See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/24058155

Micro and macro dynamics: Industry life cycles, inter-sector coordination and
aggregate growth

Article  in  Journal of Evolutionary Economics · February 2008

DOI: 10.1007/s00191-007-0077-1 · Source: RePEc

CITATIONS READS

65 350

2 authors:

Andreas Pyka Pier-Paolo Saviotti

University of Hohenheim Scuola Superiore Sant'Anna
244 PUBLICATIONS   2,777 CITATIONS    145 PUBLICATIONS   3,366 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Economic Complexity and Human Development: an efficiency analysis View project

International innovation networks and knowledge migration View project

All content following this page was uploaded by Andreas Pyka on 15 May 2014.

The user has requested enhancement of the downloaded file.

J Evol Econ (2008) 18:167–182
DOI 10.1007/s00191-007-0077-1
R E G U L A R A RT I C L E

Micro and macro dynamics: Industry life cycles,

inter-sector coordination and aggregate growth

Pier Paolo Saviotti & Andreas Pyka

Published online: 15 February 2008

# Springer-Verlag 2007

Abstract Technological change directly affects economic growth by exploiting and

exploring technological opportunities, thus determining productivity growth and
income. However, technological change also affects the composition of the
economic system, which itself constitutes an important prerequisite for economic
growth. The first aim of this paper is to show that the growing variety of the
economic system, determined by the emergence of new products and services and
leading to new industrial sectors, can allow the long term continuation of economic
development, even when the employment creating capacity of individual sectors
falls. The second aim is to illustrate the impact of micro variables on the meso-level,
that is, on the sectoral composition of an economy, as well as on its macro-economic
performance.

Keywords Economic growth and development . Industry life cycles . Variety .

Meso-level

JEL Classification O0 . O12 . O30

P. P. Saviotti (*)
INRA-GAEL, Université Pierre Mendès-France, Grenoble, France
e-mail: ppsavio@grenoble.inra.fr

P. P. Saviotti
CNRS GREDEG, University of Nice Sophia Antipolis, Nice, France

A. Pyka
Economics Department, University of Bremen, Hochschulring 4, 28359 Bremen, Germany
e-mail: pyka@uni-bremen.de

Present address:
A. Pyka
Department of Technology, Policy and Management, Technical University Delft,
Delft, The Netherlands
168 P.P. Saviotti, A. Pyka

1 Introduction

The present paper is one of a series based on a model of economic development led by
the creation of new sectors. This series shares the common theme of the relationship
between structural change and economic development, a theme which at least since
Schumpeter, Kuznets, and Clark has, for a long time, although intermittently, been
present in the economics literature. With respect to this research tradition, our work
differs in the use of the concept of variety and in the tools used in modelling. The logical
outcome of stressing the composition of the economic system as a determinant of
economic development is the need to represent analytically the composition of the
system itself. This we do by means of the concept of variety, which is used as a measure
of the overall extent of differentiation of the economic system. The concept of variety
not only measures such differentiation, but leads us to expect that economic
development proceeds in a given direction, that is, towards an increase in overall
variety. Furthermore, variety growth is not just an epiphenomenon of economic
development but a determinant of its long run sustainability. Our model of economic
development by the creation of new sectors (Saviotti and Pyka 2004a, b) was essentially
designed to test two propositions about the role of variety in economic development
(see Section 2.1). In this research program, we found inspiration in models of
economic development and structural change and in endogenous growth models.
Furthermore, the general approach of our model is broadly similar to that of history
friendly models (Malerba and Orsenigo 2002; Malerba et al. 1999).
Starting from the general premises developed in previous papers, the present
focus is on some specific issues. First, an unexpected result of our model was to
show that, under a wide range of circumstances, industrial sectors follow an industry
life cycle (ILC). Second, our model enables us to calculate macroeconomic variables
from microeconomic ones. In this respect, our model provides a bridge between
microeconomic studies of innovations and their macroeconomic performance. Thus,
an important focus of this paper is the interplay between the micro and macro
dynamics of our artificial economic system. Furthermore, the sector level is an
intermediate, or meso, level of aggregation (Dopfer and Potts 2008). Here, we study
interplay of micro, meso and macro levels of aggregation. In what follows, we first
introduce the features of our model and then explore the impact of demand, of the
industry life cycle and of inter-sector coordination on macroeconomic dynamics.

