Strategic Competition, Dynamics, and the Role of the State: A New Perspective

Strategic Competition, Dynamics, and the Role of the State

NEW DIRECTIONS IN MODERN ECONOMICS Series Editor: Malcolm C. Sawyer, Professor of Economics, University of Leeds, UK New Directions in Modern Economics presents a challenge to orthodox economic thinking. It focuses on new ideas emanating from radical traditions including post-Keynesian, Kaleckian, neo-Ricardian and Marxian. The books in the series do not adhere rigidly to any single school of thought but attempt to present a positive alternative to the conventional wisdom. For a full list of Edward Elgar published titles, including the titles in this series, visit our website at www.e-elgar.com.

Strategic Competition, Dynamics, and the Role of the State A New Perspective

Jamee K. Moudud Economics Faculty, Sarah Lawrence College, New York, USA

Foreword by Anwar M. Shaikh

NEW DIRECTIONS IN MODERN ECONOMICS

Edward Elgar Cheltenham, UK • Northampton, MA, USA

© Jamee K. Moudud 2010 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: 2009940633

ISBN 978 1 84542 923 2

02

Printed and bound by MPG Books Group, UK

To Shanaz, Laila, and Aliya and My Parents With all my love

Contents Foreword Anwar M. Shaikh Acknowledgments

x xii

1

Introduction

1

2

The microfoundations of long-run growth: controversies on capacity utilization and competition Introduction Ex ante versus ex post idle capacity Rival theories of competition and their implications for capacity utilization The persistence of excess capacity: Sraffian and Kaleckian approaches Conclusion Appendix

10 10 10 15 32 42 51

3

A review of the literature on growth Introduction Neoclassical growth models Heterodox growth models Conclusion

4

A model of disequilibrium dynamics Introduction Disequilibrium dynamics in an SFC context Conclusion Appendix 1 Appendix 2

77 77 77 98 100 106

5

Warranted growth and the role of the State Introduction Harrod’s policy insights: solutions to ambiguities and contradictions Conclusion

113 113

vii

53 53 53 63 72

116 122

viii

Strategic competition, dynamics, and the role of the State

Appendix 1 Appendix 2 6

Conclusion: the relevance of microfoundations and politics

References Index

126 130 132 142 161

Economic privation proceeds by easy stages, and so long as men suffer it patiently the outside world cares little. Physical efficiency and resistance to disease slowly diminish, but life proceeds somehow, until the limit of human endurance is reached at last and counsels of despair and madness stir the sufferers from the lethargy which precedes the crisis. Then man shakes himself, and the bonds of custom are loosed. The power of ideas is sovereign, and he listens to whatever instruction of hope, illusion, or revenge is carried to him on the air. . .But who can say how much is endurable, or in what direction men will seek at last to escape from their misfortunes? – Keynes, J.M. (1920),The Economic Consequences of the Peace, Harcourt, Brace, and Howe, pp. 250–51.

Foreword Neoclassical economics has long been divorced from the reality it purports to illuminate. But now that reality, in the form of characteristically turbulent capitalist dynamics, has asserted itself through a worldwide economic crisis. The crisis has damaged the livelihoods of hundreds of millions of people, some of whom were already living on a margin of survival. It has threatened political arrangements and overturned policy prescriptions in every country. And, in passing, it has brought the economics profession to a crossroad. Should it continue to put its reliance on the dominant paradigm, which spends almost all of its time extolling the virtues of a mythical perfect capitalism with only occasional asides on the implications of the stubborn ‘imperfections’ of the real world? Or should it reject the false dichotomy between perfect theory and imperfect reality, and work instead towards theory which is grounded in the real? This book takes the second path. The present crisis in the economics profession is not unprecedented. Much the same thing occurred in the 1930s. Then, as now, the economic orthodoxy preferred the safety of its blinkered precepts to the unpleasant reality of crumbling markets and mounting unemployment. Then, as now, government after government abandoned these theorists, who after all had little to say on such mundane matters, moving instead to a variety of practical measures based on deficit-financed public spending in which armaments eventually came to play a devastating role. But, at that time, Keynes’s General Theory had already been written, so that its prescriptions provided a most welcome theoretical foundation for the urgent needs of the time. The problem is that the Keynesian theoretical structure proved to have its own deficiencies. Its theoretical foundations were inadequately specified at the time of Keynes’s untimely death in 1946, and its subsequent theoretical expectations were confounded by the confluence of inflation and unemployment in the 1970s. Despite Keynes’s own insistence that his ideas applied to a competitive economy characterized by the ‘higgling’ of markets, Post Keynesian economics attached itself to the notion of imperfect competition and short-run equilibrium analysis. The recourse to imperfect competition tied the theory to the notion of perfect competition, which is its point of departure. The reliance on equilibrium analysis, x

Foreword

xi

on equilibrium as an attained-and-held state, removed disequilibrium dynamics from view. And the focus on the short run led to a persistent inability to treat slower dynamics. It is important to understand that these are not mere theoretical caveats. Post Keynesian economic policy rests on these very same foundations, and foundational deficiencies always have practical consequences. This book seeks to build a more robust foundation for heterodox economic theory and policy. The microfoundations of macroeconomic analysis are resituated in the theory of strategic competition developed by P.W.S. Andrews, Sir Roy Harrod and others of the Oxford Economists’ Research Group. This provides a clear and precise alternative to the perfect–imperfect competition dichotomy. The book develops an ex ante stock-flow framework which does not rely on equilibrium assumptions, leading to an explicit macroeconomic relation between excess demand in the commodity market and the excess supply of money. Disequilibrium dynamics is formally treated, so that equilibrating processes are seen to be intrinsically turbulent. This provides the grounds for the distinction between faster processes such as the perpetual chase between supply and demand (Keynes’s domain) and the slower one between supply and capacity (Harrod’s domain). The book goes on to demonstrate that Harrod’s treatment of growth provides a powerful foundation for the analysis of capitalist dynamics. Harrod himself was greatly concerned with economic policy in a dynamic context, and this book provides the valuable function of documenting his arguments and drawing out their implications for current concerns. In so doing it also restores Harrodian economics to its proper place in an expanded heterodox canon. The concerns of this book are extremely timely. In the face of the worldwide economic crisis, fiscal policy is now the rage once again. But neoclassical and Post Keynesian economists strenuously disagree on its long-term consequences. This book points to better ways to address these concerns. Anwar M. Shaikh

Acknowledgments As with any private production this book has benefited from many social inputs. At a professional level I would above all like to thank Professor Anwar Shaikh of the New School for Social Research and Professor Porus Olpadwala at Cornell University. It was the latter’s encouragement that made me pursue graduate studies in political economy at the New School after I recognized the limitations of an engineering education for the intellectual and political questions I had developed. On the other hand I owe my greatest intellectual debt to Anwar, one of the chief architects of the New School’s distinctive approach to the study of political economy, whose contributions have inspired generations of graduate students like me from all parts of the world. First as my professor, then as my thesis supervisor, and finally as a friend, Anwar has been unstinting in his generosity in helping me sort out my own thoughts on economic theory and policy. I would like to thank him for graciously agreeing to write the Foreword to this book. I would also like to thank my old friends Karl Botchway and John Sarich for providing me with important theoretical insights. A special thanks goes to Karl, who as a political scientist never ceased to remind me very convincingly about the role of political factors in shaping economic and social policies. Moreover I would like to thank my friend Michael Mavaddat, who very kindly sent me several articles and engaged in fascinating discussions about the real world of business competition. I would like to show my appreciation and thanks to Dr Cyrus Bina, Dr Mario Seccareccia, Dr Jesus Felipe, Dr Mark Setterfield and Dr Gary Mongiovi for taking the effort to read through the manuscript and write brief descriptions of the book for promotional material and the back cover. Thanks also go to Alan Sturmer and Caroline Phillips, Senior Acquisitions Editor and Managing Editor, respectively, at Edward Elgar for their guidance and support throughout the final stages of this book project. I would also like to thank Helen Moss, the copy editor, for her excellent work. It would be remiss of me not to acknowledge my many distinguished colleagues and outstanding students who constitute the vibrant intellectual atmosphere at Sarah Lawrence College. Several of my key insights on the role of the State developed during my classes and my many xii

Acknowledgments

xiii

interactions with students, too numerous to mention individually by name here. I would like to thank my colleague in economics and friend Marilyn Power for reading and commenting on Chapter 2. At a personal level I would like to begin by thanking my parents, both of whom instilled in me from a young age the love of learning, as well as my brother Roomee and sister Roohee for being encouraging and supportive throughout my life. I would like in particular to thank my mother, who encouraged me to pursue graduate studies at the New School, recognizing how vital that was to me at a personal level. A special acknowledgment of thanks goes to my gorgeous wife and best friend Shanaz, whose infectious enthusiasm and constant encouragement during difficult times kept me soldiering on. She is a busy professional herself, and I want to give her special thanks for all these years of companionship, which have been the best ones of my life, the wonderful home she has made for us, and the space she gave me (including the office she built for me!) to complete this book in the midst of our busy schedules. Our pride and joy, Laila Sophia and Aliya Mehereen, allowed me to master the art of multi-tasking, in which my scholarly pursuits somehow had to be combined with the joys and challenges of raising them and managing our four cats (Magic, Sparkles, Mischief and Troublina). It is my hope that Laila and Aliya strive for and inherit a better world.

1.

