THE JOURNAL OF ENERGY

AND DEVELOPMENT

Santosh Kumar Sahu and K. Narayanan,

“Energy Use Patterns and Firm Performance:

The Evidence from Indian Industries,”
Volume 40, Number 1

Copyright 2015
 ENERGY USE PATTERNS AND FIRM
PERFORMANCE: EVIDENCE FROM
INDIAN INDUSTRIES

Santosh Kumar Sahu and K. Narayanan*

Introduction

E nergy efficiency and conservation are critical elements of energy policy and
they are even more salient with increasing concerns regarding climate change,
establishing energy security, and meeting growing energy demand. To address

*Santosh K. Sahu, Assistant Professor at Madras School of Economics, Chennai (India),

obtained his Ph.D. on the economics of energy use in Indian industries from the Indian Institute of
Technology (IIT), Bombay. His research interests focus on industrial economics, energy economics,
and economic impacts of climate change. The author’s publications appear in The International
Journal of Energy Technology & Policy, International Journal of Energy Economics & Policy, and
Eurasian Journal of Economics and Business. Dr. Sahu also has participated in the JSPS Student
Exchange Program for East Asian Young Researchers at Nagoya University (Japan) and contributed
to research projects and studies on industrial economics, economics of climate change, and
dynamics of poverty.
K. Narayanan, Institute Chair and Professor at the Department of Humanities & Social Sciences,
IIT, Bombay, obtained his Ph.D. in economics from the Delhi School of Economics, University of
Delhi, and did postdoctoral research at the Institute of Advanced Studies United Nations University,
Tokyo. His research interests span the areas of industrial economics, international business, socio-
economic empowerment through information and communication technologies, environmental and
development economics, and economic impacts of climate change. Dr. Narayanan’s research has
been published in Research Policy, Journal of Regional Studies, Technovation, Oxford Development
Studies, International Journal of Energy Economics and Policy, and Economic and Political Weekly.
The author was also a co-editor (with N. S. Siddharthan) of Indian and Chinese Enterprises: Global
Trade, Technology, and Investment Regimes (Routledge, 2010) and has guest edited a special issue
of the IASSI Quarterly on the theme of “Human Capital and Development.”

The Journal of Energy and Development, Vol. 40, Nos. 1 and 2

Copyright Ó 2015 by the International Research Center for Energy and Economic Development
(ICEED). All rights reserved.
111
112 THE JOURNAL OF ENERGY AND DEVELOPMENT

these challenges, some call for reducing energy demand overall while others argue
for changing the energy portfolio mix, such as increasing the role of renewable
energy sources. It is not surprising that a great amount of research has been un-
dertaken to examine the economics of energy efficiency, conservation, and per-
formance at different levels of the economy.
In this paper, we have focused on Indian industries as our case study. Increased
energy efficiency and conservation will be essential for India as it was the fourth-
largest energy consumer in the world after China, the United States, and Russia in
2011.1 Despite having notable fossil-fuel resources, the nation has become in-
creasingly dependent on energy imports and, according to projections, India’s
energy demand could rise by 132 percent by the year 2035, outpacing energy de-
mand growth in China.2 This is being propelled by the fact that India’s per-capita
energy consumption today is now one-third of the global average, according to the
International Energy Agency (IEA), which indicates potentially higher energy de-
mand in the long term. Primary energy consumption in India has more than doubled
between 1990 and 2012, reaching an estimated 32 quadrillion British thermal units
(Btus).3 At more than 1.2 billion people in 2012, India’s population is a huger driver
of this growing energy demand: the country has the second-largest population in the
world, and is growing at about 1.3 percent each year since 2008, according to World
Bank data.4 Of course, with the type of population and economic growth that India
has been experiencing and that analysts are predicting in the future, pollution-related
issues arise, particularly in regard to greenhouse gases (GHGs). Thus, much attention
is being focused on how the Indian economy can grow without increasing GHGs,
such as carbon dioxide (CO2) emissions, and how increased industrial efficiency and
conservation can play a role.
We begin this study by looking at some of the most vital concepts in the field of
energy economics. First, it is important to view energy as an input into the pro-
duction of desired energy services rather than as an end in itself. Energy efficiency
is typically defined as the energy services provided per unit of energy input. At the
individual product level, energy efficiency can be thought of as one of a bundle of
product characteristics, such as product cost.5 At the aggregate level, the energy
efficiency of the economy can be measured as the level of gross domestic product
(GDP) per unit of energy consumed in its production. Regarding energy conser-
vation, it generally is defined as a reduction in the total amount of energy consumed.
Thus, energy conservation may or may not be associated with an increase in energy
efficiency. The understanding of the difference between energy conservation and
energy efficiency is imperative.6 The distinction is also key in understanding the
short- versus long-run price elasticity of energy demand, whereby short-run changes
may depend more on changes in consumption of energy services, whereas longer-
run changes may be affected more by gains in energy efficiency.
Additionally, in discussing energy and the performance of any economy or
firm, one must differentiate between energy efficiency and economic efficiency.
 INDIA: ENERGY USE & FIRM PERFORMANCE 113

Maximizing economic efficiency typically involves maximizing net benefits to so-

ciety. Market conditions may depart from efficiency if there are market failures;
nonetheless, most economic analysis of energy efficiency have taken cost-minimizing
or profit-maximizing behavior of households or firms as a given. However, some
literature has focused more closely on the decision-making behavior of economic
agents, highlighting potential “behavioral failures.” Therefore, much of the economic
literature on energy efficiency attempts to look at how energy-efficiency decisions are
made and how market or behavioral failures can be remedied.
Energy prices and markets influence consumer/firm decisions regarding how
much energy to consume and whether to invest in more energy-efficient products,
equipment, and processes. An increase in energy prices will result in some energy
conservation in the short run. However, short-run changes in energy efficiency tend to
be limited because of the slow turnover rate of energy-using equipment. Moreover,
the length of energy price increases and the perception of whether that increase is of
a more permanent nature is likely to significantly affect energy-efficiency adoption as
consumers and firms respond by replacing older equipment and changing processes.
With this background, this paper tries to form three groupings of firms based on
their energy consumption patterns and attempts to link these with firm perfor-
mance in the country of India. The motivation for such an attempt is to find out the
relationship between energy intensity (or energy efficiency) and firm performance.
Therefore, the objective is to examine whether firms that consume similar sources of
energy have similarities in determining firm profitability and hence, performance.
The rest of this study is designed as follows. In the next section we give an overview
of select literature on determinants of profitability. In the subsequent sections, we
present the data sources, econometric framework, and construction of variables,
which is followed by the empirical results and our conclusions with plausible policy
implications.