2 A model of economic development by the creation of new sectors

The main motivation underlying the creation of this model was the observation that
the changes which had taken place in the economic system in the last two centuries
did not consist uniquely in increases in efficiency of pre-existing production
processes leading to a constant set of given goods and services. In addition to the
previously described changes, which undoubtedly were very important, there was
the emergence of completely new goods and services, which changed the
composition of the economic system. We soon became convinced that these changes
in composition were not just an epiphenomenon of economic development, but one
of its main sources. A similar viewpoint had been present in the literature on
Micro and macro dynamics: Industry life cycles, inter-sector coordination and aggregate growth 169

economic development and structural change. As a consequence, our model takes

the changes in the composition of the economic system as one of its central features.
In our model, the mainspring of economic development is the endogenous creation
of new goods and services, giving rise to new industrial sectors. In this model, each
new sector is created by an important innovation, which gives rise to a new product,
which is in general highly differentiated. A sector is here defined as the population
of firms producing this highly differentiated product. The sector is generated by the
first entrepreneur who begins to produce the new product. The entrepreneur is
induced to do so by the expectation of the temporary monopoly that he will have
before imitation begins. However, if the new product is successful, imitation will
take place, gradually raising the intensity of competition within the sector and
eroding the monopolistic extra profits. As further imitation and entry occur, the
intensity of competition will rise, reaching values comparable to those of mature
sectors, thus transforming an innovating sector into a routine of the economic
system. In Schumpeterian terms, the sector will cease to be innovating and will
become part of the circular flow.
In the model, each sector is represented by an equation describing the entry and
exit of new firms. Entry is determined by the size of the potential market, as
measured by the adjustment gap, and by financial availability. Exit is determined by
the increasing intensity of competition, and by mergers and acquisitions. As a result
of the basic hypotheses on competition and on demand, each industrial sector
follows a life cycle, with the number of firms increasing first, reaching a maximum,
and then falling towards an oligopoly. Here, this cyclic behavior is stable under a
wide, but not infinite, range of conditions. In each sector, employment follows a path
very similar to that of the number of firms, increasing at first and subsequently
falling after having reached a maximum. In past work, we showed how an adequate
inter sector coordination can give rise to a stable macroeconomic employment
growth path, even when each sector provides declining employment.

2.1 Background to the model

The basic features of the model will be briefly summarized here, but a more extended
description can be found in Saviotti and Pyka (2004a, b, c). Our economic system is
constituted by an endogenously variable number of industrial sectors, each sector being
defined as the set of firms producing a unique but differentiated output. Our analysis is
in principle applicable to the production of services, but in what follows we will refer
mainly to manufacturing. One very specific feature of our model is the endogenous
creation of new sectors. In other words, the intrinsic mechanisms of development of
our economic system lead to the creation of new sectors, or alternatively, the
development of the system is not sustainable unless new sectors are created.
Each sector is created by an innovation which gives rise to an adjustment gap
AGti , a variable intended to capture the size of the potential market established by the
innovation. However, this market is initially empty because neither the production
capacity nor a structured demand for the new products exists. Both the production
capacity and the evolution of the demand will take place during a (possibly long)
period of time, by means of a gradual interaction of producers and users. Thus, the
adjustment gap measures the extent to which the market is far from saturation. When
170 P.P. Saviotti, A. Pyka

the market becomes saturated, the adjustment gap is reduced to zero. The adjustment
gap is very large right after the creation of the sector, and later it decreases gradually,
although not continuously. It is in fact possible for the adjustment gap to grow
during certain periods if innovations, following the one creating the sector, improve
either the performance of the product or the efficiency with which it is produced, or
both. In  our model, search activities affect
both the maximum possible
demand Dtmax;i and the instant demand Dti in a sector i. If we consider that
analytically the adjustment gap (Fig. 1c) is defined as the difference between these
two types of demand, we can understand that the time path of the adjustment gap
depends on those of Dtmax;i (Fig. 1b) and of Dti (Fig. 1a). During particular periods, it
is possible for Dtmax;i to grow more rapidly than Dti , thus enlarging the adjustment
gap, or delaying the saturation of the market. In the long run, we expect the
adjustment gap to be reduced to zero or to a constant value, or the market to become
saturated.
Each sector is actually created by an entrepreneur who establishes a niche induced
by the expectation of a temporary monopoly. If the innovation is successful,
imitative entry takes place, leading to the creation of a market. During this process,
imitative entry gradually raises the intensity of competition, thus reducing the
inducement to further entry. Competition plays a very important role in our model, in
two different senses. First, by its absence (temporary monopoly) it provides the
inducement for the first entrepreneur to create a niche; later, the increasing intensity
of competition due to imitative entry gradually reduces the inducement to further
entry. Second, in our model we distinguish two components of competition, intra-
and inter-sector competition. Intra-sector refers to the competition of firms within a
given sector, while the existence of inter-sector competition depends on the
possibility that different sectors provide comparable services.
The combination of a gradually increasing intensity of competition and of market
saturation transforms what was initially an innovative sector into a mature one. To
use a Schumpeterian metaphor, a new sector is created as an innovative one and
merges later into the circular flow, the set of existing routines of the economic
system (Schumpeter 1934). Our model is thus very Schumpeterian in character for
the following features:
1. a sector is created by an entrepreneur, induced by the expectation of a temporary
monopoly
2. if the innovation is successful, a bandwagon of imitators enters, eroding and
gradually abolishing the initial temporary monopoly;

demand of sector i maximum demand of sector i adjustment gap of sector i

t
1 101 201 301 401 1 101 201 301 401 t 1 101 201 301 401 t

a b c
Fig. 1 a Demand, b Maximum demand, c adjustment gap
Micro and macro dynamics: Industry life cycles, inter-sector coordination and aggregate growth 171