Introduction

The dominance of Keynesian-type and State-led development policies in much of the capitalist world during the 1950s and 1960s came under attack by the New Right in the 1970s, an onslaught that eventually culminated in the elections of Ronald Reagan and Margaret Thatcher. The free market doctrine that became the new conventional wisdom had as its intellectual foundations the theoretical views of Friedrich Hayek and Milton Friedman. It was the promise of laissez faire policies at home and abroad that persistent unemployment and stagnant growth would be eliminated once the State had been rolled back and free market policies instituted. Especially after the Berlin Wall came down and the former Soviet Union collapsed it became almost an act of political and professional suicide for both parties of the Left and non-neoclassical economists to question such policies. Henceforth, Fukuyama’s ‘end of history’ thesis pointed to a rosy future in which there would be no room for dissenting views. However, the recent triumphs of left-leaning political movements across much of Latin America and the disaster unleashed by Hurricane Katrina, which revealed the deep class fissures in American society and stark evidence of decades of neglect of its public infrastructure, show that the record of neo-liberalism has been less than stellar. Most importantly, the current depression-like world economic crisis should be embarrassing to free marketers, whose triumphalism had reached a crescendo in the 1990s. Quite simply, if the New Right analysis of markets were correct, the global march toward free markets over the last 30 years should not have culminated in the current crisis. Empirically, as Cypher and Dietz (2004, pp. 509–519) summarize the literature, free market policies implemented by the International Monetary Fund (IMF) and the World Bank have not been successful in either stimulating investment, savings and growth rates or reducing inequality and poverty. Furthermore, as Milanovic (2003) shows in a comparison of the pre- and post-1980 periods, the growth rates of 96 out of a sample of 125 countries were lower in the latter period when neo-liberalism became the conventional wisdom. These ‘stylized facts’ about the effects of global laissez faire have generated responses from several different perspectives. Perhaps the most famous critique of neo-liberalism and the IMF from

1

2

Strategic competition, dynamics, and the role of the State

eminently establishment circles has been that of the New Keynesian economist Joseph Stiglitz, the 2001 economics Nobel Prize winner, chairman of Clinton’s Council of Economic Advisers, and former senior vice president and chief economist at the World Bank. While his book Globalization and Its Discontents is rich in its empirical and institutional critique of the ‘Washington Consensus’, one of its more significant parts relates to a theoretical observation regarding the treatment of unemployment by neoclassical economists: Unfortunately, too often the training of macroeconomists does not prepare them well for the problems that they have to confront in developing countries. In some of the universities from which the IMF hires regularly, the core curricula involve models in which there is never any unemployment. After all, in the standard competitive model – the model that underlies the IMF’s market fundamentalism – demand always equals supply. If the demand for labor equals supply, there is never any involuntary unemployment. Someone who is not working has evidently chosen not to work. In this interpretation, unemployment in the Great Depression, when one out of four people was out of work, would be the result of a sudden increase in the desire for more leisure. It might be of some interest to psychologists why there was this sudden change in the desire for leisure, or why those who were supposed to be enjoying this leisure seemed so unhappy. . . . While these models might provide some amusement within academia, they seemed particularly ill suited to understanding the problems of a country like South Africa, which has been plagued with unemployment rates in excess of 25 percent since apartheid was dismantled. . . . The IMF economists could not, of course, ignore the existence of unemployment. Because under market fundamentalism – in which, by assumption, markets work perfectly and demand must equal supply for labor as for every other good or factor – there cannot be unemployment, the problem cannot lie with markets. It must lie elsewhere – with greedy unions and politicians interfering with the workings of free markets, by demanding – and getting – excessively high wages. There is an obvious policy implication – if there is unemployment, wages should be reduced. (Stiglitz, 2002, pp. 34–5)

Significantly, though, Stiglitz does not have any fundamental theoretical quarrel with neoclassical economics. The main assumption of neoclassical dissenters such as Stiglitz is that imperfections make markets behave less than smoothly. Under these circumstances State intervention, including expansionary fiscal policies, might be necessary to reduce unemployment and create an adequate social safety net. Thus, as the work of Stiglitz and other New Keynesians makes clear, one should not conclude that neoclassical economists in general necessarily prescribe a reduced role for State involvement. A new generation of neoclassical endogenous growth theory (NEGT) has become fashionable since the 1980s, in which there is some room for public policies.

Introduction

3

Neoclassical economists of all stripes rationalize interventionist public policies of various types on the basis of NEGT models. However, NEGT models are based on a framework which assumes continuous market-clearing, Say’s law, rational expectations and intertemporal optimizing behavior (see for example Greiner, Semmler and Gong, 2004). They are anchored in the production function methodology. Money and finance are not integrated into these models in any coherent way, a not surprising feature given that neoclassical theory assumes that money is neutral. Furthermore, they rest on what this book calls the false dichotomy of neoclassical theory, which is that markets are either perfect or imperfect. Finally, endogenously generated market disequilibria and effective demand considerations are alien to these models. As Solow (1994, pp. 49–51) argued, NEGT models preserve stability under the extremely stringent condition that constant returns to capital prevail; otherwise such models rest between the Scylla of increasing returns and explosive growth and the Charybdis of decreasing returns and stagnation. I will limit myself here by stating that core aspects of these models (such as the false dichotomy, Say’s law, production functions, rational expectations and general equilibrium) have been criticized by a long line of non-neoclassical authors.1 Given that these models are anchored in a general equilibrium framework one may also question their practical relevance in a world of recurrent mass unemployment, economic crises and endogenous financial instability.2 Possibly, the most influential critique of neoclassical models has come from authors in the Post Keynesian and structuralist traditions.3 Having Keynes and Kalecki as their intellectual inspiration, authors in this strand of heterodox economics4 have created a distinctive intellectual tradition within economics. The Principle of Effective Demand, as opposed to Say’s law, is the basis of these investment-constrained models (Foley and Michl, 1999) in which exogenous demand injections have unambiguously beneficial effects on output and employment because of the persistence of excess capacity. Therefore, these heterodox models have policy implications that are quite distinct from neoclassical ones. Perhaps the most fruitful contribution of the heterodox literature is in its modeling of real-financial interactions, much of it inspired by Minsky and, in recent years, by the framework developed by Richard Stone and Wynne Godley at Cambridge University. In contrast to neoclassical models, which assume that money enters the economy at the discretion of the central bank, those of Godley and Lavoie (2007), Taylor (2004) and others explicitly relate monetary stocks and flows to each other in a systematic and comprehensive stock-flow consistent (SFC) framework. The key aspect of the SFC framework is that there are no ‘black holes’

4

Strategic competition, dynamics, and the role of the State

so that the sources and uses of funds of all sectors (households, firms, banks and the government) are related explicitly to each other. Money is largely endogenously created5 by the banking system (Wray, 1990; Itoh and Lapavitsas, 1999; Taylor, 2004) to finance expenditures. In this crucial respect, the current generation of heterodox models deals with an important shortcoming of the standard textbook IS-LM model with its roots in neoclassical theory. At the core of most contemporary heterodox growth models is the assumption that excess capacity is persistent, even in the long run. Following Kalecki and Steindl, this outcome is derived from their microfoundations, in which it is assumed that the market structure is imperfectly competitive, with barriers to entry providing the basis for monopoly power. However, the Kaleckian wing of the Keynesian tradition can be distinguished from what one may call its Harrodian wing, as it was Harrod who argued that excess capacity is eliminated over the long run. Thus if the analytical framework of the Keynes/Kalecki tradition is with regard to adjustments between supply and demand, thereby establishing the multiplier, Harrod’s concern was primarily with the adjustment between supply and capacity. Drawing on the insights of P.W.S. Andrews in particular, Harrod (1952) concluded that excess capacity cannot be persistent under highly competitive conditions when firms actively minimize prices and costs in order to attempt to beat their rivals in an economic environment enshrouded in Keynesian uncertainty. This is the policy of normal-cost pricing, which corresponds to a range of output at which each firm attempts to fully utilize unexhausted economies of scale. It is important to emphasize that normalcost pricing is consistent with some deliberately held reserve capacity, both because of the need to absorb unexpected demand fluctuations and because it is not cost-effective to expand production beyond this range. The distinction between excess and reserve capacity made by authors such as P.W.S. Andrews, Elizabeth Brunner and others of the Oxford Economists’ Research Group (OERG) is consistent with an equally significant contribution by Winston (1974), who distinguishes between undesired (or redundant) idle capacity and planned (or reserve) idle capacity, chosen on the basis of cost-minimization criteria. Undesired idle capacity arises from a drop in effective demand and thus the remedy is Keynesian demand-stimulation policies. On the other hand, reserve capacity entails a certain degree of slack, which is deliberate. The idle capacity of this type allows for some degree of demand stimulation but the latter is necessarily limited because of cost constraints. The practical implication of this distinction between excess and reserve capacity is the treatment of investment. In Harrod, investment is

Introduction

5

triggered as a reaction to the gap between the actual (u) and planned or normal (u*) rates of capacity utilization, where u* is determined by costminimization criteria. This investment function produces an expansion in investment when u . u* (overutilization of capacity) and a reduction in investment when u , u* (underutilization of capacity or excess capacity). This dynamic is the basis of Harrod’s famous instability problem, with the disequilibrium situations (when u ≠ u*) corresponding to the shorterrun cycles along the long-run trend of output, which is approximately equal to capacity (when u ≈ u*). Put simply, in the Harrodian view the pressure of competition makes business investment vary endogenously to ensure that firms adjust capacity to bring it in line with output (demand), thereby producing the economy’s long-run or warranted growth path. Since the key variable is business investment, there is no guarantee that the warranted growth rate will also be consistent with the full employment of labor. On the other hand, in the Kaleckian literature (Lavoie, 1996a) the rate of capacity utilization can take on any value in response to exogenous demand injections since there is no pressure on the part of firms to minimize costs, given the persistence of entry barriers and thus monopoly power. In one version of the Kaleckian investment theory the capital accumulation rate is a function of the level of capacity utilization and animal spirits: investment and output respond passively to demand. However, in other versions Kaleckian authors make the capital accumulation rate a function of animal spirits and the gap between an actual and a redefined notion of normal capacity utilization, where the last mentioned variable is entirely endogenous and thus not regulated by cost-minimization criteria. Thus the long-run growth rate is consistent with arbitrary levels of capacity utilization precisely because there is no competitive pressure. Generally the bulk of the heterodox literature is silent on these distinctions between excess and reserve capacity utilization. In fact the Kaleckian literature usually tends to conflate these two types of idle capacity. However, it is difficult to reconcile the persistent excess (redundant) capacity argument with real-world firms that downsize and close unprofitable plants during economic crises. These issues regarding the microeconomics of firm behavior have major macroeconomic and policy implications. Put simply, traditional Keynesian demand management strategies will work only as long as there is undesired idle capacity (Winston, 1974, p. 1311; Garegnani, 1983). On the other hand, as Harrod (1973) shows, an expansion of government spending relative to the output trend produces a short-run stimulus while reducing the long-run growth rate of output. Given that authors in the classical tradition had a similar view of the long run (Eltis, 1998) it is not

6

Strategic competition, dynamics, and the role of the State

surprising that growth models in the classical tradition will yield the same results (Duménil and Lévy, 1999). To sum up, the current literature is caught between two dichotomies in which either Say’s law holds or permanent excess capacity prevails. While Post Keynesian growth models rest on monopolistic or oligopolistic competition, neoclassical models are rooted in both perfect and imperfect competition. Further, Post Keynesian models are concerned with the problem of long-term involuntary unemployment whereas the neoclassical ones do not even pretend to be concerned with this type of unemployment, as Stiglitz observed. The problem with the current literature is that it does not include a third possibility – the Harrodian Weltanschauung in which investment responds to demand but output and capacity attain rough equality while maintaining involuntary unemployment. Such an outcome, the consequence of a radically different theory of competition, raises an important policy question: what should the role of the State be to raise economic growth and lower the unemployment rate under these circumstances? This was a central concern of Harrod’s, as can be seen from several of his writings, especially Chapter 7 of his Economic Dynamics (1973). Unfortunately, these concerns of Harrod’s were swept aside by authors in the Keynesian tradition who came after him because of the instability of his growth model. Until the current crisis exploded, the Keynesian paradigm was itself marginalized by the rational expectations perspective, which does not even allow for long-run involuntary unemployment. The purpose of this book is to fill in this lacuna, which it does by making several new contributions to the literature on macrodynamics. First, by drawing on the insights of several prominent OERG authors such as P.W.S. Andrews and Sir Roy Harrod, this book provides a new microeconomic rationale for firms’ price- and cost-minimization strategies and thus for the existence of the warranted growth path. In the theory of strategic competition developed in this book, the normal-cost pricing strategy is pursued by firms of all sizes because they are under the threat of actual and potential low-cost rivals from within and outside their industries. OERG authors such as Andrews and Harrod argued that, because a firm’s capital stock lasts a relatively long time, entrepreneurs necessarily take a long-run view in their price-setting policies in an attempt to remain profitable, given potential competitive threats. Since the future is fundamentally unknown, as emphasized by Keynes, price- and cost-minimization are part of every firm’s defensive and offensive strategy, even though no final outcome is guaranteed, for example large-sized firms shielded behind entry barriers may face the ignominy of losing their market shares to smaller-sized new entrants.