Review of Literature

The work of M. Porter on the international competitiveness of industries is

seminal.7 He argues that competitive and successful industries usually occur in the
form of specialized clusters of “indigenous” or “home-base” industries, which are
linked together through vertical relationships (buyers/suppliers) or horizontal re-
lationships (common customers, technology, skills, distribution channels, etc.).
The author, who popularized the term “industry clusters,” asserted that the phe-
nomenon of industry clustering was so pervasive that it appeared to be a major
feature of advanced economies. Porter’s theories provided a platform for sub-
sequent research on competitiveness, industry structure, and firm performance.
For example, the National Economic and Social Council (NESC) commissioned
a study on industrial clusters in Ireland, examining the competitive advantages of
114 THE JOURNAL OF ENERGY AND DEVELOPMENT

three Irish sectors: dairy processing, the music industry, and the software in-
dustry.8 P. Clancy and M. Twomey concluded that the three sectors studied cannot
be regarded as part of a fully developed industry clusters as described by Porter,
although there are benefits from the presence of some form of groupings of
connected or related companies and industries and from interactions among them.9
To understand Porter’s view of why competitive and successful industries
generally occur in the form of clusters, we must refer to his “diamond model” of
competitive advantage. According to Porter’s theory, the competitive advantage of
an industry originates from the “national diamond,” that is to say, the four different
determinants of competitive advantage that are created within the home base of
a country: (1) domestic factor conditions; (2) the nature of domestic demand con-
ditions; (3) the presence of related and supporting industries; and (4) firm strategy,
structure, and rivalry in the identified industry.
In the field of regional studies, a number of works have attempted to identify
regional and industrial clusters and their characteristics.10 Their studies calculated
the relevant four correlation coefficients representing the following similarities
between two industries: (1) industries X1 and X2 have similar input purchasing
patterns; (2) industries X1 and X2 have similar output selling patterns; (3) the buying
pattern of industry X1 is similar to the selling pattern of industry X2; and (4) the
buying pattern of industry X2 is similar to the selling pattern of industry X1, with
industrial clusters identified by application of similarity matrices to principal com-
ponent factor analysis. In our case, we have assumed that industries X1 and X2 have
similar input purchasing patterns. However, the industrial cluster is mostly related to
the geographical location of firms; hence, we are not defining them as industrial
clusters based on the energy choice, but as samples of firms consuming similar
energy source as one of their inputs.
M. Hirschey and D. Wichern analyze the consistency, determinants, and uses of
accounting and market-value measures of profitability.11 They find that there exists
a significant explanatory role for research and development (R&D) intensity, ad-
vertisement, leverage, and industry growth as determinants of profitability.
G. Hansen and B. Wernerfelt integrated two sample models of firm performance—one
based on economic factors used and the other on organizational factors.12 The eco-
nomic factor model is based on traditional economic theory and focuses on the salience
of external market factors in affecting firm performance. The organizational model is
built on the behavioral paradigm and places organizational factors as the major drivers
of performance. The research results imply that, while both models’ factors are im-
portant, organizational factors explain roughly twice as much variance in firm profit
rates as economic factors.
I. Kessides found that the existence of firm effects implies inter-firm differences
in internal efficiency and that firm-specific efficiency characteristics persist across
industries (i.e., if a firm is relatively efficient in market A, it is also likely to be
relatively efficient in a randomly selected market B).13 The work of T. Brush et al.
 INDIA: ENERGY USE & FIRM PERFORMANCE 115

suggests that both corporations and industry-type influence business unit profit-
ability but the corporation influence is the greater of the two.14 Firms can gain
a comparative advantage by pursing R&D as these activities result in new products
and processes that, if implemented effectively, can gain them competitive advan-
tages. This behavior of a firm enables it to differentiate itself from its competitors.
Likewise, other economists argue that this behavior creates value for firms by
generating some intangible assets. Following the link between R&D and innovation,
for our research we assume that firms consume different energy sources based on
the technologies they adopt for production. For example, firms consume efficient
energy sources when they are technologically superior to other firms. Hence, the
performance of the firm is related to the choice of energy used and subsequently the
energy intensity.
As the cost of energy inputs rise, producers prefer to employ smaller quantities
of energy inputs and substitute cheaper inputs for more expensive energy during
the production process.15 The relationship between energy prices and technolog-
ical process is estimated by setting energy patents as a proxy for innovation. The
empirical work by J. Cornillie and S. Fankhauser focuses more extensively on the
energy intensity in emerging markets.16 They apply a decomposition technique to
macro-level data and demonstrate that energy intensity is different for regions with
different rates of privatization. The authors assert that unchanged levels of energy
intensity are associated with a large share of heavy industry in the economy. Changes
in use of energy inputs also are found to be strongly correlated with technological
development.17 As a result, investments into innovations are associated with efficient
energy use and can result in energy savings while improving technologies.18 Another
form of boosting energy efficiency through investments is presented by C. Martinez,
who argues that positive results can be achieved through a “demonstration effect” in
any business environment.19
Among the specific firm-level characteristics of overall performance are labor,
capital productivity, and their respective ratios. Frequently, these factors are
considered as significant determinants of energy efficiency.20 Firms that operate in
developing countries are likely to be characterized by a comparatively low level of
wages and, therefore, gain an advantage by using labor more intensively than other
inputs.21 At the same time, over-employment of labor can be the cause of in-
efficiency as asserted in the article by A. Couto and D. Graham.22 However, the
research of L. Lachaal et al. found that the impact of labor costs is not significant as
a measure of technical efficiency, while the share of skilled labor force is significant
and positive.23 Therefore, labor quality could be taken into consideration while
analyzing a firm’s performance with respect to energy resources. The hypothesis
that firm size can improve energy efficiency is also proven in the research of
E. Oczkowski and K. Sharma.24 Still, the relation between the firm size and effi-
ciency is not straightforward and can be either negative or positive.25 Other research
on efficiency issues looks at the role of market share or value added to the industry
116 THE JOURNAL OF ENERGY AND DEVELOPMENT

output and finds evidence that it can explain inefficiencies as a factor of market
power.26 It is worth mentioning that fossil fuels are characterized by considerable
undesirable outcomes and externalities (such as CO2 emissions); yet, their share in
total energy generation is still dominant.27
Now, let us turn our attention to the studies pertinent to our research subject of
industrial firm performance. There are several papers on firm-level energy-
intensity determinants in India, among them N. Kumar and N. S. Siddharthan, S.
Sahu and K. Narayanan, and B. Goldar.28 They have not followed the production
function approach to examine the role of energy in the production function system.
Departing from a reduced form model of the determinants of the energy intensity
in industrial firms, these studies have applied multiple regression analysis to identify
the main firm characteristics related to Indian manufacturing energy intensity.
T. Papadogonas et al. used a similar approach to analyze the energy intensity of Greek
manufacturing firms.29 The energy-intensity variable is approximated by the fuel and
power expenses over total sales ratio. The results of T. Papadogonas et al.’s research
suggest that when firms are more capital intensive, they are more energy intensive.
Capital intensity seems to be positively related with energy intensity, as well as ex-
penditures on repairs and the age of firms. It can be argued that capital-intensive
industries use more energy due to complementarities between both factors, and older
plants and equipment would require repairing as they are probably less energy effi-
cient. Thus, older firms can be characterized as more energy intensive.
The relationship between firm size and energy consumption is less obvious.
Generally speaking, larger firms have an energy cost advantage only in the low energy
consuming industries according to T. Papadogonas et al.30 A negative relationship is
found by N. Kumar and N. S. Siddharthan and B. Goldar, while S. Sahu and
K. Narayanan first find an inverted-U relationship in their cross-sectional study in 2008,
and subsequently an U-shaped relationship between both variables using pooled cross-
section data over a nine-year span in their 2010 article.31 Larger firms may benefit from
economies of scale with decreasing returns in the use of energy, but this effect is not
strongly related in those papers. Foreign firms are more energy efficient in N. Kumar
and N. S. Siddharthan, S. Sahu and K. Narayanan (2009), and B. Goldar, but not in
S. Sahu and K. Narayanan (2010).32 The impact of foreign ownership on energy con-
sumption is not clear cut regarding those study results. It could depend on the country’s
environmental regulation and energy prices. Moreover, it should be interesting to look
at the impact of ownership structure on energy efficiency as differences can emerge
between private and public structures. Surprisingly, R&D investment intensity is not
related to less energy intensity,33 and even seems to be positively correlated.34 But,
using a dummy for R&D, B. Goldar obtains the expected negative effect on energy
intensity. In addition, according to the paper by T. Papadogonas et al., the energy
intensity is smaller in high technology industries.35
Using a different approach compared to the previous set of papers, other
authors have attempted to analyze the most relevant drivers affecting energy
 INDIA: ENERGY USE & FIRM PERFORMANCE 117

