The Growth of Firms: A Survey of Theories and Empirical Evidence (New Perspectives on the Modern Corporation)

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 publication 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 performance industrial economics, organisational behaviour, innovation and competitiveness 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 Acknowledgements 1 2 3 4 5 6 7 8 9 10 11

vi vii

Introduction Firm size distributions Growth rate distributions Gibrat’s law Profits, productivity and firm growth Innovation and firm growth Other determinants of firm growth Theoretical perspectives Growth strategies Growth of small and large firms Conclusion

Notes Bibliography Index

1 14 25 39 49 76 84 100 111 129 143 152 161 191

v

Figures and tables FIGURES 2.1 2.2 2.3 2.4 3.1 3.2 3.3 3.4 5.1 5.2 5.3 10.1

Firm size distribution: French manufacturing Firm size distribution: US manufacturing Age distribution for small Indian firms Age distribution for Spanish firms Distribution of sales growth rates Distribution of employment growth rates Firm growth as the addition of discrete resources Employment growth in a model of firm growth Scatterplots of profits vs sales growth Process of firm growth and productivity growth Process of firm growth and R&D expenditure Stages of growth

15 15 21 21 27 27 33 35 50 72 74 130

TABLES 2.1 5.1 5.2 5.3 7.1

The share of activity of small firms Deriving a regression equation in a neoclassical q model Profits and growth: types of regression equation Decomposing productivity growth Values of R2 obtained from growth regressions

vi

18 52 60 67 97

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, commissioning 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 longitudinal 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 operations 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 largescale 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 heterogeneity 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 catalogue 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 different, 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 explanatory concepts to guide inquiry and make sense of evidence. A mass of undigested 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, discussing 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 contains 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 reinforced 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 therefore 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 managers 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 production and standardization has given way to an organization of production 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 productivityenhancing 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 transaction 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

4

The growth of firms

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 paradoxically, 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 importance. 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, arguably, 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 economic 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, profitability, 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 heterogeneity 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

6

The growth of firms

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 regression estimators that focus on summary point estimates corresponding to ‘the average effect for the average firm’. This approach is particularly illsuited 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 evolutionary 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 selection 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 assumption 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 investigate the evolutionary principle of ‘growth of the fitter’ and it is astonishing 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 perspective 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 theoretical 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

8

The growth of firms

still very relevant: ‘The subject of organizational growth has progressed beyond abysmal darkness. It is ready for – and badly needs – solid, systematic 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 theoretical 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 neoclassical 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 organizations. 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 actually 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 indicator 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 advantage 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

10

The growth of firms

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. Git 5

Sit 2 Si,t21 Si,t21

5

St St21 2 St21 St21

5

St 21 St21

(1.1)

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 relatively 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) # Ei,t21

(1.3)

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: DHS Indexit 5

Eit 2 Ei,t21 1/2 (Ei,t21 1 Ei,t)

(1.4)

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 primarily 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 processes 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 literature. (For an introduction to organizational decline, the reader is referred to Whetten, 1987).

12

1.4

The growth of firms

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 relative 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 determinants of firm growth have had a limited success, and the combined explanatory 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 theories of firm growth outlined in this chapter have only a limited success in explaining growth, perhaps because growth is idiosyncratic, firms are heterogeneous, and as a result of this it is difficult to make wide-reaching generalizations. 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 empirical regularities and offer descriptive, ‘appreciative’ theoretical explanations, 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 fundamental stylized facts about firm size and growth. In fact, it was while contemplating 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 lognormal 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

Pr

0.01

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

Frequency

10–4

10–7

10–10

10–13 1

Source:

10

102 103 104 Firm size (employees)

105

106

Axtell (2001).

Figure 2.2

Probability density function of the sizes of US manufacturing firms in 1997

16

The growth of firms

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 statistical 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

Firm size distributions

17

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 financial 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 distribution 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 establishments followed a skew distribution that resembled the lognormal. Gibrat considered the emergence of the firm size distribution as an outcome or

18

Source:

Note:

69.9 77.9 73.6 87.5 74.9 – 74.0 84.8 86.7 70.5

Manufacturing 87.9 90.2 78.8 96.5 – – 90.8 94.5 96.8 92.8

Business services 16.6 23.6 13.9 34.4 – – 30.2 25.8 31.2 27.7

Total economy 5.8 11.3 17.0 30.3 8.3 – 16.1 13.0 16.9 15.7

Manufacturing 20.6 33.8 12.1 46.3 – – 33.4 33.0 41.9 39.8

Business services

Share of employment (%)

26.4 17.0 33.5 10.5 – 12.7 13.3 13.0 6.5 16.8

Total economy

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

80.3 39.1 32.1 15.3 40.7 40.5 30.4 27.8 18.3 31.0

Manufacturing

21.4 11.5 35.7 6.8 – 12.0 12.7 9.9 5.3 11.4

Business services

Ave. no. employees per firm

The columns labelled ‘share of employment’ refer to the employment share of firms with fewer than 20 employees.

86.7 87.9 78.6 93.1 – – 90.0 92.6 95.8 86.3

Total economy

Absolute number (%)

The importance of small firms (i.e. firms with fewer than 20 employees) across broad sectors and countries, 1989–94

US Western Germany France Italy UK Canada Denmark Finland Netherlands Portugal

Table 2.1

Firm size distributions

19

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 probability 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 identically 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 (2.5) 2s2t 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

P (xt) 5

e2a

20

The growth of firms

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 lognormal 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 influence 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 corresponding 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.

Firm size distributions

21

1 0.1

Pr

0.01 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

Pr

0.001 le-04

le-05

le-06 0

Source:

20

40

60 Age

80

100

120

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

22

The growth of firms

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 Gibrattype 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)

Firm size distributions

23

which can be developed to obtain a lognormally distributed firm size distribution: 1

(lnxt 2e# t) 2

b (2.7) 2s2t 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 (xt) 5

e2a

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: P (xt) 5 3 lelt #

1 xt"2ps

2

e2a

(lnxt 2e# t) 2 2s2t

b

dt

(2.9)

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 distribution (section 2.1). Two distributions, in particular, stand out as candidates for approximations to the empirical distribution – the lognormal or the Pareto distribution. We then turned to theoretical models of firm

24

The growth of firms

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 empirical 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 distribution (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

26

The growth of firms

The functional form of the Laplace (symmetric exponential) density is: fL (x) 5

1 20 x2m 0/a e 2a

(3.1)

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 (introduced to the firm growth literature in Bottazzi et al., 2002). The Subbotin distribution can be formally presented by the following equation: fS (x) 5

1 1 x2m b e2 b 0 a 0 ( ) 2ab G 1/b 1 1 1/b

(3.2)

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 heavytailed, 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 employment growth rates for French manufacturing firms with over 20 employees.

Growth rate distributions

27 1998 2000 2002

Prob

1

0.1

0.01

0.001 –3

Source:

–2

–1 0 Conditional growth rate

1

2

Bottazzi et al. (2005).

Figure 3.1 Distribution of sales growth rates of French manufacturing firms

1998 2000 2002

Prob

1

0.1

0.01

0.001 –2

Source:

–1.5

–1

–0.5 0 0.5 Conditional growth rate

1

1.5

Coad (2006).

Figure 3.2

Distribution of employment growth rates of French manufacturing firms

2

28

The growth of 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 valueadded (Bottazzi et al., 2007). Other researchers have found evidence of considerable 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, multiproduct 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 distribution 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 discretion in their growth rates of assets during booms but face stricter discipline

Growth rate distributions

29

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 aggregate 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 nonGaussian 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 distribution of firm growth rates have also taken the approach of stochastic explanations. Amaral et al. (1997) develop a model in which the emergence

30

The growth of firms

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 emergence 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, competitive 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 increasing 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, effectively, required in order to reconcile the model with empirical work into the

Growth rate distributions

31

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 heterogeneity 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, complementary 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 criticality 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’

32

The growth of firms

(Dixon, 1953, p. 50). Similarly, Hannan writes: ‘changes in one organizational 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 interactions. 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 organization 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 composed 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

Growth rate distributions

Figure 3.3

33

Supervisors

Supervisors

Productive workers

Productive workers

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 organized 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.

34

The growth of firms

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 employment corresponds to the number of both productive workers and supervisors 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), liquidity 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 successive 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 following functional form: P(Δx $ |Δn 5 1) 5 a1−

(3.3)

Growth rate distributions

35

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 distribution 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

36

The growth of firms

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 P (xt 2 xt21 . 1 0 xt21 2 xt22 5 1) 5

1 a21

(3.6)

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 performance 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 investigated 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 critical 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

Growth rate distributions

37

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 dynamics 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 discovery 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 competition 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 contribution to industrial development. Quantile regression is a useful econometric 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.

