Alex Coad-The Growth of Firms - A Survey of Theories and Empirical Evidence (New Perspectives On The Modern Corporation) - Edward Elgar Pub (2009)

The Growth of Firms
NEW PERSPECTIVES ON THE MODERN CORPORATION

Series Editor: Jonathan Michie, Director, Continuing Education and President,
Kellogg College, University of Oxford, UK
The modern corporation has far reaching influence on our lives in an increasingly
globalised economy. This series will provide an invaluable forum for the publica-
tion of high quality works of scholarship covering the areas of:
● corporate governance and corporate responsibility, including environmental

sustainability
● human resource management and other management practices, and the
relationship of these to organisational outcomes and corporate perform-
ance
● industrial economics, organisational behaviour, innovation and competitive-
ness
● outsourcing, offshoring, joint ventures and strategic alliances
● different ownership forms, including social enterprise and employee
ownership
● intellectual property and the learning economy, including knowledge
● transfer and information exchange.
Titles in the series include:
Corporate Governance, Organization and the Firm

Co-operation and Outsourcing in the Global Economy
Edited by Mario Morroni
The Modern Firm, Corporate Governance and Investment

Edited by Per-Olof Bjuggren and Dennis C. Mueller
The Growth of Firms

A Survey of Theories and Empirical Evidence
Alex Coad
Knowledge in the Development Economies

Institutional Choices Under Globalisation
Edited by Roger Sugden and Silvia Sacchetti
The Growth of Firms
A Survey of Theories and Empirical
Evidence
Alex Coad
Evolutionary Economics Group, Max Planck Institute of
Economics, Germany
NEW PERSPECTIVES ON THE MODERN CORPORATION
Edward Elgar
Cheltenham, UK • Northampton, MA, USA
© Alex Coad 2009
All rights reserved. No part of this publication may be reproduced, stored in a

retrieval system or transmitted in any form or by any means, electronic,
mechanical or photocopying, recording, or otherwise without the prior
permission of the publisher.
Published by
Edward Elgar Publishing Limited
The Lypiatts
15 Lansdown Road
Cheltenham
Glos GL50 2JA
UK
Edward Elgar Publishing, Inc.

William Pratt House
9 Dewey Court
Northampton
Massachusetts 01060
USA
A catalogue record for this book

is available from the British Library
Library of Congress Control Number: 2009924303
ISBN 978 1 84844 327 3
Printed and bound by MPG Books Group, UK

Contents
List of figures and tables vi
Acknowledgements vii
1 Introduction 1
2 Firm size distributions 14
3 Growth rate distributions 25
4 Gibrat’s law 39
5 Profits, productivity and firm growth 49
6 Innovation and firm growth 76
7 Other determinants of firm growth 84
8 Theoretical perspectives 100
9 Growth strategies 111
10 Growth of small and large firms 129
11 Conclusion 143
Notes 152
Bibliography 161
Index 191
v
Figures and tables
FIGURES
2.1 Firm size distribution: French manufacturing 15

2.2 Firm size distribution: US manufacturing 15
2.3 Age distribution for small Indian firms 21
2.4 Age distribution for Spanish firms 21
3.1 Distribution of sales growth rates 27
3.2 Distribution of employment growth rates 27
3.3 Firm growth as the addition of discrete resources 33
3.4 Employment growth in a model of firm growth 35
5.1 Scatterplots of profits vs sales growth 50
5.2 Process of firm growth and productivity growth 72
5.3 Process of firm growth and R&D expenditure 74
10.1 Stages of growth 130
TABLES
2.1 The share of activity of small firms 18

5.1 Deriving a regression equation in a neoclassical q model 52
5.2 Profits and growth: types of regression equation 60
5.3 Decomposing productivity growth 67
7.1 Values of R2 obtained from growth regressions 97
vi
Acknowledgements
A large number of people helped in the preparation of this book, at various
stages of its development. I am very grateful for their help. In particular, I
am very much indebted to (listed alphabetically): David Audretsch, Erkko
Autio, Christian Cordes, Giovanni Dosi, Elizabeth Garnsey, Vojislav
Maksimovic, Rie Nemoto, Bernard Paulré, Christos Pitelis, Rekha Rao,
Angelo Secchi, Agusti Segarra, Erik Stam, Fede Tamagni, Mercedes
Teruel, and Ulrich Witt. I am also very grateful to Matt Pitman, commis-
sioning editor at Edward Elgar, for his remarkable patience and support.
Any remaining shortcomings in this book are my own responsibility,
however.
vii
1. Introduction
Research into firm growth has been accumulating at a terrific pace, and is
being published in a growing range of outlets, such as journals relating to
the disciplines of economics, management, sociology, entrepreneurship, as
well as disciplines as diverse as statistical physics and psychology.
The past few decades have witnessed much progress in empirical research
into firm growth, in particular, for a number of reasons. First, datasets
documenting economic phenomena are growing in terms of their level of
detail, sample size and availability. The rise of information technology has
played a major role in this trend. Many countries have statistical offices that
undertake censuses of business firms and establishments, creating longitudi-
nal databases that track individual firms over time, and make these records
available to researchers (under restrictions of confidentiality, of course).
Firms are required to provide information on themselves and their opera-
tions at a level of detail that is quite remarkable. For example, many firms
are required to file financial reports that describe their operations not just
at the aggregated level, but disaggregated by production plant or by line of
business. Even ‘soft’ variables, such as entrepreneurial growth aspirations,
are becoming commonplace in quantitative statistical analyses – these
variables can be measured using subjective responses of individuals to large-
scale questionnaires. A second development favouring empirical research
into economics is that econometric techniques have kept pace with the
availability of increasingly informative datasets. Modern econometric work
is able to deal with such complicated issues of endogeneity, unobserved het-
erogeneity within individuals, and sample selection bias. The progress that
has been made in this domain has been reflected by the number of Nobel
memorial prizes awarded to econometricians in recent years. Breakthroughs
in econometric theory have granted more legitimacy to empirical findings,
and have also allowed researchers to investigate more elaborate hypotheses.
Old results have been turned on their heads when new, improved statistical
methods have been applied. A third major development is that continual
increases in computational power have been able to match developments
in databases and econometric techniques. Bootstrapping methods, for
example, are particularly computationally intensive and their use has only
become feasible thanks to developments in the performance of computers.
1
2 The growth of firms
To keep up with developments in this field, we need an up-to-date cata-

logue of empirical work, as well as a coherent theoretical structure within
which these new results can be interpreted and understood. The aim of this
monograph is to provide such an overview. The need for such a book arises
because it is increasingly difficult to keep abreast of the latest developments
in the field. Empirical investigations into firm growth have multiplied, and
this has often occurred in tangential directions. The matter is complicated
further when one realizes that research methodologies can be quite differ-
ent, and hence difficult to compare, following on from the intuition that
firm growth can mean different things to different people. In reaction to
this, Garnsey et al. (2006) write that ‘it is essential to have related explana-
tory concepts to guide inquiry and make sense of evidence. A mass of undi-
gested empirical findings can be misleading.’ (p. 4). This book attempts
to take stock of the major findings in the literature, and endeavours to
provide coherent explanations for them, and to knit them together in such
a way as to give the reader an all-round appreciation of the phenomenon
of firm growth. The book also attempts to summarize the research into firm
growth that has already been done, so that future researchers can expand
upon this knowledge base to obtain further insights into the processes of
firm growth and organizational development.
The rest of this introductory chapter is organized as follows. In section
1.1 we discuss the historical context in which firm growth is placed, discuss-
ing how the roles and incentives of growing firms have changed in recent
economic history. Section 1.2 discusses the broad theoretical foundations
that we consider as a helpful starting point for the book. Section 1.3 con-
tains a practical discussion of how growth can be measured, before we
launch into the main text of the book. An outline of the book is given in
section 1.4.
1.1 FIRM GROWTH IN A HISTORICAL

PERSPECTIVE
It is instructive to place firm growth in a historical perspective. In the past,

a large size was a pre-rerequisite for security. Firms strove to become large
in order to guarantee their future. The advantages of a large size were rein-
forced by the relatively backward state of financial markets. Large firms
had the advantage of ‘deeper pockets’ into which they could delve during
adverse business conditions. Another factor to be taken into consideration
is that at the beginning of the twentieth century, the ‘Fordist’ brand of
mass-production techniques were very much in vogue. During this period,
the growth of firms was associated with economies of scale and lower unit
Introduction 3
costs. Furthermore, as firms continued to expand they began to question

the mono-product business model that had hitherto been the norm. In this
vein, Du Pont de Nemours achieved legendary success by engaging in a
diversified portfolio of activities arranged in the context of a decentralized
and multidivisional organizational form. Other firms, from science-based
industries in particular, began to diversify into new product markets, in
search of opportunities for exploiting economics of scale and scope in
production and for R&D (Chandler, 1992). In addition, it was conjectured
(for example by Schumpeter) that it was primarily the large firms that were
willing and capable of investing in R&D laboratories. Large size was there-
fore considered to be a sign of the accomplishment of a firm’s aspirations,
and as the ultimate stage in a firm’s development.
The present business climate, however, is different from what it used to
be in a number of ways. In contrast to the previous imperatives of scale
and scope, contemporary strategists place more emphasis on flexibility and
‘lean’ production. There is evidence that firms’ hierarchies are becoming
flatter, the CEO span of control is becoming broader, intermediate manag-
ers are being dispensed with, and divisional managers are receiving more
authority, higher pay, and greater incentive pay as they become closer to
the CEO (Rajan and Wulf, 2006). We are now in an age where downsizing
and refocusing are celebrated strategies. A capitalism based on mass pro-
duction and standardization has given way to an organization of produc-
tion based on customization and product differentiation. Improvements
in financial markets, and the aversion of shareholders to diversified firms
(and conglomerates in particular) has brought on the disintegration of
the large ‘Chandlerian’ firm. Information Technology has played a role in
this, allowing firms to increase the flexibility of their production lines. In
successful firms, evidence suggests that the introduction of productivity-
enhancing Information Technology has been accompanied by widespread
organizational change (Brynjolfsson and Hitt, 2000; see also Acemoglu et
al., 2007). Furthermore, Information Technology has helped reduce trans-
action costs of dealing with other firms, thereby reducing the incentives
for firms to be fully integrated along their respective production chains or
‘filières’. In the context of the ‘make-or-buy’ dilemma, firms need to be less
cautious about dealing with suppliers through the market mechanism, even
if this means the outsourcing of services from far-away continents. The fast
pace of change in markets has led to the emergence of a new stereotype –
the lean, flexible firm whose competitive advantage rests on a focus on a
small number of core competences.
More generally, the fast pace of the information age has changed the way
firms operate, bringing customers closer to their suppliers and shortening
product cycles. Industries are becoming more and more turbulent and the
competitive struggle is becoming increasingly fierce. The focus of investors

on a firm’s market value, and current incentive schemes in place within
firms, put pressure on managers to satisfy short-term goals – ‘in the eyes of
today’s top management, “long term” means about 18 months.’ (Marris,
1999, p. 56). This short planning horizon is balanced, somewhat paradoxi-
cally, against the need for large-scale R&D projects that can have pay-back
times of 10–15 years, or even more (Grabowski et al., 2002).
The globalization of business operations and the rise of multinational
enterprises (MNEs) is another challenge that is gaining increasing impor-
tance. Large firms, in particular, now face competition on a global scale.
Firms must look for business opportunities overseas, and must also face the
threats coming from overseas rivals even in their own domestic markets. In
order to maintain or create competitive advantage, firms nowadays have
to look for business opportunities in foreign markets, using strategies such
as exporting, joint ventures or even greenfield construction of production
plants overseas through foreign direct investment (FDI).
The development of financial institutions and the increasing availability
of external finance is another factor that has enabled firms to accelerate their
expansion projects – whether it be ‘organic’ growth or growth by merger
and acquisition. In the past, financial markets were not developed, and it
was suggested that firm growth came about by reinvesting retained profits
into the firm (Chandler, 1992). These days, however, small firms can turn to
venture capitalists or specialized stock markets to realize their high growth
ambitions. In addition, a large number of government schemes are in place
to ease financial constraints on new firms. We have entered the era of the
‘entrepreneurial society’ (Audretsch, 2007). Indeed, many economists now
seem to consider that access to finance is the small firm’s birthright.
In the light of this discussion, it would appear to be necessary to take a
new look at the subject of the growth of firms – a subject which still, argu-
ably, remains dominated by the seasoned works of Gibrat (1931), Penrose
(1959) and Marris (1964).
1.2 THEORETICAL FOUNDATIONS
Early theoretical work into the size and growth of firms was placed in a
comparative statics framework, and by reason of its static nature did not
really deal with the dynamic phenomenon of growth. Firms were supposed
to be at their ‘optimal size’; and if they weren’t there already, they were
assumed to grow instantaneously to reach it. In this way, firm growth
received a cursory treatment as an appendage to the optimal size theory.
Firms were considered to grow only inasmuch as this enabled them to
Introduction 5
reach their optimal size. However, dissatisfaction with this theory of firm
behaviour has grown in recent decades. Notions of an ‘optimal size’ have
been rejected in almost any interpretation of the phrase that one might
subscribe to (see section 8.1).
Growing dissatisfaction with the conventional static approach of eco-
nomic theory has led to the ascendancy of new themes in theoretical work.
Emphasis has been placed on the prevalence of uncertainty, change and
bounded rationality in the context of a turbulent and restless economy. It
has been suggested that firm growth has replaced firm size as the central
variable in industrial economics (Marris, 1999).
Uncertainty, and also bounded rationality on the part of firms, are
important foundations for an analysis of firm growth, because growth
inevitably involves expansion into new areas. Uncertainty is magnified,
of course, in dynamic markets that are continually being transformed by
technological innovation and competition. In addition, we note that the
firm itself is changing, through growth, in ways it cannot foresee.
Path-dependency is also an important theme. Firms can be seen as
bundles of specific capabilities, or as the repositories of organizational
routines (Nelson and Winter, 1982; Dosi et al., 2000). Firms tend to be
specialized in what they do and they cannot easily change from one day to
the next. What a firm did in the past defines what it can do in future, and
so a firm’s growth opportunities are very much constrained by its current
production activities. Competitive advantage rests to a large extent on
accumulated firm-specific resources as well as production capabilities that
have been carefully developed over time, and the gradual nature of this
process places a limit on the ability of firms to adapt rapidly to a changing
environment.
In addition, we feel it is necessary to recognize the great heterogeneity
that exists between firms, whether we consider productivity levels, prof-
itability, or a large number of other key dimensions. As Griliches and
Mairesse (1995) explain (p. 23):
We also thought that one could reduce aggregation biases by reducing the heter-
ogeneity as one goes down from such general mixtures as ‘total manufacturing’
to something more coherent, such as ‘petroleum refining’ or the ‘manufacture
of cement’. But something like Mandelbrot’s fractals phenomenon seems to be
at work here also: the observed variability-heterogeneity does not really decline
as we cut our data finer and finer. There is a sense in which different bakeries
are just as much different from each other, as the steel industry is from the
machinery industry.
(See also Dosi and Grazzi (2006) for further evidence of pervasive
heterogeneity of firms, even at finely disaggregated levels.) We should
be cautious of notions of a ‘representative firm’ which might lead us to

overlook this heterogeneity.
Although the notion of the ‘representative firm’ has been qualified (if
not discredited) in theoretical discourse, it can still be seen to persist in
a nuanced form in empirical work. The hypothesis of the ‘representative
firm’ in empirical research can be found implicitly in conventional regres-
sion estimators that focus on summary point estimates corresponding to
‘the average effect for the average firm’. This approach is particularly ill-
suited for looking at the relationship between innovation and firm growth,
for example, because innovating firms have fundamentally heterogeneous
performance differences – a small minority of firms doing spectacularly
well whilst in most cases R&D efforts will yield nothing substantial. As is
evident from the ‘tent-shaped’ plots of the firm growth rates distribution
(introduced into economics by Giulio Bottazzi, Giovanni Dosi, Angelo
Secchi and colleagues; see Figures 3.1 and 3.2 in Chapter 3 for an example),
we see that the average firm does not grow very much at all. We argue
that there is little point in trying to find the determinants of growth for
the ‘average firm’, because this latter grows so little that its growth could
be due to almost anything (hence the highly idiosyncratic component
in growth rates that is commonly found). Instead, it is just a handful of
extreme-growth firms that are responsible for a disproportionate share of
the turbulence and reallocation that drives industry dynamics. Focusing
on the ‘average firm’ in the case of firm growth rates would be to misplace
our attention. An important theme in Chapters 3 and 6, and also the rest
of this book, is that it is a heterogeneous minority of agents that is driving
the process of industrial evolution.
Our survey of firm growth is therefore loosely guided by the evolu-
tionary economics perspective, and this is for several reasons. First, this
perspective explicitly recognizes the heterogeneity of firms. At any time,
we can expect there to be considerable diversity in the characteristics of
firms. Whilst the least viable firms can be expected to be eliminated due to
selection pressures, there will remain at any time a marked heterogeneity
between the surviving firms, even among dimensions such as productivity
and production methods. The importance of such an evolutionary vision of
the economy has been further underlined by recent observations that selec-
tion pressures are rather weak. Second, evolutionary economics is based
on what Sid Winter has called a ‘dynamics first!’ approach. A dynamic
view of firms and industries is obviously an essential ‘point de départ’ for
our study of the growth of firms. Third, evolutionary economics embraces
the phenomenon of innovation in a way that other perspectives are not
able to do. The importance of firm-level innovative activity has grown
tremendously over the last decades, and we need a theoretical framework
Introduction 7
that will take this into account. This is especially true given that Chapter
6 focuses specifically on firm-level innovation. Fourth, the low rationality
assumptions that form the basis of the evolutionary framework strike us as
simply being far more judicious than the ‘Olympian’ rationality frequently
assumed in the neoclassical paradigm. Uncertainty is unquestionably one
of the most fundamental features of the modern economy, and it seems
to us to be also one of the defining characteristics of firm growth. Firm
growth is essentially a venture into unfamiliar territory. Indeed, in Chapter
5 (section 5.1) we criticize the mainstream literature that takes the assump-
tion of infinitely rational profit-maximizing firms as a foundation for its
empirical work into firm-level investment patterns. Instead, we delve into
evolutionary theory to obtain a guiding theory. In section 8.4 we investi-
gate the evolutionary principle of ‘growth of the fitter’ and it is astonish-
ing to observe that even this general principle, when taken literally, does
not appear to hold. It seems that even evolutionary economics, which has
genuinely mild rationality assumptions, may be overstating the capacity of
the forces of economic development.
A final motivation for basing our analysis in the evolutionary perspec-
tive is that it appears to be more or less in accordance with the empirical
facts. One of the few regularities that has emerged from research into the
growth of firms is that Gibrat’s ‘law of proportionate effect’ appears to
provide a better description of industrial development than any other
alternative theory. Although Gibrat’s law is frequently criticized as having
no theoretical content (due to the emphasis on purely stochastic shocks),
on the contrary it is our (controversial) view that Gibrat’s law does have
a theoretical basis, and that it is not too far-fetched to consider that this
basis is of an ‘evolutionary’ flavour. We have three reasons for making this
association. First, Gibrat’s law emphasizes heterogeneity between firms
that stems from the variance of the growth shocks. Second, the stochastic
nature of Gibrat’s law can be seen to emphasize the inherent uncertainty
that permeates modern capitalism. Third, Gibrat’s law accommodates the
evolutionary principle of path dependency by the fact that a firm’s current
size is viewed as the mere amalgamation of all previous growth shocks.
The evolutionary setting of this book does not go far in predicting how
much a particular firm will grow, however – instead it provides a theoreti-
cal setting in which empirical work can be grounded. The prevalence of
uncertainty and also pervasive heterogeneity of firms in the context of a
turbulent and restless economy suggests to us that the state of the economy
cannot be worked out from the armchair. Instead, our understanding of
the growth of firms must progress through solid empirical analysis. This
necessarily involves ‘getting one’s hands dirty’ and working with data.
We feel obliged to reiterate an exhortation that is dated but nonetheless
still very relevant: ‘The subject of organizational growth has progressed

beyond abysmal darkness. It is ready for – and badly needs – solid, system-
atic empirical research directed toward explicit hypotheses and utilizing
sophisticated statistical methods’ (Starbuck, 1971, p. 126).
This book can be split up, quite roughly, into empirical and theoreti-
cal chapters, and it appears that, when the chapters are tallied by sheer
headcount, the first six chapters appear to be oriented towards empirical
work while the remaining three are theoretical. This should not be taken
as an indication that theoretical work is to be marginalized in comparison
with empirical work. In contrast, we emphasize that empirical work should
be shaped and guided by theoretical hypotheses. The nature of empirical
investigation into firm growth is not a random exploratory process, but is
most effective when it refers to theoretical concepts and the interpretation
requires that theories be constructed.
In our survey of broad theoretical predictions of firm growth in Chapter
8, it appears that no single theoretical perspective is able to provide an
overview of firm growth. It seems that general, overarching theories are
not that helpful, but instead theories need to be tailored somewhat to their
specific context.
Some theoretical building blocks are clear. In contrast to the neoclassi-
cal assumptions of perfect rationality over an infinite horizon, we take the
opposite view – that firms are not rational, many fail, and that many miss
opportunities. Furthermore, many firms may shape their own destinies,
as it were, and make opportunities for themselves that did not seem to
exist before. Smaller firms, in particular, appear to be rather irrational,
although the behaviour of larger firms appears to be more predictable,
perhaps because they are more inert and lack the flexibility that small
firms have.
1.3 MEASURING SIZE AND GROWTH
A number of the chapters in this book (Chapters 2 to 7 in particular) refer

to quantitative empirical investigations into firm growth, and the basic unit
of analysis is a firm’s growth rate. In this section, we explain what is meant
by a growth rate for a firm in a given year.
In some of the later chapters, however (for example Chapters 9 and
10), we deliberately emphasize the more subtle, qualitative aspects of firm
growth, and the processes of structural change within growing organiza-
tions. These later chapters play an important role in the book, because
they remind the reader that what we call an ‘observation’ in the ensuing
econometric investigations (that is a growth rate for a firm in a given year)
Introduction 9
is not just a ‘statistic’ but actually has a much deeper significance. This is
indeed one of the dangers of empirical work – one can get so accustomed
to dealing with numbers that one may forget what the numbers actu-
ally represent. (Indeed, this has led some individuals to be unnecessarily
apprehensive about empirical work in general.)
But without further ado, we now discuss how firm size and growth can
be measured.
Measuring Size
The number of different indicators of firm size is rather vast, and is limited
only by the imagination of the researcher. Employment and total sales,
however, are the most commonly used indicators (Delmar, 1997). This is in
part because data on employment and sales is among the easiest to obtain.
In the majority of cases, it will make little difference which firm size indica-
tor is taken, as they tend to give similar results. (This should not be taken
for granted, however.)
Among the candidate indicators of firm size and growth, a major advan-
tage of employment is that, unlike financial quantities, it does not need to
be deflated. This is useful for multi-sector analyses, where sector-specific
deflators need not be sought out. It is also useful for the cross-country
analyses, or investigations involving multinational corporations, because
exchange rate complications are avoided. A drawback of employment,
however, is that indivisibilities are substantial for small firms that have
only a few employees. Sales is also frequently taken as a measure of firm
size. One disadvantage of sales, though, is that it need not necessarily
correspond to the actual value-added of a company. Consider the case
of a firm that buys an almost finished product (for example computers)
that is modified or repackaged in some minor way before being sold on to
others. Such a firm will have a high sales figure, because of the high cost
of the final product, even though its contribution to the overall economy
in terms of value-added will be low. If this firm then goes on to acquire its
upstream components suppliers, its total sales will not change but its share
of value-added will of course have increased. Value-added may thus be a
better indicator of firm size than sales, because it takes into account the cost
of materials used in the production process. In practice, however, data on
value-added is not always available, and the peculiar scenario described
above does not occur very often.
There are also many other measures of firm size in use. Another popular
measure is total assets – although this indicator encounters difficulties if
the firms in the sample have different capital intensities. Some authors (for
example Little, 1962; Baumol et al., 1970) speak of firm growth as referring
to growth of profits. Among the least conventional indicators, one finds

‘acres of land’ or ‘head of cattle’ (Weiss, 1998). In this survey we consider
growth in terms of a range of indicators, although we devote relatively little
attention to the growth of profits. This is because total profits is more of a
financial than an economic variable, and it often takes on negative values
(although the concept of a negative firm size has little meaning for the
empirical researcher).1
Measuring Growth
The most common way of measuring growth is by taking log-differences

of size.
Sit 2 Si,t21
Git 5 (1.1)
Si,t21
St St21
5 2
St21 St21
St
5 21
St21
where Sit is the size of firm i at time t. Taking logs, and remembering that
log(1) 5 0, we obtain:
log(Git) 5 log((Sit/Si,t−1) 2 1) 5 log(Sit/Si,t−1) 5 log(Sit) 2 log(Si,t−1)

(1.2)
There are also a few other ways of measuring growth rates.

Some investigations into the growth of young firms simply take size as
an indicator of growth (see for example Eisenhardt and Schoonhoven,
1990; Storey, 1994; Colombo and Grilli, 2005). The justification for this is
that the initial size Si,t−1 is zero since these young firms had zero size in the
initial period (that is shortly before they came into existence). A drawback
of this methodology is that it confounds the effects of start-up size and
subsequent growth.
Some authors, such as Delmar et al. (2003) and Shepherd and Wiklund
(2009), make the distinction between relative growth (that is the growth
rate in percentage terms) and absolute growth (usually measured in the
absolute increase in numbers of employees). Absolute growth is used rela-
tively frequently in the literature on the growth of small entrepreneurial
firms, where the firms being analysed can be very small. Absolute growth
may be preferred by policy makers, for example, who tend to be more
Introduction 11
concerned with the number of jobs in a region than the performance of

individual firms.
Measuring growth in relative or absolute terms can indeed give different
results (Almus, 2002). In this vein, we can mention the ‘Birch index’ (Birch,
1987) which is a weighted average of both relative and absolute growth
rates (this latter being taken into account to emphasize that large firms, due
to their large size, have the potential to create many jobs).
The Birch index can be presented like this:
Eit
Birch Indexit 5 (Eit 2 Ei,t21) # (1.3)
Ei,t21
where E represents the total employment of firm i at time t.
The usual method of measuring growth (as presented in equations 1.1
and 1.2), is that the growth increment is measured relative to the initial size.
The initial size may be a poor indicator of a firm’s actual size, however. If
the firm’s initial size is low because of an unusual temporary shock, then
the growth achieved over the period will be abnormally high when it is
scaled down by initial size. The sorting of growing entities according to
their size is not an easy statistical task, and, as a result, other measures of
scaling down growth, instead of using initial size, have been put forward.
For example, Friedman (1992) recommends that growth rates are scaled
down by average size, or perhaps by final size.
The growth rate index popularized by Davis et al. (1996) measures
growth relative to average size rather than initial size. This index can be
written as follows:
Eit 2 Ei,t21
DHS Indexit 5 (1.4)
1/2 (Ei,t21 1 Ei,t)
The growth increment, Eit 2 Ei,t−1, is thus scaled down by the average size
over the period of analysis.
It is trivial to verify that the growth rates obtained from this procedure
range from a growth rate of 12 (in the case of a firm starting from zero
size at t 2 1), to a growth rate of 22 (in the case of an exiting firm that has
zero size at time t).
This survey focuses predominantly on relative growth rates. It is prima-
rily a discussion of firm growth and does not discuss the processes of entry
and exit in any great detail. Furthermore, in our description of the proc-
esses of expansion (for example Chapters 9 and 10) we emphasize positive
growth rather than negative growth. This is a reflection of the fact that
organizational decline has received relatively little attention in the litera-
ture. (For an introduction to organizational decline, the reader is referred
to Whetten, 1987).
1.4 STRUCTURE OF THE BOOK

In the Simonian spirit of scientific investigation (see for example Simon,
1968) we begin by gathering together some stylized facts and empirical
insights (Chapters 2–7) before looking for theoretical explanations of these
results. In the empirically-oriented chapters, we begin by considering first
the distributions of firm size and growth rates, before moving on in search
of the determinants of growth rates. We then present some broad theories
of firm growth and evaluate their performance in explaining the stylized
facts that emerge from empirical work. We then move on from these
general theoretical perspectives to discuss some more descriptive theories
of firm growth processes in the later chapters.
To begin with, we take a non-parametric look at the distributions of firm
size and growth rates, before moving on to results from regressions that
investigate the determinants of growth rates. Chapter 2 discusses research
into the firm size distribution, which can be considered to be one of the
oldest stylized facts in the industrial economics literature.
We then proceed to discuss research into the growth rates distribution
(Chapter 3). The characteristic shape of the growth rates distribution
was discovered only recently, but offers unique insights into the growth
patterns of firms.
Our survey of the determinants of firm growth begins in Chapter 4,
where Gibrat’s law of proportionate effect is presented. This law predicts
that firm growth is a purely random phenomenon and that firm growth is
independent of firm size. Chapter 5 examines the relationship between rela-
tive performance (that is profits or relative productivity) and firm growth,
and an inquiry into the relationship between innovation and firm growth
can be found in Chapter 6. Chapter 7 considers the role of other variables
in explaining the variation in firm growth rates.
One of the main results that emerges from the literature review into the
determinants of firm growth is that, in line with Gibrat’s law, the random
element of growth rates is predominant. Efforts to identify the determi-
nants of firm growth have had a limited success, and the combined explan-
atory power of the explanatory variables (summarized by the R2 statistic)
is typically low, usually below 10 per cent (see Table 7.1).
Bearing the newly-discovered empirical regularities in mind, we turn
to Chapter 8 in search of some theoretical explanations. The broad theo-
ries of firm growth outlined in this chapter have only a limited success in
explaining growth, perhaps because growth is idiosyncratic, firms are het-
erogeneous, and as a result of this it is difficult to make wide-reaching gen-
eralizations. Moving on, therefore, Chapters 9 and 10 contain qualitative,
theoretical accounts of firm growth processes that are more tailored to the
Introduction 13
data and more descriptive in nature. Chapter 9 describes firms’ attitudes

to growth, as well as the modes of growth available to a firm that seeks
growth. Chapter 10 focuses on the transformations and organizational
stresses that accompany firm growth.
Chapter 11 concludes the book. One of our main conclusions is that,
while researching into firm growth, one should seek to discover new empir-
ical regularities and offer descriptive, ‘appreciative’ theoretical explana-
tions, rather than trying to derive the state of the economy from unfounded
mathematical axioms.
2. Firm size distributions
A suitable starting point for studies into industrial structure and dynamics
is the firm size distribution, which is one of the oldest and most fundamen-
tal stylized facts about firm size and growth. In fact, it was while contem-
plating the empirical size distribution that Robert Gibrat (1931) proposed
the well-known ‘Law of Proportionate Effect’ (also known as ‘Gibrat’s
law’), which has arguably been the most influential model of firm growth.
Even today, the firm size distribution continues to receive a lot of attention
from both empirical researchers and theoretical modellers.
In this chapter we begin by reviewing empirical evidence on the firm size
distribution (section 2.1). Empirical work seems to suggest that the log-
normal or the Pareto are useful approximations to the aggregate firm size
distribution (de Wit, 2005). In section 2.2 Gibrat’s model of the lognormal
firm size distribution is presented. Section 2.3 presents some preliminary
evidence on the age distribution of firms, and section 2.4 combines a Gibrat
process within cohorts with the distribution of firm age to derive a Pareto
firm size distribution.
2.1 SIZE DISTRIBUTIONS
The observation that the firm size distribution is positively skewed proved
to be a useful point of entry for research into the structure of industries. (See
Figures 2.1 and 2.2 for some examples of aggregate firm size distributions.)
Gibrat (1931) considered the size of French firms in terms of employees
and concluded that the lognormal distribution was a valid heuristic. Hart
and Prais (1956) presented further evidence on the size distribution, using
data on quoted UK firms, and also concluded in favour of a lognormal
model. The lognormal distribution, however, can be viewed as just one of
several candidate skew distributions. Although Simon and Bonini (1958)
maintained that the ‘lognormal generally fits quite well’ (p. 611), they
preferred to consider the lognormal distribution as a special case in the
wider family of ‘Yule’ distributions. The advantage of the Yule family of
distributions was that the phenomenon of arrival of new firms could be
incorporated into the model. Steindl (1965) applied Austrian data to his
14
Firm size distributions 15
1
1998
2000
2002
0.1
0.01
Pr
0.001
le-04
–4 –2 0 2 4 6
S
Source: Bottazzi et al. (2008a).
Figure 2.1 Kernel estimates of the density of firm size (total sales) in
1998, 2000 and 2002, for French manufacturing firms with
more than 20 employees
10–1
10–4
Frequency
10–7
10–10
10–13
1 10 102 103 104 105 106
Firm size (employees)
Source: Axtell (2001).
Figure 2.2 Probability density function of the sizes of US manufacturing

firms in 1997
analysis of the firm size distribution, and preferred the Pareto distribution
to the lognormal on account of its superior performance in describing the
upper tail of the distribution. Similarly, Ijiri and Simon (1964, 1971, 1974)
apply the Pareto distribution to analyse the size distribution of large US
firms.
Efforts have been made to discriminate between the various candidate
skew distributions. One problem with the Pareto distribution is that the
empirical density has many more middle-sized firms and fewer very large
firms than would be theoretically predicted (Vining, 1976). This leads to
a shape for the firm size distribution that is concave to the origin when
plotted on log–log axes. Other research on the lognormal distribution
has shown that the upper tail of the empirical size distribution of firms is
too thin relative to the lognormal (Stanley et al., 1995), which implies, a
fortiori, that the tails are too thin relative to the Pareto (evidence on this
can be found in Rossi-Hansberg and Wright (2007); see also Ramsden
and Kiss-Haypal (2000) for international comparisons of the upper tail
of the firm size distribution). Marsili (2005) reports that the Pareto
is a good fit for the upper tail of the aggregate firm size distribution,
whereas the lognormal seems to be a better fit for the smaller firms.
Quandt (1966) compares the performance of the lognormal and three
versions of the Pareto distribution, using data disaggregated according
to industry. He reports the superiority of the lognormal over the three
types of Pareto distribution, although each of the distributions produces
a best-fit for at least one sample. Furthermore, it may be that some
industries (for example the footwear industry) are not fitted well by any
distribution.
More generally, Quandt’s results on disaggregated data lead us to
suspect that the regularities of the firm size distribution observed at the
aggregate level do not hold with sectoral disaggregation. Silberman (1967)
also finds significant departures from lognormality in his analysis of 90
four-digit SIC sectors in his analysis of US firms. Similarly, Marsili (2005)
observes significant differences across sectors in the firm size distribution.
It has been suggested that, while the firm size distribution has a smooth
regular shape at the aggregate level, this may merely be due to a statis-
tical aggregation effect rather than a phenomenon bearing any deeper
economic meaning (Dosi et al., 1995; Dosi, 2007). Empirical results lend
support to these conjectures by showing that the regular unimodal firm
size distributions observed at the aggregate level can be decomposed
into much ‘messier’ distributions at the industry level, some of which are
visibly multimodal (Bottazzi and Secchi, 2003a; Bottazzi et al., 2008a).
For example, Bottazzi and Secchi (2005) present evidence of significant
bimodality in the firm size distribution of the worldwide pharmaceutical
industry, and relate this to a cleavage between the industry leaders and
fringe competitors.
Other work on the firm size distribution has focused on the evolution
of the shape of the distribution over time. It would appear that the initial
size distribution for new firms is particularly right-skewed, although the
log-size distribution tends to become more symmetric as time goes by.
This is consistent with observations that small young firms grow faster
than their larger counterparts. As a result, it has been suggested that the
lognormal can be seen as a kind of ‘limit distribution’ to which a given
cohort of firms will eventually converge. Lotti and Santarelli (2001)
present support for this hypothesis by tracking cohorts of new firms
in several sectors of Italian manufacturing. Cabral and Mata (2003)
find similar results in their analysis of cohorts of new Portuguese firms.
However, Cabral and Mata interpret their results by referring to finan-
cial constraints that restrict the scale of operations for new firms, but
become less binding over time, thus allowing these small firms to grow
relatively rapidly and reach their preferred size. The empirical analysis
presented in Angelini and Generale (2008) is similar to that of Cabral and
Mata (2003), in the sense that they observe that the firm size distribu-
tion for an entering cohort is extremely skewed, but that the distribution
becomes less skewed over time and approaches the lognormal. Angelini
and Generale (2008) then examine the financial constraints hypothesis
put forward by Cabral and Mata (2003), but they observe that financial
constraints play only a minor role in explaining this evolution, especially
in developed countries.
Although the skewed nature of the aggregate firm size distribution is a
robust finding, there may be some other features of this distribution that
are specific to countries. Table 2.1, taken from Bartelsman et al. (2005),
highlights some differences in the structure of industries across countries.
For instance, one observes that large firms account for a considerable
share of French industry, whereas in Italy firms tend to be much smaller
on average. (These international differences cannot simply be attributed to
differences in sectoral specialization across countries.)
2.2 GIBRAT’S LOGNORMAL FIRM SIZE

DISTRIBUTION
Robert Gibrat’s (1931) theory of a ‘law of proportionate effect’ was hatched

when he observed that the distribution of French manufacturing establish-
ments followed a skew distribution that resembled the lognormal. Gibrat
considered the emergence of the firm size distribution as an outcome or
Table 2.1 The importance of small firms (i.e. firms with fewer than 20 employees) across broad sectors and countries,
1989–94
Absolute number (%) Share of employment (%) Ave. no. employees per firm
Total Manufac- Business Total Manufac- Business Total Manufac- Business
economy turing services economy turing services economy turing services
US 86.7 69.9 87.9 16.6 5.8 20.6 26.4 80.3 21.4
Western Germany 87.9 77.9 90.2 23.6 11.3 33.8 17.0 39.1 11.5
France 78.6 73.6 78.8 13.9 17.0 12.1 33.5 32.1 35.7
18
Italy 93.1 87.5 96.5 34.4 30.3 46.3 10.5 15.3 6.8
UK – 74.9 – – 8.3 – – 40.7 –
Canada – – – – – – 12.7 40.5 12.0
Denmark 90.0 74.0 90.8 30.2 16.1 33.4 13.3 30.4 12.7
Finland 92.6 84.8 94.5 25.8 13.0 33.0 13.0 27.8 9.9
Netherlands 95.8 86.7 96.8 31.2 16.9 41.9 6.5 18.3 5.3
Portugal 86.3 70.5 92.8 27.7 15.7 39.8 16.8 31.0 11.4
Note: The columns labelled ‘share of employment’ refer to the employment share of firms with fewer than 20 employees.
Source: Based on Bartelsman et al. (2005), Tables 2 and 3.

explanandum, and wanted to see which underlying growth process could

be responsible for generating it.
In its simplest form, Gibrat’s law maintains that the expected growth
rate of a given firm is independent of its size at the beginning of the period
examined. Alternatively, as Mansfield (1962) puts it (p. 1030), ‘the prob-
ability of a given proportionate change in size during a specified period is
the same for all firms in a given industry – regardless of their size at the
beginning of the period.’
More formally, we can explain the growth of firms in the following
framework. Let xt be the size of a firm at time t, and let et be a random
variable representing an idiosyncratic, multiplicative growth shock over
the period t 2 1 to t. We have
xt 2 xt−1 5 etxt−1 (2.1)
which can be developed to obtain
xt 5 (1 1 et)xt−1 5 x0(1 1 e1)(1 1 e2) . . . (1 1 et) (2.2)
It is then possible to take logarithms in order to approximate log(1 1 et)

by et to obtain1
t
log (xt) < log (x0) 1 e1 1 e2 1 . . . 1 et 5 log (x0) 1 a es (2.3)
s51
In the limit, as t becomes large, the log(x0) term will become insignificant,
and we obtain
t
log (xt) < a es (2.4)
s51
In this way, a firm’s size at time t can be explained purely in terms of its
idiosyncratic history of multiplicative shocks. If we further assume that all
firms in an industry are independent realizations of independent and iden-
tically distributed (i.i.d.) growth shocks, then this stochastic process leads
to the emergence of a lognormal firm size distribution, that is:
1 (lnxt 2e# t) 2
b
P (xt) 5 e2a 2s2t
(2.5)
xt"2ps2
There are of course several serious limitations to such a simple vision of
industrial dynamics. We have already seen that the distribution of growth
rates is not normally distributed, but instead resembles the Laplace or
‘symmetric exponential’. Furthermore, contrary to results implied by
Gibrat’s model, it is not reasonable to suppose that the variance of firm
size tends to infinity (Kalecki, 1945).2 In addition, we do not observe the

secular and unlimited increase in industrial concentration that would
be predicted by Gibrat’s law (Caves, 1998). Whilst a ‘weak’ version of
Gibrat’s law merely supposes that expected growth rate is independent of
firm size, stronger versions of Gibrat’s law imply a range of other issues.
For example, Chesher (1979) rejects Gibrat’s law due to the existence of
an autocorrelation structure in the growth shocks. Bottazzi and Secchi
(2006b) reject Gibrat’s law on the basis of a negative relationship between
growth rate variance and firm size. Reichstein and Jensen (2005) reject
Gibrat’s law after observing that the annual growth rate distribution is not
normally distributed.
Another objection to Gibrat’s growth model is that the resultant lognor-
mal distribution has no steady state – the variance increases to infinite over
time and the parameters of the distribution change continuously (de Wit,
2005). However, there are several ‘stability devices’ that can lend stability
to a suitably modified Gibrat model (de Wit, 2005). For example, one could
introduce a negative dependence of growth rate on size, one could allow for
a constant stream of small new entrants at the minimum firm size, or one
could put a limit on firm size below which firms cannot decline.
2.3 AGE DISTRIBUTION
The age distribution of a population of firms is also worth mentioning.

It is of direct interest in theoretical models of the firm size distribution
(see section 2.4) and it also provides indirect information on a number of
phenomena such as entry rates, survival rates, and possibly even the age of
technology used in production.
Although a number of researchers have been interested in the influ-
ence of firm age on growth (see the survey in section 7.1, Chapter 7),
very few have focused explicitly on the empirical age distribution. To
the best of our knowledge, however, the only representations of the
age distribution of firms can be found in Coad and Tamvada (2008)
and Segarra et al. (2008). Coad and Tamvada (2008) focus on census
data covering around 700 000 small-scale firms in India,3 while Segarra
et al. (2008) focus on a sample of about 85 000 firms taken from the
Spanish Mercantile Register. Figures 2.3 and 2.4 show the correspond-
ing age distributions. The shape of the distribution appears to follow an
approximately straight line with negative slope over most of the support.
Given that the y-axis is expressed in logarithms, this straight line suggests
that an exponential distribution would be a valid approximation of the
empirical age distribution.
0.1
0.01
Pr
0.001
le-04
le-05
–20 0 20 40 60 80 100 120
Age
Source: Taken from Coad and Tamvada (2008).
Figure 2.3 Kernel density of the age distribution of Indian small-scale

industries. Kernel density computed for equispaced points using
an Epanenchnikov kernel
0.1
0.01
0.001
Pr
le-04
le-05
le-06
0 20 40 60 80 100 120
Age
Source: Author’s elaboration of the data in Segarra et al. (2008), page 104.
Figure 2.4 Kernel density of the age distribution of Spanish firms.

Kernel density computed for equispaced points using an
Epanenchnikov kernel
The dearth of studies into the age distribution of firms could be because
data on age is difficult to obtain, or alternatively because data on age is not
entirely reliable due to the subjective nature of the reporting of age (Phillips
and Kirchhoff, 1989). In any case, more work on the age distribution is
clearly warranted. We anticipate that future work will show more of an
interest in the age distribution, and will also be able to clarify the nature of
the firm age distribution.
2.4 EXPLAINING THE PARETO FIRM SIZE

DISTRIBUTION
Early models of industrial dynamics focused on firms above a certain size

threshold, because data on large firms was easier to obtain. These studies
generally observed a lognormal size distribution, which can be explained
by a Gibrat process. More recently, however, work that takes young, small
firms into account has increasingly emphasized the Pareto distribution as
a suitable approximation for the empirical distribution of firm size (Axtell,
2001; de Wit, 2005; Luttmer, 2007).
When compared to the lognormal, the Pareto distribution has more
weight at the lower tail (corresponding to a larger number of very small
firms), and the upper tail of the distribution decays less rapidly than the
lognormal. In practice, however, under certain conditions, and over a
large range of their support, the lognormal and the Pareto are quite similar
(Mitzenmacher, 2003; de Wit, 2005).
The explanation suggested in the influential article by Axtell (2001)
consists of applying a Kesten process (Kesten, 1973) in which firm sizes
are bounded from below. The combination of a Gibrat random growth
model and lower bounds on firm sizes produces the Zipf distribution
(which is a special case of a Pareto distribution, when the parameter is
equal to unity). A drawback of this mechanism, however, is that firms
are implicitly assumed to be all of the same age, which is of course quite
unrealistic. In the following model, however, we begin with a Gibrat-
type process, but we relax the restriction that firms are all of the same
age. Instead, guided by the results from the previous section, we posit
an exponential distribution of firm age. Mixing these two distributions
(Gibrat process for incumbents and an exponential distribution of firm
age) yields the observed Pareto distribution (Huberman and Adamic,
1999; Reed, 2001).
Let us consider again Gibrat’s model of random growth shocks:
xt 2 xt−1 5 etxt−1 (2.6)

which can be developed to obtain a lognormally distributed firm size

distribution:
1 (lnxt 2e# t) 2
b
P (xt) 5 e2a
2s2t
(2.7)
xt"2ps2
In this section, we will no longer assume that t has the same value for all
firms. Instead, we suggest that t is itself a random variable. In the light of
the previous discussion (in section 2.3) it seems reasonable to assume the
distribution of firm age to be exponentially distributed. If t is exponentially
distributed, we have:
P(t) 5 lelt (2.8)
In order to obtain the mixture of these two distributions, we apply the

following rule: if the distribution of a variable a, p(a, b), depends on a
parameter b which in turn is distributed according to its own distribution
r(b), then the distribution of a is given by p(a) 5 ∫r(b) · p(a, b)db (Adamic
and Huberman, 1999).
This gives us the following:
1
P (xt) 5 3 lelt #
(lnxt 2e# t) 2
b
e2a 2s2t
dt (2.9)
xt"2ps 2
and, as in Adamic and Huberman (1999), this can be developed to yield:
P (xt) 5 C # x2b
t (2.10)
where C is a constant and is given by C 5 l/s ( ! (e/s) 2 1 2l) . The

∞
exponent b is in the range [1, ] and is determined by b 5 1 2 (e/s2) 1
( ! (e2 1 2ls2) /s2) . When the mean growth rate is close to 0 per cent, e
will be close to 1. As a result, if l is small (implying that the exponential
decay is relatively weak, i.e. that it is not uncommon to find firms with an
age much greater than one),4 and if s is small (which is not implausible
either), then the exponent b will be close to Zipf’s value of 1, which has
been observed in empirical work (Axtell, 2001).
2.5 CONCLUSION
We began this chapter by looking at empirical work into the firm size dis-
tribution (section 2.1). Two distributions, in particular, stand out as can-
didates for approximations to the empirical distribution – the lognormal
or the Pareto distribution. We then turned to theoretical models of firm
growth in an attempt to explain how such size distributions could emerge.