3. in the whole process, the initially innovative sector is transformed into a routine
and becomes part of the ‘circular flow’.

However, a number of other authors have influenced our model. A central

preoccupation of ours is the effects of and the inducements for the changing
composition of the economic system. We find that this preoccupation is not
sufficiently present in most models of economic growth developed heretofore. Some
recent models of growth have included a representation of the composition of the
economic system, although in ways that we do not find completely adequate. For
example, Romer (1987, 1990) considers the creation of sectors producing new types
of capital goods as the logical outcome of the activities of research and development.
In his model, these new sectors are added to the previous ones. In Aghion and
Howitt (1998), the composition of the system changes due to the innovations created
by R&D, but new innovations can substitute for old ones. In both cases, the direction
in which the composition of the economic system is going to change is not clear. We
are assuming that the output variety of our economic system increases in the course
of time, and that this increase is a necessary requirement for the continuation of long
term growth. Thus, although some old products and technologies may be displaced
by innovations, more new ones will be added, either by radical innovations or by the
specialization of pre-existing products. The emphasis we place on the changing
composition of the economic system relates our model closely to the literature on
structural change and economic development, for example to the work of Salter
(1960), Cornwall (1977), Pasinetti (1981, 1993), Fagerberg (2000) Fagerberg and
Verspagen (1999, 2002), Verspagen (1993, 2002) and more recently to Metcalfe et
al. (2005). On the other hand, the role played by the emergence of new products
relates our model to those of Stokey (1988, 1991) and Grossman and Helpman
(1991a, b). We find inspiration in the work of all these authors, but we depart from
most of these for a number of assumptions, such as the role played by variety and by
equilibrium. Here we agree with Schumpeter, that even if the economic system were
to start in a position of equilibrium, innovations would perturb this equilibrium. In
the presence of a constant stream of innovations, the system would most of the time
be in a steady state which, while ordered and showing the invariance of some state
properties, is not equivalent to the equilibrium of a conservative system, but rather to
the steady states of dissipative systems (Nicolis and Prigogine 1989).
A feature of our model is that increasing efficiency is not the only factor driving
economic growth. On the contrary, we think that increasing efficiency in pre-existing
sectors and the increasing variety created by the emergence of new sectors are two
complementary trends in economic development. These ideas are more extensively
described in the following two hypotheses (Saviotti 1996, 2007):
Hypothesis 1: The growth in variety is a necessary requirement for long-term
economic development.
Hypothesis 2: Variety growth, leading to the development of new sectors, and
efficiency growth in pre-existing sectors, are complementary and not
independent aspects of economic development.
These two hypotheses can be justified by the imbalance between productivity
growth and demand growth (Pasinetti 1981, 1993). If productivity keeps increasing
172 P.P. Saviotti, A. Pyka

while the demand for new goods and services reaches a saturation point, an
imbalance arises. If the economy were constituted by a constant set of activities, in
the presence of growing productivity it would become possible to produce all
demanded goods and services with a decreasing proportion of the resources used as
inputs, including labor. This imbalance would then constitute a bottleneck for
economic development. The addition of new goods and services to the economic
system, that is, a change in composition leading to a growth in variety, can be a form
of compensation for the potential displacement of labor and of other resources.
Variety growth is then required for the long term continuation of economic
development. On the other hand, new goods and services leading to new industrial
sectors can only be generated by means of search activities. The resources required
for these activities can only come from the increases in productivity in pre-existing
sectors in a way similar to what happened during the process of industrialization
when productivity growth in agriculture created the resources required for industrial
development (Kuznezts 1965). Similarly, productivity growth in pre-existing sectors
creates the resources required for search activities and thus for the generation of new
products and services. In Schumpeterian fashion, the growing productivity of the
routines constituting the circular flow creates the resources required for innovation,
without which economic development would come to a halt.
Both growing efficiency and growing variety are required if economic development
is to proceed. We call the trend towards growing variety the creativity of the economic
system. Thus, we can reformulate our previous statement by saying that efficiency
growth and variety growth are two complementary trends in economic development
(Saviotti 1996). Without creativity, the system would have a constant composition
and it would run into a bottleneck by becoming able to produce all the demanded
output with a declining proportion of the resources required (Pasinetti 1981).
Without efficiency, the resources required for search activities, and thus for the
creation of innovations, could not be found. Furthermore, the growing efficiency of
new sectors after they have been created allows these new sectors to acquire their
‘economic weight’, that is, the market size that only becomes available when the
prices of new products fall substantially and their performance increases.