38

The growth of firms

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 empirical 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 investigate 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

40

The growth of firms 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 distributed 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 immediate reaction, in recent decades it has spawned a flood of empirical work. Nowadays, Gibrat’s ‘Law of Proportionate Effect’ constitutes a benchmark 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, considers 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 production (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 manufacturing 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 frequently 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 relationship 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.

42

The growth of firms

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). TeruelCarrizosa (2008) observes that while small manufacturing firms tend to experience fast growth relative to their larger counterparts, the inequality 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 convincingly 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 rendition. 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 essentially 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 selection 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 negative 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 selection bias has only a small influence on the Gibrat coefficient. In short, the ‘problem of sample selection does not seem to significantly affect the relationship 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 asymptotic 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 concludes 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

44

The growth of firms

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 existence 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 understanding 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 scalingof-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 ‘submarkets’. 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 interestingly 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 correlation 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 autocorrelation was later reported in comparable studies by Kumar (1985) and Dunne and Hughes (1994).

46

The growth of firms

More recently, availability of better datasets has encouraged the consideration 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 pharmaceutical industry, and Bottazzi and Secchi (2003a) for US manufacturing. 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 statistically 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 insignificantly 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 construct 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 employment 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 negative 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 experience 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 experience positive autocorrelation irrespective of their growth rate in the previous 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).

48

4.5

The growth of firms

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 foundation. 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 generalize, 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 datasets 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 productivity) and its growth rate. In fact, a number of theoretical authors tend to take it for granted that firms with higher performance will reinvest 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

50

The growth of firms

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 growth rate (t-1:t)

growth rate (t-1:t)

1 0.5 0

–1

0

–1 –0.5

0.5 0 profit rate (t-1)

1

–0.5

ISIC24: chemicals and chemical products 1 0.5 0

0.5 0

–1 –0.5

0.5 0 profit rate (t-1)

1

–0.5

ISIC28: fabricated metal products

0.5 0 profit rate (t-1)

1

ISIC29: machinery and equipment 1 growth rate (t-1:t)

1 growth rate (t-1:t)

1

ISIC25: rubber and plastic products

–1

0.5 0

–1

0.5 0

–1 –0.5

Source:

0.5 0 profit rate (t-1)

1 growth rate (t-1:t)

growth rate (t-1:t)

0.5

0.5 0 profit rate (t-1)

1

–0.5

0.5 0 profit rate (t-1)

1

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 associated 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 theoretical perspectives on firm growth, and also the rather different policy implications that emerge from them.1 The three perspectives are the neoclassical 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 desirable outcome. 5.1.1

q Theory

Neoclassical q-theory states that firm-level investment should be determined by future prospects of return. Assuming that stock prices can accurately 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,

52

The growth of firms

Table 5.1

An example of a neoclassical q model: how Blundell et al. (1992) derive the regression equation

Equation

Description

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13)

Intertemporal capital market arbitrage condition Solving (1) on an infinite horizon Defining the discount factor b over an infinite horizon Substituting for dividend payments in the firm’s stock market value Defining the firm’s after-tax net revenue First-order condition for investment The evolution of the shadow price of capital Rearranging (6) to obtain marginal q Rearranging (8) assuming a quadratic functional form for adjustment costs Rewriting marginal q assuming linear homogeneity of production and adjustment costs Expressing the expected depreciation allowances on an infinite horizon Expressing the expected present value of all cash flows associated with debt Regression equation

1982): that firms operate in perfectly competitive product and factor markets, that firms also have linear homogeneous production and adjustment 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 expectations’, 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 investigations 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

Profits, productivity and firm growth

53

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 outperformed 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,

54

The growth of firms

2000; Gomes, 2001). This may occur if the stock market is not perfectly efficient 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 investment 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 catchphrases 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:

Profits, productivity and firm growth

55

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) elaborated upon the implications for policy. They underlined the importance of investment opportunities being forgone due to credit market imperfections. 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 literature 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

56

The growth of firms

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 presented 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 sensitivity 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 classification 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

Profits, productivity and firm growth

57

‘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– – 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 evolutionary 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 manufacturing. 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 paneldata instrumental variable techniques to account for a lag structure, firmspecific 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

S

58

The growth of firms

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 empiricallybased ‘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 ‘managerial’ 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 shareholders. Growth is thus maximized subject to certain constraints (that is a minimum value for the firm’s shares). Under these circumstances, investment will respond positively to improvements in current financial performance. 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

Profits, productivity and firm growth

59

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 performance 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 motivations 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

60

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).

Cash flow taken as a proxy for financial constraints (without controlling for q). Any sensitivity of sales growth to cash flow is interpreted as financial constraints. Sales growth should be associated with operating margin according to the principle of ‘growth of the fitter’.

Investment dynamics should follow the optimal investment path in the context of parametric adjustment costs. Marginal costs of 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 explanation could be that firms are financially constrained. Any explanatory power of cash flow over and above that of q indicates financial constraints.

(I/K ) it 5 b1 (I/K ) i,t21 2 b2 (I/K) 2i,t21 2 b3 (P/K) i,t21 1 b4 (Y/K) i,t21 1 eit

Imperfect markets (I/K)it 5 b1qit 1 b2 (CF/K)it 1 eit (e.g. Fazzari et al., 1988a) Reduced form models (DS/S) 5 b2 (CF/S) it 1 eit (e.g. Fagiolo and Luzzi, 2006) Evolutionary approach (DS/S) 5 b2 (OM/S) it 1 eit (e.g. Coad, 2007c)

If the assumptions hold, investment should be entirely explained by q.

(I/K)it 5 a 1 b1qit 1 eit

q-theory (e.g. Blundell et al., 1992) Euler equation (e.g. Bond et al., 2003b)

Remarks

Basic regression equation

Types of regression equation associated with the different theoretical perspectives

Theoretical approach

Table 5.2

Profits, productivity and firm growth

61

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 financial 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 fundamental 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

62

The growth of firms

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 suggests 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 startups 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 underachieving 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 creditrationed as 6.36 per cent, and conclude that ‘the extent of true credit rationing 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).

Profits, productivity and firm growth

63

Summary Neoclassical economists hold that the existence of any sensitivity of investment 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 consider 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 theoretical 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

64

The growth of firms

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 manufacturing 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 deregulation. 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 manufacturing 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 conglomerate firms. They observe that the growth rates of plants in both singleand 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 profitability 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

Profits, productivity and firm growth

65

predicting growth rates. Put differently, it appears that selection only operates 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 ‘suboptimal’ 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 discriminating 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 differential 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: (LPi,t21 2 LPt21) Dqit DLPt 5 a qi,t21DLPit 1 a i[C i[C 1 a DLPitDqit i[C

(5.2)

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

i[N

qit (TFPit2TFPt21) 2 a

(5.3) i[X

qi,t21 (TFPi,t212TFPt21)

66

The growth of firms

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) establishments. 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 positive 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 productivity 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 techniques. The between-plant component of productivity growth (Column

67

UK manufacturing

US retail US manufacturing

Disney et al. (2003)

Foster et al. (2006) Foster et al. (2008)

TFP LP TFP LP LP TFP

Prod. Measure 1977–1987 1977–1987 1980–92 1980–92 1987–1997 1977–1997

Years 0.48 0.45 0.05 0.48 0.16 0.36

Withinplant 20.08 20.13 0.15 0.04 0.24 20.17

Betweenplant

(2)

0.34 0.37 0.26 20.01 20.39 0.50

Crossplant

(3)

0.26 0.31 0.54 0.49 0.98 0.30

Net entry

(4)

1.00 1.00 1.00 1.00 1.00 1.00

Total

sum (1)–(4)

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.

US manufacturing

Data

(1)

Decomposing productivity growth. Productivity growth decompositions calculated using the Foster et al. (1998) technique (see equations 5.2 and 5.3)

Foster et al. (1998)

Table 5.3

68

The growth of firms

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 establishments 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 experienced tremendous productivity growth over the period of analysis. The effect of entry and exit on productivity growth is quite remarkable in magnitude (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 singleplant firms. Foster et al. (2008) report results for productivity growth decompositions 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 withinplant 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.

Profits, productivity and firm growth

5.3

69

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 variable 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 associated 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 profitable 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

70

The growth of firms

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 production. 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 applying 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 instrumentalvariables 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 causality 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

71

Profits, Productivity and Firm Growth Coad (2007b) applies a reduced-form VAR model to data on French manufacturing 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 subsequent growth of sales and of profits. As could be expected, sales growth and productivity growth also appear to make a relatively large contribution to the subsequent growth of profits. Indeed, sales growth has a sizeable 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 subsequent 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 coefficients 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 relatively 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,

72

The growth of firms

Empl. growth

Empl. growth

Sales growth

Sales growth

Profits growth

Profits growth

Prod. growth

Prod. growth

t Source:

time

t+1

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 following 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 percentage 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 consistent 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.