intensity at the firm level. K. Fisher-Vanden et al. utilize a structural model of

a Cobb-Douglas cost function to identify what was reducing energy intensity in
Chinese industrial firms.36 These authors found that changes in relative energy
prices and R&D expenditures are the major causes of declining energy intensity in
their case study. Another factor effecting energy intensity was studied by
M. Morikawa who discovered a positive relationship between population density
and the energy efficiency in the services sector.37 When the population density of
the locality doubles, the author estimates a 12-percent decrease of firm-level
energy intensity in the services sector. The research also uncovered a negative link
between capital and labor intensities and energy efficiency.
A large amount of economic literature attempts to understand the so-called
“energy-efficiency gap,” which is the difference between cost-effective energy-
efficient investments and the level of such investments actually implemented. There
are three reasons cited as the causes for the energy-efficiency gap: market failures,
market barriers, and management practices.38 Market failures refer to all situations
violating the neoclassical assumptions of rationality, perfect information, and no
transaction costs. The market barriers to increased energy efficiency include a low
priority of energy issues, incomplete markets for energy efficient products, and
capital market obstacles, with the latter point being emphasized as a critical chal-
lenge. The adoption of energy-efficiency technologies and investments necessitates
funding; thus, with a lack of capital allocation combined with their lower placement
on management’s priority list, significant impediments can form.39 The paper by
A. Trianni and E. Cagno focuses on Italian small- and medium-sized enterprises
(SMEs) in the manufacturing sector and barriers to the implementation of energy-
efficient technologies.40 The authors find that capital availability is a major hindrance
along with economic and information barriers, suggesting that information about
technology, regulations, and opportunities for financing are perceived as compli-
cated and fragmented or not trustworthy, especially those coming from external
sources such as the government and financial institutions.
In addition to the barriers of capital availability, recent research has focused on
the critical role of organizational structures and management best practices for
increasing the adoption of energy-efficient technologies and processes at the firm
level. N. Bloom et al. address this issue in their study of over 300 manufacturing
firms in the United Kingdom.41 Using information about firm’s managerial quality
and census data containing energy consumption expenditures, they find that better
managed plants are significantly less energy intensive. This effect is substantial:
going from the 25th to the 75th percentile of management practices is associated
with a 17-percent reduction in energy intensity.
The paper of R. Martin et al. provides further evidence about the negative link
between management practices and energy intensity.42 The authors find that “climate
friendly” management practices are associated with lower energy intensity. Addi-
tionally, they show that the variation in management practices across firms can be
118 THE JOURNAL OF ENERGY AND DEVELOPMENT

explained in part by organizational structure. Their results support the view that the
“energy-efficiency paradox” can be explained by managerial factors and highlight
their importance for private-sector innovation that will sustain future growth in
energy efficiency.
N. Bloom et al. and R. Martin et al. employ two different proxies for energy
intensity. Both energy costs over total sales and energy costs over total variable
costs are used in their regressions. In order to ensure the robustness of our results,
we also run our analysis with both variables. In an attempt to relate technology to
energy intensity, S. Sahu and K. Narayanan computed the CO2 emissions from
fossil-fuel consumption for firms in India’s manufacturing sector from 2000 to
2011 by adopting the Intergovernmental Panel on Climate Change (IPCC) Reference
Approach.43 Their results indicate that there are differences in firm-level emission
intensity and, in turn, they are systematically related to identifiable firm-specific
characteristics. They found size, age, energy intensity, and technology intensity as
the major determinants of CO2 emissions of Indian manufacturing firms.

A Macro- and Micro-Perspective Overview

From a macro perspective, analyzing India’s energy efficiency and firm in-
tensity is extremely important. In the case of India, there are numerous factors
driving the search for ways to promote energy efficiency and reducing energy
intensity for the manufacturing sector. India is the fourth largest industrial energy
consumer with a 5-percent share of total industrial energy use, surpassed only by
China, the United States, and Russia. Globally, industry accounts for one-third of
all the energy used and for almost 40 percent of global CO2 emissions. In 2010, the
total final energy use in industry amounted to 3,019 million tons of oil equivalent
(Mtoe). Direct emissions of CO2 in industry amounted to 7.6 gigatons of CO2 (Gt
CO2) and indirect emissions were 3.9 Gt CO2. The International Energy Agency
analysis shows that industry will need to reduce its current direct emissions by about
24 percent of their 2010 levels if it is to halve global emissions by 2050. The sheer size
of this sector, combined with its energy usage demands and CO2 emissions, neces-
sitates our giving this issue serious attention. The industrial energy use in India reached
150 Mtoe in 2007, accounting for 38 percent of the country’s final energy used. The
five most energy-intensive industrial sectors (iron and steel, cement, chemicals and
petrochemicals, pulp and paper, and aluminum) accounted for 56 percent of India’s
industrial energy consumption in 2010.
From a micro-level viewpoint, as was mentioned previously, industrial clusters
are based on certain geographical locations and mostly are characterized by small-
and medium-scale industries. In the process of integration (horizontal or vertical)
firms also form clusters. Additionally, firms form clusters around resource avail-
ability for production. As the motive of any firm is either profit seeking or growth,
 INDIA: ENERGY USE & FIRM PERFORMANCE 119

the issue of long-run resource availability becomes one of the major determinants of
firm performance. Selection of certain energy sources is directly related to the
technology and machinery installed or the R&D capability of any enterprise. In this
context, the choice of energy source utilized for production and the performance of
firms are important factors that require special attention.
To summarize, energy intensity is related to some firm characteristics, such as
input composition, firm’s size and age, the ownership structure, and the population
density. Moreover, several economic factors, e.g., financial constraints, manage-
ment practices, or other market barriers, seem to be important drivers of firm-level
energy efficiency. From the review we can conclude that, apart from input choices,
firm characteristics matter for improving energy efficiency and performance. With
these motivations, we focus on analyzing the determinants of profitability of firms
consuming different energy sources for Indian manufacturing.