2.2 The structure of the model

The central feature of our model is a system of equations, one for each sector, which
specifies how the number of firms in the sector changes following entry and exit (Eq. 1).
Entry is determined by the size of the adjustment gap, that is, of the potential market
created by the innovation, and by financial availability FAti : FAti is not just the
quantity of money available in the economic system, but measures the amount of
resources the economic system is prepared to allocate to a new and unknown sector.
Thus, FAti depends both on resources and on the availability of institutions that can
appropriately evaluate the potential of new sectors, often created by radical
innovations. In the very early phases of an innovation, the knowledge required to
asses its economic prospects is likely to be in very short supply, possibly being all tied
up in the creation of the innovation itself. In these conditions, new institutions may be
required to assess the potential of innovations. Furthermore, changing conditions
during the process of economic development may induce financial innovations
Micro and macro dynamics: Industry life cycles, inter-sector coordination and aggregate growth 173

adapted to these changes. Venture capital firms and business angels are examples of
just such financial innovations, but many others may emerge during the course of
economic development (Perez 2002, 2007). In summary, FAti includes both the
availability of resources and complementary institutional innovations in finance.
Financial availability can be considered as a particular instance of co-evolution of
technologies and institutions.
ΔNit ¼ k1  FAti  AGti  IC ti  MAti ð1Þ
Exit from a sector is determined by increasing intensity of competition ICti and by
mergers and acquisitions MAti . As previously pointed out, the inducement to enter a
new sector decreases as the intensity of competition rises as a result of imitative
entry starting from the initial situation of temporary monopoly enjoyed by the first
entrepreneur. Eventually no further entry will take place and net-exit will begin. Any
process that reduces the number of firms in the sector contributes to exit. Although
the term is called MAti , for mergers and acquisitions, it can include failures because
they are determined by the same factors.

2.3 Demand

In the previous versions of our model (Saviotti and Pyka 2004a, b, c), we had a very
simplified representation of demand. Essentially demand entered
the model in two
ways: first, by means of the concept of adjustment gap AGti , defined as equal to
the difference between the maximum and the actual demand at a given time; second,
by means of the assumption that all output produced in the economy was to be
consumed. The second assumption is equivalent to Say’s law. Those assumptions
were included in the model in order to simplify it. Of course, they were very
restrictive and limited unnecessarily the model. In the present version, we introduce
the following demand form:
C
Yit  ΔYit
Di ¼
t
ð2Þ
pti
where Yit , ΔYit and pti are the average value of the services produced, the degree of
differentiation and the average price of the products of sector i, respectively. In this
section, we describe the explorations of this demand function employed. In order to
explain how these experiments were carried out, we have to give the time
dependence of Yit , ΔYit and pti (Eqs 3, 4, and 5).
1
Yit ¼
ð3Þ
1 þ exp k14  k15 SE ti

1
ΔYit ¼
ð4Þ
1 þ exp k16  k17 SEti

pti ¼ 1 þ exp k18  k19 SEti ð5Þ



SE ti ¼ SE0 þ k4i  1  exp k5  Daccti ð6Þ
174 P.P. Saviotti, A. Pyka

where SEti are the search activities carried out in sector i. In our model, the term
search activities is a generalized analogue of research and development (R&D) and
indicates all the activities which scan the external environment searching for
replacements of or additions to existing routines (Nelson and Winter 1982, pp14–15,
149, 155–157). In our model, search activities SEti increase with accumulated
demand Daccti , but at a decreasing rate. This time path depends on the presence of
limited technological opportunities in a specific industry, and is determined by the
sector specific constant k4i and by a starting value SE0. Equations 4 and 5 show how
the factors affecting demand depend on search activities. k14 k19 are constants, to be
considered as parameters affecting the time paths of Yit , ΔYit and pti . Equations 4 and
5 have the form of logistic equations. They tell us that both the services performed
by the products of sector i and their degree of differentiation rise with search
activities at a rate which starts slowly, accelerates until reaching a maximum, and
then falls gradually. This assumption is not completely ad-hoc. The logistic form of
the equation can be obtained by assuming that (1) Yit has an upper boundary, and
that (2) the rate of growth of Yit on the product of its instant value times the
difference between its upper boundary and its instant value. In turn, these
assumptions are equivalent to admitting that improvements in the performance of a
new technology can be very rapid and even accelerate in the initial period of its life
cycle, when learning effects, etc., can provide a source of increasing returns to
adoption, but that eventually decreasing returns to further improvements will be
encountered, leading to an upper boundary. The constants k14 and k16 give us the
time when the logistic curve starts growing, while the constants k15 and k17 give us
the slope of the central part of the curve, during which the rate of growth of either
Yit or ΔYit reaches a maximum. Varying k14 k17 we can vary the rate of growth of Yit
and ΔYit . Using these forms for Yit and ΔYit we assume that both the services
produced and the degree of differentiation of the products of sector i will increase as
a consequence of search activities. The rate of growth of Yit and ΔYit will start
slowly, will accelerate later; in the end it will slow due to the exhaustion of the
possibilities inherent in the technology used to modify Yit and ΔYit . The equation for
pti has a different form because we expect pti to fall as a result of search activities.