Profits, productivity and firm growth

73

Coad and Broekel (2007) undertake a similar analysis with an alternative indicator of firm-level productivity, using non-parametric frontier analysis 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 followed 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 relatively ‘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 economic theory. This relationship has important implications for the allocation 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.

74

The growth of firms

Empl. growth

Empl. growth

Sales growth

Sales growth

Profits growth

Profits growth

R&D growth

R&D growth

t Source:

time

t+1

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.

Profits, productivity and firm growth

75

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 opportunities, 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 production 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

6.1

77

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 innovation, 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 innovation 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 pioneering 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

78

The growth of firms

probability that, given technical success, the resulting product or process will be commercialized (y); and finally the probability that, given commercialization, 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 investigate the relationship between innovation and sales growth. Our gleaning of this literature yields a rather motley harvest. (This may be due to difficulties 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 corporations 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’ innovation) 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 corporate 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 innovation has a significant positive effect on sales growth when included in an

Innovation and firm growth

79

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) investigates 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 manufacturing 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 innovation 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 innovative 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 innovation 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 limitation 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

80

The growth of firms

R&D expenditures as well as patenting activity to create a synthetic ‘innovativeness’5 variable for each firm year. In this way we extract the common variance associated with each of these indicators while discarding the idiosyncratic 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 innovation. 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.

Innovation and firm growth

81

“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 influence 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 replacing 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 products). 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 reduction 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 production. Fifth, an increase in incomes may translate into higher consumption and thus higher employment. Sixth, there may be compensation via new

82

The growth of firms

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 different 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 publications 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 distinction 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 combined 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 manufacturing 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

Innovation and firm growth

83

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, generating 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 generally 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 employment 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 associated. 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 investigated 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 attention 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 opportunities 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 experience 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).

86

7.2

The growth of firms

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 relationship 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

Other determinants of firm growth

87

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 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 (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). 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 statistical 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 theoretical 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 Interfirm 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

88

The growth of firms

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’ consumption 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, university 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 founders 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

Other determinants of firm growth

89

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 entrepreneur 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 entrepreneurs 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 opportunities 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

90

The growth of firms

(Catley and Hamilton, 1998; Hisrich and Ozturk, 1999). Female entrepreneurs 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 manufacturing 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 manufacturing 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 provides 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 capitalintensive 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 comparisons of capital-intensive and labour-intensive sectors, and, consistent with their model, they find some evidence of different regimes of growth across

Other determinants of firm growth

91

sectors. Capital-intensive sectors seem to have a stronger scale dependence, such that smaller firms grow much faster than larger ones. In labourintensive 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 management. 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 analyses 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 diversification 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 advertising-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

92

The growth of firms

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 countries, 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 convincing evidence that uncertainty of demand plays a significant role in reducing 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 investment. 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 determined the appropriate conduct of incumbent firms, which in turn determined 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,

Other determinants of firm growth

93

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 concentration. 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 containing many fast-growing firms also contain a large number of firms experiencing 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

94

The growth of firms

fairly constant across industries, there is more variation when one compares 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 consequence, 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 countries in his dataset. Although it has been observed that more of the variation 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 variation) 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

Other determinants of firm growth

95

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 metropolitan 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

96

The growth of firms

analyses a size-stratified firm-level survey database covering over 4000 firms in 54 countries. They observe that firms in richer, larger, and fastergrowing 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 productivity) 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 competition, 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 empirical 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 industry 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

97

Reichstein and Dahl (2004)

Robson & Bennett (2000) Geroski & Gugler (2004)

Liu et al. (1999)

Harhoff et al. (1998)

Geroski et al. (1997)

1671 small firms in 5 Southern African countries 271 large quoted UK firms 1976–82 About 10 000 West German firms Over 900 Taiwanese manufacturing plants Over 1000 SMEs in Britain in 1997 Large firms in 14 European countries, over 100 000 obs. 1994–98 8739 Danish firms, 1994–96

Around 700–800 quoted UK companies 422 small businesses in Georgia, USA 209 leading UK firms

Kumar (1985)

Variyam & Kraybill (1992) Geroski & Toker (1996) McPherson (1996)

Data

Study

Size, age, regional specialization, concentration, industry

Age, size, industry dummies, capital–labour ratio, sales per worker, dummies for R&D and exporting activity Size, age, exports, profits, industry, innovation and technology, use of external advice, strategy variables Size, age, subsidiaries, diversification, growth of rivals

32%

Size, innovation, advertising, industry growth, industry concentration Firm age and size, dummies for sector and location, human capital and socio-economic variables Market value, lagged firm growth, innovations, patents, industry growth Size, age, subsidiary, diversification, legal status, industry

1–2%

5–6%

4–8%

19–22%

8%

17–19%

13–20%

11–17%

1–4%

R2

Size, age, multi-plant firms, industry

Size, lagged growth

Control variables

Table 7.1 A survey of R2 values obtained from regressions where the dependent variable is the growth rate of a firm or plant

98

About 1000 Spanish firms

Calvo (2006)

Dummies for government/foreign ownership, export status, subsidies, sector of activity; controls for number of competitors, GDP, GDP per capita, GDP growth, financial/ legal/corruption obstacles Size, age, legal liability, product/process innovation, technology, sample selection Size, age, cash flow, dummies for multi-plant firms, year and industry dummies Gross operating margin, lagged growth, lagged size, industry and year dummies

Control variables

4–8%

2–3%

9%

2–3%

R2

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.

8405 French manufacturing firms, 1996–2004

14 277 Italian firms 1995–2000

Survey data covering over 4000 firms of all sizes, in 54 countries

Beck et al. (2005b)

Fagiolo and Luzzi (2006) Coad (2007d)

Data

(continued)

Study

Table 7.1

Other determinants of firm growth

99

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 characteristics 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 neoclassical 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 neoclassical perspective is that firms are attracted to some sort of ‘optimal size’ (Viner, 1932). This optimal size is the profit-maximizing level of production, 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 coordination 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 coordinated 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 specificity, 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 lognormal 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 organization 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.

102

8.2

The growth of firms

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 valuecreating growth opportunities (and in particular, the training of new managers). 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 successfully 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 nonetheless 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, fastgrowing 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

Theoretical perspectives

103

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 associated 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 abilities 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 literature – 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 resourcebased 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

104

The growth of firms

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 growthmaximization. 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 quadratic 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 humpshaped relationship between growth rate and profit rate, with additional growth having a negative effect on profits only beyond a certain ‘profitmaximizing’ 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 ownercontrolled 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 surprisingly, 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

Theoretical perspectives

105

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 associated 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 performance, 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 competition 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 selection sets the economy on the path of progress, as fitter firms survive and

106

The growth of firms

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 environment. 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 evolutionary 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 ‘viability’). 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– – 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

S

Theoretical perspectives

107

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 business opportunities and sought to restrict growth expenditure even when GM was a highly profitable company. If this is the case, then stricter internal 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 competence. 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 productivity, 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 suggested. 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

108

The growth of firms

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 principle of ‘growth of the fitter’ (Baily and Farrell, 2006). If high performance firms were observed to have the fastest growth rates, then selective processes 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 competition 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 organizations in the context of niche-specific growth patterns without focusing as much on heterogeneity between organizations occupying the same niche.

Theoretical perspectives

109

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 literature 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 ideological 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 organizations have become less efficient and less well adapted to the current business 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

110

The growth of firms

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 principle 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 exceptionally 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 diversification, 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 distinction 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

112

The growth of firms

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 promotion 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 challenging 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 commitment 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 diversified 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 diversified 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 hierarchical 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

114

The growth of firms

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 bureaucratic 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 coordinate a more routinized method of production (Witt, 1998, 2000, 2007; Cordes et al., 2008). A common ideology and a cooperative working environment is substituted 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 diseconomies 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 manufacturing firms in developing countries describes how small firms have incentives to stay small and informal to avoid taxes. In contrast, mediumsized 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

Growth strategies

115

scientists that founded the firms are often not well prepared for the management 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 congenial 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 perspectives on firm growth, such as Gibrat’s law, view it as a passive absorption 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, considers 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 independent 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

116

The growth of firms

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 promotion). 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 carrying 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 managerial decision and ‘human will’ rather than being a response to technological factors.4 If, on the other hand, these unused managerial services are

Growth strategies

117

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 automatically 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 anticipate 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 rationality 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

118

The growth of firms

markets. There are thus a number of issues and complications that accompany 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 selecting 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 newlyintroduced 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

Growth strategies

119

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 temperature 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 overwhelming competitive benefit to introducing a change’ (MacDonald, 1998, p. 2) (emphasis in the original). Furthermore, if a modification has been suggested 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 diversification. 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 limiting firm growth. As a firm continues its operations, incumbent managers gradually gain experience, and new managers can be trained and integrated

120

The growth of firms

into the firm, thus expanding the firm’s resource base. In this way, managerial 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 specific 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 commencing 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 synergies 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 diversification 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 refocusing, 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

Growth strategies

121

focus on exploiting certain specific capabilities. Refocusing should not be seen as a ‘return’ to the firm’s previous condition, however, but as a strategic 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 diversification. (In fact, empirical work on diversification has mainly focused on testing the hypothesis that diversification is detrimental to firm performance.) 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 encouraged 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, diversification 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 pleasures of empire-building’ (Montgomery, 1994, p. 166). This point is clearly illustrated by Hyland and Diltz (2002), who compare managerial compensation 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 shareholders, 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

122

The growth of firms

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 associated 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 businesses 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 diversification 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 capabilities did not give them a distinct competitive advantage (Chandler, 1992). The 1970s were also characterized by unrelated acquisitions and overdiversification. 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).