Introduction

7

Thus in the OERG view firms set competitive mark-ups (as Andrews emphasized) and not monopoly mark-ups over their production costs. As in oligopolistic competition, pricing is based on strategic behavior. However, unlike game theory-based models of oligopolistic competition which are rooted in rational expectations, in the strategic competition view strategic behavior by firms takes place under Keynesian uncertainty. Unlike all models of oligopolistic competition, in the strategic competition view uncertainty necessarily drives firms to eliminate excess capacity in order to minimize selling prices and costs. My new interpretation of the OERG is that the latter’s arguments regarding the porosity of entry barriers can be generalized to be a critique of both monopolistic and oligopolistic competition, whereas the OERG’s critique was primarily aimed at monopolistic competition. In other words, the position taken in this book differs from Andrews’s position that the modern industrial economy is characterized by the presence of competitive oligopolies. In fact, I argue, the powerful Andrews/Brunner critique of entry barriers is a slippery slope that makes it impossible to sustain the notion of oligopolistic competition. My interpretation of the OERG analysis is further bolstered by actual case studies in business history and experience. I conclude that the OERG analysis of competition parallels the classical view of competition as articulated by a number of authors (McNulty, 1967; Clifton, 1977, 1983; Shaikh, 1982, 2008; Semmler, 1984; Bina, 1985, 1989a, 1989b, 2006; Botwinick, 1993). In other words, the long-run growth rate in both traditions is characterized by the attainment of a normal rate of capacity utilization driven by the cost-minimizing investment decisions of firms. Second, on the basis of what may be called the extended Harrodian cyclical growth model derived by Shaikh (1989), the book also deals with the nature of fiscal policy to raise the warranted growth rate, since in Harrod’s framework there is no automatic convergence between the warranted and natural (full employment) growth rates. In this regard too the extended Harrodian model draws on and extends Harrod’s policy analysis by focusing on the relationship between capital budgeting, public investment, taxation policy and the warranted growth rate. Harrod himself had not clarified the nature of these components of fiscal policy. This part of the book directly addresses the role of the State in raising the long-run growth path and strengthening the social safety net via suitable taxation policies. It provides a new alternative to neo-liberal policies. Third, the extended Harrodian model is rooted in an SFC framework. Because of the pervasiveness of uncertainty in the sense of Keynes and Knight I distinguish between the ex ante plans and expectations of

8

Strategic competition, dynamics, and the role of the State

different sectors and ex post outcomes. The ex ante–ex post distinction is vital in a world in which uncertainty rules, spending plans by households and firms take place on the basis of expectations, and production takes time. In such a world, there is likely to be a mismatch between the output which a firm plans to produce on the basis of expected sales and the actual demand it confronts. Thus macroeconomic disequilibria between demand and supply as well as output and capacity arise quite naturally from this framework. In other words, unlike most other heterodox models,6 shortrun equilibrium between aggregate demand and supply is not assumed. By the same token, the long-run approximate equalization of output and capacity should not be construed as some putative steady state. An important implication of the ex ante versus ex post distinction is that macroeconomic disequilibria arise quite naturally from the way in which the model is set up. The resulting non-linearities and endogenous instability thus capture the essence of Harrod’s perspective on cyclical growth (Besomi, 2001). Economic cycles are treated as stable disequilibrium adjustment processes. Two distinct types of cycles, both empirically grounded, are modeled. The first is a short-run cycle which is a reflection of imbalances between aggregate demand and supply, while a longer cycle arises from discrepancies between output and capacity. Such cycles are absent in the Lavoie–Godley and Taylor models. Indeed, most heterodox growth models assume short-run equilibrium between demand and supply but long-run disequilibrium between output and capacity. The SFC framework that is at the core of the extended Harrodian model differs from what is found in the existing literature (Taylor, 2004; Godley and Lavoie, 2007) in another important way. Because I specify the SFC system in ex ante terms I show how it can produce disequilibria between money supply and money demand. By contrast, most of the heterodox literature is cast entirely in ex post terms and thus does not model monetary disequilibria. Chapter 2 discusses rival theories of capacity utilization and competition. The core of this chapter is dedicated to discussing the OERG’s analysis of firm behavior and how the implications of this analysis differ from those of the neoclassical, Sraffian and Kaleckian literatures.7 This chapter establishes the microeconomic rationale for the Harrodian warranted growth rate. In Chapter 3 I review the current literature on growth. The Solow and NEGT models as well as growth models in the Post Keynesian tradition will be reviewed. The chapter will also discuss Harrod’s own growth framework, as well as Shaikh’s relatively recent solution to the knife-edge problem. Shaikh’s model is elaborated in Appendix 2 of Chapter 4. Chapter 4 derives the extended Harrodian model on the basis of an ex ante social accounting matrix (SAM). An important outcome

Introduction

9

of this chapter is the derivation of the economy’s aggregate or social savings rate on the basis of the SAM. The social savings rate is one of the key determinants of the warranted growth rate. Chapter 5 exploits the properties of the social savings rate to revisit Harrod’s taxation-cumpublic-investment policies. Finally Chapter 6 relates the analysis carried out in Chapters 4 and 5 to the broader literature on public policies and the capitalist developmental state. It ends on the cautionary note that actual development policies cannot and should not be mechanically related to a mathematical model given the complex ways in which the dynamics of class and power relations shape the development process.

NOTES 1. One may of course begin with the historical giants of the profession such as Marx or Keynes who have criticized Say’s law and the quantity theory of money while proposing very different theories of accumulation, the role of effective demand, and expectations. But in the modern era there have been devastating criticisms of the neoclassical theory of money (Wray, 1990; Itoh and Lapavitsas, 1999), production functions (Shaikh, 1974, 1980; McCombie and Thirlwall, 1994; Felipe and Fisher, 2003; Felipe and McCombie, 2005a, 2005b, 2006, 2007) and neoclassical macroeconomics. Perhaps the most recent critique and alternative is Taylor’s Reconstructing Macroeconomics (2004). See also various articles in Davidson and Kregel (1994), as well as Cesaratto (1999), for critiques of the NEGT framework. 2. Wolfson’s (1994) book is an excellent comparison of heterodox (Marx, Keynes, Minsky, etc.) and neoclassical analyses of finance and business cycles. 3. In what follows I will refer to these as the Keynes/Kalecki tradition. 4. The term heterodox economics embraces a wide range of non-neoclassical perspectives rooted in Marx, Keynes, Harrod, Kalecki, Sraffa and others. 5. This basically means that the central bank does not exogenously control the money supply. The supply of credit, which determines the creation of money, is largely demanddetermined. 6. The exception is the Shaikh (1989) model. 7. These are two major schools of thought in heterodox economics. Sraffians, drawing on the insights of the classical/Marxian tradition and of Sraffa, take the position that market prices are regulated by more fundamental prices, which Smith called natural prices and Ricardo and Marx called prices of production. The prices of production in each sector correspond to an industry-wide general rate of profit, which itself is established via the inflows and outflows of capital across the different sectors in search of the highest rate of profit. By basing themselves primarily on the Polish economist Michal Kalecki, Kaleckians emphasize the importance of mark-up pricing under conditions of imperfect markets. As discussed in Chapter 2, authors in both schools take the position that excess capacity is persistent.

2.

1.

The microfoundations of long-run growth: controversies on capacity utilization and competition INTRODUCTION

In the analysis of economic growth a central controversy deals with the nature of the economy’s long-run growth path and the characteristics of the forces that either make output and capacity converge or keep them persistently apart. In this chapter it will be shown how the analysis of longrun capacity utilization is shaped by theoretical controversies regarding the nature of competition. For example, in neoclassical theory perfectly competitive markets, characterized by passive price-taking firms, ensure the attainment of minimum average total costs and the elimination of excess capacity. On the other hand, in neoclassical theory price-setting behavior is considered to be a defining feature of what it calls imperfect competition, in which firms maintain persistent excess capacity. This second view of the competitive process underpins both Kaleckian growth models and neo-Schumpeterian neoclassical endogenous growth models (P. Romer, 1994). This chapter will discuss how the nexus between competition and capacity utilization is analyzed in four broad traditions: the Oxford Economists’ Research Group (OERG), classical/Marxian, Kaleckian and neoclassical. It will be demonstrated that there are significant areas of overlap between the analysis of competition in the classical/Marxian tradition and the contributions of Sir Roy Harrod, P.W.S. Andrews and Elizabeth Brunner, and others of the OERG tradition. This approach to competition, which we will call the theory of strategic competition, will be shown to be quite different from models of perfect and imperfect competition.

2.