3 Model results

3.1 Basic model results

In this section, we present what we consider to be the most basic results of the
model. Within a wide range of conditions, the number of firms in each sector grows
initially, reaches a maximum and then falls to a fairly low value (Fig. 2). Under these
conditions, each sector seems to follow a life cycle, similar to the ones detected by
Klepper (1996), Jovanovic and MacDonald (1994), Utterback and Suarez (1993).
However, in our model, this industry life cycle (ILC) is created by variables very
different from those used by the previous authors, and include increasing returns to
R&D, radical innovations and the emergence of dominant designs. In our case, the
cyclical behavior is caused only by the combined dynamics of competition and of
market saturation. We do not wish to say that cyclical behavior cannot arise under
Micro and macro dynamics: Industry life cycles, inter-sector coordination and aggregate growth 175

Fig. 2 Time path of the number #

of firms created in each sector 50
during the process of economic
40
development
30

20

10

-
1 51 101 151 201 251 301 351 401 451 t

the conditions identified by the previous authors. We simply say that cyclical
behavior can arise also from the interplay of competition and of market saturation.
As in the previous models, we call the time at which the number of firms in a sector
reaches a maximum the shake-out.
The adjustment gap can start falling immediately after the emergence of a new
sector, or it can increase for a while as a result of technological progress, but in the
end it will fall to a low or zero value (Fig. 3).
The intensity of competition (Fig. 4) within each sector rises during the initial
period of fast entry, reaches a maximum and then falls. However, the presence of
inter-sector competition prevents the intensity of competition of each sector from
falling to very low or zero values as a consequence of the growing industrial
concentration after the shake-out.
Starting from the behavior of microeconomic variables, such as those previously
described, we can calculate the curves for aggregate variables. Figure 5 shows the
time path of aggregate employment, obtained by aggregating the employment curves
of individual sectors.
As can be seen in Fig. 5, the aggregate employment curve, constituted by the
superposition of the individual sectors employment curves, can give rise to a
constant or growing employment even when the ability of each sector to create
employment declines. Using the number of sectors in the economic system as an
approximate (by defect) measure of variety, it can be seen that these results support
hypothesis No. 1, showing that variety growth is a necessary, although not sufficient,
condition for the long term continuation of economic development. In the rest of the
paper, the slope of the quasi-linear part of the aggregate employment curve will be
used as a measure of the performance of the economic system.

Fig. 3 Time paths of the adjust- AG

ment gaps of different sectors 1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
-
1 51 101 151 201 251 301 351 401 451 t
176 P.P. Saviotti, A. Pyka

Fig. 4 Time path of the intensi- IC

ty of competition 5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
-
1 51 101 151 201 251 301 351 401 451 I

3.2 Demand dynamics, industry life cycles (ILC) and macro-economic trends

In this section, we explore the influence of micro variables on the macroeconomic

performance of the economic system. Our model is highly interdependent, with most
variables influencing one another. In particular, search activities play a fundamental
role and affect most other variables. Thus, the time path of the number of firms in
each sector is influenced by the dynamics of search activities. As it was previously
pointed out, under a wide range of circumstances, the number of firms in each sector
follows an Industry Life Cycle (ILC), increasing first, reaching a maximum and then
falling. In our experiment, we study the influence of model parameters both on the
aggregate employment curve and on the ILC. As previously pointed out, the sector
level of aggregation can be considered at an intermediate, or meso, level (Dopfer and
Potts 2008) and the ILC a meso phenomenon. The interest of focusing not only on
the micro and macro levels of aggregation derives from the potential policy interest
of meso phenomena. For example, sectoral policies are of interest to various types of
policy makers.
We know that search activities affect and are affected by demand (Eq. 6). Thus,
one of our experiments consists of varying the parameters affecting search activities
and of finding out the effect this has on (1) the shape of the ILC and (2) on the
aggregate employment curve. We carry out the experiment by varying the
parameters affecting the dynamics of search activities. For this purpose, we vary
the two parameters k4 and k5, which we previously called technological opportunity
and rate of learning, and which affect the rate of growth and the level attained by
search activities in a sector. We carry out the experiment by calculating (1) the slope
of the linearized part of the aggregate employment curve, and (2) the curve for the