Growth strategies

123

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 multisegment 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 diversification 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 underachievers 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 strategy 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).

124

9.3

The growth of firms

INTERNAL GROWTH VS GROWTH BY ACQUISITION

Both internal growth and acquisition can be used as means of either expanding market share in a particular industry or of diversifying into new industries. 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 industry. 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 successfully 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 contrast, 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.

Growth strategies

125

Nevertheless, a strategy of growth by acquisition is particularly difficult 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) estimate, 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 economists 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 shareholder 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 managerial 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.

126

9.4

The growth of firms

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 products 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, consumers 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 knowledge 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

Growth strategies

127

construction of new facilities. FDI presents several advantages, such as the possibility of choosing a location with favourable labour markets or proximity 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 commitment, however, and so firms that face uncertainty and have limited knowledge 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 uncertainty 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 gingerly 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

128

The growth of firms

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 experience 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 valuable knowledge and allowing them to augment their existing capabilities (Kuemmerle, 2002; Zahra et al., 2000). Indeed, internationalization is a significant 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 replication 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, knowledge 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 generating 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 statistical 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 innovative, 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

130

The growth of firms

large

Phase 1

2

3

4

5 collaboration

coordination

Size of Organization

“?”

delegation red tape direction

creativity

leadership

small

young

Source:

control

autonomy evolution: stages of growth revolution: stages of crisis

Age of Organization

mature

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 focusing on the dichotomous distinction between small and large firms (sections 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 decision-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 countries (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 unemployment. 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 independent 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

132

The growth of firms

competition, however. In fact, survey evidence for small businesses indicates 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 association between employer size and wages, which holds at both the establishment 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 conditions, 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

Growth of small and large firms

133

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 struggle 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

134

The growth of firms

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 performance 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 ownermanagers 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 successful 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 counterparts, 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

Growth of small and large firms

135

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 separated 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 opportunities 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 probability 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 fastgrowth firms,5 the majority of small firms do not experience fast growth.

136

The growth of firms

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 instrumental 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 compromise 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 businesses 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 subsequent 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

Growth of small and large firms

137

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 departments. 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 versions 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 animated 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

138

The growth of firms

the firm undergoes a radical metamorphosis as it grows, with the entrepreneur’s vision and dynamism gradually being replaced by a more bureaucratic 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, priorities 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 maturity. These stages correspond to configurations of problems, strategies and priorities that firms are likely to face as they grow, as well as describing 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 management 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’.

Growth of small and large firms

139

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 conception. 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 cognitive 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 knowledge, 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 entrepreneur’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 entrepreneur 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 interests. Individualistic cognitive frames can then spread, favouring opportunistic 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

140

The growth of firms

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 transformation 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 processes a prelude to every new problem they encounter’ (p. 541). As time goes by, however, firms learn about their business and develop problemsolving 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 consequence, 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 different challenges and deal with them in different ways (Hugo and Garnsey, 2005; Stam and Garnsey, 2005). Firms may struggle with the same recurring 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;

Growth of small and large firms

141

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 organizational 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 episodic because they take a distant perspective of organizations and focus on general trends in their long-term development over their life span.

142

10.4

The growth of firms

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 transformations 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 difficulties (such as issues of delegation, monitoring and coordination) as they move from one stage to the next, eventually becoming large-scale bureaucratic 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 experiences. 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

144

The growth of firms

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 crosssection) 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 financial 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) recommendation: ‘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 explaining growth, empirical studies have not really picked up on this in a satisfactory 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 quantile 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 productivity 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 consequence, 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 proposition 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 population 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 resourcebased perspective is particularly rich with insights. In our view, it is

146

The growth of firms

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 application 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’ theorizing 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 possible. 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 stylized 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 particularly robust and that has led to the construction of new theoretical models (such as Bottazzi and Secchi, 2006a). The remarkable lumpiness of investment 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 conditional 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 elements (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

148

The growth of firms

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 researchers 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, systematic 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 accustomed 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 satisfactory 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 consistent 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 administrative 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 opportunities 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

150

The growth of firms

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 complacent 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 increasingly due to employees, many of whom the original entrepreneur has little time to meet. The firm grows further and becomes a bureaucratic organization; it turns from a small team into an impersonal economic instrument, 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 possibilities 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 allocation 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 significantly lower than the Laplace value of 1. This highlights the importance 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 indivisibilities 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 steadystate 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

154

3.

4. 5.

6.

7.

The growth of firms

upon the distinction between replacement investment and expansionary 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). 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 decision 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. Note that Whited (2006) uses cash flow as a proxy variable for investment opportunities. 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 mortgaged 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 payas-you-go basis, financing our growth primarily out of earnings rather than by borrowing money’ (Packard, 1995, p. 85). A wealth of evidence on this topic is provided in Beck et al. (2005b). In particular, they observe that while financial constraints can be significant for small firms in developing countries, they are not important for large firms in developed countries (see also Angelini and Generale, 2008). 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

8.

9.

10.

11.

12.

13. 14. 15.

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. 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). Unpublished calculations on French and US data show that the effect of profits on growth tends to increase in magnitude when firms of different 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. 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. One difference between the ‘managerial’ and ‘free cash flow’ perspectives and the evolutionary perspective, however, is that the observed investment–cash flow sensitivities are signs of value-reducing investment in the first case but are more likely to be value-creating in the latter. 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. The Small Business Administration (SBA) provided $2.8 billion in guaranteed loans to small firms in 1986 alone. Notwithstanding this latter result, Bottazzi et al. (2008b) observe a strong positive relationship between productivity and profitability. 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 profitability that predicts exit. The reason, he suggests, is that high-profit

156

16.

17.

18.

19.

The growth of firms

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. Alternative schemes of decomposing productivity growth can be found in Griliches and Regev (1995), Olley and Pakes (1996) and Aw et al. (2001). Traditional productivity indicators measure output in terms of total sales rather than production, which makes establishments that charge higher prices appear more productive. Conventional econometric techniques that are applied to investigate 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. 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 innovation 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 pharmaceutical 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 confidentlytitled 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 elasticity’ (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 government 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 unionization and firm growth disappears, such that, for a given firm, unionization 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 innovation 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

158

3.

4. 5. 6.

7.

8.

9.

10.

11.

12. 13.

The growth of firms

ne réussissant jamais à atteindre un état d’équilibre complet dans l’utilisation de ses resources’ (Lesourne, 1973, p. 92). Winter writes ‘routines clearly qualify as resources, given the expansive 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). Penrose’s analysis considers that firms operate in a world of constant returns to scale. For a more complete reappraisal of Edith Penrose’s contribution to economics, the reader is referred to Pitelis (2002). 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 promotions all tend to be more closely related to the size or changes in size of the firm than to its profits’ [emphasis added]. This quadratic specification of the relationship between profits and growth, however, has not been investigated in the empirical literature in a satisfactory way. 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 neoclassical ‘optimum’, thereby vindicating the predictions of neoclassical theory. See also Metcalfe (2007), who generalizes his model of industry evolution by attributing different propensities to grow to different firms. 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). There is ample evidence that the population ecology perspective explicitly 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 organizations?’ Furthermore, Hannan (2005) opens his literature review with this very same question. As Geroski (2001, p. 535) notes, there is a ‘heavy reliance on density dependence to drive dynamics.’ 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.

160

The growth of firms

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; survival; 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

162

The growth of firms

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: redirecting 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 distributions’, 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 district, 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 industry 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 productivity 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 investment 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 business 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 analysis 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 retention, 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.

