expected or postulated mechanism, and to compare
it with data as described by growth trajectories. This simple law allows for a mathematical formulation
of postulated mechanism and for translating this mechanism into growth trajectories, which can be
readily tested by data. Thus, this law allows for testing various mechanisms of growth by data.
The growth rate G is defined as
1 dW
G , (4)
W dt
where t is time.
If we now insert the postulated driving force of economic growth defined by the eqn (2) into the eqn
(3), we shall get the following equation describing economic growth.
1 dW
= kW , (5)
W dt
where k c / r .
We have now linked the driving force with economic growth trajectory. The parameter k is inversely
proportional to the resistance to growth r and could be called the compliance factor or simply the
compliance. In the formulation of the general law of growth (Nielsen, 2016d) we have defined 1/ r as
compliance. However, k differs only by a constant c so it plays the same role as 1/ r . The larger is the
parameter k, the more efficient is the generation of wealth and the faster is the growth of W. We could
easily extend this model by considering that c or r or both of them depend on time, but at this stage it
is preferable to use the simplest possible assumption.
The eqn (5) does not describe the growth of an individual economic unit but the average economic
growth of a country, region or globally. Economic growth of a single unit might be affected by many
random forces but for a large assembly of such units, random forces might be averaging out. If they are
not or if there is some other strong force not included in our simple assumption, then our predictions of
growth will be contradicted by data and we shall have to modify our assumed mechanism of growth.
We can check whether our assumption is correct by comparing the calculated trajectory with data.
The eqn (5) can be solved using the substitution W = Z 1 . Its solution is
1
W= . (6)
C kt
This is hyperbolic growth. Data describing historical economic growth (Maddison, 2010) and their
analysis (Nielsen, 2016b) show that our choice of the driving force was correct and that there is no need
to assume the presence of any other type of forces. An example of comparing calculations with data is
presented in Figure 1.
Data and their analysis show that the historical economic growth was indeed hyperbolic and now
we can understand why. Historical economic growth was prompted by a single dominant force directly
proportional to the existing volume of wealth, expressed usually as the GDP. Hyperbolic economic
growth describes the net historical growth of a large number of economic units. The larger was the
existing wealth of a country or a region or globally, the larger was the driving force of economic growth.
281
__________________________________________________________________________________
Currently, economic growth is no longer hyperbolic. It is no longer controlled by the simple force given
by the eqn (2). Driving forces appear now to be now more complicated.
Figure 1. World economic growth as described by the Gross Domestic Product (Maddison, 2010) compared
with hyperbolic distribution. Single and simple driving force explains the mechanism of growth. This force was
so strong that even the Industrial Revolution had no impact on changing the growth trajectory.
It should be noted that the growth described by compound interest is of a different kind. It is not a
spontaneous and unconstrained growth controlled by the net driving force proportional to the size of
the existing wealth. The force controlling the growth described by compound interest is constrained. It
is dictated by human-imposed regulations. No bank in the world would pay interest increasing in the
direct proportion to the balance of our deposits. For the money deposited in the bank, interest varies
within a small range of values and consequently it is approximately constant. This type of growth is
described by a constant or approximately constant force of growth, which generates exponential growth,
the growth described by compound interest. Likewise, no bank in the world would give a loan with
interest decreasing with the decreasing balance. These two types of transactions are controlled by man-
made regulations. They are not controlled by the assumed by us, and confirmed by data, force describing
the spontaneous and unconstrained historical economic growth. However, it does not mean that the
current economic growth cannot be exponential. It can and it often is because, as indicated by data, the
current economic growth is no longer prompted by the historically prevailing single force.
2.2. Mechanism of the historical growth of population
The most obvious and essential force, which has to be considered to explain the mechanism of the
growth of population is obviously the biologically-controlled or prompted force of procreation, which
is defined here as the difference between biologically controlled birth and death rates. Other forces
might be included, if necessary, but this force is indispensable.
Let us assume that on average, the biologically controlled force of procreation is constant per
person. Biologically controlled birth and death rates may vary over time but we assume that on average
and per person the difference remains the same. This is a very simple assumption but again in science
it is always advisable to use the simplest possible assumptions and make them more complicated only
if necessary. Under this assumption,
F
= c, (7)
S
282
__________________________________________________________________________________
where F is the biologically controlled force of procreation, S is the size of the population and c is certain
average constant. It describes how, on average, each person contributes to the growth of population.
If we use this force in the general law of growth given by the eqn (3) we shall get
cS = rG , (8)
where G is now given by
1 dS
G= , (9)
S dt
which leads to the following differential equation describing the growth of population
1 dS
= kS . (10)
S dt
Solution to this equation is
1
S= . (11)
C kt
It is also a hyperbolic distribution, which gives excellent description of data (Nielsen, 2016a, 2016c).
Example is shown in Figure 2.
Figure 2. Growth of the world population (Maddison, 2010) compared with hyperbolic distribution. Single and
simple driving force explains the mechanism of growth. This force was so strong that even the Industrial
Revolution had no impact on changing the growth trajectory.
The mechanism of historical hyperbolic growth of population is explained as an unconstrained

growth prompted solely by the biologically controlled force of procreation. This force is given by the
average difference between biologically controlled birth and death rates and is assumed to be constant
per person. This simple mechanism explains global and regional historical growth of population
(Nielsen, 2016a, 2016c).
2.3. Mechanism of demographic transitions
283
__________________________________________________________________________________
If we include in our analysis a wider range of data describing the growth of the world population
(Biraben, 1980; Clark,1968; Cook,1960; Durand, 1974; Gallant, 1990; Haub, 1995; Livi-Bacci, 1997;
Maddison, 2010; McEvedy & Jones, 1978; Taeuber & Taeuber, 1949; Thomlinson, 1975; Trager, 1994,
United Nations, 1973, 1999, 2013; US Bureau of Census, 2016) we shall soon discover certain
interesting details showing two demographic transitions in the past and the current ongoing transition
(Nielsen, 2016a).
As we have shown earlier (Nielsen, 2016a), growth of the world population was hyperbolic between
10,000 BC and 500 BC, between AD 500 and 1200, and between AD 1400 and 1950. During these
large sections of time, taking approximately 90% of the past 12,000 years, the mechanism of growth of
population can be explained as being prompted by the simple, biologically controlled, force of
procreation, which was on average constant per person. All other forces, even if present, had no
influence on the growth of global population. They were either too weak or they were averaged out.
The time when the prevailing hyperbolic growth was significantly disturbed in the past 12,000 years
was only between 500 BC and AD 500, between and AD 1200 and 1400 and now after around 1950.
These are the only recorded demographic transitions in the past 12,000 years. The first transition was
from a fast to a slow hyperbolic trajectory. The second transition was from a slow to a slightly faster
trajectory and the current transition is to a yet unknown trajectory.
The first transition appears to coincide with the massive and widespread changes in the style of living
associated with the intensified changes in the political landscape in various parts of the world,
graphically and comprehensively explained by Teeple (2002). It is also probably not without
significance that this transition coincides with the rise and fall of Roman Empire, the longest lasting
political system in history, which by the first century BC ruled already over vast areas of land
surrounding Mare Nostrum (the Mediterranean). After its fast expansion and after subjugating many
independently-living societies under its rule, this powerful and seemingly unconquerable political
structure disintegrated into many fragments of independent countries. However, during that long time,
significant changes in the political landscape were also occurring outside the realm of the Roman
Empire.
Between 10,000 and 500 BC, growth of population is described by a fast-increasing hyperbolic
trajectory, as defined by the parameter k. After the BC/AD transition, the growth was directed to a
significantly slower trajectory characterised now by the parameter k, which was about 6.4 times smaller.
(The resistance to growth was now significantly larger.) Thus, the proposed explanation of the BC/AD
transition is that it was caused by strong endogenous forces of political nature, forces causing the wide-
spread and profound changes in the style of living. During that time, the resistance to growth was
changing and eventually settled along a significantly larger value.
Demographic transition between AD 1200 and 1400 is much easier to explain. During that time,
there was a temporary delay in the growth of human population. When closely inspected, it can be found
that this delay coincides with the most unusual convergence of demographic catastrophes. It appears to
have been caused by a combined impact of five large demographic catastrophes (Nielsen, 2013a):
Mongolian Conquest (1260-1295) with the total estimated death toll of 40 million; Great European
Famine (1315-1318), 7.5 million; the 15-year Famine in China (1333-1348), 9 million; Black Death
(1343-1352), 75 million; and the Fall of Yuan Dynasty (1351-1369), 7.5 million.
During this transition, hyperbolic growth changed to a slightly faster trajectory, characterised by k
only about 30% higher. This is the only available evidence that the growth of human population might
have been affected by demographic catastrophes. However, their combined impact was small. The
transition to a faster trajectory quickly compensated for the loss of time in the growth of population.
This quick process of recovery could be explained by the regenerating impacts of Malthusian positive
checks (Malthus, 1798; Nielsen, 2016f).
Currently, after a minor boosting around 1950, the growth of human population is slowing down.
The possible explanation of the current diversion to a slower trajectory appears to be of endogenous
nature associated with human choices and motivations, voluntary or enforced. While in many countries
there is an increasing tendency to opt for smaller families, in China, small families have been enforced
by legislation. This additional force appears to be the force of preventive checks (Malthus, 1798). They
may have been active in the past but they were too weak to shape the growth trajectories.
So, the three demographic transitions in the past 12,000 years, including the ongoing transition, can
be probably explained by three different forces: political forces active during the first transition, which
284
__________________________________________________________________________________
lasted for about 1000 years; forces of demographic catastrophes, which were active for about 200 years;
and the endogenous forces of personal choices, either voluntary or enforced by law during the current
transition.
It would be difficult to describe mathematically all these complex forces. However, as already
mentioned, the first two transitions were between hyperbolic trajectories characterised by different k
factors. During these transitions, the k factor was changing. During the first transition, k factor
dramatically decreased, which means that the resistance to growth dramatically increased. It increased
by a massive factor of about 6.4. During the second transition, k factor slightly increased. The resistance
to growth decreased by about 30%. The description of the past and present demographic transitions can
be reduced to the description of changes in the compliance factor or in the corresponding resistance to
growth. Resistance to growth was changing and we can study how it was changing. Such a study will
not give a complete mathematical explanation of the mechanism of demographic transitions but will
reduce this explanation to a single parameter: to changes in the compliance factor k or in the
corresponding resistance to growth.
We can study these changes using a slightly modified eqn (8). If we assume that the resistance to
growth was dependent on time (or equivalently that k depended on time), then we shall have the
following equation describing growth trajectories during demographic transitions:
1 dS (t )
= k (t ) S (t ) . (12)
S (t ) dt
Now, for better clarity, we are showing explicitly the dependence on time.
Solution of this equation is
1
S (t ) = k (t )dt . (13)

If we assume that k (t ) is represented by an n-order polynomial,
n
k (t ) = ai t i , (14)
i =0
then
1
n +1
S (t ) = b j t j , (15)
j =0
where b j = a j 1 / j for j > 0 and b0 is the constant of integration.

Even though we cannot describe mathematically the mechanism of growth during the demographic
transitions, we can understand them a little better by studying changes in the growth factor k (t ) , whose
reciprocal values represent resistance to growth. Results are shown in Figure 3. The corresponding
parameters are listed in Table 1. These calculations do not explain why the resistance to growth was
changing (they do not explain the mechanism of the demographic transitions) but at least they are
describing how the resistance to growth (or the compliance factor) was changing.
In the lower section of Figure 3, we show the growth trajectory during the AD era. It is made of two
hyperbolic trajectories, between AD 500 and 1200 and between AD 1400 and 1950. The remaining
segments of time represent demographic transitions described by the reciprocal values of polynomials,
as given by the eqn (15). This section shows also one of the projected trajectories.
In the middle section, we show the overall fit to the data, which is represented by hyperbolic
trajectories between 10,000 BC and 500 BC, between AD 500 and 1200 and between AD 1400 and
285
__________________________________________________________________________________
1950. The remaining segments of time represent demographic transitions described by the reciprocal
values of polynomials [see eqn (15)].
In the top section, we show time dependence of the compliance factor k (t ) , which can be calculated
using the fitted S (t ) . As we can see from the eqn (13)
dZ (t )
k (t ) = , (16)
dt
1
where Z (t ) S (t ) .
Figure 3. Growth of the world population in the past 12,000 including mathematical description of the past two
demographic transitions between hyperbolic trajectories and the ongoing transition to a yet unknown trajectory.
In Figure 3, we show the compliance factor k (t ) only down to 2000 BC. However, this factor was
constant between 10,000 BC and 500 BC but then started to decrease. The compliance was decreasing,
the resistance to growth was increasing and the growth of population was slowing down. Around 80
286
__________________________________________________________________________________
BC, the compliance factor decreased to zero, the resistance to growth increased to infinity and the
growth of population reached its maximum. The compliance factor continued to decrease and the size
of population was decreasing. When the compliance factor reached its minimum, around AD 200, there
was a turning point in the growth of population. The compliance factor was still negative but now it
was increasing. Slowly, the deceleration in the growth of population was decreasing. Around AD 450
the compliance factor reached its second value of zero. The size of the population reached a minimum
value and started to increase. By around AD 500, this demographic transition was over and the growth
of population settled again along an unconstrained hyperbolic trajectory, but now is was a significantly
slower trajectory characterised by a significantly smaller compliance factor or equivalently by the
significantly larger resistance to growth.
Table 1. Parameters describing the growth trajectory of the world population in the past 12,000
years.
Unconstrained, hyperbolic growth Demographic transitions

(~89% of the total combined time) (~11% of the total combined time)
k = const n
k (t ) = a t
i =0
i
i
10,000 BC 500 BC 500 BC AD 500

a = 2.282=
; k 2.210 102 2.347 103 , a1 =
a0 = 2.659 105 ,
= a2 7.479 108
AD 500 1200 AD 1200 1400
a = 6.940=
; k 3.448 103 , a1 2.618 103 ,
a0 = 1.022=
2.198 106=
a2 = , a3 6.068 1010
AD 1400 1950 1950 present

a = 9.123=
; k 4.478 103 , a1 1.891 103 ,
a0 = 1.820 =
a2 4.899 107
=
The onset of the second demographic transition occurred around AD 1200. Again, the compliance
factor started to decrease and the growth of population started to slow down and even briefly decline.
However, the growth quickly recovered and by around AD 1400 this short-lasting transition was over.
Growth of population resumed its spontaneous and preferred hyperbolic trajectory, which was even a
little faster than before, as indicated by the slightly larger k factor.
Around 1950, the compliance factor was boosted but only for a short time. The growth of population
started to be a little faster than before, but very soon this temporary boosting was halted and the growth
of population started to slow down as expressed in the continually decreasing compliance factor k.
3. Characteristic properties of hyperbolic growth

Hyperbolic growth might be more common than we think. In order to understand this type of growth
it is useful to compare it with other processes and particularly with the more familiar exponential
growth.
For the exponential growth, the size added per fixed unit of time is directly proportional to the total
size of the growing entity, e.g. to the size of the population or the GDP (if they are assumed to increase
exponentially). If the total size doubles, then the added size per unit of time also doubles. For the
hyperbolic growth, to size added per fixed unit of time depends quadratically on the total size of the
growing entity. If the size of the growing entity doubles, the added size per fixed unit of time quadruples.
If the size triples, the added size per fixed unit of time increases nine-folds.
For the exponential growth, the doubling time is constant. For the hyperbolic growth, it decreases
linearly with time. As the size of the growing entity increases, the doubling time decreases. Each
287
__________________________________________________________________________________
consecutive doubling time is twice as small as the immediately preceding doubling time. So, for
instance, if at a certain stage of growth, the doubling time is 24 years, then after 24 years it will be
reduced to 14 years, after 14 years to 7 years, and so on. That is why, hyperbolic growth, or any other
type of growth but exponential, should never be characterised by the doubling time. Constant doubling
time applies exclusively to the exponential growth.
For the exponential growth, the total driving force is constant. No matter how large is the growing
entity, the force remains unchanged. Driving force per single unit decreases exponentially. If
exponential growth were to describe economic growth then the driving force per unit of invested wealth,
e.g. the driving force per invested dollar, would decrease exponentially with the size of the investment.
The potential to generate economic growth per dollar would be decreasing exponentially with the size
of the GDP. If exponential growth were to describe the growth of population then the biologically driven
force of procreation (the difference between the biologically generated or controlled birth and death
rates) per person would be decreasing exponentially.
For the hyperbolic growth, the total driving force increases hyperbolically, i.e. in the direct
proportion with the size of the growing entity, which means that the driving force per unit or per element
of the whole assembly of hyperbolically growing entity, e.g. per person or per dollar, is constant. Each
unit, on average and for a large assembly of growing units, contributes equally to support growth. For
the hyperbolic growth, the potential of each invested unit of wealth, e.g. the potential of each dollar to
create more wealth is constant. It does not depend on the size of invested wealth; it does not depend on
the size of the GDP. For the hyperbolic growth of human population, the force of procreation per person
remains constant; it does not decrease with the size of population.
For the hyperbolic growth, each element, each added component, makes on average, a fixed
contribution to the overall driving force. Individual contributions may vary, but on average the
contribution of each component is constant over time. The larger is the size of the growing entity the
larger is the combined force pushing the growth forward. It is the growth that propels itself in a very
specific way. In the unconstrained hyperbolic growth, the growth is propelled by the approximately
equal contribution of all individual members of the growing entity. It is an interesting and distinct
process where growth generates growth in a very specific way, i.e. where the driving force of growth
per person or per unit of the growing entity is constant. In contrast, for the exponential growth, the
combined driving force is constant but the driving force per unit of the growing entity decreases
exponentially.
Now, we can see that there might be more examples of hyperbolic growth. Take, for instance,
technology or knowledge. Knowledge generates knowledge by stimulating new ideas. Technology
generates technology by stimulating new solutions to technological problems. This is the well-known
process, which even a single person can experience. The more we learn, the easier it is to learn more.
The more problems we solve, the easier it is to solve new problems. Ideas create new ideas, solutions
create new solutions, and knowledge creates new knowledge. It is, therefore, not surprising that
knowledge and technological innovations appear to have been increasing hyperbolically (Kurzweil,
2006; Vinge, 1993). There is a close correlation between the growth of population and technology
(Kremer, 1993). The two processes are similar but they are prompted by different kind of forces.
Technology is certainly not prompted by the force of procreation (the biologically prompted sex
drive and the biologically prompted process of aging and dying). The growth of population is obviously
controlled by these processes. It could be also controlled by some additional forces but the historical
growth of population shows that these other forces were either too weak or that they were averaging
out.
Technology is prompted by concepts, solutions and by research activities. Growth of population is
definitely not prompted by technological concepts, solutions and by research activities but by the force
of procreation. Economic growth is similar to the growth of population but it is obvious that economic
growth is not prompted by the biologically controlled force of procreation.
Another example of hyperbolic growth could be the growth of biodiversity. We could expect that
biodiversity should generate greater biodiversity through competition, adaptation and biological
solutions based on life-supporting mutations. We can also expect that the force driving the growth of
biodiversity is proportional to the existing biodiversity. If it is directly proportional, then the growth of
biodiversity is hyperbolic. Even if we consider minor or major extinctions of species one might expect
that over a sufficiently long time the prevailing trend might be hyperbolic. If we think in terms of driving
288
__________________________________________________________________________________
forces, we could probably identify other examples of hyperbolic growth. We can also understand easier
the distinctions between various types of growth.
For processes described by hyperbolic trajectories, each system will be prompted by its own
mechanism reflected in a specific driving force, but each system will be prompted by the same type of
force. In each case, the force per unit of the growing entity will be constant. Hyperbolic similarities and
close correlations between hyperbolic systems should never be interpreted as necessarily reflecting
precisely the same mechanism of growth represented by precisely the same driving force. In general,
each hyperbolic process will be expected to be propelled by a distinctly different force reflecting a
distinctly different mechanism, but all these forces will be of the same type: their intensity will increase
in the direct proportion to the size of the growing entity; their intensity per person, per biological object,
per unit of measurement (such a dollar, for instance) will be always constant during the entire time of
the unconstrained growth.
Hyperbolic growth is characterised by singularity where the growth escapes to infinity at a fixed
time. Such a growth is deemed impossible. However, historical economic growth and historical growth
of populations were hyperbolic so obviously, they were possible. Growth trajectories can change and
there is nothing unusual about that. A new force may be added to the existing force or the previously
active force might be replaced by a new force. In the growth of global population there were only two
instances in the past 12,000 years when a new force of growth was added temporarily to the force of
procreation. First time, this additional force appears to be of political nature changing radically and on
a large scale the style of living. Second time, it was in the form of demographic catastrophes, the only
known case when demographic catastrophes were reflected in the trajectory describing the growth of
population. Currently, there is also a diversion to a new trajectory. The force of procreation continues
to be active but the new and significant force added to the force of procreation appears to be the force
of preventive checks (Malthus, 1798).
It is also absolutely not necessary to imagine that in order to avoid the problem of singularity we
have to find some mathematically-described trajectory, which over a certain time would mimic
hyperbolic growth but at around a certain time would gradually become non-hyperbolic, and that this
unusual and yet unknown mathematical distributions would also reproduce the growth of human
population. It is absolutely not necessary to eliminate the unrealistic demographic explosion from
the model (Karev & Kareva, 2014, p. 76), because it is not at all unusual for a trajectory to remain
undisturbed over a certain time but then to be diverted to a new trajectory. The mechanisms of growth
can change or can be modified by adding new type of force to the already existing force. We do not
have to imagine that we should have a single force, which over a long time would describe hyperbolic
growth and then would also describe a diversion to a new, non-hyperbolic growth. Karev attempted to
find such a force but failed (Karev, 2005). He tried two such forces but they did not explain the
mechanism of growth because they were incomprehensibly complicated (Nielsen, 2016g). They were
also unsuccessful in describing the growth of population. A single and easy to understand force of
procreation results in a far better description of data.
Current growth of population and economic growth is no longer described so consistently by a single
type of force. For instance, economic growth in Greece was logistic over a certain time but then it
changed to a pseudo-hyperbolic growth with singularity in 2017 (Nielsen, 2016h). This fast growth
could not have been supported in any way and it collapsed. The current global economic growth is
exponential (Nielsen, 2016i). Such a growth is insecure because it does not lead to a maximum or to a
safe and sustainable level of the GDP. It continues to grow until it can be no longer supported.
The current global growth of population is less clearly defined and its projections are less certain.
Analysis of the growth rate shows that growth of population may reach a certain maximum but it may
also continue to increase for as long as it can be supported by the availability of natural resources
(Nielsen, 2006).