In section 2.2 we saw how a Gibrat process leads to a lognormal firm
size distribution. A drawback of Gibrat’s law, however, is that firms are
assumed to be all of the same age. Under closer examination, we observed
that the age of firms tends to follow an exponential distribution (section
2.3). Combining an exponential distribution of firm age with a Gibrat-type
proportional growth model leads to the emergence of a Pareto firm size
distribution (section 2.4).
3. Growth rate distributions
The previous chapter presented one of the oldest and best-known empirical
regularities in industrial organization – the skewed firm size distribution. In
this chapter, we focus on the growth rate distribution. Regularities in the
distribution of firm growth rates were discovered only recently, a little over
ten years ago, but the characteristic ‘tent-shaped’ distribution has been
observed in a wide variety of databases and appears to be a remarkably
robust feature of firm growth and industrial dynamics.
We begin this chapter by surveying the empirical literature investigating
the distribution of firm growth rates (section 3.1) before moving on to some
theoretical models that attempt to explain the emergence of the observed
distribution (section 3.2).
3.1 GROWTH RATE DISTRIBUTIONS
It has long been suspected that the distribution of firm growth rates is
fat-tailed. In an early contribution, Ashton (1926) considers the growth
patterns of British textile firms and observes that ‘In their growth they
obey no one law. A few apparently undergo a steady expansion . . . With
others, increase in size takes place by a sudden leap’ (Ashton, 1926, pp.
572–3). Little (1962) investigates the distribution of growth rates, and
also finds that the distribution is fat-tailed. Similarly, Geroski and Gugler
(2004) compare the distribution of growth rates to the normal case and
comment on the fat-tailed nature of the empirical density. Recent empiri-
cal research, from an ‘econophysics’ background, has discovered that the
distribution of firm growth rates closely follows the parametric form of
the Laplace density. Using the Compustat database of US manufacturing
firms, Stanley et al. (1996) observe a ‘tent-shaped’ distribution on log-log
plots that corresponds to the symmetric exponential, or Laplace distribu-
tion (see also Amaral et al., 1997 and Lee et al., 1998). The quality of the
fit of the empirical distribution to the Laplace density is quite remarkable.
The Laplace distribution is also found to be a rather useful representation
when considering growth rates of firms in the worldwide pharmaceutical
industry (Bottazzi et al., 2001).
25
The functional form of the Laplace (symmetric exponential) density is:

1 20 x2m 0/a
fL (x) 5 e (3.1)
2a
where μ is the location parameter and b . 0 is the scale parameter.
Giulio Bottazzi and co-authors extend these findings by considering
the Laplace density in the wider context of asymmetric exponential power
distributions – also known as the Subbotin family of distributions (intro-
duced to the firm growth literature in Bottazzi et al., 2002). The Subbotin
distribution can be formally presented by the following equation:
1 1 x2m b
fS (x) 5 1/b
e2 b 0 a 0 (3.2)
(
2ab G 1/b 1 1)
where G(x) is the Gamma function. The distribution has three parameters
– the mean μ, the dispersion parameter a and the shape parameter b. As
the shape parameter b decreases, the tails of the density become fatter. The
density is leptokurtic for b , 2, and platykurtic for b . 2. Two noteworthy
special cases of the Subbotin distribution are the Gaussian distribution (for
which b 5 2) and the Laplace distribution (with b 5 1).
Bottazzi and Secchi find that, for the Compustat database, the Laplace
is indeed a suitable distribution for modelling firm growth rates, at both
aggregate and disaggregated levels of analysis (Bottazzi and Secchi,
2003a). The exponential nature of the distribution of growth rates also
holds for other databases, such as Italian manufacturing (Bottazzi et al.,
2007). The growth rates of French manufacturing firms have also been
studied, and roughly speaking a similar shape was observed, although it
must be said that the empirical density was noticeably fatter-tailed than the
Laplace (see Bottazzi et al., 2008a).1 Research into Danish manufacturing
firms presents further evidence that the growth rate distribution is heavy-
tailed, although it is suggested that the distribution for individual sectors
may not be symmetric but right-skewed (Reichstein and Jensen, 2005).
Generally speaking, however, it would appear that the shape of the growth
rate distribution is more robust to disaggregation than the shape of the firm
size distribution. In other words, whilst the smooth shape of the aggregate
firm size distribution may be little more than a statistical aggregation
effect, the ‘tent-shapes’ observed for the aggregate growth rate distribution
are usually still visible even at disaggregated levels (Bottazzi and Secchi,
2003a; Bottazzi et al., 2008a). This means that extreme growth events can
be expected to occur relatively frequently, and make a disproportionately
large contribution to the evolution of industries.
Figures 3.1 and 3.2 show plots of the distribution of sales and employ-
ment growth rates for French manufacturing firms with over 20 employees.
Growth rate distributions 27
1998
2000
2002
1
Prob
0.1
0.01
0.001
–3 –2 –1 0 1 2
Conditional growth rate
Source: Bottazzi et al. (2005).
Figure 3.1 Distribution of sales growth rates of French manufacturing firms
1998
2000
2002
1
Prob
0.1
0.01
0.001
–2 –1.5 –1 –0.5 0 0.5 1 1.5 2
Conditional growth rate
Source: Coad (2006).
Figure 3.2 Distribution of employment growth rates of French

manufacturing firms
Note the appearance of a ‘tent shape’ on a plot of the distribution of (log)

growth rates2 when the y-axis is expressed in logs. If the slopes are straight
lines, we have the Laplace density.
The heavy-tailed Laplace distribution of growth rates appears to hold
across a variety of firm growth indicators, not only for sales growth and
employment growth (as in Figures 3.1 and 3.2) but also growth of value-
added (Bottazzi et al., 2007). Other researchers have found evidence of con-
siderable lumpiness in the dynamics of plant-level investment (Doms and
Dunne, 1998; Cooper et al., 1999). Plant-level investment appears to occur
in large investment episodes that are known as ‘investment spikes’. Doms
and Dunne (1998) analyse a large sample of US manufacturing plants in
1972–1988 and observe that, on average, half of a plant’s total investment
over this period was performed in just three years. They also observe that,
while 52.9 per cent of plants increase their capital stock by less than 2.5 per
cent in a year, 11 per cent of plants increase their capital stock by more than
20 per cent. These authors do not undertake any parametric estimation of
the empirical distribution of investment in capital, but it may well be that
the distribution is even more skewed than the growth rate distributions for
other series such as employment or sales.
Although the Laplace density provides a good representation of the
growth rates distribution, some refinements should be mentioned. First,
it appears that the Laplacian nature of the distribution tends to fade over
time, such that the distribution becomes less heavy-tailed and approaches
the normal when growth is measured over periods of time longer than one
year (Bottazzi and Secchi, 2006a; Buldyrev et al., 2007). Second, it has been
suggested that the Laplace distribution is a better fit for larger, multiprod-
uct firms while the growth rate distribution of smaller firms has even fatter
tails that appear to be Pareto-distributed (Fu et al., 2005).
Time-varying moments of the growth rate distribution

Research suggests that both the size distribution and the growth rate distri-
bution are relatively stable over time, although it should be noted that there
is great persistence in firm size but much less persistence in growth rates on
average (more on growth rate persistence is presented in section 4.4). As a
result, it is of interest to investigate how the moments of the growth rates
distribution change over the business cycle. Indeed, several studies have
focused on these issues and some preliminary results can be mentioned
here. It has been suggested that the variance of growth rates changes over
time for the employment growth of large US firms (Hall, 1987) and that
this variance is procyclical in the case of growth of assets (Geroski et al.
(2003)). This is consistent with the hypothesis that firms have a lot of discre-
tion in their growth rates of assets during booms but face stricter discipline
during recessions. Higson et al. (2002) and Higson et al. (2004) consider
the evolution of the first four moments of distributions of the growth of
sales, for large US and UK firms over periods of 30 years or more. They
observe that higher moments of the distribution of sales growth rates have
significant cyclical patterns. In particular, evidence from both US and UK
firms suggests that the variance and skewness are countercyclical, whereas
the kurtosis is procyclical. Higson et al. (2002) explain the countercyclical
movements in skewness in these words (p. 1551):
The central mass of the growth rate distribution responds more strongly to
the aggregate shock than the tails. So a negative shock moves the central mass
closer to the left of the distribution leaving the right tail behind and generates
positive skewness. A positive shock shifts the central mass to the right, closer to
the group of rapidly growing firms and away from the group of declining firms.
So negative skewness results.
The procyclical nature of kurtosis (despite their puzzling finding of

countercyclical variance) emphasizes that economic downturns change the
shape of the growth rate distribution by reducing a key parameter of the
‘spread’ or ‘variation’ between firms.
3.2 LUMPS, BUMPS AND GROWTH SPURTS
There is something of a tradition in Industrial Organization (IO) modelling

to represent growth processes in purely stochastic terms. Gibrat’s law is a
well-known example (Gibrat, 1931). According to Gibrat’s law, the aggre-
gate size distribution is explained by referring to a growth process in which
the growth rates of firms are purely random variables. Relatedly, Ijiri and
Simon (1977) offered an explanation of the skewed firm size distribution
in terms of a random process in which the probability of a firm taking
up an additional business opportunity is conditional upon its size. The
Ijiri–Simon model, dubbed the ‘island’ model because of the independent
arrival of the growth opportunities, has been widely accepted, and interest
in it was recently revived by Sutton (1998).
Although there is a need for theoretical models that can explain non-
Gaussian phenomena (McKelvey and Andriani, 2005), few explanations
of the heavy-tailed growth rates distribution have been proposed in the
literature to date – no doubt because the exponential nature of the firm
growth rates distribution is a ‘stylized fact’ that has been discovered only
recently. Some of the models that have focused on the exponential distri-
bution of firm growth rates have also taken the approach of stochastic
explanations. Amaral et al. (1997) develop a model in which the emergence
of the distribution rests on a particular specification of the functional form

of the stochastic growth process. However, there is little justification of
the choice of such a functional form, and so it could be argued that their
model is more of a tautology than an explanation. The model in Bottazzi
and Secchi (2006a) also conceives of firm growth as a random process,
and refers to the concept of competition between firms and the struggle
for business opportunities to explain the distribution of growth rates. The
model presented in section 3.2.2, however, attempts to explain the emer-
gence of the empirically-observed growth rates distribution by referring to
common features of business organizations – that firms are composed of
resources that are indivisible in nature and subject to interactions.
3.2.1 Interfirm Competition and Increasing Returns to Growth
Giulio Bottazzi and Angelo Secchi have played a major role in bringing
attention to the empirical growth rate distribution, and their theoretical
model is also an important contribution to the literature.
The Bottazzi and Secchi (2006a) model3 is in line with previous ‘island
models’ of industry evolution in the sense that it conceives of firm growth
as a random process – ‘in our model luck is the principal factor that
finally distinguishes winners from losers among the contenders’ (Bottazzi
and Secchi, 2006a, p. 236). The main point of departure from the ‘island’
models, however, is that their model emphasizes the role of competition
between organizations in shaping the growth rate distribution. While
previous models tended to treat firms as independent entities, competi-
tive effects between firms are prominent in the Bottazzi and Secchi model.
According to this model, firms compete in a given industry for a finite
number of pre-existing business opportunities, that, once obtained, can
be translated into growth. Of central importance is the positive feedback
mechanism which assigns the growth opportunities to the different firms.
This assignment mechanism postulates the existence of dynamic increas-
ing returns to growth (because of economies of scale, economies of scope,
network externalities, knowledge accumulation, and so on) and posits that
market success is cumulative or self-reinforcing. Firms that have already
received a growth opportunity are more likely to obtain another one. As a
result, many opportunities tend to concentrate in a few firms. Bottazzi and
Secchi show that, under certain conditions, the growth rate distribution
approaches the empirically-observed Laplace distribution.
It should be noted, however, that the positive feedback mechanism is
assumed to operate in the short term only (that is for periods of up to one
year). This short duration of the increasing returns mechanism is, effec-
tively, required in order to reconcile the model with empirical work into the
correlation of year-on-year firm growth rates, which displays little evidence

of increasing returns to growth over periods longer than one year. (More
on growth rate autocorrelation can be found in section 4.4).
3.2.2 Firms as Bundles of Discrete Interdependent Resources
The choice of stochastic models to describe industrial evolution bears

witness to a reluctance to generalize across firms. Firms grow for a wide
variety of different reasons, they are indeed heterogeneous, and it is
believed that the best or only way to model growth may be by treating it as
purely stochastic. To move beyond describing industry dynamics in terms
of purely random shocks, we need to address the following question: ‘Can
we generalize across firms?’ Our answer is: ‘Yes we can, to some degree’.
Without denying the complexity of commercial organizations or the het-
erogeneity that exists between firms from different sectors of the economy,
we maintain that there are some general features that are present in firms.
(Indeed, Simon (1962) suggests that there are some broad features that
appear to be common not only to all firms but to all complex systems!)
The theoretical explanation proposed here is rooted in the ‘resource-based
approach’, which views firms as being composed of discrete, complemen-
tary resources (Penrose, 1959). In addition, we allow for the possibility
of growth being accommodated by organizational slack. Organizational
slack is a widely-recognized characteristic of business firms – indeed, a
firm’s resources will not be fully utilized at any given time for a number of
reasons.4 However, managers will seek to use a firm’s resources efficiently,
to have them as close as possible to ‘full utilization’. If a firm’s resources are
under-utilized, then growth can feed off these slack resources.5 On the other
hand, if resources are already more or less fully employed, then growth
will only be possible with the addition of new resources. In the former
case, growth requires no additional investment, whilst in the latter case,
firm growth will be accompanied by potentially wide-scale investment.6
This depiction of firm growth can be expressed in terms of self-organizing
criticality. The firm can be seen as a system which tends to a ‘critical state’
of full utilization of its resources, as managers strive to organize the firm’s
resources efficiently within the firm’s hierarchical framework. Depending
upon the criticality of the system, the addition of an activity during growth
will result in a (marginally) increased strain for many associated resources,
thus potentially triggering off a chain reaction of subsequent growth across
the whole of the organization. In this vein, Dixon comments on the criti-
cality of a firm at a more general level: ‘the later addition of one person to
regular activities can bring into operation a chain of reactions in the form
of salaried employee increases, salary increases, and fixed asset additions’
(Dixon, 1953, p. 50). Similarly, Hannan writes: ‘changes in one organi-

zational feature often generate cascades of additional changes, because
of the interdependence among parts of an organization’ (Hannan, 2005,
p. 61). Weick and Quinn put it this way: ‘Small changes can be decisive
if they occur on the edge of chaos . . . in interconnected systems, there is
no such thing as marginal change’ (Weick and Quinn, 1999, p. 378). The
‘avalanche’ will only stop if there is sufficient slack capacity to absorb the
extra workload associated with the additional resources.
To illustrate this idea, Coad (2008c) proposes a model which is capable
of generating the symmetric exponential (that is, Laplace) growth rate
distributions within the time series of a single firm. In this model, firms
are seen as being composed of resources, which are indivisible in nature.
Firms grow by adding discrete resources to a complex of interdependent
resources that they already possess. The indivisible resources that form the
basis for a firm’s productive potential are not perfect substitutes, but they
need to be combined in certain proportions in order for the firm to use
these resources to produce its output. As a consequence, firms strive to find
those combinations of resources that reduce slack. Penrose explains that
‘[u]nused productive services are, for the enterprising firm, at the same time
a challenge to innovate [and] an incentive to expand . . .’ (Penrose, 1959,
p. 85). In other words, the resources in a firm are interdependent because,
under circumstances where firms strive for the most efficient combination
of indivisble resources, the addition of one indivisible resource may well
have consequences on the desirable levels of other resources. In other
words, the resources in a firm are interdependent and subject to local inter-
actions. This may lead to non-linearities in the growth process as firms add
indivisible resources to arrive at an efficient level of production.
A simplified model
This model considers the special case of the propagation of employment
growth throughout the various levels of a firm’s hierarchy. The organiza-
tion of production in a hierarchy is indeed a general feature of all firms – in
fact, in the Transaction–Cost–Economics literature, the words ‘firm’ and
‘hierarchy’ are used almost interchangeably. The firm is characterized as
being composed of a relatively large number of hierarchies.7 The bottom
layer of the firm (that is, the very lowest hierarchical level) is composed
exclusively of productive workers, whilst all of the other levels are com-
posed of managers whose task is to supervise either productive workers or
subordinate managers (see Figure 3.3 for an illustration). A firm grows by
adding a productive worker. The number of managers is determined by the
number of productive workers and also by limits on the efficient span of
control, a, which correspond to the maximum number of subordinates that
Supervisors Supervisors
Productive Productive
workers workers
Figure 3.3 An illustration of the underlying intuition of the model, where

the span of control is a53. Depending upon the ‘criticality’ of
the system, the addition of a productive worker may lead to an
increase in the number of supervisors further up the hierarchy.
If there is some slack in the system, a productive worker can
be added and new supervisors need not be added (see left).
If, however, the attention of supervisors is already at full
utilization, the addition of a productive worker will require the
addition of a supervisor (see right).
a manager can effectively supervise. ‘At executive levels [the span of control]
is seldom less than three, and seldom more than ten, and usually lies within
narrower bounds – particularly if we take averages over all executives in an
organization at a given level.’ (Simon, 1957, p. 32). In this model, though,
we do not need to attribute any specific numerical value to a and so we leave
it in algebraic form. It is computationally helpful, and also theoretically
meaningful, however, to assume that a is a whole number that is strictly
greater than unity (that is, a [ N1, a . 1). For analytical simplicity, we
assume that a is a constant and does not vary either within a hierarchical
level or across levels (for a discussion of the plausibility of this assumption,
see Williamson, 1967, p. 128). For the purposes of this model, we also must
assume that adjustment of the firm’s hierarchical organization to additional
productive workers occurs within one time period. Finally, we assume that
the firm is initially at a stable state, such that it is already efficiently organ-
ized in the sense that it is not possible for it to employ fewer managers given
the number of productive workers and its given value of a (that is the limit
on the efficient span of control). The reader may notice major similarities
between the model developed here and the executive compensation model
of Simon (1957) and the information flows model of Williamson (1967).
The fact that the same hierarchical model has been applied in quite different
contexts lends credibility to its use here – indeed, we cannot be accused of
having conclusions that emerge from ad hoc modelling assumptions.
A summary understanding can be obtained by looking at Figure 3.3.
Two important points should be emphasized. First, there is a distinction

between total production n and total employment x. Total production
corresponds to the number of productive workers (n), while total employ-
ment corresponds to the number of both productive workers and supervi-
sors combined (x). Second, it should be noted that we do not attempt to
generalize on the sources of growth opportunities, but rather we focus
on how firms build upon given growth opportunities. We argue that the
fat-tailed distribution of growth rates does not come from the distribution
of opportunities available to firms, but rather on the reactions of firms to
growth stimuli. The model is admittedly a gross simplification and does
not take into account such factors as the interdependence of growth rates
between firms, flexibility of a (the span of control parameter), liquid-
ity constraints that limit growth, or limits on the availability of suitable
workers. Nonetheless, its simplicity will make it clear to what properties
we owe the emergence of the distribution.
Formal model Let us begin with the simplest possible case, considering
one firm that grows by adding just one productive worker (Δn 51). If new
productive workers can be integrated without having to add a supervisor,
we have Δn 5 Δx; that is the number of productive workers added is equal to
change in total employment. It is possible, however, that all of the managers
in the second hierarchical level (that is, those that supervise the productive
workers) are already fully occupied. This will occur when the number of
productive workers (before adding the new one) is exactly a multiple of a.
If this is the case, the arrival of the supplementary worker will require that
one supplementary manager be hired at the next hierarchical level. This
scenario will occur with probability 1/a. However, the arrival of this new
manager at the second level may add to the workload of managers on the
third hierarchical level, and so on. The probability that the addition of a
productive worker leads to at least two managers being hired at two succes-
sive levels is 1/a 3 1/a 5 1/a2. We can continue with this reasoning to end
up with the following distribution of employment growth:
Prob. (Δx $ 1|Δn 5 1) 5 1

Prob. (Δx $ 2|Δn 5 1) 5 1/a
Prob. (Δx $ 3|Δn 5 1) 5 1/a2
...
and so on. Formally, we have an exponential distribution with the follow-

ing functional form:
P(Δx $ |Δn 5 1) 5 a1− (3.3)

or, expressed differently,
P(Δx 5 |Δn 5 1) 5 a1− (1 − 1/a) (3.4)
where is a positive integer ( $ Δn). We therefore observe that the distri-

bution of total employment growth increments (Δx) of a firm that grows
by adding one productive worker will follow an exponential distribution.
It is trivial to show that an exponential distribution of growth increments
is equivalent to an exponential growth rates distribution.
It is also possible to generalize for the case where a firm grows by adding
Dn [ N1 productive workers (with, of course, Δx $ Δn). For Δn , a, we
obtain the following distribution:
P (Dx 5 g 0 Dn) 5 1 2 Dn/a if g 5 Dn
P (Dx 5 g 0 Dn) 5 Dn # aDn2g (1 2 1/a) if g . Dn (3.5)
where equation (3.4) corresponds to the special case where Δn 5 1. An

illustration is offered in Figure 3.4. Analogical reasoning can be applied to
the case where a firm shrinks in size.
(Log) Prob.
1-(1/)
-1[1-(1/)]
-2[1-(1/)]
-3[1-(1/)]
1 2 3 4
Growth of total employment (x)
Figure 3.4 The distribution of growth of total employment if a firm grows

by Δn 5 1
Autocorrelation dynamics It is also possible to derive the conditional

autocorrelation dynamics of the firm’s growth dynamics. Consider the case
where one production worker is added in each period t, that is, nt 5 nt21
1 1. The conditional growth autocorrelation can be written as:
P (xt 2 xt21 . 1 0 xt21 2 xt22 . 1) 5 0

1
P (xt 2 xt21 . 1 0 xt21 2 xt22 5 1) 5 (3.6)
a21
If the firm experienced a growth spurt in the previous period (that is,
xt−1 − xt−2 . 1), it has a probability of zero of repeating this growth per-
formance in the following period. If, however, the firm added a productive
worker in the previous period but this did not trigger off the addition of a
supervisor, then the addition of a productive worker in this period has a
positive probability of leading to the further addition of a supervisor.
Considering that E(xt − xt−1|xt−1 − xt−2 . 1) 5 1 and E(xt − xt−1|xt−1 − xt−2
5 1) . 1, the model generates negative growth rate autocorrelation in the
case where a growth spurt of x was triggered in the previous period (that
is, when xt−1 − xt−2 . 1).
Coad (2008c) extends the model to include not only employment, but
also capital goods and production plants. This extended model is investi-
gated through simulations, and a heavy-tailed growth rates distribution is
observed to emerge.
Discussion The model is admittedly far too simple to be realistic, yet its
simplicity makes for greater visibility of the source of the emergence of
the heavy-tailed distribution. The model can be seen as the simplest model
in a family of possible models that view firms as coherent collections of
resources that are complementary and discrete. These latter are subject to
localized interactions and embedded in an organization that tends to a crit-
ical state of full utilization of its resources. In this context, a small growth
stimulus working through local interaction channels can be transmitted
throughout a firm to produce potentially large-scale effects. We argue that
it is these properties that explain the emergence of the observed fat-tailed
growth rate distributions.
The model describes the dynamics of a single, ‘autistic’ organization and
makes no attempt to account for competitive interactions between firms.
In our view, this is not a serious flaw. Other explanations of the fat-tailed
growth rate distribution have emphasized the complex nature of inter-firm
competition as the source of the emergence of the observed distribution
(for example Bottazzi and Secchi, 2006a; McKelvey and Andriani, 2005).
Recent empirical work has nonetheless cast doubt on the importance of
inter-firm competition. Sutton (2007) analyses the dynamics of market

shares of the largest and second largest firms in a number of Japanese
industries, and finds (perhaps surprisingly) that their market share dynam-
ics can be modelled as statistically independent. Only in the case where the
combined market share of an industry’s two largest firms is at least 90 per
cent of the industry total does inter-firm competition leave a detectable
statistical footprint. Geroski and Gugler (2004) consider the impact of the
growth of rival firms on a firm’s employment growth, using a database
on several thousand of the largest firms in 14 European countries. Rival
firms are defined as other firms in the same 3-digit industry. In their main
regression results (their Table 2) they are unable to detect any significant
effect of rival’s growth on firm growth, although they do find a significant
negative effect in specific industries (that is, differentiated good industries
and advertising-intensive industries). In our model, it is the complex
nature of interactions between the resources within a firm (rather than the
competitive struggle between firms) that accounts for the emergence of the
observed growth rates distribution.
3.3 CONCLUSION
The Laplace, or symmetric exponential distribution of firm growth rates is

a robust stylized fact that has been discovered but very recently. This dis-
covery constitutes a fascinating opportunity to improve our understanding
of the firm growth process. Several theoretical models have already been
constructed in order to explain this regularity, according to which the
heavy-tailed growth rates distribution is associated with inter-firm compe-
tition or, alternatively, because firms are organized as bundles of indivisible
and interacting resources. Further explanations and models will no doubt
be put forward in the near future.
The heavy-tailed nature of firm growth is also a challenge to empirical
work. Econometric techniques need to take the growth rates distribution
into account, by renouncing Gaussian estimators (such as OLS) in favour
of more robust estimators (such as median regressions). Empirical work
might also benefit by seeking out the characteristics of the few high-growth
firms that grow particularly rapidly and thus make a disproportionate con-
tribution to industrial development. Quantile regression is a useful econo-
metric tool for such work (Coad and Rao, 2008). Another approach would
be to treat high growth events as discrete events and to try to predict the
probability of such a high-growth event occurring. Such an approach has
been used by Whited (2006), for example, who investigates the probability
of an investment spike taking place.
Investment is indeed a lumpy process. The data suggests that much of

a firm’s investment activity is concentrated in a very short time period.
How do these investment spikes relate to other aspects of firm growth? Are
investment spikes followed by sudden changes in employment, or perhaps
productivity leaps? Salter (1960) develops a theory according to which
investment in recent capital vintages is associated with increases in labour
productivity. Empirical evidence in Power (1998), however, fails to detect
the expected relationship between investment and productivity growth in
her sample of US manufacturing plants. Further work relating the lumpy
nature of investment to other aspects of firm growth would, we speculate,
be extremely valuable.
4. Gibrat’s law
Gibrat’s law continues to receive a huge amount of attention in the empiri-
cal industrial organization literature, more than 75 years after the seminal
publication of Gibrat (1931).
We begin by presenting the ‘Law’, and then review some of the related
empirical literature. We do not attempt to provide an exhaustive survey
of the literature on Gibrat’s law, because the number of relevant studies is
indeed very large. (For other reviews of empirical tests of Gibrat’s law, the
reader is referred to the survey by Lotti et al., 2003); for a survey of how
Gibrat’s law holds for the services sector see Audretsch et al., 2004.)
Instead, we try to provide an overview of the essential results. We inves-
tigate how expected growth rates and growth rate variance are influenced
by firm size, and also investigate the possible existence of patterns of serial
correlation in firm growth.
4.1 GIBRAT’S MODEL
Let us briefly return to Gibrat’s model of firm growth presented earlier in

section 2.2. As before, we define xt to be the size of a firm at time t, and let
et be random variable representing an idiosyncratic, multiplicative growth
shock over the period t – 1 to t. We have
xt – xt−1 5 etxt−1 (4.1)
which can be developed to obtain
xt 5 (1 1 et)xt−1 5 x0(1 1 e1)(1 1 e2). . .(1 1 et) (4.2)
It is then possible to take logarithms in order to approximate log(1 1 et)

by et to obtain1
t
log (xt) < log (x0) 1 e1 1 e2 1 . . . 1 et 5 log (x0) 1 a es (4.3)
s51
In the limit, as t becomes large, the log(x0) term will become insignificant,
and we obtain
39
t
log (xt) < a es (4.4)
s51
In this way, a firm’s size at time t can be explained purely in terms of its
idiosyncratic history of multiplicative shocks. If we further assume that all
firms in an industry are independent realizations of i.i.d. normally distrib-
uted growth shocks, then this stochastic process leads to the emergence of
a lognormal firm size distribution.
4.2 FIRM SIZE AND AVERAGE GROWTH
Although Gibrat’s (1931) seminal book did not provoke much of an imme-
diate reaction, in recent decades it has spawned a flood of empirical work.
Nowadays, Gibrat’s ‘Law of Proportionate Effect’ constitutes a bench-
mark model for a broad range of investigations into industrial dynamics.
Another possible reason for the popularity of research into Gibrat’s law,
one could suggest quite cynically, is that it is a relatively easy paper to
write. First of all, it has been argued that there is a minimalistic theoretical
background behind the process (because growth is assumed to be purely
random). Then, all that needs to be done is to take the IO economist’s
‘favourite’ variable (that is, firm size, a variable which is easily observable
and readily available) and regress the difference on the lagged level. In
addition, few control variables are required beyond industry dummies and
year dummies, because growth rates are characteristically random.
Empirical investigations of Gibrat’s law rely on estimation of equations
of the type:
log(xt) 5 a 1 blog(xt−1) 1 e (4.5)
where a firm’s ‘size’ is represented by xt, a is a constant term (industry-wide

growth trend) and e is a residual error. Research into Gibrat’s law focuses
on the coefficient b. If firm growth is independent of size, then b takes the
value of unity. If b is smaller than 1, then smaller firms grow faster than
their larger counterparts, and we can speak of ‘regression to the mean’.
Conversely, if b is larger than 1, then larger firms grow relatively rapidly
and there is a tendency to concentration and monopoly.
A significant early contribution was made by Edwin Mansfield’s (1962)
study of the US steel, petroleum and rubber tyre industries. In particular
interest here is what Mansfield identified as three different renditions of
Gibrat’s law. According to the first, Gibrat-type regressions consist of
both surviving and exiting firms and attribute a growth rate of −100 per
cent to exiting firms. However, one caveat of this approach is that smaller
Gibrat’s law 41
firms have a higher exit hazard which may obfuscate the relationship
between size and growth. The second version, on the other hand, consid-
ers only those firms that survive. Research along these lines has typically
shown that smaller firms have higher expected growth rates than larger
firms. The third version considers only those large surviving firms that
are already larger than the industry Minimum Efficient Scale of produc-
tion (with exiting firms often being excluded from the analysis). Generally
speaking, empirical analysis corresponding to this third approach suggests
that growth rates are more or less independent from firm size, which lends
support to Gibrat’s law.
The early studies focused on large firms only, presumably partly due to
reasons of data availability. A series of papers analysing UK manufactur-
ing firms found a value of b greater than unity, which would indicate a
tendency for larger firms to have higher percentage growth rates (Hart,
1962; Samuels, 1965; Prais, 1974; Singh and Whittington, 1975).
However, the majority of subsequent studies using more recent datasets
have found values of b slightly lower than unity, which implies that, on
average, small firms seem to grow faster than larger firms. This result is fre-
quently labelled ‘reversion to the mean size’ or ‘mean-reversion’.2 Among
a large and growing body of research that reports a negative relationship
between size and growth, we can mention here the work by Kumar (1985)
and Dunne and Hughes (1994) for quoted UK manufacturing firms, Hall
(1987), Amirkhalkhali and Mukhopadhyay (1993) and Bottazzi and Secchi
(2003a) for quoted US manufacturing firms (see also Evans, 1987a and
1987b, for US manufacturing firms of a somewhat smaller size), Gabe
and Kraybill (2002) for establishments in Ohio, Goddard et al. (2002) for
quoted Japanese manufacturing firms, and Sleuwaegen and Goedhuys
(2002) for manufacturing firms from Côte d’Ivoire.
Studies focusing on small businesses have also found a negative rela-
tionship between firm size and expected growth – see for example Yasuda
(2005) for Japanese manufacturing firms, Segarra and Callejon (2002) and
Calvo (2006) for Spanish manufacturing, McPherson (1996) for Southern
African micro businesses, and Wagner (1992) and Almus and Nerlinger
(2000) for German manufacturing. Dunne et al. (1989) analyse plant-level
data (as opposed to firm-level data) and also observe that growth rates
decline along size classes. Research into Gibrat’s law using data for specific
sectors also finds that small firms grow relatively faster (see for example
Barron et al. (1994) for New York credit unions, Weiss (1998) for Austrian
farms, Liu et al. (1999) for Taiwanese electronics plants, and Bottazzi and
Secchi (2005) for an analysis of the worldwide pharmaceutical sector).
Indeed, there is a lot of evidence that a slight negative dependence of
growth rate on size is present at various levels of industrial aggregation.
Although most empirical investigations into Gibrat’s law consider only

the manufacturing sector, some have focused on the services sector. The
results, however, are often qualitatively similar – there appears to be a
negative relationship between size and expected growth rate for services
too (see Variyam and Kraybill, 1992; Johnson et al., 1999). Teruel-
Carrizosa (2008) observes that while small manufacturing firms tend to
experience fast growth relative to their larger counterparts, the inequal-
ity between the growth of small and large firms in the service industry is
less pronounced. This comparison of firm growth in manufacturing and
service industries is consistent with the evidence in Phillips and Kirchhoff
(1989, Table V), as well as the theoretical model in Rossi-Hansberg and
Wright (2007).
In a number of cases, a weak version of Gibrat’s law cannot be convinc-
ingly rejected, since there appears to be no significant relationship between
expected growth rate and size (see the analyses provided by Bottazzi
et al. (2008a) for French manufacturing firms, Droucopoulos (1983)
for the world’s largest firms, Hardwick and Adams (2002) for UK Life
Insurance companies, and Audretsch et al. (2004) for small-scale Dutch
services). Notwithstanding these latter studies, however, we acknowledge
that in most cases a negative relationship between firm size and growth is
observed. Indeed, it is quite common for theoretically-minded authors to
consider this to be a ‘stylised fact’ for the purposes of constructing and
validating economic models.3 Furthermore, John Sutton refers to this
negative dependence of growth on size as a ‘statistical regularity’ in his
revered survey of Gibrat’s law (Sutton, 1997, p. 46).
A number of researchers maintain that Gibrat’s law does hold for firms
above a certain size threshold. This corresponds to acceptance of Gibrat’s
law according to Mansfield’s third rendition, although ‘mean reversion’
leads us to reject Gibrat’s law as described in Mansfield’s second rendi-
tion. Mowery (1983), for example, analyses two samples of firms, one of
which contains small firms while the other contains large firms. Gibrat’s
law is seen to hold in the latter sample, whereas mean reversion is observed
in the former. Hart and Oulton (1996) consider a large sample of UK
firms and find that, whilst mean reversion is observed in the pooled data,
a decomposition of the sample according to size classes reveals essen-
tially no relation between size and growth for the larger firms. Similarly,
Bigsten and Gebreeyesus (2007) observe a negative association between
lagged size and growth in their sample of Ethiopian manufacturing firms,
although size seems to be independent of growth for the very largest firms.
Their results suggest that size and growth are inversely related until the
firm reaches a relatively large size of around 400 employees. Lotti et al.
(2003) follow a cohort of new Italian start-ups and find that, although
Gibrat’s law 43
smaller firms initially grow faster, it becomes more difficult to reject the
independence of size and growth as time passes. Similarly, results reported
by Becchetti and Trovato (2002) for Italian manufacturing firms, Geroski
and Gugler (2004) for large European firms and Cefis et al. (2007) for
the worldwide pharmaceutical industry also find that the growth of large
firms is independent of their size, although including smaller firms in the
analysis introduces a dependence of growth on size. After digging around
in the available evidence, Caves (1998) concludes his survey of industrial
dynamics with the ‘substantive conclusion’ that Gibrat’s law holds for
firms above a certain size threshold, whilst for smaller firms growth rates
decrease with size. You surveyed the literature and arrived at a similar
conclusion (You, 1995).
Concern about econometric issues has often been raised. Sample selec-
tion bias, or ‘sample attrition’, is one of the main problems, because
smaller firms have a higher probability of exit. Failure to account for the
fact that exit hazards decrease with size may lead to underestimation of
the regression coefficient (that is, b). Hall (1987) was among the first to
tackle the problem of sample selection, using a generalized Tobit model.
She concludes that selection bias does not seem to account for the nega-
tive relationship between size and growth in her data. Similar conclusions
were reached by McPherson (1996). An alternative way of correcting for
sample selection is by applying Heckman’s two-step procedure. This is the
methodology used by Harhoff et al. (1998), who also observe that selec-
tion bias has only a small influence on the Gibrat coefficient. In short, the
‘problem of sample selection does not seem to significantly affect the rela-
tionship between growth rate and size of firm’ (Marsili, 2001, p. 15). The
possibility of heteroskedasticity is also frequently mentioned, although it
can be corrected for quite easily, for example by applying White’s (1980)
procedure. In any case, heteroskedasticity does not introduce any asymp-
totic bias in the coefficient estimates. Serial correlation in growth rates
can lead to biased estimates, although Chesher (1979) proposes a simple
framework for dealing with this. Finally, Hall (1987) investigates whether
‘errors-in-variables’ may be influencing the regression results, but con-
cludes that measurement error does not appear to be an important factor
in her dataset.
4.3 FIRM SIZE AND GROWTH RATE VARIANCE
Hymer and Pashigian (1962) were among the first to draw attention to the
negative relationship between growth rate variance and firm size. If firms
can be seen as a collection of ‘components’ or ‘departments’, then the
overall variance of the growth rate of the firm is a function of the growth
rate variance of these individual departments. In many cases, the variance
of the firm’s growth rate will decrease with firm size. For example, in the
case where these departments (i) are of approximately equal size, such that
the size of the firm is roughly proportional to the number of components;
and (ii) have growth rates that are perfectly independent from each other,
then Central Limit Theorem leads us to expect a decrease in growth rate
variance that is proportional to the inverse square root of the firm’s size.
However, Hymer and Pashigian (1962) were puzzled by the fact that the
rate of decrease of growth rate variance with size was lower than the rate
that would be observed if large firms were just aggregations of independent
departments. At the same time, they found no evidence of economies of
scale. They saw this as an anomaly in a world of risk-averse agents. Why
should firms grow to a large size, if there are no economies of scale, and if
the growth rate variance of a large firm is higher than the corresponding
variance of an equivalent group of smaller firms? Subsequent studies did
not attempt to answer this question, but they did bear in mind the exist-
ence of a negative relationship between growth rate variance and firm size.
As a consequence, empirical analyses of Gibrat’s law began to correct for
heteroskedasticity in firm growth rates (for example Hall, 1987; Evans,
1987a,b; Dunne and Hughes, 1994; Hart and Oulton, 1996; and Harhoff
et al., 1998).
In recent years efforts have been made to quantify the scaling of the
variance of growth rates with firm size. This scaling relationship can be
summarized in terms of the following power law: s (gi) ~ebsi; where s(gi)
is the standard deviation of the growth rate of firm i, b is a coefficient
to be estimated, and si is the size (total sales) of firm i. Values of b have
consistently been estimated as being around 20.2 for US manufacturing
firms (Amaral et al., 1997, 1998; Bottazzi and Secchi, 2003a) and also for
firms in the worldwide pharmaceutical industry (Bottazzi et al., 2001; De
Fabritiis et al., 2003; Matia et al., 2004; Bottazzi and Secchi, 2006b). Lee
et al. (1998) find that a scaling exponent of 20.15 is able to describe the
scaling of growth rate variance for both quoted US manufacturing firms
and the GDP of countries. For French manufacturing firms, the analysis
in Bottazzi et al. (2008a) yields estimates of b of around 20.07, although
in the case of Italian manufacturing firms Bottazzi et al. (2007) fail to find
any relation between growth rate variability and size.
The discussion in Lee et al. (1998, p. 3277) gives us a better understand-
ing of the values taken by b, the scaling exponent. If the growth rates of
divisions of a large diversified firm are perfectly correlated, we should
expect a value of b 5 0. On the other hand, if a firm can be viewed as an
amalgamation of perfectly independent sub-units, we expect a value of b
Gibrat’s law 45
5 20.5. The fact that the estimated exponents are between these extreme
values of 0 and −0.5 suggests that the constituent departments of a firm
have growth patterns that are somewhat correlated.
Matia et al. (2004) and Bottazzi and Secchi (2006b) return to the scaling-
of-variance puzzle by considering firms as being composed of a certain
number of products that correspond to independent sub-markets. The
average size of the sub-markets increases with firm size, but the growth
rates are independent across sub-markets. These authors provide support
for their model by examining evidence from the worldwide pharmaceutical
industry, where a firm’s portfolio of activities can be decomposed to a fine
level of aggregation. As a result,
the explanation of the relationship between the variance of the growth rates
distribution and the size of the firm based on the Central Limit Theorem is valid,
as long as one considers the actual number of sub-markets a firm operates in,
instead of assuming that this number is somehow proportional to the size of the
firm. (Bottazzi and Secchi, 2006b, p. 860)
The model described in Matia et al. (2004) and Bottazzi and Secchi
(2006b) bears a certain similarity with the model in Amaral et al. (1998,
2001), who explain scaling of variance in terms of firms being composed of
independent ‘divisions’ in a diversified firm, rather than independent ‘sub-
markets’. Relatedly, Sutton (2002) explains the scaling of variance with size
by considering that firms are composed of independent business lines that
display great heterogeneity in size. Coad (2008b) conducts an empirical
investigation that can be situated along these lines, relating the variance of
growth rates to a firm’s multiplant structure. The variance of growth rates
is observed to decline with number of production plants, although interest-
ingly enough this decline is not monotonic. It may be that the tendency for
large size to be associated with lower variance is partially offset by a greater
propensity to take risks in larger firms.
4.4 AUTOCORRELATION OF GROWTH RATES
Early empirical studies into the growth of firms considered serial correla-
tion when growth was measured over a period of four to six years. Positive
autocorrelation of 33 per cent was observed by Ijiri and Simon (1967) for
large US firms, and a similar magnitude of 30 per cent was reported by
Singh and Whittington (1975) for UK firms. However, much weaker auto-
correlation was later reported in comparable studies by Kumar (1985) and
Dunne and Hughes (1994).
More recently, availability of better datasets has encouraged the con-

sideration of annual autocorrelation patterns. Indeed, persistence should
be more visible when measured over shorter time horizons. However, the
results are quite mixed. Positive serial correlation has often been observed,
in studies such as those of Chesher (1979) and Geroski et al. (1997) for
UK quoted firms, Wagner (1992) for German manufacturing firms, Weiss
(1998) for Austrian farms, Bottazzi et al. (2001) for the worldwide phar-
maceutical industry, and Bottazzi and Secchi (2003a) for US manufactur-
ing. On the other hand, negative serial correlation has also been reported
– some examples are Boeri and Cramer (1992) for German firms, Goddard
et al. (2002) for quoted Japanese firms, Bottazzi et al. (2007) for Italian
manufacturing, and Bottazzi et al. (2008a) for French manufacturing. Still
other studies have failed to find any significant autocorrelation in growth
rates (see Almus and Nerlinger (2000) for German start-ups, Bottazzi et al.
(2002) for selected Italian manufacturing sectors, Geroski and Mazzucato
(2002) for the US automobile industry, and Lotti et al. (2003) for Italian
manufacturing firms). To put it mildly, there does not appear to be an
emerging consensus.
Another subject of interest (also yielding conflicting results) is the
number of relevant lags to consider. Chesher (1979) and Bottazzi and
Secchi (2003a) found that only one lag was significant, whilst Geroski et al.
(1997) find significant autocorrelation at the third lag (though not for the
second). Bottazzi et al. (2001) find positive autocorrelation for every year
up to and including the seventh lag, although only the first lag is statisti-
cally significant.
To summarize these regression-based investigations, then, it would
appear that decades of research into growth rate autocorrelation can best
be described as yielding ‘conflicting results’ (Caves, 1998, p. 1950). It is
perhaps remarkable that the results of the studies reviewed above have
so little in common. It is also remarkable that previous research has been
so little concerned with this question. Indeed, instead of addressing serial
correlation in any detail, often it is ‘controlled away’ as a dirty residual, a
blemish on the ‘natural’ growth rate structure. The baby is thus thrown out
with the bathwater. One reason for this confusion could be that, if indeed
there are any regularities in the serial correlation of firm growth, they are
more complex than the standard specification would be able to detect (that
is, there is no ‘one-size-fits-all’ serial correlation coefficient that applies for
all firms). A fresh approach is needed.
The analysis in Bottazzi et al. (2002) begins with the observation that
the mean autocorrelation coefficient for a given industry is either insig-
nificantly different from zero, or else very small in magnitude. However,
the authors go on to calculate firm-specific autocorrelation coefficients
Gibrat’s law 47
and observe that firms do in fact have idiosyncratic growth patterns that
are not visible simply by looking at averages across firms. They create a
purely random ‘benchmark’ case in which the growth rates of all firms
are pooled together and then growth rates are extracted randomly to con-
struct growth patterns for ‘artificial firms’. Bootstrap resampling methods
allow them to generate a distribution of autocorrelation coefficients for
this random scenario. They then compare this stochastic benchmark with
the empirical distribution of autocorrelation coefficients (see for example
Figure 5 in Bottazzi et al. (2002) for the case of autocorrelation of employ-
ment growth). The differences between the distributions are supported by
formal statistical tests (that is, Kolmogorov–Smirnov tests). The authors
conclude that firm growth patterns are indeed idiosyncratic, that they do
have a memory process, and that there are indeed persistent asymmetries
in growth dynamics across firms.
Coad (2007a) also explores the issue of heterogeneous growth profiles
across firms and goes some way towards finding regularities in growth rate
autocorrelation patterns. A firm’s growth dynamics are seen to depend
on two dimensions – a firm’s size and its lagged growth rate. First of all,
it is demonstated that smaller firms are more prone to experience nega-
tive autocorrelation, whilst larger firms have a tendency towards positive
autocorrelation. This is consistent with propositions that small and large
firms operate on a different ‘frequency’ or timescale, with the actions of
large firms unfolding over a longer time horizon. This dependence of
autocorrelation on firm size helps to explain why the studies reviewed
above yielded different autocorrelation coefficients for databases with
different firm size compositions. Second, Coad (2007a) demonstrates that
the autocorrelation coefficient depends on the growth rate. Firms whose
growth rate is close to the average in one year are likely not to experience
any autocorrelation in the following year. However, those firms that expe-
rience extreme growth rates (either extreme positive or negative growth
rates) are likely to experience considerable negative autocorrelation. This
is especially true for fast-growth small firms, whose growth patterns are
particularly erratic. These findings are consonant with work by Garnsey
and Heffernan (2005), who highlight the ‘prevalence of interruptions to
growth’ (p. 675) for small firms, after observing that only a small fraction
of small firms grew continuously over their period of analysis. Large firms,
however, undergo a smoother growth experience – they are likely to experi-
ence positive autocorrelation irrespective of their growth rate in the previ-
ous period. Mrs Penrose offered this explanation: ‘Large firms often plan
further ahead than do smaller firms, partly because their greater financial
strength enables them to afford it, and partly because the nature of their
operations more or less forces them to do so.’ (Penrose, 1959, p. 244).
4.5 CONCLUSION
Like it or not, Gibrat’s law still reigns supreme as the leading model of
firm growth. Gibrat’s Law is essentially a random statistical process, and
as such it is often criticized by economists as having no theoretical founda-
tion. In this chapter we began by introducing Gibrat’s model of the ‘law
of proportionate effect’ (section 4.1) before reviewing empirical work on
the law and its implications. In section 4.2 we considered the relationship
between firm growth and size. The empirical literature is huge – much work
has been done on many different datasets. If it were indeed possible to gen-
eralize, then we would suggest the following recapitulation of this work. It
appears that smaller firms tend to grow faster than larger firms, although
above a certain size threshold these differences fade out, such that expected
growth rates are more or less independent of firm size. Gibrat’s law also
yields some other testable implications. In section 4.3 we looked at the
relationship between firm size and growth rate variance. In nearly all data-
sets considered, growth rate variance decreases with firm size. In section
4.4 we searched for an autocorrelation in firm growth rates. In many cases
there appears to be an autocorrelation structure in growth rates, although
the results are far from harmonious and suggest that autocorrelation can
either be negative or positive, or even in between (that is, no significant
autocorrelation). These conflicting results are reconciled somewhat by the
observation that small firms tend to display negative autocorrelation while
that of larger firms is more positive. Furthermore, it appears that small
firms that experienced rapid growth in one year are unlikely to be able to
repeat this in the following year.
5. Profits, productivity and firm
growth
Theoretical work has long been interested in the relationship between a
firm’s relative performance (measured either in terms of profits or pro-
ductivity) and its growth rate. In fact, a number of theoretical authors
tend to take it for granted that firms with higher performance will rein-
vest their profits into growth, such that the more efficient firms will have
higher growth rates. Empirical work into the matter, however, suggests
that the expected positive relationship between performance and growth
is generally lower than expected, or even non-existent.
We begin this chapter by looking at the relationship between profits
and growth (section 5.1). There remains a controversy, however, over
how the relationship between financial performance and growth should be
interpreted. Neoclassical economists, for example, think that in a perfect
world, current financial performance should not be related to investment.
If investment is related to financial performance, then financial constraints
are preventing firms from undertaking their optimal expansion plans.
Evolutionary economists take an opposite view – ideally, firm expansion
should be related to current profits. This is because progressive industrial
development requires the reallocation of scarce resources towards the more
efficient producers.
In addition to profits, many theorists have focused on the relationship
between productivity and firm growth. In fact, profits and productivity are
quite closely correlated and they can be considered as alternative indicators
of relative performance. The relationship between productivity and growth
is addressed in section 5.2.
In section 5.3 we look at models that look at the coevolution of both
profits, productivity, and other aspects of firm growth. These models
consider not only the effect of profits and productivity on subsequent firm
growth, but also the association between firm growth and subsequent
growth in profits or productivity. Section 5.4 concludes.
49
5.1 PROFITS AND GROWTH: FINANCIAL

CONSTRAINTS AND SELECTION EFFECTS
A number of empirical studies have looked for statistical relationships
between profits and firm growth. As an introduction to this literature,
Figure 5.1 plots the relationship between profits and sales growth for
ISIC17: textiles ISIC22: publishing and printing

1 1
growth rate (t-1:t)
growth rate (t-1:t)

0.5 0.5
0 0
–1 –1
–0.5 0 0.5 1 –0.5 0 0.5 1
profit rate (t-1) profit rate (t-1)
ISIC24: chemicals and chemical products ISIC25: rubber and plastic products
1 1
growth rate (t-1:t)
growth rate (t-1:t)
0.5 0.5
0 0
–1 –1
–0.5 0 0.5 1 –0.5 0 0.5 1
ISIC28: fabricated metal products ISIC29: machinery and equipment

1 1
growth rate (t-1:t)
growth rate (t-1:t)
0.5 0.5
0 0
–1 –1
–0.5 0 0.5 1 –0.5 0 0.5 1
Source: Coad (2007d).
Figure 5.1 The relationship between profit rate (t 2 1) and sales

growth (t 2 1 : t) where t5 2001, for selected 2-digit French
manufacturing sectors
Profits, productivity and firm growth 51
French manufacturing firms. Casual inspection of Figure 5.1 does not

suggest that there is any particularly strong association between these
two variables. Robson and Bennett (2000) look at the growth of British
Small and Medium Enterprises (SMEs) and observe a positive relationship
between profitability and both employment and sales growth, although it
is only statistically significant in the case of sales growth. Guariglia (2008)
observes that, among more profitable firms, higher profits are associ-
ated with higher levels of investment. Among the least profitable firms,
however, it appears that lower profits are associated with higher levels of
investment.
What do these results mean? Should any relationship between financial
performance and growth simply be taken at face value? Or is there any
deeper significance behind such results? How does firm investment and
growth react to current-period financial performance? How should it?
In this section we highlight the differences between competing theo-
retical perspectives on firm growth, and also the rather different policy
implications that emerge from them.1 The three perspectives are the neo-
classical q-theory of investment (and the related Euler equation approach
– see Chirinko (1993) and Schiantarelli (1996) for surveys), the ‘imperfect
capital markets’ theory (following on from Fazzari et al. (1988a); see
Hubbard (1998) for a survey), and also what we could call the evolutionary
perspective (along the lines of the theory that will be presented in section
8.4).
Research into the relationship between financial performance and firm
expansion has traditionally taken the view that any sensitivity between
financial performance and investment2 signals the problem of financial
constraints and imperfect capital markets. We begin by explaining how this
interpretation became predominant (sections 5.1.1 and 5.1.2). However,
drawing on evolutionary theory we argue that firms are heterogeneous
and that not all firms deserve to grow (section 5.1.3). As such, we conclude
that any positive association between profits and growth may be a desir-
able outcome.
5.1.1 q Theory
Neoclassical q-theory states that firm-level investment should be deter-

mined by future prospects of return. Assuming that stock prices can accu-
rately summarize future profits, the viability of investment opportunities
can be determined by the firm’s value of marginal q (that is, market value
of assets / book value of assets). However, data on marginal q is difficult to
obtain, and is usually proxied by average q. Average q has been shown to
be a valid proxy for marginal q when four assumptions are met (Hayashi,
Table 5.1 An example of a neoclassical q model: how Blundell et al.