Data Source, Model, and Construction of Variables

We use data from the Center for Monitoring Indian Economy (CMIE)
PROWESS from 2005–2013.44 According to the database, Indian manufacturing
firms use 44 types of energy sources, which are classified into seven categories (as
primary, secondary, etc.). There is also evidence that firms shift in choice of fuel in
two different time periods. Therefore, choice of energy source is dynamic at the firm
level when we consider the energy mix. For the empirical analysis, we have selected
firms that are consistent in their choice of energy input from 2005–2013. This study is
restricted to the primary source of energy consumption that includes (1) natural gas,
(2) petroleum, and (3) coal. The sample consists of 23,434 firms from 2005–2013.
The sample is divided into three categories based on the primary energy demand
(henceforth, energy groups). From figure 1, we can observe that 38 percent of the
sample are in the petroleum group, 36 percent in that of natural gas, and the remaining
26 percent in the coal group. However, more importantly, firms in the coal group are
larger in size (based on sales) relative to other classifications.
In other research, S. Sahu and K. Narayanan and B. Goldar estimated determinants
of energy intensity for sample firms in Indian manufacturing industries.45 Analyzing
growth and profit behavior of large-scale Indian firms, N. Siddharthan et al. have
estimated determinants of profitability.46 These previous studies have used the
structure conduct performance theory in determining factors for energy intensity and
profitability. Based on the aforementioned research, the econometric specification
takes the following functional form:

Pit = ait + b1 CI it + b2 EIit + b3 RDit + b4 S it + b5 S 2 it + b6 Ait + b7 M it + mit ð1Þ

where P = profitability of firms, CI = capital intensity of firms, EI = energy in-

tensity of firms, RD = research and development intensity of firms, S = firm size,
120 THE JOURNAL OF ENERGY AND DEVELOPMENT

Figure 1
DISTRIBUTION OF FIRMS IN EACH ENERGY GROUP IN INDIAN MANUFACTURING
(by percent)

Source: Authors’ calculation from the Center for Monitoring Indian Economy (CMIE)
PROWESS database.

S2 = square of firm size, A = age of firms, and M = multinational enterprises

affiliation of firms.
Equation (1) is estimated four times for the full sample and each of the energy
groups. Panel data econometrics are applied for the full sample as well as for the
three energy groups separately. Fixed effects and random effects models are esti-
mated and based on the result of Hausman statistics methodology with random
effects preferred over fixed effects.47 In equation (1), energy intensity is a proxy for
energy efficiency, measured as a summation of all possible sources of energy
consumed by a firm in Btus as a proportion of net sales. As a firm becomes energy
efficient, its performance is likely to improve. D. Roberts and J. Tybout found that
the most productive firms find it profitable to incur the sunk costs in export mar-
kets.48 Higher profit-earning firms can more easily face competitiveness in foreign
markets. The existence of fixed production costs implies that the firms producing
below the zero-profit productivity cut-off would make negative profits if they
produce and, therefore, they choose to exit the industry. We define profitability as
the ratio of net profits to net sales. Because of economies of scale, larger firms may
have lower average and/or marginal costs, which would increase the likelihood of
better performance. Firm size is measured by the natural log of net sales. R&D
expenditure has the potential to enhance quality and to generate economy in the
production process and these factors that may increase the likelihood of entering the
 INDIA: ENERGY USE & FIRM PERFORMANCE 121

export market and, thus, enhance firm performance. We define R&D intensity as the
ratio of R&D expenditures to net sales. Firms can gain a technological advancement
not only through their own innovation but also through purchases of new capital or
intermediate goods from other sectors. Capital intensity is measured in terms of net
fixed asset (i.e., total fixed assets net of accumulated depreciation) as a proportion of
net sale. Net fixed assets include capital, work-in-progress, and revalued assets. The
age of the firm is calculated as the difference between years of the study to year of
the incorporation of the firm as reported in the CMIE database. Through “learning
by doing” firms may improve their energy efficiency, thereby becoming more
profitable relative to younger enterprises. According to H. Faruq and D. Yi, there is
empirical evidence that foreign-owned companies tend to be more efficient in en-
ergy conservation and, at the same time, there is evidence provided by V. Zelenyuk
and V. Zheka that reveals a negative correlation between foreign ownership and
firm’s efficiency level.49 Multinational enterprise (MNE) affiliation of the firm is
defined as a dummy where the firm belonging to the foreign affiliate takes the value
of 1 and the domestic firms takes the value of 0. This study utilizes three primary
sources of fuel choice: coal, natural gas, and petroleum.

Determinants of Profitability for Full Sample and Three Energy Groups

This section provides an overview of the sample descriptions and summary sta-
tistics. First, we divide the sample based on MNE affiliations. Energy intensity and
profitability are calculated for both the subsamples and for the full sample as well.
Table 1 gives the results where firm characteristics are compared between the MNEs
and domestic firms. Firms consuming coal as the primary energy input are the most
profitable ones, whereas firms consuming natural gas are the least profitable, with
petroleum-consuming firms in between. Capital intensity is higher for firms using
natural gas. Firms become energy efficient when they consume natural gas and en-
ergy intensive when they use coal as their primary energy source. R&D intensity is
higher for firms using natural gas. From the MNE-affiliated firms, 141 use natural
gas, 127 use petroleum, and 99 use coal as their primary energy source. The mean
profitability of MNE-affiliated firms was found to be higher for those using coal when
compared to other energy sources. Capital intensity of the MNE-affiliated firms that
use natural gas is higher as compared to others. Energy intensity and R&D intensity
are found to be the lowest for firms using natural gas and highest for firms using coal.
The above results are also similar for the domestic firms. From the sample of
domestic firms, 8,427 use natural gas, 8,694 use petroleum, and 5,946 use coal as
their primary source of energy. The mean profitability is higher for domestic firms
using coal as compared to those using natural gas or petroleum. The capital in-
tensity of domestic firms using natural gas is found to be higher relative to the
other two energy sources. Firms using natural gas are more energy efficient. R&D
122 THE JOURNAL OF ENERGY AND DEVELOPMENT

Table 1
COMPARISON OF INDICATORS FOR FULL SAMPLE, MULTINATIONAL ENTERPRISE-
a
AFFILIATED FIRMS, AND DOMESTIC FIRMS IN THREE ENERGY GROUPS

Natural Gas Group Petroleum Group Coal Group

Variables Mean SD Mean SD Mean SD

Full sample
Profitability 1.006 2.446 1.141 2.900 1.324 3.802
Capital intensity 0.833 1.369 0.058 0.084 0.056 0.079
Energy intensity 0.059 0.076 0.841 1.298 0.903 1.512
R&D intensity 0.084 1.295 0.073 0.498 0.081 0.533
Number of observations 8,568 8,821 6,045
Multinational Enterprise-Affiliated Firms
Profitability 5.055 8.152 6.347 10.310 7.618 13.142
Capital intensity 4.405 5.688 0.065 0.089 0.071 0.092
Energy intensity 0.055 0.081 4.055 4.617 4.559 6.325
R&D intensity 0.460 1.362 0.379 1.261 0.455 1.485
Number of observations 141 127 99
Domestic Firms
Profitability 0.938 2.168 1.065 2.567 1.219 3.344
Capital intensity 0.773 1.072 0.058 0.084 0.056 0.079
Energy intensity 0.059 0.076 0.794 1.117 0.842 1.200
R&D intensity 0.078 1.293 0.068 0.476 0.075 0.500
Number of observations 8,427 8,694 5,946

a
SD denotes standard deviation; R&D denotes research and development.
Source: Authors’ calculations based upon the Center for Monitoring Indian Economy (CMIE)
PROWESS database.