Historical economic growth and historical growth of population were hyperbolic (Nielsen, 2016a,
2016b, 2016c). We have explained their mechanism by postulating simple forces of growth. Hyperbolic
growth is mathematically simple and its mechanism of growth is also simple.
289
__________________________________________________________________________________
For the economic growth, the mechanism of the historical hyperbolic growth is explained by the net
market force, which on average was directly proportional to the invested wealth usually expressed as
the Gross Domestic Product. For the growth of population, the mechanism of the historical hyperbolic
growth is explained by the biologically prompted force of procreation defined as the difference between
the biologically prompted birth rate and the biologically controlled process of aging and dying. It is
assumed that this force was on average constant per person.
We do not explain the net market force and neither do we explain the biological force of procreation.
We do not dissect these processes, isolate their components, study minute interactions between them
and then put them together to derive the net driving force. We only describe these forces in the simplest
possible way using simple mathematical expressions based on simple and readily acceptable
assumptions. We then use these simplified forces to explain the mechanisms of growth.
This type of approach is common in scientific investigations. For instance, we do not understand the
force of gravity. We do not really know what it is. However, we can represent this force using a simple
mathematical expression (Newton, 1687) and then use it to explain the mechanism of the movement of
celestial bodies, we can land a man on the Moon and bring him back to Earth, explore our solar system,
land our probes on Mars, detect the presence of the invisible matter and in general explain the dynamics
of the Universe.
We do not understand nuclear forces but we can describe them mathematically and use this
description to study, for instance, the mechanism of nuclear reactions and nuclear structure (Nielsen,
2011). Nobody understands quantum mechanics (Feynman, 1967) but this does not stop us from
describing mathematically various quantum phenomena, explain them and even use the acquired
knowledge to try to apply them, for instance, in quantum computing or cryptography. We do not
understand the weak force and yet we can explain the process of radioactive decay and use radioactive
isotopes in many applications, primarily in medicine but also industry and agriculture.
We do not understand why matter reveals itself as mass or energy. We do not understand the intricate
details of this peculiar phenomenon but we can describe it by a simple and well-known equation
(Einstein, 1905a). We can then use this simple equation to calculate how much energy will be released
if a certain amount of mass manifests itself as energy. We can use this knowledge, combined with our
fundamental knowledge of nuclear processes, to explain the mechanism of fusion and fission reactions.
We can then go a step further and construct (unfortunately) a nuclear bomb or more and (maybe
similarly unfortunately) to construct a controversial nuclear reactor to produce energy. However, we
can also explore how this huge amount of energy locked in the mass could be used in a controlled fusion
reaction and maybe at last to construct a clean and practically inexhaustible source of energy. We can
also use this simple mass-energy relation to explain the mechanism of the production of energy in our
Sun and in the distant stars. We do not know everything but what we already know can be useful.
We do not understand why electromagnetic radiation reveals itself as waves or particles, the
property, which turns out to apply not only to electromagnetic radiation but also to all matter, but we
can describe this relationship by simple mathematical expressions (Einstein, 1905b; de Broglie, 1924)
and explain not only why the rainbow looks so nice but also the strange phenomenon of photoelectricity
(Einstein, 1905b). Einstein is well known for his theory of relatively and for his mass-energy equation
but he received his Nobel Prize for explaining photoelectricity, which demonstrates that light can
manifest itself as being made of tiny particle.
We may not know all the details how nature works but we can still explain many phenomena we
observe and even represent our explanations by useful and often simple mathematical expressions. We
might not be able to explain everything. We might not answer every single question but we can still
explain many phenomena in a satisfactory manner and answer many questions. A deeper understanding
might come much later but only if we make sure that our current knowledge is not based on illusions
and impressions but on the methodically checked interpretations of observed phenomena.
The fundamental principle in scientific research is to look for the simplest explanations of observed
phenomena. These few examples from physics show that even complicated processes can be often
represented by simple mathematical descriptions and that the interpretation of their mechanism can be
significantly simplified.
Distributions describing historical growth of population and the historical economic growth look
complicated, so complicated that they are routinely interpreted as being made of two distinctly different
components, slow and fast, stagnant and explosive, each component governed by distinctly different
290
__________________________________________________________________________________
and complicated mechanisms. The illusion is so persuasive that even most prominent researchers are
easily misguided, particularly if the data are not properly analysed or if they are presented in a grossly
distorted way (Ashraf, 2009; Galor, 2005a, 2005b, 2007, 2008a, 2008b, 2008c, 2010, 2011, 2012a,
2012b, 2012c; Galor & Moav, 2002; Snowdon & Galor, 2008).
The first indication that these distributions are not complicated is demonstrated when they are
mathematically analysed. The analysis is trivially simple (Nielsen, 2014) and it shows that these
distributions are hyperbolic (Nielsen, 2016a, 2016b, 2016c). Hyperbolic distributions look complicated
but they are described by an exceptionally simple mathematical formula: a reciprocal of a linear function
containing just two adjustable parameters.
This remarkable simplicity of hyperbolic distributions representing the historical growth of
population and the historical economic growth suggests a simple mechanism of growth. We have now
demonstrated that the mechanism of these two processes is indeed remarkably simple. They were
prompted by the well-known and simple forces.
Data describing the growth of global population allow for a study of growth over an exceptionally
long time. They show that for the most part of the past 12,000 years, growth of global population was
hyperbolic: between 10,000 BC and around 500 BC, between around AD 500 and 1200 and between
around AD 1400 and 1950. The remaining time of the past 12,000 years was taken by demographic
transitions: between around 500 BC and AD 500, between around AD 1200 and 1400, and from around
1950.
We have proposed the explanation of the mechanism of these transitions. The first transition is
explained by the dramatic and wide-spread changes in the style of living associated with significant
changes in the political landscape. The second transition is explained as being caused by the combined
impact of five major demographic catastrophes. This is the only example when demographic
catastrophes appear to have had impact on shaping the population growth trajectory. However, this
impact was insignificant. The slight delay in the growth of population was soon compensated because
the growth of population was diverted to a slightly faster trajectory. We can explain the mechanism of
this quick recovery by the regenerating effects of the Malthusian positive checks. The mechanism of
the ongoing transition is explained by the Malthusian preventive checks.
A partial mathematical explanation of these transitions is by assuming that the growth of human
population was still prompted by the biologically controlled force of procreation but that the resistance
to growth (or equivalently the compliance factor) was changing. This simple assumption does not allow
us to predict growth trajectories during demographic transitions but only to determine how the
resistance to growth (or compliance factor) was changing during each transition.
Currently, neither the growth of population nor the economic growth can be described by the
historically simple driving force. Generally, we have to use different descriptions for each specific case.
For instance, current global economic growth can be described by a relatively simple but non-
hyperbolic trajectory, which is now converging into the exponential growth (Nielsen, 2016i). Economic
growth in Greece was logistic but then it was converted to a fast-increasing pseudo-hyperbolic growth,
which inevitably resulted in the economic collapse because it came too close to the point of singularity
(Nielsen, 2016h). Global growth of population can be described using different trajectories, each
trajectory giving different prediction of growth (Nielsen, 2006).
The general law of growth (Nielsen, 2016d) helps to understand mechanisms of growth because it
links growth trajectories with driving forces, which are usually easier to visualise and to understand.
We have used this general law of growth and the simplest driving forces to explain the mechanism of
the historical growth of population and the historical economic growth.
References
Cobb, C. W. & Douglas, P. H. (1928). A Theory of Production. American Economic Review, 18, 139-165.
291
__________________________________________________________________________________
475.
0062-7
117, 1133-1191. doi. 10.1162/003355302320935007
de Broglie, L. (1924). Recherches sur la thorie des quanta, PhD Thesis. Paris. Masson.
Feynman, R. (1987). The Character of Physical Law. Massachusetts: The MIT Press.
Publishing Company.
Einstein, A. (1905a). Ist die Trgheit eines Krpers von seinem Energieinhalt abhngig? Annalen der Physik, 18, 639643.
Einstein, A. (1905b). ber einen die Erzeugung und Verwandlung des Lichtes betreffenden heuristischen Gesichtspunkt.
Annalen der Physik, 17 (6), 132148.
Kapitza, S. P. (2006). Global Population Blow-up and After, Hamburg: Global Marshall Plan Initiative.
Karev, G. P. (2005). Dynamics of inhomogeneous populations and global demography models. http://arxiv.org/ftp/q-
bio/papers/0505/0505039.pdf
Kurzweil, R. (2005). The Singularity is Near. New York: Viking Adult.
Kremer, M. 1993. Population Growth and Technological Change: One Million B.C. to 1990. The Quarterly Journal of
Economics, 108: 681-716.
Ltd.
Newton, I. (1687). Philosophiae Naturalis Principia Mathematica. London: Benjamin Motte.
Nielsen, R. W. (2011). Nuclear Reactions: Mechanism and Spectroscopy.
http://www.scribd.com/doc/46997082/RonNielsen-NuclearReactions
http://dx.doi.org/10.4236/am.2014.513188
Nielsen, R. W. (2016c). Unified Growth Theory contradicted by the mathematical analysis of the historical growth of human
Nielsen, R. W. (2016d). The Law of Growth. Journal of Economic and Social Thought, 3(4), 349-357.
Nielsen, R. W. (2016e). Industrial Revolution did not boost the economic growth or the growth of population even in the
Nielsen, R. W. (2016f). The dichotomy of Malthusian positive checks: Destruction and even more intensified regeneration.
Nielsen, R. W. (2016g). Interpretations of Hyperbolic Growth. Journal of Economics and Political Economy, 3(4), 594-626.
Nielsen, R. W. (2016h). Early warning signs of the economic crisis in Greece: A warning for other countries and regions.
Journal of Economics and Political Economy, 2(4), 460-466.
Nielsen, R. W. (2016i). The Insecure Future of the World Economic Growth. Journal of Economic and Social Thought, 2(4),
242-255.
292
__________________________________________________________________________________
von Hoerner, S. J. (1975). Population Explosion and Interstellar Expansion. Journal of the British Interplanetary Society, 28,
691-712.
Pub.
therein.
Vinge, V. (1993). The coming technological singularity: How to survive in the post-human era. In Vision21:
Interdisciplinary Science and Engineering in the Era of Cyberspace. (pp. 11-22). NASA. Lewis Research Center.
http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19940022855.pdf
293
__________________________________________________________________________________
18
Population and Economic Growth in Australia:

8,000 BC AD 1700 and Earlier
By Ron W. NIELSEN
Abstract. Data for the number of rock-shelter sites in Australia between 1000 and 10,000 years BP (before present
or approximately between 8,000 BC and AD 1000) are analysed. These data were interpreted by Johnson and
Brook (2011) as representing the growth of population. Their claim of the Mid-Holocene intensification of the
growth of population around 5000 years BP is contradicted by their own data. The puzzle of the Mid-Holocene
turning point has been solved: there was no turning point. Their claim of the Mid-Holocene turning point relies
on a precise position of a single point in the set of the already inaccurate data. Based on the fitted distribution to
the data, the size of human population has been estimated between 8,000 BC and AD 1700 and tentatively
extended down to 60,000 BC. The absolute values of the size of population were determined in relation to
Maddisons data between AD 1 and 1700 (Maddison, 2010). Data of Maddison show also that the value of income
per capita was constant below AD 1700. Using this information and the fit to the population data, economic
growth in Australia was estimated down to 8,000 BC and tentatively extended to 60,000 BC.
Keywords. Australia, Population growth, Economic growth, Gross Domestic Product, Hyperbolic growth,
Unified Growth Theory, Malthusian stagnation
JEL. A10, A12, C12, C20, Y80
1. Introduction
While data describing economic growth during the AD era are readily available (Maddison, 2001,
2010), similar data for the BC era are hard to find. However, De Long (1998) pointed out that if data
for the growth of population are available, they can assist in calculating the Gross Domestic Product
(GDP) during the BC era by using the income per capita (GDP/cap) values during the AD era because,
income per capita values during the AD era converge quickly to an approximately constant value with
the decreasing time (De Long, 1998; Nielsen, 2015). This property, which is nothing more than the
mathematical property of dividing hyperbolic distributions (Nielsen, 2015), is mistakenly interpreted as
stagnation.
A perfect example of such incorrect interpretation of data is the Unified Growth Theory (Galor,
2005a, 2011). It is a theory based fundamentally on distorted presentation of data and on using
impressions created by such distorted presentations. It is an unreliable and misleading theory. The data
were used in their distorted way but they were never analysed.
When data are presented in a grossly distorted way (Ashraf, 2009; Galor, 2005a, 2005b, 2007, 2008a,
2008b, 2008c, 2010, 2011, 2012a, 2012b, 2012c; Galor & Moav, 2002; Snowdon & Galor, 2008), they
quickly lead to incorrect conclusions. However, when precisely the same data are analysed, they tell a
diametrically different story (Nielsen, 2014, 2015, 2016a, 2016b, 2016c). They show that the Unified
Growth Theory and all other similar interpretations of the economic and population growth are
contradicted by data. In particular, they show that the epoch of Malthusian stagnation did not exist and
that there was no escape from the Malthusian trap because there was no trap. Analysis of data shows
Australia.
. +61407201175
Published as: Nielsen, R. W. (2017). Population and Economic Growth in Australia: 8,000 BC AD 1700 and Extended to
60,000 BC. Journal of Economic and Social Thought, 4(1), 41-54.
http://www.kspjournals.org/index.php/JEST/article/download/1156/1206
294
__________________________________________________________________________________
that the claimed by Galor mysteries of growth did not exist (Nielsen, 2016d, 2016e). It shows that the
origin of the claimed mysteries was the distorted presentation of data. Galor created these mysteries by
distorting data.
The aim of our discussion presented in this publication is to analyse data for the growth of human
population in Australia. As demonstrated earlier (Nielsen, 2016a, 2016f), the growth of population and
economic growth are closely correlated. They follow nearly identical trajectories. Correct
understanding of the growth of population helps also in the correct interpretation of the economic
growth.
2. Data analysis
2.1. Rock shelters
Johnson and Brook (2011) analysed the time-dependent distribution of the number of rock-shelter
sites in Australia, which they interpreted as representing the growth of ancient human population. The
data, as obtained from Brook (2013), are displayed in Figure 1. They are also listed in Table 1. They
represent the relative number of rock shelters because they were normalised to 100 at 10,000 years BP.
Furthermore, it should be pointed out that in their Figure 4 (Johnson & Brook, 2011, p. 3752) data were
shifted by 500 years (Brook, 2013). For instance, the number of rock shelters in 10,000 years BP was
assumed to represent the number of rock shelters in 9,500 years BP. In our analysis, we shall use the
data as supplied by Brook (2013) and as listed in Table 1.
Figure 1. Data (black dots) for the relative number of rock-shelter sites (Brook, 2013; Johnson & Brook,
2011) representing the growth of human population in Australia. The red dot represents a slightly shifted point
at 6,000 years BP illustrating that their claim of the intensification of growth around that time depends entirely
on the precise position of this single point belonging to the already inaccurate set of data.
Table 1. The relative number of rock shelters, N (t ) , in Australia (Brook, 2013).

Year (BP) N(t) Year (BP) N(t) Year (BP) N(t)
1000 1263 4000 432 7000 168
2000 968 5000 405 8000 189
3000 547 6000 184 9000 126
10000 100
These data seem to suggest a slow growth until around 6,000 years BP and a faster growth after that
year. With their arbitrarily displaced presentation of data by 500 years, the apparent change in the
growth pattern could be claimed for 5,000 years BP. Johnson and Brook (2011) concluded that the
growth of human population was slow or negligible before 5000 years BP, and faster since then
(Johnson & Brook 2011, p. 1752). This observation led them inevitably to the question what might have
triggered such a dramatic change in the growth pattern. Whatever the trigger, our results provide new
support for the view, advocated by some Australian archaeologists but contested by others, that
something important happened to the human population of Australia during the Holocene, and that the
Mid-Holocene in particular was a turning point in Australian prehistory (Johnson & Brook 2011, p.
1753).
295
__________________________________________________________________________________
So now, the vital questions are: Is their conclusion acceptable? Was there or was there not a
significant change in the growth pattern of human population in Australia in the distant past? Was there
really a turning point in the Australian prehistory? Did something important happen during the
Holocene that affected dramatically the growth of population, and consequently also the economic
growth?
If there was a change, we have the research field wide open and we can look for answers? However,
if the interpretation of data was in some way incorrect and if there was no change, we will have saved
a great deal of time, effort and financial resources by not pursuing the suggested line of investigation.
We can then divert our efforts into more productive channels.
Before we go any further we should notice that this claim of a sudden intensification of growth
around 5,000 years BP (or rather around 6,000 years BP if we plot the data correctly without shifting
them by 500 years) depends entirely on the precise position of just a single point at 6,000 years BP in
the already inaccurate set of data. If this point is shifted only slightly up, as shown in Figure 1, the
claim of the intensification is not justified because the data follow then an approximately monotonically
increasing distribution. There is no justification for claiming the intensification of growth around 5,000
years BP or around 6,000 years BP. We could terminate our discussion at this stage and conclude that
the data give no support to the claim of the intensification of growth. However, data for the growth of
human population during the BC era are so rare that, if they become available, it is interesting to analyse
them to gain perhaps new information on a related topic.
In order to understand data, it is useful to look at them from a new angle. For instance,
semilogarithmic display of data is useful because it identifies easily exponential growth. If data follow
approximately a straight line, then the growth is approximately exponential. Data analysed by Johnson
and Brook (2011) are presented in Figure 2 using logarithmic scale for the vertical axis.
Figure 2. The number of rock-shelter sites N (t ) shown in Figure 1 is now plotted using semilogarithmic
display. The data follow closely exponential distribution. There is no justification for claiming the
intensification of growth around 6,000 years BP
We can now see clearly that the data follow a monotonically increasing trajectory with no sign of
any unusual acceleration or intensification. The two phases of growth, fast and slow, did not exist. There
was no transition from a slow to a fast growth and there was nothing unusual in the growth pattern
around 6,000 years BP [or around 5,000 years BP if we use the arbitrarily shifted data of Johnson and
Brook (2011)]. Trying to explain the unusual change in the number of sites around that time or around
any other time is irrelevant because there is no convincing evidence that there was a change. On the
contrary, in this display, the data follow closely a straight line suggesting exponential growth over the
entire range of time.
Exponential distribution is described by the following equation:
N (t ) = aert (1)
296
__________________________________________________________________________________
where, for the distribution presented in Figure = 2, a 1.114 103 and r = 2.790 104 . The growth
rate r is in fact positive but in this equation, it is expressed as negative because the time is expressed in
years before present. The number of rock shelters was increasing with time.
Another useful way to examine data and to understand their trend is to plot and to analyse their
reciprocal values (Nielsen, 2014). This type of display is shown in Figure 3. In this representation, an
unusual acceleration or intensification in the number of rock shelters would be indicated by a clear
downward change in the growth pattern. In contrast, data show that the trajectory of the reciprocal
values was gradually bending upwards. There is no sign of any intensification of growth claimed by
Johnson and Brook (2011), not only around 6000 years BP (or around 5000 years BP, depending on
how the data are plotted) but also at any time during this section of time. The reciprocal values of data
for the number of rock-shelter sites in Australia decrease monotonically with time indicating a
monotonic increase in the number of rock shelters. The best and the simplest fit to the reciprocal values
of data is by using the second-order polynomial.
Figure 3. Reciprocal values, 1 / N (t ) , of the number of rock-shelter sites in Australia. There is no sign of
any intensification in the number of rock shelters claimed by Johnson and Brook (2011). The best and the
simplest fit to the reciprocal values of data is by the second-order polynomial.
We can now combine our analysis of rock shelters in Australia in one figure. Results are presented
in Figure 4.
Figure 4. Mathematical analysis of the number of rock shelters in Australia. The best description of data is
by using the reciprocal of the second-order polynomial.
Initially, the growth of the number of rock shelters can be described well using exponential function
or the reciprocal of the second-order polynomial but the reciprocal of the second-order polynomial gives
a better overall description of data. This distribution is given by the following equation:
( )
1
N (t ) = a0 + a1t + a2 t 2 (2)
297
__________________________________________________________________________________
a0 4.882 104 ,
where t is the time in years BP, N (t ) is the number of rock-shelter sites,=
=a1 1.861107 and= a2 7.255 1011 .
So now the puzzling conundrum, acknowledged by some Australian archaeologists (Lourandos
1997) but contested by others (Hiscock 2008) has been solved, and the approach is so simple: just a
different way of plotting the same set of data and by carrying a simple mathematical analysis of data.
Nothing important happened to the human population in Australia during the Holocene (Johnson &
Brook, 2011, p. 3753) and there was no turning point in Australian prehistory (Johnson & Brook,
2011, p. 3753), at least no turning point with respect to the number of rock-shelter sites. There was no
trigger and no transition requiring explanation. The number of rock shelters was increasing
monotonically over the whole time. The mechanism of the sudden intensification of growth does not
have to be explained because there was no intensification.
2.2. Growth of population
We can now go a step further and analyse the historical growth of human population in Australia.
To this end, we have to translate the number of rock shelters into the size of human population. We
shall assume that the size of population was proportional to the number of rock shelters. This
approximation works well even if an approximate fixed fraction of the population did not live in rock
shelters. For the calibration purpose, we shall use Maddisons data (Maddison, 2010). They overlap the
data for the rock shelters at 1000 and 2000 years BP, i.e. at approximately AD 1000 and 1, respectively.
The combined data are listed in Table 2 and are also shown in Figure 5 as dots. They extend only to AD
1700 because between AD 1700 and 1800 the steady growth of population was interrupted by the British
colonisation. The population in Australia decreased from the estimated 450,000 in AD 1700 to 334,000
in 1820. From around 1840 it started to increase rapidly reaching the first million in 1856 and two
million in 1877 (Maddison, 2010). This pattern appears to represent the initial decrease in the aboriginal
population followed by the intensified increase in the number of people arriving in Australia.
Table 2. Growth of human population in Australia, 8,000 BC AD 1700.

The size S (t ) is in thousands.
Year S(t) Year S(t) Year S(t)
8000 BC 34 4000 BC 63 AD 1 346
7000 BC 43 3000 BC 139 AD 1000 417
6000 BC 65 2000 BC 149 AD 1500 450
5000 BC 58 1000 BC 188 AD 1600 450
AD 1700 450
Figure 5. Growth of human population in Australia, 8,000 BC AD 1700. The BC years are represented by
the negative numbers. The size of population was increasing monotonically. There was no intensification of
growth at any time. Growth of population is described remarkably well by the reciprocal of the second-order
polynomial [eqn (3)].
298
__________________________________________________________________________________
Population in Australia was increasing monotonically between 8,000 BC and AD 1500. There was
no intensification of growth at any time. The growth is described well by the reciprocal of the second-
order polynomial:
( )
1
S (t ) = b0 + b1t + b2 t 2 (3)
where S (t ) is the size of population and t is the time (negative for the BC era). Parameters reproducing
b0 3.524 103 , b1 =
the growth of population between 8,000 BC and AD 1500 are:= 1.256 106 and
=b2 2.254 1010 .
This formula reproduces the data so well that it can be used to calculate the size of population at any
time between 8,000 BC and AD 1500 or even to extend the estimations to AD 1700 and below 8,000
BC. The calculated values are listed in Tables A1 and A2 in the Appendix. They are close to the
empirical values listed in Table 5. There is a certain degree of discrepancy between the predicted values
in AD 1600 and 1700. Maddisons data give 450,000 for these two years while the predicted values are
474,000 and 485,000 respectively.
We can also use the determined parameters to calculate the growth rate, which is given by the
following formula:
1 dS (t ) dZ (t )
R (t ) =
S (t ) , (4)
S (t ) dt dt
1
where Z (t ) = S (t ) . (5)
Explicitly, for the eqn (3), the growth rate
b1 + 2b2t
R(t ) =(b1 + 2b2t ) S (t ) = . (6)
b0 + b1t + b2t 2
Calculated size of human population in Australia (in thousands) and the corresponding growth rate
(in per cent) are shown in Figure 6. The growth rate was increasing steadily but it reached a maximum
around AD 1.
Figure 6. Calculated size of human population in Australia between 60.000 BC and AD 1700 (in thousands)
and the corresponding growth rate (in per cent).
299
__________________________________________________________________________________
We should also notice that the parameter= b2 2.254 1010 is small. Mathematical description of the
growth of human population in Australia is, therefore, similar to the mathematical description of the
historical growth of global and regional populations and to the mathematical description of the historical
economic growth (Nielsen, 2016a, 2016f, 2016g). They are all described well using the first-order
hyperbolic distributions given by the following simple equation:
S (t=) (a kt )1 . (7)
Considering that b2 b1 ,
( )
1
S (t ) = b0 + b1t + b2 t 2 (b0 + b1t ) 1 = (a kt ) 1 . (8)
Distribution, given by the eqn (3) and shown in Figures 5 and 6, is similar to the well-known, and
ever-present hyperbolic distribution, given by the eqn (7), which describes so well economic and
population growth, global and regional (Nielsen, 2016a, 2016f, 2016g), even down to 10,000 BC for
the growth of population. These similarities are shown in Figure 7. The distribution labelled as the
Second-order Hyperbola (the reciprocal of the second-order polynomial) describes the growth of human
population in Australia. It was calculated using the eqn (3) and the empirically determined parameters
b0 , b1 and b2 listed under this equation. The distribution labelled as the First-order Hyperbola (the
reciprocal of the first-order polynomial, i.e. the reciprocal of the linear function) was calculated using
the eqn (7) and parameters a = b0 and k = b1 . The two distributions differ only by the presence (or
absence) of the parameter b2 . For the first-order hyperbolic distribution, b2 = 0 . For the second-order
10
hyperbolic distribution= b2 2.254 10 . Another essential difference is that, for this set of parameters,
the distribution describing the growth of ancient population in Australia does not escape to infinity at a
fixed time.
Figure 7. Characteristic features of the second-order hyperbolic distribution [eqn (3)] describing the
growth of population in Australia are similar to the characteristic features of the first-order hyperbolic
3 6
distribution [eqn (7)]. Parameters used in the calculations are b0= a= 3.524 10 , b1 =k =1.256 10 and
b2 2.254 1010 .
=
Considering the omnipresence of hyperbolic distributions (Nielsen, 2016a, 2016f, 2016g) and that
the growth of population in Australia can be so well described using a similar distribution between
8,000 BC and AD 1500 or even 1700, estimation of the size of the population listed in Table A1 was
extended tentatively down to 60,000 BC. The widely-accepted date for the arrival of humans in
300
__________________________________________________________________________________
Australia is around 40,000 years ago (Hiscock, 2008) but it could have been also as early as 60,000
years ago (Lourandos, 1997).
There is also a close similarity between the growth of population in Australian and the growth rate
calculated using a simpler, first-order hyperbolic distribution. The growth rate for the first-order
hyperbolic distribution given by the eqn (7) is
R(t ) = kS (t ) . (9)
However, considering that for the growth of population in Australia b2 b1 , the corresponding
growth rate, given by the eqn (6)
R (t ) = (b1 + 2b2 t ) S (t ) b1S (t ) = b1 S (t ) , (10)
because b1 < 0 .
3. Economic growth
According to Maddison (2010), income per capita in Australia between AD 1 and 1700 was constant.
Expressed in terms of the 1990 International Geary-Khamis dollars, it was $400. The approximately
constant values of income per capita can be easily explained as simply representing the mathematical
property of dividing hyperbolic (or hyperbolic-like) distributions (Nielsen, 2015). Using the suggestion
of De Long (1998), this property can be used to calculate the past GDP values from the estimated size
of population. Economic growth can be assumed to be directly proportional to the size of the population.
The scaling factor for Australia is $400 (1990 International Geary-Khamis dollars). Thus, for
instance, the estimated size of Australian population around 40,000 BC is 2,000 and, consequently, the
estimated GDP is $800,000. The estimated size of population between 60,000 BC and AD 1700 is listed
in Tables A1 and A2. These values can be used to calculate the size of the GDP. The corresponding
values after AD 1700 are listed by Maddison (2010).
It is obvious that no-one in Australia, or in any other region for that matter, was calculating the
values of the GDP, let alone calculating them in the 1990 International Geary-Khamis dollars in that
distant time. The listed values for Australia and for other regions or countries, published by Maddison
(2010) for such remote time can serve only as a guide for the relative size of the common wealth. Thus,
for instance we cannot claim that the value of the GDP in 40,000 BC in Australia was indeed $800,000
but we can estimate that the common wealth in Australia in AD 1700 was about 250 times larger than
in 40,000 BC and about 20 times larger than in 10,000 BC. Using the listed values and the values
published by Maddison (2010) we can also estimate that the GDP in Australia in AD 2000 was about
5,000 times larger than the common wealth of the aboriginal population around 40,000 BC and about
400 times larger than in 10,000 BC. The estimated growth rate of the GDP below AD 1700, or
equivalently the estimated growth rate of the common wealth in Australia is, of course, given by the
estimated growth rate of population listed in Tables A1 and A2.
Economic growth in Australia was slow but the growth rate was increasing monotonically until
around AD 1, when it started to decrease (see Figure 6). From around that time, the size of the common
wealth, as expressed now in terms of the estimated GDP, continued to increase but at the ever-
decreasing growth rate. Such a pattern could lead either to a maximum or to the levelling off of the size
of the GDP. The use of natural resources by the aboriginal population was exceptionally prudent and
parsimonious. Such economic growth could have been sustained indefinitely.
Even if the growth rate stopped to decrease from AD 1700 and remained constant, the doubling time
for the corresponding exponential growth would have been around 3000 years. The GDP would have
increased from $180 million in AD 1700 to only $360 million in around 4700. There was obviously
much room for improving the living conditions without the excessively rapid economic growth.
The invasion of Australia changed everything and soon the GDP started to increase rapidly. Rather
than doubling in about 3000 years, it doubled in only 135 years soon after AD 1700. By the year 2000,
the GDP in Australia increased to $414,058 million. Measured in the constant currency of the 1990
301
__________________________________________________________________________________
International Geary-Khamis dollars, it was 2300 times larger than in AD 1700. The current growth of
the GDP doubles approximately every 22 years. Such a rapid growth is unsustainable.