(1992) derive the regression equation
Equation Description
(1) Intertemporal capital market arbitrage condition
(2) Solving (1) on an infinite horizon
(3) Defining the discount factor b over an infinite horizon
(4) Substituting for dividend payments in the firm’s stock market value
(5) Defining the firm’s after-tax net revenue
(6) First-order condition for investment
(7) The evolution of the shadow price of capital
(8) Rearranging (6) to obtain marginal q
(9) Rearranging (8) assuming a quadratic functional form for
adjustment costs
(10) Rewriting marginal q assuming linear homogeneity of production
and adjustment costs
(11) Expressing the expected depreciation allowances on an infinite
horizon
(12) Expressing the expected present value of all cash flows associated
with debt
(13) Regression equation
1982): that firms operate in perfectly competitive product and factor

markets, that firms also have linear homogeneous production and adjust-
ment cost technologies, that capital is homogeneous, and that investment
decisions are separable from other real and financial decisions. Assuming
that firms seek to maximize shareholder value and possess ‘rational expec-
tations’, it is possible to take the first-order condition of a mathematical
model as the basis for a regression model. In this final model, q should be
the only predictor for investment (Chirinko, 1993).
As an example of a prominent empirical study based on the neoclassical
q model, Table 5.1 shows how Blundell et al. (1992) derive their regression
equation. This table illustrates how the interpretation of the empirical
results obtained from regression analysis of their equation (13) is framed by
a rather long list of previous assumptions. One could argue that investiga-
tions such as these are ‘semi-empirical’ because their results are only open
to identification within the ‘straitjacket’ of a complicated mathematical
model. Perhaps unsurprisingly, we observe that ‘Q models have not been
noticeably successful in accounting for the time series variation in aggregate
investment’ (Blundell et al., 1992, p. 234).
An alternative to the q model is the Euler equation model. The Euler
equation describes the optimal path of investment in a parametric
adjustment costs model. Although it is derived from the same dynamic

optimization problem as the q-theory model, it has the advantage of
avoiding the requirement of measuring q. ‘It states that the value of the
marginal product of capital today, net of adjustment costs, must equal
the cost of a new machine minus the cost savings due to the fact that the
firm can invest less tomorrow and still maintain the capital stock on its
optimal path’ (Schiantarelli, 1996, p. 75). Euler equation studies also tend
to derive the regression equation from a long list of preceding theoretical
equations. For example, in the influential article by Whited (1992), the
regression equation is presented as equation (14) after being derived from
a complicated theoretical model. (For other examples of Euler equation
studies, see Bond and Meghir ,1994; Galeotti et al., 1994; and Bond et al.,
2003b.) Again, we direct the reader’s attention to how the regression results
are placed squarely in the context of the preceding mathematical models.
Any interpretation of the results as evidence of ‘suboptimal’ behaviour on
the part of firms is thus precluded.
Empirical research into investment decisions based on q models, and
the related Euler equation models, have typically produced disappointing
results (Barnett and Sakellaris, 1998). The explanatory power is typically
rather low (Blundell et al., 1992). Also, contrary to the theory, other variables
enter significantly into the investment equation, such as lagged q (Chirinko,
1993) cash flow (Fazzari et al., 1988a), and output (Blundell et al., 1992).
Furthermore, the implied adjustment costs of investment are generally so
high that they seem economically implausible (Schaller, 1990). Different
versions of the same underlying theory (that is, q models and Euler equation
models) sometimes give quite different results (Whited, 2006). It has also
been suggested that tests of the q-theory of investment have been outper-
formed by simpler ‘accelerator’ models of investment (Whited, 1992).
Geroski et al. (1997) investigate the influence of market value on sales
growth in a panel of large UK firms. They are able to detect a positive
relationship between market value and subsequent sales growth. The
regression coefficient is relatively small in magnitude, however, which leads
them to conclude that Gibrat’s model of random growth is an accurate
description of the firm growth process.
We can conclude from the preceding discussion that the q-theory of
investment performs unsatisfactorily. However, we don’t exactly know
why. Estimation of regression equations such as (13) in Table 5.1 is not
just a test of a single null hypothesis, but instead it is essentially a joint test
of the whole series of previous assumptions. The failure of the model to
produce results in line with the theory could be due to the failure of any
of these assumptions. One problem is that average q may not be a good
indicator of expected future profit (Chirinko, 1993; Erickson and Whited,
2000; Gomes, 2001). This may occur if the stock market is not perfectly effi-
cient at foreseeing a firm’s fortunes or allocating resources. Furthermore,
the denominator of q includes only fixed capital, and regrettably it does
not include those elements that are truly valued by shareholders and that
cannot be easily bought or sold on asset markets, such as management
skill, human capital or R&D capital. Furthermore, q may not be a good
predictor of investment if managers are boundedly rational, or if they just
don’t choose to grow on the basis of maximizing shareholder value. q may
also fail to predict investment if the other assumptions mentioned above
do not hold.
5.1.2 Imperfect Markets Theory
In the light of the disappointing performance of q-models, Fazzari et

al. (1988a) consider US manufacturing firms that are listed on the stock
market,3 include cash flow in the investment equation and observe that it
is significant. Why is cash flow a significant determinant of investment?
Predictions based on the neoclassical model (which is built on a large
number of unreasonable assumptions such as perfect competition, perfect
foresight, perfectly efficient financial markets, managers that are selfless
and optimizing, linear homogeneous production technologies, and so on)
do not allow for cash flow to be a predictor of investment. The real reason
why cash flow is significant is not really known. For example, in an uncertain
environment it could be that combining cash flow and average q may yield
a better proxy for marginal q than just average q alone. If firms are unable
to predict the future, they may prefer to base their investment decisions on
current-period indicators rather than speculative stock market indices.4
Alternatively, it could be because firms are wary of becoming dependent
on external finance.5 It could also be because managers are reluctant to
distribute dividends and prefer to spend free cash flow on additional invest-
ment projects (Jensen, 1986). However, the interpretation that Fazzari et
al. (1988a) gave is that any sensitivity of investment to cash flow is due to
financial constraints. The authors associated any such sensitivity to catch-
phrases such as ‘market imperfections’, ‘asymmetric information’, and the
‘lemons’ problem. In other words, any dependence of investment on cash
flow is seen as a failure of the capital markets, a source of inefficiency akin
to the problems raised in Akerlof (1970) and Stiglitz and Weiss (1981), and,
as such, a welfare-reducing problem in need of policy intervention.
One caveat of the Fazzari et al. (1988a) analysis is their choice of sample
of firms. As they introduce the concept of ‘financial constraints’, they
explain that small firms should be subject to such constraints whereas
larger firms should not:
only the largest and most mature firms are likely to face a smoothly increasing
loan interest rate . . . Small and medium-sized firms are less likely to have access
to impersonal centralized debt markets. . . . during periods of tight credit, small
and medium-sized borrowers are often denied loans in favor of better-quality
borrowers. (Fazzari et al., 1988a, p. 153)6
However, it is perhaps ironic that their final sample consists of large

firms that are quoted on the stock market. The authors do this because
they require values of Tobin’s q for these firms. However, the snag is that
these firms can hardly be described as small. In fact, Fazzari et al. (1988a)
acknowledge this, observing that even the smallest firms in their study
are ‘still large relative to US manufacturing corporations in general’. I
therefore suggest that problems of asymmetric information, which affect
smaller firms much more than larger firms, is not a useful interpretation
for investment–cash flow sensitivities in their study of large listed US
firms.
Following on from their empirical findings, Fazzari et al. (1988a) elabo-
rated upon the implications for policy. They underlined the importance
of investment opportunities being forgone due to credit market imperfec-
tions. As a consequence, they prescribed policy intervention to provide
finance for liquidity-constrained firms. They also highlighted the influence
of average tax rates (and not just marginal tax rates) on investment in
financially-constrained firms (see also Fazzari et al., 1988b).
Fazzari et al. (1988a) has since spawned a large stream of subsequent lit-
erature and is nowadays often branded as a ‘seminal paper’. The Fazzari et
al. (1988a) regression strategy (and interpretation of investment–cash flow
sensitivities) has been replicated, and extended in a number of ways, on a
large number of datasets. As one author remarked, ‘[t]he last two decades
have seen a flood of empirical studies of the effects of external finance
constraints on corporate investment’ (Whited, 2006, p. 467). Among this
large body of research, we can mention Hu and Schiantarelli (1998), Oliner
and Rudebusch (1992), Whited (1992), Gilchrist and Himmelberg (1995),
Hadlock (1998) and Carpenter and Petersen (2002) for US firms, Bond
and Meghir (1994), Bond et al. (2004) and Guariglia (2008) for UK firms,
Hoshi et al. (1991) for Japanese firms, Schaller (1993) for Canadian firms,
Galeotti et al. (1994) for Italian firms, Bond et al. (2003b) for European
firms, and Audretsch and Elston (2002) for German firms. Lamont (1997)
analyses investment undertaken by US oil companies, while Bond et al.
(2003b) and Guariglia (2008) consider samples of unlisted firms. Some
scholars have attempted to replace average q with financial analyst-based
measures of investment opportunities (see inter alia Bond et al., 2004 and
Cummins et al., 2006; see also the critique by Carpenter and Guariglia,
2007). Others have attempted to link other forms of investment to cash
flow, such as R&D investment (see for example Himmelberg and Petersen,
1994; Bougheas et al., 2003; Bond et al., 2003a; and Brown et al., 2007).
A common theme in many of these studies is that, whenever investment
(or firm growth) is associated with changes in cash flow, it tends to be pre-
sented as ‘bad news’. It is implicitly assumed that any investment–cash flow
sensitivities are signs of financial constraints, that investment opportunities
have been forgone, and also that these investment opportunities would
have been ‘optimal’.7 An interpretation based on market imperfection is
evoked, and policy makers have frequently been urged to intervene to help
constrained firms to grow.
However, the Fazzari et al. (1988a) approach to investment research has
recently met an extensive criticism by Kaplan and Zingales (1997, 2000).8
To begin with, Kaplan and Zingales present a theoretical model to show
that any sensitivity of investment to cash flow should not be interpreted
as evidence of financial constraints (see also the theoretical model by Alti,
2003). They also re-examine the original Fazzari et al. (1988a) database in
conjunction with a scrutiny of annual company reports of these companies,
and observe that the highest investment–cash flow sensitivities actually
belong to those firms that seem to be the least financially constrained.
Indeed, ‘wrong-way’ differential investment–cash flow sensitivities (that is
where firms classified as more constrained a priori display lower sensitiv-
ity of investment to cash flow) have also been found by a number of other
researchers, such as Gilchrist and Himmelberg (1995), Kadapakkam et al.
(1998), Cleary (1999) and Erickson and Whited (2000). One notable example
mentioned by Kaplan and Zingales (2000) is that, in 1997, Microsoft would
have been labelled as ‘financially constrained’ according to the classifica-
tion schemes of Fazzari et al. (1988a, 2000) even though it had almost $9
billion in cash, corresponding to 18 times its capital expenditures!
5.1.3 Evolutionary Theory
The basic evolutionary prediction is that expansion of operations should

be the domain of the ‘fitter’ firms (but not necessarily only the ‘fittest’). In
constrast, the weakest should decline and exit. In this view, a population
of firms cannot be represented in terms of a single maximizing firm, but
instead there is considerable heterogeneity between firms, such that high
productivity firms can be found alongside low productivity firms even in
narrowly defined industries. Not all firms should grow. Resources should
be allocated to the more productive firms, whereas growth of the least
productive should be discouraged (Coad, 2008a).
Evolutionary economics also stresses the importance of the Simonian
notion of ‘bounded rationality’. A firm’s future is not known, it cannot be
‘rationally anticipated’, and its course can be changed by luck or human

will. As a result, a firm cannot make its investment decisions on discounted
expected future returns on an infinite horizon. Instead, firms are myopic,
and their investment is largely determined by the firm’s current financial
performance. The existence of any significant explanatory power of market
value (reflected in Tobin’s q) over and above that of current financial
performance does not undermine the fundamental relationship between
growth and current profitability; instead it would probably be welcomed
as supplementary information.
The evolutionary mechanism of ‘selection via differential growth’ (which
is discussed in more detail in section 8.4) can be presented formally by
Fisher’s ‘fundamental equation’, which states that:
–
xi 5 xi (Fi 2 F ) (5.1)
where d stands for the variation in the infinitesimal interval (t, t 1 dt), and
xi represents the market share of firm i in a population of competing firms.
Fi is the level of ‘fitness’ of the considered firm, where fitness corresponds
to relative financial performance or perhaps some measure of relative pro-
– –
S
ductivity. F is the average fitness in the population, i.e. F 5 ixiFi, and a
is a parameter. It is straightforward to see that this equation favours the
above-average firms with increasing market share, whilst reducing that of
‘weaker’, less profitable firms.
Although empirical investigations of the evolutionary principle of
‘growth of the fitter’ are rather scarce, some empirical studies of evolu-
tionary inspiration can be found in Dosi (2007), Bottazzi et al. (2008b)
and Coad et al. (2008) for Italian manufacturing firms; Coad (2007b,d)
for French manufacturing firms; and Coad and Rao (2009) for US manu-
facturing. Dosi (2007) and Bottazzi et al. (2008b) present scatterplots of
the relationships between profits, productivity and sales growth. While
there appears to be a relationship between profits and productivity, these
variables seem to be unrelated to firm growth. Coad (2007d) applies panel-
data instrumental variable techniques to account for a lag structure, firm-
specific time invariant effects, and endogeneity in the relationship between
profits and growth. He finds a statistically significant relationship between
financial performance and sales growth for French manufacturing firms.
Nevertheless, the magnitude of the coefficient is so small that he concludes
‘it may be more useful to consider a firm’s profit rate and its subsequent
growth rate as entirely independent’ (p. 385). Coad (2007d) also observes
that the effect of profits on firm growth appears to be overshadowed by
the effect of firm growth on subsequent profitability. Coad (2007b), Coad
and Rao (2009) and Coad et al. (2008) focus on the co-evolution over time
of a number of variables relating to firm growth and performance (these

models are presented in more detail in section 5.3). In these models, profits
and productivity have little association with subsequent firm growth.
In contrast, the association between employment and sales growth, on
the one hand, and subsequent growth of profits, on the other hand, is
considerably larger. A common finding in these approaches, however, is
that financial performance does not seem to be an important determinant
of firm growth, whether this latter is measured in terms of investment or
sales growth. Although the coefficients on financial performance are often
statistically significant, there is a large amount of unexplained variation in
growth rates.9 Firms appear to have a large amount of discretion in their
growth behaviour.
These studies look for associations between growth and operating
margin, whilst including controls for other potentially significant factors.
It should be noted, however, that the regression methodology is slightly
different from the neoclassical-based studies reviewed above. First of all,
firm growth is measured in terms of sales growth rather than investment in
fixed assets, because evolutionary theory emphasizes the important role of
firm-specific capabilities and intangible capital (rather than fixed tangible
assets) in economic change. Furthermore, operating margin is used instead
of cash flow as a measure of current-period financial performance: these
two indicators are nonetheless closely related.10
Predictions from evolutionary economics are also in line with those
originating in the behavioural finance literature. Consider the empirically-
based ‘financial pecking-order’ theory (Myers, 1984), which supposes
there is an imperfect substitutability of internal and external sources of
finance. In this view, firms are quite willing to spend free cash flow on
investment projects but are much less enthusiastic about having to resort
to external finance (see also Hines and Thaler, 1995). As a result, changes
in cash flow would be positively associated with changes in investment.
Furthermore, evolutionary economics is able to accommodate ‘manage-
rial’ theories of firm growth (see for example Marris, 1964), which posit
that managers attach positive utility to their firm’s size. In this perspective,
managers pursue growth even when this is not in the interest of sharehold-
ers. Growth is thus maximized subject to certain constraints (that is a
minimum value for the firm’s shares). Under these circumstances, invest-
ment will respond positively to improvements in current financial per-
formance. Relatedly, the ‘agency theory’ of free cash flow (Jensen, 1986)
should be mentioned. This theory predicts that managers will be reluctant
to distribute available cash flow as dividends but will prefer to spend it
on investment projects (even if these are likely to generate low returns).11
Recently, however, attempts have been made by mainstream economists
to introduce these aforementioned ‘behavioural finance’ considerations

into the Fazzari et al. (1988a)-based financial constraints literature (see
the promising work by Goergen and Renneboog, 2001 and Degryse and
de Jong, 2006).
Table 5.2 presents the regression equations investigated in the different
theoretical approaches. The early regression equations, such as those in
neoclassical q models and Fazzari et al. (1988a)-type imperfect market
equations, have been succeeded by reduced form models, such as Fagiolo
and Luzzi (2006) and Sarno (2008). These reduced form models have
regression specifications that are very similar to ‘evolutionary’ models, and
the results obtained are of course also very similar: financial performance
is positively associated with firm growth. It should be recognized that the
issue here cannot be resolved by a simple ‘horse-race’ between the perform-
ance of different regression models, because the difference lies not in the
regression model nor the results obtained, but in the interpretation of the
results. The former models tend to interpret their results as a failure of
financial markets, and lean towards suggesting that policy should intervene
to help provide funding for financially constrained firms. Evolutionary
models, however, interpret any such positive association as evidence that
selection pressures are alive and well, and the economy is reallocating
market share towards the more productive firms. Indeed, evolutionary
authors may even lament that the positive association between financial
performance and growth is not stronger in magnitude.
5.1.4 Evaluating the Importance of Financial Constraints
In this section it is argued that the basic neoclassical assumptions, which

also form the basis of the ‘asymmetric information’ models, find their way
into the policy conclusions. In particular, I criticize the assumption of
perfectly rational, shareholder-wealth-maximizing managers. The motiva-
tions behind this choice of assumption are technical in nature and have
little to do with the underlying economic reality; this assumption exists
mainly to aid tractability of the mathematical construct. However, this
assumption has an important role in the framing of the research question.
In discussions of the empirical results, questions relating to the quality
of managers’ investment projects are no longer posed. Instead, when
the q-model is observed to perform poorly and cash flow is observed to
be statistically significant, all too often buzzwords such as ‘asymmetric
information’ and the ‘lemons’ problem are automatically applied. In
many empirical studies, it appears to be implicitly accepted that firms
should have the right to realize their investment opportunities, and that
the government should intervene to make sure these firms get the finance
Table 5.2 Types of regression equation associated with the different theoretical perspectives
Theoretical approach Basic regression equation Remarks

q-theory (I/K)it 5 a 1 b1qit 1 eit If the assumptions hold, investment should be entirely explained by q.
(e.g. Blundell et al.,
1992)
Euler equation (I/K ) it 5 b1 (I/K ) i,t21 2 Investment dynamics should follow the optimal investment path
(e.g. Bond et al., b2 (I/K) 2i,t21 2 b3 (P/K) i,t21 1 in the context of parametric adjustment costs. Marginal costs of
2003b) b4 (Y/K) i,t21 1 eit investment in time t are set equal to marginal costs of foregone
investment in t 1 1. Theory predicts that b1 $ 1, b2 $ 1, b3 . 0
and b4 $ 0. If the Euler equation regressions perform poorly, one
60
explanation could be that firms are financially constrained.
Imperfect markets (I/K)it 5 b1qit 1 b2 (CF/K)it 1 eit Any explanatory power of cash flow over and above that of q
(e.g. Fazzari et al., indicates financial constraints.
1988a)
Reduced form models (DS/S) 5 b2 (CF/S) it 1 eit Cash flow taken as a proxy for financial constraints (without
(e.g. Fagiolo and controlling for q). Any sensitivity of sales growth to cash flow is
Luzzi, 2006) interpreted as financial constraints.
Evolutionary approach (DS/S) 5 b2 (OM/S) it 1 eit Sales growth should be associated with operating margin according
(e.g. Coad, 2007c) to the principle of ‘growth of the fitter’.
Notes: I is investment for firm i at time t, K is fixed assets, q is Tobin’s q, Y is output, CF is cash flow, (DS/S) is the growth rate of sales, OM is
operating margin, e is the residual error term. ∏it 5 pitF(Kit, Lit) − pitG(Iit, Kit) − witLit (see Bond et al., 2003b, p. 156).
they want. To sum up, one caricature of the neoclassical approach could
be: ‘Assume firms are efficient. Financial-market imperfections prevent
them from getting enough funding. Policy should intervene, perhaps by
subsidizing firm investment.’
A major caveat of the mainstream neoclassical literature is that it takes
as a starting point the assumption that firms are perfectly rational and
will invest only if this increases their long-term profits. In this stream of
literature, questions regarding the existence of imperfections in the finan-
cial system take centre stage, while questions regarding suboptimality of
firms are largely ignored. Evolutionary economics, in constrast, discards
assumptions of hyper-rationality and starts from the hypothesis that firms
are heterogeneous, boundedly rational, and that not all firms deserve to
grow.
But how important are financial constraints for economic development,
really? In an attempt to answer this question, we focus our investigations
where financial constraints are alleged to matter the most: the case of
small young entrepreneurial firms. We begin by referring to a growing
literature on the theme of excessive start-up (that is, entry beyond the
socially optimum level) which is surveyed extensively in Santarelli and
Vivarelli (2007). These authors observe that ‘at the end of the 90s, in
the UK the incidence of people starting a firm not because of a market
opportunity but just because they had no better choice was about 22%’ (p.
461). This can be compared to statistics (admittedly from other sources)
reporting that there is only ‘a minority of firms (about 15–20%) indicating
the desire to introduce product and/or process innovation as a funda-
mental reason to start a new independent economic activity’ (Santarelli
and Vivarelli, 2007, p. 463). Santarelli and Vivarelli (2007) emphasize
the important role that market selection has to play, and consider that
‘(early) failure should be seen as socially optimal rather than the result of
either financial market imperfections or other market failures.’ (p. 473).
As a result, they judge that financial constraints are not a major problem
affecting entrepreneurship and the growth of (small) firms, and conclude
that ‘modern developed economies are affected by too many start-ups
and that policy subsidies have contributed to an overall “excess of entry”
which – far from fostering economic growth – may just fuel turbulence
and market churning.’ (p. 475).
Other contributions, of a theoretical nature, have related over-entry
to over-optimistic forecasts of entrepreneurs (Dosi and Lovallo, 1998;
Camerer and Lovallo, 1999 and Arabsheibani et al., 2000). These articles
go on to suggest that entry of over-confident low-quality entrepreneurs
may even crowd out higher-quality entrepreneurs. It has been argued
that marginal entrepreneurs can free-ride on the credentials of more able
entrepreneurs, thus bringing down the average quality of the credit pool
(de Meza and Webb, 1987, 1999; de Meza, 2002). Marginal entrants should
thus be discouraged from entering. This body of theoretical work also sug-
gests that the use of internal finance to fund start-ups has beneficial effects
on start-up survival rates (through ‘incentive effects’) and also plays a role
in reducing moral hazard. As a result, it has been suggested that start-
ups should not be subsidized.12 Empirical evidence on excessive start-up
should also be taken into consideration (for example Dunne et al., 1988;
Bartelsman et al., 2005). These studies highlight the waste associated with
entry of new firms, by showing that a large proportion of entrants can be
expected to fail only a few years.
It is nonetheless rather unsettling to observe that the recommendations
emerging from the neoclassical literature have, to a certain extent, been
able to guide policy.
A belief that capital-market failure has held back enterprise has been a factor
behind policies designed to encourage start-ups and business expansion.
Examples include the grants and subsidies provided by the Federal Small
Business Administration in the United States and the Loan Guarantee Scheme
in the United Kingdom. (de Meza and Webb, 1999, p. 153)
In the United States, for example, there have been public initiatives
to provide finance to small firms that are suspected of being ‘financially
constrained’. According to Lerner (2002), these ‘public venture capital
programmes are often characterised by a considerable number of under-
achieving firms. . . . The end result can be a stream of government funding
being awarded to companies that consistently underachieve.’ (pp. 81–82).
Levenson and Willard (2000) are also critical of schemes such as the
provision of guaranteed loans to small firms.13 They remark that ‘there is
no direct evidence that small firms are, in fact, credit rationed in formal
capital markets’ (p. 84). Using data from a national survey in 1988–89, they
calculate an upper bound for the share of small businesses that were credit-
rationed as 6.36 per cent, and conclude that ‘the extent of true credit ration-
ing appears quite limited’ (p. 83). Likewise, Hughes (1997) focuses on UK
SMEs and concludes that ‘the evidence for general equity or debt gaps in
the UK is weak. If anything, SME funding was too easy in the boom of
the late 1980s.’ (p. 151). Cressy (1996) reaches a similar conclusion: ‘an
appropriate government policy should be to make business startups more
difficult, rather than less’ (Cressy, 1996, p. 1266).
Finally, it should also be noted that apart from being a potential waste of
funds, government initiatives to alleviate financial constraints also have the
drawback of encouraging socially wasteful rent-seeking behaviour (Little,
1987; Lerner, 2002).
Summary
Neoclassical economists hold that the existence of any sensitivity of invest-

ment to cash flow is interpreted as a signal of market failure, a problem
which is worthy of a policy intervention. Evolutionary economists stand
out from their neoclassical counterparts by abandoning the assumption of
rational profit-maximizing firms; instead, evolutionary economists con-
sider that firms are heterogeneous and that not all firms deserve to grow.
As a result, we suggest that the problem of financial constraints impeding
firm growth has been exaggerated by much of the neoclassical literature.
We argue in favour of an evolutionary interpretation of the relationship
between financial performance and growth, according to which growth
opportunities should ideally be given to the more profitable firms. In this
way, the economies’ scarce resources would be reallocated to the more
efficient producers, which would result in overall productivity growth.
From empirical work on the matter, however, it is clear that financial
performance is not a major determinant of firm growth rates. Firm growth
appears to be a remarkably idiosyncratic phenomenon.
5.2 RELATIVE PRODUCTIVITY
In this section we begin by discussing previous work on the relationship

between productivity and firm growth (section 5.2.1), before moving on to
recent attempts at decomposing productivity growth into the contributions
of within-firm learning, between-firm reallocation of resources, productivity
growth via firm growth, and entry and exit (section 5.2.2).
5.2.1 Productivity and Growth
It is perhaps quite natural to assume that the most productive firms will
grow while the least productive will decrease in size. A number of theo-
retical contributions have made strong statements along these lines (see
in particular the evolutionary literature in section 5.1.3). This assumption
does not seem to be borne out by empirical work, however. A number of
studies have cast doubt on the validity of the evolutionary principle of
‘growth of the fitter’, when relative productivity is taken as a proxy for
fitness. One explanation for this is that while some firms become more
productive through expansion, others become more productive through
downsizing. An illustration of this is provided by Baily et al. (1996) who
observe that, among plants with increasing labour productivity between
1977 and 1987, firms that grew in terms of employees were balanced out
by firms that decreased employment. They find that about a third of labour
productivity growth is attributable to growing firms, about a third to
downsizing firms, and the remaining third is attributable to the processes
of entry and exit. Similarly, Foster et al. (1998) also fail to find a robust
significant relationship between establishment-level labour productivity or
multifactor productivity and growth (see also the review in Bartelsman and
Doms, 2000, pp. 583–4). In addition, using a database of Italian manufac-
turing firms, Bottazzi et al. (2002, 2008b) fail to find a robust relationship
between productivity and growth.14
Other researchers working on other datasets have been able to detect a
positive association between relative productivity and firm growth. Pavcnik
(2002) investigates productivity growth among Chilean manufacturing plants
as the Chilean economy was undergoing significant liberalization and deregu-
lation. She observed that aggregate productivity increased by 19 per cent
over a seven-year period, and that most of this productivity growth was due
to the real-location of resources from the less to the more efficient producers.
Sleuwaegen and Goedhuys (2002) also observe a positive relationship between
productive efficiency and sales growth in their sample of Ivorian manufactur-
ing firms (although the effect is not always statistically significant). Likewise,
Liu et al. (1999) observe that labour productivity has a positive effect on
growth in their sample of Taiwanese electronics plants. Maksimovic and
Phillips (2002) find evidence that selection effects also operate within con-
glomerate firms. They observe that the growth rates of plants in both single-
and multiple-segment firms are significantly and positively related to their
productivity. The sensitivity of plant growth to productivity is greater for
single-segment firms than it is for multiple-segment firms, though.
There is ample evidence suggesting that low productivity helps to predict
exit.15 Bellone et al. (2008) observe that both productivity and profitabil-
ity are positively related to the probability of survival, although the chief
factor that influences survival is profitability. Schlingemann et al. (2002)
report that conglomerates are more likely to divest peripheral segments
that are performing poorly (see their Table 6). Maksimovic and Phillips
(2008) perform a plant-level analysis; they find that more productive plants
are less likely to be shut down (see in particular their Table 11).
Relatedly, many researchers have provided indirect evidence on the
matter, by reporting that industries experience positive productivity
growth that is due, in part, to the exit of relatively inefficient establishments
(see for example Griliches and Regev, 1995; Foster et al., 2006; and Foster
et al., 2008). Although Foster et al. (2008) find that low productivity is
associated with exit, they nonetheless stress that it is profitability, rather
than productivity, that is the true determinant of survival.
Nonetheless, it seems that productivity levels are not very helpful in
predicting growth rates. Put differently, it appears that selection only oper-
ates via elimination of the least productive firms or establishments, while
the mechanism of selection via differential growth does not appear to be as
strong. As a result, the mechanism of selection appears to be rather ‘sub-
optimal’ in the sense that its effectiveness is lower than it could conceivably
be. For Baily and Farrell (2006), the lack of a positive relationship between
relative productivity and growth corresponds to a lack of competition.
In an ideal scenario, firms would compete for growth opportunities, and
selective pressures would attribute these growth opportunities discrimi-
nating in favour of the most productive firms. In this way, there would be
some sort of dynamic efficient reallocation at work, whereby an economy’s
scarce resources are redistributed to those firms that are able to employ
them most efficiently. In reality, however, this mechanism does not seem
to be operating. Instead, the evidence is consistent with the hypothesis that
many of the more productive firms may not actually seek to grow, or may
be unable to grow. As a consequence, the absence of selection via differen-
tial growth is evidence of missed productivity growth opportunities for the
economy as a whole. Whilst we can put forward here that stimulating the
growth of high-productivity firms might constitute an objective for policy,
it is evident that there are large question marks surrounding how such a
policy intervention might be engineered.
5.2.2 Decomposing Productivity Growth
In this section we refer to recent work that decomposes productivity

growth and attributes it to underlying processes of learning, growth, and
entry and exit processes of establishments. In particular, we focus on the
decomposition of productivity growth pioneered by Baily et al. (1992) and
modified by Foster et al. (1998).16 Foster et al. (1998) propose the following
decomposition of productivity growth:
DLPt 5 a qi,t21DLPit 1 a (LPi,t21 2 LPt21) Dqit

i[C i[C
1 a DLPitDqit (5.2)
i[C
1 a qit (LPit 2 LPt21) 2 a qi,t21 (LPi,t21 2 LPt21)
i[N i[X
DTFPt 5 a qi,t21DTFPit 1 a (TFPi,t21 2 TFPt21) Dqit

i[C i[C
1 a DTFPitDqit (5.3)
i[C
1a qit (TFPit2TFPt21) 2 a qi,t21 (TFPi,t212TFPt21)
i[N i[X
where ΔLPt and ΔTFPt correspond to share-weighted growth of labour

productivity (LP) and total factor productivity (TFP), respectively, for
the period ending at time t. Establishments are indexed by i and may be
classified as either continuing (C), entering (N) or exiting (X) establish-
ments. q corresponds to the activity shares attributed to establishment i.
Bars over variables indicate that the average has been taken over all the
establishments.
The five terms on the right-hand side of equations (5.2) and (5.3) can
be explained as follows. The first term corresponds to the within-plant
learning effect, which is the component of productivity growth that occurs
within existing plants. The second term corresponds to between-plant
reallocation. This term accounts for changes in the shares of plants. If
this term is positive, then production capacity is being reallocated from
the least efficient to the more efficient plants. If negative, then the least
productive plants are growing faster than the more productive ones. The
third term is the cross term, the interpretation of which is not so simple.
This term relates to the association between relative growth and changes
in relative productivity. The cross term will be positive if growing firms
experience productivity growth, and negative if growth is accompanied
by decline in productivity. The last two terms correspond to changes in
productivity due to entry and exit, and are often combined to obtain the
overall effect of net entry. The fourth term will be positive if entrants are
more productive than continuing plants, and the fifth term will be posi-
tive if it is the least efficient plants that contribute to productivity growth
by exiting.
Table 5.3 presents a comparison of the results obtained from studies
that decompose productivity growth according to the technique presented
in equation (5.2) or (5.3). Care should be taken while reading the results,
because of differences in methodology. In some cases, the establishment
shares (qit) are weighted by employment (Disney et al., 2003; Foster et
al., 2006), whereas in others they are weighted by output (Foster et al.,
1998, 2008). In some cases, productivity is measured as labour produc-
tivity whereas in others a multifactor productivity indicator is preferred.
The different studies analyse data on different firms over different time
periods, and these time periods are not all of the same length. Bearing
these differences in mind, however, some interesting comparisons can be
made.
Table 5.3 shows that within-plant share of productivity growth (Column
1) is positive and in most cases contributes between a third and a half
of total productivity growth. Plants become more productive over time
as they gain experience and learn about more efficient production tech-
niques. The between-plant component of productivity growth (Column
Table 5.3 Decomposing productivity growth. Productivity growth decompositions calculated using the Foster et al.
(1998) technique (see equations 5.2 and 5.3)
(1) (2) (3) (4) sum (1)–(4)

Data Prod. Years Within- Between- Cross- Net entry Total
Measure plant plant plant
Foster et al. (1998) US manufacturing TFP 1977–1987 0.48 20.08 0.34 0.26 1.00
67
LP 1977–1987 0.45 20.13 0.37 0.31 1.00
Disney et al. (2003) UK manufacturing TFP 1980–92 0.05 0.15 0.26 0.54 1.00
LP 1980–92 0.48 0.04 20.01 0.49 1.00
Foster et al. (2006) US retail LP 1987–1997 0.16 0.24 20.39 0.98 1.00
Foster et al. (2008) US manufacturing TFP 1977–1997 0.36 20.17 0.50 0.30 1.00
Notes: columns (1) to (4) may not add to 1.00 because numbers in columns (1) to (4) are rounded off. In some cases, the establishment shares (qit)
are weighted by employment (Disney et al., 2003; Foster et al., 2006), whereas in others they are weighted by output (Foster et al., 1998, 2008).
Foster et al. (2008) estimates are pooled for the four five-year periods over the interval 1977–1997.
2) varies considerably across the specifications, although it does not seem

to make a major contribution to productivity growth. In half the cases,
the between-plant share is positive and in the others it is negative. The
contrasting results obtained for the between-plant share are reminiscent
of the discussion in section 5.2.1, in which it appeared that the existing
literature has found no clear-cut relation between productivity and firm
growth. The reallocation of productive resources from the less efficient
to the more efficient establishments does not seem to be a robust feature
of growth of productivity in industry, and, judging by the negative sign
observed in half of the cases, it may sometimes be that less efficient estab-
lishments grow relatively faster than the others. Productivity growth,
it seems, tends to come mainly from channels other than the dynamic
reallocation of productive capacity from the least efficient to the more
efficient firms. Column 3 shows the cross term, and in most cases this
term is positive. Growing firms, in many cases, become more productive
as they grow and thus contribute to overall productivity growth. Column
4 corresponds to the combined share of entry and exit processes. Entry
and exit make a considerable contribution to productivity growth in each
of the studies reported.
Foster et al. (2006) focuses on the retail sector, a sector which experi-
enced tremendous productivity growth over the period of analysis. The
effect of entry and exit on productivity growth is quite remarkable in mag-
nitude (0.98), and suggests that almost all of the growth of productivity
in the retail sector over the period of analysis can be attributed to entry
and exit processes. Closer analysis reveals that much of this productivity
growth comes from new establishments opened by continuing multi-plant
firms (such as Wal-Mart), and the closure of establishments from single-
plant firms.
Foster et al. (2008) report results for productivity growth decomposi-
tions using traditional productivity measures (shown in Table 5.3), and
also alternative measures of productivity that take into account the fact
that establishments may charge different prices (not reported here). The
estimates they obtain from the alternative price-adjusted productivity
measures attribute a larger share of productivity growth to the within-
plant share. They also argue that the productivity growth attributed to
net entry should be higher, the reason being that entrants tend to charge
lower prices and so their true productivity is not accurately measured
using traditional productivity indicators.17 They also observe that the cross
term decreases in magnitude when different prices across establishments
are taken into account, which is consistent with the hypothesis that young
producers start with low prices but tend to increase prices as they grow
and age.
5.3 VAR MODELS OF FIRM GROWTH PROCESSES

Firm growth can be seen as a multidimensional phenomenon, and while
no single indicator of firm size is perfect, different indicators of firm size
provide information on different aspects of a firm’s size and expansion.
The correlation between different indicators of firm size (such as sales
growth and employment growth) can be relatively low (Delmar et al., 2003;
Shepherd and Wiklund, 2009), and may even be low enough to consider
these variables as being distinct or independent.
The reduced-form VAR (vector autoregression) models presented in
this section offer new insights into the firm growth process by looking
at the dynamic relationships of variables such as employment growth,
sales growth and growth of profits, and describe the complex structure of
interactions between these variables as well as the lead-or-lag associations
between them.
The VAR models summarize the co-movements of the different variables
as the growth process unfolds. No particular variable is to be preferred
as the main dependent variable, but the starting point is that each vari-
able depends on each other variable. Theoretical reasoning has suggested
that while profits influence firm growth, that firm growth also influences
profits (Dobson and Gerrard, 1989; Coad, 2007d). It is possible to identify
several mechanisms through which firm growth may be negatively associ-
ated with rates of profit. The classical, ‘Ricardian’ stance is that if a firm
is enjoying relatively high profit rates, it will expand to exploit additional
business opportunities that are less profit-intensive but that nonetheless
generate profit. In neoclassical terms, such a firm grows until its marginal
cost of production is equal to the marginal revenue on goods sold. Such
a firm starts by exploiting its most profitable business opportunities, and
then includes less and less profitable opportunities until the marginal
profit on the last opportunity exploited is equal to zero. Thus, a profit-
able firm that expands in this way maximizes its overall levels of profits,
but experiences a decrease in its profit rate when profits are divided by
scale of production. Edith Penrose (1959) also suggests that, above a
certain point, growth may lead to a reduction in the profit rate, although
for different reasons. Firm growth requires managerial attention, and if
managers focus on the expansion of their firm, their attention is diverted
from keeping operating costs down. Thus, ‘Penrose effects’ occur when
costs inflate, as managers focus not on operating efficiency but instead on
exploiting new opportunities. On the other hand, the notion of ‘increasing
returns’ predicts that growth will lead to a higher, not lower, profit rate.
Increasing returns may allow a firm to achieve gains from specialization
and build up economies of scale in production, thus reducing the unit cost
of its products. Dynamic increasing returns, as described by Kaldor and

Verdoorn (see for example McCombie, 1987), might also be applicable at
a firm level, such that firm growth leads to increases in productivity and
thus increases in profit rates. For example, expanding firms may invest
in new technologies and learn about more efficient methods of produc-
tion. Their growth may also be an anticipation of medium-term demand
prospects, which (if correctly anticipated) would allow them to earn large
profits in the future. Finally, from the resource-based perspective, growth
may lead to increases in profits if it feeds off organizational slack and
puts resources that were previously idle or underutilized to good use. An
implication of learning-by-doing is that managerial (and other) resources
are continually being freed up as time passes and experience accumulates.
Large profits can be earned if these newly-liberated resources are used to
grow the firm.
Growth of profits is therefore not just a final outcome for firms but it can
also be an input, providing firms with the means for expansion. Similarly,
we have reason to suspect that changes in productivity can be both the
antecedent and the outcome of firm growth. Employment growth can be
seen as an input (in the production process) but also as an output if, for
example, the policy maker is interested in the generation of new jobs.
Another valuable feature of the VAR models is that firm growth is
seen as an ongoing process. Organizational growth and development
is often portrayed as a continuous phenomenon, a process that has no
particular beginning or end, in which no particular event can be seen
as entirely exogenous (Weick, 1995). Reduced-form VARs are suitable
modelling devices because we want to avoid taking a strong position as
to the causal structure of the firm growth process. Such reduced-form
models impose no a priori causal structure on the econometric model, but
instead they favour a descriptive approach. At this early stage of apply-
ing VAR models to the analysis of firm growth, we are mainly interested
in summarizing the co-movements of the main variables, and describing
the time profile of firm growth processes. Future work, however, will no
doubt seek to firmly establish the causal direction between the variables.18
Nonetheless, Coad (2007d) applies dynamic panel data instrumental-
variables GMM (Generalized Method of Moments) techniques to the
relationship between profits and firm growth for this dataset of French
manufacturing firms, in an attempt to unravel the direction of causal-
ity between profits, on the one hand, and sales growth or employment
growth, on the other. The GMM results are close to those obtained from
simpler regression models, however, and the results suggest that growth
of profits has only a small, and perhaps even negligible, influence on sales
or employment growth.
Profits, Productivity and Firm Growth
Coad (2007b) applies a reduced-form VAR model to data on French manu-

facturing firms with 20 or more employees, over the period 1996–2004. The
variables of interest are growth of employment, growth of sales, growth of
absolute profits (where profits are measured in terms of gross operating
margin),19 and growth of labour productivity.
The results suggest that growth of a firm’s employment is associated
with previous growth of sales and of labour productivity. Sales growth
and labour productivity growth have a relatively small positive effect, and
the magnitude is of a similar order even at the second lag. Employment
growth, however, appears to be relatively strongly associated with subse-
quent growth of sales and of profits. As could be expected, sales growth
and productivity growth also appear to make a relatively large contribu-
tion to the subsequent growth of profits. Indeed, sales growth has a size-
able impact on GOS (Gross Operating Surplus) growth even at the second
lag.
It is also interesting to observe the dynamic associations between growth
of labour productivity, on the one hand, and growth of employment and
of sales, on the other. In effect, we observe some sort of ‘positive feedback
loop’ whereby growth of productivity is positively associated with sub-
sequent growth of sales/employment, and vice versa. This aspect of firm
growth is consistent with ‘increasing returns’ theories of firm growth. The
importance of this effect should not be exaggerated, however. The coef-
ficients are relatively small in magnitude, and this feedback phenomenon
appears to operate in the shadow of stronger negative autocorrelation
dynamics that are visible in the time series of sales growth and productivity
growth.
In addition, it appears that growth of profits is associated with a rela-
tively small subsequent growth in sales, and an even smaller growth of
employment. Growth of profits may have a more persistent effect on
employment growth than on sales growth, however. (Growth of sales,
on the other hand, is very strongly associated with subsequent growth of
profits.)
Figure 5.2 presents a summary of the main intertemporal relationships
between the variables. This illustration shows how employment growth
is strongly associated with subsequent growth of profits, and to a lesser
extent, with the subsequent growth of sales. Labour productivity growth,
and also sales growth, are associated with subsequent growth of profits.
Although many other minor relationships between variables exist, they are
much smaller in magnitude and so they are not shown in the diagram. In
contrast to several theoretical contributions on the theme of firm growth,
Empl. growth Empl. growth
Sales growth Sales growth
Profits growth Profits growth
Prod. growth Prod. growth
t t+1
time
Source: Coad (2007b).
Figure 5.2 A stylized depiction of the process of firm growth, based on

1-lag VAR regression results. The thickest lines correspond to
‘major’ associations (coefficients of magnitude $0.15), while
the thinner lines correspond to relatively ‘minor’ associations
(coefficients of magnitude $0.10). Coefficients lower than
0.05 in magnitude, as well as autocorrelation coefficients, are
not represented here
we do not observe much in the way of feedback from growth of profits to

subsequent growth of employment or sales. This is made clear in the fol-
lowing numerical example: the results suggest that if we were to observe
an increase in the employment growth rate of 1 percentage point, then
ceteris paribus we can expect growth of profits to rise by about 0.2 percent-
age points in the following year. On the other hand, a 1 percentage point
increase in growth of profits can be expected to be followed by a −0.002
percentage point increase in employment growth.
It is also interesting to observe that employment growth has less of an
effect on subsequent productivity growth for larger firms, which is consist-
ent with the idea that small firms have to struggle to reach the industry
minimum efficient scale (MES), and until they reach the MES, increases in
employment will be associated with increases in productivity.
Coad and Broekel (2007) undertake a similar analysis with an alternative

indicator of firm-level productivity, using non-parametric frontier analy-
sis to estimate a multifactor productivity score for each firm year. This
multifactor productivity indicator is an alternative to the simpler labour
productivity measure. Their findings are similar to those presented in Coad
(2007b), although they find evidence of a negative relationship between
employment growth and subsequent growth of multifactor productivity.
Coad et al. (2008) also obtain comparable results in their analysis of Italian
manufacturing firms.
Profits, Growth and R&D Investment
Coad and Rao (2009) investigate the processes of firm-level investment in

research and development expenditures (R&D), using data on large listed
US firms. Their VAR model focuses on growth of employment, growth
of sales, growth of (absolute) profits and growth of (absolute) R&D
expenditure.
Figure 5.3 shows a concise representation of the results. Of particular
interest is the finding that employment growth and sales growth are fol-
lowed by growth of R&D expenditure, while growth of profits has little
discernible effect on the subsequent growth of R&D. Coad and Rao (2009)
also investigate asymmetric effects for growing and shrinking firms, using
quantile regression techniques. This seems appropriate because R&D is rel-
atively ‘sticky’, and R&D levels cannot easily be started or stopped – and so
plans to decrease R&D may not be as easy as plans to pursue or expand upon
an existing R&D project. They conclude that firms behave ‘as if’ they follow
two behavioural rules concerning R&D investment. First: if employment
or sales have grown recently, aim to keep R&D levels at a roughly constant
ratio with respect to these two firm size indicators. Second: if the firm has
decreased in size, try to decrease R&D expenditure by as little as possible. In
contrast to many traditional conjectures about R&D investment, firms do
not appear to reinvest much of their profits into R&D.
5.4 CONCLUSION
The relationship between firm performance (whether it be profitability or

productivity) and firm growth has received a lot of attention from eco-
nomic theory. This relationship has important implications for the alloca-
tion of scarce resources between heterogeneous firms in an industry (Baily
and Farrell, 2006). In an ideal world, we suggest, scarce resources would
be redirected to the more efficient firms.
Empl. growth Empl. growth
Sales growth Sales growth
Profits growth Profits growth
R&D growth R&D growth
t t+1
time
Source: Coad and Rao (2009).
Figure 5.3 A stylized depiction of the process of firm growth and R&D
investment, based on 1-lag VAR regression results. The
thickest line corresponds to the most significant association
(i.e. a coefficient of $0.3), while the other thinner lines
correspond to relatively ‘minor’ associations (coefficients of
magnitude $0.10). Coefficients lower than 0.05 in magnitude,
as well as autocorrelation coefficients, are not represented
Theoretical presumptions of a positive influence of relative performance

on firm growth are not firmly grounded, however. Empirical work seems
to offer only limited support: productivity and profitability are not major
determinants of firm growth.
We need to rethink our preconceptions about what firms actually do
with their profits. We cannot take it for granted that efficient firms will
reinvest their profits in further expansion. Neither should we presume that
there is a ‘virtuous’ mechanism in place whereby firms reinvest their profits
into R&D projects (for example Scherer, 2001). The evidence suggests that,
once made, profits are not reinvested into firm growth to any great extent.
One might consider that profits need not lead to an increase in the scale of
operations any more than, say, a higher calorific content (especially among
adults) should lead to an increase in height.
Much work remains to be done on the topics raised in this chapter.

We speculate that cohort studies may be a particularly fruitful avenue of
research, because the availability of finance, and also of growth opportu-
nities, vary considerably over a firm’s lifetime. Small firms generally have
little cash but lots of growth potential, whereas larger firms may have fewer
growth prospects but more funds. It may be that comparing firms at a
similar stage in their life cycle might reveal more of a relationship between
relative performance and growth.
6. Innovation and firm growth
Innovation plays an increasingly important role in our modern economy,
transforming it from within, and bringing about a tremendous amount
of structural change and turbulence (Metcalfe, 1998). New sectors are
born, new products and techniques replace their older counterparts, and
some firms can harness the power of their innovations to experience
spectacular growth, while less innovative firms appear to wither away
and perish.
The influence of innovation on firm growth has been of great interest
to both theoretical and empirical scholars. However, as we will see in the
rest of this chapter, the strong predictions emerging from theoretical work
cannot easily be reconciled with the available empirical evidence.
When discussing the relationship between innovation and firm
growth, however, it is meaningful to distinguish between employment
growth and sales growth. Employment growth is an input, while sales
growth is an output. Innovation, it is anticipated, can lead to the produc-
tion of a higher level of output through a more efficient use of inputs.
Are firms becoming capable of producing more by hiring fewer workers?
Perhaps worse, are workers being made unemployed, being replaced
by machines? Firms and strategists are usually more concerned about
the impact of innovation on sales growth or growth of profits, while
economists and policy makers are more concerned about employment
growth. Given the distinction between these two indicators, therefore,
we will discuss them separately. Innovation and sales growth is discussed
in section 6.1, while the relation between innovation and employment
growth (also known as the ‘technological unemployment’ literature) is
treated in section 6.2.
This is not the place to consider how innovative activity affects other
aspects of firm performance such as stock market success. For a survey
of the literature on innovation and market value, see the surveys in Hall
(2000) and Coad and Rao (2006). Furthermore, we don’t consider here
the impact of firm growth on increases in R&D investment – the reader is
instead referred to section 5.3 in the previous chapter.
76
Innovation and firm growth 77
6.1 INNOVATION AND SALES GROWTH

The relationship between innovation and sales growth can be described as
something of a paradox – on the one hand, a broad range of theoretical and
descriptive accounts of firm growth stress the important role innovation
plays for firms wishing to expand their market share. For example, Carden
(2005, p. 25) presents the main results of the McKinsey Global Survey
of Business Executives, and writes that ‘[e]xecutives overwhelmingly say
that innovation is what their companies need most for growth.’ Another
survey focusing on SMEs reports that investment in product innovation
is the single most popular strategy for expansion, a finding which holds
across various industries (Hay and Kamshad, 1994). Economic theorizing
also recognizes the centrality of innovation in growth of firm sales (see for
example the discussion in Geroski, 2000, 2005, or the theoretical models
in Nelson and Winter, 1982; Aghion and Howitt, 1992; and Klette and
Griliches, 2000).1 On the other hand, empirical studies have had difficulty
in identifying any strong link between innovation and sales growth, and
the results have often been something of a let-down. Indeed, some studies
fail to find any influence of innovation on sales growth at all. Commenting
on the current state of our understanding of firm-level processes of innova-
tion, Cefis and Orsenigo (2001) write: ‘Linking more explicitly the evidence
on the patterns of innovation with what is known about firms’ growth and
other aspects of corporate performance – both at the empirical and at the
theoretical level – is a hard but urgent challenge for future research’ (Cefis
and Orsenigo, 2001, p. 1157).
A major difficulty in observing the effect of innovation on growth is that
it may take a firm a long time to convert increases in economically valuable
knowledge (that is, innovation) into economic performance. Even after
an important discovery has been made, a firm will typically have to invest
heavily in product development. In addition, converting a product idea
into a set of successful manufacturing procedures and routines may also
prove costly and difficult. Furthermore, even after an important discovery
has been patented, a firm in an uncertain market environment may prefer
to treat the patent as a ‘real option’ and delay associated investment and
development costs (Bloom and Van Reenen, 2002). There may therefore
be considerable lags between the time of discovery of a valuable innova-
tion and its conversion into commercial success.2 Another feature of the
innovation process is that there is uncertainty at every stage, and the
overall outcome requires success at each step of the process. In a pioneer-
ing empirical study, Mansfield et al. (1977) identify three different stages of
innovation that correspond to three different conditional probabilities of
success: the probability that a project’s technical goals will be met (x); the
probability that, given technical success, the resulting product or process

will be commercialized (y); and finally the probability that, given commer-
cialization, the project yields a satisfactory return on investment (z). The
overall success of the innovative activities will be the product of these three
conditional probabilities (x 3 y 3 z). If a firm fails at any of these stages,
it will have incurred costs without reaping benefits. We therefore expect
that firms differ greatly both in terms of the returns to R&D (measured
here in terms of post-innovation sales growth) and also in terms of the time
required to convert an innovation into commercial success. However, it is
anticipated that innovations will indeed pay off on average and in the long
term, otherwise commercial businesses would obviously have no incentive
to perform R&D in the first place.
A number of empirical investigations have been undertaken to investi-
gate the relationship between innovation and sales growth. Our gleaning
of this literature yields a rather motley harvest. (This may be due to dif-
ficulties in linking firm-level innovation data to other firm characteristics.)
Mansfield (1962) considers the steel and petroleum sectors over a 40-year
period, and finds that successful innovators grew more quickly, especially
if they were initially small. Moreover, he asserts that the higher growth
rate cannot be attributed to their pre-innovation behaviour. Another
early study by Scherer (1965) looks at 365 of the largest US corpora-
tions and observes that inventions (measured by patents) have a positive
effect on company profits via sales growth. Furthermore, he observes that
innovations typically do not increase profit margins but instead increase
corporate profits via increased sales at constant profit margins. Mowery
(1983) focuses on the dynamics of US manufacturing over the period
1921–46 and observes that R&D employment only has a significantly
positive impact on firm growth (in terms of assets) for the period 1933–46.
Using two different samples, he observes that R&D has a similar effect
on growth for both large and small firms. Geroski and Machin (1992)
look at 539 large quoted UK firms over the period 1972–83, of which 98
produced an innovation during the period considered. They observe that
innovating firms (that is firms that produced at least one ‘major’ innova-
tion) are both more profitable and grow faster than non-innovators. Their
results suggest that the influence of specific innovations on sales growth
are nonetheless short-lived (p. 81) – ‘the full effects of innovation on cor-
porate growth are realized very soon after an innovation is introduced,
generating a short, sharp one-off increase in sales turnover.’ In addition,
and contrary to Scherer’s findings, they observe that innovation has a
more noticeable influence on profit margins than on sales growth. Geroski
and Toker (1996) look at 209 leading UK firms and observe that innova-
tion has a significant positive effect on sales growth when included in an
OLS regression model amongst many other explanatory variables. Roper

(1997) uses survey data on 2721 small businesses in the UK, Ireland and
Germany to show that innovative products introduced by firms made a
positive contribution to sales growth. Freel (2000) considers 228 small
UK manufacturing businesses and, interestingly enough, observes that
although it is not necessarily true that ‘innovators are more likely to grow’,
nevertheless ‘innovators are likely to grow more’ (that is they are more
likely to experience particularly rapid growth). Corsino (2008) investi-
gates the relationship between product innovations and sales growth both
at the firm level and also at the level of its constituent lines of business.
Although he detects a small positive relationship between innovation and
sales growth at the firm level, the magnitude of this effect increases slightly
when the analysis is performed at the level of individual lines of business.
Finally, Del Monte and Papagni (2003) report a positive relationship
between R&D activity and sales growth in their analysis of Italian manu-
facturing firms, although they also refer to previous research on Italian
data that did not find evidence of any relationship.
Not all investigations were able to find a positive relation between inno-
vation and subsequent performance. Geroski et al. (1997) fail to observe
any relationship between ‘major innovations’ and number of patents on
sales growth in their sample of 271 large listed UK firms. Bottazzi et al.
(2001) study the dynamics of the worldwide pharmaceutical sector and do
not find any significant contribution of a firm’s ‘technological ID’ or inno-
vative position to sales growth.3 Freel and Robson (2004) actually observe
a negative relationship between product innovation and the sales growth
of manufacturing firms, in their sample of small businesses in Scotland and
Northern England.
One observation that emerges from the preceding survey is that inno-
vation can be measured in several ways, although the most common
approach is to use R&D statistics or patent counts. However, each of these
indicators has its drawbacks. R&D statistics are typically quite smoothed
over time, which contrasts with the lack of persistence frequently observed
in patent statistics. Furthermore, R&D expenditure is an innovative input
and it gives only a poor indication of the value of the resulting innovative
output that a firm can take to market. Patent statistics are very skewed in
value, with many patents being practically worthless whilst a fraction of
patents generate the lion’s share of the economic value. Another limita-
tion is that many previous studies have lumped together firms from all
manufacturing sectors – even though innovation regimes (and indeed
appropriability regimes) vary dramatically across industries.4 To deal with
these difficulties of quantifying firm-level innovative activity, the analysis
in Coad and Rao (2008) combines information on a firm’s recent history of
R&D expenditures as well as patenting activity to create a synthetic ‘inno-

vativeness’5 variable for each firm year. In this way we extract the common
variance associated with each of these indicators while discarding the idi-
osyncratic noise and measurement error. We also focus on four two-digit
‘complex technology’ manufacturing industries that were hand-picked
because of their relatively high intensities.
Using semi-parametric quantile regressions, Coad and Rao (2008)
explore the influence of innovation at a range of points of the conditional
growth rate distribution. Their results indicate that most firms don’t grow
very much, and their growth is hardly related to their attempts at innova-
tion. Nevertheless, innovation is seen to be of critical importance for a
handful of fast-growth firms. This emphasizes the inherent uncertainty in
firm-level innovative activity – whilst for the ‘average firm’ innovativeness
may not be very important for sales growth, innovativeness is of crucial
importance for the ‘superstar’ high-growth firms. Standard regression
techniques which implicitly give equal weights to both high-growth and
low-growth firms, and that yield a summary point estimate for the ‘average
firm’, are unable to detect this relationship. Similar results were observed
by Goedhuys and Sleuwaegen (2008) in their analysis of manufacturing
firms in 11 African countries. For most firms, product innovation has no
significant effect on growth, while it has a strong positive effect on growth
for the fastest-growing firms.
In other cases innovation can be seen to have a negative impact on firm
performance. This can be linked to the costly and uncertain nature of
innovation. In the pharmaceutical sector, for instance, Grabowski et al.
(2002) observe a skewed distribution of returns to R&D, with the mean
industry internal rate of return only slightly in excess of the cost of capital.
Although some fortunate firms can earn huge profits from blockbuster
drugs, only the top one-third of new drug introductions over the period
1990–94 have positive net present values, with the median drug having
a net present value that is below R&D costs. In the quantile regression
analysis in Coad and Rao (2008) and Goedhuys and Sleuwaegen (2008),
innovative activity is observed to have a negative association with growth
for those firms at the lowest quantiles of the growth rates distribution. In
other words, among firms facing rapid decline, attempts at innovation can
aggravate this decline. While innovation can propel some firms into fast
growth, it should also be recognized that failed attempts at innovation
can leave the firm worse off than if it had not attempted innovation in the
first place. Freel comments on this phenomenon in the following words:
‘firms whose efforts at innovation fail are more likely to perform poorly
than those that make no attempt to innovate. To restate, it may be more
appropriate to consider three innovation derived sub-classifications – i.e.
“tried and succeeded”, “tried and failed”, and “not tried”’ (Freel, 2000, p.
208; see also Freel and Robson, 2004).
6.2 INNOVATION AND EMPLOYMENT GROWTH