164

The growth of firms

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 refocusing 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 econometric 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 patterns 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 industrial structure: some evidence from the Italian manufacturing industry’, Industrial and Corporate Change, 11, 705–23. Bottazzi, G., A. Secchi and F. Tamagni (2008b), ‘Productivity, profitability 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 constrained? 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.

166

The growth of firms

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 distribution: 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, observable 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 constrained 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 barriers: conjectural decisions and the contrived deterrence to new competition’, 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 strategies, 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 constraints: 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 explanation of the exponential distribution of firm growth rates’, Documents de travail du Centre d’Economie de la Sorbonne, Université PanthéonSorbonne (Paris 1), Centre d’Economie de la Sorbonne.

168

The growth of firms

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 intellectual 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 naturalistic 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 behavior, 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 computer hardware industry in India’, International Journal of Industrial Organization, 13(1), 111–26. Davidsson, P. and F. Delmar (2006), ‘High-growth firms and their contribution 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 taxonomy 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 asymmetric 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 distributions 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: asymmetric 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

170

The growth of firms

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 managers’ 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 highgrowth firm’, Journal of Business Venturing, 18(2), 189–216. Dickerson, A.P., H.D. Gibson and E. Tsakalotos (2000), ‘Internal vs external 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 sustainability of competitive advantage’, Management Science, 35(12), 1504–11. Disney, R., J. Haskel and Y. Heden (2003), ‘Restructuring and productivity 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 manufacturing 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 Journal of Industrial Organization, 13, 523–42. Dosi, G. (2004), ‘A very reasonable objective still beyond our reach: economics as an empirically disciplined social science’, in M. Augier and J. March (eds), Models of a Man: Essays in Memory of Herbert A. Simon, Cambridge, MA: MIT Press. 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 procedures 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 discount’, 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 companies 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 relationship between investment and q’, Journal of Political Economy, 108(5), 1027–57. Ericson, R. and A. Pakes (1995), ‘Markov-perfect industry dynamics: a framework 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.

172

The growth of firms

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 of Political Economy, 95(4), 657–74. Fagiolo, G. and A. Luzzi (2006), ‘Do liquidity constraints matter in explaining firm size and growth? Some evidence from the Italian manufacturing 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, financing 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 industry 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, reallocation, 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?’, American Economic Review, 98(1), 394–425. Freel, M. and P. Robson (2004), ‘Small firm innovation, growth and performance: 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-innovators?’, 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 framework 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 development incentives on employment growth of establishments’, Journal of Regional Science, 42(4), 703–30. Galeotti, M., F. Schiantarelli and F. Jaramillo (1994), ‘Investment decisions 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 organizational 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 Organization, 14(2), 141–58. 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 differences in firm size transitory or permanent?’, Journal of Applied Econometrics, 18(1), 47–59.

174

The growth of firms

Geroski, P.A. (1995), ‘What do we know about entry?’, International Journal of Industrial Organization, 13, 421–40. 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 Organization, 20(3), 415–33. 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 diversification 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’, Quarterly Journal of Economics, 114(1), 185–227. 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 learning 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 organizations’, 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 automobile 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,

176

The growth of firms

construction, trade and service industries’, Journal of Industrial Economics, 46(4), 453–88. Harrison, R., J. Jaumandreu, J. Mairesse and B. Peters (2005), ‘Does innovation 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, implementation and impediments’, Business Strategy Review, 5(3), 49–68. Hayashi, F. (1982), ‘Tobin’s marginal q and average q: a neoclassical interpretation’, 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 dynamics 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 developing 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 equilibrium’, 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’, Quarterly Journal of Economics, 106(1), 33–60. Hu, X. and F. Schiantarelli (1998), ‘Investment and capital market imperfections: 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 worldwide 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 creation 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 of Political Economy, 70(6), 556–69. 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 business 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.

178

The growth of firms

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 internationalisation’, 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 innovation’, Journal of Political Economy, 112(5), 986–1018. Kogut, B. and U. Zander (2003), ‘Knowledge of the firm and the evolutionary 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 international 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.

180

The growth of firms

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’, Quarterly Journal of Economics, 122(3), 1103–44. 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 measuring 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, acquisitions 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 evidence’, Academy of Management Journal, 35(2), 398–412. Markides, C. (1995), ‘Diversification, restructuring and economic performance’, Strategic Management Journal, 16(2), 101–18. Marris, R. (1963), ‘A model of the managerial enterprise’, Quarterly Journal of Economics, 77(2), 185–209. 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 shareholder 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: interpretations from follow-up data’, Journal of Applied Psychology, 79(4), 627–30.

182

The growth of firms

Minkoff, D. (1999), ‘Bending with the wind: strategic change and adaptation 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 Journal of Economics, 83(4), 643–59. 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 telecommunications 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 improvements: evidence from Chilean plants’, Review of Economic Studies, 69(1), 245–76. Pavitt, K. (1984), ‘Sectoral patterns of technical change: towards a taxonomy 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 biotechnology’, Administrative Science Quarterly, 41(1), 116–45. Power, L. (1998), ‘The missing link: technology, investment, and productivity’, 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.

184

The growth of firms

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 Business Economics, 30(4), 385–403. 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 comparison of the strategies of German, UK and Irish companies’, Small Business Economics, 9(6), 523–37. Rossi-Hansberg, E. and M. Wright (2007), ‘Establishment size dynamics 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 backward regions of the [sic] south Italy’, Applied Financial Economics, 18, 821–33. Sastry, M. (1997), ‘Problems and paradoxes in a model of punctuated 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 pharmaceutical R&D spending’, Health Affairs, 20(5), 216–20. Schiantarelli, F. (1996), ‘Financial constraints and investment: methodological 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 turbulence: 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 accumulation 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.

186

The growth of firms

Silberman, I. (1967), ‘On lognormality as a summary measure of concentration’, American Economic Review, 57(4), 807–31. Simon, H. and C. Bonini (1958), ‘The size distribution of business firms’, American Economic Review, 48(4), 607–17. 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, NorthHolland, Amsterdam, pp. 439–59. Simons, T. and P. Ingram (2004), ‘An ecology of ideology: theory and evidence 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 developing 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, innovation, 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, Cambridge, MA: MIT Press. 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 leadership” 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 technologybased 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.

188

The growth of firms

Van Reenen, J. (1997), ‘Employment and technological innovation: evidence from UK manufacturing firms’, Journal of Labor Economics, 15(2), 255–84. Variyam, J.N. and D.S. Kraybill (1992), ‘Empirical evidence on determinants of firm growth’, Economics Letters, 38, 31–6. Villalonga, B. (2004), ‘Does diversification cause the “diversification discount”?’, 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 development’, 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 Management Journal, 5(2), 171–80. 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 hypercompetition 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 business 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’, Quarterly Journal of Economics, 85(2), 237–61. Winter, S.G. (2003), ‘Understanding dynamic capabilities’, Strategic Management Journal, 24(10), 991–5. Winter, S.G. (1995), ‘Four Rs of profitability: rents, resources, routines, 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 Organization, 35(2), 161–77. 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 Business Economics, 24(1), 39–51. Yasuda, T. (2005), ‘Firm growth, size, age and behavior in Japanese manufacturing’, 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

190

The growth of firms

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 acquisitions: post-acquisition strategies and integration capability in US bank mergers’, Strategic Management Journal, 25(13), 1233–56.

Index absolute growth 10–11, 20 Acemoglu, D. et al. 3 acquisitions 94, 101, 150 and competition between firms 125 conglomerate mergers 104, 124–5 and diversification 122, 124 and FDI 126, 127 and financial performance 105 vs internal growth 124–5, 147 Acs, Z. and D. Audretsch 131 Adamic, L. and B. Huberman 22, 23 advertising 86, 87, 91, 97 age distribution 20–22 agency theory 58, 103, 121 Aghion, P. and P. Howitt 77 Akerlof, G. 54 Alchian, A. 105–6 Allen, F. et al. 132 Almus, M. 11, 88, 92 Almus, M. and E. Nerlinger 41, 46 Amaral, L. 29–30, 44, 45 Amihud, Y. and B. Lev 113, 121 Amirkhalkhali, S. and A. Mukhopadhyay 41, 95 Andriani, P. and B. McKelvey 137 Angelini, P. and A. Generale 17 Ansoff, I. 114, 120, 125, 126 antitrust 91, 121, 122, 150 Aoki, M. 112 appreciative theorizing 13, 111, 146 Arabsheibani, G. et al. 61 Ashton, T. 25 Audretsch, D. 4, 39, 42, 86, 93, 114, 131 Audretsch, D. and J. Elston 55 Audretsch, D. and T. Mahmood 90, 93 Austria 41, 46 Autio, E. et al. 128 autocorrelation dynamics 36 Axtell, R. 15, 22, 23