We have analysed the time dependence of the relative number of rock shelters in Australia. They
were assumed by Johnson and Brook (2011) to represent the growth of aboriginal population.
We have found that the growth of population can be best described using the reciprocal of the
second-order polynomial. Our analysis shows that within the range of analysable data between 8,000
BC and AD 1700, the generally claimed mythical epoch of the so-called Malthusian stagnation did not
exist even in Australia and even in this distant time when early humans must have encountered
numerous adverse conditions. Growth of population in Australia was increasing monotonically. It was
slow, but definitely not stagnant.
Using the fitted distribution, we have calculated the size of the aboriginal population between 8,000
BC and AD 1700. The calculated values are close to the values determined from the study of the number
of rock shelters. However, calculations based on the fitted curve allow for filling in the gaps between
the data.
We have shown that the reciprocal of the second order polynomial, which reproduces the growth of
population in Australia, is in the same class as the hyperbolic distributions describing global and
regional economic growth and the growth of population (Nielsen, 2016a, 2016f, 2016g). Considering
the common presence of hyperbolic distributions and the excellent fit to the data between 8,000 BC and
AD 1700, we have tentatively extended our estimates of the size of population in Australia down to
60,000 BC.
It should be remembered, however, that the estimated historical size of Australian population is
based on the assumption that it is directly proportional to the relative number of rock shelters. If this
assumption is incorrect, then obviously, the estimated size of the population is also incorrect. However,
this is the simplest assumption and in science simplest assumptions are usually preferable.
Our analysis solves the puzzle of the so called Mid-Holocene turning point. According to Johnson
and Brook (2011), there was a turning point in the growth of human population in Australia around
5,000 years BP. The growth of population was supposed to have been slow or negligible before 5000
years BP, and faster since then (Johnson & Brook 2011, p. 1752). Whatever the trigger, our results
provide new support for the view, advocated by some Australian archaeologists but contested by others,
that something important happened to the human population of Australia during the Holocene, and that
the Mid-Holocene in particular was a turning point in Australian prehistory (Johnson & Brook 2011,
p. 1753).
This puzzle has now been solved: there was no Mid-Holocene turning point in the growth of
aboriginal population in Australia. The number of rock shelter sites and the corresponding size of
population were increasing monotonically between 8,000 BC (approximately 10,000 years BP) and AD
1500 or even 1700. The so-called evidence about the Mid-Holocene turning point is based totally on
the position of just a single point in the distribution of the already imprecise data (see Figure 1). Relying
on just a single point to draw far-reaching conclusions is unacceptable, particularly if, as it is in this
case, the data are already inaccurate. Our analysis of data shows that there is nothing remarkable about
this single point. It is as close to the calculated distributions as all other points (see Figures 2-5).
With the exception of the recent surge, growth of human population in Australia over the past 10,000
years was remarkably stable and was following closely the distribution described by the reciprocal of
the second-order polynomial, which is similar to the commonly observed hyperbolic distributions.
Splitting this monotonically increasing growth of population into two distinct segments, as attempted
by Johnson and Brook (2011), and trying to explain them by assuming different mechanisms of growth
is not only unnecessary but also incorrect. There is nothing to explain about the change in the
mechanism of growth because there was no change. However, the data suggest a remarkable and
perhaps unexpected feature which could be further investigated. Why was the growth of the ancient
human population in Australia so stable, so robust and so resilient to any variable forces over such a
long time of around 10,000 years but maybe even over around 60,000 years?
Historical growth of population and historical economic growth in Australia fit well into the
generally observed pattern of hyperbolic growth (Nielsen, 2016a, 2016f, 2016g; von Foerster, Mora &
302
__________________________________________________________________________________
Amiot, 1960). Many serious mistakes have been made with the interpretation of such distributions and
a good example is the Unified Growth Theory (Galor, 2005a, 2011). These distributions are seen as
being made of two distinctly different components, slow and fast. Sometimes a third component is
inserted between these two. The perceived slow component is then interpreted as stagnation and the
perceived fast component as explosion or takeoff. However, such interpretations are incorrect because
hyperbolic distributions increase monotonically. The two distinct components (or stages of growth, or
regimes of growth) do not exist. Each hyperbolic distribution or hyperbolic-like distribution, as it is in
the case of the growth of ancient population in Australia, has to be interpreted as a whole and the same
mechanism has to be applied to the slow and fast growth.
Similar mistake was made by Johnson and Brook (2011) who claimed the intensification of growth
around 5,000 years BP. They also divided the monotonically increasing distribution into two stages,
slow and fast with an apparent intensification at a certain time. This intensification never happened.
They made the same mistake as it is repeatedly made with the interpretation of the historical growth of
population and the historical economic growth when the apparent but non-existent intensification is
described as takeoff, explosion, sprint or spurt, the features contradicted by the methodical analysis of
data.
It is curious that in many publications excellent data are used but they are never analysed (Ashraf,
2009; Galor, 2005a, 2005b, 2007, 2008a, 2008b, 2008c, 2010, 2011, 2012a, 2012b, 2012c; Galor &
Moav, 2002; Snowdon & Galor, 2008). It is also curious that the mistake of failing to analyse data is
compounded by presenting them in a grossly distorted manner. It is as if data were deliberately
manipulated to support erroneous preconceived ideas. Such an approach to research is scientifically
unacceptable. It cannot lead to reliable conclusions and in these cases, it did not. All these publications
are contradicted by the same data, which in their distorted way were used to promote the erroneous
concepts.
Theories such as, the Unified Growth Theory and the Demographic Transition Theory are
consistently contradicted by data and by their analyses (Biraben, 1980; Clark,1968; Cook,1960; Durand,
1974; Gallant, 1990; Haub, 1995; Kapitza, 2006; Kremer, 1993; Lehmeyer, 2004; Livi-Bacci, 1997;
Maddison, 2001, 2010; Mauritius, 2015; McEvedy & Jones, 1978; Nielsen, 2014, 2015, 2016a, 2016b,
2016c, 2016d, 2016e, 2016f, 2016g, 2016h, 2016i, 2016j, 2016k, 2016l, 2016m; Podlazov, 2002;
Shklovskii, 1962, 2002; Statistics Mauritius, 2014; Statistics Sweden, 1999; Taeuber & Taeuber, 1949;
Thomlinson, 1975; Trager, 1994, United Nations, 1973, 1999, 2013; von Hoerner, 1975, von Foerster,
Mora & Amiot, 1960; Wrigley & Schofield, 1981). There is no gain in continuing to use these theories.
They have to be replaced by new theories which incorporate scientific evidence. In particular, there is
no gain in continuing to use the concepts of the so-called Malthusian stagnation, Malthusian trap, escape
from the Malthusian trap and all other associated concepts because they are contradicted by data and
they do not help to explain the mechanism of growth. These concepts are incorrect and misleading. Any
attempt to explain the mechanism of the past growth of population or the economic growth should be
based on accepting the monotonically increasing hyperbolic distributions
For distributions describing the growth of population and the economic growth, even though the
growth was slow it was not stagnant. Even though over a sufficiently long time the growth becomes
significantly faster, there is no sudden takeoff or explosion. Hyperbolic distributions can be misleading
but their analysis is trivially simple (Nielsen, 2014). Anyone can do it to avoid being misguided by their
deceptive features. Hyperbolic distributions are slow over a long time and fast over a short time but
they increase monotonically and they cannot be divided into distinctly different stages of growth. The
characteristic features of hyperbolic distributions describing the historical economic growth and the
historical growth of population should be correctly recognised and accepted in the demographic and
economic research.
References
Brook, B. W. (2013). Private communication.
303
__________________________________________________________________________________

Publishing Company.
475.
http://dx.doi.org/10.1007/s11698-011-0062-7
117, 1133-1191. http://dx.doi.org/10.1162/003355302320935007
Hiscock, P. (2008). Archaeology of ancient Australia. London, UK: Routledge.
Economics, 108(3), 681716.
Lourandos, H. (1997). Continent of huntergatherers: new perspectives in Australian prehistory. Cambridge, UK:
Ltd.
http://arxiv.org/ftp/arxiv/papers/1511/1511.09323.pdf ; https:/www.yumpu.com/user/notrelluf
Nielsen, R. W. (2016b). The postulate of the three regimes of economic growth contradicted by data. Journal of Economic
Nielsen, R. W. (2016c). Unified Growth Theory contradicted by the absence of takeoffs in the Gross Domestic Product.
Nielsen, R. W. (2016d). Puzzling properties of the historical growth rate of income per capita explained. Journal of
Nielsen, R. W. (2016e). The unresolved mystery of the great divergence is solved. Journal of Economic and Social Thought,
Nielsen, R. W. (2016f). Unified Growth Theory contradicted by the mathematical analysis of the historical growth of human
Nielsen, R. W. (2016g). Growth of the world population in the past 12,000 years and its link to the economic growth.
Nielsen, R. W. (2016h). Demographic Transition Theory and its link to the historical economic growth. Journal of
Economic and Political Economy, 3(1), 32-49. http://www.kspjournals.org/index.php/JEPE/article/view/618/798
304
__________________________________________________________________________________
Nielsen, R. W. (2016i). Puzzling properties of the historical growth rate of income per capita explained. Journal of
Nielsen, R. W. (2016j). Mathematical analysis of the historical income per capita distributions. Economic Review, 3(2), 300-
319. http://www.kspjournals.org/index.php/TER/article/view/766/889
Nielsen, R. W. (2016k). The unresolved mystery of the great divergence is solved. Journal of Economic and Social Thought,
Nielsen, R. W. (2016l). The dichotomy of Malthusian positive checks: Destruction and even more intensified regeneration.
Nielsen, R. W. (2016m). Industrial Revolution did not boost the economic growth or the growth of population even in the
United Kingdom. Journal of Economic Bibliography, Journal of Economic Bibliography, 3(4), 434-446.
Pub.
691-712.
(Publishers) Ltd.
305
__________________________________________________________________________________
Appendix
Table A1: Growth of human population in Australia, 60,000 100 BC.
The size of population, S (t ) , is in thousands. The growth rate, R (t ) , is in
per cent (%).
Year S(t) R(t) Year S(t) R(t)

(BC) (000) (%) (BC) (000) (%)
60000 1 0.0032 3200 102 0.0274
50000 2 0.0038 3100 104 0.0277
40000 2 0.0047 3000 107 0.0280
30000 4 0.0061 2900 110 0.0283
20000 8 0.0086 2800 114 0.0286
15000 14 0.0110 2700 117 0.0289
10000 26 0.0149 2600 120 0.0292
9500 28 0.0155 2500 124 0.0295
9000 30 0.0161 2400 128 0.0298
8500 33 0.0167 2300 131 0.0301
8000 36 0.0174 2200 136 0.0305
7500 39 0.0181 2100 140 0.0308
7000 43 0.0189 2000 144 0.0311
6500 47 0.0197 1900 149 0.0314
6000 52 0.0207 1800 153 0.0317
5800 54 0.0210 1700 158 0.0320
5600 57 0.0214 1600 164 0.0324
5400 59 0.0219 1500 169 0.0327
5200 62 0.0223 1400 175 0.0330
5000 65 0.0227 1300 181 0.0333
4800 68 0.0232 1200 187 0.0336
4600 71 0.0237 1100 193 0.0338
4400 75 0.0241 1000 200 0.0341
4200 78 0.0247 900 207 0.0344
4000 82 0.0252 800 214 0.0346
3900 84 0.0254 700 222 0.0348
3800 87 0.0257 600 229 0.0350
3700 89 0.0260 500 238 0.0352
3600 91 0.0263 400 246 0.0354
3500 94 0.0265 300 255 0.0355
3400 96 0.0268 200 264 0.0356
3300 99 0.0271 100 274 0.0356
To calculate the GDP, expressed in the 1990 International Geary-
Khamis dollars, multiply the size of population, S (t ) , by $400. The
GDP values after AD 1700 are listed by Maddison (2010). Growth rate
is the same for the growth of population and for the economic growth.
306
__________________________________________________________________________________
Table A2: Growth of human population in Australia, AD 1- 1700. The

size of population, S (t ) , is in thousands. The growth rate, R (t ) , is in
per cent (%).
Year S(t) R(t) Year S(t) R(t)

(AD) (000) (%) (AD) (000) (%)
1 284 0.0357 850 382 0.0333
50 289 0.0356 900 388 0.0330
100 294 0.0356 950 395 0.0327
150 299 0.0356 1000 401 0.0323
200 305 0.0355 1050 408 0.0319
250 310 0.0355 1100 414 0.0315
300 316 0.0354 1150 421 0.0310
350 321 0.0353 1200 427 0.0306
400 327 0.0352 1250 434 0.0301
450 333 0.0351 1300 440 0.0295
500 339 0.0349 1350 447 0.0289
550 345 0.0348 1400 453 0.0283
600 351 0.0346 1450 460 0.0277
650 357 0.0344 1500 466 0.0270
700 363 0.0342 1550 472 0.0263
750 369 0.0339 1600 478 0.0256
800 376 0.0336 1650 484 0.0248
1700 490 0.0240
To calculate the GDP, expressed in the 1990 International Geary-
Khamis dollars, multiply the size of population, S (t ) , by $400. The
GDP values after AD 1700 are listed by Maddison (2010). Growth rate
is the same for the growth of population and for the economic growth.
307
__________________________________________________________________________________
19
Puzzling Features of the Historical Income per

Capita Distributions Explained
By Ron W. NIELSEN
Abstract. Distributions describing growth of the Gross Domestic Product per capita (GDP/cap) are puzzling.
They show that income per capita was approximately constant over hundreds of years, maybe even over thousands
of years, but then, as if suddenly it started to increase. The growth was changed apparently rapidly from
approximately horizontal to approximately vertical. We shall show that these puzzling features represent nothing
more than purely mathematical property of dividing two hyperbolic distributions. Historical growth of income per
capita can be explained as having been controlled by the simple and familiar forces of growth.
Keywords. Income per capita, Gross Domestic Product, Growth of population, Hyperbolic growth.
JEL. A10, A12, A20, B41, C02, C12, C20, C50, Y80
1. Introduction
yperbolic distributions appear to be creating significant problem with their interpretation. They
H are routinely seen as being made of two distinctly different components, slow and fast, jointed
perhaps by a transition stage. However, these distributions are easy to understand if they are
represented by their reciprocal values (Nielsen, 2014) because in this representation the confusing
features disappear and hyperbolic distributions are represented by straight lines.
It is always convenient to reduce the analysis of data to a straight line, if possible, for two reasons:
(1) straight lines are easy to understand and (2) any deviation from a straight line can be easily observed.
For the exponential growth, the analysis can be reduced to a straight line by calculating the logarithm
of data. For the hyperbolic growth, a straight line is produced by calculating the reciprocal values of
data. However, for the income per capita, this simple method cannot be applied and we have to use a
different approach. Furthermore, distributions describing income per capita are even more confusing
than hyperbolic distributions because features, which were already difficult to understand for hyperbolic
distributions, are even more confusing.
Incorrect interpretation of the historical GDP/cap data is a serious problem and the prominent
example is the Unified Growth Theory (Galor, 2005a, 2011). Using the reciprocal values of the GDP
data, it has been already demonstrated (Nielsen, 2014) that the fundamental postulates of this theory are
contradicted by empirical evidence. We shall now demonstrate that the same conclusion can be reached
by the analysis of the GDP/cap data coming from precisely the same source as used in developing this
theory.
Unified Growth Theory tries to explain the apparent different stages of growth but we shall
demonstrate that this explanation is grossly incorrect because the GDP/cap data follow a single,
monotonically increasing, trajectory, which should be interpreted as a whole. We shall demonstrate that
the three regimes of growth, postulated in the Unified Growth Theory and generally accepted in other
related publications did not exist and that there was no generally claimed takeoff in the economic growth
at any time.
Griffith University, Environmental Futures Research Institute, Gold Coast Campus, Qld, 4222, Australia.
. +61407201175
Published as: Nielsen, R. W. (2017). Puzzling Features of Income per Capita Explained. Journal of Economics Bibliography.
4(1), 10-24. http://www.kspjournals.org/index.php/JEB/article/download/1155/1216
308
__________________________________________________________________________________
2. Crude representation of data

The GDP/cap distributions are frequently displayed in a grossly simplified way by selecting just four
strategically-located points (Ashref, 2009; Galor, 2005a, 2005b, 2007, 2008a, 2008b, 2008c, 2010,
2011, 2012a, 2012b, 2012c; Galor and Moav, 2002; Snowdon & Galor, 2008) as shown in the top panel
of Figure 1.
Figure 1. Gross Domestic Product (GDP) per capita (Maddison, 2001) as frequently presented in numerous
publications (Ashref, 2009; Galor, 2005a, 2005b, 2007, 2008a, 2008b, 2008c, 2010, 2011, 2012a, 2012b,
2012c; Galor and Moav, 2002; Snowdon & Galor, 2008). Strongly misleading impressions created by such
presentations of data are the basis for promoting erroneous interpretations of the mechanism of economic
growth and the prominent example is the Unified Growth Theory (Galor, 2005a, 2011)
In this figure, we show an example for the world economic growth but similar plots are also used
for regional data. Such displays are strongly suggestive and misleading, and they serve as a perfect
prescription for drawing incorrect conclusions. This is a good example of the unscientific approach to
research and it is hardly surprising that such handling of data leads to incorrect conclusions. Galors
Unified Growth Theory and all other associated publications are not based on science. They are
unreliable and strongly questionable. Indeed, when closely analysed they are found to be repeatedly
contradicted by data (Nielsen, 2014, 2016a, 2016b, 2016c, 2016d, 2016e, 2016f, 2016g, 2016h), even
by the same data, which in their distorted way were used to support these numerous publications.
The GDP/cap distributions are already sufficiently confusing even if all data are plotted (see the
lower panel in Figure 1). They do not have to be distorted to create even greater confusion. They have
to be methodically and carefully analysed. Displays such as shown in Figure 1 are not helpful because
they reinforce incorrect impressions and interpretations.
Impressions can be misleading and every effort should be taken to avoid being guided by their
deception. Science is not based on impressions but on a rigorous analysis of data. Unified Growth
309
__________________________________________________________________________________
Theory (Galor, 2005a, 2011) describes ideas based on impressions created by such displays as shown
in Figure 1 or by quoting certain data without making any effort to analyse them scientifically. In this
theory, many complicated but rather primitive mathematical formulations are presented, but incorrect
concepts remain incorrect even if translated into mathematical formulae.
3. Explaining the GDP/cap ratio

The GDP/cap ratio combines two time-dependent distributions: (1) the time-dependent GDP growth
and (2) the time-dependent population growth. In order to understand the GDP/cap distributions we
have to understand their two components: the growth of the GDP and the growth of population.
Over 50 years ago, von Foerster, Mora and Amiot (1960) demonstrated that the world population
was increasing hyperbolically during the AD era. Recent analysis shows that that the world population
was increasing hyperbolically for thousands of years not only during the AD era but also during the BC
era (Nielsen, 2016i). Hyperbolic growth of population applies not only to the global but also to regional
populations (Nielsen, 2016d). Contrary to the expectation of Malthus (1798), when unchecked,
population does not increase exponentially but hyperbolically. Furthermore, the growth of population
was hardly ever checked. Historical GDP values, global and regional, were also following hyperbolic
distributions (Nielsen, 2014, 2016a, 2016h).
Even though hyperbolic distribution appears to be made of two different components, slow and fast,
joined by a transition component, it has been shown (Nielsen, 2014) that such interpretation is based on
strongly misleading impressions. Reciprocal values of a hyperbolic distribution describing growth
follow a decreasing straight line and it is then obvious that it makes no sense to divide a straight line
into arbitrarily selected sections and claim different mechanisms of growth for each section. It also
makes no sense to look for a point marking a takeoff on such a monotonically decreasing straight line
because a monotonically decreasing straight line remains a monotonically decreasing straight line and
there is no justification in selecting a certain point on such a line and claim that there is a change of
direction at this point because there is no change of direction.
In order to understand the GDP/cap distributions, the first and essential step in the past studies should
have been to understand mathematical properties of their two components (GDP and population). Now
we know that that they follow hyperbolic distributions. Consequently, in order to understand the
historical GDP/cap data we have to understand the mathematical process of dividing two hyperbolic
distributions.
We are going to demonstrate that the characteristic features of the GDP/cap distributions, which
were used in the formulation of the Unified Growth Theory (Galor, 2005a, 2011), represent purely
mathematical properties of dividing two hyperbolic distributions. They do not represent different socio-
economic conditions describing different mechanisms of growth for different perceived sections of
these distributions as claimed erroneously in the Unified Growth Theory.
f (t=) (a kt )1 , (1)
where f (t ) is the size of the growing entity, t is the time, and a and k are positive constants.
A reciprocal of hyperbolic distribution, [ f (t )]1 , is represented by a decreasing straight line:
1
[ f (t )]1 =
a kt . (2)
f (t )
Hyperbolic distributions should not be confused with hyperbolic functions ( sinh(t ) , cosh(t ) , etc).
Furthermore, reciprocal distribution or functions, [ f (t )]1 , should not be confused with inverse
functions, f 1 (t ) . Mathematical symbol for the inverse function, f 1 (t ) , is similar to the mathematical
symbol for the reciprocal function, [ f (t )]1 , but the concepts are different.
310
__________________________________________________________________________________
In the inverse functions, the roles of variables are inversed. In the reciprocal functions, they remain
the same. Thus, for instance, for the distribution given by the equation (1), the aim of using its inverse
function would be to calculate how the time depends on the size of the growing entity. The inverse
function of the eqn (1) is
a 1
f 1 (t =
) , (3)
k kt
where t is now the size of the growing entity and f 1 (t ) is the time. For the reciprocal function given
by the eqn (2), t is still the time as in the eqn (1). From the eqn (3) we can see that when the size of the
growing entity, t, increases to infinity, the time, f 1 (t ) , reaches its terminal value of a / k .
The characteristic feature of hyperbolic distributions is that they increase slowly over a long time
and fast over a short time, escaping to infinity at a certain fixed time ts = a / k , i.e. when the
denominator in the eqn (1) approaches its zero value. However, as we have already pointed out and as
discussed earlier (Nielsen, 2014), it is a mistake to interpret such distributions as being made of two
distinctly different components joined by a transition component. It is one and continuous distribution,
which has to be interpreted as a whole. If such a distribution represents a certain mechanism of growth,
it is the same mechanism for the whole distribution.
Let us now take two, purely mathematical, hyperbolic distributions, f (t ) and g (t ) , and let us divide
them. Results are presented in Figure 2.
Figure 2. Two, mathematically-defined, hyperbolic distributions, f (t ) and g (t ) , and their ratio f (t ) / g (t ) .

The time of the perceived, but non-existent, takeoff is indicated.
Parameters describing hyperbolic distributions displayed in Figure 2 are: a = 4.5 and = k 2.2 103
for f (t ) and a = 7 and= k 3.35 103 for g (t ) . These distributions are purely mathematical entities.
They have nothing to do with the growth of the population or with the economic growth. However, they
satisfy a simple condition: the singularity of the f (t ) distribution occurs earlier than the singularity of
the g (t ) distribution. For the curves displayed in Figure 2 singularities are at ts 2045 for f (t ) and
ts 2090 for g (t ) . The point of singularity for the f (t ) / g (t ) ratio is, of course, at ts 2045 .
311
__________________________________________________________________________________
When the distribution f (t ) is divided by g (t ) they produce a distribution, which resembles closely
a typical GDP/cap distribution (see the lower panel of Figure 1). The characteristic features of this
distribution are a long stage of nearly constant values of the f (t ) / g (t ) ratio followed by a nearly
vertical increase.
It is important to notice that for the ratio of two hyperbolic distributions, the difference between slow
and fast growth is much more clearly pronounced than for the corresponding hyperbolic distributions.
The nearly horizontal part is flatter and the nearly vertical part is even more vertical. That is why, if the
hyperbolic distributions are already so confusing, the distributions representing the ratio of two
hyperbolic distributions are even more confusing and their interpretation is even more difficult. They
have to be analysed with extra care and their analysis cannot be simplified by using their reciprocal
values because the reciprocal of the ratio of two hyperbolic distributions is also a ratio of two hyperbolic
distributions. Their analysis is significantly more difficult than the analysis of hyperbolic distributions.
They represent a well-concealed trap suggesting strongly the existence of two or even three different
components and even the most experienced researcher, who is not familiar with hyperbolic distributions
or who is reluctant to accept them because of their singularity, can be easily misguided.
So we can see now that by dividing two, mathematically defined and monotonically increasing
hyperbolic distributions, which have nothing to do with the economic growth, we have generated the
fundamental features, which inspired the creation of the grossly incorrect Unified Growth Theory
(Galor, 2005a, 2011) propagating such erroneous concepts as the Malthusian Regime represented by
the flat part, Sustained-Growth Regime represented by the steep part, the Post-Malthusian
Regime, represented by the middle part and a takeoff, represented by the apparent but non-existent
fast transition from the flat to the steep growth. All these parts and the takeoffs do not exist because
distributions representing the ratios of monotonically increasing hyperbolic distributions increase also
monotonically. We could devote volumes on discussing the mechanism of growth of these imagined
parts and trying to explain the triggering mechanism of the non-existent takeoffs but our discussions
would have no scientific merit. Unified Growth Theory is made of such unscientific explanations but
we can find them in numerous other publications, all creating the undesirable confusion and all of them
diverting attentions from the correct interpretation of the mechanism of economic and population
growth.
The puzzling and apparently peculiar features observed in the GDP/cap distributions can be
reproduced using purely mathematical, monotonically-increasing, hyperbolic distributions. These
features reflect purely mathematical properties of a single distribution representing the f (t ) / g (t ) ratio.
They do not describe different stages of growth. Furthermore, it is clear that these features cannot be
attributed uniquely to the GDP/cap distributions. The division of two hyperbolic distributions may
represent a certain mechanism of growth but it is still a single mechanism.
We have created an unusual and perhaps puzzling distribution but it would be incorrect to be so
mesmerised by this simple mathematical operation as to propose different regimes of growth for
different perceived parts of the f (t ) / g (t ) ratio. We can see that the features observed for the GDP/cap
distributions can be easily replicated by dividing two mathematically-defined hyperbolic distributions.
It is, therefore, clear that hasty assumptions about different socio-economic conditions for the different
perceived parts of the GDP/cap distributions can be questioned, which means that the whole Unified
Growth Theory based on such assumptions can be not only questioned but indeed shown to be grossly
incorrect and scientifically unacceptable (Nielsen, 2014, 2016a, 2016b, 2016c, 2016d, 2016e, 2016f,
2016g, 2016h). There is no point in presenting elaborate descriptions of different socio-economic
conditions if these descriptions are contradicted by data. Even if the described socio-economic
conditions did exist, they obviously had no impact on shaping economic growth trajectories, at least as
expressed by the GDP or by the GDP/cap values. Such theories, as the Unified Growth Theory, could
be regarded as interesting collections of stories but these stories do not assist in understanding the
mechanism of economic growth.
The next step in explaining the GDP/cap distributions is now to explain why the division of two
hyperbolic distributions generates such a puzzling trajectory, which appears to be made of two distinctly
different components and why these apparently different components are so strongly pronounced.
312
__________________________________________________________________________________
4. Explaining the ratio of hyperbolic distributions

Using the eqns (1) and (2) we can see that the ratio of two hyperbolic distributions can be represented
also in two other ways:
f (t )[ Hyperbolic] [ g (t )]1[ Linear ]