Whilst firm-level innnovation can be expected to have a positive influ-
ence on sales growth, the overall effect on employment growth is a priori
ambiguous. Innovation is often associated with increases in productivity
that lower the amount of labour required for the production of goods and
services. In this way, an innovating firm may change the composition of
its productive resources, to the profit of machines and at the expense of
employment. As a result, the general public has often expressed concern
that technological progress may bring about the ‘end of work’ by replac-
ing manpower with machinery. Economists, on the other hand, are usually
more optimistic.
To begin with, theoretical discussions have found it useful to decompose
innovation into product and process innovation. Product innovations are
often associated with employment gain, because the new products create
new demand (although it is possible that they might replace existing prod-
ucts). Process innovations, on the other hand, often increase productivity by
reducing the labour requirement in manufacturing processes (for example
via the introduction of robots, Fleck, 1984). Thus, process innovations are
often suspected of bringing about ‘technological unemployment’.
The issue becomes even more complicated, however, when we consider
that there are not only direct effects of innovation on employment, but
also a great many indirect effects operating through various ‘substitution
channels’. For example, the introduction of a labour-saving production
process may lead to an immediate and localized reduction in employees
inside the plant (the ‘direct effect’), but it may lead to positive employment
changes elsewhere in the economy because of the indirect effects. Six such
indirect ‘substitution channels’ can be mentioned here (following on from
Spiezia and Vivarelli, 2000). First, there may be an increased demand for
new machines, which leads to the creation of jobs in upstream capital
goods sectors. Second, a labour-saving innovation may lead to a reduc-
tion in prices, and as a consequence there may be an increase in demand,
production and employment. Third, a reduction in costs brought about
by technological progress may lead some firms to make new investments.
Fourth, compensation may occur via a decrease in wages, which will then
encourage the adoption of more labour-intensive techniques of produc-
tion. Fifth, an increase in incomes may translate into higher consumption
and thus higher employment. Sixth, there may be compensation via new
products if new goods are developed and are associated with the creation
of new jobs. As a result, the direct labour-saving effects of innovation need
to be weighed up against the indirect effects, and so the overall effect of
innovation on employment needs to be investigated empirically. Although
Van Reenen recently lamented the ‘dearth of microeconometric studies on
the effect of innovation on employment’ (Van Reenen, 1997, p. 256), the
situation has improved over the last decade.
Research into technological unemployment has been undertaken in dif-
ferent ways. As a consequence, the results emerging from different studies
are far from harmonious – ‘[e]mpirical work on the effect of innovations on
employment growth yields very mixed results’ (Niefert, 2005, p. 9). Doms
et al. (1995) analyse survey data on US manufacturing establishments, and
observe that the use of advanced manufacturing technology (which would
correspond to process innovation) has a positive effect on employment. At
the firm level of analysis, Hall (1987) observes that employment growth is
related positively and significantly to R&D intensity in the case of large
US manufacturing firms. Coad and Rao (2007) also observe a positive and
significant influence of innovation on employment growth in their analysis
of four high-tech US manufacturing industries. In addition, they observe
that when weights are given to firms according to their size, the coefficient
increases in magnitude – which suggests that larger firms, who have the
power to create or destroy a larger absolute number of jobs, are more likely
to convert innovative activity into employment gain. Greenhalgh et al.
(2001) observe that R&D intensity and also the number of patent publica-
tions have a positive effect on employment for British firms. Nevertheless,
Evangelista and Savona (2002, 2003) observe a negative overall effect of
innovation on employment in the Italian services sector. When the distinc-
tion is made between product and process innovation, the former is usually
linked to employment creation whereas the consequences of the latter are
not as clear-cut. Evidence presented in Brouwer et al. (1993) reveals a small
positive employment effect of product-related R&D although the com-
bined effect of innovation is imprecisely defined. Relatedly, work by Van
Reenen (1997) on listed UK manufacturing firms and Smolny (1998) for
West German manufacturing firms show a positive effect on employment
for product innovations. Smolny also finds a positive employment effect
of process innovations, whereas Van Reenen’s analysis yields insignificant
results. Harrison et al. (2005) consider the relationship between innovation
and employment growth in four European countries (France, Italy, the
UK and Germany) using data for 1998 and 2000 on firms in the manufac-
turing and services industries. Whilst product innovations are consistently
associated with employment growth, process innovation appears to have a
negative effect on employment, although the authors acknowledge that this
latter result may be attenuated (or even reversed) through compensation

effects. To summarize, therefore, we can consider that product innovations
generally have a positive impact on employment, whilst the role of process
innovations is more ambiguous (Hall et al., 2008).
6.3 CONCLUSION
Successful innovation enables firms to become more productive, generat-

ing an increase in output while lowering the requirements of inputs. When
one is searching for the influence of innovation on firm growth, then, it is
useful to distinguish between growth of employment (where employment is
one of the firm’s inputs) and growth of sales (which is an output).
We began by looking at the relationship between innovation and sales
growth (section 6.1). Theoretical work, as well as questionnaire evidence,
have both suggested a strong link between innovation and firm growth.
Empirical work on the matter is not scarce, and this body of work gener-
ally is able to detect a positive effect of innovation on growth, although the
magnitude of this effect is not very large. One explanation for this result
is that, while innovation is not very important in explaining the growth of
the average firm (which doesn’t grow very much), innovation is of crucial
importance for a small number of fast-growing firms.
We then considered the role that innovation plays concerning employ-
ment growth (section 6.2). In this context, it is helpful to distinguish
between product innovation and process innovation. Product innovation
tends to be positively associated with employment growth, while the effect
of process innovation is less clear and may even be negatively associ-
ated. There are many substitution channels, however, through which the
economy can adjust to innovation and productivity growth by relocating
employees to new jobs.
7. Other determinants of firm
growth
In the previous sections we investigated some of the determinants of firm
growth rates. The relationship between firm size and growth rate was inves-
tigated in Chapter 4. In Chapter 5, we considered the associations between
profits and productivity and firm growth. In Chapter 6, we focused on the
role of innovation in explaining firm growth rates.
In this chapter we look at other factors that have been shown to exert
an influence on growth rates. At the end of this chapter, however, we
take stock of the findings of the last four chapters and acknowledge that,
although there are some factors that are significantly associated with firm
growth rates, that firm growth has a preponderant random aspect. The
combined explanatory power of explanatory variables on firm growth is
nonetheless rather modest.
7.1 AGE
The relationship between size and growth has received a great deal of atten-
tion in empirical work, as we discussed above in section 4.2. Relatedly,
the relationship between a firm’s age and its growth rate has also been
frequently investigated. Age and size are certainly closely related, and in
some cases they are both taken to represent what is essentially the same
phenomenon (see for example the model in Greiner, 1972).
Older firms are often assumed to be more inert and less capable of
adapting to a changing environment. Older firms may lack the drive and
entrepreneurial spark that is required to observe new business opportu-
nities and then build upon them. The routinized nature of production
in firms may lead a firm to stick to what it knows best and become
increasingly distanced from external developments. Inertia can also
accumulate as people in the organization become entrenched in their
positions and resist change. Nonetheless, it is also conceivable that age
can be an advantage for small, young firms: since older firms are more
experienced, they have an established reputation and a proven track
record that confer more credibility in factor markets. This latter point
84
Other determinants of firm growth 85
notwithstanding, age is often observed to exert a negative influence on

firm growth.
One of the earliest investigations of the influence of age on growth
was made by Fizaine (1968), who examined the growth of establishments
from the French region of Bouches-du-Rhône. She observed that age
has a negative effect on the growth of establishments, and also that the
variance of growth rates decreases with age. Fizaine (1968) also argued
that the correct causality runs from age to growth, rather than from size
to growth as supposed by many investigations into firm growth based
on Gibrat’s law (this argument was subsequently reiterated by Evans,
1987a). Dunne et al. (1989) analyse US establishments and concur with
Fizaine’s findings that both the expected growth rate and also the growth
variance decrease with age. Age is also observed to have a negative effect
on growth at the firm level, as a large number of studies have testified –
see inter alia Evans (1987a,b) for US manufacturing firms, Variyam and
Kraybill (1992) for US manufacturing and services firms, Robson and
Bennett (2000) for UK SMEs, Liu et al. (1999) for Taiwanese electronics
plants, Goedhuys and Sleuwaegen (2000) and Sleuwaegen and Goedhuys
(2002) for Ivorian manufacturing firms, Reichstein and Dahl (2004) for
Danish limited liability companies, Geroski and Gugler (2004) for large
European companies, and Yasuda (2005) for Japanese manufacturing
firms.
Generally speaking, then, the negative dependence of growth rate on
age appears to be a robust feature of industrial dynamics. This is not
always observed, however. Two studies of firm dynamics in India (Das,
1995; Shanmugam and Bhaduri, 2002) apply panel data techniques to the
data and observe that, although larger firms have lower growth rates, age
is significantly positively related to firm growth. Das (1995) examines the
growth of firms in a young, fast-growing computer hardware industry in
India, while Shanmugam and Bhaduri (2002) focus on a sample of 392
Indian manufacturing firms.
It has also been suggested that the relationship between age and growth
may vary over the age distribution. Evidence presented in Barron et al.
(1994) would support this hypothesis. Barron et al. (1994) analyse the
growth of New York Credit Unions and observe that older firms grow
faster than adolescent firms, although it is the very young firms that experi-
ence the fastest growth. Similarly, Bigsten and Gebreeyesus (2007) analyse
Ethiopian census data on manufacturing firms with over 10 employees,
and observe that ‘[g]rowth and age are inversely related only in the first few
years after entry and stay constant for most of the age group until it starts
to have a positive relation beyond age 50’ (p. 831).
7.2 COMPETITION BETWEEN FIRMS

The concept of competitive struggle between firms has long been discussed in
the industrial organization literature. Interest in inter-firm rivalry received
new impetus, however, with the development of the analytical apparatus of
game theory. Game theory is, of course, to a great extent concerned with
situations involving two players whose final pay-off depends on the choice
of the other player. The analysis of inter-firm competition seemed to be a
natural candidate scenario in which to apply these new theoretical ideas.
The diffusion of game theory into the analysis of inter-firm relations has
thus had the effect of taking the conception of interactions between rival
firms to new extremes – zero-sum games between two players, where one
player’s gain is equal to the other player’s loss. Game-theoretic conceptions
of firm behaviour have indeed become prominent.1
By way of an illustration, let us consider the case of the reaction of
incumbents to the entry of new firms. The game-theoretic literature frames
this situation as a one-on-one strategic game whereby entrants take market
share from incumbents, and incumbents make strategic investments in
capacity to deter potential entrants from entering (see amongst others the
influential work of Salop, 1979 and Dixit, 1980). This vision of the rela-
tionship between incumbents and entrants is far from realistic, however. In
reality, entrants are often far too small to be of any threat, their growth is
too slow, their exit hazard too high, and their entry into the market is too
erratic. Furthermore, in the unlikely event that incumbents are genuinely
concerned about defending their market share from entrants, the available
evidence suggests that they are unlikely to do so by investing in additional
capacity, but rather through the use of strategies such as advertising or
licensing deals (Geroski, 1995). Small and large firms operate in different
‘strategic groups’ (Caves and Porter, 1977) and so they do not engage in
direct competition (Audretsch et al., 1999). Furthermore, it seems that
small firms grow first and foremost through increases in demand in their
market niches rather than by struggling against competitors over market
share in existing markets (Wiklund, 2007). As such, some predictions
emerging from the theoretical literature seem to be quite irrelevant to the
actual workings of the economy.
Interest in inter-firm rivalry has also appeared in several other types of
theoretical model. Bottazzi and Secchi (2006a) explain the heavy-tailed
distribution of firm growth rates by suggesting a model whereby firms
fight each other to obtain growth opportunities. Similarly, McKelvey and
Andriani (2005) suggest that this heavy-tailed distribution of firm growth
rates emerges from interdependencies between firms. (The model in section
3.2.2, however, shows how the heavy-tailed nature of the growth rate
distribution can be explained by focusing on the internal structure and

dynamics of one isolated firm.)
It would appear that the emphasis placed on competition by theoretical
work has not been duly investigated in empirical work. The small body
of such research has usually measured competition in terms of industry
concentration, or rents obtained by incumbents, which are poor indicators
of actual inter-firm competition (Boone et al., 2007).2 Geroski and Gugler
(2004) consider the impact of the growth of rival firms on a firm’s employ-
ment growth, using a database on several thousand of the largest firms in
14 European countries. Rival firms are defined as other firms in the same
3-digit industry. In their main regression results (Table 2) they are unable
to detect any significant effect of rival’s growth on firm growth, although
they do find a significant negative effect in specific industries (that is dif-
ferentiated good industries and advertising intensive industries). Recent
headway has also been made by Sutton (2007), who analyses the dynamics
of market shares of leading Japanese firms. Sutton uses simple statistical
techniques to show that the changes in market share of the first and second
largest firms in any industry are, in all but a few exceptional circumstances,3
statistically independent. These few empirical investigations into inter-firm
competition seem to suggest that market share dynamics are best described
by firm independence, rather than firm interdependence.
If inter-firm competition for growth opportunities exists, where can sta-
tistical evidence for it be found? The work surveyed above looked at firm
growth rates for rival firms in related industries, and found that in most
cases these firms act independently. Further work is needed to complement
these lonely studies. More detailed analysis would be especially valuable
if it is able to observe inter-firm competition at the level of disaggregated
business units, or competition in narrowly-defined geographical areas.
For the time being, however, it might be preferable to consider the growth
of firms to be independent of the growth of other firms – in spite of theoreti-
cal models that view the growth of one firm to occur to the detriment of the
growth of its rivals. Schumpeter might have had this in mind when he wrote
the following line: ‘The business man feels himself to be in a competitive
situation even if he is alone in his field’ (Schumpeter, 1942, p. 85).4 Inter-
firm competition effects are definitely in need of more work. While the
evidence in favour of inter-firm competition may be easier to detect when
other indicators of firm performance are used (such as profitability – see
for example Wiggins and Ruefli, 2005), there does not seem to be any trace
of inter-firm competition when firm growth rates are considered. Amid the
present fog we may hazard the following conjecture. Although we often
hear about the competitive struggle that growing firms face, this may refer
to the struggle to improve, increase and expand themselves rather than a
fight or struggle against any particular rival. Competition, we suggest, is

against an imagined shadow rather than against any physical rival; it is a
fight against slack, a struggle to improve oneself; it is a war on inefficiency.
It is also a struggle to create growth opportunities where they may not have
existed before. Given that consumers tend to get locked in to ‘sticky’ con-
sumption habits that display remarkable persistence over time (Fishman
and Rob, 2003), competition between firms may take the shape of creating
new products and niches, rather than fighting with other firms over existing
market share. As such, the growth of one firm need not manifest itself as a
loss of market share for a ‘rival’ firm.
7.3 CHARACTERISTICS OF THE ENTREPRENEUR
In this section we look at the influence of the entrepreneur’s human capital

and sex on the growth of the founder’s enterprise.
Human Capital
The human capital embodied in the proprietor has long been suspected
of having an effect on firm growth. Education may be instrumental for
entrepreneurs to carry out their growth aspirations (Wiklund, 2007), or
indeed it may help entrepreneurs nurture such aspirations in the first place.
It may also make the founders more aware of their business environment
and the opportunities available to them, or it may put entrepreneurs in
a better standing vis-à-vis lenders (Robson and Obeng, 2008) or other
stakeholders.
Almus (2002) identifies a positive effect of human capital (that is, univer-
sity degree or above) on growth for fast-growing German firms. Robson
and Bennett (2000), however, fail to find a significant effect of skill level
in explaining employment or profitability growth in their sample of UK
small businesses.
Research on firm growth undertaken in developing countries has also
shown an interest in the human capital of the founding entrepreneur.
McPherson (1996) investigates small firms in five southern African nations
and observes that the level of human capital embodied in the proprietor
has a positive and significant influence on the growth of micro and small
businesses in five Southern African nations. Robson and Obeng (2008)
focus on small businesses in Ghana and report that better educated found-
ers faced fewer obstacles to expansion. They observe that education had a
stronger effect than age or sex on the importance of barriers facing growing
small businesses. Mead and Liedholm’s (1998) survey of the evidence
for developing countries provides further support to the hypothesis that

an entrepreneur’s education tends to have a positive influence on firm
growth.
Wiklund and Shepherd (2003) take the investigations a step further by
considering the roles of education and experience of the founding entre-
preneur alongside the entrepreneur’s growth aspirations. Although they
observe that previous studies have found a small positive relationship
between growth aspiration and actual growth, it is the interaction between
growth aspirations, on the one hand, and education and experience, on
the other, that makes the most significant contribution to growth. In other
words, education and experience magnify any positive effect on growth
that an entrepreneur’s growth aspirations may have, because education
and experience enable the entrepreneur to realize his or her growth plans.
Sex
The sex of the founding entrepreneur has also been linked to firm growth,
and the main finding seems to be that businesses headed by female entrepre-
neurs experience slower growth. This broad relationship has received much
attention in samples from developed countries (see Catley and Hamilton,
1998 for a survey) and has also been found among less developed countries
such as Indonesia (Singh et al., 2001), India (Coad and Tamvada, 2008),
and South Africa, Swaziland and Botswana (McPherson, 1996). Although
women entrepreneurs tend to be found in lower growth industries, the
lower growth of these entrepreneurs still remains after controlling for
industry (Mead and Liedholm, 1998).
Why is it that businesses led by female entrepreneurs tend to grow at
slower rates? One reason for this could be that females are less ambitious
than their male counterparts. Questionnaire evidence from Ghana suggests
that while women entrepreneurs encounter barriers preventing them from
starting their business, these ventures do not face higher barriers to growth
once they are already in business (Robson and Obeng, 2008). If established
female businesses do not encounter higher barriers to growth, then their
lower growth rates may well be due to less ambitious attitudes towards
growth. Another possible explanation is that females are more sensitive to
both domestic and professional responsibilities, and that since they rely on
their business for their household income, they will not seek to take risks
with their business and may be more willing to pass up business opportuni-
ties if these are perceived as too risky (Mead and Liedholm, 1998). There
may be still other differences between male and female entrepreneurs.
For example, while female entrepreneurs have strong skills in dealing
with people, their financial skills would appear to be weak in comparison
(Catley and Hamilton, 1998; Hisrich and Ozturk, 1999). Female entrepre-
neurs seem to be less interested in making money (Cromie, 1987).
7.4 OTHER FIRM-SPECIFIC FACTORS

A number of other firm-specific variables have been associated with growth
rates. Ownership structure appears to be a relevant factor because there is
evidence that multi-plant firms have higher growth rates, on average, than
single-plant firms. This appears to be the case for US small businesses
(Variyam and Kraybill, 1992; Audretsch and Mahmood, 1994), large
European corporations (Geroski and Gugler, 2004), and also manufactur-
ing firms in Italy (Fagiolo and Luzzi, 2006) and France (Coad, 2008b). In
their analysis of West German firms, Harhoff et al. (1998) identify that
subsidiary firms grow faster than non-subsidiaries in construction and
trade industries, although no difference can be found for manufactur-
ing and services. Furthermore, a plant-level analysis reveals that plants
which belong to large companies are observed to have higher growth than
stand-alone plants (Dunne et al., 1989). Whilst there is weak evidence that
foreign-owned firms experience faster growth rates, government-owned
firms seem to grow more slowly (Beck et al., 2005b). Unionization has also
been investigated as a factor affecting firm growth, but it appears that the
unionization status of a firm has no impact on its growth.5
A firm’s legal status is also proposed as a determinant of its growth
rate (Harhoff et al., 1998; Storey, 1994). Harhoff et al. (1998) examine the
growth of West German firms and observe that firms with limited liability
have significantly higher growth rates in comparison to other companies.
However, these firms also have significantly higher exit hazards. These
results are in line with theoretical contributions, along the lines of Stiglitz
and Weiss (1981), that emphasize that the limited liability legal form pro-
vides incentives for managers to pursue projects that are characterized by
both a relatively high expected return and a relatively high risk of failure.
Capital intensity is another factor whose influence on firm growth has
been explored. Sleuwaegen and Goedhuys (2002) observe that capital-
intensive firms (defined as firms with more intensive use of electricity,
fuel, water and telephone services) grew significantly faster than others, in
their sample of manufacturing firms from the Côte d’Ivoire. On the other
hand, Liu et al. (1999) found a positive but statistically insignificant effect
of the capital–labour ratio on the growth of electronics firms in Taiwan.
Rossi-Hansberg and Wright (2007) perform some preliminary compari-
sons of capital-intensive and labour-intensive sectors, and, consistent with
their model, they find some evidence of different regimes of growth across
sectors. Capital-intensive sectors seem to have a stronger scale depend-

ence, such that smaller firms grow much faster than larger ones. In labour-
intensive industries, in contrast, there is a smaller gap between the growth
of small and large firms.
Another approach has been to consider the characteristics of the man-
agement. The ‘managerial’ theory (surveyed in section 8.3) suggests that
managers attach utility to the size and growth of their firms, such that they
will pursue growth above the shareholder-value-maximizing level. This
leads to the hypothesis that owner-controlled firms will have lower growth
rates (and perhaps higher profits) than manager-controlled firms. Whilst
Radice (1971) and Holl (1975) find no support for this claim in their analy-
ses of large UK firms, Hay and Kamshad (1994) find that owner-controlled
SMEs have lower growth rates than non-owner-controlled SMEs.
It has also been shown that characteristics relating to the nature of the
firm’s activity have an influence on firm growth. The level of diversifica-
tion appears to have a negative overall association with the growth of large
European corporations (Geroski and Gugler, 2004), although a positive
and significant influence can be detected in the particular cases of advertis-
ing-intensive industries (Geroski and Gugler, 2004) and the life insurance
industry (Hardwick and Adams, 2002). (The negative association between
diversification and firm growth could be due to diversified firms being
more likely to operate in low-growth niches or stagnating sub-sectors,
however.) Advertising intensity is another factor that is associated with
sales growth, according to Geroski and Toker’s (1996) analysis of leading
UK firms. In addition, whilst previous firm-level analyses have mainly
associated exporting activity with increases in productivity, some authors
have identified a positive relationship between exports and firm growth
(Robson and Bennett, 2000; Beck et al., 2005b). The degree of centrality, or
the amount of experience in a network of firms also contributes to a firm’s
(employment) growth rate, according to Powell et al. (1996). It has also
been observed that firms with inter-firm partnership arrangements with
members of their supply chain tend to grow faster and experience sustained
growth (Wynarczyk and Watson, 2005).
Threshold effects of various kinds are also thought to dampen the growth
of firms. In the past, when antitrust legislation was relatively obsessed with
firm size per se, large firms sought to limit their growth to avoid antitrust
intervention. Furthermore, large firms may be reluctant to implement a
strategy of rapid growth (and especially forward integration) because of
the threat of a reaction from competitors (see for example Penrose’s (1960)
biography of the Hercules powder company). In developed countries, there
is often a size threshold above which firms face a sudden increase in firing
costs. As a result, there may be a slight self-imposed restriction on growth
for small firms whose size is close to this threshold. This usually affects
firms whose size is somewhere in the range of 8–15 employees, depending
upon the country (see Schivardi and Torrini, 2008). In developing coun-
tries, firms can avoid or evade taxes by remaining small and informal.
Larger firms, on the other hand, can effectively lobby governments to
reduce their tax burden. As a result, the size distribution has a lot of weight
corresponding to small firms and large firms, and with a ‘missing middle’
which testifies to the disadvantages associated with a medium-sized scale of
operations (Tybout, 2000; Sleuwaegen and Goedhuys, 2002). In this case,
small firms will tend to allay their growth aspirations, while medium-sized
firms will have incentives to grow.
Still other determinants of firm growth can be mentioned here. Almus
(2004) observes that small German firms have lower growth rates when
there is ‘the shadow of death sneaking around the corner’ (Almus, 2004, p.
199). Employment growth rates are observed to be significantly lower up to
three years before a firm’s exit. There is also some evidence that uncertainty
may dampen a firm’s investment. Guiso and Parigi (1999) present convinc-
ing evidence that uncertainty of demand plays a significant role in reduc-
ing firm-level investment in the case of Italian manufacturing firms. Their
measure of demand uncertainty is constructed by referring to the subjective
probability distribution of future demand for a firm’s products according
to the firm’s leading managers. Relatedly, Lensink et al. (2005) use survey
data on Dutch SMEs to show that uncertainty has a mixed effect on invest-
ment. They observe that uncertainty increases the probability of investing
(in the context of a binary ‘invest or not’ model); it is seen to reduce the
overall amount of investment. Finally, Robson and Bennett (2000) show
that the use of external business advice is also associated with superior
growth, although the direction of causality is not easily ascertained. They
also present evidence that firms with an ‘established reputation’ experience
lower employment growth and higher turnover growth.
7.5 INDUSTRY-SPECIFIC FACTORS
According to the traditional ‘structure–conduct–performance’ paradigm

developed by the Harvard school, it was the market structure that deter-
mined the appropriate conduct of incumbent firms, which in turn deter-
mined the performance of these firms. Industry structure was thus ascribed
a central role in explaining firm behaviour and performance.
There are several reasons to expect that the characteristics of an industry
will determine the growth rates of its incumbent firms. Firms in mature
industries are likely to have lower average growth rates, ceteris paribus,
because of the lower level of opportunity in mature industries. Firms in

high-technology industries may have high growth rates due to the rapid
pace of technological progress and the appearance of new products.
Innovation regimes are also supposed to differ considerably across sectors,
which may have an impact on the growth patterns of firms in different
industries.6 In addition, it is reasonable to expect that the growth of firms
is somehow linked to sector-specific degrees of competition and concentra-
tion. More generally, the population ecology literature (surveyed in section
8.5) emphasizes the prevalence of industry-specific factors in explaining
growth of firms, because they share the same resource pool.
In most empirical research into firm growth, industry-specific factors are
controlled away by using industry dummies that take into consideration
the total combined influence of all industry-specific variables put together.
The list of industry dummy variables are not usually reported alongside
the main regression results, partly because of space limitations, and
partly because these industry-specific effects are amalgamations of many
industry-specific factors, which makes their interpretation difficult. In
any case, the inclusion of industry-specific dummy variables does little to
improve the overall explanatory power of the regression model (that is, the
R2 statistic; see Table 7.1). Even within industries, there is a considerable
idiosyncratic component in firm growth rates. For example, industries con-
taining many fast-growing firms also contain a large number of firms expe-
riencing rapid decline (Headd and Kirchhoff, 2007). However, some efforts
have been made to identify the sources of industry-wide differences in firm
growth rates. Audretsch (1995) reports a positive correlation between the
minimum efficient scale (MES) and growth of new firms. It appears that
the post-entry growth rate of surviving firms tends to be spurred on by the
extent to which there is a gap between the MES and the size of the firm.
Similarly, Gabe and Kraybill’s (2002) analysis of 366 Ohio establishments
provides (albeit inconclusive) evidence that the growth of firms is positively
associated with the average size of plants in the same 2-digit industry.
Industry growth, perhaps unsurprisingly, is observed to have a positive
effect on firm growth (Audretsch and Mahmood, 1994; Audretsch, 1995).
Geroski and Toker (1996) examine the growth of firms that are leaders
in their respective industries and find that growth of industry sales has a
positive effect on firm growth. Nonetheless, total industry innovation does
not appear to have a significant effect. Furthermore, Geroski and Toker
observe that the degree of market concentration is positively related to the
growth of these firms.
Apart from differences in growth rates of firms in different industries,
there are differences in how they grow. Maksimovic and Phillips (2008)
observe that, although levels of investment via capital expenditure are
fairly constant across industries, there is more variation when one com-
pares the levels of growth via acquisition across industries.
Some investigators have expressed reservations about the standard
industrial classification scheme, because considerable heterogeneity can
be observed between firms in the same sector. If firms can be grouped
together with comparable firms, however, then it may well be easier to find
regularities and to generalize across firms within these groups. As a con-
sequence, other ways of grouping together similar firms have been sought.
Techniques such as Principal Components Analysis and Cluster Analysis
have been applied in attempts to group like firms together, often resulting
in new taxonomies of firms. In this spirit, Delmar et al. (2003) and Birley
and Westhead (1994) consider the case of small firms, while de Jong and
Marsili (2006), Leiponen and Drejer (2007) and Srholec and Verspagen
(2008) examine the case of innovating firms.
7.6 MACROECONOMIC FACTORS
Some scholars have attempted to uncover country-specific components of

firm growth. For example, McPherson (1996) observes that firms in South
Africa have higher expected growth rates than the four other African coun-
tries in his dataset. Although it has been observed that more of the varia-
tion in firm growth rates is between industries rather than across countries
(Geroski and Gugler, 2004), it is nonetheless instructive to continue our
literature review by considering the influence of macroeconomic factors
on firm growth rates.
Several studies have discussed how firm growth varies over the business
cycle. In this vein, Higson et al. (2002, 2004) analyse US and UK firms
over periods of 30 years and more and observe that the mean growth rate
is indeed sensitive to macroeconomic fluctuations. Furthermore, higher
moments of the growth rate distribution appear to be sensitive to the
business cycle (more on this in section 3.1). Hardwick and Adams (2002)
investigate changes in the Gibrat law coefficient over the business cycle
(that is, the coefficient b in equation 4.5), and they obtain some evidence of
a countercyclical variation of this coefficient. In other words, smaller firms
appear to grow relatively faster during booms, whereas larger firms grow
faster during recessions and recoveries.
Davis et al. (2006) investigate the existence of any long-term trends in
the dispersion (between-firm variation) and volatility (within-firm varia-
tion) of the growth of firms, using an extensive database on US businesses
over the period 1976–2001. They present evidence of a large secular decline
in both dispersion and volatility of firm growth rates. Although publicly
traded firms have experienced a rise in volatility over this period, this is
overwhelmed by declining volatility among privately held firms.
Regional effects have also been observed to have an influence on firm
growth. McPherson (1996) reports that Southern African small businesses
grow faster in urban areas than in rural areas. Similarly, Sleuwaegen
and Goedhuys (2002) observe that firms in the relatively industrialized
Abidjan region experience a faster growth than firms from rural areas of
Côte d’Ivoire, presumably because of the better availability of resources.
Reichstein and Dahl (2004) analyse the growth of Danish limited liability
firms and find that firms have higher expected growth rates when they are
located in a region that is increasing its specialization in the firm’s specific
industry. The aforementioned studies notwithstanding, however, Gabe
and Kraybill (2002) observe that both the county growth rate and a metro-
politan area dummy do not appear to have a statistically significant effect
on growth rates, for their sample of plants in Ohio.
Hart and Pearce (1986) analyse the growth patterns of very large firms
in several countries and observe that firms in Japan and Germany differed
from firms hailing from the USA and the UK, in that their Japanese and
German firms showed no tendency for Galtonian regression towards the
mean. In other words, while small firms grow faster than larger ones in the
US and UK, this was not observed in Japan or Germany. Subsequent work
found that small firms tended to grow faster than larger ones, for both US
data (Hall, 1987; Evans, 1987a,b; Amirkhalkhali and Mukhopadhyay, 1993;
Bottazzi and Secchi, 2003a) and also UK data (Kumar, 1985; Dunne and
Hughes, 1994). In other countries, Galtonian regression to the mean may
be less of an issue. For instance, regression to the mean is not observed in
the cases of Italian (Bottazzi et al., 2007) or French firms (Bottazzi et al.,
2008a). Bartelsman et al. (2005) contribute to this literature by exploring
differences in firm growth rates in a number of developed countries. They
observe that the post-entry growth of successful entrants is much higher in
the USA than in Europe. In particular, they observe that ‘[a]fter 7 years of
life, the average cohort of firms in manufacturing experience more than 60%
growth in employment, while in European countries the increase is in the
5–35% range’ (Bartelsman et al., 2005, p. 386). This is partly because new
firms tend to be relatively smaller upon entry in the US, thus having a larger
gap between their entry size and the industry minimum efficient scale (MES).
The authors suggest that this difference in post-entry growth rates is due to
institutional barriers to growth that are in place in Europe, such as the lack
of market-based financial systems, relatively high administrative costs that
may deter smaller firms at entry, and tighter hiring-and-firing restrictions.
Several other interesting results relating to cross-country differences in
firm growth rates can be found in the study by Beck et al. (2005b), which
analyses a size-stratified firm-level survey database covering over 4000

firms in 54 countries. They observe that firms in richer, larger, and faster-
growing countries have significantly higher growth rates. The growth rate
of GDP is positively correlated with firm growth, which indicates that firms
grow faster in an economy with greater growth opportunities. Inflation
appears to have a positive impact on growth rates, although the authors
admonish that this most likely reflects the fact that firm sales growth is
given in nominal terms. Furthermore, indicators of financial and legal
obstacles, as well as the prevalence of corruption, are obtained from the
questionnaire data. These obstacles vary in importance across countries
and are observed to be negatively correlated with firm growth rates.
7.7 DETERMINANTS OF FIRM GROWTH: A

DISCUSSION
In the last few chapters we sought to look for the determinants of firm
growth. Chapter 4 presented the literature relating firm size to growth rate,
while Chapters 5 and Chapter 6 sought to clarify the relationship between
growth, on the one hand, and relative performance (profits and productiv-
ity) and innovation, on the other. In this chapter, we sought to find other
factors that might influence firm growth, such as firm age, inter-firm com-
petition, characteristics of the entrepreneur, and other firm-specific factors
(such as multi-plant structure or legal status) and also industry-specific and
macroeconomic factors. Although many variables are associated with firm
growth, to an extent that is statistically significant, it is not easy to predict
a firm’s future growth rate with much precision.
Without doubt, the main result that emerges from our survey of empiri-
cal work into firm growth is that the stochastic element is predominant.
Marsili (2001) summarizes in this way: ‘In short, the empirical evidence
suggests that although there are systematic factors at the firm and indus-
try levels that affect the process of firm growth, growth is mainly affected
by purely stochastic shocks’ (Marsili, 2001, p. 18). Geroski (2000) makes
an even bolder statement: ‘The most elementary “fact” about corporate
growth thrown up by econometric work on both large and small firms is
that firm size follows a random walk.’ (p. 169). The R2 statistic in growth
rate regressions is characteristically low, especially for databases containing
many small firms whose growth is particularly erratic. Including a long list
of explanatory variables and lags does little to help raise the R2 value, as is
evident from the survey provided in Table 7.1. Firm growth thus appears to
be remarkably idiosyncratic, even if the assumption of a purely stochastic
process of firm growth is often rejected on purely statistical grounds. Even
Table 7.1 A survey of R2 values obtained from regressions where the dependent variable is the growth rate of a firm or
plant
Study Data Control variables R2

Kumar (1985) Around 700–800 quoted UK Size, lagged growth 1–4%
companies
Variyam & Kraybill 422 small businesses in Georgia, Size, age, multi-plant firms, industry 11–17%
(1992) USA
Geroski & Toker 209 leading UK firms Size, innovation, advertising, industry growth, industry 32%
(1996) concentration
McPherson (1996) 1671 small firms in 5 Southern Firm age and size, dummies for sector and location, human 13–20%
African countries capital and socio-economic variables
97
Geroski et al. (1997) 271 large quoted UK firms Market value, lagged firm growth, innovations, patents, 17–19%
1976–82 industry growth
Harhoff et al. (1998) About 10 000 West German Size, age, subsidiary, diversification, legal status, industry 8%
firms
Liu et al. (1999) Over 900 Taiwanese Age, size, industry dummies, capital–labour ratio, sales per 19–22%
manufacturing plants worker, dummies for R&D and exporting activity
Robson & Bennett Over 1000 SMEs in Britain in Size, age, exports, profits, industry, innovation and 4–8%
(2000) 1997 technology, use of external advice, strategy variables
Geroski & Gugler Large firms in 14 European Size, age, subsidiaries, diversification, growth of rivals 5–6%
(2004) countries, over 100 000 obs.
1994–98
Reichstein and Dahl 8739 Danish firms, 1994–96 Size, age, regional specialization, concentration, industry 1–2%
(2004)
Table 7.1 (continued)
Study Data Control variables R2

Beck et al. (2005b) Survey data covering over 4000 Dummies for government/foreign ownership, export 2–3%
firms of all sizes, in 54 countries status, subsidies, sector of activity; controls for number of
competitors, GDP, GDP per capita, GDP growth, financial/
legal/corruption obstacles
Calvo (2006) About 1000 Spanish firms Size, age, legal liability, product/process innovation, 9%
98
technology, sample selection
Fagiolo and Luzzi 14 277 Italian firms 1995–2000 Size, age, cash flow, dummies for multi-plant firms, year and 2–3%
(2006) industry dummies
Coad (2007d) 8405 French manufacturing Gross operating margin, lagged growth, lagged size, industry 4–8%
firms, 1996–2004 and year dummies
Notes: Control variables include the constant term (though this is not mentioned above). Where fixed-effect regressions have been employed, we
refer to the overall R2 and not the within R2 or between R2. Where we have the choice, we prefer the adjusted R2 to the basic R2. Although growth
rates are mostly obtained by measuring size at annual intervals, this is not always the case. For example, McPherson (1996) takes the average
annual growth rate for the whole of the period since start-up, whereas Liu et al. (1999) take a yearly average of the growth rate over four years.
if we were to compare firms of a similar size and age in the same industry,
with similar financial resources and similar patterns of innovation activity,
the growth rates of these firms would be largely idiosyncratic.
Research into firm growth has made progress on several fronts, however.
Gibrat’s model of proportionate growth shocks (described in section
4.1) serves as a useful benchmark model for describing firm growth. The
empirical literature has suggested a list of determinants of growth, and
they seem to go some way (although admittedly not very far) in explaining
differences in firm growth rates. Statistical analyses that focus on specific
hypotheses have provided nuances to our understanding of both the char-
acteristics of growth rates, and also the determinants and consequences
of growth. Unexpected results have also highlighted puzzles that deserve
more investigation in future.
With these insights in mind, in the following chapter we now turn to
compare our insights with theoretical work which, we hope, will be able to
piece these findings together in a coherent conceptual framework, as well
as providing explanations of the associations. Theoretical approaches to
firm growth can also play a role in suggesting which causal relationships
might be at work, as well as providing further testable hypotheses that can
be investigated in future work.
8. Theoretical perspectives
In the following we briefly present five distinct theoretical perspectives,
discussing their predictions for firm growth and judging them according
to the available empirical evidence. These five theories are the neoclassi-
cal theory (in particular, propositions based on the notion of an ‘optimal
size’), Penrose’s (1959) ‘theory of the growth of the firm’, the managerial
approach, evolutionary economics and its principle of ‘growth of the
fitter’, and also the population ecology approach.
8.1 NEOCLASSICAL NOTIONS OF AN ‘OPTIMAL

SIZE’
Although the term ‘neoclassical’ encompasses a large and vaguely defined

body of literature, for the purposes of our discussion on firm growth we
consider that the main prediction emerging from the traditional neoclas-
sical perspective is that firms are attracted to some sort of ‘optimal size’
(Viner, 1932). This optimal size is the profit-maximizing level of produc-
tion, in which economies of large-scale production are traded off against
the costs of coordinating large bureaucratic organizations. In this view,
firm growth is merely a means of attaining this ‘optimal size’, and it is
of no interest per se. Once firms have reached their optimal size, they are
assumed to grow no more.1
It is relevant to mention here the well-known transaction costs theory of
the firm, which began with Ronald Coase’s seminal article (Coase, 1937).
To summarize briefly, this theory considers that the optimal boundaries of
the firm are determined in a trade-off between the advantages of coordi-
nation via authority in a hierarchy versus the advantages of coordination
through the price mechanism. If transaction costs are relatively large, then
firms will find it worthwhile to expand upstream or downstream in order
to acquire strategic assets. In this way, the production chain can be coordi-
nated by the use of authority in the context of a hierarchical organization.
If transaction costs are low, however, the optimal boundaries of the firm
are smaller because the firm can interact with suppliers and customers via
the market mechanism. Factors affecting the desirability of integration
100
Theoretical perspectives 101
are the frequency of transactions, uncertainty, the degree of asset specifi-

city, and the possibility of opportunistic behaviour. We observe that the
predictions made by the transaction costs literature most often concern
growth by acquisition in the context of vertical integration (Kay, 2000).
You remarked that transaction cost theory is most effective when it is used
to explain cross-country differences (You, 1995). As a result, transaction
cost economics appears to have a limited scope in explaining other aspects
of firm growth.
Another variation on the optimal size theme is in Lucas (1978), who
‘explains’ the lognormal distribution of firm sizes by assuming a lognor-
mal distribution of managerial talent. These managers are then assumed
to be successfully matched to firms with a size that corresponds to their
skill level. Large firms are large because their managers are particularly
talented and can accomplish the difficult task of running a large organiza-
tion with reasonable success. On the other hand, small firms are supposed
to remain small because of the relative incapacity of their managers.
Although managers of large firms would be happy to endorse this idea,
we consider that the practical value of such a model is questionable. (For
instance, we can probably all think of leaders of large organizations who
appear to be outrageously incompetent.) We must acknowledge, however,
that Lucas’s model has proven to be quite influential in the literature on
firm size distributions.
The concept of an optimal size has received (and still receives) a great
deal of attention, despite a blatant lack of empirical support. The notion
of an industry-specific optimal size is at odds with observations on the
wide support and the prominent skewness of the firm size distribution
which can be found even at finely disaggregated levels of analysis. Even
the concept of a firm-specific optimal size appears to be inconsistent with
time-series analysis of the patterns of firm growth (Geroski et al., 2003;
Cefis et al., 2007). In contrast, Gibrat’s model of stochastic drift in firm
size performs much better in empirical analysis of firm growth rates than
do the neoclassical optimizing models we have mentioned.
It seems to us that attaching the word ‘optimal’ in front of the phrase
‘firm size’ is quite fruitless; it seems to accomplish little more than act as
a statement of faith on the part of the writer that the economy operates
according to optimal design, the optimality of which is forever obscured
by the sheer complexity of the economic system. By way of conclusion to
this section, therefore, we suggest that the notion of ‘optimal size’ is of
little use in understanding why firms grow, and that it would be better to
un-learn it quickly.
8.2 PENROSE’S THEORY OF THE GROWTH OF THE

FIRM
Penrose’s (1959) seminal book contains several important contributions to
our discussion on firm growth. We first present her idea of ‘economies of
growth’ before moving on to the ‘resource-based view’ of the firm.
Penrose’s fundamentally dynamic vision of firms holds that firm
growth is led by an internal momentum generated by learning-by-doing.
Managers become more productive over time as they become accustomed
to their tasks. Executive functions that initially posed problems because
of their relative unfamiliarity soon become routinized. As managers gain
experience, therefore, their administrative tasks require less attention
and less energy. As a result, managerial resources are continually being
released. This excess managerial talent can then be used to focus on value-
creating growth opportunities (and in particular, the training of new man-
agers). Firms are faced with strong incentives to grow, because while ‘the
knowledge possessed by a firm’s personnel tends to increase automatically
with experience’ (Penrose, 1959, p. 76), there is a challenge to take full
advantage of this valuable firm-specific knowledge.
It takes time and effort to integrate new managerial resources success-
fully within the firm, but once this is done these new recruits will be able to
execute managerial tasks and, in turn, train managers themselves. In this
way, a firm will grow in order to create value from its unused resources,
which in turn will create new resources.2 Growth in any period is nonethe-
less limited by the amount of available managerial attention. Managers
who spend too much time focusing on the firm’s expansion divert their
attention from operating efficiency. As a result, above a certain point
corresponding to what we might call an ‘optimal growth rate’ (Slater,
1980), increases in growth will lead to higher operating costs. Although
temporary ‘economies of growth’ provide incentives for firms to grow, fast-
growing firms will have higher operating costs than their slower-growing
counterparts. This latter proposition is commonly known as the ‘Penrose
effect’.
Another key concept in Penrose’s theory of firm growth is that firms are
composed of idiosyncratic configurations of ‘resources’. These resources
can play a role in ensuring durable competitive advantage if they are
valuable, rare, inimitable and non-substitutable (Dierickx and Cool, 1989;
Eisenhardt and Martin, 2000). Examples of resources are brand names,
in-house knowledge of technology, employment of skilled personnel,
trade contracts, machinery, and efficient procedures (Wernerfelt, 1984).
Other examples of ‘resources’ have also been put forward. Montgomery
(1994) suggests that Disney’s cast of animated characters can be viewed as
a resource, which has been observed to fuel diversification. Winter (1995)

comments on the similarity of the Penrosian concept of ‘resources’ and
the evolutionary notion of ‘organizational routines’ and concludes that
even routines can be considered as resources.3 Somewhat more unusual is
Feldman’s (2004; p. 304) affirmation that even emotions such as anger and
frustration can be considered to be organization-specific ‘resources’.
A firm can decide upon the direction of a growth project by examining
the strengths and weaknesses of its existing resource base (Barney, 1986).
Economies of growth may emerge from exploiting the strengths associ-
ated with the unique collection of productive opportunities available to
each firm. The indivisible and interdependent nature of these resources
can also be seen to add impetus to a firm’s growth (as we saw in the model
in section 3.2.2). In fast-changing markets, however, a firm’s competitive
advantage may erode if it relies too heavily on certain specific resources. In
such circumstances, a firm’s performance depends on its abilities to create
or release resources and to reconfigure their resource portfolio. These abili-
ties are known as ‘dynamic capabilities’ (Teece et al., 1997; Eisenhardt and
Martin, 2000; Winter, 2003).
Penrose’s vision of firm growth considers that firms grow because of
‘economies of growth’ that are inherent in the growth process, and not
because of any advantage linked to size per se.4 A firm’s size is merely a
by-product of past growth. Although there may be limits to firm growth,
there is no limit to firm size a priori. Penrose’s approach therefore contrasts
greatly with the mainstream neoclassical perspective, in which firms only
grow in order to reach an ‘optimal size’ in static equilibrium, and in which
there are limits to firm size (on this last point, see for example the model in
Williamson, 1967). It is perhaps because of this that Penrose’s contribution
has, unfortunately, been marginalized in the industrial organization litera-
ture – as Montgomery (1994: 167) notes, ‘[a]lthough The Theory of the
Growth of the Firm was published in 1959, it has not had a strong impact
on the direction of economic discourse.’5 Nonetheless, Penrose’s resource-
based perspective has been quite influential in the strategic management
literature.
8.3 MARRIS AND ‘MANAGERIALISM’
The fundamental observation of the ‘managerial’ theory of the firm is that

managers attach utility to the size of their firms (for pioneering work on
the ‘managerial’ perspective, see Marris (1963, 1964) and also the books
by Baumol (1959) and Williamson (1964); see also the ‘agency theory’
proposed by Jensen and Meckling (1976) and Jensen (1986)). A manager’s
compensation, bonuses, and other perquisites very often increase with firm
size.6 Furthermore, non-pecuniary incentives such as prestige, likelihood
of promotion, social status and power are also associated with firm size.
As a result, firm size (and firm growth) are seen to be important factors in
the ‘managerial utility function’, alongside the financial performance of
the firm. For some firms, such as small young firms, the pursuit of growth
maximization may coincide with that of profit maximization, so that a
manager has no conflict of interest between his duties to shareholders and
his own objectives (Mueller, 1969). In other cases, however, managers
have to choose between fulfilling their mandate of profit-maximization
(in service of shareholders) or pursuing their own interests of growth-
maximization. According to the managerial theory, utility-maximizing
managers are assumed to maximize the growth rate of the firm subject to
the constraint of earning a satisfactory profit rate, which should be large
enough to avoid being dismissed by shareholders or being taken over by
stock market ‘raiders’.
In the influential managerial model developed by Marris (1963, 1964),
firms are assumed to grow by diversification only. Marris posits a quad-
ratic relationship between profits and firm growth.7 Above a certain level
of growth, additional diversification has a lower expected profitability
because managers have less time and attention to devote to the operating
efficiency of existing activities and the development of new activities.
The managerial theory has also been extended to the case of growth
by conglomerate merger (Mueller, 1969). Mergers are a faster (and more
expensive) way of growth than internal growth – so managerial arguments
are a fortiori relevant for this type of growth.
Testing the ‘managerial hypothesis’ is a difficult task because the
theoretical models (for example Marris, 1964) propose a non-linear hump-
shaped relationship between growth rate and profit rate, with additional
growth having a negative effect on profits only beyond a certain ‘profit-
maximizing’ growth rate. Nonetheless, one basic prediction that emerges is
that the growth rates of manager-controlled firms will be higher than those
of owner-controlled firms, whilst profit rates are likely to be lower. Some
early studies thus tried to find performance differences between owner-
controlled and manager-controlled firms. The results, however, did not
offer unequivocal support in favour of the theoretical predictions. Radice
(1971) tests the hypothesis that owner-controlled firms have lower growth
rates and higher profit rates than management-controlled firms, using a
sample of 89 large UK firms over the period 1957–67. Perhaps surpris-
ingly, he observes that owner-controlled firms have both higher growth
rates and profit rates. Holl’s (1975) analysis also focuses on large UK firms,
but he fails to detect any significant difference in performance between
owner-controlled and manager-controlled firms. If SMEs are considered,

however, there is some survey evidence that management-controlled firms
have stronger preferences for growth than owner-controlled firms (Hay and
Kamshad, 1994). More specifically, it appears that the largest difference
between the strategies of management-controlled and owner-controlled
firms concerns the area of geographical expansion.
Another body of research, predominantly from the financial economics
literature, has investigated the managerial hypothesis by evaluating the
performance of diversifying firms. This is a meaningful way of investigating
managerialism because the original model proposed by Marris (1963, 1964)
considers that growth takes place exclusively through diversification. The
theoretical prediction, then, is that high levels of diversification are associ-
ated with lower performance. These studies are surveyed in more detail
in section 9.2.2, which focuses on growth by diversification. Many early
studies found diversification to be detrimental to overall financial perform-
ance, which provides some indirect support for the managerial hypothesis.
This evidence came from both ‘event studies’ of the stock market’s response
to diversification announcements, and also analysis of ex post profits of
diversifying firms. Conversely, over-diversified firms that subsequently
refocus are seen to improve their performance. More recent evidence on
growth by diversification has called previous work into question, however
– it seems that diversification can be a valuable strategy for those firms that
choose it. Finally, the evidence suggests that growth by acquisition appears
to be negatively related to a firm’s financial performance (Dickerson et al.,
2000).
8.4 EVOLUTIONARY ECONOMICS AND THE

PRINCIPLE OF ‘GROWTH OF THE FITTER’
The modern economy is increasingly characterized by turbulent competi-

tion and rapid technical change, and as a consequence a dynamic theory
of competitive advantage may well be more relevant to understanding the
economics of industrial organization than the more neoclassical concepts
of equilibrium and static optimization. Evolutionary economics has thus
been able to make a significant impact on IO thinking, because it proposes
a dynamics first! conceptualization of the economy. Evolutionary theory
has its foundations in Schumpeter’s vision of capitalism as a process of
‘creative destruction’, and borrows the notions of diversity creation and
selection to account for the dynamics of economic development. Alchian’s
(1950) theoretical paper argues that the evolutionary mechanism of selec-
tion sets the economy on the path of progress, as fitter firms survive and
grow whilst less viable firms lose market share and exit.8 The notion of
selection via differential growth is also a central theme in the books by
Downie (1958) and Nelson and Winter (1982). Downie (1958) models
industrial development by assuming that firms grow by reinvesting their
earnings. Growth rates thus rise with profitability. Nelson and Winter’s
(1982) influential book contains a formal micro-founded simulation model
in which firms compete against each other in a turbulent market environ-
ment. In this model, firms can gain competitive advantage through either
the discovery of cost-reducing innovations or by imitating the industry
best practice. Firms that are more profitable are assumed to grow, whilst
firms that are less successful are assumed to lose market share. Agent-based
simulation modelling has since remained a dominant tool in the evolution-
ary literature (see, among others, Chiaromonte and Dosi, 1993; Dosi et
al., 1995; Marsili, 2001; and Dosi et al., 2006; see also Kwasnicki, 2003;
and Dawid, 2006 for surveys). In addition to computer simulation models,
the principle of ‘growth of the fitter’ has also formed the foundations of
analytical evolutionary models (see, for example, Winter, 1964, 1971; and
Metcalfe, 1993, 1994, 1998;9 see also Jovanovic, 1982; Hopenhayn, 1992;
and Melitz, 2003 for more elaborate mathematical models).
The evolution of firms and industries, as depicted in this family of
models, is guided by the mechanism of ‘replicator dynamics’, by which
growth is imputed according to some broad measure of ‘fitness’ (or ‘viabil-
ity’). This mechanism can be presented formally by Fisher’s ‘fundamental
equation’, which states that:
–
xi 5 xi (Fi 2 F ) (8.1)
where d stands for the variation in the infinitesimal interval (t, t 1 dt), and
xi represents the market share of firm i in a population of competing firms.
Fi is the level of ‘fitness’ of the considered firm, where fitness corresponds
to relative financial performance or perhaps some measure of relative pro-
– –
S
ductivity.10 F is the average fitness in the population, i.e. F 5 ixiFi, and
a is a parameter. It is straightforward to see that this equation favours the
above-average firms with increasing market share, whilst reducing that of
‘weaker’, less profitable firms.
This ‘replicator dynamics’ does sound intuitively appealing, because
implicit in it is the idea that selective pressures act with accuracy, that
financial constraints prevent inefficient firms from growing, and that the
economic system adapts so as to allocate resources efficiently amongst
firms, such that firms ‘get what they deserve’. However, these assumptions
may not find empirical validation for a number of reasons. First of all, it
cannot be assumed that all firms have the same propensity to grow. Some
high-profit firms may not be interested in business opportunities that are

instead taken up by less demanding firms. Freeland (2001), for example,
documents how GM’s shareholders resisted investing in additional busi-
ness opportunities and sought to restrict growth expenditure even when
GM was a highly profitable company. If this is the case, then stricter inter-
nal selection will cause high-profit firms to overlook opportunities that are
instead taken up by less profitable competitors. In this way, growth may
be negatively related to profitability. An extension of this idea is presented
by the managerial literature (see section 8.3), which identifies a tension
between profits and growth – this arises when managers seek to grow at
a rate higher than that which would be ‘optimal’ for the firm as a whole,
with the resulting growth rate being limited by shareholder supervision. If
shareholders monitor management closely, growth rates are predicted to
be low and profit rates high. If shareholders are ineffective at monitoring
and discipline, however, the growth rate may be high and profit rates low.
Second, high profits may be made by firms that can exercise market power
by restricting their production to obtain a higher price per unit sold. In this
case, a firm which has sufficiently inelastic demand for its goods will have
a higher profit rate if it reduces its capacity. In this case too, increases in
profits would be associated with negative growth. Third, if a firm occupies
a highly profitable niche market, it may not have opportunities to expand,
despite its high profits. Fourth, a firm may experience a higher profit rate
due to efficiency gains by downsizing and concentrating on its core com-
petence. Here again, we have no reason to suppose a positive association
between profits and firm growth. (Further reasons why firms may not all
want to grow are discussed in section 9.1 on ‘Attitudes to Growth’.) As a
result, the existence of a relationship between profitability and growth is
an empirical question.
The principle of ‘growth of the fitter’, despite its eloquence, does not
appear to receive much support from empirical analyses. Let us consider
the two usual candidates for ‘fitness’, namely profitability and productiv-
ity, in the light of the survey of empirical work in section 5.1. To begin
with, we observed that profitability and sales growth appear to be largely
independent from each other, when we consider the available evidence
from studies of French and Italian manufacturing industries. Similarly,
research based on data for US, UK and Italian manufacturing firms fails to
find that the more productive firms grow faster than the others. Although
profitability and productivity are perhaps the most obvious indicators of
‘fitness’, others such as product quality or cost levels have also been sug-
gested. These latter variables are usually more difficult to observe, and so
they are not often used in empirical work (although it can be anticipated
that they should be positively correlated with both profitability and
productivity). However, we can mention here the work by Hardwick and