intensity is found higher for firms using natural gas and least for firms using pe-
troleum as the primary source of energy. From tables 1 and 2, we can observe that
profitability is higher for firms using coal for both MNE-affiliated and domestic
firms. Both domestic and MNE-affiliated firms that use natural gas as their primary
source of energy report the least profitability and are more capital intensive. Firms
are categorized as energy efficient when they use natural gas and energy intensive
when they use coal as their primary source of energy.
The comparison of firm characteristics for MNE-affiliated and domestic firms
is presented in table 2. We can observe that MNE-affiliated firms are more profitable
when compared to domestic firms; however, the standard deviation for profitability
among MNEs is higher relative to the domestic ones. MNE-affiliated firms are also
more capital intensive. Whereas domestic firms are more energy intensive, R&D
intensity is higher for MNE-affiliated firms.
Table 3 presents the correlation matrix of select variables for the full sample.
From the results, we can observe that profitability is positively related to energy
 INDIA: ENERGY USE & FIRM PERFORMANCE 123

Table 2
COMPARISON OF VARIABLES FOR MULTINATIONAL ENTERPRISE-AFFILIATED
a
FIRMS (MNE) AND DOMESTIC FIRMS (FULL SAMPLE)

MNE-Affiliated Firms Domestic Firms

Variables Mean SD Mean SD

Profitability 6.193 10.455 1.058 2.663

Capital intensity 1.734 4.104 0.319 0.737
Energy intensity 0.538 4.724 2.654 0.988
R&D intensity 0.431 1.360 0.074 0.872
Number of observations 367 23,067

a
SD denotes standard deviation; R&D denotes research and development.
Source: Authors’ calculations based upon the Center for Monitoring Indian Economy (CMIE)
PROWESS database.

intensity, R&D intensity, size, and age of the firms. The positive relationship of
profitability suggests that with increases in profitability, there might be positive
change for those variables. However, as the sample is further divided into the three
groups based on the primary source of energy consumption, it will be interesting to
observe the correlation between energy intensity and firm characteristics for each
of the groups.
Table 4 gives the correlation coefficient of the natural gas, petroleum, and coal
groupings. From the table it is evident that firms consuming natural gas as their
primary energy source have negative relationships with profitability, R&D in-
tensity, and firm size; yet, a positive relationship is found with age of the firms.
However, for the two other energy groups—petroleum and coal—the result is
similar to the full sample result and positively related to energy intensity.

Table 3
a
CORRELATION MATRIX (FULL SAMPLE)

Energy R&D Size of Age of

Variables Profitability Intensity Intensity Firm Firm

Profitability 1.000
Energy intensity 0.480 1.000
R&D intensity 0.151 0.158 1.000
Size of firm 0.485 0.418 0.125 1.000
Age of firm 0.134 0.260 0.026 0.205 1.000

a
R&D denotes research and development.
Source: Authors’ calculations based upon the Center for Monitoring Indian Economy (CMIE)
PROWESS database.
124 THE JOURNAL OF ENERGY AND DEVELOPMENT

Table 4
CORRELATION BETWEEN ENERGY INTENSITY AND FIRM CHARACTERISTICS IN
a
ENERGY GROUPS

Energy Intensity of Energy Intensity of Energy Intensity of

Variables Natural Gas Group Petroleum Group Coal Group

Profitability –0.003 0.543 0.599

R&D intensity –0.020 0.335 0.408
Size of firm –0.093 0.548 0.524
Age of firm 0.067 0.380 0.331

a
R&D denotes research and development.
Source: Authors’ calculations based upon the Center for Monitoring Indian Economy (CMIE)
PROWESS database.

Figure 2 presents the mean energy intensity and profitability for the three
energy groups. From the graph, it is evident that firms using natural gas as their
primary energy source are more energy efficient when compared to those using
petroleum and coal. In comparing differences between the petroleum and coal
categories, we can observe that firms in the coal group are more energy intensive
relative to the petroleum group. Regarding profitability, firms that use coal have
the highest profitability whereas those using natural gas are the least profitable.
The cross tabulation and correlation matrix try to depict the relationship of energy
intensity and other firm characteristics with profitability for the full sample as well
as for three energy groups.

Figure 2
MEAN ENERGY INTENSITY AND PROFITABILITY FOR THREE ENERGY GROUPS

Source: Authors’ calculations based upon the Center for Monitoring Indian Economy (CMIE)
PROWESS database.
 INDIA: ENERGY USE & FIRM PERFORMANCE 125

Further, we have estimated the determinants of profitability for the three en-
ergy groups and the full sample where we assume that the determinants of firm
profitability differ for natural gas, petroleum, and coal. The result of the full
sample is given in table 5. The sample size for analysis is 23,434. The minimum
profitability is found to be 1.0 percent, with average profitability of 4.6 percent
and maximum profitability of 9.0 percent across the groups. The overall model R2
is found to be 0.47. Wald chi2 at 9 degrees of freedom is found to be highly
statistically significant at 1 percent. Equation (1) is estimated using both the fixed
and random effects model. The Hausman test statistics of 0.98 rejects the effi-
ciency of fixed effects estimates and, hence, the random effects model is selected.
The robustness of the random effects model is evident from the LM Chi2 test
(significant at the 1-percent level).
From the estimates given in table 5, we can observe that capital and energy
intensity are positively related to profitability and significant at the 1-percent
level, indicating that with increases in capital and energy intensity, profitability of
firms increase. This means that firms that are capital and energy intensive are
profitable. Energy intensity represents all forms of energy use at the firm level and

Table 5
a
ESTIMATES OF FULL SAMPLE
(Dependent variable: profitability)

Independent Variables Coefficient Standard Error z statistics

Capital intensity 0.039 0.023 2.690***

Energy intensity 0.315 0.021 14.820***
R&D intensity 0.026 0.013 2.030**
Firm size –3.782 0.090 –41.830***
Square of firm size 1.524 0.025 61.000***
Age of the firm –0.004 0.001 –3.240***
MNE affiliation of firms (dummy) –0.366 0.221 –1.660*
Constant 2.452 0.238 10.310***
sigma_u 1.589 Number of observations 23,434
2
sigma_e 1.445 R : Within 0.24
2
Rho 0.547 R : Between 0.40
2
Obs per group: Min 1.000 R : Overall 0.47
2
Obs per group: Avg 4.600 Wald chi (9) 9706.58***
2
Obs per group: Max 9.000 LM Chi 21733.40***
2
Hausman Chi 0.98

a
*** = statistically significant at the 1-percent level; ** = statistically significant at the 5-percent
level; * = statistically significant at the 10-percent level; MNE = multinational enterprises; R&D =
research and development.
Source: Authors’ calculations based upon the Center for Monitoring Indian Economy (CMIE)
PROWESS database.
126 THE JOURNAL OF ENERGY AND DEVELOPMENT