=[ g (t )]1[ Linear ] f (t )[ Hyperbolic] = . (3)
g (t )[ Hyperbolic] [ f (t )]1[ Linear ]
These operations are represented graphically in Figure 3. We can see that all these mathematical
operations create the same distribution representing the ratio f (t ) / g (t ) . It does not matter which
pathway we take results are the same.
Dividing two monotonically-increasing hyperbolic distributions is the same as multiplying
hyperbolic distribution by a decreasing linear function and the same as dividing two decreasing linear
functions. It is all just as simple as that. There are no hidden mysteries that need to be explained by
some kind of complicated theories and mechanisms, but we still want to understand why these simple
operations generate such a peculiar distribution, which appears to be made of two distinctly different
components: horizontal and vertical.
The easiest way to understand the division of hyperbolic distributions is probably by looking at the
middle section of Figure 3. The effect of the multiplication of hyperbolic distribution by the decreasing
linear function is to lift up the left-hand part of the slowly increasing section of hyperbolic distribution
and suppress the right-hand part. However, if f (t ) escapes to infinity earlier than g (t ) , f (t ) will be
escaping to infinity when [ g (t )]1 is still positive. The values of [ g (t )]1 will be small but the
multiplication of the rapidly increasing values of f (t ) by small values of [ g (t )]1 will have no effect
on the escape to infinity. The product of such numbers will be also rapidly escaping to infinity. The
combined effect of such a multiplication of a decreasing straight line by the increasing hyperbolic
distribution is to flatten the slowly increasing section of the hyperbolic distribution without significantly
changing the large values. The initial slow increase is made even slower and the perceived transition to
the steep part is even more pronounced. However, there is no mathematically-defined transition at any
time between these two perceived components.
The ratio of two hyperbolic distributions can be described simply as the linearly-modulated
hyperbolic distribution. Thus, in our example the ratio of f (t ) / g (t ) can be described as the linearly-
1
modulated hyperbolic f (t ) distribution. The linear modulation is done by the linear function [ g (t )]
representing the reciprocal values of the hyperbolic g (t ) distribution.
Likewise, the distribution representing the historical GDP/cap growth can be described as the
linearly-modulated hyperbolic GDP distribution. The linear modulation is done by the linear
distribution representing the reciprocal values of the hyperbolic distribution describing the growth of
human population.
The ratio of two hyperbolic distributions looks as if being made of two different components, slow
and fast, but it is still the same, uninterrupted, monotonically increasing distribution. It is still a single
mathematical distribution. It is the distribution, which is not made of two different sections. It is the
distribution that it is impossible to divide into two distinctly different parts represented by two different
functions. This distribution increases slowly over a long time and fast over a short time but the transition
from the perceived slow to the perceived fast growth occurs over the entire range of time. It is
impossible to determine the time of this perceived transition. It is impossible to determine the time of
the perceived takeoff because the takeoff does not exist even if it appears to exist. The perceived takeoff
is an illusion. There is a slow growth over a long time and a fast growth over a short time but there is
no transition at any time between the slow and the fast growth. The slow and the fast growth are
represented by the same, monotonically increasing distribution, which is not made of distinctly different
components.
Even though the ratio of hyperbolic distributions, f (t ) / g (t ) , looks as if being made of two or three
components (see Figures 2 and 3), even though the distribution represented by this ratio increases slowly
over a long time and fast over a short time, even though it increases to infinity at a fixed time and even
313
__________________________________________________________________________________
though it appears to be characterised by a takeoff at a certain time, it is still just a single, monotonically-
increasing distribution, which is impossible to divide into different components. We have to accept it
and learn to live with it.
Figure 3. Graphic representation of the eqns (3).
Perhaps the easiest way to dispel the strong illusion of the distinctly different components of growth
is to examine the lowest part of Figure 3. It would be obviously unreasonable to claim that each of these
straight lines is made of two or three distinctly different components, because these straight lines are
obviously not made of different components. It would be unreasonable to claim different mechanisms
of growth for various, arbitrarily-selected parts of these straight lines. At which point located on a
straight line one mechanism of growth is supposed to end and a new mechanism to begin? It is
impossible to claim two or three distinctly different sections on the monotonically decreasing straight
lines. There is also obviously no feature on such straight lines that could be claimed as marking a
takeoff.
We can also take a different approach and demonstrate again that the ratio f (t ) / g (t ) represents a
single, monotonically-increasing distribution and that there is no takeoff at any time. This different
approach consists in calculating the gradient and the growth rate of the f (t ) / g (t ) ratio. Results are
presented in Figure 4 around the time of the perceived takeoff, i.e. when the f (t ) / g (t ) reaches the value
of 2 (see Figure 2). For better clarity, results are plotted as a function of the size of the f (t ) / g (t ) ratio.
These calculations show clearly that both the gradient and the growth rate of the hyperbolic ratio
f (t ) / g (t ) increase monotonically. The perceived takeoff never happened. What looks like a takeoff in
314
__________________________________________________________________________________
Figure 2 is in fact just the continuation of the undisturbed and monotonically-increasing distribution
representing the f (t ) / g (t ) ratio. It is impossible to claim different components for any of the
distributions displayed in Figure 4, representing the f (t ) / g (t ) distribution, which in Figure 2 looks very
deceptively as being made of two different components. It is impossible to claim a takeoff for any of
these two distributions. The two components simply do not exist and the takeoff is just an illusion.
Figure 4. The gradient and growth rate of the ratio of hyperbolic distributions f (t ) / g (t ) . The onset of the
perceived takeoff shown in Figure 2 is indicated. This figure shows that the takeoff never happened and that the
distribution representing the ratio f (t ) / g (t ) is not made of different components. It is a single, monotonically-
increasing distribution.
5. Analysis of the historical GDP/cap data

The GDP and population data (Maddison, 2001) [the same data as used but not analysed during the
formulation of the Unified Growth Theory (Galor, 2005a, 2011)] together with their fitted hyperbolic
distributions are shown in Figure 5. Indicated in the figure is the time of the Industrial Revolution 1760-
1840 (Floud & McCloskey, 1994), which is generally claimed as the time of the alleged takeoff in the
economic growth (Galor, 2005a, 2008a, 2011, 2012). Parameters fitting the GDP data are:
= a 1.716 102 and = k 8.671 106 while parameters fitting the population data are a = 8.724 and
3
= k 4.267 10 .
Points of singularity are: ts 1979 for the world GDP and ts 2045 for the population data. The
point of singularity for the world GDP is before the point of singularity for the growth of the world
population. Consequently, the GDP/cap ratio should display the same features as shown in Figure 2 for
the f (t ) / g (t ) ratio and indeed it does.
In Figure 6 we present the data for the GDP/cap and the corresponding fit to the data calculated by
dividing the corresponding hyperbolic distributions shown in Figure 5. The calculated curve and the
data shown in Figure 6 follow a similar distribution as displayed in Figure 2. The characteristic features
of the nearly horizontal growth over a long time and the nearly vertical growth over a short time of the
GDP/cap distribution are nothing more than the mathematical property of dividing two hyperbolic
distributions.
If the point of singularity for the GDP trajectory was located higher than the point of singularity for
the population trajectory, the growth of the GDP/cap would also have remained nearly constant over a
315
__________________________________________________________________________________
long time but it would eventually decrease to zero at the time of the singularity for the growth of
population. Income per capita would not have been increasing with the size of the population. On the
contrary, it would have been decreasing. For hyperbolic distributions, the growth of income per capita
depends on the relative positions of singularities of the two components.
Figure 5. Hyperbolic distributions are compared with the world GDP and population data (Maddison, 2001).
The GDP is expressed in billions of 1990 International Geary-Khamis dollars and the population in billions.
316
__________________________________________________________________________________
According to Galor (2008a, 2012a), the so called Malthusian Regime, represented allegedly by the
nearly constant income per capita, commenced around 100,000 BC. There is, of course, no justification
for this date because the Malthusian Regime did not exist (Nielsen, 2014, 2016a, 2016b, 2016c, 2016d,
2016e, 2016g, 2016h, 201gi; von Foerster, Mora & Amiot, 1960. However, if we wanted to claim a
certain date for this mythical regime, one would imagine that the usually claimed date of 200,000 BC
for the onset of the existence of Homo Sapiens would have been more suitable.
Figure 6. Calculated, linearly-modulated hyperbolic GDP distribution, representing the GDP/cap

ratio, is compared with the world GDP/cap data (Maddison, 2001). The GDP/cap is expressed in the
1990 International Geary-Khamis dollars.
He also claims that Malthusian Regime was terminated in AD 1750 for developed countries and in
1900 for less-developed countries. The Post-Malthusian Regime was supposed to have existed between
1750 and 1870 for developed countries and from 1900 for less-developed countries. The Sustained-
Growth Regime was supposed to have commenced in 1870 and is supposed to continue until the present
time. It is impossible to determine such specific landmarks for the monotonically increasing
distributions. These imagined dates are contradicted by data. There were no takeoffs in the growth of
the GDP, and the historical GDP trajectory cannot be divided into two or three different regimes
(Nielsen, 2014, 2016b, 2016c). We also know that the growth of human population was hyperbolic and
that it was never characterised by a sudden takeoff (Nielsen, 2016d, 2016i). Consequently, even though
the GDP/cap data might be suggesting the existence of different stages of growth governed by different
mechanisms of growth, their scientific analysis clearly demonstrates that different regimes of growth
did not exist. Each historical GDP/cap distribution, global or regional, has to be interpreted as a whole
and the same mechanism has to be applied to the slow and fast growth. Under these conditions, the
interpretation of the mechanism of growth appears to be complicated because we have to use the same
mechanism to explain the slow and fast growth. However, the explanation turns out to be exceptionally
simple (Nielsen, 2016j).
By following our earlier approach, which we used for the division of arbitrary hyperbolic
distributions, we can demonstrate that there was no takeoff in the GDP/cap distribution and that the
317
__________________________________________________________________________________
three regimes of growth did not exist. We shall do this by calculating the gradient and the growth rate
for the calculated GDP/cap trajectory. These calculations are presented in Figures 7 and 8.
A takeoff in the GDP/cap trajectory would be marked by a clear change in the gradient and in the
growth rate around the time of the Industrial Revolution when a transition to a new economic growth
regime was supposed to have happened (Galor, 2005a, 2008a, 2011, 2012a). The shape of the
trajectories describing the gradient and growth rate would have to be distinctly different before and after
the Industrial Revolution. There should be a certain clear discontinuity.
Figure 7. Gradient of the world GDP/cap calculated using the fitted, linearly-modulated hyperbolic distribution
shown in Figure 6. The GDP/cap is expressed in the 1990 International Geary-Khamis dollars. There was no
takeoff at any time and the three regimes of growth postulated by Galor (2005a, 2011) did not exist.
The gradient and the growth rate of the fitted curve increase monotonically confirming that the fitted,
linearly-modulated hyperbolic distribution increases also monotonically. The calculated curve gives
excellent fit to the GDP/cap data and consequently the gradient and the growth rate of the fitted curve
represent also the gradient and the growth rate of the data.
Figures 7 and 8 clearly demonstrate that there is no reason for terminating the alleged Malthusian
Regime around AD 1750 and for starting a new regime because there was no unusual change in the
gradient and in the growth rate of the GDP/cap around that time, but there was also no scientifically-
justified reason for assuming the existence of the Malthusian Regime. There is no reason for terminating
the equally imaginary Post-Malthusian Regime around 1870 and starting the alleged Sustained-Growth
Regime. There is no reason for slicing the monotonically-increasing distributions into three arbitrarily-
selected sections. There is no reason for proposing three regimes of growth governed by distinctly
different mechanism. There is no reason for claiming a takeoff at any time. There has been no
scientifically justified reason for creating the Unified Growth Theory and three is no scientifically
justified reason for adopting such concepts in the interpretations of economic growth and of the growth
of population.
These calculations, supported by data, clearly demonstrate that the Industrial Revolution had no
impact on the economic growth trajectory. Impacts were of different kind but the data show that the
Industrial Revolution did not boost the global economic growth. It did not even boost the economic
growth in Western Europe (Nielsen, 2014), or in any other region (Nielsen, 2016a, 2016g) or even in
318
__________________________________________________________________________________
the United Kingdom (Nielsen, 2016h), the very centre of this revolutions, where its effects on the growth
trajectory should have been most clearly pronounced. There was no impact whatever on the growth
trajectories. Economic growth must have been prompted and controlled by some other force, which was
much stronger than any other forces, including the force of the Industrial Revolution and this force is
discussed in a separate publication (Nielsen, 2016j). Furthermore, Galors three regimes of growth did
not exist.
Figure 8. Growth rate of the GDP/cap calculated using the fitted, linearly-modulated hyperbolic distribution
shown in Figure 6. The GDP/cap is expressed in the 1990 International Geary-Khamis dollars. There was no
takeoff at any time and the three regimes of growth postulated by Galor (2005a, 2011) did not exist.
Fundamental postulates of the Unified Growth Theory (Galor, 2005a, 2011) are contradicted by the
analysis of data, the same data as used but not analysed during the formulation of his theory. Unified
Growth Theory describes and explains phenomena that did not exist and consequently it does not
explain the historical economic growth. It is an incorrect and misleading theory.
The discussion of socio-economic conditions presented by Galor might be interesting for some other
reason but there is a clear evidence in the GDP or GDP/cap data that his discussion has no relevance to
explaining the mechanism of economic growth. His discussed associations and correlations are not just
questionable but plainly incorrect because they are contradicted by data he used but never analysed.
Economic growth was indeed slow over a long time and fast over a short time but it is incorrect to
divide this monotonically increasing distribution into three regimes and claim distinctly different
mechanisms for the arbitrarily selected sections. It is also incorrect to claim that there was a takeoff at
a certain time. The data and their analysis give no scientific basis for such claims.
Historical economic growth has to be explained using a single mechanism. Such a mechanism
should describe the slow and fast growth including the apparent transition. All these parts should be
treated as one. Only then we could claim that we have explained the mechanism of the historical
economic growth.
Dividing the past growth into three different regimes and claiming three different mechanisms is
unsupported by data and it does not explain the mechanism of the historical economic growth. A truly
unified growth theory will have to be based on a single mechanism. Such an explanation is proposed in
a separate publication (Nielsen, 2016j).
319
__________________________________________________________________________________

The aim of our discussion was to explain the puzzling features of the GDP/cap distributions. They
show a slow growth over a long time, followed by a rapid increase. These features create a significant
problem with their interpretations, and the outstanding example of the created confusion is the Unified
Growth Theory (Galor, 2005a, 2011). Our discussion was based on precisely the same data which were
used, but not analysed, during the formulation of this theory. It is both surprising and disappointing that
while using excellent sets of data published by the world-renown economist (Maddison, 2001), Galor
made no attempt to adopt scientific approach to developing his theory.
Historical economic and population growth, global and regional, show a clear preference for
increasing along hyperbolic trajectories (Nielsen, 2014, 2016a, 2016d, 2016g, 2016h, 2016i).
Hyperbolic growth contains singularity, when a growing entity escapes to infinity at a fixed time. We
might think that such a growth is impossible but we have to accept the evidence in data. The past growth
of the GDP and of population were hyperbolic. There is absolutely no problem with accepting
hyperbolic growth for two reasons: (1) hyperbolic growth is obviously possible because it is
demonstrated convincingly by data and (2) growth trajectories can change and there is nothing strange
or unusual about it. Indeed, recently, hyperbolic growth was diverted to slower trajectories (Nielsen,
2016a, 2016d, 2016i).
It is remarkable, that this apparently impossible (because of its singularity) hyperbolic growth was
possible for the most part of the past 12,000 years (Nielsen, 2016i). Every time it was interrupted, and
it happened only twice in the past, it was converted again to a hyperbolic growth. Now, it is interrupted
again but the future trajectory is yet unknown.
We have discussed mathematical properties of the historical GDP/cap distributions. We have
explained how they should be analysed and interpreted.
If both components of the GDP/cap indicator increase hyperbolically, then the GDP/cap distributions
represent a ratio of hyperbolic trajectories. We have a consistent evidence in data that the economic
growth and the growth of population were hyperbolic (Nielsen, 2016a, 2016d, 2016i; von Foerster,
Mora & Amiot, 1960). The characteristic features created by the division of hyperbolic distributions
may be confusing but they can be easily explained. Data have to be analysed. Presenting them in a
grossly distorted way is self-defeating and it leads to incorrect conclusions (Ashref, 2009; Galor, 2005a,
2005b, 2007, 2008a, 2008b, 2008c, 2010, 2011, 2012a, 2012b, 2012c; Galor and Moav, 2002; Snowdon
& Galor, 2008).
We have explained how to understand the confusing features of the historical GDP/cap distributions.
They can be interpreted simply as the linearly-modulated hyperbolic GDP distributions. Linear
modulation is by the reciprocal values of population data. We have discussed how these distributions
can be analysed, how their features can be explored and explained.
As an illustration of our discussion, we have investigated the data (Maddison, 2001) used in
developing the Unified Growth Theory (Galor, 2005a, 2011). In his theory, Galor discusses various
socio-economic concepts of growth but his theory does not explain the mechanism of economic growth
because it is based firmly on the misinterpretation of the purely mathematical features of hyperbolic
distributions. His discussion of socio-economic issues might be interesting, for various reasons, but it
has no relevance to explaining the mechanism of the economic growth because changes in socio-
economic conditions had no effect on the economic growth trajectory as manifested by the available
data (Maddison, 2001), the same data, which were used, but not analysed, during the formulation of the
Unified Growth Theory.
Galors speculations about socio-economic processes are strongly guided by phantom features
created by the deliberately distorted presentations of data (Ashref, 2009; Galor, 2005a, 2005b, 2007,
2008a, 2008b, 2008c, 2010, 2011, 2012a, 2012b, 2012c; Galor and Moav, 2002; Snowdon & Galor,
2008). Different stages of growth claimed in this theory did not exist. Their claimed presence is
contradicted by the same data, which were used during the development of the Unified Growth Theory
and in all other related publications, as listed above.
In general, the GDP and population, global, regional and even in individual countries, were
increasing monotonically and consequently the GDP/cap ratios are also represented by monotonically
increasing distributions governed by a single mechanism of growth.
320
__________________________________________________________________________________
Unified Growth Theory does not explain the historical economic growth because it is critically and
inflexibly based on the deliberately constructed phantom features, which are contradicted by data. In
particular, the three regimes of growth claimed by this theory did not exist and there were no takeoffs
in the economic growth or in the growth of population. This theory describes a phantom world but
presents it as real. Stories and explanations presented in the Unified Growth Theory might sound
plausible but they are contradicted by data.
Historical GDP/cap distributions might look puzzling and complicated but they are in fact simple
distributions. Their puzzling features are nothing more than just the mathematical features created by
dividing two hyperbolic distributions. Their mechanism might also look complicated but hyperbolic
distributions are described by exceptionally simple mathematical formula and the mechanism of these
distributions, as representing the historical economic growth and the historical growth of population, is
also simple (Nielsen, 2016j).
Historical hyperbolic economic growth can be explained as having been propelled by the simplest
possible market force where the growth of the GDP (or the on average growth of the common wealth)
is prompted by the force directly proportional to the already existing size of the GDP. On average,
wealth was generating wealth directly proportionally to the existing wealth. Historical hyperbolic
growth of the population can be explained as having been propelled by the simplest force of procreation
(the combination of the natural sex drive combined with the natural process of aging and dying), which
on average was constant per person. Historical growth of income per capita, expressed as the GDP/cap,
can be explained as having been prompted by the combination of these two forces, and the puzzling
features of the GDP/cap distributions turn out to be nothing more than the mathematical properties of
dividing two hyperbolic distributions. If these simplest forces of growth are combined with some other
strong forces, as it is now, the economic growth and the growth of population are no longer hyperbolic.
The current GDP/cap values are still increasing but the shapes of their distributions cannot be
explained by the mathematical properties of diving two hyperbolic distributions because we are no
longer dealing with hyperbolic distributions. However, in principle, their shapes could be reproduced
by dividing mathematical distributions describing the current growth of the GDP and population.
However, the underlying mechanism of any of them is now no longer simple and the mechanism of the
current growth of income per capita is also no longer simple.
References
Press.
475.
http://dx.doi.org/10.1007/s11698-011-0062-7
321
__________________________________________________________________________________
http://dx.doi.org/10.4236/am.2014.513188
Nielsen, R. W. (2016b). Unified Growth Theory contradicted by the absence of takeoffs in the Gross Domestic Product.
Nielsen, R. W. (2016f). The unresolved mystery of the great divergence is solved. Journal of Economic and Social Thought,
Nielsen, R. W. (2016g). Mathematical analysis of the historical income per capita distributions. Economic Review, 3(2),
300-319. http://www.kspjournals.org/index.php/TER/article/view/766/889
Nielsen, R. W. (2016i). Growth of the world population in the past 12,000 years and its link to the economic growth.
Nielsen, R. W. (2016j). Mechanism of Hyperbolic Growth Explained. Journal of Economic Library, 3(4), 411-428.
322
__________________________________________________________________________________
20
Economic Growth and the Growth of Human

Population in the Past 2,000,000 Years
By Ron W. NIELSEN
Abstract. Growth of human population in the past 2,000,000 years is analysed. It is shown that the growth was
in three major stages: (1) 2,000,000 to 27,000 BC, (2) 27,000 BC to AD 510 and (3) AD 510 to present. Each
stage is described by hyperbolic distribution followed by a significantly shorter, non-hyperbolic transition to a
new stage. Data show also a minor disturbance in the third hyperbolic stage. Each hyperbolic stage was prompted
by a single force, the biologically-controlled force of procreation expressed as the difference between the
biologically-controlled force of sex drive and the biologically-controlled process of aging and dying. The
fundamental parameter describing hyperbolic growth is given by the ratio of the force of growth and of the
resistance to growth. It is assumed that during transitions, this fundamental force remained the same but the
resistance to growth was changing. All these three stages, and the minor disturbance in the middle of the third
stage, are now described mathematically and explained. The derived parameters are used to calculate the size of
the world population in the past 2,000,000 years and to fill in the gaps between data. These parameters can be
used to calculate the growth rate at any time in the past 2,000,000 years. Analysis of population data and the
earlier analysis of the Gross Domestic Product (GDP) per capita allow also for the evaluation of the economic
growth in the past 2,000,000 years. The size of the population and the GDP values are tabulated.
Keywords. Growth of human population, Economic growth, Hyperbolic growth, Mechanism of hyperbolic
growth
JEL. A10, A12, A20, C12, Y80
1. Introduction
he aim of this publication is to analyse the growth of human population and the associated
T economic growth in the past 2,000,000 years. This work is an extension of our previous analysis
of the growth of human population in the past 12,000 years (Nielsen, 2016a) and of the analysis
of economic growth during the AD era (Nielsen, 2016b). These earlier studies demonstrated that the
natural tendency for the growth of human population and for the economic growth is to follow
hyperbolic distributions. Hyperbolic growth can be faster or slower but it is always prompted by the
fundamentally the same mechanism.
We have shown that the mechanism of hyperbolic growth of human population can be easily
explained (Nielsen, 2016c). It is a growth prompted by just one indispensable force, the biologically-
controlled force of procreation expressed as a difference between the biologically-controlled force of
sex drive and the biologically-controlled process of aging and dying. No other forces are needed. A
change in the growth trajectory occurs only if other forces interfere substantially with this fundamental
force of growth. In the past 12,000 years, there was only one strong interference, around AD 1, and one
minor interference, around AD 1300. Each time, the fundamental character of the growth trajectory was
not changed. There was only a transition from one hyperbolic trajectory to another. The first time, it
was a transition from a fast to a slow hyperbolic growth, while the second time it was a transition to
Australia.
. +61407201175
Published as: Nielsen, R. W. (2017). Economic Growth and the Growth of Human Population in the Past 2,000,000 Years.
323
__________________________________________________________________________________
only a slightly faster growth. With the exception of these two, relatively brief transitions, the growth
was always hyperbolic. In addition to these past transitions, we now experience a new strong
interference reflected in the gradual slowing down growth. The growth of population is no longer
hyperbolic but it is still close to the historical hyperbolic growth. Now, we are going to demonstrate
that hyperbolic growth prevailed not only in the past 12,000 years but also during the past 2,000,000
years.
2. Data for the early growth of population

If data for the BC era down to 10,000 BC are scarce, data beyond that time are even more difficult
to find. However, we now have a few estimates from reputable sources and we can use them to extend
the analysis of the growth of human population down to 2,000,000 million years ago.
The earliest estimates were made by Deevey (1960). He estimated that during the Lower Palaeolithic
(around 1,000,000 years ago) the size of population was 0.125 million, during the Middle Palaeolithic
(around 300,000 years ago) it was 1 million and 3.34 million during the Upper Palaeolithic (around
25,000 years ago). Birdsell (1972) estimated 0.4, 1 and 2.2 million for the same years, respectively,
while Hassan (1981) estimated 0.6, 1.2 and 6. In 2002, he estimated 0.4, 0.8, 1.2 and 3.3 million at
1,500,000, 1,000,000, 100,000 and 14,000 years ago, respectively (Hassan, 2002). In our calculations,
we shall use his updated estimates (Hassan, 2002). Incidentally, it should be noted that his two values
listed in his Table 17.2 (Hassan, 2002, p. 684) are clearly misplaced. The values of 0.4 and 0.8 million
should have been aligned with 1,500,000 and 1,000,000 respectively. However, his diagram presented
as Figure 17.2 is correct.
All these estimates are listed in Table 1. The corresponding years are expressed as BC. The
expression years ago or before present are interpreted as before 2000. The years 1,500,000, 1,000,000
and 300,000 years ago or before present are interpreted as 1,500,000 BC, 1,000,000 BC and 300,000
BC. The values for the years after 100,000 BC were reduced by 2000.
Table 1. Estimates of the size of population before 10,000 BC (in million)

Year Deevey Birdsell Hassan Average
(BC) (1960) (1972) (2002) Values
1,500,000 0.4 0.4
1,000,000 0.125 0.4 0.8 0.44
300,000 1 1 1
100,000 1.2 1.2
23,000 3.34 2.2 2.77
12,000 3.3 3.3
Reciprocal values of these data are shown in Figure 1. This figure shows also data used in the earlier
analysis (Nielsen, 2016a). As pointed out earlier (Nielsen, 2014), linearly decreasing reciprocal values
identify hyperbolic growth, because hyperbolic growth is described by the reciprocal of a decreasing
straight line:
1
S (t ) = (1)
a kt
where S (t ) is the size of the growing entity, in our case the size of population or the size of economic
growth, t is the time and a and k are positive constants.
In Figure 1 we see two straight lines. They cross at 34,350 BC. Around that time there was a
transition from a slow hyperbolic trajectory to a significantly faster growth, also described by
hyperbolic trajectory. This transition was one of only two major transitions in the past 2,000,000 years.
The later major transition was around AD 1 (Nielsen, 2016a). A closer view of this first transition is
shown in Figure 2.
324
__________________________________________________________________________________
Figure 1. Reciprocal values of the size of population between 1,500,000 BC and 1000 BC. Decreasing straight
lines for the reciprocal values identify hyperbolic trajectories (Nielsen, 2014). The two trajectories cross at
34,350 BC marking a transition from a slow to a fast hyperbolic trajectory. The late BC trajectory was
discussed earlier (Nielsen, 2016a). The BC years are represented by negative numbers.
Figure 2. The first major demographic transition. It occurred between 46,000 BC and 27,000 BC. It was a
transition from a slow hyperbolic growth to a significantly faster hyperbolic growth, which prevailed until 425
BC to be replaced in AD 510 by a significantly slower hyperbolic growth during the AD era (Nielsen, 2016a).
The BC years are represented by negative numbers.
This early transition commenced around 46,000 BC and continued until around 27,000 BC. From
around that year, the growth of the world population started to follow a significantly faster hyperbolic
trajectory. These new data confirm that the natural tendency for the growth of population is to follow
hyperbolic distributions.
2. Growth of human population in the past 2,000,000 years