Baily, M. et al. 63, 65 Baily, M. and D. Farrell 65, 73, 108 Barnett, S. and P. Sakellaris 53 Barney, J. 103 Barron, D. et al. 41, 85, 109 Bartelsman, E. et al. 17, 18, 62, 95, 134 Bartelsman, E. and M. Doms 64 Batsch, L. 120 Baumol, W. et al. 9–10, 103 Becchetti, L. and G. Trovato 43 Beck, T. et al. 90, 91, 95–6, 98, 131 Bellone, F. et al. 131 Berger, P. and E. Ofek 123 Bhagat, S. et al. 122 Bigsten, A. and M. Gebreeyesus 42, 85 Birch, D. 11, 129 Birley, S. and P. Westhead 94 Bloom, N. and J. Van Reemen 77 Blundell, R. et al. 52, 53, 60 Boeri, T. and U. Cramer 46 Bond, S. and C. Meghir 53, 55 Bond, S. et al. 53, 55, 60 booms and recessions 28–9, 94 Boone, J. et al. 87 ‘born global’ firms 127, 150 Bottazzi, G. et al. 6, 15, 16, 25, 26, 27, 42, 44, 46–7, 57, 64, 79, 95 Bottazzi, G. and A. Secchi 16–17, 19, 26, 28, 30, 36, 41, 44, 45, 46, 86, 95, 115, 147 bounded rationality 5, 54, 56–7, 61 Brock, W. 146 Broekel, T. (and Coad) 73 Bronars, S. and D. Deere 112 Brouwer, E. et al. 82 Brown, C. and J. Medoff 132 Brynjolfsson, E. and L. Hitt 3 Buldyrev, S. et al. 28 Cabral, L. and J. Mata 17 Calvo, J. 41, 98

191

192

The growth of firms

Camerer, C. and D. Lovallo 61 Campa, J. and S. Kedia 123 Canada 18, 55 Carden, S. 77 Carpenter, R. and B. Petersen 55 Catley, S. and R. Hamilton 89, 90 Caves, R. 43, 46, 131, 144 Caves, R. and M. Porter 86 Cefis, E. et al. 43, 101 Cefis, E. and L. Orsenigo 77 Central Limit Theorem 44 Chandler, A. 3, 4, 122 Chapin, F. 136–7 Chesher, A. 20, 43, 46 Chile 64 Chirinko, R. 51, 52, 53 Churchill, N. and V. Lewis 139, 140 Coase, R. 100 cognitive leadership 139 Colombo, M. and L. Grilli 10 competition 30–31, 37, 93 competitive advantage 4, 5, 103, 124 and employment growth 87 and game theory 86 and growth by acquisition 125 and profits 65 and size of firm 86, 91 and small firms 86, 131–2 conglomerate mergers 104, 124–5 Cooper, R. et al. 28 copy EXACTLY! policy 118–19 Cordes, C. et al. 114, 139 corporate refocusing 120–21 Corsino, M. 79 Côte d’Ivoire 41, 64, 85, 90, 95 credit market imperfections 55 Cressy, R. 62 Cromie, S. 90 Cummins, J. et al. 55 Das, S. 85 Davidsson, P. et al. 124, 129, 135, 147 Davis, S. et al. 11, 94, 129 De Fabritiis, G. et al. 44 de Jong, J. and O. Marsili 94 de Meza, D. and D. Webb 62 de Wit, G. 20, 22 Degryse, H. and A. de Jong 59 Del Monte, A. and E. Papagni 79 Delmar, F. 9, 10, 69, 94, 124, 129

Delmar, F. and J. Wiklund 136, 151 Denmark 18, 26, 85, 95, 97 developing countries 88–9, 92, 114, 131 see also individual countries Dickerson, A. 105, 125 Dierickx, I. and K. Cool 102 Disney, R. et al. 66, 67 diversification 3, 91, 104, 105, 122–3, 124, 126–8 growth strategies 112–13, 114, 117–18, 119–23 and management 105 and risk reduction 121–2 Dixit, A. 86, 112 Dixon, R. 31–2, 117, 149 Dobson, S. and B. Gerrard 69 Doms, M. et al. 28, 64, 82, 147 Dosi, G. 6, 16, 57, 106, 148 Dosi, G. et al. 5, 16, 106 Dosi, G. and M. Grazzi 5 Dosi, G. and D. Lovallo 61 Doukas, J. and O. Kan 123 Downie, J. 106 downsizing 3, 63–4, 107 Droucopoulos, V. 42 Druilhe, C. and E. Garnsey 140 Dunne, P. and A. Hughes 41, 44, 45, 95 Dunne, T. et al. 18, 41, 62, 85, 90, 147 economies of growth 102, 103, 113, 114, 135, 145 economies of scale 2–3, 44, 69, 112, 135, 149 Eisenhardt, K. and J. Martin 102, 103 Eisenhardt, K. and C. Schoonhoven 10 employment levels 9, 10–11, 18, 28 and competition between firms 87 and employee behaviour 116 growth propagation 32–7 and innovation 76, 81–3 and sales growth 58, 70, 71–2 and uncertainty 92 and worker morale 112 entrepreneurial characteristics 88–90 entry rates 86, 108, 131 and competitive advantage 112 entry costs and internal growth 124 excess of new firms 61–2 failure rate 62 and FDI 127

Index and job creation 129 and productivity 64, 65, 66, 67, 68, 134 survival 93, 95, 133, 134 Erickson, T. and T. Whited 53–4 Ericson, R. and A. Pakes 134–5 Ethiopia 42, 85 Euler equation model 52–3, 60 Evangelista, R. and M. Savona 82 Evans, D. 41, 44, 85, 95 evolutionary economics 6–7, 105–8 evolutionary theory of growth 56–9, 60, 61, 63 exit rates 56, 92 and diversification 123 exit hazards 86, 90, 106, 133, 134, 155–6 and productivity 63, 64, 65, 66, 68 small firms 40–41, 43, 129, 133, 142 survival of the fittest 7, 107–8, 144 exporting 4, 91, 126–8 Fagiolo, G. and A. Luzzi 59, 60, 90, 98 family-owned firms 113, 114 Fazzari, S. et al. 51, 53, 54–5, 59, 60 FDI 4, 126–7 Feldman, M. 102 financial constraints capital intensity and growth 90–91 evaluating importance of 59–62 and profits, productivity and firm growth 54–5, 58 and selection effects on profits 50–63 small firms 17, 62, 132, 136 ‘financial pecking-order’ theory 58 Finland 18 Fishman, A. and R. Rob 88 Fizaine, F. 85 Fleck, J. 81 Fluck, Z. and A. Lynch 123 Foster, L. et al. 64, 65–6, 67, 68, 131 France 18, 50–51, 82, 85, 90, 95 growth rate distribution in manufacturing 14, 15, 17, 26–8, 42, 44, 46, 57, 71, 98, 107, 113 Freel, M. 79, 81 Freeland, R. 107 Friedman, M. 11

193

future research discrete growth events, impact of 147 econometric investigations into firm growth 147 firm-specific variables 144 Gibrat’s law 148 growth and ‘fitness’, relationship between 144 inter-firm competition 87 investment, lumpy nature of 38 performance and ambition 144 profits, productivity and firm growth 75 theoretical perspective on firm growth 145–6 VAR models of firm growth 70 Gabe, T. and D. Kraybill 41, 93, 95 Galeotti, M. et al. 53, 55 game theory 86 Garnsey, E. 2, 40, 47, 134, 140 Gartner, W. 149 Gaussian distribution 26, 29 GDP 96 Germany 18, 46, 55, 79, 82, 88, 90, 92, 95 growth rates of manufacturing firms 41, 46, 82, 97 Geroski, P. 77, 86, 96, 108 Geroski, P. et al. 28, 46, 53, 79, 97, 101 Geroski, P. and K. Gugler 25, 37, 43, 85, 87, 90, 91, 94, 97, 144 Geroski, P. and S. Machin 78 Geroski, P. and M. Mazzucato 46 Geroski, P. and S. Toker 78–9, 91, 93, 97 Gibrat’s law 7, 14, 17–20, 29, 39–48, 147–8 autocorrelation of growth rates 45–7 and business cycles 94 econometric issues 43 and economies of scale 44 firm size and average growth 40–43, 53, 85 firm size and growth rate variance 43–5 heterogeneous growth and autocorrelation patterns 47 and independent sub-markets 45