therefore can be a catchall factor. R&D intensity is also found to be positively

related and significant to profitability, indicating an increase in R&D intensity
increases the firm’s profitability. We found a nonlinear relationship between
profitability and firm size, indicating a U-shaped relationship. This suggests that
both larger- and smaller-sized firms are more profitable relative to medium-sized
firms. Further, older firms are found to be less profitable relative to younger ones.
The MNE-affiliated firms are found to be more profitable (estimate for MNE
affiliation is significant at the 10-percent level and negative; however, as this
variable is constructed as a dummy, adding to the coefficient a constant it gives
a positive relationship) as compared to the domestic firms. This estimate of
full sample based upon the panel data random effects model gives the de-
terminants of profitability of firms. As we aim to find the determinants of
profitability of firms for the three energy groups, we have also modeled similar
econometric applications in determining factors affecting profitability for the
energy groups.
Table 6 presents the estimates for the three subsamples of energy groups. For
the natural gas group (model 2, second column of table 6), the sample size is 8,568.
The minimum profitability in this group is found to be 1.0 percent with an average
profitability of 2.1 percent and maximum profitability of 5.0 percent across groups.
The overall model R2 is found to be 0.47. Wald chi2 at 9 degrees of freedom is
highly statistically significant at the 1-percent level. From the estimates for the
natural gas group, we observe that capital intensity is positive and significant at the
1-percent level with profitability, indicating an increase in capital intensity increases
firm profitability. Capital-intensive firms are also more profitable. The result for
capital intensity is similar to the full sample estimates. However, the relationship of
energy intensity to profitability is negative and significant at the 1-percent level,
which is a deviation from the full sample estimates. This suggests that firms that use
natural gas are profitable and energy efficient. In other words, increases in energy
efficiencies make firms profitable. Hence, shifting to natural gas as a primary en-
ergy source could be beneficial for firms in achieving profitability and energy ef-
ficiency. Further, we found a nonlinear relationship between profitability and the
size of firm, indicating a U-shaped relationship. This suggests that larger- and
smaller-sized firms are more profitable relative to medium-sized firms. This result is
similar to the outcomes of the full sample estimates. Similarly, older firms are found
to be less profitable relative to younger ones.
In the case of the petroleum group (model 3, third column in table 6), the
sample size is 8,821. The minimum profitability is found to be 1.0 percent with an
average profitability of 2.1 percent and maximum profitability of 4.0 percent
across the group. The overall model R2 is found to be 0.49. Wald chi2 at 9 degrees
of freedom is found to be highly statistically significant at the 1-percent level.
From the estimates of the petroleum grouping, we can find that capital intensity is
positively significant at the 1-percent level, indicating that when capital intensity
 Table 6
a
ESTIMATES OF DETERMINANTS OF PROFITABILITY FOR NATURAL GAS, PETROLEUM, AND COAL GROUPS
(Dependent variable: profitability)

Model 2: Natural Gas Group Model 3: Petroleum Group Model 4: Coal Group
Independent Variables Coefficient S.E. z Statistics Coefficient S.E. z Statistics Coefficient S.E. z Statistics

Capital intensity 0.291 0.023 12.490*** 1.088 0.288 3.780*** 0.220 0.447 2.490**
Energy intensity –0.468 0.287 –2.630** 0.262 0.028 9.370*** 0.800 0.035 22.880***
R&D intensity 0.008 0.013 0.610 0.383 0.046 8.320*** –0.131 0.073 –2.120**
Firm size –2.995 0.108 –27.680*** 3.406 0.115 29.540*** –4.055 0.165 –24.610***
Square of firm size 1.204 0.030 40.170*** –1.365 0.032 –43.060*** 1.453 0.043 33.470***
Age of the firm –0.005 0.001 –4.090*** –0.004 0.002 –2.900*** –0.011 0.002 –5.510***
MNE affiliation (dummy) –0.294 0.208 –1.420 –0.508 0.240 –1.110 0.409 0.314 1.300
Constant 2.088 0.227 9.190*** 2.412 0.260 9.260*** 2.191 0.347 6.320***
sigma_u 1.193 1.326 1.530
sigma_e 1.222 1.464 1.951
2
R : within 0.21 0.26 0.24
2
R : between 0.45 0.46 0.50
2
R : overall 0.47 0.49 0.50
Rho 0.488 0.451 0.381
Obs per group: Min 1.000 1.000 1.000
Obs per group: Avg 2.100 2.100 1.600
Obs per group: Max 5.000 4.000 3.000
2
Wald Chi (9) 4982.440*** 5743.800*** 5018.38***
Number of observations 8568 8821 6045
2
Hausman Chi 1.28 1.08 1.21
INDIA: ENERGY USE & FIRM PERFORMANCE

2
LM Chi 1753.23*** 2133.34*** 1736.12***

a
*** = statistically significant at 1%; ** = statistically significant at 5%; * = statistically significant at 10%; S.E. = standard error; R&D = research and development.
Source: Authors’ calculations based upon the Center for Monitoring Indian Economy (CMIE) PROWESS database.
127
128 THE JOURNAL OF ENERGY AND DEVELOPMENT

increases the profitability of firms also increases. This result is similar to the re-
sults of the full sample as well as for the natural gas sample. R&D intensity is
found to be positively significant at the 1-percent level, indicating that an increase in
R&D intensity also increases the firm profitability. We found a nonlinear re-
lationship between profitability and size of firm, indicating an inverted U-shaped
relationship. This suggests that larger- and smaller-size firms are less profitable as
compared to medium-sized firms. However, for the full and the natural gas group
we found an opposite relationship between firm size and profitability. That means
firm size matters until a threshold level of profitability is achieved and, beyond that
threshold level, profitability declines for firms in the petroleum grouping. Further,
older firms are found to be less profitable compared to younger firms, which is
similar to the estimates of the full and the coal grouping. The estimate of energy
intensity is similar to the estimates of the full sample. Energy intensity is positively
related to profitability. This implies energy-intensive firms in this grouping are
profitable.
Now, turning our attention to the coal group (model 4, fourth column in table
6), we have a sample size of 6,045. The minimum profitability is found to be 1.0
percent with an average profitability of 1.6 percent and maximum profitability of
3.0 percent across this grouping. The overall model R2 is found to be 0.50. Wald chi2
at 9 degrees of freedom is found to be highly statistically significant at the 1-percent
level. From the estimates for this grouping, we can observe that capital intensity is
positively significant at the 1-percent level, indicating that profitability increases
when capital intensity increases, which is similar to the results of the full sample as
well as the natural gas and petroleum groupings. R&D intensity is found to be
negatively significant at the 1-percent level; thus, firms with less R&D intensity are
also profitable. We found a nonlinear relationship between profitability and the size
of firm, demonstrating a U-shaped relationship. This indicates that larger- and
smaller-sized firms are more profitable relative to medium-sized firms. Further,
older firms are found to be less profitable relative to younger ones. The results of
firm size and age are similar to the full and the natural gas group. The estimate of
energy intensity is also similar to the estimates of the full sample, which is positively
related to profitability. This implies that energy-intensive firms in the coal grouping
are profitable.