EX ANTE VERSUS EX POST IDLE CAPACITY

The conventional view is that if perfect competition is consistent with full capacity utilization (and full employment of labor) the existence of 10

The microfoundations of long-run growth

11

considerable degrees of idle capacity implies some kind of market imperfection, a position taken by Kalecki (Sawyer, 1985, p. 28). Among heterodox authors there is widespread consensus that idle capacity (as well as varying degrees of unemployment) constitutes the normal state of affairs in the capitalist economy. As Kurz (1992) points out, the capitalist industrial system maintains a considerable degree of flexibility, thereby allowing the rate of capacity utilization to vary over wide ranges with the level of effective demand. The question of elasticity was perhaps best summarized by Marx, who observed: So soon . . . as the factory system has gained a certain breadth of footing and definite degree of maturity, and, especially, so soon as its technical basis, machinery, is itself produced by machinery . . . this mode of production acquires an elasticity, a capacity for sudden extension by leaps and bounds that finds no hindrance except in the supply of raw material and in the disposal of the produce. (Marx, 1954, p. 424, cited from Kurz, 1992, pp. 75–76, emphasis added)

Marx further added with regard to fluctuations in the capacity utilization rate: When there is a hitch in production, when the markets are overstocked . . . the normal outlay of circulating capital is restricted – once the pattern of fixed capital is set – by cutting down working time to, say, one half. On the other hand, in times of prosperity, the pattern of fixed capital given, there is an abnormal expansion of the circulating capital, partly through the extension of working time and partly through its intensification. (Marx, 1956, p. 262, cited from Kurz, 1992, p. 76, emphasis added)

The existence of considerable idle industrial capacity is an empirical fact in both the developed and the developing world (Foss, 1963; Marris, 1964; Winston, 1974; Lim, 1976). However, the question is why? It is important not to confuse two broad reasons for the existence of idle capacity (Winston, 1974). Unintended or ex post idle capacity arises because of circumstances that the entrepreneur had not foreseen when designing the plant. Ex post she may find a shortfall of demand (i.e. a classical Keynesian scenario) or a shortage of key inputs. For instance, a plant may have been designed with the mistaken expectation that adequate amounts of skilled labor, working capital and foreign exchange reserves to purchase imported inputs would be available. Thus the entrepreneur is forced to use her plant and equipment below that level for which it was designed. This is the type of idleness ‘to escape from as quickly as possible’ (ibid., p. 1303), something that could be accomplished via injections of demand. On the other hand, intended or ex ante idle capacity is deliberately

12

Strategic competition, dynamics, and the role of the State

planned. Part of this idle capacity is some reserve capacity to absorb unexpected jumps in demand and/or unexpected repairs and breakages (Andrews, 1949a). Another reason for ex ante idle capacity is that to operate beyond it the firm would have to incur higher prices of certain inputs. For example, given institutional and labor market norms, workers might have to be paid a premium over their basic wage rate to work at nighttime or on weekends. Furthermore, operating at higher rates of capacity utilization may raise repair and maintenance costs (Andrews, 1949a, p. 261). Of course, total fixed capital cost per unit output will fall as output rises as a consequence of nighttime or weekend operations. If the higher costs that are thus incurred from additional shifts and/or additional repair/ maintenance work exceed the falling average fixed costs then the plant will be deliberately kept idle during nighttimes or weekends. The normal or optimal level of capacity utilization that is chosen ex ante by the entrepreneur is that range of output at which these two opposing sets of unit costs more or less balance each other. Under- or overutilization of the capital stock is defined with respect to this normal rate of capacity utilization (Foss, 1963, p. 25; Kurz 1986, pp. 37–8, 43–4; Shapiro, 1989, p. 184). Winston’s arguments can be shown to be substantially consistent with those made by Andrews (1949a, 1949b). On the basis of the survey work carried out by the OERG of British firms during the inter-war period, Andrews argued that the average total cost (ATC) curve is not U-shaped but, for a given shift, declines steadily in the form of a rectangular hyperbola. The reason is that, given fixed costs, increases in output will necessarily lower average fixed costs (AFC). On the other hand, if the materials used up per unit of output as well as direct labor productivity are constant per shift, then the average variable cost (AVC) curve will be constant.1 Then the combination of these two costs (ATC 5 AVC 1 AFC) will also yield a rectangular hyperbola for every shift. On the other hand, additional shifts will entail the payment of overtime wages, which will raise the AVC curve (Andrews, 1949a, pp. 108–9). These propositions can be illustrated by means of a numerical example. We will first start out with the assumption that (a) a firm may operate up to three shifts and in the second and third shifts workers get a shift premium (overtime) relative to their base wage rate and (b) materials cost and labor costs per output are constant within a given shift although shift premiums make labor costs per output rise through the day. In all cases the AFC curve will be downward-sloping as the scale of output increases. In the absence of shift premiums the AVC curve will be a horizontal line. The ATC, AFC and AVC curves are shown in Figure 2.1. In this scenario minimum unit costs will be attained if the firm produces a

The microfoundations of long-run growth

13

Unit Costs (AVC, AFC, and ATC)

30 25

Shift 1

Shift 2

Shift 3

20 15 10

ATC

5

AFC

AVC

0 5

10

15

20

25

30

Output (Y)

Figure 2.1

Unit cost curves with no shift premiums

maximum output using three shifts. In this situation the firm’s optimum (or normal) rate of capacity utilization will be close to maximum engineering capacity (see Winston, 1974, p. 1310 for a discussion of actual and maximum capacity utilization). Next, assume that workers get overtime pay on the second and third shifts so that higher rates of capacity utilization raise labor costs per unit output and thus average variable costs.2 Given labor productivity, this makes the firm’s AVC curve an upward-sloping step function. The firm’s unit curves become as shown in Figure 2.2. Finally, consider the following two situations. In one scenario the shift premium on the second shift is 15 percent on the base wage rate and on the third shift it is 75 percent. These yield the curves AVC1 and ATC1. In the second scenario workers get 100 percent and 150 percent of their base wage on the second and third shifts respectively. These yield the curves AVC2 and ATC2. Figure 2.3 plots these curves (on logarithmic scales). Note that the AFC curve remains unchanged and thus is not shown. In these situations the minimum cost is either when output is 10 (ATC2 curve) or when it is 20 (ATC1 curve). The reason is that the minimum range will correspond to the opposing effects of the AVC and the AFC on the ATC. Thus, depending on the shift premium the firm may operate with only one or with two shifts in order to minimize costs. The consequence is that the cost-minimizing level of output or the ‘practicable optimum’

14

Strategic competition, dynamics, and the role of the State

Unit Costs (AVC, AFC, and ATC)

30 25

Shift 1

Shift 2

Shift 3

20 15 ATC 10

AVC

5 AFC 0 5

10

15

20

25

30

Output (Y)

Figure 2.2

Unit cost curves with shift premiums 30

Unit Costs (AVC and ATC)

25

Shift 1

Shift 2

Shift 3

20 15 ATC2

AVC2

10 ATC1 AVC1 5 5

10

15

20

25

30

Output (Y)

Figure 2.3

Average variable and total costs with different shift premiums

(Brunner, 1975a, p. 26) rate of capacity utilization may be consistent with varying degrees of deliberately planned idle capacity because of cost criteria (Winston, 1974).3 It should be emphasized that the practicable optimum generally involves some varying range of output rather than being a unique level. The key reason why it is the optimum is that

The microfoundations of long-run growth

15

it involves the exhaustion of all economies of scale and thus entails minimum costs (Brunner, 1952a). Brunner (1952a, p. 522) makes the important distinction between excess and reserve idle capacity. As discussed in section 3 of this chapter, from the OERG perspective competition between firms producing the same type of product will tend to create very similar prices.4 Competition drives each firm to its practicable optimum rate of capacity utilization. However, this practicable optimum position is consistent with a certain degree of reserve capacity in order to allow for unexpected demand fluctuations or breakages and repairs. On the other hand: Neither the short run nor the long run reserve capacity of the normal cost theory has anything to do with excess capacity or unexhausted economies of scale adduced in the theory of monopolistic competition. The former reflect the need for flexibility in a dynamic world, whereas the latter are due to multiplications of firms in a limited market, leading to too many firms all too small for efficiency. (Brunner, 1952a, p. 522, emphases added)

One can see in this passage the same distinction that Winston (1974) made between ex ante and ex post idle capacity where the latter is the classically Keynesian scenario involving a shortfall of demand. In a survey of US businesses based on a questionnaire, Eiteman and Guthrie (1952) found that the overwhelming majority of firms chose curves in which average total costs declined steadily to a minimum either at capacity or shortly before it. Not one firm chose the curve that showed constant unit costs across the entire range of output below capacity after an initial decline.5 Capacity in their paper was defined as the maximum possible output without the payment of overtime.6

3.

RIVAL THEORIES OF COMPETITION AND THEIR IMPLICATIONS FOR CAPACITY UTILIZATION

It may be asked, why should firms attempt to minimize costs? In conventional analysis, this condition only occurs under perfect competition, although it must be recalled that in perfect competition firms are entirely passive and make no active attempts to minimize prices and costs: firms are passive price-takers. The reason is because there are a large number of firms within an industry producing identical products, so that none has any market power. On the other hand in imperfect competition firms are under no such pressure because they have market power since they produce differentiated products: they are price-setters. In monopolistic

16

Strategic competition, dynamics, and the role of the State

competition there are many firms within an industry with no barriers to entry. In oligopolistic competition there are fewer firms within each industry and there are barriers to entry. It must be noted that perfect competition is the benchmark against which all other market structures are compared (Hahn, 1970, p. 5, 1973). Current debates generally assume this dichotomy between perfect and imperfect competition. And yet, as McNulty (1967) discussed in a paper that should have gotten far more attention, the rise of neoclassical theory in the late nineteenth century led to an analysis of competition that was radically different from that in Adam Smith and earlier authors such as John Stuart Mill. The key feature of this older view of competition is that all firms actively seek to lower prices and costs in order to undersell their opponents. In this older view, because the firm was not seen as facing a horizontal demand curve, the price charged was a variable rather than a parameter from the standpoint of an individual firm. This approach to pricing and output stands in sharp contrast to the essentially passive firm in the theory of perfect competition in which the demand curve faced by the firm is horizontal (Stigler, 1957, p. 5) so that the price becomes a parameter. Perfect competition, as Frank Knight pointed out, involves no ‘presumption of psychological competition, emulation, or rivalry’ (Knight, 1946, p. 102, cited from McNulty, 1967, p. 397). In the literature that followed McNulty’s seminal paper, most authors have generally focused on Marx’s theory of competition in rejecting the standard perfect–imperfect competition dichotomy (Clifton, 1977, 1983; Shaikh, 1982, 2008; Semmler, 1984; Bina, 1985; Botwinick, 1993). At the core of the distinction between this classical/Marxian literature and conventional analyses (both neoclassical and radical) is the very different economic environment faced by firms in the former perspective. The key feature in this framework is that all firms, large and small, are aggressive, price-setting and cost-cutting entities that seek to expand their market shares. Authors in this literature reject the market power view of mature capitalism and emphasize its highly competitive nature – in the classical sense of that term. The dynamics of competition, rather than market imperfections, tend to reproduce unequal profit rates within industries and roughly equal profit rates between industries (Shaikh, 2008). Perhaps unwittingly, scholars who were involved in the OERG study developed a new theory of the firm that was very similar to the classical view. The OERG rejected perfect and monopolistic competition because of their lack of relevance with regard to the real world. As with classical political economy, barriers to entry were seen as temporary, while firms large and small were modeled as aggressive, price-setting entities. While the OERG authors never went so far as to reject the theory of oligopolistic

The microfoundations of long-run growth

17

competition, it is argued in what follows that the analytical content of their contributions cannot be consistent with oligopoly theory. Rather, it is concluded, the OERG view can be married quite comfortably to the classical/Marxian theory of competition. Andrews started his new theory of competition by contrasting it with that of Kalecki: ‘The upshot of the theory certainly differs considerably from that of the theories of such economists as Kalecki. They think in terms of monopoly where I think in terms of competition, and I do not see the gross profit margin as a simple index of monopoly power’ (Andrews, 1949b, p. 54). Andrews went on to construct a model of competitive behavior in which price- and cost-minimization are central strategies adopted by a firm under highly competitive pressures. At the heart of the new theory of competition that Andrews and Brunner were attempting to craft there were three main propositions: a.