2.1. Overview
In Figure 3 we show the average values of data describing the growth of the world population in the
past 2,000,000 years (Biraben, 1980; Birdsell, 1972; Clark,1968; Cook,1960; Deevey, 1960; Durand,
1974; Gallant, 1990; Hassan, 2002; Haub, 1995; Livi-Bacci, 1997; McEvedy & Jones, 1978; Taeuber
325
__________________________________________________________________________________
& Taeuber, 1949; Thomlinson, 1975; Trager, 1994, United Nations, 1973, 1999, 2013; US Census
Bureau, 2017). The time scale is in years before 2100. We also display the best fit to the data, which
most of the time is hyperbolic. We can see these hyperbolic distributions more clearly in Figure 4. The
fit presented in Figure 3, combined with the exceptionally slow growth during the first stage, allows for
the extension of the growth of population to 2,000,000 years before 2100 or to approximately to
2,000,000 BC.
Figure 3. Growth of human population in the past 2,000,000 years.
Growth of human population in the past 2,000,000 was in three major stages but it was not in the
stages imagined by Deevey (1960). It is remarkable that based on a strongly limited information he did
realise that the growth of population was in three major stages. However, while being close to the correct
interpretation of the growth of population, Deevey imagined the three stages incorrectly. He imagined
that each stage was leading to an equilibrium, i.e. to a plateau in the growth of population as shown in
Figure 5. This figure is based on his conceptual diagram (Deevey, 1960, p. 198). If we compare his
interpretation of growth shown in Figure 5 with the growth presented in Figure 3, we can see that the
growth was indeed in three stages as suggested by Deevey but the details of the growth trajectories are
clearly different. Only the first stage looks similar to the stage suggested by Deevey. However, as we
shall explain later, this stage also did not lead to an equilibrium.
It is remarkable, that the currently established knowledge (Nielsen, 2016g), which is based on the
doctrine of the so-called Malthusian stagnation, ignores not only results of von Foerster, Mora and
Amiot (1960) but also the observation of Deevey (1960). These two early studies clearly demonstrated
that there was no stagnation in the growth of population. They indicated that there was a regular and
well-defined pattern of growth, which contradicts the doctrine of the so-called Malthusian stagnation.
Results of von Foerster, Mora and Amiot (1960) demonstrated that the growth of human population
during the AD era was hyperbolic, and consequently not stagnant. It followed a monotonically
increasing trajectory. Hyperbolic growth is in the direct contradiction of the concept of the so-called
Malthusian stagnation.
Hyperbolic growth is slow over a long time (but not stagnant) and fast over a short time, so fast that
it escapes to infinity at a fixed time. It is the growth, which is governed by the same mechanism when
it is slow and when it is fast. If we want to interpret the slow growth as stagnant we should apply the
326
__________________________________________________________________________________
same interpretation to the fast growth. It would be obviously ludicrous to describe the growth escaping
to infinity at a fixed time as stagnant, but because it is always the same growth, then it is also ludicrous
to describe it as stagnant when it is slow. The concept of the so-called Malthusian stagnation is based
on the incorrect interpretation of hyperbolic growth and has no place in science.
Figure 4. The three major stages of growth of the world population in the past 2,000,000 years: (1) between
2,000,000 BC and 27,000 BC, (2) between 27,000 BC and AD 510 and (3) between AD 510 and present. The
last stage experienced a minor distortion between around AD 1195 and 1470. This distortion caused a small
shift in the hyperbolic growth.
Figure 5. The three population surges as imagined by Deevey (1960, p. 198).
Results of Deevey indicated that over a longer time, extending as far back as to 1,000,000 years ago,
the growth was in three distinctly different stages. It is again a clearly different pattern than the pattern
suggested by the concept of the so-called Malthusian stagnation. The doctrine of the so-called
Malthusian stagnation claims an endless stagnant state of growth, characterised by unpredictable,
327
__________________________________________________________________________________
random fluctuations often described as Malthusian oscillations. Hyperbolic growth is definitely

predictable and consequently it suggests an entirely different interpretation of the mechanism of growth.
Studies of von Foerster, Mora and Amiot (1960) and of Deevey (1960) indicated that there was
nothing chaotic about the growth of population. They indicated that there was a certain regular pattern.
Such a regular pattern can hardly be expected to be produced by random forces of growth.
In order to understand the growth of population in the past 2,000,000 years, it is useful to discuss
separately its three stages of growth as presented in Figure 4: (1) between 2,000,000 BC and 27,000
BC, (2) between 27,000 BC and AD 510, and (3) between AD 510 and present. Each of these stages is
described by hyperbolic growth followed by a transition to the next stage. However, the last stage
contains a fine structure expressed as a slight shift in the hyperbolic distribution.
2.2. Mathematics of growth
Parameters describing the growth of population in the past 2,000,000 years are listed in Table 2.
They are: a and k, for the hyperbolic growth and ai ( i = 0 to n) and bi ( i = 0 to n + 1 ) for transitions.
They can be used to calculate the size of population S (t ) and the growth rate R (t ) at any given time.
For these parameters, the size of population is in billions. The time is in years and it is positive for the
AD era and negative for the BC era.
Table 2 presents also the range of S (t ) and R (t ) values for hyperbolic distributions, which are also
the range of values for transitions, because the end of a given hyperbolic growth is the beginning of a
transition while the beginning of a hyperbolic growth is the end of a preceding transition.
Mathematics of growth of population is exceptionally simple. As discussed earlier (Nielsen, 2016a,
2016c) and as shown in Figures 3 and 4, the growth was hyperbolic, except when there was a relatively
brief transition. Hyperbolic growth is described by a very simple mathematical expression, presented
as eqn (1), which is a solution of a very simple differential equation:
1 dS (t )
= kS (t ) . (2)
S (t ) dt
Transitions are described by a similar differential equation:
1 dS (t )
= k (t ) S (t ) . (3)
S (t ) dt
Parameter k, whether constant or dependent on time, is the driving force divided by the resistance to
growth (Nielsen, 2016c). For the growth of population, the driving force is the force of procreation
given by the difference between the biologically controlled force of sex drive and the biologically
controlled aging and dying. It is a spontaneous, unrestrained and fundamental force of growth, which
has to be considered in any attempt to explain the mechanism of growth of human population. Other
forces may be added but only if necessary, i.e. if this fundamental force is unable to explain the
mechanism of growth. The study presented here and in earlier publications demonstrates that this force
alone explains why the growth of population was, most of the time, hyperbolic (Nielsen, 2016a, 2016c,
2016d; von Foerster, Mora & Amiot, 1960).
During transitions, the fundamental force of procreation does not change. There is no need to assume
that it does. Only the resistance to growth is changing and this change is described by k (t ) .
In the past, every change in the resistance to growth was leading to a new, constant resistance and
consequently to a new hyperbolic growth. The current transition, which commenced around 1950, also
describes a change in the resistance to growth but the future trajectory is unknown.
Solution of the eqn (3) is given by the following expression:
1
S (t ) = . (4)
k (t )dt
In the simplest case, when k (t )= k= const , the eqn (3) is the same as eqn (2) and the solution (4) is
the same as eqn (1). It is the reciprocal of a linear function.
If we assume that k (t ) is represented by the n-order polynomial, if
328
__________________________________________________________________________________
n
k (t ) = a t
i =0
i
i
(5)
then
1
n +1

S (t ) = bi t i .
i =0
(6)
Table 2. Parameters describing the growth of population in the past 2,000,000 years.
Hyperbolic Growth Transitions

1
n +1 n
) (a kt )1 ; R(t ) = kS (t )
S (t= S (t ) =
bit i

; k (t ) = a t ; R(t ) = k (t )S (t )
i
i
i =0 i =0
Years Parameters Years Parameters
2.000,000 =a 7.120 10 2
46,000 27,000 BC 9.247 102 b1 =
b0 = 9.990 102
46,000 BC =k 1.296 103 1.966 106 b3 =
b2 = 1.295 1011
2,000,000 BC =
S (t ) 3.027 105 a0 9.990 102 a1 =
= 1.808 101
R(t ) 3.923 105%

= a2 3.885 1011
=
46,000 BC =
S (t ) 1.296 106
R(t ) 1.680 104%
=
27,000 425 BC =a 2.282 100 425 BC AD 510 =
b0 3.834 10=
0
b1 2.347 103
=k 2.210 102 = 2.493 108

b2 1.330 105 b3 =
27,000 BC =
S (t ) 1.682 106 a0 = 2.659 105
2.347 103 a1 =
R(t ) 3.718 103%

= a2 7.479 108
=
425 BC =
S (t ) 1.406 108
R(t ) 3.108 101%
=
AD 510 1195 =a 6.940 100 AD 1195 1470 2.903 10=

b0 = 2
b1 1.022 100
=k 3.448 103 1.309 103=
b2 = b3 7.326 107
AD 510 =
S (t ) 1.930 108 1.517 1010 a0 =
b4 = 1.022 100
R(t ) 6.654 102%

= a1 2.618 103 a2 =
= 2.198 106
AD 1195 =
S (t ) 3.546 108 a3 6.068 1010
=
R(t ) 1.223 101%

=
AD 1470 1950 =a 9.123 100 AD 1950 2016 =
b0 2.001 103 b1 =
2.928 100
=k 4.478 103 b2 1.428 103 b3 =
= 2.323 107
AD 1470 =
S (t ) 3.935 108 = 2.856 103
a0 2.928 100 a1 =
R(t ) 1.762 101%

= a2 6.968 107
=
AD 1950 =
S (t ) 2.550 109
=
R(t ) 1.142 100%
S (t ) - the size of population. R (t ) - the growth rate. In mathematical formulae, time is in years and it has positive values for
the AD era and negative for the BC era. Furthermore, for the listed parameters, the size of population is in billions. The growth
rate given by the mathematical formulae is not expressed in per cent.
We should also notice that eqns (2) and (3) describe the growth rate R (t ) . Thus, if we know the size
of the population and k or k (t ) , we can also calculate the corresponding growth rate at a given time:
329
__________________________________________________________________________________
R (t ) = k (t ) S (t ) . (7)
For the hyperbolic growth [i.e. for the first-order hyperbolic growth given by the eqn (1)]
k (t )= k= const and the growth rate is directly-proportional to the size of population.
2.3. Stage 1: 2,000,000-27,000 BC
This stage is made of a hyperbolic growth between 2,000,000 BC and 46,000 BC followed by a
transition to the next stage. The transition was between 46,000 BC and 27,000 BC (see Figures 2 and
4).
In Figures 3 and 4, this stage looks different than the other two stages and it resembles the
distribution outlined by Deevey (see Figure 5). However, it is just an illusion created by using
logarithmic scales of reference and Deevey appears to have been misguided by this illusion. He
imagined that it was a fast growth followed by an equilibrium, or a plateau. However, the calculated
curve shown in Figures 3 and 4 is hyperbolic. It was not a fast growth followed by equilibrium but a
monotonically increasing growth.
We know that the growth in the first stage was hyperbolic because we have shown earlier (see Figure
1) that the reciprocal values of the size of the population during that time were following closely a
straight line, which identifies hyperbolic growth (Nielsen, 2014). Why then does this stage look so much
different? How to explain the peculiar shape presented in Figures 3 and 4?
First, it is important to notice that hyperbolic growth during this first stage was exceptionally slow,
so slow that if continued it would not escape to infinity until around AD 549,391. Second, we have to
remember that logarithmic scales, while being useful in displaying a wide range of data, they also
introduce unavoidable distortions. In Figure 3 and 4 we have double distortion because we are using
two logarithmic scales. The further we go back in time to stronger is the compression of the displayed
data, but there is also an increasing compression of the displayed size of the population as we move up
along the vertical scale.
Every marked section of the first (left-most) cycle of the horizontal scale represents a compression
of 1,000,000 years. The vertical scale introduces similar distortion but in reverse order. Here the first
cycle represents an exceptionally stretched scale. This compressing and stretching, combined with the
exceptionally slow growth during the first stage creates an illusion of a fast growth followed by an
equilibrium, illusion so strong that it caused Deevey not only see an incorrect pattern but also to try to
explain its mechanism.
A simple way to dispel this illusion is to use linear scales as shown in Figure 6. In this figure, we
present precisely the same data (for the Stage 1) and the same hyperbolic distribution as shown in
Figures 3 and 4 but now we use linear scales for the time and for the size of the population.
Figure 6. The first stage of hyperbolic growth between 2,000,000 BC and 46,000 BC displayed using linear
scales for the time and for the size of population. The illusion of a fast growth followed by an equilibrium
created by the double-logarithmic scales used by Deevey (1960, p.198) and in Figures 3 and 4 has now
330
__________________________________________________________________________________
disappeared. It is now clear that the growth is increasing monotonically and that it does not lead to an
equilibrium.
We can now see clearly that the growth of population was increasing monotonically. There is
obviously no sign of any plateau or equilibrium and no hope of having such a plateau in the future
because the growth was hyperbolic, escaping to infinity at a fixed time. Deeveys claim of plateaus and
his attempts to explain their mechanism was based on illusion.
During this early stage of the BC growth, the size of population increased from the estimated 0.4
million in 1,500,000 BC to 1.3 million in 46,000 BC. The calculated value in 1,500,000 BC is 0.38
million. If we extend the fitted hyperbolic distribution to 2,000,000 BC, then the calculated size of
population in that year is 0.3 million. If continued, the size of population would increase to one billion
in AD 548,620.
Reciprocal values of data shown in Figure 1 demonstrates that this exceedingly slow hyperbolic
growth was replaced by a much faster growth. The transition occurred between 46,000 BC and 27,000
BC (see Figure 2).
The size of the population during this transition increased from 1.3 million in 46,000 BC to 1.7
million in 27,000 BC. This transition converted the exceedingly slow hyperbolic growth during the first
stage to a 17 times faster hyperbolic growth (as measured by the parameter k) during the second stage,
the difference in the intensity of growth reflected in the distinctly different values of the gradient of the
reciprocal values of the size of the population shown in Figure 1. During that time, the resistance to
growth decreased by a huge factor of around 17 and starting from around 27,000 BC the growth of
population was much faster than before 46,000 BC.
The timing of this transition agrees well with archaeological and anthropological data. Even though
the emergence of modern humans is claimed to have been between 150,000 and 200,000 years ago
(Mellars, Boyle, Bar-Yosef & Stringer, 2007) the progress in their development was slow until around
50,000 BC, as demonstrated by archaeological evidence (Klein, 1989, 1995; Mellars, 1989; Stringer &
Gamble, 1993). Human evolution appears to have experienced a great leap forward around that time.
For a long time since their emergence, modern humans were not much different than other hominins
and it was only around 50,000-40,000 years ago that a major behavioural difference developed (Klein,
1995, p. 167). This first transition in the hyperbolic growth appears to coincide also with the extinction
of Neanderthals, first in Europe and later in other parts of the world, marking the beginning of the
undisputed domination of Homo sapiens (Higham, et al., 2014).
Forces operating during the first transition between 46,000 BC and 27,000 BC from an earlier large
resistance to growth before around 46,000 BC to significantly smaller resistance after around 27,000
BC were of a social and intellectual nature. The long race between different representatives of hominins
was over. One by one they were left behind and became extinct. Finally, the last two remaining were
Homo floresiensis and Homo neanderthalensis but they also were eliminated or virtually eliminated
around the time of the beginning of the first transition, i.e. around 50,000 BC. Now, only modern
humans, represented by Homo sapiens, remained. The first transition, between 46,000 BC and 27,000
BC was a transition to a new era of the exceptionally fast and long-lasting hyperbolic growth, the unique
growth which was never to be repeated.
This complete freedom of growth was eventually restricted, not by forces of nature and not by the
competition with other representatives of the genus Homo because they were extinct for a long time but
by the strong competition between humans. However, in 27,000 BC, at the end of the first transition,
this change was still long time into the future. The gained momentum of the free growth was to propel
the growth of human population for many thousands of years.
2.4. Stage 2: 27,000 BC - AD 510
Stage 2 is made of a fast hyperbolic growth between 27,000 BC and 425 BC, followed by a transition
to a slower hyperbolic trajectory during the AD era. The transition took place between 425 BC and AD
510. (In our earlier publications, we have labelled this transition as being roughly between 500 BC and
AD 500.)
Hyperbolic growth between 27,000 BC and 425 BC was the fastest growth (as defined by the
parameter k) in the past 2,000,000 years. During that time, the size of population increased from 1.7 in
27,000 BC million to 140 million in 425 BC, representing a nearly 82-fold increase. In contrast, there
was only around 38-fold increase between AD 510 and present. If continued, this fast BC growth would
331
__________________________________________________________________________________
escape to infinity at the end of 104 BC. We have come very close to experiencing the so-called
population explosion at the end of the BC era.
In deciding which hyperbolic growth is fast we should not be confused by the growth during the AD
era. It reached a higher size of population in a shorter time but we should remember that it also started
with a significantly larger size of population, around 190 million, compared with only 1.7 million for
the hyperbolic growth between 27,000 BC and 425 BC.
The transition between 425 BC and AD 510 can be described by the reciprocal of the third-order
polynomial. During this transition, the resistance to growth increased by a factor of 6.4. As discussed
earlier (Nielsen, 2016c), forces shaping this transition appear to have been of political nature. This
transition coincides with the domination of Roman Empire over large areas surrounding the
Mediterranean Sea. It also coincides with the accelerated process of the formation of countries in
various parts of the world and with the rapidly changing political landscape (Teeple, 2002). From the
complete freedom in around 27,000 BC, humans became slaves of their own design. They have invented
many ways of self-destruction, bondage and oppression, which eventually led to a new hyperbolic
growth characterised now by a larger resistance to growth. Humans appear to be their own best enemies
and they might eventually cause their own extermination.
2.5. Stage 3: AD 510 - present
This stage is also made of a hyperbolic growth followed by a transition, which commenced around
1950. We have shown earlier (Nielsen, 2016a) that the growth of population between AD 510 and 1950
can be well described using a single hyperbolic distribution. However, we have also pointed out that
there was a minor disturbance in this hyperbolic growth between around AD 1200 and 1400. This
disturbance caused only a small shift in the hyperbolic growth (see Figure 4).
The best description of data between AD 510 and 1950 is given by two, approximately parallel
hyperbolic trajectories separated by a small transition between around AD 1195 and 1470. The two
hyperbolic trajectories, before AD 1195 and after AD 1470 are virtually identical. Measured by the
parameter k, hyperbolic growth after AD 1470 was only 30% faster than the hyperbolic growth before
AD 1195.
As discussed earlier (Nielsen, 2016a), this minor transition between AD 1195 and 1470 coincides
with a unique event of a convergence of five demographic catastrophes. This is the only example
showing a correlation between demographic catastrophes and the growth of population. However, the
combined impact was small.
From around 1950, there was at first a small surge in the growth of population followed soon by a
consistently slowing down growth. The data for the world population from that year are well
documented by the US Bureau of Census (2017) but they can be also described using a third-order
polynomial with parameters listed in Table 2 (see Figure 7).
Figure 7. Population data (US Census Bureau, 2017) are compared with the third-order polynomial
distribution. Its parameters are listed in Table 2. This is just a mathematical description of data.
This current transition appears to be associated with the increasing impact of Malthusian
preventative checks (Malthus, 1798). The outcome of this transition is unknown. If the past pattern of
332
__________________________________________________________________________________
growth is repeated, it could be a transition to a new hyperbolic trajectory. However, hyperbolic growth
of population is possible only if the dominating force of growth is the biologically-controlled force of
procreation. It is unlikely that this force alone will control the future growth of population. Under new
conditions, with the increasing awareness of the need to control growth, the future growth of population
could follow an entirely different trajectory. For the first time in human existence it will probably not
be a hyperbolic growth.
The fitted distribution shown in Figure 3 with parameters listed in Table 2 can be now used to
calculate the size of population at any time in the past 2,000,000 years. The calculated values are listed
in Tables A1-A3 in the Appendix.
3. Economic growth in the past 2,000,000 years

De Long (1998) pointed out that income per capita (GDP/cap) can be used to estimate the past
economic growth expressed in terms of the Gross Domestic Product (GDP). It is because the GDP/cap
values quickly converge to an approximately constant value when we move back in time (see Figure
8). This property is nothing more than the mathematical property of dividing two hyperbolic
distributions (Nielsen, 2017a) but it is useful for calculating the GDP values from the population data.
What it simply means is that as we move back in time, the size of the GDP becomes directly proportional
to the size of the population. They follow virtually the same trajectories but displaced by an
approximately constant factor.
Parameters describing the fitted GDP/cap distribution shown in Figure 8 are = a 1.684 102 and
= k 8.539 106 for the GDP expressed in billions of the 1990 international Geary-Khamis dollars and
= a 7.739 100 and= k 3.765 103 for the Maddisons population data expressed in billions.
The fitted curve is a linearly modulated hyperbolic distribution (Nielsen, 2017a), which increases to
infinity at a fixed time. For the distribution displayed in Figure 8, the point of singularity is in 1971.
The growth of income per capita came close to this critical point but it bypassed it by a small margin of
about 20 years. Income per capita continues now to increase along a new trajectory.
We can see that the calculated curve and the data representing the GDP/cap values quickly converge
to a constant value when we move back in time. We can use this property to estimate the size of the
GDP down to 2,000,000 BC. Results are presented in Tables A4-A6 in the Appendix.
Figure 8. Growth of the Gross Domestic Product per capita (GDP/cap) during the AD era. The full circles
represent Maddisons data (Maddison (2010) and the line is the best fit to the data (Nielsen, 2016e). The
calculated curve is the linearly modulated hyperbolic distribution (Nielsen, 2017a).
333
__________________________________________________________________________________
De Long (1998) carried out similar calculations. However, he used population data listed by Kremer
(1993), which were taken from two sources: McEvedy and Jones (1978) and Deevey (1960). Our results
are based on the analysis of all available data.
Furthermore, De Long assumed a constant GDP/cap value below AD 1500. This is good
approximation below AD 1000 but not between AD 1000 and 1500. We shall use consistently the fitted
trajectories but only below 1950 for two reasons: (1) good year-by-year data for the GDP starting from
AD 1950 are already available (Maddison, 2010; GGDC, 2013) and (2) the calculated distribution of
income per capita reproduces the data only up to 1950. From 1950, the GDP/cap values do not follow
the linearly modulated hyperbolic distribution.
The GDP values presented in Tables A4-A6 are based on the best fits to the population data and to
the GDP/cap data up to 1950. From 1950, the GDP values are as listed by Maddison (2010) for the
years of up to 2008 and as calculated from the GDP/cap data listed by GGDC (2013) for 2009 and 2010
by using population data of the US Census Bureau (2017).

We have carried out analysis of the growth of population in the past 2,000,000 years using data from
a variety of sources (Biraben, 1980; Birdsell, 1972; Clark,1968; Cook,1960; Deevey, 1960; Durand,
1974; Gallant, 1990; Haub, 1995; Hassan, 2002; Livi-Bacci, 1997; McEvedy & Jones, 1978; Taeuber
& Taeuber, 1949; Thomlinson, 1975; Trager, 1994, United Nations, 1973, 1999, 2013; US Census
Bureau, 2017). We have confirmed the earlier observation of Deevey (1960) that the growth of the
world population was in three major stages. However, our analysis reveals that Deevey made a mistake
by imagining that each stage was at first fast but then was reaching a certain equilibrium. Our analysis
shows that each stage was hyperbolic. Each stage was increasing monotonically and was never levelling
off to any form of equilibrium. On the contrary, if not terminated, hyperbolic distributions increase to
infinity at a fixed time.
Nothing can increase to infinity. Consequently, any hyperbolic growth has to be, at a certain stage,
terminated, which is not unusual because many other types of growth not only can but also are at a
certain stage terminated. For instance, the better known exponential growth does not increase to infinity
at a fixed time but if continued over a sufficiently long enough time, it leads to such large values that
becomes unsustainable.
The three stages of growth are: (1) 2,000,000 BC to 27,000 BC; (2) 27,000 BC to AD 510, and (3)
AD 510 to present. Each of the listed stages includes a transition to a new growth. The transitions, as
revealed by the analysis of data, are: (1) 46,000 BC to 27,000 BC, (2) 425 BC to AD 510, and (3) AD
1950 to present. During the third stage of growth, there was a minor transition between AD 1195 and
1470 but it only produced a slight shift in the hyperbolic trajectory.
Hyperbolic growth of population is generated by only one predominant force, the force of
procreation, which is expressed as the difference between the ever-present, biologically-controlled
force of sex drive and the biologically-controlled force of aging and dying (Nielsen, 2016c). This
essential force has to be included in any explanation of the mechanism of growth of human population
and it turns out this force alone generates hyperbolic growth. As long as the growth remains hyperbolic,
there is no need to include any other force. When a hyperbolic growth is being terminated or strongly
disturbed, as between AD 1195 and 1470, other forces are strong enough to interfere with the usually
dominant, biologically controlled, force of procreation.
Hyperbolic growth is characterised uniquely by parameter k [see eqn (1)]. This parameter is the ratio
of the force of growth and of the resistance to growth. Working on the fundamental scientific principle
of parsimony we can assume that during each hyperbolic growth the fundamental force of procreation
per person remains unchanged and only resistance to growth is different. Transitions are associated with
changing the resistance to growth. This change is described by the time-dependent parameter k (t ) [see
eqn (3) and Table 2].
334
__________________________________________________________________________________
Each, of the first two stages of growth of human population in the past 2,000,000 years was
terminated by a transition to a new hyperbolic growth. The third stage is now also being terminated.
This transition commenced around 1950 but its outcome is unknown.
The first hyperbolic stage of growth was slow but during the first transition the resistance to growth
decreased by a factor of around 17. The second stage was characterised by a fast hyperbolic growth, so
fast that if continued it would have escaped to infinity around 104 BC. Fortunately, this fast hyperbolic
growth was terminated. During the second transition, between 425 BC and AD 510, the resistance to
growth increased by an approximate factor of 6.4. The new hyperbolic trajectory was significantly
slower than the immediately preceding BC trajectory.
Each of the past two major transitions, as well as the current transition, appears to be associated with
significant changes in the style of living. The first transition between 46,000 BC and 27,000 BC appears
to have been associated with the surge in the evolution of Homo Sapiens (Klein, 1989, 1995; Mellars,
1989; Stringer & Gamble, 1993). Forces, which eventually reduced substantially the resistance to
growth appear to have been of social and intellectual character. The second major transition between
425 BC and AD 510 appears to have been of political nature as reflected in the apparently intensified
changes in the political landscape (Teeple, 2002). The current third major transition appears to be
moulded predominantly, if not exclusively, by the Malthusian preventative checks (Malthus, 1798).
The minor transition between AD 1195 and 1470 appears to have been of an entirely different nature.
It was not associated with the change in the style of living but rather with the one and only example of
a strong impact of demographic catastrophes caused by an unusual convergence of five major
catastrophic events (Nielsen, 2016a, 2017b). This transition caused a 30% decrease in the resistance to
growth, reflecting the efficient action of the regeneration process triggered by the Malthusian positive
checks (Malthus, 1798; Nielsen, 2016f).
Using the best fit to the data we have calculated the size of human population in the past 2,000,000
years. These values are listed in Tables A1-A3. Using results of our earlier analysis (Nielsen, 2016e) of
the Gross Domestic Product per capita (GDP/cap) and the current analysis of population data, we have
also listed the estimated values of the GDP in the past 2,000,000 years until 1950. The GDP values
from 1950 to 2008 were taken directly from the publication of Maddison (2010). The last two values,
for 2009 and 2010 were calculated using the GDP/cap values listed by GGDC (2013) and the population
data of the US Census Bureau (2017). All these values, expressed in billions of 1990 international Geary-
Khamis dollars are listed in Tables A4-A6.
335
__________________________________________________________________________________
Appendix
Table A1. Growth of human population from 2,000,000 BC to 1BC
Year Population Year Population Year Population