Adams (2002). Whilst these authors fail to find any effect of profitability
on firm growth, they do observe a negative influence of the input cost ratio
on growth, for UK life insurance companies (that is that high-cost firms
have lower growth rates). Weighing up the available evidence, though, we
must acknowledge that empirical work on the principle of ‘growth of the
fitter’ does not provide encouraging results. It may be better to suppose
that selection works only by elimination of the weaker, with growth not
being related to any notion of ‘viability’ but instead being at the discretion
of managers. In this view, we have ‘survival of the fitter’ without ‘growth of
the fitter’ (as in the simulation model of van Dijk and Nomaler (2000)).
There are also welfare implications attached to the failure of the princi-
ple of ‘growth of the fitter’ (Baily and Farrell, 2006). If high performance
firms were observed to have the fastest growth rates, then selective proc-
esses would bring about some sort of efficient dynamic allocation of the
economy’s resources between firms. Scarce productive resources would
be attributed to those firms who can best exploit them. However, since
‘growth of the fitter’ is generally not observed, economies may be far from
achieving their full productive potential. This may be an opportunity for
policy intervention.
8.5 POPULATION ECOLOGY
The ‘population ecology’ or ‘organizational ecology’ perspective hails

from sociology and follows on from the seminal contribution of Hannan
and Freeman (1977). (More on population ecology approach can be seen
in the surveys by Geroski (2001) and Hannan (2005), and some recent
developments can be found in the special issue of Industrial and Corporate
Change, Vol. 13, No. 1, 2004.) The basic theoretical prediction pertaining
to the growth of organizations is that these latter require resources which
are specific to niches, and these niches have a particular ‘carrying capacity’.
If a firm has discovered a new niche with a rich resource pool, then this firm
will be able to grow without hindrance. The number of firms in the niche
will also grow, due to entry of new organizations. If the population grows
to a level where the niche’s resource pool is saturated, however, then com-
petition between firms will limit the growth rates of firms. This relationship
between the growth of organizations and the competition for resources in
a particular niche is known as ‘density dependence’.
The population ecology perspective thus places the growth of organiza-
tions in the context of niche-specific growth patterns without focusing as
much on heterogeneity between organizations occupying the same niche.
This should not be taken to mean that the scholars deny the existence of
differences between organizations.11 Instead, this is due to the fact that
the fundamental unit of analysis here is the population of organizations
within a niche, rather than the individual organizations that make up
the population. As a consequence, population ecologists tend to explain
the performance of organizations by referring to features common to all
organizations within the same niche, rather than firm-specific factors.12 Of
course, there are clear limits to a theory of firm growth rates based solely
on industry-wide characteristics, because large differences in growth rates
can be observed between firms in the same industries. Notwithstanding
the analytical starting point, however, some work in this stream of litera-
ture relates the performance of organizations to idiosyncratic rather than
environmental factors.
Broadly speaking, the empirical strategy in the ‘population ecology’
literature takes place by gathering life-history data on populations of
organizations that are arguably in the same ‘niche’. This niche may refer
to specific industries (for example automobile producers, Hannan et al.,
1995), niches within industries (such as biotechnology drug discovery
companies, Sorensen and Stuart, 2000), or even non-commercial ideologi-
cal organizations (Minkoff, 1999; Simons and Ingram, 2004). Most studies
focus on the effects of characteristics of organizations,13 populations, and
the environment on organizational performance by examining birth and
death rates of organizations. However, efforts have been made to explain
differences in growth rates between firms in the same industry. Barron et al.
(1994) analyse data on New York Credit Unions over the period 1914–90
and observe that larger firms have lower expected growth rates than their
smaller counterparts. The interpretation they offer is that larger organiza-
tions have become less efficient and less well adapted to the current busi-
ness environment, thus being more vulnerable to young competitors. This
builds upon a key population ecology tenet that firms are fundamentally
inert (Hannan and Freeman, 1984), being both averse to and relatively
incapable of strategic or organizational change.
8.6 DISCUSSION
A number of theories of firm growth have been proposed; each of them is

eloquent in its own way and makes a contribution to our understanding of
how and why firms grow.
We began with the neoclassical conception of firms growing towards
an optimal size (section 8.1), before discussing Penrose’s theory of firm
growth. In Penrose’s view, firms can be seen as composed of idiosyncratic
resources, and managers divert time and effort into drawing up their
growth projects, taking into account the growth opportunities that are
available. The managerial perspective on firm growth was presented in
section 8.3, according to which managers take pleasure in being in charge
of a large organization and seek to grow as much as possible, beyond the
profit maximizing level. In section 8.4 we presented the evolutionary princi-
ple of ‘growth of the fitter’, which seems quite plausible at first sight but, so
far, it has not performed well in empirical investigations. Finally, in section
8.5 we presented the population ecology perspective on firm growth.
Although some theories are more useful than others, none of the theories
can provide a comprehensive explanation of firm growth. Firm growth is
indeed a multifaceted phenomenon, it has a strong idiosyncratic character,
and as a result it is difficult to generalize across firms and circumstances.
9. Growth strategies
After starting the book by reviewing the empirical evidence on firm growth,
the previous chapter contrasted the evidence with some broad-based
theoretical predictions on firm growth. These theories did not fare excep-
tionally well in accounting for the variation in growth rates across firms,
presumably because it is not easy to generalize across heterogeneous firms
facing dissimilar circumstances.
In this chapter (and also the next) we take an approach that can perhaps
be described as ‘appreciative theorizing’ (Nelson and Winter, 1982, pp.
45–8); we will seek to provide tailored descriptions of certain aspects of the
firm growth process, without trying to unify these aspects together under
one centralized theoretical monolith.
We begin by discussing the attitudes of firms towards growth (section
9.1), and then the available means of achieving growth (such as diversifica-
tion, acquisition and internationalization – see sections 9.2–9.4). It appears
useful to relate these two topics to the distinction between ‘demand’ for
growth and ‘supply’ of growth opportunities, respectively. Firm growth
requires both a willing attitude to take up growth opportunities, and also
the availability of suitable opportunities. However, in the long run, the dis-
tinction between supply and demand determinants of growth may become
blurred (Penrose, 1960). Entrepreneurs and managers with a strong desire
to grow will surely find suitable growth opportunities if they search for
them. Correspondingly, one could suppose that even firms with a marked
aversion to growth will eventually take up additional growth opportunities
if these are attractive enough.
9.1 ATTITUDES TO GROWTH
As firms get older, they generally increase in size. However, growth is

neither irresistible nor inevitable. Indeed, some firms may not wish to
pursue growth even if the opportunity presents itself. We observed in
section 5.1 that a firm’s growth rate is largely independent of its financial
performance. This is consistent with suspicions of a disconnect between a
firm’s ability to grow and its desire to grow. In this section we attempt to
111
expound why firms may or may not want to grow, as well as discussing the
intentionality of growth.
9.1.1 The Desirability of Growth
Advantages of growth
Growth of an organization can be seen as a means of alleviating tensions in
its internal management. Employees appreciate the opportunities for pro-
motion as well as the higher salaries and prestige that accompany growth.
Aoki (1990) writes that employees may even be willing to forgo current
earnings in exchange for future benefits made possible by promotion in an
expanding hierarchy. In addition, work is likely to become more challeng-
ing as the firm ‘breaks from its routines’ and expands into new business
areas. ‘Work is more fun in a growing company’ as Roberts (2004, p. 243)
bluntly puts it. Bronars and Deere (1993) point out that growing firms
are better able to maintain worker morale, and as a result they are less
susceptible to unionization activity on the part of the employees.
Conversely, a lack of growth can create an uninspiring and stultifying
business environment which depresses managerial efficiency (Hay and
Morris, 1979). As a result, in growing firms it is ‘easier to obtain com-
mitment to organizational goals and priorities from various factions and
to resolve conflicts between those factions’ (Whetten, 1987, p. 340). An
organization may thus seek a positive growth rate in order to keep its
members satisfied. Indeed, it has been conjectured that firms that take
their employees’ interests seriously are likely to have higher growth rates
(Aoki, 1990).
The managerial vision of the firm can be considered as an extension of
this line of reasoning. Managers attach positive utility to the growth rate
of the firm, because an increase in firm size is associated with increases in
compensation, power, prestige, bonuses and perquisites. One difference,
however, is that managers have the power to determine a firm’s growth
strategy themselves, and so they can pursue a growth rate above that which
would be optimal for the shareholders. For more on the managerialist
theory of the firm, see section 8.3.
Firms may also seek growth as a means of attaining other objectives
related to its production of goods and services. Lower production costs
may be achieved if expansion leads to economies of scale (due to a larger
scale of production), or economies of scope (because of a wider range of
products or services). Growth may also take place if firms wish to expand
their productive capacity or boost their output so as to deter entry from
potential competitors (Dixit, 1980).1 Furthermore, a larger, more diversi-
fied firm is better able to spread its risk among its various activities. (This
Growth strategies 113
will be an advantage for managers whose fortunes are tied to those of the
firm (Amihud and Lev, 1981), although it is not necessarily an advantage
for shareholders because they can reduce their risk by investing in a diversi-
fied portfolio including other firms.) In this way, growth can be considered
to be a basis for security (Whetten, 1987).
Other reasons have also been advanced to suggest why firms might
want to grow. One reason might be because growth is sometimes a more
suitable metric of performance than profits – this is particularly true for
high-volatility markets. A firm’s management may thus set its performance
goals in terms of percentage increases in sales rather than profit margins or
share prices. Other firms may grow for want of a better alternative. This
might be the case for firms who grow by reinvesting profits in the company,
as a means of avoiding heavy taxes (on dividends, for example).
There is some empirical evidence that demonstrates the positive effect
of growth on firm performance. Coad (2007d) analyses a large sample of
French manufacturing firms and observes that growth is associated with
short-lived increases in profit rates, whether growth is measured in terms
of employment, sales, or value-added. Perhaps surprisingly, there seems to
be a larger effect of growth on profits than that of profits on growth. This
finding of a beneficial and temporary influence of firm growth on profit
rates is consistent with the Kaldor–Verdoorn ‘dynamic increasing returns’,
Penrose’s (1959) theory of ‘economies of growth’, and Starbuck’s (1971)
‘will-o’-the-wisp’ models of firm growth.
Disadvantages of growth
Despite the aforementioned advantages linked to growth, some managers
or owner-managers may be wary of increasing the size of their firm. One
major reason for this is what we could call the ‘control-loss’ argument.
Loss of control may originate from the increased size or the rate of growth.
As a firm increases in size, as employees are added and the number of hier-
archical levels increases, the manager has less control of the firm and is less
well informed of its current state (Williamson, 1967). Problems of control
and coordination are also increasing functions of the growth rate. Whilst it
has been advanced that problems of coordination vanish under truly static
conditions (Kaldor, 1934), fast-growth firms may experience difficulties in
coordinating operations in a complex and changing environment.2
Family-owned and traditional firms may have an especially cautious
approach to growth if they are keen to keep the firm under tight control or
if they are reluctant to integrate a large number of employees and managers
from outside the family. Furthermore, they may be particularly risk-averse
because failure of the enterprise may take on connotations of ruining
the family tradition. Managers whose training and experience have been
confined to a single industry are also characteristically timid when it comes

to growth, especially growth by diversification (Ansoff, 1987). This is also
true for managers approaching retirement. In these cases, firms may prefer
not to expand, and instead remain in a ‘comfort zone’.
Larger firms are less attractive environments than smaller firms for a
number of reasons. Large firms are less adaptable and less responsive than
their smaller counterparts. Routinization replaces initiative, and bureau-
cratic ossification replaces the dynamism associated with small firms.
Large organizations tend to become less motivating environments for
employees. The initial energy and motivating enthusiasm of the founding
entrepreneur is replaced by a manager whose role is to monitor and coor-
dinate a more routinized method of production (Witt, 1998, 2000, 2007;
Cordes et al., 2008).
A common ideology and a cooperative working environment is substi-
tuted by an organizational culture in which employees are more concerned
with personal and self-centred goals. However, it should be emphasized
that a distaste for organizations of a large size does not necessarily preclude
a firm’s growth. Because of ‘economies of growth’, firms may still benefit
from taking up marginal growth opportunities even if there are disecono-
mies of large size (Penrose, 1959). Indeed, growth should not be seen as
merely a means of attaining a larger size.
A firm’s attitude to growth may also be influenced by the existence of
a certain size threshold. Schivardi and Torrini (2008) demonstrate that
Italian firms close to the threshold of 16 employees are reluctant to expand
because this would be associated with an increase in their employment
protection responsibilities. Although statistically significant, this effect
can only be detected using large databases, however, and so its economic
importance should not be exaggerated. Tybout’s (2000) survey of manu-
facturing firms in developing countries describes how small firms have
incentives to stay small and informal to avoid taxes. In contrast, medium-
sized firms have incentives to grow in order to become large enough to
be able to lobby the government. It has also been suggested that large
firms whose sales account for a significant fraction of the market may also
restrain their own growth in order to keep prices high and avoid ‘spoiling
the market’ (see, for example, Nelson, 1987).
Some empirically-minded papers have found negative attitudes to
growth in a range of situations. A lack of desire for growth has been found
by Tether (1997) in the case of UK high-tech firms as well as by Audretsch
et al. (2004) for family-owned hospitality industries in the Netherlands.
Hay and Kamshad (1994) present evidence from a survey of UK SMEs.
They find that many software firms encounter limits to growth imposed
by the scarcity of first-class programmers. In the instruments industry, the
scientists that founded the firms are often not well prepared for the man-
agement roles that larger firms require. In the printing sector, many firms
choose not to grow simply because the owners use their business as a means
to support a relaxed and independent lifestyle. More generally, Greiner
(1998) provides the following description of the ‘lifestyler’ manager’s
attitude to growth:
Top management that is aware of the problems ahead [linked to organizations

of a large size] could well decide not to expand the organization. Managers may,
for instance, prefer to retain the informal practices of a small company, knowing
that this way of life is inherent in the organization’s limited size, not in their con-
genial personalities. If they choose to grow, they may actually grow themselves
out of a job and a way of life they enjoy. (Greiner, 1998, p. 67)
9.1.2 Is Growth Intentional or Does it ‘Just Happen’?
Are growth opportunities to be passively seized or are they to be built? Is

firm growth intentional and proactive, or does it ‘just happen’? Some per-
spectives on firm growth, such as Gibrat’s law, view it as a passive absorp-
tion and accumulation of growth opportunities. Other authors, however,
talk of ‘growth strategies’, and sometimes firms include growth rate targets
among their explicit performance objectives. In this section, we discuss
different perspectives on the intentionality of firm growth.
Gibrat’s (1931) ‘law of proportionate effect’, in its simplest form, consid-
ers that the growth of firms is best modelled as a stochastic process in which
the magnitude of a random ‘growth shock’ over a specific period is inde-
pendent of a firm’s size. Relatedly, the ‘island models’ developed by Ijiri
and Simon (1977), Sutton (1998) and Bottazzi and Secchi (2006a) present
statistical processes in which firms are seen as ‘islands’, or independent
entities, whose resultant growth is simply a cumulation of the stochastic
opportunities they receive in any period. These growth opportunities are
supposed to be exogenously created and upon arrival they are randomly
allocated across firms. Firms are required to have minimal rationality,
and, more generally, these statistical models can be said to have a minimal
recourse to any economic theory because growth is entirely explained
by random factors. One advantage of this class of models, however, is
that they can explain the observed size distribution whilst demonstrating
both simplicity and generality. Whilst Gibrat’s law appears to be one of
the more useful approaches to modelling firm growth and the evolution
of industries, it should nonetheless be remembered that there is a certain
rationality and intentionality in the process of firm growth.
Another early model in Parkinson (1957) considered that the size of
an organization has an inherent and quasi-automatic tendency to drift
upwards. Parkinson focuses on the growth of public administration

organizations. The rationale of this model is that the behaviour of officials
in a bureaucracy are guided by two axioms: first, that an official prefers to
multiply subordinates rather than rivals; and second, that officials make
work for each other. Consider the case of an employee, A, who considers
himself overworked. He has three options – he may resign, he may ask to
halve his work with a colleague called B, or he may ask the assistance of
two subordinates, C and D. In fact, the third option is the only serious
one. If he were to resign, he would lose his job and all associated privileges.
Were he to ask for B to be appointed, he would merely introduce a rival
into his level of the hierarchy (which would also reduce his chances of pro-
motion). As a result, he asks for two assistants. These assistants improve
his status in the organization, and furthermore, by dividing his work into
two categories (for C and D) he will become entrenched in a position of
power because he is the only person who understands the work of both of
the assistants. The story need not end here, however:
When C complains in turn of being overworked (as he certainly will) A will, with
the concurrence of C, advise the appointment of two assistants to help C. But he
can then avert internal friction only by advising the appointment of two more
assistants to help D, whose position is much the same. With this recruitment of
E, F, G and H the promotion of A is now practically certain.
Seven officials are now doing what one did before. . . . For these seven make so
much work for each other that all are fully occupied and A is actually working
harder than ever. (Parkinson, 1957, p. 5)
As we have seen, in this particular model, the growth of the organization

has little to do with strategic decisions taken by top management, but
instead it is due to the behaviour of employees throughout the hierarchy,
with the top management having less than perfect control over the firm.
Some authors, mainly from Penrose’s camp, explain growth as being due
to the build-up of internal pressure. As time goes by, managerial resources
are continually being released as managers become more accustomed to
their work and become more productive. (More on Penrose’s Theory of
the Growth of the Firm can be found in section 8.2.) As a result, managers
can divert their attention from routine operations to planning and carry-
ing out growth projects. Unused managerial services are a key determinant
in a firm’s capacity to expand. Firms must then decide upon the direction
for growth. Managers must search for potential growth opportunities and
draw up growth plans. As a result, growth is an informed and intentional
process (Penrose, 1955).3 Growth is seen primarily as a result of manage-
rial decision and ‘human will’ rather than being a response to technologi-
cal factors.4 If, on the other hand, these unused managerial services are
involved in growth projects that are unstructured or ill-prepared, then they

are unlikely to succeed (Penrose, 1955; Dixon, 1953).
Strong rationality is often imputed to growth decisions taken by young,
small firms in particular. In this context, Wiklund (2007) writes: ‘First and
foremost, it is important to emphasize that development and growth is, for
a large part, dependent on conscious decisions made by the firm leader.’
(p. 147). This may be because the entrepreneur has more knowledge of
his firm’s day-to-day operations and is in greater control of his venture. It
could also be because all of the problems faced by young firms are new to
them, and as such they all require special attention. More mature firms, in
contrast, are more experienced and routinized, precedents have determined
the firm’s cognitive frame, and many decisions are taken almost automati-
cally without much conscious planning. Mature firms may have become
burdened with the repetitive grind that accompanies their production
routines to the extent that they have lost their original ‘drive’.
In neoclassical work, strong rationality is also attributed to the growth
of firms. In this perspective, growth is the result of a forward-looking
process in which firms adjust their current scale of production to antici-
pate future market trends. According to neoclassical q-theory, firms are
assumed to have rational anticipations, and their size is determined as the
solution to an intertemporal profit-maximization problem on an infinite
time horizon (see section 5.1).
By way of conclusion, then, we consider that firms do have some ration-
ality in their growth, although assuming perfect rationality is certainly
taking things too far. For some firms, such as small firms struggling to
reach the MES (minimum efficient scale of production), growth is very
much an intended outcome. This is in spite of what a simplistic and literal
interpretation of Gibrat’s law might suggest – firm growth is not just an
‘organizational drift’, but instead there is some rationality and planning
involved.
9.2 GROWTH STRATEGIES: REPLICATION OR

DIVERSIFICATION
‘[G]rowth is not for long, if ever, simply a question of producing more

of the same product on a larger scale; it involves innovation, changing
techniques of distribution, and changing organization of production and
management’ (Penrose, 1959, p. 161). Although in some cases firms may be
able to expand by producing ‘more of the same’ using the same resources,
the time will come when further expansion will require them to take on
new employees, build new production plants, or even diversify into new
markets. There are thus a number of issues and complications that accom-
pany a firm’s decision to grow. These issues are discussed in the following
sections.
9.2.1 Growth by Replication
In traditional economic theory, firms decide how much to produce by select-

ing a profit-maximizing output level determined by the demand curve. It is
supposed that the firm operates in a homogeneous product market and can
easily expand or contract to arrive at the optimal output level. While this
may be an acceptable description of the output of one particular factory
floor, it is unhelpful in describing more significant growth events such as
the hiring of new employees or the setting up of new production plants.
One caveat of this primitive vision of firm growth is that the production
of goods and services requires the application of a certain amount of tacit
knowledge. This tacit knowledge is difficult to transfer from one individual
to another, or from one locus of production to another. As a firm grows,
problems may arise because of the difficulty in transferring this tacit
knowledge. Although the firm may have enjoyed successful production
in the past, it may be non-trivial to replicate this past success with newly-
introduced additional productive capacity, especially where production
processes are characterized by a high degree of complexity (Rivkin, 2001).
In other words, businesses may fail when they try to reproduce a best
practice because the in-house ‘experts’ don’t truly know why it worked in
the first place (Szulanski and Winter, 2002).
Indeed, the extensiveness of tacit knowledge and the difficulty of
replication may go some way in explaining the persistent heterogeneity in
profitability and also productivity levels that are visible even between firms
in the same narrowly-defined industrial sectors.5
How then can a firm replicate its superior performance? A firm’s
replication strategy is more likely to be successful if a few guidelines are
followed (Winter and Szulanski, 2001; Szulanski and Winter, 2002). First,
the template should be kept in mind throughout the replication process,
and even after acceptable results have been obtained by the new unit.
This template should be copied as closely as possible. Changes can be
introduced only after decent results have been obtained. Managers should
focus on the activity they are trying to replicate, rather than on what the
documentation or the experts say. Finally, it is important that managers
have a meek attitude and a keenness to copy the template faithfully rather
than to attempt to improve upon it.
A more extreme approach to technology transfer, applied by Intel, is
known as the ‘copy EXACTLY!’ policy (MacDonald, 1998). Semiconductor
manufacturing is characterized by very complex production processes in

which the process steps have low tolerances and have complex interactions.
In addition, this complexity has increased with successive generations of
semiconductors. Precision in replication is thus of paramount importance.
If variables such as barometric pressure, ultra pure rinse water tem-
perature and the length of the electrode cooling hose are not copied with
utmost accuracy, the results can be catastrophic. After a period in which
new plants exhibited a dismal performance, Intel developed the ‘Copy
EXACTLY!’ philosophy according to which ‘“everything which might
affect the process, or how it is run” is to be copied down to the finest detail,
unless it is either physically impossible to do so, or there is an overwhelm-
ing competitive benefit to introducing a change’ (MacDonald, 1998, p. 2)
(emphasis in the original). Furthermore, if a modification has been sug-
gested and is applied, this idea is simultaneously implemented at all other
sites as well. As a result of this replication strategy, it is now common for
Intel’s new production plants to meet best-practice performance standards
from the very first day of production.
9.2.2 Growth by Diversification
It would not be possible to describe firm growth without discussing diversi-

fication. Indeed, the early theoretical contributions by Penrose (1959) and
Marris (1963, 1964) spoke of growth exclusively in terms of diversification.
In this section we begin by presenting some theoretical insights before
moving on to a discussion of the empirical evidence on the matter.
Theoretical perspectives
An early view of diversification considered that managerial competences
were the key to superior firm performance, irrespective of the sector of
activity. In other words, this perspective holds that ‘management is an
amorphous substance which can be applied with equal success to totally
unrelated lines of business’ (Mueller, 1969, p. 651). In order to take full
advantage of these scarce assets, successful firms sought to spread their
superior management capabilities across several different industries. In
this way, diversification was guided by a logic of synergies of managerial
competence as opposed to synergies of a technological nature. As a result,
the large diversified conglomerate became a popular organizational form,
especially in the 1950s and 1960s.
Penrose’s (1959) vision of firm growth by diversification can be placed
within this context. Managerial attention is seen to be the main factor limit-
ing firm growth. As a firm continues its operations, incumbent managers
gradually gain experience, and new managers can be trained and integrated
into the firm, thus expanding the firm’s resource base. In this way, mana-
gerial resources are continually being freed up over time. Growth thus
constitutes a responsible use for excess managerial attention – it challenges
managers to focus their attention on generating profits in new activities.
However, Penrose also gives clear recommendations as to the direction of
diversification. A key element of Penrose’s theory of firm growth is that
firms are composed of indivisible resources, which are specialized and spe-
cific to the firm. A firm’s diversification strategy should therefore focus on
how best to exploit the idiosyncracies of the firm’s current resource base.
In other words, growth by diversification is most effective when the new
activities are related to the existing resource base.
The notion of related or ‘synergistic’ diversification is central to Igor
Ansoff’s (1987) celebrated book. Ansoff advocated a prudent approach
for diversification at a time when, in retrospect, it appears that general
management synergies were overestimated. According to him, firms should
only consider diversification when they have no other option in order to
realize their growth objectives – ‘if a firm can meet all of its objectives by
measures short of diversification or internationalization, it should do so’
(Ansoff, 1987, p. 131). Indeed, in many cases a firm can discover growth
opportunities by re-evaluating and re-formulating its strategies within its
present portfolio of activities, instead of expanding the portfolio by com-
mencing new activities. Firms that choose to diversify, however, can do this
in one of three ways: by exporting the firm’s traditional products or services
into new markets, which constitues the ‘highest synergy move’ (Ansoff,
1987, p. 125), or by diversifying according to synergies of demand or syner-
gies of technology. In each case, attention must be paid to the coherence of
the diversified firm’s portfolio of activities. Candidate new businesses must
display synergies with the existing portfolio of activities along dimensions
such as operations, R&D, or marketing and distribution. These synergies
may be due to lower expected fixed costs of starting up, or alternatively
due to anticipated operating economies. Furthermore, efforts should be
made to convert the ex ante ‘potential synergy’ into ‘realized synergy’, by
actively seeking to integrate the new activity alongside the firm’s existing
activities. If these guidelines are successfully applied, synergistic diversifi-
cation allows firms to earn superior profits by leveraging their capabilities,
know-how and general experience in new markets. It should be pointed out
that synergistic diversification is not incompatible with corporate refocus-
ing, but is instead closely related (Batsch, 2003). Both of these view the
firm as a coherent portfolio of related activities based on a small number of
core competences. Refocusing can be seen as a corrective strategic measure
undertaken after excessive unrelated diversification – it is a modification
(but not necessarily a reduction) in a firm’s activities as the firm seeks to
focus on exploiting certain specific capabilities. Refocusing should not be

seen as a ‘return’ to the firm’s previous condition, however, but as a strate-
gic re-evaluation of a firm’s core competences in an ever-changing business
environment (Paulré, 2000).
‘Managerial’ or ‘agency’ theories of firm growth, as presented above in
section 8.3, have also made a considerable impact on research into diversi-
fication. (In fact, empirical work on diversification has mainly focused on
testing the hypothesis that diversification is detrimental to firm perform-
ance.) The decision to diversify is usually taken at the initiative and the
discretion of managers, and managers have strong incentives to diversify
even when this is not in the best interests of shareholders. On the one hand,
standard economic theory predicts that diversification will be in the best
interests of the firm as a whole when expansion into new activities promises
relatively high profit levels. Diversification was also historically encour-
aged for other reasons pertaining to the business environment around the
time of the 1960s – the multidivisional firm (the ‘M-form’) was lauded as
an effective organizational innovation, underdeveloped financial markets
meant that there were advantages in having an internal capital market (the
‘deep pockets’ argument), and the prevailing antitrust legislation limited
growth prospects in any one industry. On the other hand, however, diver-
sification also offers at least four other advantages that are more specific
to managers. First, managers of large and growing firms receive higher
pay (as well as increases in bonuses, ‘perks’, prestige, and ‘the pure pleas-
ures of empire-building’ (Montgomery, 1994, p. 166). This point is clearly
illustrated by Hyland and Diltz (2002), who compare managerial com-
pensation for a group of diversifying firms with a similar matched sample
of undiversifying firms – ‘the mean compensation increase over the time
interval between proxy statements for diversifying firms is $84,397 and the
median is $57,133. . . . For matched-sample firms, the mean compensation
increase is $22,642 and the median is $18,128’ (Hyland and Diltz, 2002,
p. 64). Second, managers who have vested interests in the performance
of their firm (or who are merely concerned about their reputations) may
attempt to lower the firm’s volatility by spreading the risk and diversifying
into new activities, even if this does not improve the firm’s average rate of
return (Amihud and Lev, 1981). This is against the interests of sharehold-
ers, because these latter usually prefer to reduce risk by including diverse
specialized firms in their investment portfolio, rather than by investing in
one diversified firm. Third, managers may diversify in order to ensure that
the firm will require their personal skills and services in the future – this is
known as the ‘managerial entrenchment’ argument (Shleifer and Vishny,
1989). Fourth, managers may be reluctant to distribute any spare cash
flow back to shareholders in the form of dividends, and instead they may
prefer to spend it on pet projects even if these have a low expected return
(Jensen, 1986).
Empirical evidence
A large body of research in the financial economics literature has focused
on the relative performance of diversified firms vis-à-vis stand-alone firms
or less-diversified firms, generally using data on large US firms. The general
message that emerges from this literature is that diversification is associ-
ated with inferior performance, although more recent work has challenged
this view, and some scholars have even suggested that diversification may
even be associated with superior performance (see the survey in Martin
and Sayrak, 2003).
A historical perspective provides a helpful context in which research
into diversification can be framed. In the 1950s and especially the 1960s,
diversification was actually a popular strategy, for several reasons. First,
and perhaps most important, antitrust law imposed limits on the market
shares of firms in specific industries, which meant that firms who were
willing to grow had to do so in new industries. Second, capital markets were
relatively undeveloped and firms had incentives to organize several busi-
nesses around an ‘internal capital market’ – this is also known as the ‘deep
pockets’ argument. Third, the multidivisional or ‘M-form’ organization
was growing in popularity. Fourth, there is evidence that early diversifica-
tion announcements actually received a positive stock market reaction,
which encouraged further diversification. As a result, the 1960s have been
described as a ‘wave of unrelated acquisitions’ (Montgomery, 1994, p. 170).
Unrelated diversification by acquisition became increasingly common
during the 1960s (Shleifer and Vishny, 1990), whereas in previous times
firms had been more wary of entering markets where their learned capabili-
ties did not give them a distinct competitive advantage (Chandler, 1992).
The 1970s were also characterized by unrelated acquisitions and overdi-
versification. The 1980s, however, have been associated with a ‘return to
corporate specialization’ (Bhagat et al., 1990). During this time, changes in
the business environment made diversification less appealing (in particular,
financial markets became more developed, and antitrust law changed its
stance on measures of absolute market share). Furthermore, increasing
attention was being drawn to the relatively poor financial performance
of large diversified conglomerates. As a result, the takeover wave of the
1980s can be characterized by firms refocusing on their core capabilities by
acquiring businesses in related activities. During this time, it was common
for ‘corporate raiders’ to acquire a large diversified conglomerate and to
sell on the constituent segments individually, often at a considerable profit
(Shleifer and Vishny, 1990).
In more recent times, conglomerates have continued to be painted in

a bad light. Research has found that diversified firms trade at a discount
when compared to non-diversified firms in the same industries – this
finding is commonly known as the ‘diversification discount’. Lang and
Stulz (1994) and Berger and Ofek (1995) compare a cross-section of multi-
segment to single segment firms and report that US conglomerates are
valued at a discount of around 15 per cent. Lins and Servaes (1999) find
evidence of a significant diversification discount in Japan and the UK,
although not in Germany. Other scholars have examined the performance
of diversifying firms via ‘event studies’ of stock market reactions to diver-
sification or refocusing. It appears that the stock prices respond negatively
to diversification announcements (see for example Hyland and Diltz,
2002) but positively to refocusing announcements (Berger and Ofek, 1999;
Markides, 1992). Others have analysed the effects of diversification on ex
post realized profits, again finding that diversification exerts a negative
pressure on profits (Doukas and Kan, 2004). Conversely, there is evidence
that corporate refocusing is associated with increases in ex post profits
(Markides, 1995).
The distinction between related and unrelated diversification has also
received attention from empirical work. Whilst unrelated diversification
is often detrimental to firm performance, related diversification seems to
be more successful. An early contribution by Rumelt (1982) suggested
that related diversification led to results superior to those associated with
either unrelated diversification or no diversification. As a result, despite the
negative tone of research into the performance of diversified companies, it
was suggested that the ‘optimal level of diversification’ for large firms was
above the minimum of one industry (Montgomery, 1994).
In recent years, however, the finding that diversifying firms are under-
achievers has been firmly contested. The main objection is that, even before
the diversification event, diversifying firms are often situated in mature
industries that are characterized by low returns, low investments in R&D,
and relatively high rates of exit (Campa and Kedia, 2002). Statistical
investigations should thus control for the endogeneity of the decision
to diversify, by taking into account the firm-specific characteristics that
affect both firm value and the decision to diversify. Empirical work taking
this endogeneity into account has found that the diversification discount
disappears, and that diversification can even be a value-enhancing strat-
egy for those firms that actually choose to pursue it (Campa and Kedia,
2002; Villalonga, 2004; see also Graham et al., 2002). In addition, several
theoretical contributions have shown how the decision to diversify can be
a value-increasing strategy for firms (Matsusaka, 2001; Fluck and Lynch,
1999; Maksimovic and Phillips, 2002).
9.3 INTERNAL GROWTH VS GROWTH BY

ACQUISITION
Both internal growth and acquisition can be used as means of either expand-
ing market share in a particular industry or of diversifying into new indus-
tries. Internal growth, also known as ‘organic growth’, is usually associated
with smaller firms who are not seeking to diversify. Davidsson and Delmar
(2006) present compelling evidence that the younger the firm, the more of
its total growth was organic. Growth by acquisition, on the other hand,
is usually associated with large diversified firms, who have the necessary
financial resources. For example, Maksimovic and Phillips (2008) analyse
plant-level US data and observe that acquisitions account for 36 per cent
of the growth of conglomerate segments in growth industries, whereas the
figure is 9 per cent for single-segment firms.
Internal growth is a preferable means of diversifying when there are
strong synergies between the firm’s existing activities and the target indus-
try. These synergies may take the form of reduced entry costs or reduced
operating costs, or both. Furthermore, internal growth is particularly
attractive if firms can develop and integrate their new capabilities in an
environment where time pressures are not too great. In this way they can
steadily cultivate a sound base of in-house competences that will be a source
of enduring competitive advantage. Internal growth is also a relevant
option when there are no suitable target firms available for acquisition at
a reasonable price.
Growth by acquisition of other businesses, on the other hand, is most
effective when a firm must rapidly acquire new capabilities, production
capacity or good managerial resources. Similarly, acquisition is a preferred
means of entry into industries in which market shares are already stable
and there is little space for a new entrant. Furthermore, acquisition is
more appropriate if synergies with the new activity are not expected to be
significant.
Growth by acquisition has the consequence of injecting ‘new blood’
into the organization, and acquiring firms face the challenge of success-
fully absorbing the new resources. Lockett et al. (2007) analyse a sample of
growing Swedish firms and observe that growth by acquisition is followed by
above-average organic growth. Their explanation is as follows. Acquisitions
effectively bring new knowledge and resources into the firm, and the resulting
diversity of resources can be a spur to further growth as managers become
aware of a wider set of opportunities for expansion. Organic growth, in con-
trast, usually involves pursuing the growth opportunities that are closest to
a firm’s existing operations. As a result, organic growth does not produce as
much new knowledge as growth by acquisition.
Nevertheless, a strategy of growth by acquisition is particularly diffi-

cult to make good. ‘There are more unsuccessful acquisitions than there
are successful ones’ according to John Harvey-Jones, former Chairman
of ICI (cited in Ansoff, 1987, p. 10). In reality, acquisitions are rather
expensive growth strategies. According to one (admittedly dated) esti-
mate, the typical premium paid by an acquiring firm is 10–30 per cent
above the market price of the acquired firm’s stock before the merger
(Mueller, 1969, p. 652). To this must be added the costs of assimilating
the target firm, in order to convert the ‘potential synergy’ into ‘realized
synergy’.
Acquisitions have been attributed a noble character by some econo-
mists because, in effect, they introduce an element of competition into the
‘market for corporate control’. The possibility of takeover can therefore
act as a disciplining device that gives incentives for management to run a
company with efficiency and due responsibility (see, for example, Marris,
1964). Consistent with this hypothesis, Maksimovic and Phillips (2008)
present evidence that plants acquired by conglomerate firms experience
increases in productivity after the acquisition. In reality, however, the
‘market for corporate control’ is very imperfect, takeovers are very rare,
and inefficient management can continue for long periods. The disciplining
device of takeovers is rather weak. In contrast, it seems that acquisitions
are often a source of inefficiency in the economic system – indeed, ‘quite a
bit of evidence points to the dominance of managerial rather than share-
holder motives in firms’ acquisition decisions’ (Shleifer and Vishny, 1997,
p. 747). For example, acquisitions may take place because managers act in
their own interests rather than those of the firm as a whole (Mueller, 1969).
This conflict of interests may arise if pay increases, bonuses, perquisites,
or prestige are associated with the size of the firm. In addition, managers
of mature firms (often having high cash flow but few growth prospects)
may choose to acquire businesses because they are reluctant to distribute
the earnings to shareholders (Jensen, 1986). Furthermore, managers may
undertake acquisitions because they are overconfident of their manage-
rial abilities – this is the essence of Roll’s (1986) ‘hubris hypothesis’. As
a result, empirical evidence suggests that ‘acquisitions, in general, have a
deleterious effect on company performance as measured by profitability’
(Dickerson et al., 2000, p. 424). Acquisitions may also be socially harmful
if a firm acquires a competitor as a way of obtaining market power in a
particular industry.
Reflecting upon the theoretical literature leads us to conclude that
acquisitions are more often than not associated with decreases rather than
increases in social welfare. The empirical evidence also seems to lean in
this direction.
9.4 GROWTH BY INTERNATIONALIZATION

Expansion into new geographic markets can be an attractive opportunity
for firm growth. This form of diversification can allow a firm to boost sales
of its products and may confer several advantages such as scale and scope
economies, more market power, diversification of revenues, and the ability
of spreading fixed costs such as R&D over a larger sales base.
The most basic mode of internationalization is to begin exporting
existing products into new markets. In some cases, products need not
be modified for the specificities of the export market; they can simply be
produced in greater quantity and shipped overseas. A firm can deal with
a distributor that is already in place in the foreign market, and deal with
this distributor at ‘arm’s-length’ via the market system. If this is the case,
‘[t]he highest synergy move is to offer abroad the firm’s traditional prod-
ucts or services.’ (Ansoff, 1987, p. 125).6 Several drawbacks to this kind
of exporting arrangement may arise, however. For example, there may be
problems concerning the relationship with the distributor, if the distributor
lacks the tacit knowledge required to act in accordance with the exporting
firm’s marketing strategy. The distributor may also ask high prices or may
not act in the best interests of the exporting firm. Furthermore, consum-
ers in the export market are likely to have different preferences, budgets
and consumption habits, and as a result they may not be interested in
the product being offered. While it is possible to modify the product for
the export market, additional costs and complexities may creep in during
such modifications, and they may eventually overwhelm the benefits that
initially motivated the decision to export.
Some of the agency problems and transaction costs that come between
the exporting firm and the distributor can be addressed by forming a
strategic alliance between the two parties. Strategic alliances can take a
number of forms, ranging from simple non-equity contractual agreements
to equity joint ventures. Alliances have the advantages of allowing firms to
overcome resource deficiencies, and also to benefit from outside knowledge
and experience to improve their existing capabilities. The exporting firm
can gain access to a partner’s distribution network and also benefit from
the partner’s market experience and privileged access to local information.7
However, even in strategic alliances there may still be arguments over the
division of control, goal conflicts or other problems due to lack of trust
or understanding. Furthermore, it may be easier to transfer tacit knowl-
edge within a firm’s own boundaries, rather than between two alliance
partners.
An alternative mode of internationalization is through foreign direct
investment (FDI), which can take the form of an acquisition or of greenfield
construction of new facilities. FDI presents several advantages, such as the

possibility of choosing a location with favourable labour markets or prox-
imity to other resources. Furthermore, if internationalization entails the
transfer of knowledge that is tacit, idiosyncratic and complex, then firms
may prefer to transfer this knowledge within the firm (by FDI) rather than
through a joint venture with another firm (Kogut and Zander, 2003). An
expansion plan involving FDI does entail a considerable resource commit-
ment, however, and so firms that face uncertainty and have limited knowl-
edge may be deterred from FDI. Political risk (such as the risk of forced
appropriation of facilities by host governments) will also discourage firms
from undertaking FDI. In addition, smaller firms may lack the managerial
and financial resources to undertake such projects.
A major obstacle that internationalizing firms encounter is the uncer-
tainty that stems from a lack of knowledge of foreign market conditions.
Although uncertainty is a feature in any strategy of firm growth (such as
acquisition or diversification), the uncertain conditions surrounding entry
into foreign markets are seen to be particularly severe. As a result, the
theory of firm growth by internationalization has traditionally emphasized
the gradual and time-consuming process of accumulation of knowledge
and experience. Johanson and Vahlne (1977, 1990) describe a ‘Process
Theory of Internationalization’ whereby boundedly-rational firms gin-
gerly consider internationalization opportunities and make increasing
commitments only after they gain knowledge and experience from their
existing overseas operations. In this view, exporting can be seen as a
learning opportunity, and as a first step that will be followed by increased
exporting, strategic alliances or maybe even the establishment of overseas
production facilities through FDI. Firms that embark upon an expansion
into new foreign markets may thus find that past internationalization can
be a spur to future internationalization, as they seek to benefit from the
knowledge they have acquired, and also as they try to spread the overhead
costs of international administration and coordination over a larger sales
base.
Recent years have witnessed the apparition of ‘born global’ firms,
however – SMEs with strong global ambitions. Early internationalization
will be especially attractive for small firms specializing in niche markets
for high-technology goods and services, who face insufficient demand
from national markets alone. For instance, some firms may have over
50 per cent of their revenues from international markets as from the first
year of operation (McDougall et al., 1994). The emergence of successful
young international SMEs has highlighted some of the limits of the process
theory of internationalization, such that a new stream of literature on
international entrepreneurship has emerged (see the seminal contribution
in Oviatt and McDougall, 1994; and also McDougall et al., 1994; Autio et
al., 2000 and Sapienza et al., 2006; see also Zahra and George, 2002 and
Jones and Coviello, 2005 for surveys of international entrepreneurship).
Although it is recognized that the accumulation of knowledge and experi-
ence is an incremental process, it has been argued that small firms are more
flexible and are able to learn rapidly, so they can expand in foreign markets
at a faster pace than older firms. SMEs may have established home-based
advantages (Kuemmerle, 2002) that they leverage in new markets. In this
view, SMEs can expand abroad by applying existing capabilities to profit
opportunities in new foreign markets. However, they may also treat their
expansion projects as learning opportunities, providing them with valu-
able knowledge and allowing them to augment their existing capabilities
(Kuemmerle, 2002; Zahra et al., 2000). Indeed, internationalization is a sig-
nificant event that has the power to change the firm itself. While older firms
tend to be more inert and are prone to being ‘locked in’ to the exploitation
of established capabilities, international SMEs may be more flexible and
also more receptive to knowledge gained from internationalization, using
their knowledge to develop their core capabilities further.
9.5 CONCLUSION
In this chapter we took a more descriptive approach to theorizing about

the processes of firm growth. In the first part (section 9.1), we focused the
‘demand’ for growth opportunities by discussing the attitudes of managers
to growth, and the advantages and drawbacks that accompany growth. In
the later sections we discuss the different modes of growth – growth by rep-
lication and growth by diversification (section 9.2), growth by acquisition
(section 9.3) and also growth by internationalization (section 9.4).
The degree of uncertainty varies considerably across these different
modes of growth. While in some cases growth can be achieved merely
by producing more of the same for sale in similar markets, in other cases
growth is only possible if risks are involved. In these latter cases, knowl-
edge of market opportunities is a crucial factor determining the success of
growth projects.
10. Growth of small and large firms
The growth of small firms is often seen as having a beneficent character,
often being taken as a goal for policy intervention. Small firms are often
portrayed as being dynamic and innovative, playing a key role in gen-
erating new employment opportunities. In contrast, it appears that the
growth of large firms is often implicitly put in a bad light – questions of
market power, unfair competition, or managerialist ‘empire-building’ are
frequently raised. (In our view, however, the growth of ‘good’ firms should
be encouraged and the growth of ‘bad’ firms discouraged – regardless of
whether these firms are large or small.)
An emphatic contribution in favour of small firms was made by Birch
(1987), who suggested that the majority of net new jobs in the US between
1968 and 1976 were created by firms with 20 or fewer employees. Birch
coins the term ‘gazelles’ to refer to high-growth small firms – firms that, he
argued, created a large share of new employment. Birch’s analysis has been
forcefully criticized by a number of authors, however. Prominent among
Birch’s critics are Davis et al. (1996), who identify a number of statisti-
cal problems in Birch’s analysis. It seems that whether or not small firms
generate the bulk of new jobs is very much dependent on the statistical
methodology employed (Davidsson and Delmar, 2006).
In our view, it is overly simplistic to view small firms as the main source
of new job creation. In reality, only a fraction of small firms are truly inno-
vative, their ability to generate jobs is limited, and the jobs they create often
disappear shortly afterwards. It might be better to characterize the entry
of small firms by phenomena of excessive entry, high exit rates, and a large
amount of waste of economic resources. Larger firms, on the other hand,
have the ability to generate jobs in large absolute numbers, and these jobs
appear to correspond to relatively stable positions. Furthermore, it has
been argued that the ability of large firms to diversify into new markets
helps to ensure that markets are reasonably contestable.1
Small and large firms differ in several fundamental ways that it would
not be appropriate to neglect. Indeed, the econometric analyses surveyed
in earlier chapters (for example in Chapter 4) suggested that small firms do
have different growth patterns from larger firms. The aim of this section
is to elaborate upon these differences. Whereas in previous chapters we
129
Phase
1 2 3 4 5
large
collaboration
coordination
“?”
Size of Organization
delegation
red tape
direction
control
creativity
autonomy
evolution: stages of growth

revolution: stages of crisis
leadership
small
young mature
Age of Organization
Source: Greiner (1998, p. 58).
Figure 10.1 Evolution and revolution in a model of growth stages
tended to consider firm growth as a number expressed in percentage terms,

in this chapter we aim to provide a more subtle and qualitative description
of the changes that occur in growing organizations. We begin by focus-
ing on the dichotomous distinction between small and large firms (sec-
tions 10.1 and 10.2), before taking a more detailed look at organizational
stresses that accompany the growth process in our discussion of the ‘stages
of growth’ models (section 10.3).
Firms that are small (large) very often correspond to firms that are
young (old). Some theories even posit a linear relationship between size
and age (see for example Figure 10.1 taken from Greiner, 1998). Although
this is not always the case,2 in the following, small (large) and young (old)
can be taken as more or less synonymous adjectives of firms.
10.1 DIFFERENCES BETWEEN SMALL AND

LARGE FIRMS
Small firms have the advantage of higher flexibility and responsiveness

both in terms of production technology and also in terms of organizational
Growth of small and large firms 131
structure. They benefit from efficient information flow, relatively quick deci-
sion-making and proximity to their customer base (You, 1995). As a result,
small firms may have an advantage in serving niche markets for specialized
products, whereas larger firms may be better adapted to catering for larger,
standardized markets. Small firms also tend to be less capital-intensive
(Caves, 1998; Bellone et al., 2008) or, similarly, more labour-intensive (You,
1995) than their larger counterparts. These factors notwithstanding, smaller
firms seem to be associated with lower productivity than larger firms (Idson
and Oi, 1999), especially when one considers small firms in developing coun-
tries (Little, 1987). One reason for this could be because small, young firms
tend to charge lower prices for the same goods (Foster et al., 2008).
In developed countries, entrepreneurial small firms have been ascribed
an important role in introducing new products and new techniques into
the market, through technological innovations (Pavitt et al., 1987; Acs
and Audretsch, 1990). Pavitt et al. (1987) observe that small firms in the
UK account for a share of major innovations that is disproportionately
large when compared to their R&D expenditure levels. Acs and Audretsch
(1990) analyse data on US firms and emphasize the innovative prowess of
small businesses. More specifically, small firms seem to play an especially
important role in highly innovative and skill-intensive industries which are
in early stages of their life cycles (You, 1995).
Even in developed countries, however, many entrepreneurs are not true
entrepreneurs, in the sense that they do not bring innovations or bring
about reform in stagnant markets (Santarelli and Vivarelli, 2007). Many
enter for less noble reasons, such as over-optimism on the part of the
founder, the pursuit of a relaxed lifestyle, or the flight from unemploy-
ment. In fact, many entrants are far less productive than incumbents (even
taking into account their liability of small scale). The concept of new firm
entry gathers together a particularly heterogeneous group of enterprises. In
many cases, and especially in developing countries, micro and small firms
exist because they offer individuals a livelihood and a source of independ-
ent revenue. In many cases, new small businesses are founded as a last
resort rather than as a first choice (Beck et al., 2005a). It has even been
argued that, in the case of India, better educated individuals are quick to
leave self-employment and pursue alternative career paths, whereas less
educated individuals have fewer opportunities to leave self-employment
(Nafziger and Terrell, 1996). Workers in such enterprises are relatively
badly paid, or may even be unpaid family members.
According to Penrose (1959), small firms can thrive in the ‘interstices’ of
major markets, in niche sub-markets that are not large enough to support
large firms. As a result, they are often sheltered from direct competition
with large firms. This is not to say that they are entirely protected from the
competition, however. In fact, survey evidence for small businesses indi-

cates that competitive pressures are a major factor inhibiting their growth
(Hay and Kamshad, 1994; Robson and Bennett, 2000).
Small firms also differ from larger firms in terms of their financial
account structure. Generally speaking, smaller firms have a precarious
financial structure and can only plan over a shorter time horizon. Hughes
(1997) analyses a sample of UK firms and reports that while long-term
loans account for only 20.5 per cent of all loans for small firms in the
manufacturing sector, they correspond to 61.7 per cent of all loans for
larger firms. In contrast, short-term liabilities may cover more than half of
a small firm’s assets (Sarno, 2008). Whereas larger firms can rely on funds
from equity issues, banks are by far the main source of finance for smaller
firms. In particular, small firms are more reliant on short-term loans and
overdrafts, and have a relatively high proportion of trade debt in their asset
structure (Hughes (1997)). Small firms must often pay for bank loans at a
premium, however, and during periods of tight money (‘credit crunches’)
bank lending to small firms tends to contract particularly rapidly. Smaller
firms tend not to pay dividends, but often rely on retained profits to fund
their investment projects. Internal funds are a crucial source of funds for
growing small firms (Allen et al., 2006).
Another robust difference between small and large firms is that large firms
tend to pay higher wages. Brown and Medoff (1989) analyse a number of
US datasets to investigate this relationship. They observe a positive asso-
ciation between employer size and wages, which holds at both the estab-
lishment and the firm level.3 Furthermore, they show that this employer
size–wage relationship remains significantly positive after controlling for a
number of other influences. Some possible hypotheses are that larger firms
hire higher-quality workers, that they offer less attractive working condi-
tions, or that they pay higher wages to stave off unionization attempts.
Even after controlling for these influences, however, larger employers are
still seen to pay higher wages. They also observe that employees tend to
remain with larger employers over longer periods of time.
10.2 DIFFERENCES IN GROWTH PATTERNS OF