Firms set prices on the basis of their normal average total costs (Andrews, 1949b, pp. 82–3). This is the principle of normal-cost pricing.7 b. Competitive pressures from both inside and outside the industry will tend to make similar products have similar selling prices.8 c. Barriers to entry are in the long period not likely to be persistent.9 Points (a) and (b) follow from the argument (Andrews, 1949b; Harrod, 1952) that the pressure of actual competition and the fear of potential competition in an unknown future force each firm to set a price corresponding to the minimum-cost range of its ATC curve. Thus the high-priced strategy carried out by firms under the traditional models of imperfect competition is not sustainable under the actual and potential competitive threat from low-cost firms. OERG authors were thus describing a firm’s pricing policy in terms of its strategic attempts to maintain its competitive advantage. Following (b) however it does not follow that the rough equality of prices within a given product-industry leads to a convergence of profitability of the firms in it as in perfect competition. Drawing on his argument that Marshall has been misrepresented by neoclassical theory, Andrews adopts the Marshallian view that in the long run there is no need that ‘all firms in a given industry should be covering their costs, and thus have their survival assured, even in a position of long-run equilibrium’ (Andrews, 1951, p. 145). The main reason, Andrews argues following Marshall, is that the relative positions of firms in an industry will change because of changes in efficiencies of businesses ‘so that one business may make an innovation and go ahead at one time, but others will catch up later and

18

Strategic competition, dynamics, and the role of the State

the initiative will pass to a different business man’ (ibid., p. 145). Thus as Brunner (1975b, p. 44) put it: Not all firms at any one time need to be making profits even if there appears to be an equilibrium of industrial price. Some small firms will be new entrants, hanging on and hoping to grow by special merits and endeavours to a position of practicable optimum scale. Other firms, left behind in the race, will be on their way out. . . . Larger, more aggressive firms will not be able to drive from the market those businesses which are already on a sufficient scale to have paying-out (cash) costs which are no higher than the average total costs of the larger firms.

Note that it is the dynamics of competition rather than putative market imperfections that is the determinant of these differentials. This analysis of intra-industrial competition parallels that made by authors in the classical/ Marxian tradition (Semmler, 1984; Botwinick, 1993; Shaikh, 2008) and stands in sharp contrast to models of imperfect competition. With regard to point (c) Andrews and Brunner argue that industries with widely differing product lines may often have similar product processes and equipment that could make it reasonably easy, given a sufficient span of time, for a firm in an industry to enter another one with a totally different type of product line.10 It is for this reason that Andrews argues for the need for a broader definition of the concept of an industry, one which includes firms that produce widely differing products but utilize ‘substantially the same sort of process and equipment’ (Andrews, 1952, p. 207). It is worth quoting Andrews at length on this issue: In economic literature a lot is made, and rightly, of the difficulty of new businesses getting sufficient capital and, hence, of the fact that established businesses may make large profits before new businesses start coming into the industry. When, however, the detailed history of the individual businesses is examined, one becomes aware how normal it is for an established business to take on new products requiring broadly the same equipment, even if only in one department. In the course of time it is possible, in this way, for shifts in emphasis to lead to a business changing its typical product, and crossing the frontier from one industry to another. In fact, the frontiers of an industry are rarely so firmly fixed as they appear when, at any moment, one looks at the perhaps smaller number of apparently securely established businesses. Over a period one can often find a story of several attempts to expand into that product’s market, some succeeding, some failing (Andrews, 1949b, p. 85).11

Given that firms generally aim to have multi-product lines, it therefore follows that a given firm in a particular product market is far more vulnerable to attack than conventional theory allows. A common defense of the view that barriers to entry are persistent

The microfoundations of long-run growth

19

pertains to the role of advertising and product differentiation (Bain, 1956) in keeping aspiring entrants at bay. However, in their joint book Studies in Pricing (1975), Andrews and Brunner take issue with this argument. In one of her own articles (‘Industrial Analysis Revisited’) in this book, Brunner (1975b) asserts that such an argument relies too much on the implicit assumption that the bulk of purchasers are household consumers who would presumably be more likely to be seduced by grandiose advertising campaigns and brand-name loyalty. However, Brunner continues, a large part of the output of manufacturing industries (consisting of machine tools, raw materials and other intermediate goods) is sold to other businesses. As these latter firms are as much subject to the pressures of competition they will be equally hardnosed about the quality and costs of their inputs. Thus they are unlikely to be swayed by relatively superficial attempts to sell a product. Furthermore, Brunner observes in the above chapter, product differentiation is a double-edged sword since it could facilitate the entry of a new firm which could exploit some niche area. In his discussion of Harrod and Andrews, Edwards (1955), who was a former student of Andrews, argues that the attainment of the minimumcost position in Harrod’s (1952) revised model (see below), when the firm is earning normal profits, should not be construed as an attainedand-held equilibrium situation, since new firms may enter with lower production costs. In the way that was subsequently argued by authors in the classical/Marxist literature (Shaikh, 1982; Semmler, 1984; Botwinick, 1993) Edwards argues that, unlike the case in standard oligopoly theory, the degree of competitive pressure on incumbent firms has no relationship to the actual number of firms so that even a small number of firms can be consistent with high levels of competitive pressure. Following Andrews, Brunner (1975b) emphasizes that competition is a dynamic process and that if entry barriers are breached because of innovations or because new firms have lower ATCs relative to incumbents the former are likely to have smaller market shares initially and so will at first operate at above-normal costs which may be at or below the minimum ATC of the incumbent firm (ibid., pp. 45 and 167). However, the newcomers could eventually expand their market shares and operate at their normal capacities.12 Again following Marshall, Andrews argues that new firms seek to enter an industry by looking at the ‘costs of representative established businesses and come in if existing prices seem to be yielding abnormal profits to these’ (Andrews, 1952, p. 183). Now, since competition within an industry produces approximately equal prices, it follows that the abovenormal profits correspond to the most cost-efficient firms, which become the targets of potential new investment. It is the selling prices of these most

20

Strategic competition, dynamics, and the role of the State

efficient firms that determine the long-run industry supply-price (Andrews, 1951, pp. 146–7).13 In short, it is the so-called ‘price leaders’ within each industry that may face the ignominy of losing their market shares to new entrants. This aspect of Andrews’s theory of competition bears a striking resemblance to the notion of a regulating capital (Bina, 1985; Tsoulfidis and Tsaliki, 2005; Shaikh, 2008) in the classical/Marxist literature, which refers to the most efficient firms that are the targets of new investment. The vulnerabilities of incumbent firms are likely to increase if, because of huge costs, they are unable to vacate an industry in a prolonged slump. Exit may be a very long-drawn-out process; there is the example of a firm making locomotives in the inter-war years which only made book profits in one or two years after 1921 but was still in existence in 1938. Larger, more aggressive firms will not be able to drive from the market those businesses which are already on a sufficient scale to have paying-out (cash) costs which are no higher than the average total costs of the larger firms. (Brunner, 1975b, p. 44)14

Brunner’s (1975b) chapter ‘Industrial Analysis Revisited’ in Studies in Pricing is additionally significant because of the way that she distinguishes between industries and markets and her argument that firms from different industries may be able to compete against each other to satisfy a particular customer need (or demand). For example, a person may decide to store some household item in a plastic bottle, a glass jar or a metal can. In such a situation all three items, the products of different industries, are substitutes for each other (Brunner, 1975b, p. 37; Nightingale, 1978, pp. 34–5). One implication is that a given firm may potentially be able to erode the market shares of the others even though the latter may have initially had big entry barriers (Brunner, 1975b, p. 37; Nightingale, 1978, p. 35). In this situation the threat to the incumbent firm’s market share does not come from the ability of a newer firm to replicate the former’s production conditions but from the fact that the latter produces a powerful rival substitute with which it can invade the incumbent’s market.15 If successful in conquering market territory, the invading firm may be able to change its production and organizational structures and invade the incumbent’s industrial territory. Hence Brunner’s distinction between industry and market has far-reaching consequences as far as the ability of oligopolists to maintain their market hegemony is concerned. Andrews and Brunner emphasize the fact that competition is a dynamic process in which innovation plays a key role in beating rivals (Brunner, 1975b, pp. 43–4). This Schumpeterian aspect of their framework should alert us to the possibility that in the Andrewsian notion of industry, even if a firm does not currently share the technology of a powerful incumbent firm, the prospect of high profits could push it to innovate. This would