Conclusion and Policy Implications

This paper is an attempt to understand the relationship between profitability

and energy intensity of Indian manufacturing industries, in general, and, in par-
ticular, for three energy groups of natural gas, petroleum, and coal. The de-
terminates of profitability of firms are estimated for a full sample and for three
energy subsamples. Econometric analysis of the data collected from the CMIE
 INDIA: ENERGY USE & FIRM PERFORMANCE 129

PROWESS at firm level reveals that the relationship between profitability and
energy intensity varies across groups. Energy intensity is positively related to
profitability for three models (the full sample, the petroleum grouping, and the
coal grouping), but not for the natural gas grouping. This suggests that firms
adopting petroleum and coal as their primary energy sources are both energy in-
tensive and profitable. However, firms in the natural gas group are energy efficient
and profitable. R&D intensity is positively related to profitability for the full
sample and the petroleum subsample, suggesting that firms with higher R&D
intensity are profitable. However, for the coal group, less R&D-intensive firms are
also found to be profitable. For all the cases, firm size is found to be nonlinearly
related to profitability. With the exception of the petroleum group, in all other
cases medium-sized firms are less profitable. Age of the firm has a negative re-
lationship with profitability of firms in all the cases implying younger firms are
more profitable. Furthermore, capital intensity is positively related to the profit-
ability in all the cases, indicating capital-intensive firms are profitable. Most of the
earlier literature dealing with determinants of inter-firm differences in profitability
has only examined the role of firm size, age, R&D, and capital intensity. No
specific analysis has been carried out in examining the role of energy intensity (or
efficiency) in determining profitability of firms in Indian manufacturing. This
paper is an attempt to fill this gap. Moreover, since there are large-scale differ-
ences in not only energy intensity but also firm size, R&D, and capital intensity
across different energy groups, this paper documents those differences as well.
The findings do indicate the variable role of energy as well as other firm-specific
characteristics in determining profitability.
Based on the findings, we offer the following policy suggestions for enhanced
firm performance in the Indian manufacturing sector. The econometric results
indicate that firms using natural gas are becoming energy efficient as well as
profitable. Hence, by shifting from coal or petroleum to natural gas as the primary
energy source, firms can become energy efficient and profitable. In addition, by
using natural gas there is a possibility of reducing CO2 emissions from fuel use. In
the debate of clean development mechanisms and issues surrounding climate
change, shifting from traditional fuel sources to a relatively newer fuel source,
such as natural gas, might help in reducing CO2 emissions, specifically for a de-
veloping country such as India. Greater levels of research and development and
technological advancements in the production process as well as for product de-
velopment also will help Indian manufacturing firms in achieving higher profit-
ability and energy efficiency. Fiscal incentives are effective means to stimulate
firms to realize energy conservation projects in their organization. A possible step
could be to reach an agreement between industries and the government whereby
the sector commits itself to reduce CO2 emissions and, on the other hand, the
government commits itself to provide favorable investment conditions for
adopting cleaner fuels.
130 THE JOURNAL OF ENERGY AND DEVELOPMENT