SMALL AND LARGE FIRMS
It is fitting to begin this section on the growth of small and large firms with
a well-known passage written by Alfred Marshall:
[W]e may read a lesson from the young trees of the forest as they struggle
upwards through the benumbing shade of their older rivals. Many succumb on
the way, and a few only survive; those few become stronger with every year, they
get a larger share of light and air with every increase of their height, and at last
in their turn they tower above their neighbours, and seem as though they would
grow on for ever, and for ever become stronger as they grow. But they do not.
One tree will last longer in full vigour and attain a greater size than another;
but sooner or later age tells on them all. Though the taller ones have a better
access to light and air than their rivals, they gradually lose vitality; and one after
another they give place to others, which, though of less material strength, have
on their side the vigour of youth. (Marshall, 1961, p. 263; first edition published
1890)
Marshall is quite right in pointing out the perils that afflict small young
firms. The growth of small firms indeed involves a struggle to obtain
vital resources (the light and air in Marshall’s analogy), and many firms
will not survive. One direction in which Marshall’s analogy might need
further clarification, however, concerns the fact that not all firms strug-
gle for growth. In fact, many small firms don’t seem to want to grow. In
the remainder of this section, we discuss three key aspects of the growth
of small firms: the struggle to survive, desire to grow, and also structural
change in growing organizations.
10.2.1 The Struggle to Survive
The growth of small firms is a particularly chaotic phenomenon. Entry

rates of new firms are high, regardless of the industry, and a large number
of these entrants can be expected to exit within a few years. Although not
all cases of small firm exits can be considered as failure (Headd, 2003;
Harada, 2007), it is nonetheless clear that survival rates of new firms are
very low.
A significant early investigation was made by Phillips and Kirchhoff
(1989). They begin by observing that new small US firms have an average
survival rate of 39.8 per cent over six years. However, they observe that
the growth of these small firms has a strong impact on their chances of
surviving. For example, new small firms that added no employees over
the initial six-year period had a survival probability of only 26.0 per cent,
whereas firms that added one employee or more had survival rates of 65.0
per cent; firms experiencing the fastest growth, however, had survival rates
of as much as 77.5 per cent. In short, young firms that grew were observed
to have more than twice the probability of survival when compared to that
of young, non-growing firms. While initial size was observed to have an
impact on survival, this was overshadowed by the relationship between
growth and survival.
Subsequent research into survival rates has tended to confirm these early
findings. For example, Headd and Kirchhoff (2007) observe that survival
rates hovered at about 50 per cent after around four years, in their sample
of small US firms. Bartelsman et al. (2005) examine the post-entry perform-
ance of new firms in seven OECD countries and observe that about 20–40
per cent of entering firms fail within the first two years, while only about
40–50 per cent survive beyond the seventh year.
Although survival rates are lower than one might have hoped for, not all
cases of firm exit should be considered as failures. Headd (2003) presents
evidence that up to a third of business exits are considered by owner-
managers to be successful events, corresponding to cases such as planned
closures, sale of the business, or retirement from the work force. As such,
raw figures on survival rates may actually be understating the success rate
of new businesses, because they do not take into account these cases of
successful closures. Headd (2003) analyses data on new employer firms
in the US and shows that while half of these new firms can be expected to
survive after four years, as many as one-third of cases correspond to suc-
cessful closures, with the remaining two-thirds being failures. The failure
rate remains quite high, however, even when successful closures are taken
into consideration.
New firms enter on a small scale relative to that of incumbents – around
40–60 per cent of the average size of incumbents (Bartelsman et al., 2005).
Their small size puts them at a disadvantage vis-à-vis their larger coun-
terparts, and so they must expand rapidly, as if their life depended on it.
Wiklund (2007) explains that, for new small firms, ‘growth and survival
go hand in hand’ (p. 145). Garnsey et al. (2006) acknowledge that growth
creates problems, but they add that the problems accompanying growth
are less dangerous to a firm’s survival than the absence of growth. The
larger they grow, the smaller their cost disadvantage relative to firms above
the MES, and thus the higher their chances of survival. For such firms, the
growth objective coincides with survival and the pursuit of profits. These
firms tend to have a higher average growth rate than larger firms, despite
the difficulties they may face in financing their expansion.
Some influential theoretical models have attempted to describe the
chaotic process of small firms growing larger. Jovanovic (1982) presents
what is known as the ‘passive learning’ model, in which small firms have a
fixed, firm-specific productivity level. Their growth and survival prospects
are bound to this productivity variable. Although firms do not know how
productive they are upon entry, they learn about their relative productivities
once they have entered. It is shown that this model is able to account for the
faster growth and also the higher exit hazards associated with small firms.
Hopenhayn (1992) presents a similar model in which a firm’s productivity
level evolves in random fashion, according to a Markov process. Finally,
the ‘active learning’ model (Ericson and Pakes, 1995; see also Pakes and
Ericson, 1998) investigates the evolution of a competitive industry when

firms can influence their specific productivity levels by investing in R&D.
Empirical evidence presented in Coad (2007a) provides unique insights
into differences in the growth experiences of small and large firms. The
growth of small firms appears to be marked by a negative autocorrelation
which becomes extremely negative for the fastest-growing small firms. This
is consistent with observations on the erratic nature of growth for small
firms. Some small firms may grow exceptionally fast in one year, but they
are unlikely to be able to repeat this performance in the following period.
Larger firms, on the other hand, have a much smoother growth pattern,
with a small positive autocorrelation of one year’s growth onto the next. It
appears that larger firms enjoy greater stability and are able to plan their
growth over a longer time horizon.
The growth of large firms is indeed different in several respects. While
small firms’ survival depends to some extent on their growth, for large firms
above the MES the objectives of survival, growth and profits become sepa-
rated and may even conflict. Growth of large firms takes on a new meaning
as ‘economies of growth’ become more relevant than ‘economies of scale’.
Large firms may grow not because of any long-term strategic decision, but
because of opportunities that are attractive on the margin. (These oppor-
tunities may indeed turn out to be illusory!) If these firms grow to become
very large, they begin to resemble financial investment trusts composed of
relatively autonomous divisions (Penrose, 1959).4
Given that growth of small firms has beneficial effects on survival proba-
bility for small firms, the literature on small firms tends to view firm growth
as a measure of performance (see the discussion in Davidsson et al., 2008).
For larger firms, however, for whom matters of survival are less urgent,
growth is not always a good thing. In this vein, it has been recommended
that some firms should focus on making profits before pursuing further
growth. Davidsson et al. (2008) observe that firms that are generating high
profits are the most likely to move into the group of firms experiencing both
high profits and high growth in the following period. On the other hand,
firms with high growth rates but low profits are the most likely to transit
to the group of firms with both low growth and low profits. Along similar
lines, Coad (2008a) observes that the growth of high-productivity firms is
well received on the stock market, whereas the growth of low-productivity
firms generates a much smaller positive reaction.
10.2.2 Desire to Grow
Although research into small firms has placed great emphasis on fast-
growth firms,5 the majority of small firms do not experience fast growth.
Small firms have many excuses for staying at a small size. Lack of finance
could be one reason, problems finding customers could be another, or
problems finding honest and hardworking employees could be yet another.
In many cases, however, small firms don’t grow because they just don’t
have growth ambitions.
For many owner-managers of small businesses, the main objective is
independence. These individuals simply view their enterprise as instrumen-
tal in guaranteeing an independent and relaxed lifestyle where they can ‘be
their own boss’. Owner-managers may be more interested in developing
a personal friendship with existing customers than looking for new profit
opportunities. They may also be wary of relinquishing control of the firm
and delegating tasks to new employees. If they feel that growth would com-
promise their independence or the enjoyment of their work, then they will
not be motivated to grow (Wiklund et al., 2003). Once these firms survive
infancy, therefore, they may experience no further growth; and with the
passing of time, these firms may become increasingly inert and locked in
to their present scale of operations, letting growth opportunities fly away
from under their noses. These firms have been dubbed as ‘lifestyler’ firms
by some (for example Hay and Kamshad, 1994), while others have referred
to these small firms as the ‘living dead’ (O’Farrell and Hitchens, 1988, p.
1372).
While some firms aim for high growth, others lack the ambition. Growth
aspirations, one might suppose, play a major role in separating the busi-
nesses that grow from those that don’t. A number of researchers have
used questionnaire evidence to investigate the association between growth
ambitions and realized growth, finding the expected positive association
(Miner et al., 1994; Wiklund and Shepherd, 2003; Delmar and Wiklund,
2008). Wiklund and Shepherd (2003) show that while growth ambitions are
positively associated with growth, these positive effects are magnified when
interacted with factors such as the entrepreneur’s education and business
experience. Delmar and Wiklund (2008) observe that growth motivation
is positively associated with subsequent growth, and also, interestingly
enough, that realized growth feeds back to have a positive impact on subse-
quent growth motivation. As such, growth may be an acquired taste, such
that past growth tends to increase the desire for further growth.
10.2.3 Structural Change in Growing Firms
Chapin (1957) stands out in the literature as one of the most bizarre and
esoteric attempts at finding optimality in the growth of organizations.
Chapin analyses data on the two largest sub-groups (regular membership
and Sunday-school enrolment) in 80 Minneapolis churches. Rudimentary
statistical analysis cannot reject the hypothesis that the ratio of these two
sub-groups tends to the Fibonacci proportion of 0.6180. Chapin refers
to principles of harmony and symmetry of structure, which have been
observed in the growth of snail shells and sunflower seeds, to postulate
that organizations grow by a proportional scaling up of existing depart-
ments. More specifically, he asserts that it is the Fibonacci proportion that
governs the optimal logarithmic growth spiral of the relative share of the
two largest sub-groups within organizations.
Whatever the reason, Chapin’s model did not have a major impact upon
scholars of firm growth. One major shortcoming of his model is that firms
do not stay in the same proportions as they grow, but instead they undergo
tremendous restructuring. Small firms and large firms are very different in
structure, and they should not be considered as merely scaled down ver-
sions of larger firms. Small firms are not just scaled down versions of larger
firms. As small firms grow and become larger, their growth is accompanied
by considerable organizational stresses, which leads them to undergo
substantial transformations. Hannan and Freeman (1977) provide an ani-
mated analogy of the structural change that accompanies the growth of
organizations. They write that ‘a mouse could not possibly maintain the
same proportion of body weight to skeletal structure while growing as big
as a house. It would look neither like a mouse nor operate physiologically
like a mouse’ (Hannan and Freeman, 1977, p. 938).
An alternative model of organizational growth, briefly sketched out in
Andriani and McKelvey (2007), considers firm size in terms of the concepts
of surface and volume.
Employees dealing with people outside the firm are surface employees – they
bring in the resources from the environment. Volume employees are those inside
who produce and coordinate: they are resource users. As firms grow, then, they
have to maintain the square-cube ratio by adding more surface units, or by
making them more efficient. (Andriani and McKelvey, 2007, p. 1219)
According to this model, then, firm growth leads to uneven expansion of

its volume and surface. Firms that are increasing the volume of their output
should also heed the changing nature of interactions with the external
environment.
10.3 MODELLING THE ‘STAGES OF GROWTH’
As we have seen from the previous section, small and large firms grow for
different reasons and in different ways. Indeed, it has been observed that
the firm undergoes a radical metamorphosis as it grows, with the entrepre-

neur’s vision and dynamism gradually being replaced by a more bureau-
cratic structure. A body of research along these lines, guided by ‘common
sense views of youth, adolescence, maturity, and old age’ (Whetten, 1987,
p. 337), has culminated in theoretical models of regularities in the stages
of firm growth. The main thrust of these models is that the goals, priori-
ties and issues faced by firms change considerably along their respective
trajectories of development.
The ‘stages of growth’ models view firms as growing through successive
stages of roughly sequential ordering as they evolve from birth to matu-
rity. These stages correspond to configurations of problems, strategies
and priorities that firms are likely to face as they grow, as well as describ-
ing the level of owner involvement and the organizational structure. The
resolution of one set of problems allows a firm to enjoy a period of steady
growth and prosperity, but as the firm continues to grow it encounters
new difficulties. Typically, these models contain three to six stages of firm
development, with some models focusing in particular on the early stages
of firm growth. Although the unit of analysis is usually the firm, it could
also plausibly be taken to be a subsystem of a firm in the case of a mature
organization with loosely-coupled divisions.
A prominent and early contribution to this literature was made by
Greiner (1972).6 In Greiner’s model, presented in Figure 10.1, firms
progress through episodes of evolution and revolution, with growth stages
corresponding to a series of internal crises related to leadership, control
and organizational coordination. The resolution of one crisis is seen to sow
the seeds for the next crisis. Thus, a small young firm, characterized as a
creative enterprise, will have to deal with a crisis of leadership as it grows
too big to be managed single-handedly by the founding entrepreneur (see
Figure 10.1). If the firm succeeds in introducing a capable business manager,
it will typically enjoy a period of growth characterized as the ‘direction’
stage. However, a crisis of autonomy looms as employees are torn between
following procedures and taking their own initiative – this crisis is resolved
by promoting delegation in the context of a decentralized organizational
structure. As the firm puts delegation into practice, however, top manage-
ment may feel as though it is losing control. To deal with this control crisis,
the firm enters the ‘coordination’ phase as formal coordination systems are
introduced. These latter help to alleviate control problems but they create a
gap between headquarters and operating workers. This is the bureaucratic
‘red tape’ crisis, which occurs when the organization becomes too large
to be managed using rigid, formal techniques. Spontaneous managers
capable of creating teams and encouraging teamwork help the firm move
into the final stage, the stage of ‘collaboration’.
The theory of ‘cognitive leadership’ (Witt, 1998, 2000, 2007; Cordes et

al., 2008) is not a stage model per se, but it can be mentioned here because
it provides further insights into changes in small firms as they grow. This
vision of the growth of entrepreneurial firms describes the firm as a small
team centred around a visionary entrepreneur and his/her business concep-
tion. The task of the entrepreneur is to hire employees that will be effective
in carrying out the entrepreneur’s business plan. This business plan serves
as a ‘cognitive frame’ that guides classificatory and interpretative mental
activities and is largely tacit in nature, such that it is best communicated
through observation and direct communication with the entrepreneur. If
employees adopt the entrepreneurial business conception as their own cog-
nitive frame for their firm-related activities, a firm’s organization can attain
a higher degree of cognitive coherence among its members, which affects
the interpretation of information, the coordination of dispersed knowl-
edge, and individual endeavour, as well as the motivation to contribute to
a common goal instead of private interest. Furthermore, the workers may
be more motivated to perform well if they can identify with the entrepre-
neur’s business conception. When the firm is small, the entrepreneur can
readily share his or her vision and enthusiasm with the employees, through
regular communication and face-to-face contact. This helps the entrepre-
neur overcome contractual incompleteness, because employees can share
in the entrepreneur’s vision and use their understanding of the business
plan to overcome the ambiguities that may arise as they carry out their
tasks. As the firm grows, however, the entrepreneur spends less time with
each employee and the frequency of face-to-face interactions diminishes,
and cognitive coherence is thus no longer spontaneously achieved. The
employees’ attention may then be diverted to the pursuit of separate inter-
ests. Individualistic cognitive frames can then spread, favouring oppor-
tunistic behaviour. As the firm grows, it moves away from the cognitive
leadership regime, and organizational changes (such as closer monitoring
of the employees, or the introduction of a divisional structure) will need to
be introduced if the firm is to achieve profitable growth.
Churchill and Lewis (1983) also present a five-stage ‘stages of growth’
model, although their perspective is quite different. The five stages are
those of existence, survival, success, take-off and resource maturity.7 At
the existence stage, the young firm faces problems of obtaining customers
and delivering the product. The firm requires financial resources to take it
to the ‘survival’ stage, at which the firm must demonstrate the quality of
its personnel and operating efficiency. The following stage is the ‘success’
stage, at which the firm must decide whether it wants to expand or just
maintain the status quo. At this stage, the owner still has a considerable
degree of control over the business, but will forfeit this control if the firm
expands further. If the firm does not grow, it remains at what they call
the ‘success-disengagement’ stage. If the firm decides to grow, however,
it experiences a ‘take-off’ and must deal with issues of decentralization
and delegation before reaching the ultimate stage, ‘resource maturity’.
Churchill and Lewis (1983) also emphasize a fundamental transforma-
tion that takes place in growing firms – the fact that although the owner’s
abilities are important at the start of the enterprise, they become less so as
the firm becomes mature. Conversely, delegation is not important in small
firms but it becomes increasingly important as the firm grows. It follows
that the ‘inability of many founders to let go of doing and begin managing
and delegating’ (p. 42) is a major obstacle to the development and growth
of small firms.
The model developed by Garnsey (1998) bears some similarities to
that of Churchill and Lewis (1983), although it focuses more on the early
growth and development of new firms. Garnsey places emphasis on the
high hazard rates that confront new firms, and their effort and struggle
to quickly access, mobilize and deploy resources before they can generate
resources for growth. Once a firm’s operations are set up, however, the
initial burst of energy required to get things going is no longer required,
and resources are released for growth. Garnsey (1998) also discusses the
phenomenon of routinization of operations in small growing firms. To
begin with, ‘[n]ew firms are hampered by their need to make search proc-
esses a prelude to every new problem they encounter’ (p. 541). As time
goes by, however, firms learn about their business and develop problem-
solving repertoires that make demanding situations appear more routine.
Problems can be identified as recurrent and require less time and energy,
and ‘early challenges are replaced by repetitive grind’ (p. 542). As a con-
sequence, this routinization found in growing small firms can engender
disillusionment, and growth can be hindered by morale problems (which
may even lead to spin-outs of new ventures).
Garnsey and co-authors argue in favour of a process theory of growth
rather than a theory in discrete stages, however, because although certain
developmental patterns are common in new growing firms, firms face dif-
ferent challenges and deal with them in different ways (Hugo and Garnsey,
2005; Stam and Garnsey, 2005). Firms may struggle with the same recur-
ring problems, or they may face several issues at the same time. As such,
the growth of small firms is better described in terms of processes than
successive stages (Garnsey, 1998; Druilhe and Garnsey, 2006).
Although the ‘stages of growth’ models have largely escaped empirical
attention, it is worthwhile to mention here the work by Kazanjian and
Drazin (1989).8 The essence of their test is to observe how small new firms
evolve through four discrete growth stages – conception and development;
commercialization; growth; and stability. Firms are sorted into growth

stages by a self-categorization exercise in which CEOs were requested to
select from among four alternative, unlabelled organizational descriptions
that best described their firm’s current situation. Using a longitudinal sample
of 71 technology-based new ventures, they present evidence in support of
the sequential ‘stages of growth’ model, although the statistical evidence is
rather weak.9 Their results therefore suggest that, although the evolution of
firms along a ‘stages of growth’ schema is often observed, this schema does
not have strict deterministic or uni-directional properties because, in some
cases, organizations may revert to an ‘earlier’ set of problems.
There are, however, many sceptics of ‘stages of growth’ models. For
example, these models have often been criticized because they are too
deterministic, too simple, and because they have little predictive power
(Whetten, 1987). One particular group of discontents includes those who
affirm that organizational change is pervasive and continuous rather than
discrete and episodic. Tsoukas and Chia (2002), for example, dismiss the
notion of episodic change and argue that ‘[w]e should rather start from
the premise that change is pervasive and indivisible’ (p. 569). In this
view, change is viewed as a permanent feature of organizations, without
beginning or end, emerging from the complex interaction of individuals
within an organization and the evolving environment. Even organiza-
tional routines can be said to contain the seeds of change, because they
are performed by individuals who experiment and improvise as they apply
routines to novel situations (Feldman and Pentland, 2003). How then
can these two different views of organizational change be reconciled? Is
organizational change continuous or episodic? How can some sociologists
(for example Tsoukas and Chia, 2002) view change as a pervasive feature
of organizations whilst others (for example Hannan and Freeman, 1984)
seem to view organizations as being fundamentally inert? (To complicate
matters further, other authors take an intermediate position and view
organizational dynamics as occurring in a context of punctuated change
– see for example Sastry, 1997.) The survey by Weick and Quinn (1999)
focuses on precisely this question. For them, organizational change can
be either episodic or continuous, depending on the vantage point of the
social scientist. If we consider the entire life span of organizations, it is
possible to pick out certain points, describe the characteristics of these
points and compare them. On the other hand, a more detailed look reveals
‘all the subterranean, microscopic changes that always go on in the bowels
of organizations’ (Tsoukas and Chia, 2002, p. 580). ‘Stages of growth’
models, therefore, characterize organizational growth and change as epi-
sodic because they take a distant perspective of organizations and focus on
general trends in their long-term development over their life span.
10.4 CONCLUSION
One often gets the impression from the popular press that nothing is as
exciting as the rapid growth of successful small firms. Small growing firms
do not stay small for long, however. This chapter focused on the differences
between small and large firms, as well as the organizational transforma-
tions that accompany the growth as small firms become larger. We began
in section 10.1 by highlighting some of the differences between small and
large firms, before moving on to a discussion of how small and large firms
differ both in their attitudes to growth and also the characteristics of their
growth (section 10.2). Small firms must struggle to survive, and growth can
reduce their chances of exit. Many small firms are daunted by the prospects
of growth, and prefer to stay at a small size. Once firms start to grow,
however, their growth may awaken in them the desire for further growth.
In section 10.3 we presented the ‘stages of growth’ models. According to
these models, small entrepreneurial firms must deal with a number of dif-
ficulties (such as issues of delegation, monitoring and coordination) as they
move from one stage to the next, eventually becoming large-scale bureau-
cratic businesses. An advantage of these models is that they highlight the
fundamental transformations that accompany growth. Indeed, small firms
cannot be seen as simply ‘scaled-down’ versions of larger firms, as the
growth stage models vividly illustrate. The growth stage models also have a
number of drawbacks, however. Firms do not automatically progress from
one stage to another, but may face recurring problems, or several obstacles
at the same time. Empirical investigations into growth stage models have
drawn attention to the limited practical relevance of these models, if they
are taken too literally. The stage of growth models should therefore be
taken with a grain of salt.
11. Conclusion
What have we learned about firm growth? To conclude this book, we begin
by reviewing the main themes encountered (section 11.1) and make some
final comments concerning theoretical (section 11.2) and empirical (section
11.3) work. Section 11.4 concludes with a synthetic discussion about the
nature of firm growth, guided by the main findings of the book.
11.1 TAKING STOCK
Chief among the characteristics of firm growth rates, it would appear, is

that firm growth rates are remarkably idiosyncratic and that it is quite
difficult to generalize across the growth experiences of firms. It may
well be that, after reading this book, both the econometrician and the
theorist feel like tearing their clothes in frustration and wailing ‘random,
utterly random, everything is random!’ Although the random element is
indeed prevalent, it is nonetheless possible to find ways of identifying new
regularities. This can be achieved by applying well thought-out statistical
techniques whilst using an appropriate theoretical framework that pays
special attention to the context to which they are applied. In any case,
we believe that growth rates are not so random that there is no point in
looking at them.
We began the book by looking at the distributions of firm size, firm age
and firm growth in Chapters 2 and 3, before moving on to consider the
results of regressions that had sought to reveal the determinants of firm
growth (Chapters 4 to 7). One of the main results that emerged from this
literature, however, was that firm growth appeared to be characterized
by a predominant random element, with none of the candidate variables
being able to explain more than a fraction of the variance in firm growth
rates. Even firm size proved to be of little use in explaining differences in
firm growth rates – firms of the same size have very different growth experi-
ences. This idiosyncratic nature of firm growth proved to be a stumbling
block for theories of firm growth (surveyed in Chapter 8), because firm
growth is a rather singular phenomenon which is not particularly amenable
to broad theoretical generalizations.
143
It may be that the majority of the total variation in firm growth rates is
within firms over time (Geroski and Gugler, 2004; see however Davis et al.,
2006). As a consequence, it would make sense for future empirical work
to attempt to explain growth by referring to variables that vary more over
time within particular firms than they vary between firms (in the cross-
section) at any given time. Unfortunately, however, firm-specific variables
that display such properties are not easy to think of, and data on such
variables is even harder to obtain.
Our survey has also emphasized a number of surprising and perhaps
counterintuitive findings. For example, we saw in Chapter 5 that finan-
cial performance and productivity are poor predictors of growth. The
evolutionary mechanism of selection by differential growth does not
seem to work very effectively at all. Instead, selection appears to operate
mainly via the channel of exit – that is ‘survival of the fitter’ rather than
‘growth of the fitter’ – and this considerably reduces the power of selective
forces. Although there are strong implications hinging on the relationship
between relative ‘fitness’ (usually profits or productivity) and growth, there
is nonetheless a shortage of empirical research that has been conducted in
this domain. As a result, I feel obliged to reiterate Caves’ (1998) recom-
mendation: ‘Because reallocations of activity from the less efficient to
the more efficient are so important for the optimal use of resources, more
evidence is needed on how competitive conditions within an industry affect
the speed with which the more efficient displace the less efficient.’ (Caves,
1998, p. 1977). The nature of the relationship between relative performance
(whether it be financial performance or productivity) and firm growth is
indeed bewildering in terms of the conspicuous gap between theoretical
work and empirical findings. This is a puzzle that remains to be tackled.
We suggest that cohort studies, where cohorts of firms having the same age
are tracked over time, might be a useful approach because a firm’s attitude
to growth is assumed to vary over the life cycle.
Firms are surprisingly heterogeneous with regards to their propensities
for growth, and there is some evidence to suggest that, in several cases,
poorly performing firms are often more ambitious in their growth plans
than better firms. Business firms are certainly not homogeneous, rational
profit maximizers. What are the implications of this? Should inefficient
firms be discouraged, and efficient firms be encouraged to grow? How could
such a policy be operationalized? More work on this would be welcome.
Another puzzle in the literature concerns the link between innovation
and firm growth. While much theoretical work, as well as questionnaire
evidence from managers, stresses the crucial role of innovation in explain-
ing growth, empirical studies have not really picked up on this in a satisfac-
tory manner. One explanation for this discrepancy between theoretical and
Conclusion 145
empirical work may be because the standard regression approach, which

focuses on ‘the average effect for the average firm’, is ill-appropriate for
analysing a phenomenon by which a minority of firms will grow very fast
while the average firm will barely grow at all. The semi-parametric quan-
tile regression approach employed by Coad and Rao (2008) is much more
suitable in circumstances where firms are a priori heterogeneous. Coad and
Rao (2008) observe that innovation has little effect on the growth of the
‘average firm’, but that it is of much greater importance in explaining the
growth of the fastest-growing firms. This latter group, of course, makes a
disproportionately large contribution to the overall process of industrial
development. This perspective militates in favour of a vision of productiv-
ity growth and industrial progress that is less concerned with the average
firm in a population, than what is taking place among a handful of firms
operating at the frontier (Coad, 2008a).
11.2 THEORETICAL WORK
The broad theoretical predictions that were surveyed in Chapter 8 were

not particularly helpful in predicting a firm’s growth rate, presumably
because firm growth rates are characteristically random and, as a conse-
quence, it is difficult to generalize across firms. The theories we surveyed
are certainly diverse, and sometimes they are contradictory. For example,
while neoclassical theory considers that growth is only a means to an end,
Penrose considers that growth is an end in itself, and that it may occur
even if the firm is beyond an ‘optimal size’ threshold, in the case where
‘economies of growth’ of exploiting a marginal growth opportunity offset
the diseconomies of the resultant size.
It is also striking that the theories, though intuitively appealing, do
sometimes yield predictions that are quite false. The neoclassical proposi-
tion that firms grow in an attempt to reach an ‘optimal size’ is unhelpful
at best. The evolutionary principle of ‘growth of the fitter’ fails to receive
strong empirical support. Furthermore, the main prediction from the pop-
ulation ecology perspective (that is that firm growth should be modelled
by considering industry-specific components) seems rather weak when it is
subjected to the empirical test, because factors that are common to all firms
in an industry cannot explain the tremendous variation in growth rates of
firms within the same industry.
It seems that no single theoretical perspective can explain firm growth,
but that several theories are needed in order to shed light on different facets
of the phenomenon at hand. We also consider that Penrose’s resource-
based perspective is particularly rich with insights. In our view, it is
meaningful to follow Penrose and suppose that growth is not just a means
to obtain a certain size, but rather it is an end in itself, a constructive appli-
cation of spare resources. Indeed, in the presence of learning-by-doing and
dynamic increasing returns, a lack of growth would be akin to stagnation.
As a result, we consider the notion of an ‘equilibrium growth rate’ to be
closer to the truth than that of an ‘equilibrium size’.
We also argue in favour of descriptive theorizing, ‘appreciative’ theoriz-
ing in the spirit of Nelson and Winter (1982), where the objective is not on
mathematical formalism or obtaining testable hypotheses, but an earnest
desire to provide as realistic a description of the phenomenon as possible.
Theoretical models can also play a valuable role in explaining specific
aspects of firm growth. Theoretical modellers, we suggest, should try to
compare the predictions of their model with as many stylized facts as pos-
sible. A large number of regularities can be found in the growth of firms,
as this survey has testified. For example, a model of firm growth can be
evaluated by considering the shape of the growth rate distribution, for
instance, or by looking at growth rate autocorrelation. A theoretical model
can be better understood if as many implications as possible are tried and
tested, and even if in some dimensions the model does not agree with the
empirical facts, these shortcomings should be acknowledged. In this way,
the gains and limits of the model can be better appreciated, and room is
left for further progress.
We are less interested in theoretical models that merely attempt to
explain the firm size distribution, however – there are many other styl-
ized facts available that can be useful in evaluating the contribution of a
theoretical model. There are indeed many statistical mechanisms that can
explain aggregate distributions (Brock, 1999), but the richness of a model
rests on other criteria, such as its ability to reproduce a number of stylized
facts, or its ability to describe the dynamics of the economic system. In
addition, we are sceptical of elaborate models trying to reconcile empirical
behaviour with rational maximizing behaviour, because the plain truth is
that firms are not perfectly rational profit maximizers.
11.3 EMPIRICAL WORK
Perhaps the main message that seemed to emerge from this monograph,
and especially the survey of empirical work, was that growth rates
appeared to be remarkably random by nature, reflecting the existence of
strong idiosyncratic components in the statistical series of firm growth. The
challenge researchers face, of course, is to further our knowledge of firm
growth by making original contributions to this literature that correspond
Conclusion 147
to significant advances in our knowledge. New light on the matter can be

shed when firm growth is approached from a new and different angle. This
is a call for imaginative lines of attack for econometric investigations into
firm growth.
Our understanding of firm growth has made significant headway over
the last decade by finding regularities in firm growth rate distributions
(surveyed in Chapter 3) – a new stylized fact that appears to be particu-
larly robust and that has led to the construction of new theoretical models
(such as Bottazzi and Secchi, 2006a). The remarkable lumpiness of invest-
ment over time, as shown in Doms and Dunne (1998) is another example
of how new empirical findings in the form of basic statistics can lead to a
new impetus in research into firm dynamics. Empirical work building on
the lumpy nature of firm growth has made some interesting discoveries.
Quantile regression techniques, for example, have proven valuable because
they allow the researcher to go beyond the characteristics of the ‘average
firm’ to place special attention on the fastest growing firms – the small
minority of firms that contribute the most to overall industry turbulence
and economic development (Coad and Rao, 2008). Other researchers
take the lumpy nature of firm growth into account by trying to predict
the probability of a large growth event such as an investment spike (for
example Whited, 2006). Progress has also been made by prescribing condi-
tional strategies for heterogeneous firms. For example, empirical work has
suggested that profitable firms should undertake new expansion projects
whereas firms with poor performance should not (Davidsson et al., 2008;
Coad 2008a). Other progress has been made by taking into account the
different modes of growth available to firms, considering for example the
differences between internal growth and growth by acquisition (Lockett
et al., 2007; Maksimovic and Phillips, 2008). These ideas can be taken
further, for example by investigating the impact on growth of other discrete
growth events (such as investment spikes, the construction of new plants,
diversification, entry into new export markets). Finally, progress has also
been made by decomposing firm growth into some of its constituent ele-
ments (such as sales growth, employment growth and growth of profits)1
and looking at the interactions of these variables over time as firms grow
(as surveyed in Chapter 5, section 5.3).
It is also fitting for us to begin by making a statement with regard to
validity of Gibrat’s law. The question of whether or not we should reject
Gibrat’s law has indeed been hotly debated. Whilst Mansfield (1962),
for example, voiced strong opposition to Gibrat’s law, Ijiri and Simon
(1964) take a much more favourable approach. These latter consider that
although Gibrat’s law does not hold with perfect accuracy, it is a useful first
approximation, just as Galileo’s law is approximately correct in describing
the motion of balls rolling down inclined planes (albeit without taking into
account such factors as friction, air resistance and magnetic fields). This
seems to us to be a sensible position to take.
Gibrat’s law has indeed proven to be a very useful model of firm growth,
and has spawned a truly vast empirical literature that seeks to determine
whether or not Gibrat’s law holds in any one particular database. It seems
to us that further work that aims to test Gibrat’s law is not the most fruitful
avenue of further progress, however. Instead, we recommend that research-
ers look for ways to gain new insights into firm growth, equipped with a
solid grasp of econometric techniques, and – more importantly – being
driven by a curiosity and an imagination that comes from a genuine desire
to glimpse further into the obscure cloud of confusion that surrounds the
subject of firm growth. In order to make progress in this field, therefore,
we feel obliged to reiterate an exhortation that is dated but nonetheless
still very relevant: ‘The subject of organizational growth has progressed
beyond abysmal darkness. It is ready for – and badly needs – solid, system-
atic empirical research directed toward explicit hypotheses and utilizing
sophisticated statistical methods’ (Starbuck, 1971, p. 126).
We wrap up by, once again, arguing in favour of Herbert Simon’s (1968)
research strategy, which emphasizes the need for solid empirical work to
first produce the ‘stylized facts’ that theory can then attempt to explain. At
this stage, we consider that research into the growth of firms could benefit
greatly from gathering of statistical regularities and ‘stylized facts’. We
consider that theory without any solid empirical basis – what we might call
‘armchair axiomatics’ (Dosi, 2004) – will be of little use in furthering our
knowledge of the growth of firms and the evolution of industries.
11.4 THE NATURE OF FIRM GROWTH
One of the more useful theories of firm growth was the descriptive theory
formulated by Edith Penrose in her celebrated book in 1959. The essence
of Penrose’s vision was that firms will always have internal resources for
growth because of learning-by-doing effects and, more specifically, the
freeing up of managerial attention as managers become increasingly accus-
tomed to their tasks. Unless the firm decides to grow, however, and unless
it chooses to make use of these spare resources, it appears to us that these
newly-liberated managerial resources will be absorbed as organizational
slack. Firms need to decide on the direction into which they can channel
these excess resources.
Growth can be seen as an entrepreneurial venture, no matter how large a
firm is. It is the quest for new opportunities. Growth requires imagination
Conclusion 149
and managerial involvement – managers must be alert, they must be

actively looking for growth, and planning for it. In order to succeed in their
growth projects, firms need to know themselves at the present moment (not
what they want to be, nor what they were like when they started, nor what
they were like a few years ago), and they need to have good knowledge of
the strengths and weaknesses of their existing resource base. They also need
to be aware of market developments and be able to recognize a business
opportunity when it arrives, amidst all its ambiguity and idiosyncracy.
As such, growing firms need to be masters of both the inside and outside
worlds.
Growth can be considered to be a dissatisfaction with the present scale
of operation. Growing firms must have a vision that extends beyond
their present situation, and look outward for new opportunities, thereby
embarking upon a venture into the unknown. Indeed, growth requires a
certain audacity – or, perhaps, a certain ‘ego’ (Gartner, 1997, p. 67). While
low-profit firms may be able to improve their circumstances through
growth, their poor past performance offers little support to their projects.
High-profit firms, if they desire to grow, must look beyond their satisfac-
tory performance and take a chance, without holding back out of fear of
compromising their past success. As a result, high-profit firms may not be
willing to take this risk. This may be why we observe no net effect of profits
on firm growth in Chapter 5.
In contrast, we observed that the influence of growth on profits tended
to be more important than the influence of profits on growth. This is con-
sistent with what Starbuck (1971, p. 74) calls the ‘will-o’-the-wisp’ models
of growth. According to these models, there are temporary gains that lure
firms to grow. For instance, it has been noted that there is a considerable
time lag between increases in productive capacity and the commensurate
additions to managerial resources (Starbuck, 1971, p. 54) or administra-
tive overhead (Dixon, 1953). Relatedly, Penrose speaks of these short-lived
gains in terms of her ‘economies of growth’. Firms may choose to expand
to a considerable size, even in the absence of economies of scale, simply
because there may be short-term gains from marginal growth opportuni-
ties that may be present at every step of a firm’s growth. (Clearly, we are
far from a rationalist optimal-size framework here.)
Although growth opportunities may well be available to imaginative and
enterprising managers, not everyone will take them up. It may be relevant
to endorse such a motto as ‘who dares grows’. Some managers may not be
willing to take the risks associated with expansion. At the other extreme,
it appears from our studies of autocorrelation dynamics in Chapter 4
that firms that attempt to grow too fast will not succeed. It appears that
growth requires a certain amount of time for previous growth events to be
properly internalized and ‘digested’, resulting in limits to growth within

time periods. Furthermore, we observe that in the majority of cases,
success in past growth does not in any way guarantee success in future
growth. ‘Learning to grow’ advantages do not play a significant role on
average. Growth opportunities are very different one from the other, and
they seem to be sufficiently heterogeneous that success with past growth
confers no great advantage with future expansion plans (this seems to be
especially true for all but the largest firms). Instead, it may well be the case
that past success can count against the firm, which risks becoming compla-
cent or having its cognition dulled by illusions of repetition. This point is
clearly illustrated by research into growth by acquisition, which finds that
past acquisition success has no clear effect on future acquisition success,
and may even be a liability (Haleblian and Finkelstein, 1999; Zollo and
Singh, 2004). It appears to us that future growth concerns the taking up
of opportunities that are, in some sense, new; growth involves challenges
that have not been faced by the firm previously in this particular form. This
is just as true for the small firm that ventures into new local markets as it
is for the diversified multinational that launches a new product in a new
country. This conception of growth is particularly evident in the ‘stages of
growth’ models surveyed in section 10.3, where growth occurs by resolving
one organizational crisis by introducing reforms that will, in turn, lead to
the arrival of a new crisis. In still other cases, a fortunate firm may, through
investment in innovation, happen upon a valuable discovery which propels
it into the fast-growth category (as in Coad and Rao, 2008). The common
theme here is that firm growth is an uncertain undertaking, and perhaps it
is the antithesis to the organizational routine.
We saw in Chapter 10 that many small firms don’t seem to want to
grow. They may find the uncertainty daunting and the challenge too great.
Growth in itself is often seen as stimulating and exciting, especially when we
hear media reports of young, high-tech, born-global firms. But the modes
and consequences of growth are often less attractive. Owner-managers lose
control of their firms as they grow, and the success of these firms is increas-
ingly due to employees, many of whom the original entrepreneur has little
time to meet. The firm grows further and becomes a bureaucratic organi-
zation; it turns from a small team into an impersonal economic instru-
ment, bought and sold on the stock market, a legal person of its own for
which the owners have only a limited liability. Firm growth does not stop
there, however. For large firms, it seems that they can’t get enough of firm
growth. For example, when US antitrust legislation limited the possibili-
ties of growth within industries by placing limits on the total market share,
firms began to grow by acquiring businesses in unrelated industries, even if
there was not much economic rationale behind this behaviour (Shleifer and
Conclusion 151
Vishny, 1990). Growth of conglomerates by acquisition and diversification

was common over the last decades, even when these growth strategies were
frowned upon by shareholders and economists. Firm growth may thus be
addictive in the sense that once a firm has tasted growth, it wants more
and more (Delmar and Wiklund, 2008). Larger firms, it seems, may have
difficulties resisting the challenge of becoming bigger. The proper alloca-
tion of growth opportunities among firms, according to the efficiency with
which these firms will use them, remains a significant challenge for the
economy.
Notes
CHAPTER 1
1. Note also that, in a very small number of cases, value-added can take
on negative values. This would be the case, for instance, of a firm that
sells goods for less than the cost of labour.
CHAPTER 2
1. This logarithmic approximation is only justified if et is ‘small’ enough

(i.e. close to zero), which can be reasonably assumed by taking a short
time period (Sutton, 1997).
2. Note, however, that if reversion to the mean is observed (i.e. that small
firms grow faster than large firms) then the variance of a firm growth
process operating through multiplicative shocks need not approach
infinity (Hart and Pearce, 1986).
3. The skewed age distribution in the sample of small firms in Coad and
Tamvada (2008) provides a unique illustration of the fact that not all
small firms are young.
4. This condition is trivial since the duration of a Gibrat-type ‘shock’ can
be made arbitrarily short.
CHAPTER 3
1. The observed Subbotin b parameter (the ‘shape’ parameter) is signifi-

cantly lower than the Laplace value of 1. This highlights the impor-
tance of following Bottazzi et al. (2002) and considering the Laplace
as a special case in the Subbotin family of distributions.
2. Growth rates calculated by taking log-differences correspond to log
growth rates, as explained in section 1.3 in Chapter 1.
3. This model made earlier appearances as Bottazzi and Secchi (2003b)
and Bottazzi and Secchi (2003c).
4. Here are a few possible examples. Slack may be present because indi-
visibilities of key inputs may prevent a firm from attaining perfect
152
Notes 153
productive efficiency. Also, slack may creep in as the learning-by-doing

effects that increase a worker’s productivity are not counterbalanced
by increasing demands made of the worker. Furthermore, slack may be
necessary because firms must be able to adapt and act flexibly in response
to unforeseen contingencies and the changing market environment.
5. Penrose writes ‘[a]t all times there exist, within every firm, pools of unused
productive services and these, together with the changing knowledge of
management, create a productive opportunity which is unique for each
firm.’ (Penrose, 1960, p. 2). Similarly, Lesourne writes ‘L’entreprise
cherchera à employer ces ressources inutilisées, mais en le faisant en
créera d’autres, en ne réussissant jamais à atteindre un état d’équilibre
complet dans l’utilisation de ses resources’ (Lesourne, 1973, p. 92).
6. A similar story could be imagined for growth after the arrival of an
innovation, since the innovating firm will typically have to invest
in a wide range of complementary assets in order to profit from the
innovation (Teece, 1986; see also Coad and Rao, 2008).
7. We do not need to define the number ‘large’ nor define what happens
at the very top of the hierarchy. Also, we do not need to suppose that
the number of hierarchies tends to infinity, because we only want to
explain the distribution of growth rates for a certain limited range. An
implication of this assumption is that this model is not suitable for
describing growth processes in very small firms.
CHAPTER 4
1. This logarithmic approximation is only justified if et is ‘small’ enough

(i.e. close to zero), which can be reasonably assumed by taking a short
time period (Sutton, 1997).
2. We should be aware, however, that ‘mean-reversion’ does not imply
that firms are converging to anything resembling a common steady-
state size, even within narrowly defined industries (see in particular the
empirical work by Geroski et al., 2003 and Cefis et al., 2007).
3. See for example Cooley and Quadrini (2001), Gomes (2001), Clementi
and Hopenhayn (2006), and Rossi-Hansberg and Wright (2007).
CHAPTER 5
1. This section draws on the survey in Coad (2007c).

2. The mainstream literature focuses on the relationship between current
financial performance and investment, although it does not elaborate
upon the distinction between replacement investment and expansion-

ary investment. The author is not aware of any relevant empirical
work that distinguishes replacement investment and expansionary
investment. For the purposes of the present discussion, we place more
emphasis on the latter when we speak of ‘investment’. In any case, the
distinction between the two may not be very clear-cut in the first place,
especially when we consider that firms tend to replace their exhausted
capital stock with more recent vintages (Salter, 1960).
3. Fazzari et al. (1988a) had originally intended to study small firms, as
is evident from the following quote: ‘Conventional representative firm
models in which financial structure is irrelevant to the investment deci-
sion may well apply to mature companies with well-known prospects.
For other firms, however, financial factors appear to matter in the
sense that external capital is not a perfect substitute for internal funds,
particularly in the short run’ (Fazzari et al., 1988a, p. 142). However,
given the requirement to obtain observations on market value (for
calculating q), the final sample contains only listed firms. This may be
somewhat inappropriate, because these firms have already reached a
certain size.
4. Note that Whited (2006) uses cash flow as a proxy variable for
investment opportunities.
5. David Packard, of Hewlett-Packard, relates how he was reluctant to
become dependent on external sources of finance: ‘I often helped my
father in looking up the records of those companies that had gone
bankrupt. I noted that the banks simply foreclosed on firms that mort-
gaged their assets and these firms were left with nothing . . . The firms
that did not borrow money had a difficult time, but they ended up with
their assets intact and survived . . . From this experience I decided our
company should not incur any long-term debt. For this reason Bill
[Hewlett] and I determined we would operate the company on a pay-
as-you-go basis, financing our growth primarily out of earnings rather
than by borrowing money’ (Packard, 1995, p. 85).
6. A wealth of evidence on this topic is provided in Beck et al. (2005b).
In particular, they observe that while financial constraints can be sig-
nificant for small firms in developing countries, they are not important
for large firms in developed countries (see also Angelini and Generale,
2008).
7. Consider the following example taken from a key reference on the
topic: after observing that investment–cash flow sensitivity is higher
in the UK than for several other European countries, Bond et al.
(2003b) arrive at the conclusion that ‘the market-oriented financial
system in the United Kingdom performs less well in channeling
Notes 155
investment funds to firms with profitable investment opportunities’ (p.

162, emphasis added). These authors simply presume that the forgone
investment opportunities would have been ‘profitable’, although in
fact this counterfactual presumption has no empirical basis. I argue
that this kind of speculation emerges from working too closely with
theoretical models in which firms are modelled as perfectly rational
profit-maximizers.
8. See also Fazzari et al. (2000) for a reply, and also Cleary et al. (2007)
and Guariglia (2008) who attempt to reconcile these two groups of
authors (that is, Fazzari, Hubbard and Peterson, and Kaplan and
Zingales).
9. Unpublished calculations on French and US data show that the effect
of profits on growth tends to increase in magnitude when firms of dif-
ferent sizes are given weights corresponding to their share of overall
activity. This is consistent with the interpretation that small firms’
behaviour is particularly erratic, while the behaviour of larger firms is
more sensible.
10. Cash flow can be defined simply as ‘an ambiguous term that usually
means cash provided by operations’ (Horngren, 1984, p. 776). More
specifically, the difference between cash flow and gross operating
income is a question of adding taxes and removing depreciation and
amortizement. Bougheas et al. (2003) use net profit as a proxy for cash
flow. Other studies (for example Bond et al., 2003b) build their cash
flow variable from an operating margin variable, by subtracting taxes
and adding depreciation.
11. One difference between the ‘managerial’ and ‘free cash flow’ perspec-
tives and the evolutionary perspective, however, is that the observed
investment–cash flow sensitivities are signs of value-reducing invest-
ment in the first case but are more likely to be value-creating in the
latter.
12. de Meza and Webb (1999) even go on to suggest that entrepreneurs
should be given incentives not to enter, or that they should be taxed if
they do enter.
13. The Small Business Administration (SBA) provided $2.8 billion in
guaranteed loans to small firms in 1986 alone.
14. Notwithstanding this latter result, Bottazzi et al. (2008b) observe a
strong positive relationship between productivity and profitability.
15. Note, however, that for the special case of industries facing sharp
decline, the usual positive association between poor performance and
exit may not hold. Baden-Fuller (1989) investigates exit dynamics in a
rapidly declining industry and observes that, in fact, it is high profit-
ability that predicts exit. The reason, he suggests, is that high-profit
firms can readily afford to pay the exit costs and promptly close down
the plant, and also because the higher abilities of the managers have a
higher opportunity cost and can quickly be put to good use elsewhere
in the economy.
16. Alternative schemes of decomposing productivity growth can be
found in Griliches and Regev (1995), Olley and Pakes (1996) and Aw
et al. (2001).
17. Traditional productivity indicators measure output in terms of total
sales rather than production, which makes establishments that charge
higher prices appear more productive.
18. Conventional econometric techniques that are applied to investi-
gate the causality between variables rely on instrumental variables.
Instrumental variable techniques may not be very effective in this
particular case, however, because firm growth is notoriously random
and it is unlikely that a suitable instrumental variable can be found.
For instance, in many applications of panel data instrumental variable
estimators, lagged variables are taken as instruments. It would hardly
be appropriate to take lags as instruments in the case of growth rates,
however, because of the low autocorrelation in growth rate series.
19. Due to the data construction procedure (growth rates calculated by
taking log-differences), firms with negative profits cannot be included
in the analysis.
CHAPTER 6
1. See, however, Klette and Kortum (2004) for a theoretical model in

which innovation is not correlated with firm growth.
2. However, it is reasonable to assume that the time lag from innova-
tion to superior firm-level performance is shorter when this latter is
measured in terms of stock market valuation – this line of reasoning is
pursued in Coad and Rao (2006).
3. They measure a firm’s innovative activity by either the discovery
of NCEs (new chemical entities) or by the proportion of patented
products in a firm’s product portfolio.
4. Patenting is an effective means of protecting innovations in the phar-
maceutical industry, for example, although it is not very effective in
the steel, glass or textile industries (Cohen et al., 2000). Therefore, it
is problematic to compare one patent for a pharmaceutical firm with
one patent for a steel, glass or textile firm.
5. Note that our use of the word ‘innovativeness’ does not correspond to
Mairesse and Mohnen’s (2002) use of the same word.
Notes 157
CHAPTER 7
1. For a survey of this work, see for example Carl Shapiro’s confidently-
titled article, ‘The theory of business strategy’ (Shapiro, 1989).
2. Other less common candidate variables for inter-firm competition
include import penetration (e.g. Haskel et al., 2007) and ‘profit elastic-
ity’ (Boone et al., 2007).
3. That is, when the combined market share of these two firms is greater
than 80 per cent.
4. It is interesting to also read onwards in Schumpeter (1942). The longer,
and more complete quotation, is as follows: ‘The businessman feels
himself to be in a competitive situation even if he is alone in his field
or if, though not alone, he holds a position such that investigating gov-
ernment experts fail to see any effective competition between him and
any other firms in the same or a neighboring field and in consequence
conclude that his talk, under examination, about his competitive
sorrows is all make-believe.’
5. When firms are pooled together in a cross-section, unionized firms
have lower expected growth rates, although this could simply reflect
the fact that unionized firms are more likely to be found in low-growth
sectors. When controlling for other influences (such as industry effects
and business cycle activity) the negative association between unioniza-
tion and firm growth disappears, such that, for a given firm, unioniza-
tion may not have any impact on its subsequent growth rates (Bronars
and Deere, 1993).
6. Pavitt (1984), Malerba (2002), Malerba and Orsenigo (1997) and a
number of other scholars have emphasized that innovation regimes
are sector-specific and that innovative activity undertaken by firms
varies considerably across sectors. Empirical evidence in Leiponen and
Drejer (2007) and Srholec and Verspagen (2008), however, finds that
there is a considerable amount of heterogeneity of firm-level innova-
tion strategies even within specific sub-sectors. The importance of the
sector in explaining innovative activity is therefore a matter of debate.
CHAPTER 8
1. One might see a resemblance here with some theories to be found in the
Vatican, which consider that people only have sex because they intend
to reach an ‘optimal’ family size . . .
2. Jacques Lesourne puts it this way – ‘L’entreprise cherchera à employer
ces ressources inutilisées, mais en le faisant en créera d’autres, en
ne réussissant jamais à atteindre un état d’équilibre complet dans

l’utilisation de ses resources’ (Lesourne, 1973, p. 92).
3. Winter writes ‘routines clearly qualify as resources, given the expan-
sive use of the term “resources” in the literature of the resource-based
view. . . . a routine in operation at a particular site can be conceived as
a web of coordinating relationships connecting specific resources . . .’
(Winter, 1995, pp. 148–9).
4. Penrose’s analysis considers that firms operate in a world of constant
returns to scale.
5. For a more complete reappraisal of Edith Penrose’s contribution to
economics, the reader is referred to Pitelis (2002).
6. Commenting on the contemporary business climate of the 1960s, when
managerial theories were first hatched, Mueller (1969, p. 644) ventures
to say that ‘[m]anagerial salaries, bonuses, stock options, and promo-
tions all tend to be more closely related to the size or changes in size of
the firm than to its profits’ [emphasis added].
7. This quadratic specification of the relationship between profits and
growth, however, has not been investigated in the empirical literature
in a satisfactory way.
8. Somewhat more far-fetched is Milton Friedman’s (1953) reiteration of
Alchian’s original idea, which supposes that the mechanisms of growth
of the fitter and exit of the weaker will lead the economy to the neo-
classical ‘optimum’, thereby vindicating the predictions of neoclassical
theory.
9. See also Metcalfe (2007), who generalizes his model of industry
evolution by attributing different propensities to grow to different
firms.
10. Financial performance and relative productivity are indeed closely
related (Bottazzi et al., 2008b), although some authors suggest that
selection operates on financial performance rather than productive
efficiency (Foster et al., 2008).
11. There is ample evidence that the population ecology perspective explic-
itly acknowledges interorganizational heterogeneity. For example, in
the seminal article by Hannan and Freeman (1977, p. 956), they write
‘[f]or us, the central question is, why are there so many kinds of organi-
zations?’ Furthermore, Hannan (2005) opens his literature review with
this very same question.
12. As Geroski (2001, p. 535) notes, there is a ‘heavy reliance on density
dependence to drive dynamics.’
13. Organizational heterogeneity is usually modelled using variables such
as age, size and organizational form.
Notes 159
CHAPTER 9
1. The ‘entry deterrence’ argument is of limited relevance, because
entrants are usually too small to pose a serious threat. However,
the argument may hold as long as large firms in other industries are
deterred from diversifying into the sector under consideration.
2. Using survey evidence for Dutch SMEs, Lensink et al. (2005) observe
that higher growth firms perceive that they have more idiosyncratic
uncertainty than other firms.
3. In fact, it is precisely because of the intentionality attributed to the
growth of firms that Penrose (1955) rejects biological analogies as valid
descriptions of firm growth.
4. An unpublished comparison of sectoral growth rate distribution
parameters (at the 3-digit level) for Italy and France reveals that there
is very little in common in the growth rate distributions for same
sectors across countries. This hints that the underlying sector-specific
production technology does not go far in explaining growth rates –
instead it may well be that human factors play a major role.
5. For empirical evidence on the heterogeneity of firm productivity levels,
even within narrowly-defined industrial sectors, see Dosi and Grazzi
(2006). See also Dosi (2007) for evidence on the dispersion of profit
margins within industries.
6. A similar conclusion is reached in Chandler (1992, p. 94).
7. There is evidence that joint ventures undertaken with firms from the
host country seem to do better than alliances undertaken with firms
from the same country, or with firms from a third country (Lu and
Beamish, 2001).
CHAPTER 10
1. Bain, quoted in Penrose (1959, p. 256).

2. It may be that small firms are nonetheless relatively old, if they have
a history of aversion to growth or if they face powerful obstacles to
growth.
3. The relationship between employer age and wages is less clear-cut,
however (Brown and Medoff, 2003).
4. These firms have a decentralized structure because the firm is too large
for the top management to play an active role in the activities of each
division. This decentralized structure has been observed to facilitate
spin-offs of the weakest divisions (Penrose, 1959).
5. See for example Davidsson et al. (2008) for a discussion.
6. Reprinted as Greiner (1998).