[BC] [Million] [BC] [Million] [BC] [Million]
2,000,000 0.30 7000 6.56 380 160.53
1,500,000 0.38 6000 7.67 370 165.35
1,000,000 0.50 5500 8.38 360 170.22
800,000 0.57 5000 9.24 350 175.15
600,000 0.67 4500 10.29 340 180.11
400,000 0.81 4000 11.61 330 185.09
200,000 1.03 3500 13.32 320 190.09
100,000 1.19 3000 15.62 310 195.07
80,000 1.23 2800 16.78 300 200.04
60,000 1.27 2600 18.12 290 204.96
50,000 1.29 2400 19.70 280 209.82
46,000 1.30 2200 21.58 270 214.61
42,000 1.31 2000 23.85 260 219.29
40,000 1.32 1900 25.18 250 223.85
38,000 1.35 1800 26.66 240 228.28
36,000 1.37 1700 28.33 230 232.54
34,000 1.41 1600 30.23 220 236.63
32,000 1.46 1500 32.39 210 240.51
30,000 1.53 1400 34.89 200 244.18
28,000 1.62 1300 37.80 190 247.62
27,000 1.68 1200 41.25 180 250.80
26,000 1.75 1100 45.39 170 253.73
25,000 1.82 1000 50.45 160 256.38
24,000 1.89 900 56.78 150 258.75
23,000 1.98 800 64.93 140 260.83
22,000 2.07 700 75.81 130 262.61
21,000 2.17 600 91.08 120 264.10
20,000 2.27 500 114.03 110 265.29
19,000 2.39 490 116.98 100 266.19
18,000 2.53 480 120.08 90 266.80
17,000 2.68 470 123.36 80 267.12
16,000 2.85 460 126.82 70 267.17
15,000 3.04 450 130.47 60 266.95
14,000 3.26 440 134.35 50 266.49
13,000 3.51 430 138.46 40 265.78
12,000 3.80 425 140.61 30 264.85
11,000 4.15 420 142.05 20 263.71
10,000 4.57 410 146.54 10 262.37
9000 5.09 400 151.12 1 261.01
8000 5.73 390 155.78
336
__________________________________________________________________________________
Table A2. Growth of human population from AD 1 to 1330

[AD] [Million] [AD] [Million] [AD] [Million]
1 260.69 450 188.30 900 260.63
10 259.18 460 188.67 910 262.99
20 257.36 470 189.19 920 265.40
30 255.41 480 189.87 930 267.85
40 253.35 490 190.72 940 270.34
50 251.19 500 191.75 950 272.89
60 248.95 510 192.99 960 275.48
70 246.64 520 194.28 970 278.12
80 244.28 530 195.59 980 280.82
90 241.88 540 196.92 990 283.56
100 239.45 550 198.27 1000 286.36
110 237.00 560 199.63 1010 289.22
120 234.54 570 201.02 1020 292.13
130 232.09 580 202.42 1030 295.10
140 229.66 590 203.84 1040 298.14
150 227.24 600 205.28 1050 301.23
160 224.85 610 206.75 1060 304.39
170 222.50 620 208.23 1070 307.62
180 220.19 630 209.74 1080 310.92
190 217.94 640 211.27 1090 314.29
200 215.73 650 212.82 1100 317.73
210 213.59 660 214.39 1110 321.25
220 211.51 670 215.99 1120 324.85
230 209.49 680 217.61 1130 328.53
240 207.55 690 219.25 1140 332.30
250 205.68 700 220.92 1150 336.15
260 203.90 710 222.62 1160 340.09
270 202.19 720 224.34 1170 344.12
280 200.56 730 226.09 1180 348.26
290 199.03 740 227.86 1190 352.49
300 197.58 750 229.67 1195 354.64
310 196.22 760 231.50 1200 355.85
320 194.96 770 233.36 1210 359.73
330 193.79 780 235.26 1220 363.23
340 192.72 790 237.18 1230 366.33
350 191.75 800 239.14 1240 369.01
360 190.89 810 241.13 1250 371.27
370 190.13 820 243.15 1260 373.11
380 189.48 830 245.20 1270 374.55
390 188.95 840 247.29 1280 375.59
400 188.52 850 249.42 1290 376.28
410 188.22 860 251.58 1300 376.64
420 188.05 870 253.79 1310 376.73
430 188.00 880 256.03 1320 376.58
440 188.08 890 258.31 1330 376.26
337
__________________________________________________________________________________
Table A3. Growth of human population from AD 1340 to 2016

[AD] [Million] [AD] [Million] [AD] [Million]
1340 375.83 1770 834.64 1974 3,984.30
1350 375.35 1780 867.04 1975 4,057.11
1360 374.90 1790 902.06 1976 4,130.85
1370 374.54 1800 940.03 1977 4,205.51
1380 374.36 1810 981.33 1978 4,281.06
1390 374.45 1820 1,026.44 1979 4,357.47
1400 374.90 1830 1,075.88 1980 4,434.73
1410 375.80 1840 1,130.33 1981 4,512.79
1420 377.26 1850 1,190.59 1982 4,591.63
1430 379.41 1860 1,257.63 1983 4,671.22
1440 382.38 1870 1,332.68 1984 4,751.52
1450 386.34 1880 1,417.25 1985 4,832.48
1460 391.48 1890 1,513.28 1986 4,914.08
1470 393.49 1900 1,623.26 1987 4,996.26
1480 400.54 1910 1,750.49 1988 5,078.98
1490 407.86 1920 1,899.36 1989 5,162.20
1500 415.45 1930 2,075.91 1990 5,245.86
1510 423.32 1940 2,288.63 1991 5,329.92
1520 431.50 1945 2,412.23 1992 5,414.31
1530 440.00 1950 2,538.51 1993 5,498.99
1540 448.84 1951 2,587.24 1994 5,583.89
1550 458.05 1952 2,636.93 1995 5,668.96
1560 467.64 1953 2,687.57 1996 5,754.14
1570 477.64 1954 2,739.19 1997 5,839.36
1580 488.08 1955 2,791.79 1998 5,924.57
1590 498.98 1956 2,845.38 1999 6,009.70
1600 510.39 1957 2,899.97 2000 6,094.69
1610 522.32 1958 2,955.57 2001 6,179.48
1620 534.83 1959 3,012.18 2002 6,263.99
1630 547.95 1960 3,069.82 2003 6,348.16
1640 561.73 1961 3,128.47 2004 6,431.94
1650 576.23 1962 3,188.16 2005 6,515.25
1660 591.49 1963 3,248.87 2006 6,598.04
1670 607.58 1964 3,310.62 2007 6,680.24
1680 624.57 1965 3,373.41 2008 6,761.79
1690 642.54 1966 3,437.22 2009 6,842.64
1700 661.57 1967 3,502.07 2010 6,922.73
1710 681.77 1968 3,567.94 2011 7,002.00
1720 703.24 1969 3,634.83 2012 7,080.40
1730 726.10 1970 3,702.74 2013 7,157.89
1740 750.50 1971 3,771.65 2014 7,234.42
1750 776.60 1972 3,841.55 2015 7,309.94
1760 804.57 1973 3,912.44 2016 7,384.42
338
__________________________________________________________________________________
Table A4. Economic growth from 2,000,000 BC to 1BC
Year GDP Year GDP Year GDP

2,000,000 0.13 7000 2.92 380 73.30
1,500,000 0.17 6000 3.42 370 75.51
1,000,000 0.22 5500 3.74 360 77.75
800,000 0.25 5000 4.12 350 80.01
600,000 0.30 4500 4.60 340 82.29
400,000 0.36 4000 5.19 330 84.58
200,000 0.45 3500 5.96 320 86.87
100,000 0.52 3000 7.00 310 89.17
80,000 0.54 2800 7.53 300 91.45
60,000 0.56 2600 8.14 290 93.71
50,000 0.57 2400 8.85 280 95.95
46,000 0.57 2200 9.70 270 98.16
42,000 0.58 2000 10.74 260 100.31
40,000 0.59 1900 11.34 250 102.42
38,000 0.59 1800 12.02 240 104.46
36,000 0.61 1700 12.78 230 106.43
34,000 0.62 1600 13.64 220 108.31
32,000 0.65 1500 14.63 210 110.11
30,000 0.68 1400 15.76 200 111.81
28,000 0.72 1300 17.09 190 113.40
27,000 0.74 1200 18.67 180 114.88
26,000 0.77 1100 20.56 170 116.24
25,000 0.80 1000 22.87 160 117.48
24,000 0.84 900 25.77 150 118.59
23,000 0.87 800 29.49 140 119.56
22,000 0.91 700 34.47 130 120.40
21,000 0.96 600 41.46 120 121.11
20,000 1.01 500 51.98 110 121.67
19,000 1.06 490 53.33 100 122.11
18,000 1.12 480 54.75 90 122.41
17,000 1.19 470 56.25 80 122.58
16,000 1.26 460 57.84 70 122.63
15,000 1.35 450 59.52 60 122.55
14,000 1.44 440 61.29 50 122.36
13,000 1.56 430 63.18 40 122.06
12,000 1.69 425 64.16 30 121.66
11,000 1.84 420 64.82 20 121.16
10,000 2.03 410 66.88 10 120.57
9000 2.26 400 68.98 1 119.97
8000 2.55 390 71.12
Year: BC; GDP: Gross Domestic Product, billion 1990 international Geary-Khamis
dollars.
339
__________________________________________________________________________________
Table A5. Economic growth from AD 1 to 1330

1 119.83 450 87.59 900 123.89
10 119.15 460 87.79 910 125.10
20 118.34 470 88.07 920 126.33
30 117.47 480 88.42 930 127.59
40 116.55 490 88.84 940 128.87
50 115.58 500 89.36 950 130.18
60 114.57 510 89.96 960 131.52
70 113.53 520 90.60 970 132.88
80 112.47 530 91.25 980 134.27
90 111.39 540 91.90 990 135.69
100 110.30 550 92.57 1000 137.14
110 109.19 560 93.24 1010 138.62
120 108.09 570 93.92 1020 140.14
130 106.98 580 94.62 1030 141.68
140 105.89 590 95.32 1040 143.27
150 104.80 600 96.04 1050 144.88
160 103.72 610 96.76 1060 146.54
170 102.66 620 97.50 1070 148.23
180 101.62 630 98.25 1080 149.96
190 100.60 640 99.01 1090 151.73
200 99.61 650 99.78 1100 153.55
210 98.65 660 100.56 1110 155.41
220 97.71 670 101.36 1120 157.31
230 96.81 680 102.16 1130 159.26
240 95.94 690 102.99 1140 161.26
250 95.10 700 103.82 1150 163.31
260 94.30 710 104.67 1160 165.42
270 93.53 720 105.53 1170 167.58
280 92.81 730 106.41 1180 169.79
290 92.12 740 107.30 1190 172.06
300 91.48 750 108.20 1195 173.22
310 90.88 760 109.12 1200 173.92
320 90.32 770 110.06 1210 176.04
330 89.80 780 111.02 1220 177.99
340 89.33 790 111.99 1230 179.75
350 88.91 800 112.97 1240 181.32
360 88.54 810 113.98 1250 182.69
370 88.21 820 115.00 1260 183.87
380 87.94 830 116.04 1270 184.85
390 87.72 840 117.10 1280 185.65
400 87.55 850 118.18 1290 186.29
410 87.44 860 119.28 1300 186.77
420 87.38 870 120.40 1310 187.12
430 87.39 880 121.54 1320 187.37
440 87.46 890 122.71 1330 187.55
Year: AD; GDP: Gross Domestic Product, billion 1990 international Geary-Khamis
dollars.
340
__________________________________________________________________________________
Table A6. Economic growth from AD 1340 to 2010

1340 187.67 1750 471.71 1970 13,765.94
1350 187.79 1760 495.02 1971 14,336.49
1360 187.92 1770 520.73 1972 15,018.42
1370 188.12 1780 549.22 1973 16,015.15
1380 188.42 1790 580.95 1974 16,388.00
1390 188.87 1800 616.51 1975 16,637.92
1400 189.51 1810 656.64 1976 17,449.53
1410 190.39 1820 702.24 1977 18,157.09
1420 191.59 1830 754.52 1978 18,955.43
1430 193.14 1840 815.03 1979 19,633.16
1440 195.15 1850 885.85 1980 20,029.99
1450 197.68 1860 969.81 1981 20,422.61
1460 200.85 1870 1,070.89 1982 20,648.35
1470 202.45 1880 1,194.80 1983 21,235.64
1480 206.68 1890 1,350.09 1984 22,204.27
1490 211.09 1900 1,550.16 1985 22,969.60
1500 215.69 1910 1,817.13 1986 23,781.92
1510 220.50 1920 2,190.37 1987 24,693.77
1520 225.52 1930 2,747.20 1988 25,753.18
1530 230.78 1940 3,662.62 1989 26,576.36
1540 236.29 1945 4,381.03 1990 27,134.08
1550 242.06 1950 5,335.86 1991 27,494.23
1560 248.12 1951 5,649.96 1992 28,077.30
1570 254.50 1952 5,911.28 1993 28,693.57
1580 261.20 1953 6,208.99 1994 29,697.95
1590 268.27 1954 6,421.22 1995 30,942.24
1600 275.73 1955 6,830.52 1996 31,990.50
1610 283.61 1956 7,151.72 1997 33,241.79
1620 291.96 1957 7,423.90 1998 33,803.49
1630 300.80 1958 7,662.29 1999 34,997.33
1640 310.20 1959 8,013.45 2000 36,688.28
1650 320.19 1960 8,432.82 2001 37,739.37
1660 330.85 1961 8,725.32 2002 39,021.27
1670 342.23 1962 9,136.47 2003 40,809.56
1680 354.42 1963 9,533.55 2004 42,950.18
1690 367.51 1964 10,224.89 2005 44,982.59
1700 381.58 1965 10,760.25 2006 47,340.58
1710 396.77 1966 11,346.93 2007 49,411.11
1720 413.20 1967 11,769.15 2008 50,973.94
1730 431.04 1968 12,416.76 2009 50,762.92
1740 450.47 1969 13,101.91 2010 53,650.54
Year: AD; GDP: Gross Domestic Product, billion 1990 international Geary-Khamis
dollars. From 1950, the data are as listed by Maddison (2010) up to 2008. The two values
for 2009 and 2010 were calculated using the GDP/cap values listed by GGDC (2013) and
the population data of the US Census Bureau (2017).
341
__________________________________________________________________________________
References
Birdsell, J. B. (1972). Human Evolution: An Introduction to the New Physical Anthropology. Chicago: Rand Mc Nelly.
Publishing Company.
GGDC (2013). Maddison Project: Database. Groningen Growth and Development Centre.
http://www.ggdc.net/maddison/maddison-project/data/mpd_2013-01.xlsx
Hassan, F. A. (1981). Demographic Anthropology. New York: Academic Press.
Hassan, F. A. (2002). Population Dynamics. In G. Barker (Ed), Companion Encyclopedia of Archaeology. London:
Routledge, pp. 672-713.
Higham, T., et al. (2014). The timing and spatiotemporal patterning of Neanderthal disappearance. Nature, 512 (7514): 306
309. doi:10.1038/nature13621. PMID 25143113.
Klein, R. (1989). The Human Career: Human Biological and Cultural Origins. Chicago: University of Chicago Press.
Klein, R. (1995). Anatomy, behavior, and modern human origins. Journal of World Prehistory. 9: 167198.
doi:10.1007/bf02221838.
Economics, 108(3), 681716.
Ltd.
Mellars, P. (1989). Major issues in the emergence of modern humans. Current Anthropology, 30, 349-385.
Mellars, P. A., Boyle, K., Bar-Yosef, O. & Stringer, C. (Eds.). (2007). Rethinking the Human Revolution: new behavioural
and biological perspectives on the origin and dispersal of modern humans. Cambridge: McDonald Institute for
Archaeological Research. ISBN 978-1-902937-46-5.
http://dx.doi.org/10.4236/am.2014.513188
Journal of Economics Bibliography, 3(1), 1-12. http://www.kspjournals.org/index.php/JEB/article/view/607
Nielsen, R. W. (2016c). Mechanism of Hyperbolic Growth Explained. Journal of Economic Library, 3(4), 411-428.
http://kspjournals.org/index.php/JEL/article/view/1074.
Nielsen, R. W. (2016e). Mathematical analysis of historical income per capita distributions. Economic Review, 3(2), 300-
319. http://www.kspjournals.org/index.php/TER/article/view/766/889.
Nielsen, R. W. (2016f). The dichotomy of Malthusian positive checks: Destruction and even more intensified regeneration.
Nielsen, R. W. (2016g). Scientifically unacceptable established knowledge in demography and in economic research.
Journal of Economic Library, 3(3), 429-457. http://kspjournals.org/index.php/JEL/article/view/918
Nielsen, R. W. (2017a). Puzzling Features of Income per Capita Explained. Journal of Economics Bibliography, 4(1), 10-24.
http://www.kspjournals.org/index.php/JEB/article/download/1155/1216
Nielsen, R. W. (2017b). Demographic Catastrophes Did Not Shape the Growth of Human Population or the Economic
Growth. Journal of Economic and Social Thought, 4(2), 121-141.
Stinger, C. B. & Gamble, C. (1993). In Search of Neanderthals. New York: Thames and Hudson.
Pub.
United Nations (1999). The World at Six Billion. http://www.un.org/esa/population/publications/sixbillion/sixbilpart1.pdf
342
__________________________________________________________________________________
US Census Bureau (2017). International Data Base. http://www.census.gov/ipc/www/idb/worldpopinfo.php
343
__________________________________________________________________________________
21
Demographic Catastrophes Did Not Shape the

Growth of Human Population or the Economic
Growth
By Ron W. NIELSEN
Abstract. Rigorous analysis of demographic catastrophes shows that, individually, they were too weak to shape
the growth of human population and of the associated economic growth. On average, only 6.5% of all major
demographic catastrophes, associated with the death toll larger than or equal to one million, were potentially
strong enough to cause perhaps a minor change in the growth trajectory of the world population, but as shown by
the population data, they did not produce any noticeable disturbance. The absence of impacts of demographic
catastrophes on the growth of population can be explained not only by their low relative intensity but also by the
strong and efficient regenerating process recorded for the first time by Malthus. There was, however, one unusual
event manifested in the convergence of five, major demographic catastrophes. They have caused a minor and
short-lasting change in the growth trajectory of the world population, which, however, was soon counteracted by
the process of regeneration. This analysis shows that the dominant force controlling the growth of human
population was too strong to be influenced in any substantial way by accidental forces. As explained in an earlier
publication, this strong and dominant force driving the growth of population was the force of procreation, which
was approximately constant per person, the force expressed as a difference between the ever-present, biologically-
controlled, force of sex drive and the ever-present and also biologically-controlled process of aging and dying.
Keywords. Demographic catastrophes, Growth of population, Economic growth, Malthusian stagnation,
Hyperbolic growth, Mechanism of hyperbolic growth
JEL. A10, A12, A20, B41, C12, Y80
1. Introduction
emographic catastrophes were supposed to have shaped the growth of human population
D and indirectly also the economic growth because as indicated by Maddisons data
(Maddison, 2001, 2006, 2010) these two processes are strongly correlated. Demographic
catastrophes were supposed to have been responsible for creating the alleged, but non-existent, epoch
of the so-called Malthusian stagnation in the growth of population and in the associated economic
growth. This concept, which was accepted for decades in the demographic and economic research, has
been recently reinforced by Galor and his associates by the deliberately distorted presentation of data
(Ashraf, 2009; Galor, 2005a, 2005b, 2007, 2008a, 2008b, 2008c, 2010, 2011, 2012a, 2012b, 2012c;
Galor and Moav, 2002; Snowdon & Galor, 2008). We have discussed this issues in earlier publications
and we have shown that precisely the same data, which were used in their distorted way by Galor and
his associates to support their preconceived but erroneous ideas, are in fact in the direct contradiction
of the concept of the so-called Malthusian stagnation (Nielsen, 2014, 2016a, 2016b, 2016c, 2016d,
2016e, 2016f, 2016g, 2016h).
The erroneous concept of the so-called Malthusian stagnation and takeoffs from the alleged but non-
existent Malthusian trap in the demographic and economic growth is based on the incorrect
interpretations of hyperbolic distributions. They are indeed slow over a long time and fast over a short
Australia.
. +61407201175
Published as: Nielsen, R. W. (2017). Nielsen, R. W. (2017). Demographic Catastrophes Did Not Shape the Growth of
Population or the Economic Growth. Journal of Economic and Social Thought, 4(2), 121-141.
344
__________________________________________________________________________________
time but they increase monotonically and there is no place on them where they change suddenly from
being slow to being fast. In order to explain the mechanism of hyperbolic growth, hyperbolic
distributions have to be treated as a whole. They cannot be divided into two or three different regimes
of growth as incorrectly imagined by Galor (2005a, 2011) and by many other researchers.
In the discussion presented here we shall extend our earlier discussions of the growth of human
population by concentrating our attention on the possible impacts of demographic catastrophes. We
have already explained (Nielsen, 2016i) that in harmony with the observation published by Malthus
(1798), his positive checks (demographic catastrophes and harsh living conditions) have a dichotomous
effect on the growth of population: they are destructive by increasing the death toll but they are also
constructive by triggering the process of regeneration. In the discussion presented here we are going to
demonstrate that there is also another reason why demographic catastrophes did not shape the growth
of human population: they were generally too weak to have any tangible impact. They might have been
strong enough to upset the growth of some local populations but with only one exception discussed
earlier (Nielsen, 2016j) when there was an unusual convergence of five remarkably strong demographic
catastrophes, they had no effect on the growth of global population, or even on the growth of regional
populations (Nielsen, 2016d).
2. The alleged age of pestilence and famine