194

The growth of firms

and Kolmogorov–Smirnov tests 47 model 39–40 model of random growth shocks 22–3 and negative dependence of growth on size 42, 43–4 and neoclassical optimizing models 101 objections to 20 and relevant lags 46 and sample selection bias 43 scaling of growth rate variance 44–5 and services sector 42 as stochastic process 115 Gilchrist, S. and C. Himmelberg 55 GMM (Generalized Method of Moments) techniques 70 Goddard, J. et al. 41, 46 Goedhuys, M. and L. Sleuwaegen 41, 64, 80, 85, 90, 92, 95 Goergen, M. and L. Renneboog 59 Gomes, J. 54 government schemes 4, 59, 62, 90 Grabowski, H. et al. 4 Graham, J. et al. 123 Greenhalgh, C. et al. 82 Greiner, L. 84, 115, 138 Griliches, Z. 77 Griliches, Z. and J. Mairesse 5 Griliches, Z. and H. Regev 64 growth advantages of 112–13 and age of firm 84–5 aspirations 2, 88, 89, 92, 136 attitudes to 111–17 between-plant reallocation 66–8 and capital intensity 90–91 and centrality of network 91 and competition 93 and conglomerate merger 104, 124–5 control-loss argument 113 coordination problems 113 copy EXACTLY! policy 118–19 desirability of 112–15 determinants 96–9 disadvantages of 113–15 and diversification see diversification and exports 4, 91, 126–8 and external business advice 92

and ‘growth of the fitter’ 7, 107–8, 144 growth rate distribution 25–38 growth rate distribution, heavy-tailed nature 25, 26, 28, 29, 34, 36, 86–7 industry-specific factors 92–4 inherent tendency towards 115–16 and innovation see innovation intentionality of growth 115–17 and inter-firm partnerships 91 internal 116, 124–5 and investment spikes 37–8 Laplace distribution 25–6, 28, 30, 32 macroeconomic factors 94–6 and managerial resources 102, 112–14, 115, 116–17, 119–20, 121, 125, 138–9 modelling stages of 137–41 and nature of firm activities 91 as ongoing process 70 and ownership structure 90 and population ecology 108–9 R2 values 96–9 reduced form models 59, 60 relative 10, 11, 66 spurts 29–37 strategies 111–28 ‘tent-shaped’ distribution 6, 25, 26, 27–8 time-varying moments 28–9 VAR models of firm growth processes 69–73 very large firms 95 Guariglia, A. 51, 55 Guiso, L. and G. Parigi 92 Hadlock, C. 55 Haleblian, J. and S. Finkelstein 150 Hall, B. 28, 41, 43, 44, 76, 82, 95 Hannan, M. 32, 108, 109 Hannan, M. and J. Freeman 108, 109, 137, 141 Harada, N. 133 Hardwick, P. and M. Adams 42, 91, 94, 108 Harhoff, D. et al. 43, 44, 90, 97 Harrison, R. et al. 82 Hart, P. 14, 41 Hart, P. and N. Oulton 42, 44

Index Hart, P. and R. Pearce 95 Hay, D. and D. Morris 112 Hay, M. and K. Kamshad 77, 91, 105, 114, 132, 136 Hayashi, F. 51–2 Headd, B. 133, 134 Headd, B. and B. Kirchhoff 93, 133–4 hierarchical nature of firm 3, 32–5 high-profit firms and business opportunities, lack of interest in 107 Higson, C. et al. 29, 94 Hines, J. and R. Thaler 58 Hisrich, R. and S. Ozturk 90 Holl, P. 91, 104–5 Hopenhayn, H. 106, 134 Hoshi, T. et al. 55 Hu, X. and F. Schiantarelli 55 Hubbard, R. 51 Huberman, B. and L. Adamic 22, 23 Hughes, A. 62, 132 Hugo, O. and E. Garnsey 140 human capital 54, 88–9 Hyland, D. and J. Diltz 121 Hymer, S. and P. Pashigian 43 Idson, T. and W. Oi 131 Ijiri, Y. and H. Simon 16, 29, 30, 45, 115, 147–8 imperfect markets theory 54–6, 59, 60 India 20, 21, 85, 89, 131 Indonesia 89 industrial classification scheme 94 innovation and economic performance 77–8 and employment growth 76, 81–3 failed attempts, outcome of 80–81 and industrial classification schemes 94 industry-specific factors 93 measurement methods 79–80 profit margins 78 sales growth 76, 77–81 small firms 77, 131 technological unemployment 82–3 and uncertainty 77–8, 80 see also R&D instrumental variables 156 internationalization 126–8, 150

195

investment and cash flow 54–6, 58, 59, 60 plant-level 28 and productivity growth relationship 38 and profit, relationship between 51 R&D 73, 78–9, 82 spikes 37–8 and uncertainty 92 Ireland 79 Italy 18, 42–3, 55, 79, 82, 90, 92, 95, 114 growth rate distribution in manufacturing 17, 26, 43, 44, 46, 57, 64, 73, 79, 90, 98, 107 Japan 41, 46, 55, 85, 87, 95 Jensen, M. 20, 26, 54, 58, 122, 125 Jensen, M. and W. Meckling 103 Johanson, J. and J. Vahlne 127 Johnson, P. et al. 42 Jones, M. and N. Coviello 129 Jovanovic, B. 106, 134 Kaldor, N. 70, 113 Kalecki, M. 20 Kay, N. 101 Kazanjian, R. and R. Drazin 140–41 Kesten, H. 22 Klette, T. and Z. Griliches 77 knowledge, tacit knowledge transfer 118, 126, 127 Kogut, B. and U. Zander 127 Kuemmerle, W. 128 Kumar, M. 41, 45, 95, 97 labour productivity growth 63–4, 71–3 Lamont, O. 55 Lang, L. and R. Stulz 123 Laplace distribution 25–6, 28, 30, 32 learning-by-doing 66, 67, 102 Lee, Y. et al. 44 legal status 90 Leiponen, A. and I. Drejer 94 Lensink, R. et al. 92 Lerner, J. 62 Levenson, A. and K. Willard 62 limited liability companies 85, 90, 95, 150 Little, I. 9–10, 25, 62, 131

196

The growth of firms

Liu, J. et al. 41, 64, 85, 90, 97, 98 Lockett, A. et al. 124, 147 lognormal distribution 14, 16, 20, 101 Lotti, F. and E. Santarelli 17 Lotti, F. et al. 39, 42–3 Lucas, R. 101 Luttmer, E. 22 McCombie, J. 70 MacDonald, C. 118, 119 McDougall, P. et al. 127, 128 McKelvey, B. and P. Andriani 29, 36, 86, 137 McPherson, M. 41, 43, 88, 89, 94, 95, 97, 98 Maksimovic, V. and G. Phillips 64, 93–4, 123, 124, 125, 147 management cognitive leadership 139 control and small firms 105 and diversification 105 entrenchment 121 incentives 104, 121 pursuit of growth 58, 69, 70, 104 resources 3, 54, 91, 102, 112–14, 115, 116–17, 119–20, 121, 125, 138–9 shareholder-wealth-maximizing managers 59–61 talent distribution and small firms 101 theory of the firm 103–5 Mansfield, E. 19, 40, 42, 77–8, 147 manufacturing industries see under individual countries market concentration 93 power 107 selection 61 value 4, 51, 52, 53, 57, 76, 97 Markides, C. 123 Marris, R. 4, 5, 58, 103–5, 119, 125 Marshall, A. 132–3 Marsili, O. 16, 43, 94, 96 Martin, J. and A. Sayrak 122 Matia, K. et al. 44, 45 Matsusaka, J. 123 Mead, D. and C. Liedholm 88–9 mergers and acquisitions see acquisitions Metcalfe, J. 76, 106

Miner, J. et al. 136 Minimum Efficient Scale (MES) 41, 93 Minkoff, D. 109 Mitzenmacher, M. 22 Montgomery, C. 102–3, 121, 122, 123 motivation, lack of, in larger firms 114, 117 Mowery, D. 42, 78 Mueller, D. 104, 119, 125 multiplant firms 68, 90, 96, 97, 98 Myers, S. 58 Nafziger, E. and D. Terrell 131 Nelson, R. 114 Nelson, R. and S. Winter 5, 77, 106, 111, 146 neoclassical economics 7, 8, 61, 62, 63, 69, 145 ‘optimal size’ 56, 60, 100–101, 102, 103, 107, 112, 145 q-theory 51–4, 55, 57, 59, 60, 117 Netherlands 18, 42, 92, 114 networks 91, 126 Niefert, M. 82 O’Farrell, P. and D. Hitchens 136 Oliner, S. and G. Rudebusch 55 ‘optimal size’ theory 56, 60, 100–101, 102, 103, 107, 112, 145 organizational change 3, 5, 11, 112, 114, 141 slack 31, 32, 70 Oviatt, B. and P. McDougall 128 ownership structure 90 Pareto distribution 16, 22–3, 28 Parkinson, C. 115–16 partnership, inter-firm 91 patents 77, 78, 79–80, 82, 97 Paulré, B. 121 Pavcnik, N. 64 Pavitt, K. 131 Penrose, E. 4, 31, 32, 47, 69, 91, 111, 119–20, 131, 135, 145–6, 148 theory of growth of firm 102–3, 113, 116–17 pharmaceutical industry 16–17, 41, 43, 44, 45, 46, 79 Phillips, B. and B. Kirchhoff 22, 42, 133 population ecology 108–9