7. The five-stage model in Scott and Bruce (1987) is inspired by the
Churchill and Lewis (1983) model and bears a number of similarities.
The five stages in the Scott and Bruce (1987) model are inception; sur-
vival; growth; expansion and maturity. Churchill (1996) also presents
a similar model of firm growth, this time comprising six stages. These
stages are conception/existence; survival; profitability/stabilization;
profitability/growth; take-off, and maturity.
8. For another example of empirical research into ‘stages of growth’
models, see Mitra and Pingali (1999). These authors apply the
Churchill and Lewis (1983) model to an analysis of 40 automobile
ancillaries in India.
9. The authors use the ‘del’ statistic, which is preferable to the c2 statistic
because it tests for directionality. They obtain a del statistic (analogous
to the R2 coefficient) of 0.65 (with p , 0.001). In other words, knowing
the ‘stages of growth’ rule (whereby firms advance 0 or 1 stages over
an 18-month period) leads to a 15% proportionate reduction in error
over not knowing the rule in predicting stage transitions.
CHAPTER 11
1. A further challenge would be to try to describe how investment spikes

are related to changes in other firm-level variables such as sales and
employment, and productivity.
Bibliography
Acemoglu, D., P. Aghion, C. Lelarge, J. Van Reenen and F. Zilibotti
(2007), ‘Technology, information, and the decentralization of the firm’,
Quarterly Journal of Economics, 122(4), 1759–99.
Acs, Z.J. and D.B. Audretsch (1990), Innovation and Small Firms,
Cambridge, MA: MIT Press.
Adamic, L.A. and B.A. Huberman (1999), ‘The nature of markets in
the world wide web’, SSRN Working Paper IEA5. DOI: 10.2139/
ssrn.166108.
Aghion, P. and P. Howitt (1992), ‘A model of growth through creative
destruction’, Econometrica, 60(2), 323–51.
Akerlof, G.A. (1970), ‘The market for “lemons”: quality uncertainty and the
market mechanism’, Quarterly Journal of Economics, 84(3), 488–500.
Alchian, A.A. (1950), ‘Uncertainty, evolution and economic theory’,
Journal of Political Economy, 58, 211–22.
Allen, F., R. Chakrabarti, S. De, J. Qian and M. Qian (2006), ‘Financing
firms in India’, World Bank Policy Research Working Paper 3975.
Almus, M. (2002), ‘What characterizes a fast-growing firm?’, Applied
Economics, 34, 1497–508.
Almus, M. (2004), ‘The shadow of death – an emperical [sic] analysis of the
pre-exit performance of new German firms’, Small Business Economics,
23(3), 189–201.
Almus, M. and E. Nerlinger (2000), ‘Testing “Gibrat’s law” for young
firms – empirical results for West Germany’, Small Business Economics,
15, 1–12.
Alti, A. (2003), ‘How sensitive is investment to cash flow when financing is
frictionless?’, Journal of Finance, 58(2), 707–22.
Amaral, L.A.N., P. Gopikrishnan, V. Plerou and H.E. Stanley (2001), ‘A
model for the growth dynamics of economic organizations’, Physica A,
299, 127–36.
Amaral, L.A.N., S.V. Buldyrev, S. Havlin, M.A. Salinger and H.E. Stanley
(1998), ‘Power law scaling for a system of interacting units with complex
internal structure’, Physical Review Letters, 80(7), 1385–8.
Amaral, L.A.N., S.V. Buldyrev, S. Havlin, M.A. Salinger, H.E. Stanley
and M.H.R. Stanley (1997), ‘Scaling behavior in economics: the problem
of quantifying company growth’, Physica A, 244, 1–24.
161
Amihud, Y. and B. Lev (1981), ‘Risk reduction as a managerial motive for

conglomerate mergers’, Bell Journal of Economics, 12, 605–17.
Amirkhalkhali, S. and A. Mukhopadhyay (1993), ‘The influence of size
and R&D on the growth of firms in the US’, Eastern Economic Journal,
19(2), 223–33.
Andriani, P. and B. McKelvey (2007), ‘Beyond Gaussian averages: redi-
recting international business and management research toward extreme
events and power laws’, Journal of International Business Studies, 38,
1212–30.
Angelini, P. and A. Generale (2008), ‘On the evolution of firm size distribu-
tions’, American Economic Review, 98(1), 426–38.
Ansoff, I. (1987), Corporate Strategy, Harmondsworth: Penguin Books.
Aoki, M. (1990), ‘Toward an economic model of the Japanese firm’,
Journal of Economic Literature, 28(1), 1–27.
Arabsheibani, G., D. de Meza, J. Maloney and B. Pearson (2000), ‘And a
vision appeared unto them of a great profit: evidence of self-deception
among the self-employed’, Economic Letters, 67, 35–41.
Ashton, T.S. (1926), ‘The growth of textile businesses in the Oldham dis-
trict, 1884–1924’, Journal of the Royal Statistical Society, 89(3), 567–83.
Audretsch, D.B. (2007), The Entrepreneurial Society, Oxford, UK: Oxford
University Press.
Audretsch, D.B. and J. Elston (2002), ‘Does firm size matter? Evidence
on the impact of liquidity constraints on firm investment behavior in
Germany’, International Journal of Industrial Organization, 20, 1–17.
Audretsch, D.B. (1995), ‘Innovation, growth and survival’, International
Journal of Industrial Organization, 13, 441–57.
Audretsch, D.B. and T. Mahmood (1994), ‘Firm selection and indus-
try evolution: the post-entry performance of new firms’, Journal of
Evolutionary Economics, 4, 243–60.
Audretsch, D.B., Y.M. Prince and A.R. Thurik (1999), ‘Do small firms
compete with large firms?’, Atlantic Economic Journal, 27(2), 201–09.
Audretsch, D.B., L. Klomp, E. Santarelli and A.R. Thurik (2004), ‘Gibrat’s
law: are the services different?’, Review of Industrial Organization, 24,
301–24.
Autio, E., H. Sapienza and J. Almeida (2000), ‘Effects of age at entry,
knowledge intensity, and imitability on international growth’, Academy
of Management Journal, 43(5), 909–24.
Aw, B., X. Chen and M. Roberts (2001), ‘Firm-level evidence on produc-
tivity differentials and turnover in Taiwanese manufacturing’, Journal of
Development Economics, 66(1), 51–86.
Axtell, R.L. (2001), ‘Zipf distribution of US firm sizes’, Science, 293,
1818–20.
Bibliography 163
Baden-Fuller, C. (1989), ‘Exit from declining industries and the case of

steel castings’, Economic Journal, 99(398), 946–61.
Baily, M., E.J. Bartelsman and J.C. Haltiwanger (1996), ‘Downsizing and
productivity growth: myth or reality?’, Small Business Economics, 8(4),
259–78.
Baily, M., C. Hulten and D. Campbell (1992), ‘Productivity dynamics
in manufacturing plants’, Brookings Papers on Economic Activity:
Microeconomics, 2, 187–249.
Baily, M.N. and D. Farrell (2006), ‘Breaking down barriers to growth’,
Finance & Development, 43(1), 1–9.
Barnett, S.A. and P. Sakellaris (1998), ‘Non-linear response of firm invest-
ment to q: testing a model of convex and non-convex adjustment costs’,
Journal of Monetary Economics, 42, 261–88.
Barney, J.B. (1986), ‘Strategic factor markets: expectations, luck and busi-
ness strategy’, Management Science, 32(10), 1231–41.
Barron, D.N., E. West and M.T. Hannan (1994), ‘A time to grow and a
time to die: growth and mortality of credit unions in New York, 1914–
1990’, American Journal of Sociology, 100(2), 381–421.
Bartelsman, E., S. Scarpetta and F. Schivardi (2005), ‘Comparative analy-
sis from demographics and survival: evidence from micro-level sources
in OECD countries’, Industrial and Corporate Change, 14(3), 365–91.
Bartelsman, E.J. and M. Doms (2000), ‘Understanding productivity:
lessons from longitudinal microdata’, Journal of Economic Literature,
38(3), 569–94.
Batsch, L. (2003), ‘Le recentrage: une revue des approches financières’,
Finance Control Strategie, 6(2), 43–65.
Baumol, W., P. Heim, B. Malkiel and R. Quandt (1970), ‘Earnings reten-
tion, new capital and the growth of the firm’, Review of Economics and
Statistics, 52(4), 345–55.
Baumol, W.J. (1959), Business Behavior, Value and Growth, New York:
Macmillan.
Becchetti, L. and G. Trovato (2002), ‘The determinants of growth for small
and medium sized firms. The role of the availability of external finance’,
Small Business Economics, 19(4), 291–306.
Beck, T., A. Demirguc-Kunt and R. Levine (2005a), ‘SMEs, growth, and
poverty: cross-country evidence’, Journal of Economic Growth, 10(3),
199–229.
Beck, T., A. Demirguc-Kunt and V. Maksimovic (2005b), ‘Financial and
legal constraints to growth: does firm size matter?’, Journal of Finance,
60(1), 137–77.
Bellone, F., P. Musso, L. Nesta and M. Quere (2008), ‘Market selection
along the firm life cycle’, Industrial and Corporate Change, 17(4), 753–77.
Berger, P. and E. Ofek (1995), ‘Diversification’s effect on firm value’,

Journal of Financial Economics, 37(1), 39–65.
Berger, P.G. and E. Ofek (1999), ‘Causes and effects of corporate refocus-
ing programs’, Review of Financial Studies, 12(2), 311–45.
Bhagat, S.A., A. Shleifer and R.W. Vishny (1990), ‘Hostile takeovers in
the 1980’s: the return to corporate specialization’, Brookings Papers on
Economic Activity, special issue, pp. 1–84.
Bigsten, A. and M. Gebreeyesus (2007), ‘The small, the young, and the
productive: determinants of manufacturing firm growth in Ethiopia’,
Economic Development and Cultural Change, 55(4), 813–40.
Birch, D. (1987), Job Creation in America: How our Smallest Companies
put the Most People to Work, New York: Free Press.
Birley, S. and P. Westhead (1994), ‘A taxonomy of business start-up
reasons and their impact on firm growth and size’, Journal of Business
Venturing, 9(1), 7–31.
Bloom, N. and J. Van Reenen (2002), ‘Patents, real options and firm
performance’, Economic Journal, 112, C97–C116.
Blundell, R., S. Bond, M. Devereux and F. Schiantarelli (1992), ‘Investment
and Tobin’s q’, Journal of Econometrics, 51, 233–57.
Boeri, T. and U. Cramer (1992), ‘Employment growth, incumbents and
entrants: evidence from Germany’, International Journal of Industrial
Organization, 10, 545–65.
Bond, S. and C. Meghir (1994), ‘Dynamic investment models and the firm’s
financial policy’, Review of Economic Studies, 61, 197–222.
Bond, S., D. Harhoff and J. Van Reenen (2003a), ‘Investment, R&D and
financial constraints in Britain and Germany’, Discussion Paper 0595,
Centre for Economic Performance.
Bond, S., J. Elston, J. Mairesse and B. Mulkay (2003b), ‘Financial
factors and the investment in Belgium, France, Germany and the
United Kingdom: a comparison using company panel data’, Review of
Economics and Statistics, 85, 153–65.
Bond, S., A. Klemm, R. Newton-Smith, M. Syed and G. Vlieghe (2004),
‘The roles of expected profitability, Tobin’s q and cash flow in economet-
ric models of company investment’, Working Paper No. 04/12, Institute
for Fiscal Studies.
Boone, J., J.C. van Ours and H. van der Weil (2007), ‘How (not) to measure
competition’, CEPR Discussion Paper 6275.
Bottazzi, G. and A. Secchi (2003a), ‘Common properties and sectoral
specificities in the dynamics of US manufacturing companies’, Review
of Industrial Organization, 23, 217–32.
Bottazzi, G. and A. Secchi (2003b), ‘A stochastic model of firm growth’,
Physica A, 324, 213–19.
Bibliography 165
Bottazzi, G. and A. Secchi (2003c), ‘Why are distributions of firm growth

rates tent-shaped?’, Economics Letters, 80(3), 415–20.
Bottazzi, G. and A. Secchi (2005), ‘Growth and diversification pat-
terns of the worldwide pharmaceutical industry’, Review of Industrial
Organization, 26, 195–216.
Bottazzi, G. and A. Secchi (2006a), ‘Explaining the distribution of firms
growth rates’, Rand Journal of Economics, 37, 234–63.
Bottazzi, G. and A. Secchi (2006b), ‘Gibrat’s law and diversification’,
Industrial and Corporate Change, 37(2), 235–56.
Bottazzi, G., E. Cefis and G. Dosi (2002), ‘Corporate growth and indus-
trial structure: some evidence from the Italian manufacturing industry’,
Industrial and Corporate Change, 11, 705–23.
Bottazzi, G., A. Secchi and F. Tamagni (2008b), ‘Productivity, profitabil-
ity and financial performance’, Industrial and Corporate Change, 17(4),
711–51.
Bottazzi, G., E. Cefis, G. Dosi and A. Secchi (2007), ‘Invariances and
diversities in the patterns of industrial evolution: some evidence from
Italian manufacturing industries’, Small Business Economics, 29(1),
137–59.
Bottazzi, G., A. Coad, N. Jacoby and A. Secchi (2008a), ‘Corporate
growth and industrial dynamics: evidence from French manufacturing’,
Applied Economics, forthcoming.
Bottazzi, G., G. Dosi, M. Lippi, F. Pammolli and M. Riccaboni (2001),
‘Innovation and corporate growth in the evolution of the drug industry’,
International Journal of Industrial Organization, 19, 1161–87.
Bougheas, S., H. Gorg and E. Strobl (2003), ‘Is R&D financially con-
strained? Theory and evidence from Irish manufacturing’, Review of
Industrial Organization, 22, 159–74.
Brock, W.A. (1999), ‘Scaling in economics: a reader’s guide’, Industrial and
Corporate Change, 8(3), 409–46.
Bronars, S. and D. Deere (1993), ‘Union organizing activity, firm growth,
and the business cycle’, American Economic Review, 83(1), 203–20.
Brouwer, E., A. Kleinknecht and J.O.N. Reijnen (1993), ‘Employment
growth and innovation at the firm level: an empirical study’, Journal of
Evolutionary Economics, 3, 153–59.
Brown, C. and J. Medoff (1989), ‘The employer size–wage effect’, Journal
of Political Economy, 97(5), 1027–59.
Brown, C. and J. Medoff (2003), ‘Firm age and wages’, Journal of Labor
Economics, 21(3), 677–97.
Brown, J.R., S. Fazzari and B. Petersen (2007), ‘Financing innovation and
growth: cash flow, external equity and the 1990s R&D boom’, SSRN
Working Paper, available at SSRN: http://ssrn.com/abstract51000382.
Brynjolfsson, E. and L.M. Hitt (2000), ‘Beyond computation: information

technology, organizational transformation and business performance’,
Journal of Economic Perspectives, 14, 23–48.
Buldyrev, S., J. Growiec, F. Pammolli, M. Riccaboni and H. Stanley
(2007), ‘The growth of business firms: facts and theory’, Journal of the
European Economic Association, 5(2–3), 574–84.
Cabral, L.M.B. and J. Mata (2003), ‘On the evolution of the firm size distri-
bution: facts and theory’, American Economic Review, 93(4), 1075–90.
Calvo, J.L. (2006), ‘Testing Gibrat’s law for small, young and innovating
firms’, Small Business Economics, 26, 117–23.
Camerer, C.F. and D. Lovallo (1999), ‘Overconfidence and excess entry: an
experimental approach’, American Economic Review, 89(1), 306–18.
Campa, J. and S. Kedia (2002), ‘Explaining the diversification discount’,
Journal of Finance, 57(4), 1731–62.
Carden, S.D. (2005), ‘What global executives think about growth and risk’,
McKinsey Quarterly, no. 2, 16–25.
Carpenter, R. and A. Guariglia (2007), ‘Investment behavior, observ-
able expectations, and internal funds: a comment on Cummins et al.
(American Economic Review, 2006)’, Economics Bulletin, 5(12), 1–12.
Carpenter, R. and B. Petersen (2002), ‘Is the growth of small firms con-
strained by internal finance?’, Review of Economics and Statistics, 84,
298–309.
Catley, S. and R. Hamilton (1998), ‘Small business development and
gender of owner’, Journal of Management Development, 17(1), 75–82.
Caves, R.E. (1998), ‘Industrial organization and new findings on the
turnover and mobility of firms’, Journal of Economic Literature, 36(4),
1947–82.
Caves, R.E. and M.E. Porter (1977), ‘From entry barriers to mobility bar-
riers: conjectural decisions and the contrived deterrence to new competi-
tion’, Quarterly Journal of Economics, 91(2), 241–62.
Cefis, E. and L. Orsenigo (2001), ‘The persistence of innovative activities:
a cross-countries and cross-sectors comparative analysis’, Research
Policy, 30, 1139–58.
Cefis, E., M. Ciccarelli and L. Orsenigo (2007), ‘Testing Gibrat’s legacy: a
Bayesian approach to study the growth of firms’, Structural Change and
Economic Dynamics, 18(3), 348–69.
Chandler, A. (1992), ‘Organizational capabilities and the economic history of
the industrial enterprise’, Journal of Economic Perspectives, 6(3), 79–100.
Chapin, F. (1957), ‘The optimum size of institutions: a theory of the large
group’, American Journal of Sociology, 62(5), 449–60.
Chesher, A. (1979), ‘Testing the law of proportionate effect’, Journal of
Industrial Economics, 27(4), 403–11.
Bibliography 167
Chiaromonte, F. and G. Dosi (1993), ‘Heterogeneity, competition, and

macroeconomic dynamics’, Structural Change and Economic Dynamics,
4(1), 39–63.
Chirinko, R.S. (1993), ‘Business fixed investment spending: modeling strat-
egies, empirical results, and policy implications’, Journal of Economic
Literature, 31(4), 1875–911.
Churchill, N. (1996), ‘The six key phases of company growth’, in S. Birley
and D.F. Muzyka (eds), Mastering Enterprise, London, UK: Financial
Times, Pitman, pp. 213–19.
Churchill, N. and V. Lewis (1983), ‘The five stages of small business
growth’, Harvard Business Review, 61(3), 30–50.
Cleary, S. (1999), ‘The relationship between firm investment and financial
status’, Journal of Finance, 54(2), 673–92.
Cleary, S., P. Povel and M. Raith (2007), ‘The u-shaped investment curve:
theory and evidence’, Journal of Financial and Quantitative Analysis, 42,
1–40.
Clementi, G. and H. Hopenhayn (2006), ‘A theory of financing constraints
and firm dynamics’, Quarterly Journal of Economics, 121(1), 229–65.
Coad, A. (2006), ‘A closer look at serial growth rate correlation’, LEM
Papers Series 2006/29, Laboratory of Economics and Management
(LEM), Sant’Anna School of Advanced Studies, Pisa, Italy.
Coad, A. (2007a), ‘A closer look at serial growth rate correlation’, Review
of Industrial Organization, 31(1), 69–82.
Coad, A. (2007b), ‘Exploring the “mechanics” of firm growth: evidence
from a short-panel VAR’, Documents de travail du Centre d’Economie
de la Sorbonne r07037, Université Panthéon-Sorbonne (Paris 1), Centre
d’Economie de la Sorbonne.
Coad, A. (2007c), ‘Neoclassical vs evolutionary theories of financial con-
straints: critique and prospectus’, Cahiers du Centre d’Economie de la
Sorbonne No. 2007-08, Paris, France.
Coad, A. (2007d), ‘Testing the principle of “growth of the fitter”: the
relationship between profits and firm growth’, Structural Change and
Economic Dynamics, 18(3), 370–86.
Coad, A. (2008a), ‘Distance to frontier and appropriate business strategy’,
LEM Papers Series 2008/13, Laboratory of Economics and Management
(LEM), Sant’Anna School of Advanced Studies, Pisa, Italy.
Coad, A. (2008b), ‘Firm growth and scaling of growth rate variance in
multiplant firms’, Economics Bulletin, 12(9), 1–15.
Coad, A. (2008c), ‘Firms as bundles of discrete resources towards an expla-
nation of the exponential distribution of firm growth rates’, Documents
de travail du Centre d’Economie de la Sorbonne, Université Panthéon-
Sorbonne (Paris 1), Centre d’Economie de la Sorbonne.
Coad, A. and T. Broekel (2007), ‘Firm growth and productivity growth:

evidence from a panel VAR’, Jena Economic Research Papers 2007-103,
Friedrich-Schiller-University Jena.
Coad, A. and R. Rao (2006), ‘Innovation and market value: a quantile
regression analysis’, Economics Bulletin, 15(13), 1–10.
Coad, A. and R. Rao (2007), ‘The employment effects of innovations in
high-tech industries’, Papers on Economics and Evolution 2007-05, Max
Planck Institute of Economics, Evolutionary Economics Group.
Coad, A. and R. Rao (2008), ‘Innovation and firm growth in high-tech
sectors: a quantile regression approach’, Research Policy, 37(4), 633–48.
Coad, A. and R. Rao (2009), ‘Firm growth and R&D expenditure’,
Economics of Innovation and New Technology, forthcoming.
Coad, A. and J.P. Tamvada (2008), ‘The growth and decline of small firms
in developing countries’, Papers on Economics and Evolution 2008-08,
Max Planck Institute of Economics, Evolutionary Economics Group.
Coad, A., R. Rao and F. Tamagni (2008), ‘Growth processes of Italian
manufacturing firms’, Jena Economic Research Papers 2008-039,
Friedrich-Schiller-University Jena, Germany.
Coase, R. (1937), ‘The nature of the firm’, Economica, 4(16), 386–405.
Cohen, W.M., R.R. Nelson and J.P. Walsh (2000), ‘Protecting their intel-
lectual assets: appropriability conditions and why US manufacturing
firms patent (or not)’, NBER Working Paper 7552.
Colombo, M. and L. Grilli (2005), ‘Founders’ human capital and the
growth of new technology-based firms: a competence-based view’,
Research Policy, 34(6), 795–816.
Cooley, T. and V. Quadrini (2001), ‘Financial markets and firm dynamics’,
American Economic Review, 91(5), 1286–310.
Cooper, R., J. Haltiwanger and L. Power (1999), ‘Machine replacement
and the business cycle: lumps and bumps’, American Economic Review,
89(4), 921–46.
Cordes, C., P. Richerson, R. McElreath and P. Strimling (2008), ‘A naturalis-
tic approach to the theory of the firm: the role of cooperation and cultural
evolution’, Journal of Economic Behavior and Organization, 68, 125–39.
Corsino, M. (2008), ‘Product innovation and corporate growth: the case
of integrated circuits’, LEM Working Paper, 2008-02, Laboratory of
Economics and Management (LEM), Sant’Anna School of Advanced
Studies, Pisa, Italy.
Cressy, R. (1996), ‘Are business startups debt-rationed?’, Economic Journal,
106(438), 1253–70.
Cromie, S. (1987), ‘Similarities and differences between women and
men who choose business proprietorship’, International Small Business
Journal, 5(3), 43–60.
Bibliography 169
Cummins, J., K. Hassett and S. Oliner (2006), ‘Investment behav-

ior, observable expectations, and internal funds’, American Economic
Review, 96(3), 796–810.
Das, S. (1995), ‘Size, age and firm growth in an infant industry: the com-
puter hardware industry in India’, International Journal of Industrial
Organization, 13(1), 111–26.
Davidsson, P. and F. Delmar (2006), ‘High-growth firms and their contri-
bution to employment: the case of Sweden 1987–96’, in P. Davidsson,
F. Delmar and J. Wiklund (eds), Entrepreneurship and the Growth of the
Firm, Cheltenham, UK and Northampton, MA, USA: Edward Elgar.
Davidsson, P., P. Steffens and J. Fitzsimmons (2008), ‘Growing profitable
or growing from profits: putting the horse in front of the cart?’, Journal
of Business Venturing, forthcoming.
Davis, S., J. Haltiwanger, R. Jarmin and J. Miranda (2006), ‘Volatility and
dispersion in business growth rates: publicly traded versus privately held
firms’, NBER Working Paper 12354.
Davis, S.J., J.C. Haltiwanger and S. Schuh (1996), Job Creation and
Destruction, Cambridge and London: MIT Press.
Dawid, H. (2006), ‘Agent-based models of innovation and technological
change’, in K. Judd and L. Tesfatsion (eds), Handbook of Computational
Economics, Volume 2: Agent-based Computational Economics, North
Holland: Amsterdam, Chapter 25.
De Fabritiis, G., F. Pammolli and M. Riccaboni (2003), ‘On size and
growth of business firms’, Physica A, 324(1–2), 38–44.
de Jong, J. and O. Marsili (2006), ‘The fruit flies of innovations: a tax-
onomy of innovative small firms’, Research Policy, 35(2), 213–29.
de Meza, D. (2002), ‘Overlending?’, Economic Journal, 112, F17–F31.
de Meza, D. and D. Webb (1987), ‘Too much investment: a problem of asym-
metric information’, Quarterly Journal of Economics, 102(2), 281–92.
de Meza, D. and D. Webb (1999), ‘Wealth, enterprise and credit policy’,
Economic Journal, 109(455), 153–63.
de Wit, G. (2005), ‘Firm size distributions: an overview of steady-state dis-
tributions resulting from firm dynamics models’, International Journal
of Industrial Organization, 23(5–6), 423–50.
Degryse, H. and A. de Jong (2006), ‘Investment and internal finance: asym-
metric information or managerial discretion?’, International Journal of
Industrial Organization, 24(1), 125–47.
Del Monte, A. and E. Papagni (2003), ‘R&D and the growth of firms:
empirical analysis of a panel of Italian firms’, Research Policy, 32(6),
1003–14.
Delmar, F. (1997), ‘Measuring growth: methodological considerations and
empirical results’, in R. Donckels and A. Miettinen (eds), Entrepreneurship
and SME Research: On its Way to the Next Millennium, Aldershot, VA:
Avebury, pp. 190–216.
Delmar, F. and J. Wiklund (2008), ‘The effect of small business man-
agers’ growth motivation on firm growth: a longitudinal study’,
Entrepreneurship Theory and Practice, 32(3), 437–57.
Delmar, F., P. Davidsson and W. Gartner (2003), ‘Arriving at the high-
growth firm’, Journal of Business Venturing, 18(2), 189–216.
Dickerson, A.P., H.D. Gibson and E. Tsakalotos (2000), ‘Internal vs exter-
nal financing of acquisitions: do managers squander retained profits?’,
Oxford Bulletin of Economics and Statistics, 62(3), 417–31.
Dierickx, I. and K. Cool (1989), ‘Asset stock accumulation and sustainabil-
ity of competitive advantage’, Management Science, 35(12), 1504–11.
Disney, R., J. Haskel and Y. Heden (2003), ‘Restructuring and productiv-
ity growth in UK manufacturing’, Economic Journal, 113, 666–94.
Dixit, A. (1980), ‘The role of investment in entry-deterrence’, Economic
Journal, 90(357), 95–106.
Dixon, R.L. (1953), ‘Creep’, Journal of Accountancy, July, 48–55.
Dobson, S. and B. Gerrard (1989), ‘Growth and profitability in the
Leeds engineering sector’, Scottish Journal of Political Economy, 36(4),
334–52.
Doms, M. and T. Dunne (1998), ‘Capital adjustment patterns in manufac-
turing plants’, Review of Economic Dynamics, 1, 409–29.
Doms, M., T. Dunne and M.J. Roberts (1995), ‘The role of technology
use in the survival and growth of manufacturing plants’, International
Dosi, G. (2004), ‘A very reasonable objective still beyond our reach: eco-
nomics as an empirically disciplined social science’, in M. Augier and J.
March (eds), Models of a Man: Essays in Memory of Herbert A. Simon,
Dosi, G. (2007), ‘Statistical regularities in the evolution of industries:
a guide through some evidence and challenges for the theory’, in F.
Malerba and S. Brusoni (eds), Perspectives on Innovation, Cambridge,
UK: Cambridge University Press.
Dosi, G. and M. Grazzi (2006), ‘Technologies as problem-solving pro-
cedures and technologies as input-output relations: some perspectives
on the theory of production’, Industrial and Corporate Change, 15(1),
173–202.
Dosi, G. and D. Lovallo (1998), ‘Rational entrepreneurs or optimistic
martyrs? Some considerations on technological regimes, corporate
entries, and the evolutionary role of decision biases’, in R. Garud, P.
Nayyar and Z. Shapiro (eds), Foresights and Oversights in Technological
Change, Cambridge, UK: Cambridge University Press, pp. 41–68.
Bibliography 171
Dosi, G., G. Fagiolo and A. Roventini (2006), ‘An evolutionary model of

endogenous business cycles’, Computational Economics, 27(1), 3–34.
Dosi, G., R. Nelson and S. Winter (2000), The Nature and Dynamics of
Organizational Capabilities, Oxford: Oxford University Press.
Dosi, G., O. Marsili, L. Orsenigo and R. Salvatore (1995), ‘Learning,
market selection and the evolution of industrial structures’, Small
Business Economics, 7(6), 411–36.
Doukas, J. and O. Kan (2004), ‘Excess cash flows and diversification dis-
count’, Financial Management, 33(2), 71–88.
Downie, J. (1958), The Competitive Process, London: Duckworth.
Droucopoulos, V. (1983), ‘International big business revisited: on the
size and growth of the world’s largest firms’, Managerial and Decision
Economics, 4(4), 244–52.
Druilhe, C. and E. Garnsey (2006), ‘University spin-out firms: patterns of
development based on expertise’, in M. Bernasconi, S. Harris and M.
Moensted (eds), High-tech Entrepreneurship, London: Routledge, pp.
158–173.
Dunne, P. and A. Hughes (1994), ‘Age, size, growth and survival: UK com-
panies in the 1980s’, Journal of Industrial Economics, 42(2), 115–40.
Dunne, T., M. Roberts and L. Samuelson (1988), ‘Patterns of entry and
exit in US manufacturing industries’, Rand Journal of Economics, 19(4),
495–515.
Dunne, T., M. Roberts and L. Samuelson (1989), ‘The growth and failure
of US manufacturing plants’, Quarterly Journal of Economics, 104(4),
671–98.
Eisenhardt, K. and J. Martin (2000), ‘Dynamic capabilities: what are
they?’, Strategic Management Journal, 21(10–11), 1105–21.
Eisenhardt, K. and C. Schoonhoven (1990), ‘Organizational growth:
linking founding team, strategy, environment, and growth among US
semiconductor ventures, 1978–1988’, Administrative Science Quarterly,
35(3), 504–29.
Erickson, T. and T. Whited (2000), ‘Measurement error and the relation-
ship between investment and q’, Journal of Political Economy, 108(5),
1027–57.
Ericson, R. and A. Pakes (1995), ‘Markov-perfect industry dynamics: a frame-
work for empirical work’, Review of Economic Studies, 62(1), 53–82.
Evangelista, R. and M. Savona (2002), ‘The impact of innovation on
employment in services: evidence from Italy’, International Review of
Applied Economics, 16(3), 309–18.
Evangelista, R. and M. Savona (2003), ‘Innovation, employment and skills
in services. Firm and sectoral evidence’, Structural Change and Economic
Dynamics, 14, 449–74.
Evans, D.S. (1987a), ‘The relationship between firm growth, size and
age: estimates for 100 manufacturing industries’, Journal of Industrial
Economics, 35, 567–81.
Evans, D.S. (1987b), ‘Tests of alternative theories of firm growth’, Journal
Fagiolo, G. and A. Luzzi (2006), ‘Do liquidity constraints matter in
explaining firm size and growth? Some evidence from the Italian manu-
facturing industry’, Industrial and Corporate Change, 15(1), 1–39.
Fazzari, S., R. Hubbard and B. Petersen (1988a), ‘Financing constraints
and corporate investment’, Brookings Papers on Economic Activity,
1(1988), 141–95.
Fazzari, S., R. Hubbard and B. Petersen (1988b), ‘Investment, financ-
ing decisions, and tax policy’, American Economic Review Papers and
Proceedings, 78(2), 200–205.
Fazzari, S., R. Hubbard and B. Petersen (2000), ‘Investment-cash flow
sensitivities are useful: a comment on Kaplan and Zingales’, Quarterly
Journal of Economics, 115(2), 695–705.
Feldman, M. and B. Pentland (2003), ‘Reconceptualizing organizational
routines as a source of flexibility and change’, Administrative Science
Quarterly, 48(1), 94–121.
Feldman, M.S. (2004), ‘Resources in emerging structures and processes of
change’, Organization Science, 15(3), 295–309.
Fishman, A. and R. Rob (2003), ‘Consumer inertia, firm growth and indus-
try dynamics’, Journal of Economic Theory, 109(1), 24–38.
Fizaine, F. (1968), ‘Analyse statistique de la croissance des entreprises
selon l’âge et la taille’, Revue d’économie Politique, 78, 606–20.
Fleck, J. (1984), ‘The adoption of robots in industry’, Physics in Technology,
15, 4–11.
Fluck, Z. and A. Lynch (1999), ‘Why do firms merge and then divest? A
theory of financial synergy’, Journal of Business, 72(3), 319–46.
Foster, L., J. Haltiwanger and C. Krizan (2006), ‘Market selection, real-
location, and re-structuring in the US retail trade sector in the 1990s’,
Review of Economics and Statistics, 88(4), 748–58.
Foster, L., J. Haltiwanger and C.J. Krizan (1998), ‘Aggregate productivity
growth: lessons from microeconomic evidence’, NBER Working Paper
6803.
Foster, L., J. Haltiwanger and C. Syverson (2008), ‘Reallocation, firm
turnover, and efficiency: selection on productivity or profitability?’,
Freel, M. and P. Robson (2004), ‘Small firm innovation, growth and per-
formance: evidence from Scotland and Northern England’, International
Small Business Journal, 22(6), 561–75.
Bibliography 173
Freel, M.S. (2000), ‘Do small innovating firms outperform non-innova-

tors?’, Small Business Economics, 14, 195–210.
Freeland, R. (2001), The Struggle for Control of the Modern Corporation:
Organizational Change at General Motors, 1924–1970, Cambridge, UK:
Cambridge University Press.
Friedman, M. (1953), ‘The methodology of positive economics’, in M.
Friedman (ed.), Essays in Positive Economics, Chicago, IL: University
of Chicago Press, pp. 3–43.
Friedman, M. (1992), ‘Do old fallacies ever die?’, Journal of Economic
Literature, 30(4), 2129–32.
Fu, D., F. Pammolli, S. Buldyrev, M. Riccaboni, K. Matia, K. Yamasaki
and H. Stanley (2005), ‘The growth of business firms: theoretical frame-
work and empirical evidence’, Proceedings of the National Academy of
Sciences, 102(52), 18801–06.
Gabe, T. and D. Kraybill (2002), ‘The effect of state economic develop-
ment incentives on employment growth of establishments’, Journal of
Regional Science, 42(4), 703–30.
Galeotti, M., F. Schiantarelli and F. Jaramillo (1994), ‘Investment deci-
sions and the role of debt, liquid assets and cash flow: evidence from
Italian panel data’, Applied Financial Economics, 4(2), 121–32.
Garnsey, E. (1998), ‘A theory of the early growth of the firm’, Industrial
and Corporate Change, 7(3), 523–56.
Garnsey, E. and P. Heffernan (2005), ‘Growth setbacks in new firms’,
Futures, 37(7), 675–97.
Garnsey, E., E. Stam and P. Heffernan (2006), ‘New firm growth: exploring
processes and paths’, Industry and Innovation, 13(1), 1–20.
Gartner, W.B. (1997), ‘When growth is the problem, not the solution: an
interview with Earl Eisenberg’, Journal of Management Inquiry, 6(1),
62–8.
Geroski, P.A. (2001), ‘Exploring the niche overlaps between organiza-
tional ecology and industrial economics’, Industrial and Corporate
Change, 10(2), 507–48.
Geroski, P.A. and M. Mazzucato (2002), ‘Learning and the sources of
corporate growth’, Industrial and Corporate Change, 11(4), 623–44.
Geroski, P.A. and S. Toker (1996), ‘The turnover of market leaders in UK
manufacturing industry, 1979–86’, International Journal of Industrial
Geroski, P.A., S. Machin and C. Walters (1997), ‘Corporate growth and
profitability’, Journal of Industrial Economics, 45(2), 171–89.
Geroski, P.A., S. Lazarova, G. Urga and C. Walters (2003), ‘Are dif-
ferences in firm size transitory or permanent?’, Journal of Applied
Econometrics, 18(1), 47–59.
Geroski, P.A. (1995), ‘What do we know about entry?’, International

Geroski, P.A. (2000), ‘The growth of firms in theory and practice’, in N. Foss
and V. Mahnke (eds), Competence, Governance, and Entrepreneurship,
Oxford, UK: Oxford University Press.
Geroski, P.A. (2005), ‘Understanding the implications of empirical work on
corporate growth rates’, Managerial and Decision Economics, 26, 129–38.
Geroski, P.A. and K. Gugler (2004), ‘Corporate growth convergence in
Europe’, Oxford Economic Papers, 56, 597–620.
Geroski, P.A. and S. Machin (1992), ‘Do innovating firms outperform
non-innovators?’, Business Strategy Review, 3(2), 79–90.
Gibrat, R. (1931), Les Inégalités Économiques: Applications, aux Inégalités
des Richesses, à la Concentration des Entreprises, aux Populations des
Villes, aux Statistiques des Familles, etc.: d’une Loi Nouvelle: la Loi de
l’effet Proportionnel, Paris: Recueil Sirey.
Gilchrist, S. and C. Himmelberg (1995), ‘Evidence on the role of cash flow
for investment’, Journal of Monetary Economics, 36(3), 541–72.
Goddard, J., J. Wilson and P. Blandon (2002), ‘Panel tests of Gibrat’s
law for Japanese manufacturing’, International Journal of Industrial
Goedhuys, M. and L. Sleuwaegen (2000), ‘Entrepreneurship and growth of
entrepreneurial firms in Côte d’Ivoire’, Journal of Development Studies,
36(3), 123–45.
Goedhuys, M. and L. Sleuwaegen (2008), ‘High-growth entrepreneurial
firms in least developed countries’, paper presented at the UNU-WIDER
project workshop on Entrepreneurship and Economic Development,
21–23 August, Helsinki.
Goergen, M. and L. Renneboog (2001), ‘Investment policy, internal
financing and ownership concentration in the UK’, Journal of Corporate
Finance, 7(3), 257–84.
Gomes, J. (2001), ‘Financing investment’, American Economic Review,
91(5), 1263–85.
Grabowski, H., J. Vernon and J.A. DiMasi (2002), ‘Returns on research and
development for 1990s new drug introductions’, Pharmacoeconomics,
20(3), 11–29.
Graham, J., M. Lemmon and J. Wolf (2002), ‘Does corporate diversifica-
tion destroy value?’, Journal of Finance, 57(2), 695–720.
Greenhalgh, C., M. Longland and D. Bosworth (2001), ‘Technological
activity and employment in a panel of UK firms’, Scottish Journal of
Political Economy, 48(3), 260–82.
Greiner, L. (1972), ‘Evolution and revolution as organizations grow’,
Harvard Business Review, 37–46.
Bibliography 175
Greiner, L. (1998), ‘Evolution and revolution as organizations grow’,

Harvard Business Review, May, 55–67.
Griliches, Z. and J. Mairesse (1995), ‘Production functions: the search for
identification’, NBER Working Paper 5067.
Griliches, Z. and H. Regev (1995), ‘Firm productivity in Israeli industry
1979–1988’, Journal of Econometrics, 65(1), 175–203.
Guariglia, A. (2008), ‘Internal financial constraints, external financial
constraints, and investment choice: evidence from a panel of UK firms’,
Journal of Banking and Finance, 32(9), 1795–1809.
Guiso, L. and G. Parigi (1999), ‘Investment and demand uncertainty’,
Hadlock, C. (1998), ‘Ownership, liquidity, and investment’, Rand Journal
of Economics, 29(3), 487–508.
Haleblian, J. and S. Finkelstein (1999), ‘The influence of organizational
acquisition experience on acquisition performance: a behavioral learn-
ing perspective’, Administrative Science Quarterly, 44(1), 29–56.
Hall, B.H. (1987), ‘The relationship between firm size and firm growth in
the US manufacturing sector’, Journal of Industrial Economics, 35(4),
583–600.
Hall, B.H. (2000), ‘Innovation and market value’, in R. Barrell, G. Mason
and M. O’Mahoney (eds), Productivity, Innovation, and Economic
Performance, Cambridge, UK: Cambridge University Press, pp. 177–98.
Hall, B.H., F. Lotti and J. Mairesse (2008), ‘Employment, innovation and
productivity: evidence from Italian microdata’, Industrial and Corporate
Change, 17(4), 813–39.
Hannan, M. and J. Freeman (1977), ‘The population ecology of organiza-
tions’, American Journal of Sociology, 82(5), 929–64.
Hannan, M. and J. Freeman (1984), ‘Structural inertia and organizational
change’, American Sociological Review, 49(2), 149–64.
Hannan, M.T. (2005), ‘Ecologies of organizations: diversity and identity’,
Journal of Economic Perspectives, 19(1), 51–70.
Hannan, M.T., G.R. Carroll, E. Dundon and J.C. Torres (1995),
‘Organizational evolution in multinational context: entries of automo-
bile manufacturers in Belgium, France, Germany and Italy’, American
Sociological Review, 60, 509–28.
Harada, N. (2007), ‘Which firms exit and why? An analysis of small firm
exits in Japan’, Small Business Economics, 29(4), 401–14.
Hardwick, P. and M. Adams (2002), ‘Firm size and growth in the United
Kingdom life insurance industry’, The Journal of Risk & Insurance,
69(4), 577–93.
Harhoff, D., K. Stahl and M. Woywode (1998), ‘Legal form, growth
and exits of West German firms – empirical results for manufacturing,
construction, trade and service industries’, Journal of Industrial

Economics, 46(4), 453–88.
Harrison, R., J. Jaumandreu, J. Mairesse and B. Peters (2005), ‘Does inno-
vation stimulate employment? A firm-level analysis using comparable
micro data on four European countries’, mimeo, Universidad Carlos
III, Madrid.
Hart, P.E. (1962), ‘The size and growth of firms’, Economica, 29(113), 29–39.
Hart, P.E. and R. Pearce (1986), ‘Growth patterns of the world’s largest
firms, 1962–1982’, Review of World Economics, 122(1), 65–79.
Hart, P.E. and S. Prais (1956), ‘The analysis of business concentration:
a statistical approach’, Journal of the Royal Statistical Society, 119(2),
150–91.
Hart, P.E. and N. Oulton (1996), ‘The growth and size of firms’, Economic
Journal, 106(3), 1242–52.
Haskel, J.E., S.C. Pereira and M.J. Slaughter (2007), ‘Does inward foreign
direct investment boost the productivity of domestic firms?’, Review of
Economics and Statistics, 89(3), 482–96.
Hay, D.A. and D.J. Morris (1979), Industrial Economics: Theory and
Evidence, Oxford: Oxford University Press.
Hay, M. and K. Kamshad (1994), ‘Small firm growth: intentions, imple-
mentation and impediments’, Business Strategy Review, 5(3), 49–68.
Hayashi, F. (1982), ‘Tobin’s marginal q and average q: a neoclassical inter-
pretation’, Econometrica, 50(1), 213–24.
Headd, B. (2003), ‘Redefining business success: distinguishing between
closure and failure’, Small Business Economics, 21(1), 51–61.
Headd, B. and B. Kirchhoff (2007), ‘Small business growth: searching for
stylized facts’, SBA Working Paper.
Higson, C., S. Holly and P. Kattuman (2002), ‘The cross-sectional dynam-
ics of the US business cycle: 1950–1999’, Journal of Economic Dynamics
and Control, 26(9–10), 1539–55.
Higson, C., S. Holly, P. Kattuman and S. Platis (2004), ‘The business cycle,
macroeconomic shocks and the cross-section: the growth of UK quoted
companies’, Economica, 71(282), 299–318.
Himmelberg, C.P. and B.C. Petersen (1994), ‘R&D and internal finance: a
panel study of small firms in high-tech industries’, Review of Economics
and Statistics, 76, 38–51.
Hines, J. and R. Thaler (1995), ‘The flypaper effect’, Journal of Economic
Perspectives, 9(4), 217–26.
Hisrich, R.D. and S.A. Ozturk (1999), ‘Women entrepreneurs in a develop-
ing economy’, Journal of Management Development, 18(2), 114–24.
Holl, P. (1975), ‘Effect of control type on the performance of the firm in the
UK’, Journal of Industrial Economics, 23(4), 257–72.
Bibliography 177
Hopenhayn, H. (1992), ‘Entry, exit, and firm dynamics in long run equilib-
rium’, Econometrica, 60(5), 1127–50.
Horngren, C.T. (1984), Introduction to Financial Accounting (2nd edn),
Englewood Cliffs, NJ: Prentice Hall.
Hoshi, T., A. Kashyap and D. Scharfstein (1991), ‘Corporate structure,
liquidity, and investment: evidence from Japanese industrial groups’,
Hu, X. and F. Schiantarelli (1998), ‘Investment and capital market imper-
fections: a switching regression approach using US firm panel data’,
Review of Economics and Statistics, 80(3), 466–79.
Hubbard, R. (1998), ‘Capital-market imperfections and investment’,
Journal of Economic Literature, 36(1), 193–225.
Huberman, B.A. and L.A. Adamic (1999), ‘Growth dynamics of the world-
wide web’, Nature, 401, 131.
Hughes, A. (1997), ‘Finance for SMEs: a UK perspective’, Small Business
Economics, 9(2), 151–68.
Hugo, O. and E. Garnsey (2005), ‘Problem-solving and competence crea-
tion in the early development of new firms’, Managerial and Decision
Economics, 26(2), 139–48.
Hyland, D.C. and J.D. Diltz (2002), ‘Why firms diversify: an empirical
examination’, Financial Management, 31, 51–81.
Hymer, S. and P. Pashigian (1962), ‘Firm size and rate of growth’, Journal
Idson, T. and W. Oi (1999), ‘Workers are more productive in large firms’,
American Economic Review Papers and Proceedings, 89(2), 104–8.
Ijiri, Y. and H.A. Simon (1964), ‘Business firm growth and size’, American
Economic Review, 54(2), 77–89.
Ijiri, Y. and H.A. Simon (1971), ‘Effects of mergers and acquisitions on busi-
ness firm concentration’, Journal of Political Economy, 79(2), 314–22.
Ijiri, Y. and H.A. Simon (1967), ‘A model of business firm growth’,
Econometrica, 35(2), 348–55.
Ijiri, Y. and H.A. Simon (1974), ‘Interpretations of departures from the
Pareto curve firm-size distributions’, Journal of Political Economy, 82(2),
315–31.
Ijiri, Y. and H.A. Simon (1977), Skew Distributions and the Sizes of
Business Firms, Amsterdam: North Holland.
Jensen, M. and W. Meckling (1976), ‘Theory of the firm: managerial
behavior, agency costs, and capital structure’, Journal of Financial
Economics, 3(4), 305–60.
Jensen, M.C. (1986), ‘Agency costs of free cash flow, corporate finance,
and takeovers’, American Economic Review Papers and Proceedings,
76(2), 323–9.
Johanson, J. and J. Vahlne (1977), ‘The internationalization process of

the firm – a model of knowledge development and increasing foreign
market commitments’, Journal of International Business Studies, 8(1), 23–
32.
Johanson, J. and J. Vahlne (1990), ‘The mechanism of internationalisa-
tion’, International Marketing Review, 7(4), 11–24.
Johnson, P., C. Conway and P. Kattuman (1999), ‘Small business growth
in the short run’, Small Business Economics, 12(2), 103–12.
Jones, M. and N. Coviello (2005), ‘Internationalisation: conceptualising
an entrepreneurial process of behaviour in time’, Journal of International
Business Studies, 36(3), 284–304.
Jovanovic, B. (1982), ‘Selection and the evolution of industry’, Econometrica,
50(3), 649–70.
Kadapakkam, P., P. Kumar and L. Riddick (1998), ‘The impact of cash
flows and firm size on investment: the international evidence’, Journal of
Banking and Finance, 22(3), 293–320.
Kaldor, N. (1934), ‘The equilibrium of the firm’, Economic Journal,
44(173), 60–76.
Kalecki, M. (1945), ‘On the Gibrat distribution’, Econometrica, 13(2),
161–70.
Kaplan, S. and L. Zingales (1997), ‘Do investment-cash flow sensitivities
provide useful measures of financing constraints?’, Quarterly Journal of
Economics, 112, 169–215.
Kaplan, S. and L. Zingales (2000), ‘Investment-cash flow sensitivities
are not valid measures of financing constraints’, Quarterly Journal of
Economics, 115(2), 707–12.
Kay, N. (2000), ‘The growth of firms’, in N. Foss and V. Mahnke
(eds), Competence, Governance, and Entrepreneurship, Oxford: Oxford
University Press, Chapter 9, pp. 187–206.
Kazanjian, R. and R. Drazin (1989), ‘An empirical test of a stage of growth
progression model’, Management Science, 35(12), 1489–503.
Kesten, H. (1973), ‘Random difference equations and renewal theory for
products of random matrices’, Acta Mathematica, 131(1), 207–48.
Klette, T. and Z. Griliches (2000), ‘Empirical patterns of firm growth and
R&D investment: a quality ladder model interpretation’, The Economic
Journal, 110(463), 363–87.
Klette, T. and S. Kortum (2004), ‘Innovating firms and aggregate innova-
tion’, Journal of Political Economy, 112(5), 986–1018.
Kogut, B. and U. Zander (2003), ‘Knowledge of the firm and the evolution-
ary theory of the multinational corporation’, Journal of International
Business Studies, 34(6), 516–29. First published in Journal of International
Business Studies (1993), 24, 625–45.
Bibliography 179
Kuemmerle, W. (2002), ‘Home base and knowledge management in inter-

national ventures’, Journal of Business Venturing, 17(2), 99–122.
Kumar, M. (1985), ‘Growth, acquisition activity and firm size: evidence
from the United Kingdom’, Journal of Industrial Economics, 33(3),
327–38.
Kwasnicki, W. (2003), ‘Schumpeterian modelling’, mimeo, Wroclaw
University, Poland.
Lamont, O. (1997), ‘Cash flow and investment: evidence from internal
capital markets’, Journal of Finance, 52(1), 83–109.
Lang, L. and R. Stulz (1994), ‘Tobin’s q, corporate diversification, and firm
performance’, Journal of Political Economy, 102(6), 1248–80.
Lee, Y., L.A.N. Amaral, D. Canning, M. Meyer and H.E. Stanley (1998),
‘Universal features in the growth dynamics of complex organizations’,
Physical Review Letters, 81(15), 3275–8.
Leiponen, A. and I. Drejer (2007), ‘What exactly are technological
regimes? Intra-industry heterogeneity in the organization of innovation
activities’, Research Policy, 36, 1221–38.
Lensink, R., P. Steen and E. Sterken (2005), ‘Uncertainty and growth of
the firm’, Small Business Economics, 24(4), 381–91.
Lerner, J. (2002), ‘When bureaucrats meet entrepreneurs: the design of
effective public venture capital programmes’, Economic Journal, 112,
F73–F84.
Lesourne, J. (1973), Modèles de Croissance des Entreprises, Paris:
Dunod.
Levenson, A. and K. Willard (2000), ‘Do firms get the financing they want?
Measuring credit rationing experienced by small businesses in the US’,
Small Business Economics, 14(2), 83–94.
Lins, K. and H. Servaes (1999), ‘International evidence on the value of
corporate diversification’, Journal of Finance, 54(6), 2215–39.
Little, I.M.D. (1962), ‘Higgledy piggledy growth’, Bulletin of the Oxford
University Institute of Statistics, 24(4), 387–412.
Little, I.M.D. (1987), ‘Small manufacturing enterprises in developing
countries’, World Bank Economic Review, 1(2), 203–35.
Liu, J., M. Tsou and J. Hammitt (1999), ‘Do small plants grow faster?
Evidence from the Taiwan electronics industry’, Economics Letters,
65(1), 121–9.
Lockett, A., J. Wiklund and P. Davidsson (2007), ‘Re-examining and
extending Penrose’s growth theory: updating Penrose for the 21st
century’, Proceedings of Babson College Entrepreneurship Research
Conference (BCERC), Madrid, Spain.
Lotti, F. and E. Santarelli (2001), ‘Is firm growth proportional? An
appraisal of firm size distribution’, Economics Bulletin, 12(6), 1–7.
Lotti, F., E. Santarelli and M. Vivarelli (2003), ‘Does Gibrat’s law hold
among young, small firms?’, Journal of Evolutionary Economics, 13(3),
213–35.
Lu, J. and P. Beamish (2001), ‘The internationalization and performance
of SMEs’, Strategic Management Journal, 22(6–7), 565–86.
Lucas, R. (1978), ‘On the size distribution of business firms’, Bell Journal
of Economics, 9(2), 508–23.
Luttmer, E. (2007), ‘Selection, growth, and the size distribution of firms’,
MacDonald, C.J. (1998), ‘The evolution of Intel’s copy EXACTLY!
Technology transfer method’, Intel Technology Journal, quarter 4, 1–6.
Mairesse, J. and P. Mohnen (2002), ‘Accounting for innovation and meas-
uring innovativeness: an illustrative framework and an application’,
American Economic Review Papers and Proceedings, 92(2), 226–30.
Maksimovic, V. and G. Phillips (2002), ‘Do conglomerate firms allocate
resources inefficiently across industries? Theory and evidence’, Journal
of Finance, 57(2), 721–67.
Maksimovic, V. and G. Phillips (2008), ‘The industry life cycle, acqui-
sitions and investment: does firm organization matter?’, Journal of
Finance, 63(2), 673–708.
Malerba, F. (2002), ‘Sectoral systems of innovation and production’,
Research Policy, 31(2), 247–64.
Malerba, F. and L. Orsenigo (1997), ‘Technological regimes and sectoral
patterns of innovative activities’, Industrial and Corporate Change, 6(1),
83–118.
Mansfield, E. (1962), ‘Entry, Gibrat’s law, innovation, and the growth of
firms’, American Economic Review, 52(5), 1023–51.
Mansfield, E., J. Rapoport, A. Romeo, E. Villani, S. Wagner and F. Husic
(1977), The Production and Application of New Industrial Technology,
New York: Norton.
Markides, C. (1992), ‘Consequences of corporate refocusing: ex ante evi-
dence’, Academy of Management Journal, 35(2), 398–412.
Markides, C. (1995), ‘Diversification, restructuring and economic per-
formance’, Strategic Management Journal, 16(2), 101–18.
Marris, R. (1963), ‘A model of the managerial enterprise’, Quarterly
Marris, R. (1964), The Economic Theory of Managerial Capitalism,
London: Macmillan.
Marris, R. (1999), ‘Edith Penrose and economics’, Contributions to
Political Economy, 18, 47–65.
Marshall, A. (1961), Principles of Economics, 8th edn, London, UK:
Macmillan and Co., First published 1890.
Bibliography 181
Marsili, O. (2001), The Anatomy and Evolution of Industries, Cheltenham,