Fig. 5 Aggregate employment # aggregate employment and trend

curve obtained by adding up the 120
employment created by individ-
ual sectors 100

80
60
40

20
-
1 51 101 151 201 251 301 351 401 451 t
Micro and macro dynamics: Industry life cycles, inter-sector coordination and aggregate growth 177

Fig. 6 Effect of varying k4 on 90

the number of firms in a sector 80
70
60
50
40
30
20
10
-
1 251 501 751
k4 = 5 k4 = 7.5 k4 = 10
k4 = 12.5 k4 = 14

number of firms in a sector. The results are shown in (Figs. 6, 7, 8, and 9).
The results of the experiment can be summarized as follows. Varying either k4 or
k5 affects both (1) the time of creation of new sectors and (2) the shape of the ILC. In
particular, increasing k4 or k5 tends to speed up the creation of new sectors (new
sectors are created earlier) starting from low values of the two parameters, but a
reversal of this trend occurs for higher values. For what concerns the ILC, increasing
k4 or k5 tends to raise the rate of growth of industrial concentration after the shake
out. Thus, for higher values of k4 or k5, the number of firms falls more rapidly when
the sector moves towards saturation and maturity. The impact of these changes on
aggregate employment is also non-linear: when we increase either k4 or k5, the rate
of aggregate employment creation increases first and then starts falling. These non-
linear effects can be explained not only by the highly interactive nature of our
model, but also by the contrasting effects of search activities on (1) the rate of
creation of new sectors and on (2) the number of firms in each sector: speeding up
the creation of new sectors raises employment creation, while reducing the number
of firms in each sector depresses the rate of employment creation.
The previous experiment is just an example of the many which could be carried
out with our model to investigate the relationship between micro and macro
economic dynamics. A far more thorough analysis of the effect of various model
variables and parameters can be found in (Saviotti 2007). The purpose of the
experiment used here is to show that microeconomic variables and parameters can
affect aggregate employment performance. Furthermore, this paper shows that the

Fig. 7 Effect of varying k4 on #

the linearized employment trend 200

150

100

50

0
1 251 501 751 t
Linear (k4 = 5) Linear (k4 = 7.5) Linear (k4 = 10)
Linear (k4 = 12.5) Linear (k4 = 14)
178 P.P. Saviotti, A. Pyka

Fig. 8 Effect of varying k5 on #

the number of firms in a sector 90
80
70
60
50
40
30
20
10
-
1 101 201 301 401 501 601 701 801 901 t

k5 = 0.005 k5 = 0.0075 k5 = 0.01 k5 = 0.0125 k5 = 0.025

meso-economic level of aggregation of the industrial sector is itself affected and

affects aggregate employment performance. Taking into account the sectoral level of
aggregation by focusing on the ILC not only allows us to use our model to explore
one of the most important recent developments in industrial economics, but also
shows that consideration of the ILC provides a more direct explanation of the
complex mechanisms by which micro economic variables affect macroeconomic
dynamics. In other words, our paper confirms that meso levels of aggregation are
significant features of the structure of an economic system.
Considered from a slightly different point of view, our results show that industrial
concentration seems to grow faster for higher values of the parameters affecting the
time path of search activities and measuring technological opportunity and rate of
learning. In other words, the higher the technological opportunity and the rate of
learning for a given sector, the faster the number of firms will fall after the shake out,
leading to a growing industrial concentration, a result which seems to confirm those
of Nelson and Winter (1982, pp 308–328).

3.3 Inter-sector coordination and macro-economic trends

In this section, we investigate the effect of inter-sector coordination on macroeco-

nomic growth paths. Above we have seen that the time of emergence of a new sector
can be affected by the parameters of search activities. Inter-sector coordination can be
measured by the delay between the creation of sector n and that of sector (n+1). With
an adequate inter-sector coordination, the growing employment creating capacity of
sector (n+1) can be expected to compensate for the falling capacity of sector n.