The microfoundations of long-run growth

21

allow it access to the industry even if, as Brunner points out, it is initially unable to operate at normal costs. While not discussed by Andrews and Brunner, the study of business history shows the intimate nexus between imitation and innovation as well as the role of the State in enabling small-sized firms to enter industries dominated by much larger ones (Kim, 1997; Mathews and Cho, 2000). As the late industrialization literature has discussed (Amsden, 1989; Chang, 2002, 2004) the learning–innovation–competition nexus has propelled the entry of new and smaller firms from peripheral capitalist countries into the ranks of established and putatively oligopolistic sectors such as automobiles, electronics and textiles, often leading to the decline of the erstwhile dominant firms.16 Once technological innovation and Andrews’s and Brunner’s redefinition of an industry are taken into account then that opens up the possibility that firms in radically different industries (in the Andrews and Brunner sense) could eventually develop technologies to move into other sectors. After all, given sufficient time, why would it not be possible for a cornproducing farm to move into the production of alcohol or agricultural implements? It is concluded here that the micro-analysis in Andrews and Brunner is a slippery slope since it inexorably undermines the notion of entry barriers which is at the heart of oligopoly theory. Drawing on Andrews’s insights on the threat of new entrants as well as a prior argument of Kaldor (1935, p. 72) regarding the implausibility of long-run excess capacity Harrod17 observes that a firm fixing a price arbitrarily above minimum costs in order to get a surplus profit is liable to make itself vulnerable to low-cost rivals. By charging the high price it forgoes the present opportunity of establishing itself in a somewhat larger market, and thus deliberately makes its position weaker for the time when it has to face the incursion of new entrants. Surely it will rather seek immediately to entrench itself in as large a market as it profitably can. . . . By all accounts and all hypotheses, the future is largely uncertain. No firm, which is interested in a certain line of production, wishes to sacrifice markets available to it for the sake of a fleeting surplus profit. Such a sacrifice will tend to make it weaker in facing the various contingencies of an unforeseeable future. I submit that any experienced man of business would pronounce it most ‘unsound’ to make a temporary surplus profit by charging a high price at which it is known that sales are unlikely to be capable of being maintained in the long run. It is wrong for economists to insist, on the basis of a partial theory, that this is none the less what entrepreneurs normally do. (Harrod, 1952, p. 147, emphases added)

Harrod points out that it is illogical to assume that after an entrepreneur has invested in plant and equipment of a given size he or she ceases to

22

Strategic competition, dynamics, and the role of the State

be concerned about the state of long-run demand confronting the firm because of the presence of a potential threat from rivals (ibid., p. 150). Long-period considerations in the determination of pricing and output, irrespective of the degree of ease with which new entrants can enter an industry (ibid., p. 174), necessarily imply some assumption about longperiod expectations, and Harrod in the above quote is quite explicit in emphasizing the role of uncertainty à la Keynes:18 in a non-ergodic world (Davidson, 1991) price- and cost-minimization by the firm are strategic attempts to safeguard itself against potential threats, although no outcome is guaranteed. He concludes that firms will adjust output to minimize costs, a process that will entail the to-and-fro movements of the demand curves. One may consider this to be a strategic form of profit maximization in a dynamic economic environment rife with uncertainty. This long-period strategic form of profit maximization should be contrasted with the profit maximization policy (involving marginal revenue MR 5 marginal costs MC) of neoclassical theory, which is based on the assumption that the entrepreneur pays attention only to the short-run MR curve and is knowledgeable about it (Harrod, 1952, p. 150). The problem with the MR 5 MC formula is that most entrepreneurs surveyed by the OERG were ignorant about the short-run marginal revenue curve (ibid., pp. 153 and 156). The further problem is that this short-run profit maximization seems to imply that the entrepreneur does not take into consideration the state of long-run demand even though plant and equipment last many years. OERG authors have emphasized that the long-run demand curve is likely to shift and/or become less steep with the entry of new firms (Andrews, 1949a; Harrod, 1952). From this standpoint all that the individual entrepreneur ‘knows’ is that, approximately, if she charges a higher price she will experience some loss of market share and if she lowers her price her market share will increase somewhat, given the goodwill she commands with customers relative to rivals (Harrod, 1952, p. 156). However, such a general understanding of the long-run forces of market competition is a far cry from a precise knowledge about the functional form of the long-run demand curve. Even if the textbook MR 5 MC calculation were accepted, Harrod and other OERG authors have asked, what sense does it make for an entrepreneur to invest in capital equipment lasting several years and yet content himself with a short-run profit maximization strategy? Further, the shortrun demand curve may itself be subject to frequent and unexpected fluctuations necessitating equally frequent adjustments to the firm’s prices. However, firms are generally disinclined to make such types of price adjustments because of the loss of customer goodwill, a point emphasized by the OERG authors (Harrod, 1939a), as well as its longer-term goal of

The microfoundations of long-run growth

23

remaining viable (Clifton, 1983). It is therefore not surprising that the MR 5 MC calculation is not done by actual firms, as emphasized by Hall and Hitch (1939) and Andrews (1949b).19 Full (or normal) cost pricing constitutes a more ‘practicable criterion for calculation’ (Harrod, 1952, p. 162) compared to the marginalist pricing theory. One can use the insights provided by Harrod and Andrews to analyze the adjustment mechanisms when u ≠ u*. In the slump when u , u*, the fall in demand will drive higher-cost firms out of business while making the remaining firms reduce capacity by reducing investment.20 One would expect price- and cost-cutting to be acute as firms struggle to increase sales. While it is true that each firm’s aliquot share of the industry rises (so that its demand curve shifts to the right), the fall in aggregate investment will reduce both aggregate demand and capacity. Provided the disequilibrium adjustment between aggregate demand and capacity is a stable process21 the capacity utilization rate will converge to u ≈ u*. In the event that there is a boom and u . u*, firms will expand investment (Harrod, 1952, p. 166) which will raise both aggregate demand and output. Again, cost-cutting to expand market shares is the key in the face of the competitive drive to expand market shares. Provided the disequilibrium adjustment is stable, the resultant rise of both demand and capacity will make the economy gravitate back to u ≈ u*. Harrod’s revised model met with some criticisms by a number of authors (Paul, 1954; Streeten, 1955; Weintraub, 1955). First, several of these authors argued that Harrod’s entrepreneur cannot possibly be concerned with the state of long-run demand in deciding on the optimal plant size and price for the simple reason that he or she cannot estimate this demand given the pervasiveness of uncertainty. Therefore, so the argument goes, it is more reasonable to assume that the firm maximizes short-run profits by accepting that point at which the demand curve is tangential to the downward-sloping part of the average total cost curve (see Figure 2.3). This would be an especially attractive position since ‘[I]f the profits were high and not too fleeting, and the loss of market were small, it would clearly be worth going for the profits’ (Paul, 1954, p. 34). Furthermore, it was argued that the absence of excess profits on the part of incumbent firms may not be a deterrent to incursion by new firms (Paul, 1954, pp. 33–4; Streeten, 1955, pp. 261–2). Thus, even if a firm achieved the minimum-cost point such a strategy would not shield it from invasion by new firms whose entry would drive the demand curve of the former to the left to the higher-cost and higher-price level. In short, nothing is to be gained in choosing the minimum-cost point since each firm will allow its demand curve to move to the left and ultimately end up earning excess profits (Paul, 1954, p. 35; Streeten, 1955, pp. 261–2).

24

Strategic competition, dynamics, and the role of the State

The problems with these arguments are at several different levels. First, it is precisely the presence of uncertainty that makes individual firms fear the actions of low-priced and low-cost rivals. In this war-like climate, this fear would be even higher if it had high short-run prices. The individual firm does not have to have relatively precise knowledge about the state of long-run demand in order to behave in this way.22 Second, one may make sense of Harrod’s position in light of the dynamic approach to competition articulated by Andrews and Brunner, a view that it shares with the Marxian analysis of competition.23 In this perspective, no outcome is guaranteed to the individual firm. Just because it seeks to minimize costs to increase its market share does not imply that its position is impregnable. Quite the contrary, since it may be dislodged by even lowercost rivals, as Edwards (1955) and Brunner (1975a, 1975b) discussed. From the standpoint of strategic competition the attainment of the minimum-cost point (or range) should not be seen as a steady-state situation, as Paul’s argument (1954, p. 37) implies. Given that production takes time, that investment is planned on the basis of expectations that may be wrong, and the fact that fixed investment cannot move instantaneously between industrial sectors, one would expect continual over- and under-shooting between demand/supply and supply/capacity – exactly as Smith and Marx argued. In such a dynamic and turbulent environment, no outcome is guaranteed. On the other hand, the maintenance of excess capacity with monopoly mark-ups assumes that firms which pursue such a policy feel secure about the future. One almost gets the impression that authors who subscribe to imperfect competition characterize firms as passive-aggressive entities in that while such firms are price-setters they passively let the market settle at a higher equilibrium price without being concerned about either their vulnerability under this state of affairs or their drive to expand their market shares. Pricing models inspired by Kalecki, the conventional wisdom in Post Keynesian circles, are based on the assumption that industrial firms exist in imperfectly competitive markets and mark up their prices on their constant unit prime costs. Kalecki was one of the principal originators of the view that the AVC curve (and thus the marginal cost MC curve) takes the shape of an inverted L. Firms mark up their price p on the basis of their constant marginal cost mc with the index of monopoly power given by μ 5 (p − mc)/p (Lerner, 1934). It is postulated that there is generally excess capacity so that any increases in demand are met by increases in output and employment as opposed to prices. As for the monopoly mark-up itself, Kalecki’s position was that a number of factors such as the degree of industrial concentration and barriers to entry enabled such firms to set their prices (Sawyer, 1985).

The microfoundations of long-run growth

25

Kalecki took into account competition from other oligopolists within the industry so that a firm’s price was a joint function of its own unit prime costs and some weighted average industry price p (Kalecki, 1954): p 5 mu 1 np

(1)

If p is the industry average price then p 5 mu/(1 − n)

(2)

As Kalecki put it: ‘The coefficients m and n characterising the price fixing policy of the firm reflect what may be called the degree of monopoly of a firm’s position’ (Kalecki, 1971a, p. 45, cited from Sawyer, 1985, p. 24). The coefficient m is the mark-up, while n is a reaction coefficient (0 , n , 1) that relates each firm’s price to those of rivals. Sawyer (1985, p. 25) shows that Kalecki’s particular formulation of a firm’s pricing strategy can be related to other theories of oligopolistic competition in which the price charged by a firm is a function of its marginal costs, its and rivals’ price elasticities of demand, and the interdependent nature of the prices charged by oligopolists.24 As with Harrod and Andrews, Eichner (1980) takes issue with the Chamberlin–Robinson model of monopolistic competition because of its short-run focus and the fact that it ignores the threat posed by rival firms when a given firm raises its price. Eichner extends the Kaleckian mark-up model by taking into account longer-run competitive factors. However, as with Kalecki, such competition generally precludes the threat of new entrants because of entry barriers. Eichner’s basic insight is that when an oligopolist raises its mark-up rate the higher price will make it vulnerable either because customers will switch to competing lower-priced products and/or because higher-cost firms outside the industry will find it easier to enter. The barriers to entry confronting potential entrants into the industry include the fact that incumbent firms benefit from larger economies of scale, have lower unit costs (because of access to more sophisticated technologies), and have the ability to have built up high levels of customer loyalty because of a longer presence in the market and huge advertising expenditures (Eichner, 1976, pp. 72–3). On the other hand, these barriers may be scalable by potential new entrants the higher the price charged by the incumbents. Thus oligopolists’ pricing policies are the key, in Eichner’s model, to maintaining entry barriers. As is typically the case with models of oligopolistic competition, firms within an industry do not attempt to cut prices to expand their market shares, as this provokes equally swift retaliation by rivals, an outcome that