UK and Northampton, MA, USA: Edward Elgar.
Marsili, O. (2005), ‘Technology and the size distribution of firms: evidence
from Dutch manufacturing’, Review of Industrial Organization, 27(4),
303–28.
Martin, J. and A. Sayrak (2003), ‘Corporate diversification and share-
holder value: a survey of recent literature’, Journal of Corporate Finance,
9(1), 37–57.
Matia, K., D. Fu, S. Buldyrev, F. Pammolli, M. Riccaboni and H. Stanley
(2004), ‘Statistical properties of business firms structure and growth’,
Europhysics Letters, 67(3), 498–503.
Matsusaka, J. (2001), ‘Corporate diversification, value maximization, and
organizational capabilities’, Journal of Business, 74(3), 409–31.
McCombie, J.S.L. (1987), ‘Verdoorn’s law’, in J. Eatwell, M. Milgate
and P. Newman (eds), The New Palgrave: A Dictionary of Economics
(Volume 4), London: Macmillan, pp. 804–06.
McDougall, P., S. Shane and B. Oviatt (1994), ‘Explaining the formation
of international new ventures: the limits of theories from international
business research’, Journal of Business Venturing, 9(6), 469–87.
McKelvey, B. and P. Andriani (2005), ‘Why Gaussian statistics are
mostly wrong for strategic organization’, Strategic Organization, 3(2),
219–28.
McPherson, M.A. (1996), ‘Growth of micro and small enterprises in
Southern Africa’, Journal of Development Economics, 48, 253–77.
Mead, D.C. and C. Liedholm (1998), ‘The dynamics of micro and small
enterprises in developing countries’, World Development, 26, 61–74.
Melitz, M. (2003), ‘The impact of trade on intra-industry reallocations and
aggregate industry productivity’, Econometrica, 71(6), 1695–725.
Metcalfe, J.S. (1993), ‘Some Lamarckian themes in the theory of growth
and economic selection: provisional analysis’, Revue International de
Systemique, 7(5), 487–504.
Metcalfe, J.S. (1994), ‘Competition, Fisher’s principle and increasing
returns in the selection process’, Journal of Evolutionary Economics, 4,
327–46.
Metcalfe, J.S. (1998), Evolutionary Economics and Creative Destruction,
London: Routledge.
Metcalfe, J.S. (2007), ‘Alfred Marshall’s Mecca: reconciling the theories of
value and development’, Economic Record, 83(1), S1–S22.
Miner, J., N. Smith and J. Bracker (1994), ‘Role of entrepreneurial task
motivation in the growth of technologically innovative firms: inter-
pretations from follow-up data’, Journal of Applied Psychology, 79(4),
627–30.
Minkoff, D. (1999), ‘Bending with the wind: strategic change and adapta-
tion by women’s and racial minority organizations’, American Journal
of Sociology, 104(6), 1666–703.
Mitra, R. and V. Pingali (1999), ‘Analysis of growth stages in small firms: a
case study of automobile ancillaries in India’, Journal of Small Business
Management, 37(3), 62–3.
Mitzenmacher, M. (2003), ‘A brief history of generative models for power
law and lognormal distributions’, Internet Mathematics, 1(2), 226–51.
Montgomery, C.A. (1994), ‘Corporate diversification’, Journal of Economic
Perspectives, 8(3), 163–78.
Mowery, D. (1983), ‘Industrial research and firm size, survival, and growth
in American manufacturing, 1921–1946: an assessment’, Journal of
Economic History, 43(4), 953–80.
Mueller, D.C. (1969), ‘A theory of conglomerate mergers’, Quarterly
Myers, S. (1984), ‘The capital structure puzzle’, Journal of Finance, 39(3),
575–92.
Nafziger, E. and D. Terrell (1996), ‘Entrepreneurial human capital and the
long-run survival of firms in India’, World Development, 24(4), 689–96.
Nelson, R.R. (1987), Understanding Technical Change as an Evolutionary
Process, Amsterdam: North Holland.
Nelson, R.R. and S.G. Winter (1982), An Evolutionary Theory of Economic
Change, Cambridge, MA: Belknap Press of Harvard University Press.
Niefert, M. (2005), ‘Patenting behavior and employment growth in
German start-up firms: a panel data analysis’, Discussion Paper 05-03,
ZEW Centre for European Economic Research, Mannheim.
O’Farrell, P.N. and D.M.W.N. Hitchens (1988), ‘Alternative theories of
small-firm growth: a critical review’, Environment and Planning A, 20(3),
1365–83.
Oliner, S. and G. Rudebusch (1992), ‘Sources of the financing hierarchy for
business investment’, Review of Economics and Statistics, 74(2), 643–54.
Olley, G. and A. Pakes (1996), ‘The dynamics of productivity in the tele-
communications equipment industry’, Econometrica, 64(6), 1263–97.
Oviatt, B. and P. McDougall (1994), ‘Toward a theory of international
new ventures’, Journal of International Business Studies, 25(1), 45–64.
Reprinted in Journal of International Business Studies (2005), 36(1),
29–41.
Packard, D. (1995), The HP way: How Bill Hewlett and I built our company,
New York: HarperCollins.
Pakes, A. and R. Ericson (1998), ‘Empirical implications of alternative
models of firm dynamics’, Journal of Economic Theory, 79(1), 1–45.
Parkinson, C.N. (1957), Parkinson’s Law, Boston, MA: Houghton Mifflin.
Bibliography 183
Paulré, B. (2000), ‘Les stratégies de recentrage dans les secteurs de haute

technologie’, Ministère de l’Economie et des Finances.
Pavcnik, N. (2002), ‘Trade liberalization, exit, and productivity improve-
ments: evidence from Chilean plants’, Review of Economic Studies, 69(1),
245–76.
Pavitt, K. (1984), ‘Sectoral patterns of technical change: towards a tax-
onomy and a theory’, Research Policy, 13, 343–75.
Pavitt, K., M. Robson and J. Townsend (1987), ‘The size distribution of
innovating firms in the UK: 1945–1983’, Journal of Industrial Economics,
35(3), 297–316.
Penrose, E.T. (1955), ‘Limits to the growth and size of firms’, American
Economic Review, 45(2), 531–43.
Penrose, E.T. (1959), The Theory of the Growth of the Firm, Oxford, UK:
Basil Blackwell.
Penrose, E.T. (1960), ‘The growth of the firm – a case study: the Hercules
powder company’, Business History Review, 34(1), 1–23.
Phillips, B. and B. Kirchhoff (1989), ‘Formation, growth and survival:
small firm dynamics in the US economy’, Small Business Economics,
1(1), 65–74.
Pitelis, C. (2002), The Growth of the Firm: The Legacy of Edith Penrose,
Oxford, UK: Oxford University Press.
Powell, W., K. Koput and L. Smith-Doerr (1996), ‘Interorganizational
collaboration and the locus of innovation: networks of learning in bio-
technology’, Administrative Science Quarterly, 41(1), 116–45.
Power, L. (1998), ‘The missing link: technology, investment, and produc-
tivity’, Review of Economics and Statistics, 80(2), 300–13.
Prais, S.J. (1974), ‘A new look at the growth of industrial concentration’,
Oxford Economic Papers, 26(2), 273–88.
Quandt, R. (1966), ‘On the size distribution of firms’, American Economic
Review, 56(3), 416–32.
Radice, H.K. (1971), ‘Control type, profitability and growth in large firms:
an empirical study’, Economic Journal, 81, 547–62.
Rajan, R. and J. Wulf (2006), ‘The flattening firm: evidence from panel data
on the changing nature of corporate hierarchies’, Review of Economics
and Statistics, 88(4), 759–73.
Ramsden, J. and G. Kiss-Haypal (2000), ‘Company size distribution in
different countries’, Physica A, 277(1–2), 220–27.
Reed, W. (2001), ‘The Pareto, Zipf and other power laws’, Economics
Letters, 74(1), 15–19.
Reichstein, T. and M. Dahl (2004), ‘Are firm growth rates random?
Analysing patterns and dependencies’, International Review of Applied
Economics, 18(2), 225–46.
Reichstein, T. and M. Jensen (2005), ‘Firm size and firm growth rate
distributions – the case of Denmark’, Industrial and Corporate Change,
14(6), 1145–66.
Rivkin, J. (2001), ‘Reproducing knowledge: replication without imitation
at moderate complexity’, Organization Science, 12(3), 274–93.
Roberts, J. (2004), The Modern Firm: Organizational Design for
Performance and Growth, Oxford, UK: Oxford University Press.
Robson, P. and R. Bennett (2000), ‘SME growth: the relationship with
business advice and external collaboration’, Small Business Economics,
15(3), 193–208.
Robson, P. and B. Obeng (2008), ‘The barriers to growth in Ghana’, Small
Roll, R. (1986), ‘The hubris hypothesis of corporate takeovers’, Journal of
Business, 59(2), 197–216.
Roper, S. (1997), ‘Product innovation and small business growth: a com-
parison of the strategies of German, UK and Irish companies’, Small
Rossi-Hansberg, E. and M. Wright (2007), ‘Establishment size dynam-
ics in the aggregate economy’, American Economic Review, 97(5),
1639–66.
Rumelt, R. (1982), ‘Diversification strategy and profitability’, Strategic
Management Journal, 3(4), 359–69.
Salop, S.C. (1979), ‘Strategic entry deterrence’, American Economic Review
Papers and Proceedings, 69(2), 335–38.
Salter, W.E.G. (1960), Productivity and Technical Change, Cambridge,
UK: Cambridge University Press.
Samuels, J. (1965), ‘Size and the growth of firms’, Review of Economic
Studies, 32(1), 105–12.
Santarelli, E. and M. Vivarelli (2007), ‘Entrepreneurship and the process
of firms’ entry, survival and growth’, Industrial and Corporate Change,
16(3), 455–88.
Sapienza, H., E. Autio, G. George and S. Zahra (2006), ‘A capabilities
perspective on the effects of early internationalization on firm survival
and growth’, Academy of Management Review, 31(4), 914–33.
Sarno, D. (2008), ‘Capital structure and growth of the firms in the back-
ward regions of the [sic] south Italy’, Applied Financial Economics, 18,
821–33.
Sastry, M. (1997), ‘Problems and paradoxes in a model of punctu-
ated organizational change’, Administrative Science Quarterly, 42(2),
237–75.
Schaller, H. (1990), ‘A re-examination of the q theory of investment using
US firm data’, Journal of Applied Econometrics, 5(4), 309–25.
Bibliography 185
Schaller, H. (1993), ‘Asymmetric information, liquidity constraints, and

Canadian investment’, Canadian Journal of Economics, 26(3), 552–74.
Scherer, F.M. (1965), ‘Corporate inventive output, profits, and growth’,
Journal of Political Economy, 73(3), 290–97.
Scherer, F.M. (2001), ‘The link between gross profitability and pharmaceu-
tical R&D spending’, Health Affairs, 20(5), 216–20.
Schiantarelli, F. (1996), ‘Financial constraints and investment: methodo-
logical issues and international evidence’, Oxford Review of Economic
Policy, 12(2), 70–89.
Schivardi, F. and R. Torrini (2008), ‘Identifying the effects of firing
restrictions through size-contingent differences in regulation’, Labour
Economics, 15(3), 538–67.
Schlingemann, F., R. Stulz and R. Walkling (2002), ‘Divestitures and
the liquidity of the market for corporate assets’, Journal of Financial
Economics, 64(1), 117–44.
Schumpeter, J. (1942), Capitalism, Socialism and Democracy, New York:
Harper.
Scott, M. and R. Bruce (1987), ‘Five stages of growth in small business’,
Long Range Planning, 20(3), 45–52.
Segarra, A. and M. Callejon (2002), ‘New firms’ survival and market tur-
bulence: new evidence from Spain’, Review of Industrial Organization,
20(1), 1–14.
Segarra, A., M. Teruel, J. Arauzo and S. Iranzo (2008), ‘Dinámica
empresarial, creación de empleo y productividad en las manufacturas
españolas’, Project for the Spanish government (Ministerio de Industria,
Turismo y Comercio), Madrid.
Shanmugam, K. and S. Bhaduri (2002), ‘Size, age and firm growth in
the Indian manufacturing sector’, Applied Economics Letters, 9(9),
607–13.
Shapiro, C. (1989), ‘The theory of business strategy’, Rand Journal of
Economics, 20(1), 125–37.
Shepherd, D. and J. Wiklund (2009), ‘Are we comparing apples with
apples or apples with oranges? Appropriateness of knowledge accu-
mulation across growth studies’, Entrepreneurship Theory and Practice,
forthcoming.
Shleifer, A. and R. Vishny (1989), ‘Management entrenchment: the case
of manager-specific investments’, Journal of Financial Economics, 25(1),
123–39.
Shleifer, A. and R. Vishny (1990), ‘The takeover wave of the 1980s’,
Science, 249(4970), 745–9.
Shleifer, A. and R.W. Vishny (1997), ‘A survey of corporate governance’,
Journal of Finance, 52(2), 737–83.
Silberman, I. (1967), ‘On lognormality as a summary measure of concen-

tration’, American Economic Review, 57(4), 807–31.
Simon, H. and C. Bonini (1958), ‘The size distribution of business firms’,
Simon, H.A. (1957), ‘The compensation of executives’, Sociometry, 20(1),
32–5.
Simon, H.A. (1962), ‘The architecture of complexity’, Proceedings of the
American Philosophical Society, 106(6), 467–82.
Simon, H.A. (1968), ‘On judging the plausibility of theories’, in Logic,
Methodology and Philosophy of Science; Proceedings of the International
Congress for Logic, Methodology and Philosophy of Science, North-
Holland, Amsterdam, pp. 439–59.
Simons, T. and P. Ingram (2004), ‘An ecology of ideology: theory and evi-
dence from four populations’, Industrial and Corporate Change, 13(1),
33–59.
Singh, A. and G. Whittington (1975), ‘The size and growth of firms’,
Review of Economic Studies, 42(1), 15–26.
Singh, S.P., R.G. Reynolds and S. Muhammad (2001), ‘A gender-based
performance analysis of micro and small enterprises in Java, Indonesia’,
Journal of Small Business Management, 39(2), 174–82.
Slater, M. (1980), ‘The managerial limitation to the growth of firms’,
Economic Journal, 90(359), 520–28.
Sleuwaegen, L. and M. Goedhuys (2002), ‘Growth of firms in develop-
ing countries, evidence from Côte d’Ivoire’, Journal of Development
Economics, 68(1), 117–35.
Smolny, W. (1998), ‘Innovation, prices and employment – a theoretical
model and an empirical application for West German manufacturing
firms’, Journal of Industrial Economics, 46(3), 359–81.
Sorensen, J. and T. Stuart (2000), ‘Aging, obsolescence, and organizational
innovation’, Administrative Science Quarterly, 45(1), 81–112.
Spiezia, V. and M. Vivarelli (2000), ‘The analysis of technological change and
employment’, in M. Vivarelli and M. Pianta (eds), The Employment Impact
of Innovation: Evidence and Policy, London: Routledge, pp. 12–25.
Srholec, M. and B. Verspagen (2008), ‘The voyage of the Beagle
in innovation systems land. Explorations on sectors, innova-
tion, heterogeneity and selection’, Working Papers on Innovation
Studies 20080220, Centre for Technology, Innovation and Culture,
University of Oslo.
Stam, E. and E. Garnsey (2005), ‘New firms evolving in the knowledge
economy: problems and solutions around turning points’, Papers on
Economics and Evolution 2005-05, Max Planck Institute of Economics,
Evolutionary Economics Group.
Bibliography 187
Stanley, M., S. Buldyrev, S. Havlin, R. Mantegna, M. Salinger and H.

Stanley (1995), ‘Zipf plots and the size distribution of firms’, Economics
Letters, 49(4), 453–7.
Stanley, M.H.R., L.A.N. Amaral, S.V. Buldyrev, S. Havlin, H. Leschhorn,
P. Maass, M.A. Salinger and H.E. Stanley (1996), ‘Scaling behavior in
the growth of companies’, Nature, 379, 804–06.
Starbuck, W.H. (1971), ‘Organizational growth and development’, in
W.H. Starbuck (ed.), Organizational Growth and Development,
Harmondsworth, UK: Penguin Books, Chapter 1, pp. 11–141.
Steindl, J. (1965), Random Processes and the Growth of Firms, New York:
Hafner.
Stiglitz, J. and A. Weiss (1981), ‘Credit rationing in markets with imperfect
information’, American Economic Review, 71(3), 393–410.
Storey, D. (1994), ‘The role of legal status in influencing bank financing
and new firm growth’, Applied Economics, 26(2), 129–36.
Sutton, J. (1997), ‘Gibrat’s legacy’, Journal of Economic Literature, XXXV,
40–59.
Sutton, J. (1998), Technology and Market Structure: Theory and History,
Sutton, J. (2002), ‘The variance of firm growth rates: the “scaling” puzzle’,
Physica A, 312, 577–90.
Sutton, J. (2007), ‘Market share dynamics and the “persistence of leader-
ship” debate’, American Economic Review, 97(1), 222–41.
Szulanski, G. and S. Winter (2002), ‘Getting it right the second time’,
Harvard Business Review, 80(1), 62–9.
Teece, D.J., G. Pisano and A. Shuen (1997), ‘Dynamic capabilities and
strategic management’, Strategic Management Journal, 18(7), 509–33.
Teece, D.J. (1986), ‘Profiting from technological innovation: implications
for integration, collaboration, licensing and public policy’, Research
Policy, 15, 285–305.
Teruel-Carrizosa, M. (2008), ‘Gibrat’s law and the learning process’, Small
Business Economics, forthcoming.
Tether, B.S. (1997), ‘Growth diversity amongst innovative and technology-
based new and small firms: an interpretation’, New Technology, Work
and Employment, 12(2), 91–107.
Tsoukas, H. and R. Chia (2002), ‘On organizational becoming: rethinking
organizational change’, Organization Science, 13(5), 567–82.
Tybout, J.R. (2000), ‘Manufacturing firms in developing countries: how
well do they do, and why?’, Journal of Economic Literature, 38, 11–44.
van Dijk, M. and O. Nomaler (2000), ‘Technological diffusion patterns
and their effects on industrial dynamics’, DRUID Working Paper No.
00-6, Copenhagen Business School.
Van Reenen, J. (1997), ‘Employment and technological innovation: evi-

dence from UK manufacturing firms’, Journal of Labor Economics,
15(2), 255–84.
Variyam, J.N. and D.S. Kraybill (1992), ‘Empirical evidence on determi-
nants of firm growth’, Economics Letters, 38, 31–6.
Villalonga, B. (2004), ‘Does diversification cause the “diversification dis-
count”?’, Financial Management, 33(2), 5–27.
Viner, J. (1932), ‘Cost curves and supply curves’, Journal of Economics,
3(1), 23–46.
Vining, D. (1976), ‘Autocorrelated growth rates and the Pareto law: a
further analysis’, Journal of Political Economy, 84(2), 369–80.
Wagner, J. (1992), ‘Firm size, firm growth, and persistence of chance:
testing Gibrat’s law with establishment data from Lower Saxony,
1978–1989’, Small Business Economics, 4(2), 125–31.
Weick, K.E. (1995), Sensemaking in Organizations, Thousand Oaks, CA:
Sage Publications.
Weick, K.E. and R.E. Quinn (1999), ‘Organizational change and develop-
ment’, Annual Review of Psychology, 50, 361–86.
Weiss, C. (1998), ‘Size, growth, and survival in the Upper Austrian farm
sector’, Small Business Economics, 10(4), 305–12.
Wernerfelt, B. (1984), ‘A resource-based view of the firm’, Strategic
Whetten, D.A. (1987), ‘Organizational growth and decline processes’,
Annual Reviews in Sociology, 13(1), 335–58.
White, H. (1980), ‘A heteroskedasticity-consistent covariance matrix
estimator and a direct test for heteroskedasticity’, Econometrica, 48(4),
817–38.
Whited, T. (1992), ‘Debt, liquidity constraints, and corporate
investment: evidence from panel data’, Journal of Finance, 47(4),
1425–60.
Whited, T. (2006), ‘External finance constraints and the intertemporal
pattern of intermittent investment’, Journal of Financial Economics,
81(3), 467–502.
Wiggins, R. and T. Ruefli (2005), ‘Schumpeter’s ghost: is hypercompeti-
tion making the best of times shorter?’, Strategic Management Journal,
26(10), 887–911.
Wiklund, J. (2007), ‘Small firm growth strategies’, in A. Zackarakis
and S. Spinelli (eds), Entrepreneurship and Growth: The Engine of
Growth (2nd edn), Westport, CT: Praeger, Chapter 7, pp. 135–154.
Wiklund, J. and D. Shepherd (2003), ‘Aspiring for, and achieving
growth: the moderating role of resources and opportunities’, Journal of
Management Studies, 40(8), 1919–41.
Bibliography 189
Wiklund, J., P. Davidsson and F. Delmar (2003), ‘What do they think

and feel about growth? An expectancy-value approach to small busi-
ness managers’ attitudes toward growth’, Entrepreneurship Theory and
Practice, 27(3), 247–70.
Williamson, O. (1964), The Economics of Discretionary Behavior:
Managerial Objectives in a Theory of the Firm, Englewood Cliffs, NJ:
Prentice-Hall.
Williamson, O. (1967), ‘Hierarchical control and optimum firm size’,
Journal of Political Economy, 75(2), 123–38.
Winter, S.G. (1964), ‘Economic “natural selection” and the theory of the
firm’, Yale Economic Essays, 4(1), 225–72.
Winter, S.G. (1971), ‘Satisficing, selection and the innovating remnant’,
Winter, S.G. (2003), ‘Understanding dynamic capabilities’, Strategic
Winter, S.G. (1995), ‘Four Rs of profitability: rents, resources, rou-
tines, and replication’, in C. Montgomery (ed.), Resource-Based and
Evolutionary Theories of the Firm: Towards a Synthesis, Boston, MA:
Kluwer Academic.
Winter, S.G. and G. Szulanski (2001), ‘Replication as strategy’,
Organization Science, 12(6), 730–43.
Witt, U. (1998), ‘Imagination and leadership: the neglected dimension of
an evolutionary theory of the firm’, Journal of Economic Behavior and
Witt, U. (2000), ‘Changing cognitive frames – changing organizational
forms: an entrepreneurial theory of organizational development’,
Industrial and Corporate Change, 9(4), 733–55.
Witt, U. (2007), ‘Firms as realizations of entrepreneurial visions’, Journal
of Management Studies, 44(7), 1125–40.
Wynarczyk, P. and R. Watson (2005), ‘Firm growth and supply chain
partnerships: an empirical analysis of UK SME subcontractors’, Small
Yasuda, T. (2005), ‘Firm growth, size, age and behavior in Japanese manu-
facturing’, Small Business Economics, 24(1), 1–15.
You, J.-I. (1995), ‘Small firms in economic theory’, Cambridge Journal of
Economics, 19, 441–62.
Zahra, S.A. and G. George (2002), ‘International entrepreneurship: the
current status of the field and future research agenda’, in M.A. Hitt, R.D.
Ireland, S.M. Camp and D.L. Sexton (eds), Strategic Entrepreneurship:
Creating a New Mindset, Oxford, UK: Blackwell Publishing, Chapter
12, pp. 255–88.
Zahra, S.A., R.D. Ireland and M.A. Hitt (2000), ‘International expansion
by new venture firms: international diversity, mode of market entry,

technological learning, and performance’, Academy of Management
Journal, 43(5), 925–50.
Zollo, M. and H. Singh (2004), ‘Deliberate learning in corporate acquisi-
tions: post-acquisition strategies and integration capability in US bank
mergers’, Strategic Management Journal, 25(13), 1233–56.
Index
absolute growth 10–11, 20 Baily, M. et al. 63, 65
Acemoglu, D. et al. 3 Baily, M. and D. Farrell 65, 73, 108
acquisitions 94, 101, 150 Barnett, S. and P. Sakellaris 53
and competition between firms 125 Barney, J. 103
conglomerate mergers 104, 124–5 Barron, D. et al. 41, 85, 109
and diversification 122, 124 Bartelsman, E. et al. 17, 18, 62, 95, 134
and FDI 126, 127 Bartelsman, E. and M. Doms 64
and financial performance 105 Batsch, L. 120
vs internal growth 124–5, 147 Baumol, W. et al. 9–10, 103
Acs, Z. and D. Audretsch 131 Becchetti, L. and G. Trovato 43
Adamic, L. and B. Huberman 22, 23 Beck, T. et al. 90, 91, 95–6, 98, 131
advertising 86, 87, 91, 97 Bellone, F. et al. 131
age distribution 20–22 Berger, P. and E. Ofek 123
agency theory 58, 103, 121 Bhagat, S. et al. 122
Aghion, P. and P. Howitt 77 Bigsten, A. and M. Gebreeyesus 42, 85
Akerlof, G. 54 Birch, D. 11, 129
Alchian, A. 105–6 Birley, S. and P. Westhead 94
Allen, F. et al. 132 Bloom, N. and J. Van Reemen 77
Almus, M. 11, 88, 92 Blundell, R. et al. 52, 53, 60
Almus, M. and E. Nerlinger 41, 46 Boeri, T. and U. Cramer 46
Amaral, L. 29–30, 44, 45 Bond, S. and C. Meghir 53, 55
Amihud, Y. and B. Lev 113, 121 Bond, S. et al. 53, 55, 60
Amirkhalkhali, S. and A. booms and recessions 28–9, 94
Mukhopadhyay 41, 95 Boone, J. et al. 87
Andriani, P. and B. McKelvey 137 ‘born global’ firms 127, 150
Angelini, P. and A. Generale 17 Bottazzi, G. et al. 6, 15, 16, 25, 26, 27,
Ansoff, I. 114, 120, 125, 126 42, 44, 46–7, 57, 64, 79, 95
antitrust 91, 121, 122, 150 Bottazzi, G. and A. Secchi 16–17, 19,
Aoki, M. 112 26, 28, 30, 36, 41, 44, 45, 46, 86,
appreciative theorizing 13, 111, 146 95, 115, 147
Arabsheibani, G. et al. 61 bounded rationality 5, 54, 56–7, 61
Ashton, T. 25 Brock, W. 146
Audretsch, D. 4, 39, 42, 86, 93, 114, Broekel, T. (and Coad) 73
131 Bronars, S. and D. Deere 112
Audretsch, D. and J. Elston 55 Brouwer, E. et al. 82
Audretsch, D. and T. Mahmood 90, Brown, C. and J. Medoff 132
93 Brynjolfsson, E. and L. Hitt 3
Austria 41, 46 Buldyrev, S. et al. 28
Autio, E. et al. 128
autocorrelation dynamics 36 Cabral, L. and J. Mata 17
Axtell, R. 15, 22, 23 Calvo, J. 41, 98
191
Camerer, C. and D. Lovallo 61 Delmar, F. and J. Wiklund 136, 151

Campa, J. and S. Kedia 123 Denmark 18, 26, 85, 95, 97
Canada 18, 55 developing countries 88–9, 92, 114, 131
Carden, S. 77 see also individual countries
Carpenter, R. and B. Petersen 55 Dickerson, A. 105, 125
Catley, S. and R. Hamilton 89, 90 Dierickx, I. and K. Cool 102
Caves, R. 43, 46, 131, 144 Disney, R. et al. 66, 67
Caves, R. and M. Porter 86 diversification 3, 91, 104, 105, 122–3,
Cefis, E. et al. 43, 101 124, 126–8
Cefis, E. and L. Orsenigo 77 growth strategies 112–13, 114,
Central Limit Theorem 44 117–18, 119–23
Chandler, A. 3, 4, 122 and management 105
Chapin, F. 136–7 and risk reduction 121–2
Chesher, A. 20, 43, 46 Dixit, A. 86, 112
Chile 64 Dixon, R. 31–2, 117, 149
Chirinko, R. 51, 52, 53 Dobson, S. and B. Gerrard 69
Churchill, N. and V. Lewis 139, 140 Doms, M. et al. 28, 64, 82, 147
Coase, R. 100 Dosi, G. 6, 16, 57, 106, 148
cognitive leadership 139 Dosi, G. et al. 5, 16, 106
Colombo, M. and L. Grilli 10 Dosi, G. and M. Grazzi 5
competition 30–31, 37, 93 Dosi, G. and D. Lovallo 61
competitive advantage 4, 5, 103, 124 Doukas, J. and O. Kan 123
and employment growth 87 Downie, J. 106
and game theory 86 downsizing 3, 63–4, 107
and growth by acquisition 125 Droucopoulos, V. 42
and profits 65 Druilhe, C. and E. Garnsey 140
and size of firm 86, 91 Dunne, P. and A. Hughes 41, 44, 45, 95
and small firms 86, 131–2 Dunne, T. et al. 18, 41, 62, 85, 90, 147
conglomerate mergers 104, 124–5
Cooper, R. et al. 28 economies of growth 102, 103, 113,
copy EXACTLY! policy 118–19 114, 135, 145
Cordes, C. et al. 114, 139 economies of scale 2–3, 44, 69, 112,
corporate refocusing 120–21 135, 149
Corsino, M. 79 Eisenhardt, K. and J. Martin 102, 103
Côte d’Ivoire 41, 64, 85, 90, 95 Eisenhardt, K. and C. Schoonhoven 10
credit market imperfections 55 employment levels 9, 10–11, 18, 28
Cressy, R. 62 and competition between firms 87
Cromie, S. 90 and employee behaviour 116
Cummins, J. et al. 55 growth propagation 32–7
and innovation 76, 81–3
Das, S. 85 and sales growth 58, 70, 71–2
Davidsson, P. et al. 124, 129, 135, 147 and uncertainty 92
Davis, S. et al. 11, 94, 129 and worker morale 112
De Fabritiis, G. et al. 44 entrepreneurial characteristics 88–90
de Jong, J. and O. Marsili 94 entry rates 86, 108, 131
de Meza, D. and D. Webb 62 and competitive advantage 112
de Wit, G. 20, 22 entry costs and internal growth 124
Degryse, H. and A. de Jong 59 excess of new firms 61–2
Del Monte, A. and E. Papagni 79 failure rate 62
Delmar, F. 9, 10, 69, 94, 124, 129 and FDI 127
Index 193
and job creation 129 future research

and productivity 64, 65, 66, 67, 68, discrete growth events, impact of
134 147
survival 93, 95, 133, 134 econometric investigations into firm
Erickson, T. and T. Whited 53–4 growth 147
Ericson, R. and A. Pakes 134–5 firm-specific variables 144
Ethiopia 42, 85 Gibrat’s law 148
Euler equation model 52–3, 60 growth and ‘fitness’, relationship
Evangelista, R. and M. Savona 82 between 144
Evans, D. 41, 44, 85, 95 inter-firm competition 87
evolutionary economics 6–7, 105–8 investment, lumpy nature of 38
evolutionary theory of growth 56–9, performance and ambition 144
60, 61, 63 profits, productivity and firm growth
exit rates 56, 92 75
and diversification 123 theoretical perspective on firm
exit hazards 86, 90, 106, 133, 134, growth 145–6
155–6 VAR models of firm growth 70
and productivity 63, 64, 65, 66, 68
small firms 40–41, 43, 129, 133, 142 Gabe, T. and D. Kraybill 41, 93, 95
survival of the fittest 7, 107–8, 144 Galeotti, M. et al. 53, 55
exporting 4, 91, 126–8 game theory 86
Garnsey, E. 2, 40, 47, 134, 140
Fagiolo, G. and A. Luzzi 59, 60, 90, Gartner, W. 149
98 Gaussian distribution 26, 29
family-owned firms 113, 114 GDP 96
Fazzari, S. et al. 51, 53, 54–5, 59, 60 Germany 18, 46, 55, 79, 82, 88, 90, 92,
FDI 4, 126–7 95
Feldman, M. 102 growth rates of manufacturing firms
financial constraints 41, 46, 82, 97
capital intensity and growth 90–91 Geroski, P. 77, 86, 96, 108
evaluating importance of 59–62 Geroski, P. et al. 28, 46, 53, 79, 97, 101
and profits, productivity and firm Geroski, P. and K. Gugler 25, 37, 43,
growth 54–5, 58 85, 87, 90, 91, 94, 97, 144
and selection effects on profits Geroski, P. and S. Machin 78
50–63 Geroski, P. and M. Mazzucato 46
small firms 17, 62, 132, 136 Geroski, P. and S. Toker 78–9, 91, 93,
‘financial pecking-order’ theory 58 97
Finland 18 Gibrat’s law 7, 14, 17–20, 29, 39–48,
Fishman, A. and R. Rob 88 147–8
Fizaine, F. 85 autocorrelation of growth rates 45–7
Fleck, J. 81 and business cycles 94
Fluck, Z. and A. Lynch 123 econometric issues 43
Foster, L. et al. 64, 65–6, 67, 68, 131 and economies of scale 44
France 18, 50–51, 82, 85, 90, 95 firm size and average growth 40–43,
growth rate distribution in 53, 85
manufacturing 14, 15, 17, 26–8, firm size and growth rate variance
42, 44, 46, 57, 71, 98, 107, 113 43–5
Freel, M. 79, 81 heterogeneous growth and
Freeland, R. 107 autocorrelation patterns 47
Friedman, M. 11 and independent sub-markets 45
and Kolmogorov–Smirnov tests 47 and ‘growth of the fitter’ 7, 107–8,

model 39–40 144
model of random growth shocks growth rate distribution 25–38
22–3 growth rate distribution,
and negative dependence of growth heavy-tailed nature 25, 26, 28,
on size 42, 43–4 29, 34, 36, 86–7
and neoclassical optimizing models industry-specific factors 92–4
101 inherent tendency towards 115–16
objections to 20 and innovation see innovation
and relevant lags 46 intentionality of growth 115–17
and sample selection bias 43 and inter-firm partnerships 91
scaling of growth rate variance internal 116, 124–5
44–5 and investment spikes 37–8
and services sector 42 Laplace distribution 25–6, 28, 30, 32
as stochastic process 115 macroeconomic factors 94–6
Gilchrist, S. and C. Himmelberg 55 and managerial resources 102,
GMM (Generalized Method of 112–14, 115, 116–17, 119–20,
Moments) techniques 70 121, 125, 138–9
Goddard, J. et al. 41, 46 modelling stages of 137–41
Goedhuys, M. and L. Sleuwaegen 41, and nature of firm activities 91
64, 80, 85, 90, 92, 95 as ongoing process 70
Goergen, M. and L. Renneboog 59 and ownership structure 90
Gomes, J. 54 and population ecology 108–9
government schemes 4, 59, 62, 90 R2 values 96–9
Grabowski, H. et al. 4 reduced form models 59, 60
Graham, J. et al. 123 relative 10, 11, 66
Greenhalgh, C. et al. 82 spurts 29–37
Greiner, L. 84, 115, 138 strategies 111–28
Griliches, Z. 77 ‘tent-shaped’ distribution 6, 25, 26,
Griliches, Z. and J. Mairesse 5 27–8
Griliches, Z. and H. Regev 64 time-varying moments 28–9
growth VAR models of firm growth
advantages of 112–13 processes 69–73
and age of firm 84–5 very large firms 95
aspirations 2, 88, 89, 92, 136 Guariglia, A. 51, 55
attitudes to 111–17 Guiso, L. and G. Parigi 92
between-plant reallocation 66–8
and capital intensity 90–91 Hadlock, C. 55
and centrality of network 91 Haleblian, J. and S. Finkelstein 150
and competition 93 Hall, B. 28, 41, 43, 44, 76, 82, 95
and conglomerate merger 104, 124–5 Hannan, M. 32, 108, 109
control-loss argument 113 Hannan, M. and J. Freeman 108, 109,
coordination problems 113 137, 141
copy EXACTLY! policy 118–19 Harada, N. 133
desirability of 112–15 Hardwick, P. and M. Adams 42, 91, 94,
determinants 96–9 108
disadvantages of 113–15 Harhoff, D. et al. 43, 44, 90, 97
and diversification see diversification Harrison, R. et al. 82
and exports 4, 91, 126–8 Hart, P. 14, 41
and external business advice 92 Hart, P. and N. Oulton 42, 44
Index 195
Hart, P. and R. Pearce 95 investment

Hay, D. and D. Morris 112 and cash flow 54–6, 58, 59, 60
Hay, M. and K. Kamshad 77, 91, 105, plant-level 28
114, 132, 136 and productivity growth relationship
Hayashi, F. 51–2 38
Headd, B. 133, 134 and profit, relationship between 51
Headd, B. and B. Kirchhoff 93, 133–4 R&D 73, 78–9, 82
hierarchical nature of firm 3, 32–5 spikes 37–8
high-profit firms and business and uncertainty 92
opportunities, lack of interest in Ireland 79
107 Italy 18, 42–3, 55, 79, 82, 90, 92, 95,
Higson, C. et al. 29, 94 114
Hines, J. and R. Thaler 58 growth rate distribution in
Hisrich, R. and S. Ozturk 90 manufacturing 17, 26, 43, 44,
Holl, P. 91, 104–5 46, 57, 64, 73, 79, 90, 98, 107
Hopenhayn, H. 106, 134
Hoshi, T. et al. 55 Japan 41, 46, 55, 85, 87, 95
Hu, X. and F. Schiantarelli 55 Jensen, M. 20, 26, 54, 58, 122, 125
Hubbard, R. 51 Jensen, M. and W. Meckling 103
Huberman, B. and L. Adamic 22, 23 Johanson, J. and J. Vahlne 127
Hughes, A. 62, 132 Johnson, P. et al. 42
Hugo, O. and E. Garnsey 140 Jones, M. and N. Coviello 129
human capital 54, 88–9 Jovanovic, B. 106, 134
Hyland, D. and J. Diltz 121
Hymer, S. and P. Pashigian 43 Kaldor, N. 70, 113
Kalecki, M. 20
Idson, T. and W. Oi 131 Kay, N. 101
Ijiri, Y. and H. Simon 16, 29, 30, 45, Kazanjian, R. and R. Drazin 140–41
115, 147–8 Kesten, H. 22
imperfect markets theory 54–6, 59, Klette, T. and Z. Griliches 77
60 knowledge, tacit knowledge transfer
India 20, 21, 85, 89, 131 118, 126, 127
Indonesia 89 Kogut, B. and U. Zander 127
industrial classification scheme 94 Kuemmerle, W. 128
innovation Kumar, M. 41, 45, 95, 97
and economic performance 77–8
and employment growth 76, 81–3 labour productivity growth 63–4, 71–3
failed attempts, outcome of 80–81 Lamont, O. 55
and industrial classification schemes Lang, L. and R. Stulz 123
94 Laplace distribution 25–6, 28, 30, 32
industry-specific factors 93 learning-by-doing 66, 67, 102
measurement methods 79–80 Lee, Y. et al. 44
profit margins 78 legal status 90
sales growth 76, 77–81 Leiponen, A. and I. Drejer 94
small firms 77, 131 Lensink, R. et al. 92
technological unemployment 82–3 Lerner, J. 62
and uncertainty 77–8, 80 Levenson, A. and K. Willard 62
see also R&D limited liability companies 85, 90, 95,
instrumental variables 156 150
internationalization 126–8, 150 Little, I. 9–10, 25, 62, 131
Liu, J. et al. 41, 64, 85, 90, 97, 98 Miner, J. et al. 136
Lockett, A. et al. 124, 147 Minimum Efficient Scale (MES) 41, 93
lognormal distribution 14, 16, 20, 101 Minkoff, D. 109
Lotti, F. and E. Santarelli 17 Mitzenmacher, M. 22
Lotti, F. et al. 39, 42–3 Montgomery, C. 102–3, 121, 122, 123
Lucas, R. 101 motivation, lack of, in larger firms 114,
Luttmer, E. 22 117
Mowery, D. 42, 78
McCombie, J. 70 Mueller, D. 104, 119, 125
MacDonald, C. 118, 119 multiplant firms 68, 90, 96, 97, 98
McDougall, P. et al. 127, 128 Myers, S. 58
McKelvey, B. and P. Andriani 29, 36,
86, 137 Nafziger, E. and D. Terrell 131
McPherson, M. 41, 43, 88, 89, 94, 95, Nelson, R. 114
97, 98 Nelson, R. and S. Winter 5, 77, 106,
Maksimovic, V. and G. Phillips 64, 111, 146
93–4, 123, 124, 125, 147 neoclassical economics 7, 8, 61, 62, 63,
management 69, 145
cognitive leadership 139 ‘optimal size’ 56, 60, 100–101, 102,
control and small firms 105 103, 107, 112, 145
and diversification 105 q-theory 51–4, 55, 57, 59, 60, 117
entrenchment 121 Netherlands 18, 42, 92, 114
incentives 104, 121 networks 91, 126
pursuit of growth 58, 69, 70, 104 Niefert, M. 82
resources 3, 54, 91, 102, 112–14, 115,
116–17, 119–20, 121, 125, 138–9 O’Farrell, P. and D. Hitchens 136
shareholder-wealth-maximizing Oliner, S. and G. Rudebusch 55
managers 59–61 ‘optimal size’ theory 56, 60, 100–101,
talent distribution and small firms 102, 103, 107, 112, 145
101 organizational
theory of the firm 103–5 change 3, 5, 11, 112, 114, 141
Mansfield, E. 19, 40, 42, 77–8, 147 slack 31, 32, 70
manufacturing industries see under Oviatt, B. and P. McDougall 128
individual countries ownership structure 90
market
concentration 93 Pareto distribution 16, 22–3, 28
power 107 Parkinson, C. 115–16
selection 61 partnership, inter-firm 91
value 4, 51, 52, 53, 57, 76, 97 patents 77, 78, 79–80, 82, 97
Markides, C. 123 Paulré, B. 121
Marris, R. 4, 5, 58, 103–5, 119, 125 Pavcnik, N. 64
Marshall, A. 132–3 Pavitt, K. 131
Marsili, O. 16, 43, 94, 96 Penrose, E. 4, 31, 32, 47, 69, 91, 111,
Martin, J. and A. Sayrak 122 119–20, 131, 135, 145–6, 148
Matia, K. et al. 44, 45 theory of growth of firm 102–3, 113,
Matsusaka, J. 123 116–17
Mead, D. and C. Liedholm 88–9 pharmaceutical industry 16–17, 41, 43,
mergers and acquisitions see 44, 45, 46, 79
acquisitions Phillips, B. and B. Kirchhoff 22, 42, 133
Metcalfe, J. 76, 106 population ecology 108–9
Index 197
Portugal 17, 18 risk reduction, and diversification

Powell, W. et al. 91 121–2
Power, L. 38 Rivkin, J. 118
Prais, S. 14, 41 Roberts, J. 112
process theory of internationalization Robson, P. 79, 81
127 Robson, P. and R. Bennett 51, 85, 88,
production 91, 92, 97, 132
Minimum Efficient Scale and growth Robson, P. and B. Obeng 88, 89
rates 41 Roll, R. 125
objectives 112–13 Roper, S. 79
productivity Rossi-Hansberg, E. and M. Wright 16,
and growth 63–5, 71–3 42, 90–91
growth, decomposing 65–8 Rumelt, R. 123
relative 63–8
profits sales growth 9, 28, 29, 93
and competition 65 and ‘growth of the fitter’ 7, 107–8,
and diversification 123 144
and growth, financial constraints and innovation 76, 77–81
and selection effects 50–63 Salop, S. 86
innovation and firm growth 78 Salter, W. 38
and investment, relationship between Samuels, J. 41
4, 5, 10, 51 Santarelli, E. and M. Vivarelli 17, 61,
and productivity 57–8 131
Sapienza, H. et al. 128
q-theory 51–4, 55, 57, 59, 60, 117 Sarno, D. 59, 132
Quandt, R. 16 Sastry, M. 141
Schaller, H. 53, 55
R&D 3, 4, 6, 54, 56, 73, 74, 78–80, 82, Scherer, F. 74, 78
120, 123, 126, 131, 135 Schiantarelli, F. 51, 53, 55
see also innovation Schivardi, F. and R. Torrini 92, 114
R2 values 96–9 Schumpeter, J. 3, 87, 105
Radice, H. 91 Secchi, A. 6, 16–17, 19, 26, 28, 30, 36,
Rajan, R. and J. Wulf 3 41, 44, 45, 46, 86, 95, 115, 147
Ramsden, J. and G. Kiss-Haypal 16 Segarra, A. et al. 20, 21
Rao, R. (and Coad) 37, 57, 73, 74, 76, Segarra, A. and M. Callejon 41
79–80, 82, 145, 147, 150 selection
reallocation 6, 49, 65, 66–8, 144 bias 1, 43
Reed, W. 22 and conglomerates 64
regional effects 95 and differential growth 65, 106, 107,
Reichstein, T. and M. Dahl 85, 95, 97 108, 144
Reichstein, T. and M. Jensen 20, 26 and financial constraints 50–63
relative growth 10, 11 service industry 18, 42, 81, 82, 85, 90,
replication 117–19 112, 118, 120, 127
replicator dynamics 106 sex, and entrepreneurial characteristics
resources 89–90
and growth 102–3 Shanmugam, K. and S. Bhaduri 85
management 3, 54, 91, 102, 112–14, Shepherd, D. and J. Wiklund 10, 69,
115, 116–17, 119–20, 121, 125, 89, 136
138–9 Shleifer, A. and R. Vishny 121, 122,
sharing 93 125, 150–51
Silberman, I. 16 Southern Africa 41, 88, 89, 94, 95, 97

Simon, H. 12, 16, 29, 30, 31, 33, 45, Spain 20, 21, 41, 98
56–7, 115, 147–8 specialization 5, 95, 131
Simon, H. and C. Bonini 14 Spiezia, V. and M. Vivarelli 81
Simons, T. and P. Ingram 109 Srholec, M. and B. Verspagen 94
Singh, A. and G. Whittington 41, 45 Stam, E. and E. Garnsey 140
Singh, S. et al. 89 Stanley, M. et al. 16, 25
size distributions 14–24 Starbuck, W. 8, 113, 148, 149
lognormal model 14, 16, 17 Steindl, J. 14–16
size threshold 114 Stiglitz, J. and A. Weiss 54, 90
upper tail shape problems 16, 22 Storey, D. 10, 90
Slater, M. 102 structure–conduct–performance
Sleuwaegen, L. and M. Goedhuys 41, paradigm 92
64, 80, 85, 90, 92, 95 Subbotin distribution 26
small firms survival rates 20, 93, 95, 133–5
advantages of 130–31 Sutton, J. 29, 37, 42, 45, 87, 115
ambition, lack of 136 Sweden, growth by acquisition 124
and cognitive leadership 139 Szulanski, G. and S. Winter 118
and competition 86, 131–2
financial account structure 17, 62, Taiwan 41, 64, 85, 90, 97
132, 136 Tamvada, J. 20, 21, 89
growth desire 135–6 taxation 55, 92, 113, 114
growth rates 41 technological progress 1, 3–4, 20, 92–3
hazard rates 140 Teece, D. 103
independence of 136 Teruel-Carrizosa, M. 42
and innovation 77, 131 Tether, B. 114
and internal growth 124 theoretical perspectives 100–110
and internationalization 127–8 threshold effect 91–2, 114
and large firms, differences between Tobin’s q 51–4, 55, 57, 59, 60, 117
130–32 transaction costs theory 3, 100–101,
and large firms, growth pattern 126
differences 132–7 Tsoukas, H. and R. Chia 141
and management control 105 Tybout, J. 92, 114
and management talent distribution
101 UK
and MES (minimum efficient scales) advertising and sales growth 91
72 cash flow and investment 55
modelling stages of firm growth employment growth and innovation
137–41 82
and negative autocorrelation 135 excessive start-ups 61, 62
productivity levels 131, 134–5 financial structure of small firms
regional effects 95 132
routinization of operations 140 firm growth and age of firms 85
and specialization 131 firm growth and business cycles 94
structural change in growing firms firm growth and human capital 88
136–7 firm growth of large firms 95
survival struggle 133–5 firm growth and management
and threshold effect 91–2 characteristics 91
Smolny, W. 82 growth rates of life insurance
Sorensen, J. and T. Stuart 109 companies 42, 108
Index 199
growth rates of manufacturing firms size-wage relationship 132

41, 42, 45, 46, 67, 79, 82, 97, small firms 18, 62, 90, 97, 129, 131,
107, 132 133–4
high-tech firms and lack of desire for survival rates of small firms 133–4
growth 114–15 time-varying moments of growth
innovation and small firms 131 rate 28, 29
profit rates and management control
104–5 van Dijk, M. and O. Nomaler 108
R&D investment 78–9, 82 Van Reenen, J. 77, 82
small firms 18, 51, 79, 85, 88, 91, 97, VAR (vector autoregression) models of
114–15, 131, 132 firm growth processes 69–73
software firms, limits to growth 114 Variyam, J. and D. Kraybill 42, 85, 90,
time-varying moments of growth 97
rate 29 Villalonga, B. 123
uncertainty 5, 7, 92, 127 Viner, J. 100
unionization 90, 112, 132 Vining, D. 16
US
cash flow and investment 55 wage levels 3, 132
employment growth and innovation Wagner, J. 41, 46
82 Weick, K. 70
firm growth and age of firms 85 Weick, K. and R. Quinn 32, 141
firm growth and business cycles 94 Weiss, C. 10, 41, 46
firm growth and dispersion and Wernerfelt, B. 102
volatility factors 94–5 Whetten, D. 11, 112, 113, 138, 141
firm growth of large firms 95 White, H. 43
firm growth and ownership structure Whited, T. 37, 53–4, 55, 147
90 Wiklund, J. 10, 69, 86, 88, 89, 117, 134,
firm growth and plant size 93 136
growth by acquisition 125 will-o’-the-wisp models 113, 149
growth rates of manufacturing firms Williamson, O. 33, 103, 113
14, 15, 16, 40, 41, 44, 45, 46, 54, Winter, S. 5, 6, 77, 103, 106, 111, 118,
57, 67, 78, 82, 85, 107 146
growth rates of retail firms 67, 68 Witt, U. 114, 139
innovation and small firms 131 Wynarczyk, P. and R. Watson 91
investment and productivity growth
relationship 38 Yasuda, T. 41, 85
New York Credit Unions 41, 85, 109 You, J.-I. 43, 101, 130
oil company investment 55 ‘Yule’ distributions 14
plant-level investment and growth
rate distribution 28 Zahra, S. et al. 128
post-entry growth rates 95 Zipf distribution 22, 23
R&D investment 73, 78, 82 Zollo, M. and H. Singh 150

Alex Coad-The Growth of Firms - A Survey of Theories and Empirical Evidence (New Perspectives On The Modern Corporation) - Edward Elgar Pub (2009)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Alex Coad-The Growth of Firms - A Survey of Theories and Empirical Evidence (New Perspectives On The Modern Corporation) - Edward Elgar Pub (2009)

Uploaded by

Copyright:

Available Formats

The Growth of Firms

NEW PERSPECTIVES ON THE MODERN CORPORATION

● corporate governance and corporate responsibility, including environmental

Titles in the series include:

Corporate Governance, Organization and the Firm

The Modern Firm, Corporate Governance and Investment

The Growth of Firms

Knowledge in the Development Economies

NEW PERSPECTIVES ON THE MODERN CORPORATION

All rights reserved. No part of this publication may be reproduced, stored in a

Edward Elgar Publishing, Inc.

A catalogue record for this book

Library of Congress Control Number: 2009924303

ISBN 978 1 84844 327 3

Printed and bound by MPG Books Group, UK

2.1 Firm size distribution: French manufacturing 15

2.1 The share of activity of small firms 18

To keep up with developments in this field, we need an up-to-date cata-

1.1 FIRM GROWTH IN A HISTORICAL

It is instructive to place firm growth in a historical perspective. In the past,

costs. Furthermore, as firms continued to expand they began to question

competitive struggle is becoming increasingly fierce. The focus of investors

1.2 THEORETICAL FOUNDATIONS

be cautious of notions of a ‘representative firm’ which might lead us to

still very relevant: ‘The subject of organizational growth has progressed

1.3 MEASURING SIZE AND GROWTH

A number of the chapters in this book (Chapters 2 to 7 in particular) refer

to growth of profits. Among the least conventional indicators, one finds

The most common way of measuring growth is by taking log-differences

log(Git) 5 log((Sit/Si,t−1) 2 1) 5 log(Sit/Si,t−1) 5 log(Sit) 2 log(Si,t−1)

There are also a few other ways of measuring growth rates.

concerned with the number of jobs in a region than the performance of

1.4 STRUCTURE OF THE BOOK

data and more descriptive in nature. Chapter 9 describes firms’ attitudes

2.1 SIZE DISTRIBUTIONS

Source: Axtell (2001).

Figure 2.2 Probability density function of the sizes of US manufacturing

2.2 GIBRAT’S LOGNORMAL FIRM SIZE

Robert Gibrat’s (1931) theory of a ‘law of proportionate effect’ was hatched

Source: Based on Bartelsman et al. (2005), Tables 2 and 3.

explanandum, and wanted to see which underlying growth process could

xt 2 xt−1 5 etxt−1 (2.1)

which can be developed to obtain

xt 5 (1 1 et)xt−1 5 x0(1 1 e1)(1 1 e2) . . . (1 1 et) (2.2)

It is then possible to take logarithms in order to approximate log(1 1 et)

size tends to infinity (Kalecki, 1945).2 In addition, we do not observe the

2.3 AGE DISTRIBUTION

The age distribution of a population of firms is also worth mentioning.

Source: Taken from Coad and Tamvada (2008).

Figure 2.3 Kernel density of the age distribution of Indian small-scale

Figure 2.4 Kernel density of the age distribution of Spanish firms.

2.4 EXPLAINING THE PARETO FIRM SIZE

Early models of industrial dynamics focused on firms above a certain size

xt 2 xt−1 5 etxt−1 (2.6)

which can be developed to obtain a lognormally distributed firm size

P(t) 5 lelt (2.8)

In order to obtain the mixture of these two distributions, we apply the

and, as in Adamic and Huberman (1999), this can be developed to yield:

where C is a constant and is given by C 5 l/s ( ! (e/s) 2 1 2l) . The

growth in an attempt to explain how such size distributions could emerge.

3.1 GROWTH RATE DISTRIBUTIONS

The functional form of the Laplace (symmetric exponential) density is:

Source: Bottazzi et al. (2005).