In one of his publications, Lagerlf stated that Throughout human history, epidemics, wars and
famines have shaped the growth path of population (Lagerlf, 2003a, p. 435). He studied the growth
of population in England, France and Sweden using his model of growth, which incorporated the
concept of the so-called Malthusian stagnation. Similar calculations were carried out earlier by
Artzrouni and Komlos (1985) for the world population. These two studies are most interesting because
when closely examined they show that the mechanism of stagnation does not produce expected results
(Nielsen, 2016k). It did not produce a stagnant state of growth. Lagerlf missed the opportunity of
seeing it because he did not compare his model calculations with data. Artzrouni and Komlos (1985)
produced a distribution for the growth of the world population but did not notice that their model
generated exponential growth with no signs of stagnation. Furthermore, their results are contradicted
by data (Nielsen, 2016k) because the world growth of population was never exponential (Nielsen,
2016d, 2016j).
Lagerlf carried out Monte-Carlo calculations, which were supposed to confirm the existence of the
epoch of the so-called Malthusian stagnation in the growth of population allegedly caused by the effects
of demographic catastrophes, such as epidemics, wars and famines. He incorporated explicitly the
mechanism of stagnation in his model. Consequently, his model should have been expected to produce
the process of stagnation but it did not. Before the publication of Lagerlfs paper, data for the United
Kingdom, France and Sweden were already available (Maddison, 2001) but unfortunately Lagerlf did
not compare his model-generated calculations with these most essential data.
These data are shown in Figure 1. Their analysis demonstrates that data for the UK and France follow
hyperbolic trajectories. For Sweden, there was a change from a hyperbolic distribution to exponential
growth. All these data and their analysis demonstrate that there was no stagnation in the growth of
population and that contrary to the original assumption of Lagerlf, epidemics, wars and famines did
not shape the growth path of population. The past growth may have been slow but it was not stagnant.
It was slow because it was hyperbolic. It then became fast because it was hyperbolic. Only in Sweden
it was diverted to a faster new trajectory but it was not a transition from stagnation to growth but a
transition from growth to growth, from a hyperbolic growth to an exponential growth.
For France, the growth of population was following closely hyperbolic distribution at least until
around 2000. For the United Kingdom, the growth was hyperbolic until around 1820, when it started to
be diverted to a slower trajectory. According to the generally accepted interpretations of the mechanism
of the growth of human population, we should expect a significant boosting (takeoff or explosion)
around the time of the Industrial Revolution, 1760 and 1840, (Floud & McCloskey, 1994). This takeoff
should be clearly indicated in the United Kingdom, the very centre of this revolution, where its impacts
should have been most clearly demonstrated. The takeoff did not happen. On the contrary, in the direct
contradiction of these usually claimed expectations, the growth of population in the UK started to be
345
__________________________________________________________________________________
diverted to a slower trajectory at around 1820, right at the time when it was supposed to have been
boosted.
Figure 1. Population growth in the United Kingdom, France and Sweden. Data (Maddison, 2001) are
compared with hyperbolic distributions. Population growth was increasing monotonically. It was slow in the
past because it was hyperbolic. For Sweden, there was a change from a slow hyperbolic growth to a faster
exponential growth around AD 1600. There was no stagnation. The growth of population was not shaped by
demographic catastrophes, as claimed by Lagerlf (2003a).
In Sweden, the growth of population was boosted but it was boosted at a wrong time, around AD
1600, i.e. well before the Industrial Revolution. The boosted growth follows an exponential trajectory,
as indicated by the straight line in this semilogarithmic display.
Hyperbolic distributions displayed in Figure 1 are described by the following simple equation:
1
S (t ) = , (1)
a kt
where S (t ) is the size of the population, t is time, while a and k are the parameters determined by fitting
hyperbolic distributions to data.
For the hyperbolic distributions displayed in Figure 1, parameters are:=a 1.221 103 and
=k 6.511 107 for the UK,
= a 2.085 104 and
= k 9.635 108 for France and
= a 4.935 103 and
=k 2.221 106 for Sweden.
Exponential distribution describing the growth of population in Sweden from AD 1600 is given by
the following equation:
S (t ) = bert , (2)
346
__________________________________________________________________________________
with parameters= b 5.798 102 and = r 5.973 103 . From AD 1600, population in Sweden was
increasing at an approximately constant rate of 0.6%. Just before the transition to this new trend, the
growth rate for the preceding hyperbolic distribution was only 0.16%. The new exponential trajectory
was approximately 3.7 times faster than the preceding hyperbolic trajectory at the time of the transition.
Hyperbolic growth is often described as faster-than-exponential or hyper-exponential. Such
descriptions should be avoided. They are inaccurate and misleading. The concept of the faster-than-
exponential growth was introduced, or at least strongly promoted, by Bartlett (1993). However, he has
readily admitted that he was wrong: Thanks for your thoughtful analysis of my writing about faster
and slower than exponential. You are right! My wording is unclear and confusing and wrong. I have
used these terms for years and you are the first person to point out this error to me (Bartlett, 2011). If
only the erroneous concepts adopted in the economic and demographic research could be so readily
corrected we would see progress in these two fields of study, rather than the existing and long-lasting
stagnation.
In the example presented in Figure 1, exponential growth in Sweden after around AD 1600, is faster
than the preceding hyperbolic growth and thus, in this case, hyperbolic growth (the so-called faster-
than-exponential growth) is in fact slower than exponential.
We can only compare specific distributions and see which of them are faster or slower. Faster-than-
exponential distributions do not exist because we can always design exponential growth, which over a
certain time will be faster than some other incorrectly claimed faster-than-exponential growth.
If we have to use the expression faster-than-exponential we have to be specific. We have to
describe clearly, which specific distributions are being compared and over specifically what range of
the independent variable. Thus, for instance, for the distributions shown in Figure 1 for Sweden we
could say that over the range of the displayed time, the exponential growth, which commenced in 1600
was faster than the preceding hyperbolic growth. However, we obviously cannot claim that the
hyperbolic growth before 1600 was faster-than-exponential or hyper-exponential because we have
already demonstrated that in this particular case this so called faster-then-exponential or hyper-
exponential growth was obviously slower than the exponential growth, which replaced this hyperbolic
growth.
Lagerlf did not invent the concept of the so-called Malthusian stagnation, which is supposed to be
caused by the lethal effects of demographic catastrophes. He just accepted it without any criticism
maybe because going with the flow increases the chance of publishing new results. When in one of his
papers, published also in 2003, Lagerlf was associating the hypothetical epoch of the so-called
Malthusian stagnation with epidemic shocks he was quickly corrected by a referee for missing the
effects of wars: As suggested by a referee, this process could possibly be interpreted in terms of wars,
instead of epidemics (Lagerlf, 2003b, p. 766).
Both, Lagerlf and his referee were wrong. The process of the so-called Malthusian stagnation
cannot be interpreted in terms of wars, instead of epidemics because as shown by data presented in
Figure 1, the so-called Malthusian stagnation did not exist. However, neither Lagerlf nor his referee
cared to consult the relevant data. Data appear to be of lesser importance than the mantra of stagnation.
Unfortunately, this mantra is repeated without any convincing justification in the economic and
demographic research, and every effort is made to make sure that it is repeated faithfully and as required.
As mentioned earlier, Lagerlf did not compare his model-calculations with data, but his referee was
also misguided because the doctrine of the so-called Malthusian stagnation is repeatedly contradicted
by data (Nielsen, 2016a, 2016d, 2016j).
According to the established knowledge in demography and in economic research, The age of
pestilence and famines lasted until 1875 (Rogers & Hackenberg, 1987, p. 234) when there was
supposed to have been a transition from stagnation to a fast growth, the transition described usually as
a takeoff or explosion. It is unclear how this precise date was determined but it might have been
suggested by the generally accepted but erroneous notion of the alleged transition from stagnation to
growth around the Industrial Revolution, 1760 and 1840 (Floud & McCloskey, 1994). Analysis of data
shows convincingly that there was no stagnation in the growth of population and in the economic
growth, and that there was no transition, which could be described as a takeoff or explosion. What is
interpreted as an explosion is just the natural continuation of hyperbolic growth (Nielsen, 2016a, 2016d,
2016j).
347
__________________________________________________________________________________
The mythical age of pestilence and famines was supposed to have been characterised by what is
known as Malthusian oscillations. According to this doctrine periodic epidemics of plague, cholera,
typhoid and other infectious diseases would in one or two years wipe out the gains made over decades.
Over long periods of time there would, consequently, be almost no population growth at all (van de
Kaa, 2010, p. 87). The pattern of growth [of human population] until about 1650 is cyclic (Omran,
1971, Table 4, p. 533). Here we have a different date for the termination of the age of pestilence and
famine, which is hardly surprising because these dates are based on impressions combined with a good
dose of imagination. The age of pestilence and famine, with its assumed strong effects on the growth
of population and on the economic growth, did not exist.
The alleged, but non-existent, transition from stagnation to growth was supposed to have been
associated with the transitions in the birth and death rates. It is interesting, however, that while Omran
shows examples of the claimed transitions in birth and death rates, his examples (for Sweden, England,
Japan, Ceylon and Chile) show clearly and convincingly that these transitions had absolutely no impact
on the growth of population (Omran, 2005).
Changes in birth and death rates are not necessarily reflected in changes in the growth of population.
The growth of population is not determined by the birth and death rates alone but the average difference,
i.e. by the average gap, between these two quantities. Birth and death rates might be changing from
high to low but such changes will not be reflected in the growth of population unless the average
difference between them is also changing. Furthermore, small changes in the difference between birth
and death rates are also not reflected as the associated changes in the growth of population (Nielsen,
2016l).
Birth and death rates might be decreasing but if they are decreasing in such a way that the average
difference between them is approximately constant, the growth of population will be approximately
exponential. If the difference increases systematically, then the growth of population will be described
by a non-exponential trajectory. For instance, if the difference increases, on average, hyperbolically,
then the growth of population will be hyperbolic.
To produce stagnation, the average difference between birth and death rates has to be approximately
zero. To produce a stagnant but slowly increasing population, the average difference between birth and
death rates would have to be changing in a very specific and complicated way. It would have to be on
average zero over a long time but then it would have to be on average non-zero to generate growth.
Then again it would have to revert back to zero to produce stagnation. This process would have to be
repeated over a long time for thousands of years. We do not have data to demonstrate that such a process
ever existed. We do not have data for birth and death rates extending over thousands of years. The claim
that birth and death rates were high and that they were producing stagnation is unscientific because we
do not have data to prove it. However, we have a large body of data describing the growth of population
and we can show that the growth of population was in general hyperbolic, which by inference means
that the average difference between birth and death rates was increasing hyperbolically. There was no
stagnation.
To demonstrate a dramatic change from stagnation to growth in the growth of population we would
have to demonstrate that there was a dramatic change in the average difference between birth and death
rates from zero to a clearly and systematically larger value, and it does not matter whether birth and
death rates were decreasing or increasing. What determines the growth of population is the average
difference between birth and death rates.
If we want to claim some kind of transitions in birth and death rates, there is nothing to stop us from
doing it. However, we should remember that, in general, such studies will not help us to understand the
mechanism of growth of population. They might be interesting and stimulating for another reason but
unless we pay close attention to how the difference between these two quantities is changing, how large
or how systematic are these changes, and how these changes can be explained, we shall not explain the
mechanism of growth of human population. We also should remember that only significant changes in
the difference between birth and death rates are reflected as changes in growth trajectories.
The frequently used example in support of the concept of stagnation followed by explosion is the
growth of population in Sweden between around AD 1750 and 2000. It shows changes in the difference
between birth and death rates but no-one seems to have noticed that these changes are relatively small.
Furthermore, no-one seems to have noticed that these small changes are not reflected in the growth of
population, even though data for the birth and death rates and for the growth of population come from
348
__________________________________________________________________________________
exactly the same source (Statistics Sweden, 1999). These data are selectively and consistently ignored
in order to preserve the perfect intonation of the mantra of the so-called Malthusian stagnation.
Small changes in the average values of birth and death rates are repeatedly but incorrectly interpreted
as a proof of the existence of the epoch of the so-called Malthusian stagnation and of a transition from
stagnation to growth while data presented in the same primary source show clearly that the growth of
population in Sweden was increasing monotonically without any signs of stagnation and without any
sign of a transition from stagnation to growth. These issues were discussed earlier (Nielsen, 2016l).
Demographic research concentrating on the study of birth and death rates might be important but it
is incorrect to think that such a research can be necessarily useful for explaining the mechanism of
growth of human population. The two mechanisms are related only via the average difference between
birth and death rates. There might be strong fluctuations in birth and death rates but these fluctuations
are generally not reflected in the growth of population. They might be reflected only as minor variations
in the growth trajectory describing the growth of population.
In conformity with the established knowledge, Komlos claimed that Malthusian positive checks
(mortality crises) maintained a long-run equilibrium between population size and the food supply
(Komlos, 1989, p. 194). Here we have a hinted link to the specific type of demographic catastrophes:
famines. He also claimed that the food-controlled homeostatic equilibrium had prevailed since time
immemorial (Komlos, 2000, p. 320). Komlos appears to have been guided by the generally accepted
consensus. However, science never relies on any generally accepted consensus. It is not unusual in
science to show that the generally accepted consensus is scientifically unacceptable.
The postulate of the so-called Malthusian stagnation in the economic growth and in the growth of
human population, as well as all other related postulates, are scientifically unacceptable. because they
are systematically contradicted by data (Nielsen, 2014, 2016a, 2016b, 2016c, 2016d, 2016e, 2016f,
2016g, 2016h; von Foerster, Mora & Amiot, 1960). Growth of population, global or regional, was
hyperbolic, (Nielsen, 2016d, 2016j; von Foerster, Mora & Amiot, 1960). Economic growth was also
hyperbolic (Nielsen, 2016a).
In the case of the growth of human population we can extend our study to 10,000 BC. It is
remarkable, that over the past 12,000 years the growth of population was not only hyperbolic but also
exceptionally stable (Nielsen, 2016j) because over this long time there was only one major transition
around AD 1 from a fast to a slow hyperbolic trajectory. There was also another but only minor
transition around AD 1300 from a slower to a slightly faster hyperbolic growth. Currently we are
experiencing a new transition to a yet unknown trajectory but the growth is still close to the historical
hyperbolic trajectory.
We can extend the analysis of the growth of population even further, over the past 2,000,000 years
and show that the growth was hyperbolic (Nielsen, 2017). There is nothing in the data to support the
claim that Throughout human history, epidemics, wars and famines have shaped the growth path of
population (Lagerlf, 2003a, p. 435).
We already know that Malthusian positive checks, which include demographic catastrophes, trigger
the process of regeneration (Malthus, 1798; Nielsen, 2016i). This process alone, explains the
remarkable stability of the growth of human population. However, in order to understand even better
why demographic catastrophes had generally no impact on the growth of population we shall now
investigate their relative intensity and other parameters defining their possible impact.
3. Preliminary remarks
Impacts of demographic catastrophes depend on the death toll, their duration and on the size of
population. Death toll for a given demographic event might be high but to understand its impact we
have to express it as the relative impact by comparing the death toll with the size of population, which
could be the size of local population directly affected by a demographic crisis or it could be the size of
a regional or global population, depending on whether we are interested in the study of local, regional
or global impacts.
Impacts of demographic catastrophes depend also on the historical time. In the distant past, when
the population was small, local impacts of demographic catastrophes could be large. However, people
were living in greater isolation so the global or even regional impacts could have been small. Likewise,
at the other end of the historical time scale, when the population increased to a certain large size, relative
349
__________________________________________________________________________________
impacts were small even if the number of people killed by a given demographic catastrophe was large.
It can be, therefore, expected that there is only a relatively small window of time, mainly during the AD
era until around 1800, when the global population reached its first billion, or maybe until around 1900,
that the demographic catastrophes could have had a noticeable impact on the growth of population.
However, the study of human population shows that in general they had no damaging impact, with the
exception of the already mentioned minor disturbance around AD 1300 (Nielsen, 2016j).
The further we go back in time with our investigation the less we know about the intensity of
demographic catastrophes but we have enough information for the AD era to assess their possible
impacts.
In order to understand human population dynamics, it is essential to identify the main and the most
obvious driving force of growth and add to it any other force or forces only if the assumed main force
cannot explain growth. The fundamental force of growth of human population is obviously the force of
procreation expressed as the difference between the biologically-controlled force of sex drive and the
biologically-controlled process of aging and dying. This force cannot be dismissed and it turns out that
this force alone explains why the spontaneous and unconstrained growth of human population is
hyperbolic and why for the most part of the past human history it was hyperbolic (Nielsen, 2016m).
In the past 12,000 years, other forces were playing a significant role only during the major
demographic transition around AD 1 and during the minor transition around AD 1300. They are also
strong and active during the current transition. With the exception of these rare events, the growth was
hyperbolic in the past 12,000 years. Furthermore, with the exception of the minor disturbance around
AD 1300, there is no evidence that demographic catastrophes were ever shaping the growth of human
population (Nielsen, 2016d, 2016j).
It should be also noted that the recorded impacts of demographic catastrophes are likely to be
exaggerated. Recorded death rates are largest when the supporting evidence is skimpiest. When data
are better, the death rates are usually lower and the percentage increases less (Watkins & Menken,
1985, p. 651). For instance, both Durand (1960) and Fitzgerald (1936, 1947) claim that impact of the
An Lu-Shan Rebellion (AD 756-763) is probably exaggerated. Likewise, Russel (1968) and Twigg
(1984) believe that the number of casualties caused by the Justinian Plague (AD 541-542) is also grossly
overestimated.
Another example is the Antonine Plague (AD 166-270), which was first estimated to have killed
about 50% of the population of the Roman Empire (Seeck, 1921). However, this estimate was later
downgraded to 1-2%, or to the total number of casualties of 500,000-1,000,000 (Gilliam, 1961) and
then upgraded to 7-10% or to a maximum of 5 million (Littman & Littman, 1973), the last estimate
being still significantly smaller than the original estimate. It appears that the further back in time we go
the larger is the possibility of exaggerated claims of the number of casualties.
We shall describe demographic catastrophes in the way they are reported in the literature. However,
labelling them with just a single cause might not be accurate. For instance, a war considered as the main
cause of a crisis might include famine but famine might be linked with pestilence. For example, during
the Madras famine in the 1870s, about 40% of casualties were caused by smallpox and cholera
(Lardinois, 1985). The Justinian Plague was also accompanied by smallpox, diphtheria, cholera and
influenza (Shrewsbury, 1970) and was perhaps aided by wars, famines, floods and earthquakes Scott
and Duncan (2001, p. 5). Likewise, a number of epidemics in France were preceded by famine,
sometimes in conjunction with bad weather conditions (Scott & Duncan, 2001, p. 105) whereas
frequent and virulent outbreaks in France during 1520-1600 were accompanied by food shortages,
famines, flooding, peasant uprisings and religious wars (Scott & Duncan, 2001, p. 291).
While drawing from primary sources about the frequency and intensity of demographic catastrophes,
the presented here survey has been also assisted by some useful compilations (Austin Alchon, 2003;
Kohn, 1995; Spignesi, 2002; White, 2011).
4. Examples of prominent demographic catastrophes

One of the earliest recorded devastating plagues was the Asiatic disease identified now as
tularaemia, a bacterial disease caused by Francisella tularensis, first recorded around the early 1700s
BC. It spread over a large area between Cyprus and Iraq and between Palestine and Syria. This disease
appears to be also the first recorded example of the use of biological weapon when it was introduced
350
__________________________________________________________________________________
deliberately to Anatolia (Trevisanato, 2004, 2007). The same disease has been also probably recorded
in the Bible as causing a great number of deaths among Philistines in the city of Ashdod, the event dated
either to around 1000 BC (Khan, 2004) or to 1320 BC (Cunha & Cunha, 2006).
Early recorded plagues include also a viral haemorrhagic fever in Egypt between 1500-1350 BC
(Duncan & Scott, 2005) but it might have been the same disease as recorded earlier in Egypt and the
same plague that decimated Philistines. Incidentally, Duncan and Scott (2005) claim that Black Death
was not a bubonic plague caused by bacterium Yersinia pestis, as traditionally claimed, but rather that
is was a viral haemorrhagic fever, which according to them includes also the plagues of Mesopotamia
(700-350 BC), the Plague of Athens (430-427 BC), the Plague of Justinian (AD 541-542), Plagues of
Islam (AD 627-744), plagues in Asia minor (1345-1348), and the plague of Denmark and Sweden
(1710-1711).
The epidemic of Athens (460-399 BC) is claimed to have killed 25% of Athenian army and a great
number of civilians (Austin Alchon, 2003). It created a turning point in the history of Greece (Ross,
2008). It is also claimed that this plague killed 50% of the army of Pericles and 50% of the navy coming
to the rescue from Piraeus (Beran, 2008). The plague was triggered by the overcrowding of Athens
when Spartans attacks prompted rural population to seek shelter in that city, which was already housing
a relatively large number of people, an estimated 300,000 citizens and around 3 million slaves.
The earliest large demographic catastrophe in the AD era appears to have been associated with the
Red Eyebrows Revolt, which commenced around AD 2. The estimated size of Chinese population at
that time is claimed to have been 59.6 million but it might have been reduced to 21 million in AD 57
(Durand, 1960). However, Durand also discusses possible inaccuracies in these estimates and presents
corrected numbers of 74 million in AD 2 and 45 million in AD 88, for the entire Chinese Empire. He
also estimates 71 million and 43 million, respectively, for the China proper (Durand, 1960, p. 221). By
China proper he means the current 18 provinces. He uses this estimate in his graph (Durand, 1960, p.
247). In both cases, the relative death toll is approximately 39% of the original population in China but
only a maximum of 12% of the global population, too weak to produce any noticeable impact.
The Red Eyebrows Revolt and the associated dramatic decrease in the size of population in China
was in the middle of a massive demographic transition, one and only major demographic transition in
the past 12,000, a transition from a fast to a slow hyperbolic trajectory, the transition which lasted for
approximately 1000 years. This transition is shown in Figure 2. The dramatic event in China had no
impact on the growth trajectory of the world population.
Durand points out also that estimates of the size of the population at the time of demographic
catastrophes might be inaccurate. Even if such huge loss were conceivable, it would be nave to
suppose that accurate count of the survivors could have been carried out in the midst of the ensuing
chaos (Durand, 1960, p. 224). White (2011) attributes only 10 million of casualties to the Red
Eyebrows Revolt. However, to estimate the impact of this demographic catastrophe we shall use the
revised estimate of Durand (1960) representing the total death toll of 29 million over 87 years.
Similar uncertainty in the estimated death toll applies also to the An Lu-Shan Rebellion (AD 756-
763). Acceptable records appear to show the death toll of 36 million but White (2011) attributes only
13 million.
Between A.D. 705 and 755 to all appearances the census machinery functioned much more
effectively; but after 755 it broke down again. The recorded number of persons dropped from nearly
53 millions in the year 755 to only 17 millions in 760. During this time, China was torn by revolts
which were suppressed with bloody force, including the notorious rebellion of An Lu-Shan. Many
historians have affirmed that 36 million lives were lost as a result of these violent events, but
Fitzgerald and others have shown that this is incredible (Durand, 1960, p. 223; Fitzgerald, 1936,
1947).
In order to maximise the possible impact of this demographic crisis, we shall assume that the death
toll was 36 million.
The impact of the Plague of Justinian is hard to estimate because of the incomplete information
combined with conflicting claims. The plague is claimed to have reduced the population of
Constantinople by 40% between AD 541 and 542 (Austin Alchon, 2003). Cunha and Cunha (2006)
estimated a 30% reduction of the population of the Roman Empire between AD 542 and 590, or a
maximum of about 14 million out of the total of 48 million (Maddison, 2006; Seeck, 1921). The plague
so weakened the Roman Empire that not long after the plague had passed, Roman borders were overrun
351
__________________________________________________________________________________
by Huns, Goths, Moors, and other barbarians (Cunha & Cunha, 2006). Rosen (2007) estimates that
this plague killed 25 million people in a short time of only between AD 541 and 542. In around AD 549
the same plague emerged also in Britain (Carmichael, 2009). It is also claimed that The Plague of
Justinian recurred in discernible cycles of about nine to twelve years (Dols, 1974, p. 373).
Figure 2. Demographic transition in the growth of global population around AD 1. The transition can be
described by the monotonically changing distribution (Nielsen, 2016j, 2017). Rebellion of An Lu-Shan, which
caused a massive reduction in the size of population in China, had no impact on the growth of global
population. For the reference to the sources of data and for the description of their analysis see Nielsen (2016j,
2017).
There is also one claim, which is distinctly different than all other estimates. Assisted by the Eurasian
Silk Road, this plague was supposed to have spread to China in around AD 610, (Ross, 2008) continuing
its devastation until around AD 700 (Duncan & Scott, 2005) and killing probably a maximum of about
100 million people (Ross, 2008), which would represent a 50% reduction in the world population. Even
if we consider the regenerating effects of Malthusian positive checks (Malthus, 1798; Nielsen, 2016i),
such a huge reduction should be reflected in the growth trajectory but it is not. By AD 500, the growth
of the world population was at the end of its transition (see Figure 2) and commenced its new hyperbolic
trend. In AD 500, the estimated size of the world population was only 190 million (Nielsen, 2016j and
references therein). The claimed massive death toll of 100 million was supposed to have occurred
between AD 610 and 700, i.e. when the growth of the world population settled already along a new
hyperbolic trajectory, but we see no sign of such a disturbance. This claim of such a large death toll is
almost certainly incorrect.
In our survey, in order to maximise the evidence in favour of the concept of the so-called Malthusian
stagnation, we are considering the strongest impacts, which for the Plague of Justinian appears to be
the death toll of 25 million in a very short time, between AD 541 and 542. We shall see later that, under
this assumption, this plague had the strongest overall impact of all demographic catastrophes ever
recorded, as manifested by four out of five indicators, and yet it caused no noticeable disturbance in the
growth of the world population (see Figure 2).
Black Death (1343-1351) is another example of a massive demographic catastrophe and is claimed
to have killed over 60% of the urban population in Asia, about 30% of the population of the Middle
East and 30-60% of the population of Europe (Hawas, 2008). Beran (2008) claims that in many cities
the death toll was over 90%, creating a severe hardship for the surviving population and adding to the
total death toll caused also by the lack of food and lack of access to safe drinking water. The decaying
corpses were also reducing the chance of survival. About 20% of the population of England died
352
__________________________________________________________________________________
between AD 1348 and 1350 and a total of 50% by AD 1400 (Gilliam, 1961). Depending on the affected
area, mortality rates varied between 25% and 70% (Cunha & Cunha, 2006). In terms of the total and
relative death toll, Black Death was the greatest single demographic catastrophe ever recorded.
As mentioned earlier, plagues were also used as biological weapons by employing a gruesome
practice of catapulting infected corpses at the walls of fortifications or hurling them over the walls by
using trebuchets. This ghastly method was used by Greeks, Romans and other attackers between 300
BC and AD 1100, and by Tartars in AD 1346 against the residents of Genoa (Cunha & Cunha, 2006;
Khan, 2004).
Other examples of large local casualties caused by demographic catastrophes include smallpox in
Japan (AD 812-814), killing about half of the population of that country (Austin Alchon, 2003); the
1696 famine killing between 25% and 30% of the population of Finland (Jutikkala, 1955); the 1770
famine in Bengal, killing about 30% of the population (or a total of 10 million) and the 1376 famine in
Italy, killing 60% of the population (Ghose 2002; Keys, Brozek, Henschel, Mickelsen & Taylor, 1950;
Walford, 1878).
According to Mallory (1926), 18 provinces of China experienced 1015 draughts between AD 620
and 1619, or about one per year. However, they were unevenly distributed, illustrating that while the
number of casualties and impacts of demographic catastrophes might be high in small and isolated
regions, their effects could be much less severe when averaged over a larger number of population.
There was a total of 443 draughts in the Northern Division, 352 in the Central Division and 220 in
the Southern Division. However, even within the same division, the number of draughts varied
significantly between various districts. For instance, in the Northern Division, Honan District
experienced a total of 112 draughts but Kansu Division only 4. In the Central Division, the largest
number of draughts (113) was in the Chekiang District and the smallest (28) in the Anhwei District. In
the Southern Division, the number of draughts varied between 4 and 59 per district.
The list of significant lethal events in China includes: 60-70% of troops killed during a single
military engagement in AD 16; 70% of Mongolians killed by hunger in AD 46; 30-40% of troops killed
in AD 162; about 70% of troops killed in a single military engagement and by famine and epidemic;
close to 100% killed by locusts and famine in AD 312 in the northern and central China; over 30%
killed in Shantung in AD 762; over 50% in Chekiang in AD 806; 30-40% in Hupeh, Kinagsu and Anhui
in AD 891; 90% in Hopei in 1331; 50% of troops between 1351-1352; over 70% in Shansi in AD
135; 60-70% in Hupeh in 1354, and 100% in various towns and villages in Hunan in 1484 (Austin
Alchon, 2003; McNeill, 1976)
It is claimed that in Mexico, 25-50% of the population died of smallpox (1520-1521), 60-90%
probably of typhus (1531-1532), and over 50% of either the bubonic plague or typhus between 1576
and 1581 (Austin Alchon, 2003; Motolina aka Fray Toribio de Benavente o Motolina, 1971; del Paso
y Troncoso, 1940; Prem, 1992).
The estimated death toll in the Andes between 1524 and 1591 includes 30-50% by smallpox (1524-
1527), 25-30% by measles or bubonic plague (1531-1533), 15-20% by influenza, measles and smallpox
(1558-1559), and about 50% by influenza, measles, smallpox and typhus between 1585 and 1591
(Cook, 1981; Dobyns, 1963). Dobyns (1993) gives also many examples of large death tolls, sometimes
as high as 98% but most often close to 80-90%, caused by diseases among Native American population.
So, it appears that humans always lived with the threats and with deadly effects of demographic
catastrophes strong enough to reduce often substantially the size of local populations. We shall now
investigate their potential impact on the growth of the world population.
4. Indicators of impact
In order to study the potential impacts of demographic catastrophes we have to introduce a few
useful gauge indicators. Their definition is assisted by the diagram presented in Figure 3.
Before the onset of a demographic catastrophe, population increases along the trajectory f (t ) . It
reaches the size S0 at the time t0 , which marks the beginning of the demographic catastrophe.
Demographic crisis lasts for number of years, between the time t0 and t1 . Depending on the intensity
of the demographic catastrophe and on the efficiency of the process of regeneration (Malthus, 1798,
Nielsen, 2016i), the growth of the population during the demographic crisis may be diverted to a new
353
__________________________________________________________________________________
trajectory h(t ) , which might be still increasing, remain constant or decreasing. At the end of crisis, the
size of the population is S r while S1 , is the size of the population, which would have been reached if
the crisis did not occur.
When the crisis is over, the growth of population continues along a new trajectory g (t ) The quantity
A is the death toll and T is the recovery time, i.e. the time required for the population to reach the size
S1 .
Figure 3. Schematic diagram describing the composition of a demographic catastrophe. The leading
parameters are: the duration of demographic catastrophe; T the recovery time; A the death toll.
The recovery time depends on the growth rate, R, during the time of crisis. Over a relatively small
span of time associated with demographic catastrophes we can use linear approximations of the relevant
trajectories.
1 dg (t ) 1 A
R , (3)
S r dt Sr T
which gives
a
T = , (4)
R
where
354
__________________________________________________________________________________
A A
a (5)
Sr S0
is the relative impact, i.e. the number of people killed by the demographic catastrophe as compared
with the size of the population at the onset of crisis.
The growth rate R can be estimated by examining the population data around the time of crisis while
the quantity a can be easily calculated using the reported number of people A killed by the crisis and
the estimated size of the population at the beginning of crisis. Using these readily accessible quantities
we can then calculate the recovery time T, which together with a (the relative number of people killed
during demographic crisis) will help to gauge the intensity of the demographic catastrophe.
Another way of calculating the recovery time T is to use the exponential rather than linear
approximation for the function g (t ) . Under this assumption and using the well-known expression for
the exponential function [eqn (2)] we can easily show that
1 S1 1
T = ln = ln(1 + a ) (6)
R S r R
This is a general formula that does not have to be related to a demographic crisis. It is simply a
formula for calculating the time needed for the exponential growth to increase from S r to S1 , which
happens to be precisely what we want to use to calculate the recovery time. For small a, the recovery
time calculated using eqn (6) is virtually the same as by using the eqn (4). Thus, for instance, for
a = 20% , the recovery time calculated using eqn (6) is only 10% smaller than using eqn (4). For lower
values of a, the discrepancy is even smaller. It increases to 23% for a = 50% . As we shall soon see, in
our survey of demographic catastrophes we shall be dealing with a values of up to only 20%.
If we use the hyperbolic approximation, then referring to the eqn (1), the recovery time is given by
A
T= . (7)
( S r + A)kS r
If we want to use an approximate expression incorporating the relative impact a, then using the eqn
(1) and (5) we get
a2
T= . (8)
A(1 + a)k
So now rather than using just the parameter A (the total death toll) we have two additional gauge
indicators, a and T, the parameters that give us additional information about the intensity of crisis.
However, we can also introduce yet another useful gauge indicator, which compares the recovery
time with the duration of the demographic catastrophe. We shall call it the intensity indicator (I) and
we shall define it simply as
T
I . (9)

If the recovery time T is large when compared with the duration of crisis, then we are dealing with
a potentially strong demographic catastrophe. The larger was the recovery time compared with the
duration of crisis the stronger was the devastating impact of crisis.
However, there is also another hidden information in this indicator. Using the diagram presented in
Figure 3 and assuming that gradients of functions f (t ) and g (t ) are approximately the same, we can see
that
355
__________________________________________________________________________________
A
I . (10)
S1 S 0
If A < S1 S0 , then I < 1 , which then indicates that population size continued to increase during
crisis. If A S1 S0 , then I 1 , which then indicates that the size of population remained approximately
constant during crisis. If A > S1 S0 , then I > 1 , which indicates that the size of population was
decreasing during crisis.
Thus, by looking at the I indicator we can tell not only whether the crisis was weak or strong but
also whether the population was still increasing, remained constant or decreasing during the crisis.
However, even if I > 1 it does not mean that we a dealing with a potentially strong crisis, because
depending on the duration of crisis the size of the population could still remain approximately constant.
A potentially strong demographic crisis will be characterised by I 1 . A guide to the intensity indicator
is presented in Table1.
Finally, we can also introduce two other indicators: the per annum relative impact ( ) and the per
annum intensity indicator ( ).
a
, (11)

I
. (12)

The complete list of indicators used to evaluate the effects of demographic catastrophes is presented
in Table 2.
Table 1. A guide to the interpretation of the intensity indicator (I)
I <1 Weak crisis. Population continues to increase during crisis.