26

Strategic competition, dynamics, and the role of the State

is said to leave relative market shares unchanged (ibid., p. 39). Instead they set prices and sell quantities determined by aggregate demand (ibid., p. 38). Prices are determined on the basis of the full-cost principle and are set to cover the AVC, AFC and an adequate proportion of internal cash flow (ibid., p. 63) at some standard rate of capacity utilization. Eichner defines the ‘capacity’ of a machine as that level of output that minimizes its average variable cost, that is the most efficient scale of production (ibid., p. 36). The standard rate of capacity utilization is that percentage of capacity that the firm has used on average over the course of the business cycle (ibid., p. 62). The optimal price thus set by the oligopolist corresponds to the flat portions of the MC and AVC curves but is arbitrarily to the left of the minimum point of the ATC curve (ibid., p. 64, Figure 6) because of the firm’s goal of maintaining an adequate level of idle capacity. In the final instance then the firm sets a price which it holds constant, and the rate of capacity utilization is determined by changes in aggregate demand (ibid., pp. 43–4). The absence of cost-minimization and price-cutting strategies in models of oligopolistic competition can only be true either if firms within the industry have identical cost curves or if they make no attempt to lower their unit costs by spatially relocating to low-wage regions and/or increasing labor productivity. Eichner’s claim that once an industry’s price leader has set the price ‘[T]he individual megacorp has no control over what that price will be nor, once the price has been determined, can it unilaterally alter that price’ (ibid., p. 43) is particularly puzzling. His claim that such a cooperative solution to what he considers mutually destructive price wars was well known since Adam Smith is problematic, as both Smith and Marx had addressed this issue. In discussing Marx’s analysis of collusion between a small number of firms, Botwinick (1993, p. 138) observes that the former considered such arrangements to be inherently unstable, as each firm would have the incentive to break free of collusive activities in order to capture rivals’ market shares. The common interest is appreciated by each only so long as he gains more by it than without it. And unity of action ceases the moment one or the other side (buyers or sellers) becomes the weaker, when each tries to extricate himself on his own as advantageously as he possibly can. Again, if one produces more cheaply and can sell more goods, thus possessing himself of a greater place in the market by selling below the current market-price, or market-value, he will do so, and will thereby begin a movement which gradually compels the others to introduce the cheaper mode of production. (Marx, [1894] 1967, p. 194, cited from Botwinick, 1993, p. 138)

In a similar vein, writing about competition in Adam Smith (1937, p. 343), McNulty (1967, p. 397) observes:

The microfoundations of long-run growth

27

As far as the concept of competition is related to market structure, we should have to say that Smith, by suggesting that the individual seller could sell more by lowering price and less by raising it, presented a theory of imperfect competition. But, in fact, Smith’s use of the term seems to have been largely independent of market structure. Of duopoly, he wrote: ‘If . . . capital [in the amount required to satisfy the demand for groceries] is divided between two different grocers, their competition will tend to make them both sell cheaper.’ Although Smith specified that competition would be the more active, the greater was the number of competitors, the essence of competition in duopoly was evidently what it was in any other market structure, namely, the attempt to undersell one’s rival in the market by lowering price.

In the neoclassical textbook rendition of oligopolistic competition (Varian, 2006, Chapters 27 and 28) strategic behavior between firms is modeled in a game theoretic framework. On the basis of the decision of new entrants to enter or stay out of an industry, incumbent firms choose different strategies to maintain entry barriers. One entry barrier is the construction of excess capacity as a deterrent to entry since it allows incumbent firms to expand output and thus lower their unit costs of production. In this situation the potential entrant will make no profits if he does enter, while the incumbent firm will make profits albeit at a level which is lower than the situation in which the potential entrant stays out. On the other hand, with regard to competition between oligopolies within an industry, excess capacity may or may not be present, since that would depend on the pricing strategies of the rivals. For example, in the duopoly case if both firms tacitly (or overtly) agree not to engage in a price war there will be excess capacity. However, if one of them chooses to break from this agreement the ensuing price war will drive their prices to their minimum average total costs. This ‘tit-for-tat’ strategy leads to what is called a Bertrand equilibrium. The above two scenarios however lead to a problematic implication. If a price war between oligopolists within an industry drives outputs to their respective minimum-cost points and thereby eliminates excess capacity, how are incumbent firms in a position to shield themselves from new entrants via the maintenance of excess capacity? Put simply, the key characteristics of oligopolistic markets – limited entry and market power – break down once allowance is made for both of these competitive processes. The only way by which potential new entrants may still stay out in this situation is if the incumbents’ unit costs are lower than those of the former. However, this situation does not require excess capacity and, as discussed above, is unlikely to be sustainable over the longer run if potential entrants somehow lower their unit costs. Leaving aside intra-industrial competition, the maintenance of barriers to entry in the game theory version of oligopolistic competition leads to

28

Strategic competition, dynamics, and the role of the State

two issues. First, such a strategy is susceptible to the same criticism that Harrod made against the Chamberlin–Robinson model of monopolistic competition. Harrod’s argument was that it was illogical for a firm to pursue a high-priced strategy and thus maintain excess capacity (presumably to snatch temporary short-run profits) without taking into account the possible longer-run threats from lower-priced firms (Harrod, 1952, Essay 8). Second, a high-priced (or excess capacity) strategy will make sense only if the firm does not expect any long-run threat by lower-cost firms. Only in this situation will the excess capacity strategy by the incumbent be a Nash equilibrium.25 This however raises an important question: how are expectations modeled in the game theory framework in order for individual agents to compute the appropriate payoffs from their output and pricing strategies?26 Game theory takes it as axiomatically given that agents are rational where rationality makes each player maximize his or her payoff. This in turn implies that each player knows the number of rivals that confront her or him, all the strategies that opponents are likely to use, and all possible payoffs. Further expectations about rivals’ business strategies are made on the basis of known probabilistic expectations, the standard way in which neoclassical theory models what it calls uncertainty (Dockner et al., 2000, p. 11). The key assumption is that each player has to be fully knowledgeable about the rules of the game and the utility functions of all the players. Furthermore, ‘[E]ach player must also be aware of this fact, i.e. of the awareness of all the players; moreover each player must be aware that each player is aware that each player is aware, and so on ad infinitum’ (Aumann, 1987, p. 473). This assumption of common knowledge is the basis of all game theory, including games involving imperfect information (ibid., p. 473), since ‘otherwise the model is insufficiently specified, and the analysis incoherent’ (ibid., p. 473). The assumption about consistently aligned beliefs implies that since rational people have access to exactly the same information set they will also come to have identical thought processes with regard to the particular game being played (Foss, 2000). In short, some extreme assumptions need to be made in order to ensure that the players have access to an enormous amount of information.27 But as with many market phenomena it is a tall order to believe that individual firms can anticipate correctly (even in a stochastic sense) the range of competitive strategies used by both actual and potential rivals. Furthermore, it is fairly realistic to assume that business investments in the current period set in motion processes that irrevocably change the economy. For example, such investments could stimulate the development

The microfoundations of long-run growth

29

of new products, processes and skills, which in turn would lead to further economic mutations. In the absence of rational expectations or probabilistic calculations under imperfect information, how is the firm to set up a payoff in such a constantly evolving environment? Construction of a payoff matrix becomes no simpler for the firm once one replaces the small-group assumption of oligopoly theory with the argument that individual firms are subject to competitive pressures from many other actual or potential competitors of all sizes. On the other hand, as any standard textbook depiction of a duopoly game shows (see for example Taylor, 2004), it is curious that one contingency which is ruled out is any kind of aggressive price- and cost-cutting behavior to wipe out the opponent by an across-the-board lowering of average total costs. Such a contingency could arise if a firm successfully succeeds in lowering its unit costs of production (by raising labor productivity and/or finding low-wage labor in some other country). Neoclassical firms, even oligopolistic ones, are so well behaved that they passively accept the Nash equilibrium. But the notion of the Nash equilibrium is correct only because of the way the problem has been set up, that is the absence of aggressive behavior to wipe out one’s opponent. This curious passivity of oligopolists persists even when an attempt is made to dynamize the analysis through repeated games over multiple time periods (Krugman and Wells, 2006).28 Illustrations of the use of game theory for practical public policy discussions are equally striking in their essentially passive view of oligopolists’ competitive strategies. Depending on the example, oligopolists either passively reduce their output when rivals cut prices and expand output (Brander, 1986, pp. 28–9) or do not attempt to retaliate by cutting prices and costs, by say outsourcing, when rivals benefit from public subsidies (Krugman, 1987). One may relate this discussion more concretely to the issue of global competition, where any given firm, say in the auto industry, confronts many competitors. In this situation the firm confronts two problems in attempting to set up a payoff matrix. First, while it may know of the actual number of competitors at any given time period, it would be virtually impossible for it to gauge the actual number of new competitors coming into the industry in the following years from all over the world. On the other hand, no one would seriously claim that businesses take into account only the current period in deciding on investment plans, which very often span many years. Second, the payoff matrix would become extremely large, because of the individual payoffs of each firm, and the optimal situation from the standpoint of an individual firm would be very difficult to compute. The question is, is it credible to believe that businesses actually perform such calculations?