Fig. 9 Effect of varying k5 on #

60
the linearized employment trend
55
50
45
40
35
30
1 101 201 301 401 501 601 701 801 901 t
Linear (k5 = 0.005) Linear (k5 = 0.0075) Linear (k5 = 0.01)
Linear (k5 = 0.0125) Linear (k5 = 0.025)
Micro and macro dynamics: Industry life cycles, inter-sector coordination and aggregate growth 179

Fig. 10 Effect of inter-sector employment trends for different entry speeds of new sectors
coordination on the aggregate 100
employment growth path 90
80
70
60
50
40
30
20
10
-
1 101 201 301 401 t

slow entry semi slow standard

semi fast fast entry

Thus, not only new sectors need to be created, but they need to be created at the right
times. Furthermore, we can expect a long delay between sectors n and (n+1) to
hinder the process of compensation, with employment falling in sector n before it
recovers due to the influence of sector (n+1). This process will interact with the
duration of the ILC, since, when there is a ‘long’ ILC in sector n, macro-economic
stability will be more tolerant of a delay in the creation of sector (n+1). In principle,
the factors that influence the shape and duration of the ILC can be expected to
interact with the inter-sector delay. Studying the joint influences of the factors
affecting the ILC and of inter sector delays should enable us to derive the most
favorable conditions for the emergence of a sustainable macro-economic growth
path. We can expect that the shorter are ILCs, the faster the economic system will
have to create new sectors in order to provide a balanced growth path, one with a
high trend rate of growth and with limited fluctuations. Figure 10 shows that the rate
of growth of aggregate employment is considerably affected by inter-sector delays.
Figure 10 shows that the rate of growth of employment rises systematically as the
rate of entry of new sectors into the economy increases. Furthermore, in addition to
affecting the slope of the employment growth path, inter-sector coordination affects
also the fluctuations in employment arising from the patterns of growth and decay of
different sectors. Figure 11 shows the combined effects of inter-sector coordination
on the slope and on the fluctuations of the employment growth path. As we can see
from Fig. 10, the rate of the employment growth rises with a growing rate of entry of
new sectors, but the fluctuations behave in a more complex way, increasing first,
then falling and subsequently increasing again. Considering that a sustainable
economic development is likely to require a high rate of employment growth but low
employment fluctuations, we are tempted to say that the most sustainable economic
development paths are likely to be found at the center of Fig. 11.
In summary, these experiments, although still limited, prove that inter-sector
coordination can exert an important influence on the rate of growth and on the
fluctuations of employment, and therefore on the sustainability of economic
development. More precisely, a growing rate of entry of new sectors into the
economy raises the rate of employment growth, but has a more complex effect on
180 P.P. Saviotti, A. Pyka

Fig. 11 Influence of the rate of trend volatility and trend of different entry fluctuations
entry of new sectors on the rate 0.2 1.25
of growth and on the fluctua-
tions of employment 1
0.1
0.75

0.5
0
0.25

-0.1 0
slow fast

trend fluctuation

the fluctuations of employment. To explore the conditions required to give rise to the
most sustainable economic development paths requires further work.

3.4 Variety

Going back to our hypotheses about variety and efficiency, we find a further
confirmation of the first hypothesis. It is not only that an adequate succession of new
sectors can compensate for the growing inability of older ones to provide employment,
but in addition faster entry of new sectors leads to a higher rate of growth of variety and
to a higher rate of growth of employment. However, growth in variety is a necessary but
not a sufficient condition for the sustainability of the economic development process.
Given that now we measure variety as the number of sectors in the economic system,
variety grows when the number of sectors grows. However, if the new sectors created
are very small, they may not compensate for the loss of employment entailed by the
evolution of pre-existing and mature sectors. The measure of variety we use now is
limited since it does not take into account the internal variety of each sector, arising for
example from product differentiation. The use of a more accurate measure of variety
could help us to improve our understanding of the role of variety in economic
development. This will be the object of future work on this model.

4 Summary and conclusions

In this paper, we use a modified version of a model of economic development by the

creation of new sectors that we previously developed (Saviotti and Pyka 2004a, b, c).
The main initial objective of this model is to show that the growing variety of the
economic system determined by the emergence of new sectors can allow the con-
tinuation of economic development in the long run, a result which had already been
obtained in the initial versions of the model (ibid). In this paper, we find confirmation of
the role of variety in economic development with a more complex version of our model,
containing, for example, a more sophisticated demand function. Furthermore, we
analyze the interactions between the micro and the macroeconomic dynamics of the
model by varying microeconomic variables and parameters and by measuring their
effect on the rate of growth of aggregate employment. Here we do not carry out a full
exploration of parameter space, but rather study the effect of the parameters affecting
Micro and macro dynamics: Industry life cycles, inter-sector coordination and aggregate growth 181

search activities on the rate of aggregate employment growth. Search activities are found
to clearly affect the rate of aggregate employment growth, although in ways that are
often non-linear. We enrich the analysis by taking into account the meso level of
aggregation constituted by an industrial sector. We do this by focusing on the influence
of microeconomic variables and parameters on the shape and duration of the Industry
Life Cycle. Taking into account the ILC allows us to show that its shape and duration are
affected by microeconomic variables and parameters and considerably facilitates our
interpretation of the relationships between micro and macroeconomic dynamics. Our
paper fits into the trend by which economic growth models have moved away from an
exclusively macroeconomic approach towards a growing recognition of microeconomic
phenomena. Within this trend, we differentiate ourselves by focusing explicitly on
structural change as a determinant of economic development and by using variety as a
variable to treat analytically the composition of the economic system. We also show that
the interaction of different levels of aggregation, starting from micro variables and
passing through industrial sectors, affects macroeconomic development paths.