I 1 Moderate crisis. Population size remains approximately constant during crisis.
I >1 Moderate or potentially strong crisis depending on the value of I. Population size
decreases during crisis.
I 1 Potentially strong crisis
Table 2. Indicators used to process information about demographic catastrophes
Symbol/Definition Description Unit

t0 The onset of crisis year
A Death toll 106
Duration of crisis year
a = A / S0 The fraction of the population killed by crisis %
= a / Per annum fraction of population killed by crisis %/year
T Recovery time year
t2 The year of the full recovery year
I = T / Intensity indicator
= I /
year 1
The per annum intensity indicator
Demographic catastrophe may be considered potentially strong if indicators a = A / S0 , = a / , T, I = T / and

= I / are large.
356
__________________________________________________________________________________
5. Evaluation of impacts of demographic catastrophes

The survey of demographic catastrophes and their impacts is presented in Table 3. The summary of
all impacts is shown in Table 4. In order to maximise evidence in favour of the postulate of the so-called
Malthusian stagnation we have considered only the most significant demographic catastrophes
characterised by the death toll of A 1 million. Had we included smaller demographic catastrophes, the
fraction of potentially strong catastrophes, which could have had noticeable impact on the growth of
the world population, would have been significantly reduced.
Table 3. Survey of major demographic catastrophes AD 1-1900. The most significant values of gauge
indicators are indicated by bold characters and moderately significant by italics. (Symbols are explained in
Table 2.)
Event t0 t2 A a T I
Red Eyebrows Revolt 2 245 29.0 87 11.5 0.13 157.5 1.8 0.02
Antonine Plague 166 214 5.0 15 2.2 0.15 34.1 2.3 0.15
Plague of Justinian 541 756 25.0 2 12.5 6.23 214.2 107.1 53.54
An Lu-Shan Rebellion 756 845 36.0 8 15.4 1.93 227.7 28.5 3.56
N. Egypt Earthquake 1201 1206 1.5 1 0.5 0.46 4.8 4.8 4.80
Mongolian Conquest 1260 1405 40.0 35 11.3 0.32 110.6 3.2 0.09
Great European Famine 1315 1336 7.5 3 2.0 0.68 19.1 6.4 2.13
Famine in China 1333 1369 9.0 15 2.4 0.16 21.8 1.5 0.10
Black Death 1343 1530 75.0 9 19.7 2.19 178.7 19.9 2.21
Fall of the Yuan Dynasty 1351 1385 7.5 18 1.9 0.11 17.4 1.0 0.05
Sweating Sickness 1485 1556 3.0 67 0.6 0.01 4.6 0.1 0.00
Mexico Smallpox Epidemic 1520 1527 4.0 2 0.8 0.42 6.0 3.0 1.50
French Wars of Religion 1562 1602 3.0 37 0.6 0.02 3.7 0.1 0.00
Russias Time of Trouble 1598 1619 5.0 16 0.9 0.06 5.6 0.3 0.02
Fall of the Ming Dynasty 1618 1669 25.0 27 4.3 0.16 25.2 0.9 0.03
Thirty Years War 1618 1655 7.0 31 1.2 0.04 7.0 0.2 0.01
Deccan Famine in India 1630 1633 2.0 2 0.3 0.17 2.0 1.0 0.50
Famine in France 1693 1696 2.0 2 0.3 0.15 1.5 0.8 0.38
Bengal Famine 1769 1778 10.0 5 1.2 0.23 4.7 0.9 0.19
Napoleonic Wars 1803 1816 4.0 13 0.4 0.03 1.4 0.1 0.01
Famines in China 1810 1819 22.5 2 2.3 1.13 7.9 3.9 1.96
Great Irish Famine 1845 1850 1.0 6 0.1 0.01 0.2 0.0 0.01
Famine in China 1846 1849 11.3 1 1.0 0.96 2.8 2.8 2.83
Taiping Rebellion 1850 1868 20.0 15 1.6 0.11 4.5 0.3 0.02
Famine in India 1866 1866 1.0 1 0.1 0.08 0.2 0.2 0.20
Famine in Rajputana 1869 1869 1.5 1 0.1 0.11 0.3 0.3 0.29
Famine in Persia 1870 1871 2.0 2 0.1 0.07 0.4 0.2 0.10
Famine in N. China 1876 1880 13.0 3 0.9 0.31 2.3 0.8 0.25
British India Famine 1876 1903 17.0 25 1.1 0.05 2.6 0.1 0.00
Yellow River Flood 1887 1887 2.0 1 0.1 0.13 0.3 0.3 0.31
Famine in India 1896 1902 8.3 6 0.5 0.08 1.1 0.2 0.03
The remarkable feature of this survey is that, in general and as revealed by the values of the
introduced gauge indicators, even large catastrophic events had much smaller impact on the growth of
the world population than it might have been expected by looking just at the death toll or at their reported
local impacts. Indeed, we only have a few of events that might have had a tangible impact, and they are
all clustered around the early years of the AD era when the estimates of the total number of casualties
were probably grossly exaggerated (Durand, 1960; Fitzgerald, 1936, 1947; Gilliam, 1961; Littman &
Littman, 1973; Russel, 1968; Twigg, 1984, Watkins & Menken, 1985).
The leading indicator is the relative impact a because it gives the direct information about how the
growth trajectory might have been affected by a given individual demographic catastrophe. Events for
which a is less than or equal to around 10% can be ignored, because such displacements would be
hardly noticeable on the trajectories describing the growth of population. The corresponding
demographic catastrophes could be described as negligible. Even events with a up to around 20% could
be expected to have only relatively small effect. However, in this survey we have two events (An Lu-
Shan Rebellion and Black Death), with the relative impact of 15.4% and 19.7%, which we shall describe
as having a potentially strong impact. They account for only 6% of all impacts. Thus 94% of all large
demographic catastrophes, i.e. catastrophes with A 1 million, were individually too weak to have a
significant impact on the growth of the world population.
357
__________________________________________________________________________________
Table 4. Summary of impacts of demographic catastrophes.
Indicator Impact Number Fraction Insignificant

of Events of Total [%]
[%]
Relative impact (a) Strong 2 6
Moderate 3 10
Negligible 26 84 94
Per annum relative impact ( ) Strong 1 3
Moderate 2 6
Negligible 28 91 97
Recovery time (T) Strong 5 16
Moderate 5 16
Negligible 21 68 84
Intensity indicator (I) Strong 1 3
Moderate 11 35
Negligible 19 62 97
Per annum intensity indicator ( ) Strong 1 3
Moderate 7 23
Negligible 23 74 97
The average of all five Strong 2.0 6.5
Moderate 5.6 18.1
Negligible 23.4 75.4 93.5
The attribute described as strong should be interpreted as potentially strong or the strongest of all impacts. This
attribute does not identify impacts, which had a strong impact on the growth of population but only impacts, which
were potentially strong enough to have a noticeable impact.
We should remember, however, that we are ignoring the spontaneous process of regeneration
(Malthus, 1798; Nielsen, 2016i). By describing a crisis as strong we are only distinguishing it from
other catastrophes. A strong crisis is only relatively strong or potentially strong. It is a crisis, which
could have been reflected in the growth of population but considering the ever-present mechanism of
regeneration its impact is likely to be significantly reduced.
If we consider the per annum impact measured by the indicator , we can see that there was
possibly only one event (Plague of Justinian) that might have had a relatively strong impact on the
growth of the population and two (An Lu-Shan Rebellion and Black Death) that might have had a
marginal impact. Thus, when measured by this indicator, 97% of all large demographic catastrophes
had insignificant effect on the growth of the world population.
The recovery time (T) shows five significant events (Red Eyebrow Revolt, Plague of Justinian, An
Lu-Shan Rebellion, Mongolian Conquest and Black Death). For all of them, the estimated recovery
time was between around 100 and 200 years. They represent 16% of all demographic catastrophes, the
largest fraction in this survey. However, even for this indicator, the fraction of negligible events is high,
84%. The majority of all large critical events could have potentially inflicted only negligible impact on
The intensity indicator (I) suggests only one prominent event (Plague of Justinian) and possibly 11
moderately strong events. This indicator, therefore, shows that 97% of all large demographic
catastrophes could have had, at best, only small impact on the growth of the world population. For the
per annum intensity indicator ( ), the fraction of insignificant impacts is the same, 97%.
If we consider the average values of all five indicators we can see that only 6.5% of all demographic
catastrophes with the death toll larger or equal to 1 million might have had a tangible impact on the
growth of the world population. The remaining 93.5% were too weak to have any significant impact. It
is, therefore, clear that demographic catastrophes were too weak to shape the trajectory of growth of the
world population, particularly if we consider that demographic catastrophes trigger also a strong process
of regeneration (Malthus, 1798; Nielsen, 2016i).
The generally large percentage of insignificant impacts is an overwhelming evidence contradicting
the concept of the so-called Malthusian stagnation but confirming conclusions based on the analysis of
distributions describing the growth of population and the economic growth (Nielsen, 2016a, 2016d,
2016j), the analysis showing the absence of convincing evidence of frequent impacts of demographic
catastrophes.
358
__________________________________________________________________________________
It is also useful to notice certain correlations between gauge indicators because such correlations
could give a closer insight into the process of demographic catastrophes. They can reveal what was
happening during a given crisis. Thus, for instance, the intensity indicator (I) for the Mongolian
Conquest shows that population was decreasing during this crisis but the per annum intensity indicator
shows that the population was approximately constant. The intensity indicator was also not excessively
large. The size of the population was decreasing but slowly. However, the recovery time was
exceptionally high. We can explain it by noticing that the duration of the crisis was long.
Our survey shows also a unique convergence of five demographic catastrophes. They were: the
Mongolian Conquest (1260-1295) with the total estimated death toll of 40 million; Great European
Famine (1315-1318), 7.5 million; the 15-year Famine in China (1333-1348), 9 million; Black Death
(1343-1352), 75 million; and the Fall of Yuan Dynasty (1351-1369), 7.5 million. Their combined
maximum death toll was 139 million. The estimated size of the world population in AD 1250 was
around 380 million. The combined maximum relative impact of these five catastrophes was, therefore,
around 37%. Such a strong impact should be reflected in the growth of the world population and indeed
it was but not as strongly as we could have expected (Nielsen, 2016j). It caused only a minor
disturbance. During this crisis, the population was decreasing but very slowly to reach 360 million at
the termination of these five catastrophes, illustrating the efficient process of regeneration even during
this combined crisis. This crisis was followed by a faster growth and the lost time was soon recovered,
the faster growth illustrating again the efficient process of spontaneous regeneration (Malthus, 1798;
Nielsen, 2016i).
Before the crisis, the growth of population was following hyperbolic trajectory characterised by
= k 3.448 103 . If continued undisturbed, it would have reached the size S1 = 470 million in around
AD 1400. However, the actual size, S r , at that time was 360 million. If the growth of population after
the crisis continued along the same hyperbolic trajectory as before the crisis, then the recovery time,
calculated using the eqn (7), would have been 224 years. However, after the crisis, the growth of
population was following a faster trajectory, characterised by = k 4.478 103 . So, if we use the eqn
(7) again we can calculate that the corresponding recovery time for this faster trajectory was 173 years.
The actual recovery time, as recorded by data, was around 165 years, which is in good agreement with
the calculated value. The process of regenerations decreased the recovery time by 50-60 years.

The study presented here adds to the explanation why demographic catastrophes did not shape the
growth of population and the associated economic growth.
The currently accepted interpretation of the historical growth of population is succinctly summarised
in the following statement: Throughout human history, epidemics, wars and famines have shaped the
growth path of population (Lagerlf, 2003a, p. 435). If such is the case we should have no problem
with showing many examples of this mechanism but we cannot find them. We can analyse data going
as far back as 2,000,000 years ago and we can see that with the exception of just one minor disturbance
around AD 1300 there is no evidence of such effects (Nielsen, 2016j, 2017). We also see no evidence
in the distributions describing regional growth of population (Nielsen, 2016d).
This imagined, but never proven mechanism, was supposed to have been responsible for creating an
endless epoch of the so-called Malthusian stagnation characterised by irregular and generally stagnant
state of growth of population and of economic growth, but data are in clear contradiction of this doctrine
(Nielsen, 2016a, 2016d, 2016j, 2017; von Foerster, Mora & Amiot, 1960). It is a doctrine, which is
based on the incorrect interpretation of hyperbolic growth.
The growth of population and economic growth were hyperbolic. It is a monotonically increasing
growth. It is slow over a long time and fast over a short time but there is no stagnation and no takeoff
or explosion at any time. Stagnation and explosion are just illusions, which readily disappear when we
use the method of reciprocal values (Nielsen, 2014) to analyse data. What we see as a stagnation is just
a monotonically increasing growth and what we see as an explosion is just the natural continuation of
hyperbolic growth.
We have demonstrated that with the exception of just one event in the past 12,000 years (Nielsen,
2016j), and indeed in the past 2,000,000 years (Nielsen, 2017), there is no evidence that demographic
catastrophes were ever shaping the growth of the world population. This unique event occurred around
359
__________________________________________________________________________________
AD 1300 and coincides with five strong demographic catastrophes: the Mongolian Conquest (1260-
1295) with the total estimated death toll of 40 million; Great European Famine (1315-1318), 7.5 million;
the 15-year Famine in China (1333-1348), 9 million; Black Death (1343-1352), 75 million; and the Fall
of Yuan Dynasty (1351-1369), 7.5 million. The combined death toll caused by them is estimated at a
maximum of 139 million. At the onset of this unique event the world population was only about 380
million, so the relative impact should have been strong. This combined crisis lasted for about 280 years
but it caused only a minor disturbance in the growth of population. At the end of this crisis, the size of
population was reduced to only 360 million. There is also no convincing evidence that demographic
catastrophes were shaping the growth of regional populations (Nielsen, 2016d). Likewise, there is no
convincing evidence that they had any tangible impact on the economic growth, global or regional
(Nielsen, 2016a).
We have already explained why demographic catastrophes did not shape the growth of population.
We have demonstrated (Nielsen, 2016i) that, as first observed by Malthus (1798), his so-called positive
checks (demographic catastrophes and many forms of harsh living conditions) are responsible not only
for increasing the death toll but also for triggering the process of regeneration, reflecting the well-known
phenomenon observed commonly in nature. Thus, the destructive action of even strong demographic
catastrophes is quickly compensated by this process, which is likely to produce even faster growth than
before.
We can now understand why a combination of five strong demographic catastrophes were needed
to cause only minor and relatively short-lasting disturbance in the growth of population around AD
1300. This was one and only example in the past 2,000,000 years (Nielsen, 2017) when we can see a
correlation between the growth of population and demographic catastrophes. Now we have added to
this explanation by showing that individually, demographic catastrophes were generally too weak to
have a tangible impact on the growth of population. On rare occasions, when they were strong enough
to cause some minor damage, their action was quickly counteracted by the spontaneous and efficient
process of regeneration (Malthus, 1798; Nielsen, 2016i).
We have defined a series of gauge indicators allowing for a study of impacts of demographic
catastrophes. We have also concentrated our attention on the strongest catastrophes, thus maximising
the fraction of potentially destructive impacts. Even then, this fraction turned out to be small. On
average, only 6.5% of all major demographic catastrophes could have had a certain impact but as
demonstrated by the analysis of relevant data (Nielsen, 2016d, 2016j, 2017) they had no impact. They
were only relatively strong but even if they were stronger, such isolated actions could have been hardly
expected to cause lasting disturbances in the growth trajectory, particularly if we consider the apparently
ever-present process of regeneration (Malthus, 1798; Nielsen, 2016i).
Any negative impact on the growth of population could be expected to be reflected also in the
economic growth but the analysis of data shows that the economic growth remained also undisturbed
(Nielsen, 2016a). The growth of population and economic growth were exceptionally stable and
generally uninterrupted. Demographic catastrophes did not shape the economic growth or the growth
of population.
References
Austin Alchon, S. (2003). A pest in the land: new world epidemics in a global perspective. Albuquerque: University of New
Mexico Press.
Bartlett, A. A. (1993). The Arithmetic of Growth: Methods of Calculation. Population and Environment, 14(4), 359-387.
Bartlett, A. A. (2011). Private communication.
Beran, G. W. (2008). Disease and density mystery and mastery. Preventive Veterinary Disease, 86, 198-207.
Carmichael, A. G. (2009). Plague, historical. In M. Schaechter (Ed.), Encyclopedia of Microbiology (3rd ed., pp. 58-72). San
Diego, CA: San Diego University.
Cook, N. D. (1981). Demographic collapse, Indian Peru, 1520-1620. New York: Cambridge University Press.
Cunha, C. B., & Cunha, B. A. (2006). Impact of plague on human history. Infectious Disease Clinics of North America, 20,
253-272.
del Paso y Troncoso, F. (1940). Epistolario de Nueva Espana, 1505-1818. Mexico: Antigua Libreria Robredo.
Dobyns, H. F. (1963). An outline of Andean epidemic history to 1720. Bulletin of the History of Medicine, 37, 493-515.
360
__________________________________________________________________________________
Dobyns, H. F. (1993). Disease transfer at contact. Annual Review of Anthropology, 22(1), 273-291.
Dols, M. W. (1974). Plague in early Islamic history. Journal of the American Oriental Society, 94 (3), 37183.
Duncan, C., & Scott, S. (2005). What caused the Black Death? Postgrad Medical Journal, 81, 315-320.
Durand, J. D. (1960). The population statistics of China. Population Studies, 13(3), 209-256.
Fitzgerald, C. P. (1936). Further historical evidence of the growth of Chinese population. Sociological Review, 28, 267-273.
Fitzgerald, C. P. (1947). The consequences of the rebellion of An Lu-shan upon the population of the Tang Dynasty.
Philobiblon, 2(1), 4-11.
Press.
475.
http://dx.doi.org/10.1007/s11698-011-0062-7
Galor, O. and Moav, O. (2002). Natural Selection and the Origin of Economic Growth. The Quarterly Journal of Economics,
117, 1133-1191. http://dx.doi.org/10.1162/003355302320935007
Ghose, A. K. (2002). Food supply and starvation: A study of famines with reference to the Indian continent. Oxford
Economic Papers, 34, 368-389.
Gilliam, J. F. (1961). The plague under Marcus Aurelius. The American Journal of Philology, 82(3), 225-251.
Hawas, T. (2008). Anthropode-borne disease. In B. W. Lerner & K. L. Lerner (Eds.), Infectious diseases: In context (pp. 63-
66). Detroit: Gale Carnegie Learning.
Jutikkala, E. (1955). The great Finish famine 1696-97. Scand. Econ. Hist. Rev., 3, 48-63.
Keys, A., Brozek, J., Henschel, A., Mickelsen, O., & Taylor, H. L. (1950). The biology of human starvation. Minneapolis:
University of Minnesota Press.
Khan, I. A. (2004). Plague: the dreadful visitation occupying the human mind for centuries. Transactions of the Royal
Society of Tropical Medicine and Hygiene, 98, 270-277.
Kohn, G. C. (1995). The Wordsworth encyclopedia of plague and pestilence. New York: Facts on File.
History, 29, 307-331.
Lagerlf, N-P. (2003a). Mortality and early growth in England, France and Sweden. Scand. J. of Economics, 105(3), 419
439.
Lagerlf, N-P. (2003b). From Malthus to modern growth: Can epidemics explain three regimes? International Economic
Review, 44, 755-777.
Lardinois, R. (1985). Famine, epidemics and mortality in South India: A reappraisal of the demographic crisis of 1876-78.
Economic and Political Weekly, 20(11), 454- 465.
Littman, R. J., & Littman, M. L. (1973). Galen and the Antonine Plague. The American Journal of Philology, 94(1), 243-
255.
Maddison, A. (2006). The World Economy. Paris: OECD.
Maddison, A. (2010). Historical statistics of the world economy: 1-2008 AD. http://www.ggdc.net/maddison/Historical
McNeill, W. H. (1976). Plagues and people. Garden City, NY: Anchor Books.
Motolina aka Fray Toribio de Benavente o Motolina (1971). Memoriales; o, Libro de las cosas de la Nueva Espana y de
los naturales de ella. Mexico City: Universidad Nacional Autonoma de Mexico.
http://dx.doi.org/10.4236/am.2014.513188
361
__________________________________________________________________________________
Nielsen, R. W. (2016b). Unified Growth Theory contradicted by the absence of takeoffs in the Gross Domestic Product.
Nielsen, R. W. (2016f). The unresolved mystery of the great divergence is solved. Journal of Economic and Social Thought,
Nielsen, R. W. (2016g). Mathematical analysis of the historical income per capita distributions. Economic Review, 3(2), 300-
319. http://www.kspjournals.org/index.php/TER/article/view/766/889
http://kspjournals.org/index.php/JEB/article/view/1053
Nielsen, R. W. (2016i). The dichotomy of Malthusian positive checks: Destruction and even more intensified regeneration.
Nielsen, R. W. (2016j). Growth of the world population in the past 12,000 years and its link to the economic growth. Journal
of Economics Bibliography, 3(1), 1-12. http://www.kspjournals.org/index.php/JEB/article/view/607
Nielsen, R. W. (2016k). Interpretations of Hyperbolic Growth. Journal of Economics and Political Economy, 3(4), 594-626.
http://kspjournals.org/index.php/JEPE/article/download/1063/1099
Nielsen, R. W. (2016l). Demographic Transition Theory and its link to the historical economic growth. Journal of Economic
and Political Economy, 3(1), 32-49. http://www.kspjournals.org/index.php/JEPE/article/view/618/798
Nielsen, R. W. (2016m). Mechanism of Hyperbolic Growth Explained. Journal of Economic Library, 3(4), 411-428.
Nielsen, R. W. (2017). Economic Growth and the Growth of Human Population in the Past 2,000,000 Years. Journal of
Economic Library, 4(2), 128-149.
Quarterly, 83(4), 731-757.
Prem, H. J. (1992). Disease Outbreaks in Central Mexico during the Sixteenth Century. In N. D. Cook & W. G. Lowell
(Eds.), Secret Judgments of God: Old World Disease in Colonial Spanish America, pp. 20-48. Norman: University of
Oklahoma Press.
Rogers, R. G., & Hackenberg, R. (1987). Extending epidemiologic transition theory, A new stage. Social Biology, 34(3-4),
234-243.
Rosen, W. (2007). Justinian's flea: Plague, empire, and the birth of Europe. London: Random House.
Ross, M. E. (2008). Plague, early history. In B. W. Lerner & K. L. Lerner (Eds.), Infectious Diseases: In Context (pp. 627-
634). Detroit: Gale Carnegie Learning.
Russel, J. C. (1968). That earlier plague. Demography, 5, 174-184.
Scott, S., & Duncan, C. J. (2001). Biology of plagues: Evidence from historical populations. Cambridge: Cambridge
University Press.
Seeck, O. (1921). Geschichte des Untergangs der antiken Welt. Stuttgart: J. B. Metzlersche verlag.
Shrewsbury, J. F. D. (1970). A History of Bubonic Plague in the British Isles. Cambridge: Cambridge University Press.
Spignesi, S. J. (2002). Catastrophe! The 100 greatest disasters of all time. New York: Kensington Publishing Corp.
Trevisanato, S. I. (2004). Did the epidemic of tularemia in Ancient Egypt affect the course of world history? Medical
Hypotheses, 63, 905-910.
Trevisanato, S. I. (2007). The biblical plague of the Philistines now has the name, tularaemia. Medical Hypotheses, 69, 1144-
1146.
Twigg, G. (1984). The Black Death: A biological reprisal. London: Batsford Academic and Educational.
Walford, C. (1878). On the famines of the world: Past and present. Journal of the Royal Statistical Society, 41, 433-535.
Watkins, S. C., & Menken, J. (1985). Famines in historical perspective. Population and Development Review, 11, 647-676.
White, M. (2011). The great big book of horrible things: The definite chronicle of historys 100 worst atrocities. New York:
W. W. Norton & Company.
362

Mechanism of Growth

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Mechanism of Growth

Uploaded by

Copyright:

Available Formats

Griffith University, Gold Coast, Australia

Prof. Ron W. Nielsen

Prof. Ron W. Nielsen

About the Author: http://home.iprimus.com.au/nielsens/ronnielsen.html

1. Scientifically unacceptable established knowledge in demography and in economic

2. Changing the direction of the economic and demographic research... 25

5. Demographic Transition Theory contradicted by data 75

7. The postulate of the three regimes of economic growth contradicted by

8. Unified Growth Theory contradicted by the mathematical analysis of the historical

9. Puzzling properties of the historical growth rate of income per capita

11. The unresolved mystery of the great divergence is solved.. 172

15. Interpretations of hyperbolic growth. 241

16. The law of growth 270

17. Mechanism of hyperbolic growth explained 279

19. Puzzling Features of Income per Capita Distributions Explained 308

Economic growth was hyperbolic.

Industrial Revolution had no impact on shaping the GDP/cap trajectory.

2. The established knowledge

5. The origin of the concept of stagnation

6. Examples of questionable claims

7. Summary and conclusions

Changing the direction of the economic and

2. The common problem

2. The method of reciprocal values

3.2. Economic growth in Western Europe

3.3. Global economic growth

4. Summary and conclusions

Mathematical Analysis of the Historical Economic Growth with a

2. World economic growth

9. Summary and conclusions

Table 1. Summary of the mathematical analysis or the historical economic growth

Denmark, France, Netherlands and Sweden

Asia (including Japan)

3. Analysis of population data

where a is a constant related to the constant of integration.

It is just a reciprocal of a linearly-decreasing function. Consequently,

4. Detailed analysis of the AD data

10,000 BC 500 BC 500 BC AD 500

5. Implications for the economic growth

6. Summary and conclusions

Demographic Transition Theory Contradicted by Data

2. Demographic Transition Theory

3. Examination of data for Sweden

4. Examination of data for Mauritius

5. Examination of data for England

Figure 5. Growth of human population in England (Wrigley & Schofield, 1981).

6. Summary and conclusions

Unified Growth Theory Contradicted by the Absence of Takeoffs

2. World economic growth

9. Summary and conclusions

The Postulate of the Three Regimes of Economic Growth

2. Unified Growth Theory

3. Method of analysis and related issues

4. Analysis of data for Western Europe

5. Analysis of data for Eastern Europe

6. Analysis of data for Asia

7. Analysis of data for the former USSR

8. Analysis of data for Africa

9. Analysis of data for Latin America

10. Summary and conclusions

Nielsen, R. W. (2015b). Unified Growth Theory Contradicted by the GDP/cap Data.

Schematic diagrams of the signatures of Malthusian stagnation and takeoffs

Asia (excluding Japan)

Unified Growth Theory Contradicted by the Mathematical