Index Portugal 17, 18 Powell, W. et al. 91 Power, L. 38 Prais, S. 14, 41 process theory of internationalization 127 production Minimum Efficient Scale and growth rates 41 objectives 112–13 productivity and growth 63–5, 71–3 growth, decomposing 65–8 relative 63–8 profits and competition 65 and diversification 123 and growth, financial constraints and selection effects 50–63 innovation and firm growth 78 and investment, relationship between 4, 5, 10, 51 and productivity 57–8 q-theory 51–4, 55, 57, 59, 60, 117 Quandt, R. 16 R&D 3, 4, 6, 54, 56, 73, 74, 78–80, 82, 120, 123, 126, 131, 135 see also innovation R2 values 96–9 Radice, H. 91 Rajan, R. and J. Wulf 3 Ramsden, J. and G. Kiss-Haypal 16 Rao, R. (and Coad) 37, 57, 73, 74, 76, 79–80, 82, 145, 147, 150 reallocation 6, 49, 65, 66–8, 144 Reed, W. 22 regional effects 95 Reichstein, T. and M. Dahl 85, 95, 97 Reichstein, T. and M. Jensen 20, 26 relative growth 10, 11 replication 117–19 replicator dynamics 106 resources and growth 102–3 management 3, 54, 91, 102, 112–14, 115, 116–17, 119–20, 121, 125, 138–9 sharing 93

197

risk reduction, and diversification 121–2 Rivkin, J. 118 Roberts, J. 112 Robson, P. 79, 81 Robson, P. and R. Bennett 51, 85, 88, 91, 92, 97, 132 Robson, P. and B. Obeng 88, 89 Roll, R. 125 Roper, S. 79 Rossi-Hansberg, E. and M. Wright 16, 42, 90–91 Rumelt, R. 123 sales growth 9, 28, 29, 93 and ‘growth of the fitter’ 7, 107–8, 144 and innovation 76, 77–81 Salop, S. 86 Salter, W. 38 Samuels, J. 41 Santarelli, E. and M. Vivarelli 17, 61, 131 Sapienza, H. et al. 128 Sarno, D. 59, 132 Sastry, M. 141 Schaller, H. 53, 55 Scherer, F. 74, 78 Schiantarelli, F. 51, 53, 55 Schivardi, F. and R. Torrini 92, 114 Schumpeter, J. 3, 87, 105 Secchi, A. 6, 16–17, 19, 26, 28, 30, 36, 41, 44, 45, 46, 86, 95, 115, 147 Segarra, A. et al. 20, 21 Segarra, A. and M. Callejon 41 selection bias 1, 43 and conglomerates 64 and differential growth 65, 106, 107, 108, 144 and financial constraints 50–63 service industry 18, 42, 81, 82, 85, 90, 112, 118, 120, 127 sex, and entrepreneurial characteristics 89–90 Shanmugam, K. and S. Bhaduri 85 Shepherd, D. and J. Wiklund 10, 69, 89, 136 Shleifer, A. and R. Vishny 121, 122, 125, 150–51

198

The growth of firms

Silberman, I. 16 Simon, H. 12, 16, 29, 30, 31, 33, 45, 56–7, 115, 147–8 Simon, H. and C. Bonini 14 Simons, T. and P. Ingram 109 Singh, A. and G. Whittington 41, 45 Singh, S. et al. 89 size distributions 14–24 lognormal model 14, 16, 17 size threshold 114 upper tail shape problems 16, 22 Slater, M. 102 Sleuwaegen, L. and M. Goedhuys 41, 64, 80, 85, 90, 92, 95 small firms advantages of 130–31 ambition, lack of 136 and cognitive leadership 139 and competition 86, 131–2 financial account structure 17, 62, 132, 136 growth desire 135–6 growth rates 41 hazard rates 140 independence of 136 and innovation 77, 131 and internal growth 124 and internationalization 127–8 and large firms, differences between 130–32 and large firms, growth pattern differences 132–7 and management control 105 and management talent distribution 101 and MES (minimum efficient scales) 72 modelling stages of firm growth 137–41 and negative autocorrelation 135 productivity levels 131, 134–5 regional effects 95 routinization of operations 140 and specialization 131 structural change in growing firms 136–7 survival struggle 133–5 and threshold effect 91–2 Smolny, W. 82 Sorensen, J. and T. Stuart 109

Southern Africa 41, 88, 89, 94, 95, 97 Spain 20, 21, 41, 98 specialization 5, 95, 131 Spiezia, V. and M. Vivarelli 81 Srholec, M. and B. Verspagen 94 Stam, E. and E. Garnsey 140 Stanley, M. et al. 16, 25 Starbuck, W. 8, 113, 148, 149 Steindl, J. 14–16 Stiglitz, J. and A. Weiss 54, 90 Storey, D. 10, 90 structure–conduct–performance paradigm 92 Subbotin distribution 26 survival rates 20, 93, 95, 133–5 Sutton, J. 29, 37, 42, 45, 87, 115 Sweden, growth by acquisition 124 Szulanski, G. and S. Winter 118 Taiwan 41, 64, 85, 90, 97 Tamvada, J. 20, 21, 89 taxation 55, 92, 113, 114 technological progress 1, 3–4, 20, 92–3 Teece, D. 103 Teruel-Carrizosa, M. 42 Tether, B. 114 theoretical perspectives 100–110 threshold effect 91–2, 114 Tobin’s q 51–4, 55, 57, 59, 60, 117 transaction costs theory 3, 100–101, 126 Tsoukas, H. and R. Chia 141 Tybout, J. 92, 114 UK advertising and sales growth 91 cash flow and investment 55 employment growth and innovation 82 excessive start-ups 61, 62 financial structure of small firms 132 firm growth and age of firms 85 firm growth and business cycles 94 firm growth and human capital 88 firm growth of large firms 95 firm growth and management characteristics 91 growth rates of life insurance companies 42, 108

Index growth rates of manufacturing firms 41, 42, 45, 46, 67, 79, 82, 97, 107, 132 high-tech firms and lack of desire for growth 114–15 innovation and small firms 131 profit rates and management control 104–5 R&D investment 78–9, 82 small firms 18, 51, 79, 85, 88, 91, 97, 114–15, 131, 132 software firms, limits to growth 114 time-varying moments of growth rate 29 uncertainty 5, 7, 92, 127 unionization 90, 112, 132 US cash flow and investment 55 employment growth and innovation 82 firm growth and age of firms 85 firm growth and business cycles 94 firm growth and dispersion and volatility factors 94–5 firm growth of large firms 95 firm growth and ownership structure 90 firm growth and plant size 93 growth by acquisition 125 growth rates of manufacturing firms 14, 15, 16, 40, 41, 44, 45, 46, 54, 57, 67, 78, 82, 85, 107 growth rates of retail firms 67, 68 innovation and small firms 131 investment and productivity growth relationship 38 New York Credit Unions 41, 85, 109 oil company investment 55 plant-level investment and growth rate distribution 28 post-entry growth rates 95 R&D investment 73, 78, 82

199 size-wage relationship 132 small firms 18, 62, 90, 97, 129, 131, 133–4 survival rates of small firms 133–4 time-varying moments of growth rate 28, 29

van Dijk, M. and O. Nomaler 108 Van Reenen, J. 77, 82 VAR (vector autoregression) models of firm growth processes 69–73 Variyam, J. and D. Kraybill 42, 85, 90, 97 Villalonga, B. 123 Viner, J. 100 Vining, D. 16 wage levels 3, 132 Wagner, J. 41, 46 Weick, K. 70 Weick, K. and R. Quinn 32, 141 Weiss, C. 10, 41, 46 Wernerfelt, B. 102 Whetten, D. 11, 112, 113, 138, 141 White, H. 43 Whited, T. 37, 53–4, 55, 147 Wiklund, J. 10, 69, 86, 88, 89, 117, 134, 136 will-o’-the-wisp models 113, 149 Williamson, O. 33, 103, 113 Winter, S. 5, 6, 77, 103, 106, 111, 118, 146 Witt, U. 114, 139 Wynarczyk, P. and R. Watson 91 Yasuda, T. 41, 85 You, J.-I. 43, 101, 130 ‘Yule’ distributions 14 Zahra, S. et al. 128 Zipf distribution 22, 23 Zollo, M. and H. Singh 150