References

Aghion P, Howitt P (1998) Endogenous growth theory. MIT Press, Cambridge, Mass
Cornwall J (1977) Modern capitalism: its growth and transformation. Martin Robertson, London
Dopfer K, Potts J (2008) The general theory of economic evolution. Routledge, London
Fagerberg J (2000) Technological progress, structural change and productivity growth: a comparative
study, Working paper N . 5/2000, Centre for Technology, Innovation and Culture, University of Oslo
Fagerberg J, Verspagen B (1999) Productivity, R&D spillovers and trade, Working Paper N . 3/1999,
Centre for Technology, Innovation and Culture, University of Oslo
Fagerberg J, Verspagen B (2002) Technology-gaps, innovation diffusion and transformation: an
evolutionary interpretation. Res Policy 31:1291–1304
Grossman GM, Helpman E (1991a) Quality ladder and product cycles. Q J Econ 106:557–586
Grossman GM, Helpman E (1991b) Innovation and growth in the global economy. MIT Press, Cambridge,
Mass
Jovanovic B, MacDonald GM (1994) The life cycle of a competitive industry. J Polit Econ 102:322–347
Klepper S (1996) Entry, exit growth and innovation over the product life cycle. Am Econ Rev 86:562–583
Kuznezts S (1965) Economic growth and structure. New York, Norton
Malerba F, Orsenigo L (2002) Innovation and market structure in the dynamics of the pharmaceutical
industry and biotechnology: towards a history friendly model. Ind Corp Change 11:667–703
Malerba F, Nelson RR, Orsenigo L, Winter S (1999) History friendly models of industry evolution: the
case of computer industry. Ind Corp Change 8:3–40
Metcalfe JS, Foster J, Ramlogan R (2005) Adaptive economic growth. Cambridge Journal of Economics
Nelson RR, Winter SG (1982) An evolutionary theory of economic change. Harvard University Press,
Cambridge MA
Nicolis G, Prigogine I (1989) Exploring complexity. Freeman, New York
Pasinetti LL (1981) Structural change and economic growth. Cambridge University Press, Cambridge
Pasinetti LL (1993) Structural economic dynamics. Cambridge University Press, Cambridge
Perez C (2002) Technological revolutions and financial capital: the dynamics of bubbles and golden ages.
Edward Elgar, Cheltenham
Perez C (2007) Finance and technical change: a long-term view. In: Hanusch H, Pyka A (eds) The Elgar
companion to neo-schumpeterian economics. Edward Elgar, Cheltenham, pp 820–839
Romer P (1987) Growth based on increasing returns due to specialization. Am Econ Rev 77:565–62
Romer P (1990) Endogenous technical progress. J Polit Econ 98:71–102
Salter WEG (1960) Productivity and technical change. Cambridge University Press, Cambridge
Saviotti PP (1996) Technological evolution, variety and the economy. Edward Elgar, Aldershot
182 P.P. Saviotti, A. Pyka

Saviotti PP (2007) Qualitative change and economic development. In: Hanusch H, Pyka A (eds) The Elgar
companion to neo-schumpeterian economics. Edward Elgar, Cheltenham, pp 820–839
Saviotti PP, Pyka A (2004a) Economic development by the creation of new sectors. J Evol Econ 14:1–35
Saviotti PP, Pyka A (2004b) Economic development, qualitative change and employment creation. Struct
Chang Econ Dyn 15:265–287
Saviotti PP, Pyka A (2004c) Economic development, variety and employment. Revue Economique 55(6)
Special issue on structural change and growth (Gaffard JL, Saviotti PP eds), (November)
Schumpeter J (1934) The theory of economic development. Harvard University Press, Cambridge, Mass
(original edition 1912)
Stokey NL (1988) Learning by doing and the introduction of new goods. J Polit Econ 96:701–717
Stokey NL (1991) Human capital, product quality and growth. Q J Econ 106:587–616
Utterback JM, Suarez FF (1993) Innovation, competition and industry structure. Res Policy 22:1–21
Verspagen B (1993) Uneven growth between interdependent economies. Avebury, Aldershot
Verspagen B (2002) Structural change and technology. a long view. Paper presented at the 2002 DRUID
Conference

View publication stats