NOTES
1
U.S. Department of Energy, Energy Information Administration (EIA), India: Country Analysis
(Washington, D.C.: EIA, 2014).
2
BP, BP Energy Outlook 2035 (London: BP, 2014).
3
U.S. Department of Energy, Energy Information Administration (EIA), op. cit.
4
Ibid.
5
R. Newell, A. Jaffe, and R. Stavins, “The Induced Innovation Hypothesis and Energy-saving
Technological Change,” Quarterly Journal Economics, vol. 114, no. 3 (1999), pp. 941–75.
6
Defined as the demand for energy services that may increase in response to energy efficiency-
induced decline in the marginal cost of energy services.
7
M. Porter, The Competitive Advantage of Nations (London: Macmillan, 1990).
8
L. O’Connell, C. Van Egeraat, and P. Enright., “The Irish Dairy Processing Industry: An
Application of Porter’s Cluster Analysis,” Research Series Paper no. 1, National Economic and
Social Council, Dublin, Ireland, 1997; P. Clancy and M. Twomey, “The Irish Popular Music In-
dustry: An Application of Porter’s Cluster Analysis,” Research Series Paper no. 2, National Eco-
nomic and Social Council, Dublin, Ireland, 1997; and, for further discussion of their broader
implications, see National Economic and Social Council, Proceedings of NESC Seminar, Research
Series: Sustaining Competitive Advantage (Dublin, Ireland: National Economic and Social Council,
1998).
9
P. Clancy and M. Twomey, op. cit.
10
C. L. Kelton, M. K. Pasquale, and R. P. Rebelein, “Using the North American Industry
Classification System (NAICS) to Identify National Industry Cluster Templates for Applied Re-
gional Analysis,” Regional Studies, vol. 42, no. 3 (2008), pp. 305–21.
11
M. Hirschey and D. W. Wichern, “Accounting and Market-Value Measures of Profitability:
Consistency, Determinants, and Uses,” Journal of Business and Economic Statistics, vol. 2, no. 4
(1984), pp. 375–83.
12
G. Hansen and B. Wernerfelt, “Determinants of Firm Performance: The Relative Importance
of Economic and Organizational Factors,” Strategic Management Journal, vol. 10, no. 5 (1989), pp.
399–411.
13
I. N. Kessides, “Internal versus External Market Conditions and Firm Profitability: An Ex-
ploratory Model,” Economic Journal, vol. 100, no. 402 (1990), pp. 773–92.
14
T. H. Brush, B. Philip, and H. Margaretha, “The Relative Influence of Industry and Corpo-
ration on Business Segment Performance: An Alternative Estimate,” Strategic Management
Journal, vol. 20, no. 6 (1999), pp. 519–47.
15
S. H. Schurr, “Energy Efficiency and Productive Efficiency: Some Thoughts Based on
American Experience,” The Energy Journal, vol. 3, no. 3 (1982), pp. 1–14, and D. W. Jorgenson,
“The Role of Energy in Productivity Growth,” The Energy Journal, vol. 5, no. 3 (1984), pp. 11–26.
 INDIA: ENERGY USE & FIRM PERFORMANCE 131
16
J. Cornillie and S. Fankhauser, “The Energy Intensity of Transition Countries,” Energy
Economics, vol. 26, no. 3 (2004), pp. 283–95.
17
A. Rose and C. Y. Chen, “Sources of Change in Energy Use in the U.S. Economy, 1972–1982:
A Structural Decomposition Analysis,” Resources and Energy, vol. 13, no. 1 (1991), pp. 1–21, and
L. R. Murillo-Zamorano, “The Role of Energy in Productivity Growth: A Controversial Issue?” The
Energy Journal, vol. 26, no. 2 (2005), pp. 69–88.
18
H. L. F. de Groot, E. T. Verhoef, and P. Nijkamp, “Energy Saving by Firms: Decision-making,
Barriers and Policies,” Energy Economics, vol. 23, no. 6 (2001), pp. 717–40.
19
C. I. P. Martı́nez, “Energy Efficiency Development in German and Colombian Non-energy-
intensive Sectors: A Non-parametric Analysis,” Energy Efficiency, vol. 4, no. 1 (2010), pp. 115–31.
20
C. I. P. Martı́nez, op. cit., and H. A. Faruq and D. T. Yi, “The Determinants of Technical
Efficiency of Manufacturing Firms in Ghana,” Global Economy Journal, vol. 10, no. 3 (2010), pp.
1–23.
21
E. Oczkowski and K. Sharma, “Determinants of Efficiency in Least Developed Countries:
Further Evidence from Nepalese Manufacturing Firms,” Journal of Development Studies, vol. 41,
(2005), pp. 617–30.
22
A. Couto and D. Graham, “The Determinants of Efficiency and Productivity in European
Railways,” Applied Economics, vol. 41, no. 22 (2009), pp. 2827–851.
23
L. Lachaal, K. Boubaker, B. Dhehibi, and A. Chebil, “Technical Efficiency Measures and Its
Determinants for Olive Producing Farms in Tunisia: A Stochastic Frontier Analysis,” African
Development Review, vol. 17, no. 3 (2005), pp. 580–91.
24
E. Oczkowski and K. Sharma, op. cit.
25
H. A. Faruq and D. T. Yi, op. cit.
26
N. Hrovatin and S. Ursic, “The Determinants of Firm Performance after Ownership Trans-
formation in Slovenia,” Communist and Post-Communist Studies, vol. 35, no. 2 (2002), pp. 169–90.
27
P. Zhou, B. W. Ang, and K. L. Poh, “A Survey of Data Envelopment Analysis in Energy and
Environmental Studies,” European Journal of Operational Research, vol. 189, no. 1 (2008), pp.
1–18.
28
N. Kumar and N. S. Siddharthan, “Technology, Firms Size and Export Behavior in De-
veloping Countries: The Case of Indian Enterprises,” Journal of Development Studies, vol. 31, no. 2
(1994), pp. 289–309; and B. Goldar, “Energy Intensity of Indian Manufacturing Firms: Effect of
Energy Prices, Technology and Firm Characteristics,” Working Paper no. 2483, eSocialSciences,
2010.
29
T. Papadogonas, J. Mylonakis, and D. Georgopoulos, “Energy Consumption and Firm
Characteristics in the Hellenic Manufacturing Sector,” International Journal of Energy, Technology
and Policy, vol. 5, no. 1 (2007), pp. 89–96.
30
Ibid.
132 THE JOURNAL OF ENERGY AND DEVELOPMENT
31
N. Kumar and N. S. Siddharthan, op. cit.; S. Sahu and K. Narayanan, “Determinants of Energy
Intensity: A Preliminary Investigation of Indian Manufacturing,” MPRA Paper no. 16606, Uni-
versity Library of Munich, Germany, 2009; S. Sahu and K. Narayanan, “Determinants of Energy
Intensity in Indian Manufacturing Industries: A Firm Level Analysis,” MPRA Paper no. 21646,
University Library of Munich, Germany, 2010; and B. Goldar, op. cit.
32
N. Kumar and N. S. Siddharthan, op. cit.; B. Goldar, op. cit.; S. Sahu and K. Narayanan,
“Determinants of Energy Intensity: A Preliminary Investigation of Indian Manufacturing,” MPRA
Paper no. 16606, 2009; and S. Sahu and K. Narayanan, “Determinants of Energy Intensity in Indian
Manufacturing Industries: A Firm Level Analysis,” MPRA Paper no. 21646, 2010.
33
N. Kumar and N. S. Siddharthan, op. cit., and S. Sahu and K. Narayanan, “Determinants of
Energy Intensity: A Preliminary Investigation of Indian Manufacturing.”
34
S. Sahu and K. Narayanan, “Determinants of Energy Intensity in Indian Manufacturing In-
dustries: A Firm Level Analysis.”
35
T. Papadogonas et al., op. cit.
36
K. Fisher-Vanden, G. H. Jefferson, H. Liu, and Q. Ta, “What is Driving China’s Decline in
Energy Intensity?” Resource and Energy Economics, vol. 26, no. 1 (2004), pp. 77–97.
37
M. Morikawa, “Population Density and Efficiency in Energy Consumption: An Empirical
Analysis of Service Establishments,” Energy Economics, vol. 34, no. 5 (2012), pp. 1617–622.
38
S. J. DeCanio, “The Efficiency Paradox: Bureaucratic and Organizational Barriers to Profit-
able Energy-saving Investments,” Energy Policy, vol. 26, no. 5 (1998), pp. 441–54; M. A. Brown,
“Market Failures and Barriers as a Basis for Clean Energy Policies,” Energy Policy, vol. 29, no. 14
(2001), pp. 1197–207; and R. Martin, M. Muuls, L. B. de Preux, and U. J. Wagner, “Anatomy of
a Paradox: Management Practices, Organizational Structure and Energy Efficiency,” Journal of
Environmental Economics and Management, vol. 63 (2012), pp. 208–23,
39
S. J. DeCanio, op. cit.
40
A. Trianni and E. Cagno, “Dealing with Barriers to Energy Efficiency and SMEs: Some
Empirical Evidences,” Energy, vol. 37, no. 1 (2012), pp. 494–504.
41
N. C. Bloom, C. Genakos, R. Martin, and R. Sadun, “Modern Management: Good for the
Environment or Just Hot Air?” The Economic Journal, vol. 120, no. 544 (2010), pp. 551–72.
42
R. Martin et al., op. cit.
43
S. K. Sahu and K. Narayanan, “Carbon Dioxide Emissions from Indian Manufacturing In-
dustries: Role of Energy and Technology Intensity,” Madras School of Economics Working Paper
no. 82, Madras School of Economics, Chennai, India, 2013, available at http://www.mse.ac.in/pub/
Working%20Paper%2082.pdf.
44
Since a large number of small- to medium-sized enterprises do not report/appear in the Center
for Monitoring Indian Economy (CMIE) database, the analysis is restricted to the sample of large
firms listed in the CMIE database. The internet database of the CMIE PROWESS was accessed on
February 15, 2014.
 INDIA: ENERGY USE & FIRM PERFORMANCE 133
45
S. K. Sahu and K. Narayanan, “Determinants of Energy Intensity in Indian Manufacturing
Industries: A Firm Level Analysis,” Eurasian Journal of Business and Economics, vol. 4, no. 8
(2011), pp. 13–30, and B. Goldar, op. cit.
46
N. S. Siddharthan, B. L. Pandit and R. N. Agarwal, “Growth and Profit Behaviour of Large
Indian Firms,” The Developing Economies, vol. 32, no. 2 (1994), pp. 188–209.
47
J. A. Hausman, “Specification Tests in Econometrics,” Econometrica, vol. 46, no. 6 (1978),
pp. 1251–271.
48
D. M. Roberts and J. Tybout, “Market Entry Costs, Producer Heterogeneity and Export Dy-
namics,” Econometrica, vol. 75, no. 3 (2007), pp. 837–73.
49
H. A. Faruq and D. T. Yi, op. cit., and V. Zelenyuk and V. Zheka, “Corporate Governance and
Firm’s Efficiency: The Case of Transitional Country,” Journal of Productivity Analysis, vol. 25,
nos. 1–2 (2006), pp. 143–57.