30

Strategic competition, dynamics, and the role of the State

In fact any kind of realistic analysis of entrepreneurial behavior has to recognize that the competitive or innovative enterprise arises from a complex combination of historical and institutional factors, including the role of the State (Amsden, 1989; Chang, 2002, 2004; Chibber, 2003; Lazonick, 2008). The development of new technologies, a process which is at the heart of the dynamics of competition, cannot surely be anticipated on the basis of rational expectations – how could many such technologies be characterized as innovations? In fact, historians of technology have argued that technological change takes place in a non-ergodic manner (Rosenberg, 1994; Nelson, 1997).29 At the core of the use of game theory to study oligopolistic competition is an assumption about the formation of expectations which in turn hinges on the distinction between ergodic and non-ergodic processes. In a non-ergodic world the future is fundamentally unknown, so that future trends of many variables cannot be estimated in terms of time-invariant probability distributions (Davidson, 1982, 1991). Thus many economic time series are non-stationary because ‘the economic environment is not homogeneous over a period of time (perhaps because non-statistical factors are involved’ (Keynes, 1939, p. 560, cited from Davidson, 1991, p. 132).30 A sufficient but not necessary condition for non-ergodicity is non-stationarity where the process of historical and social evolution generally precludes the possibility of mean-reversion or constancy of variances. Historical evolution implies that the distribution functions of stochastic processes are time-dependent so that rational expectations of future trends, based on ‘all’ available past and present information, will tend to generate misleading forecasts.31 On the other hand, neoclassical theory models expectations by assuming that there is an objective and time-invariant probability distribution for all variables such that the past can be extrapolated in a stochastic sense into the future. Thus economic processes are fundamentally ergodic stochastic processes so that the past average of a variable cannot be systematically different from future average outcomes of that variable. To quote Davidson (1991, p. 132), ‘In the ergodic circumstances of objective probability distributions, probability is knowledge not uncertainty.’ The non-stationary versus stationary distinction implies that there is a difference between probability and uncertainty (Keynes, 1937, pp. 213– 14), as Keynes argued in his restatement of the principal claims made in The General Theory: By ‘uncertain knowledge’, let me explain, I do not mean merely to distinguish what is known for certain from what is only probable. The game of roulette is not subject, in this sense, to uncertainty. . . . The sense in which I am using the

The microfoundations of long-run growth

31

term is that in which the prospect of a European war is uncertain, or the price of copper and the rate of interest twenty years hence, or the obsolescence of a new invention, or the position of private wealth-owners in 1970. About these matters there is no scientific basis on which to form any calculable probability whatever. We simply do not know.

Perhaps Hicks summed up best the issue of uncertainty (Hicks, 1979, p. vii, cited from Davidson, 1991, p. 133): ‘One must assume that the people in one’s models do not know what is going to happen, and know that they do not know just what is going to happen.’ One cannot find a clearer implicit critique of the treatment of knowledge in the game theory framework! The existence of barriers to entry is the sine qua non of monopoly power since it allows incumbent firms to shield themselves from potential entrants into the industry. The barriers to entry are said to arise from the fact that incumbent firms benefit from a combination of factors that potential new entrants do not possess: absolute cost advantages, economies of scale from large-scale production, and the greater availability of finance. Following Bain, Sylos-Labini and Modigliani, incumbent firms set a price that deters the entry of potential new firms (Zamagni, 1987). Theoretical issues aside, the problem with the persistent-barriers-toentry argument is that it flies in the face of historical evidence. Business history provides ample evidence regarding the evolution of technological learning within different industrial sectors which at some point has enabled firms to leapfrog into new sectors along the lines argued by Andrews. Both Toyota and Nissan owed the development of their worldclass automobiles to technologies developed in the cotton textile and aircraft industries (Daito, 2000); Toyota in fact moved into the automotive sector after its successful development of textile machinery (Mass and Robertson, 1996). As with the erosion of the market share of US Steel because of competition from low-cost domestic and international minimills,32 the loss of US car manufacturers’ market shares to Japanese (and to some extent South Korean) car producers is hardly consistent with the existence of persistent entry barriers. In fact practical analysts of real-world business behavior are quite aware that incumbent firms even in sectors that are difficult to enter can sometimes be challenged quite successfully by highly competitive new entrants – although the ability to enter is not an inevitability it would be a stretch to assume that it is an impossibility. In their article in the Harvard Business Review, Bryce and Dyer (2007) discuss the attack strategies used by firms to enter new markets, a process that requires complicated methods of organizing technologies, management, labor and finance. Sometimes, provided other divisions are profitable enough, multi-product and multi-divisional firms are willing to tolerate losses for considerable

32

Strategic competition, dynamics, and the role of the State

periods as they seek to gain a beachhead into new industries. Finally, as a number of authors have written, powerful developmental states were often central players in facilitating the entry of private firms into new industries, especially knowledge-intensive ones (Kim, 1997; Mathews and Cho, 2000; Lazonick, 2004, 2008). In short, if even heavy industries can become porous over long stretches of time because a wide range of factors, often unpredictable to incumbents, determine the competitiveness of potential entrants we will be forced to move beyond conventional models of competition to the Weltanschauung of Andrews, Brunner, Harrod and others of the OERG as well as the insights of classical political economy. As Andrews (1949b, pp. 88–9, emphasis added) put it: Profits sometimes seem generous in the cases of well-established businesses which have been able to build up for themselves a rather specialized market, even in what may appear to be a competitive industry. Experience, however, suggests that such cases do not last, and that long-term forces do readjust the size of the margin. The tide of competition may leave little pools of abnormal profit behind it, but in the end they tend to disappear. In general then, experience of industry does suggest that the business man is right when he sees his gross margin and his price as competitively determined. In our view, the newer developments in theory have caused economists to be too ready to regard manufacturing industry as a network of monopolies, and of deliberate restriction of output.

4.

THE PERSISTENCE OF EXCESS CAPACITY: SRAFFIAN AND KALECKIAN APPROACHES

In neoclassical theory the persistence of excess capacity follows from the profit maximization condition MR 5 MC in the Chamberlin–Robinson model of monopolistic competition or via a game of strategic deterrence by incumbent firms in oligopolistic competition. Bearing in mind that heterodox authors eschew marginalist principles as well as game theory, in the current section heterodox arguments for the persistence of excess capacity are reviewed. (a)

Sraffian Perspectives

Following the classical and Marxian tradition Sraffian authors emphasize the centrality of long-period positions in which market prices gravitate around prices of production. Drawing on the insights of Garegnani (1992), Palumbo and Trezzini (2003) describe one version of the Sraffian school’s approach to capacity utilization in the long run. These authors

The microfoundations of long-run growth

33

remind us that neo-Keynesian growth models of Kaldor (1956) and Robinson (1956, 1962) relied on the steady-state assumption such that output growth was continuously equal to capacity growth. This would necessarily lead to Harrod’s (1939b) warranted growth rate. Palumbo and Trezzini (2003) object to the distinctly anti-Keynesian nature of this steady-state framework because, they argue, it undermines the positive role that autonomous demand is said to play in raising long-run growth. Following Garegnani, Palumbo and Trezzini (2003) take the view that in fact the system shows far greater degrees of elasticity than is implied by the neo-Keynesian growth models. They argue that, while there are inherent forces that make the actual rate of capacity utilization tend toward the normal one, this is necessarily a very slow process and indeed the adjustment process that reduces this deviation generates countervailing tendencies that move output away from capacity. For them, ‘the normal utilization hypothesis, which is so central in the analysis of prices, plays no analogous role in the analysis of quantities because, unlike the situation with respect to prices, no gravitation of actual quantities towards “normal” quantities may be assumed’ (2003, p. 110). In the final instance then, in this version of the Sraffian framework, actual rates of capacity utilization will be different from the normal one for extremely long periods, thereby ensuring the relevance of Keynes’s Principle of Effective Demand even in the long run. In order to revive the demand-led growth to accumulation, Palumbo and Trezzini (2003) discuss a number of what they call ‘offsetting forces’ that retard capacity from catching up with demand. First, given the nature of fixed investment, they make the intuitively correct point that firms consider changes in their capital stocks only when the deviation of capacity from demand is systematic rather than transitory. This necessarily involves time. Second, following Harrod and Domar, they point out that fixed investment has a dual nature in that it is both a source of demand and a source of capacity. They argue that, while adjusting capacity to demand, the induced injection of investment in turn provides a stimulus to demand ‘something that is likely both to slow down the completion of the whole process and to generate endogenous forces that may hamper the tendency to adjustment itself’ (ibid., p. 118). Thus, they conclude, it is quite likely that the fully adjusted position will never be reached, or take a very long time to do so. They do not offer any empirical evidence to substantiate this argument. Third, they argue that adjustment from one steady state to another one at normal capacity utilization would require both perfect foresight and implausible behavioral reactions on the part of firms in a decentralized

34

Strategic competition, dynamics, and the role of the State

and competitive market economy. Suppose there is a rise in autonomous demand, caused say by an increase in government spending, so that output and thus the capacity utilization rate rise above the normal level. Trezzini (1995) argues that if full adjustment between the higher aggregate demand and capacity has to happen firms will have to be able to distinguish changes in aggregate demand brought about by the persistent change in autonomous demand from transitional changes caused by the adjustment process. Furthermore, each firm must also be able to estimate the share of the additional aggregate demand that is aimed at its own products. Moreover, Palumbo and Trezzini (2003) claim, the increase in demand has to be constant and not be subject to random fluctuations, in order for firms to make the “correct” investment decision. They conclude that it is this inability to distinguish permanent from transitory additions to demand that makes it impossible for firms to choose the ‘correct’ capacity to meet the demand. On the other hand, they claim, models in which capacity does catch up to demand implicitly rest on the assumption that firms have perfect foresight. Trezzini (1995) argues that, since the higher levels of above-normal capacity utilization rates have been reached with high levels of induced investments, it seems implausible that firms would eventually reduce the absolute level of investment at a time that capacity utilization is above normal. That is to say, Trezzini takes the view that in an out-of-steadystate situation it seems unlikely that the representative firm would reduce investment to bring itself on to the underlying steady-state growth path. The only factor that might induce the firm to do so would be if it was certain that all the other firms were planning to do so. Such is however not the case under decentralized competitive markets when each firm would have the incentive to raise, and not cut back, investment. On the other hand, Trezzini concludes, with imperfect foresight and knowledge when firms’ investment decisions are influenced by the current state of the market they would ‘certainly continue to expand their capital stocks at a rate higher than that required by the adjustment. If even only a few firms did not reduce investment, actual demand would remain at levels implying overutilisation and this situation could continue forever’ (ibid., p. 53, emphasis added). It may be asked whether this assumption of a permanent state of disequilibrium is a realistic description of the market economy. For one thing, what is to prevent this disequilibrium from producing knife-edge instability? No one, least of all Harrod (1973), would seriously claim that actual capitalist economies, unstable as they really are, actually experience this kind of instability. Furthermore, Palumbo and Trezzini do not discuss the economic

The microfoundations of long-run growth

35

consequences of the state of quasi-permanent disequilibrium and/or knife-edge instability. Would there not be a pressure on prices and wages? Moreover, such a quasi-permanent situation would in turn entail persistent above-normal costs. That being the case, what about each firm’s competitive position vis-à-vis actual and potential rivals? As the OERG authors stressed, the attainment of a normal (or practicable optimum) rate of capacity utilization follows from the potential threat of new entrants. The adjustments to fixed and circulating investments to minimize costs arise as safeguard measures precisely because of the potential for attack by lowcost firms in an uncertain future (Harrod, 1952). ‘Incorrect’ investment adjustments are bound to be made by some firms which will get wiped out. Moreover, some of the arguments made by Palumbo and Trezzini could equally well be used to rationalize permanent demand–supply imbalances, an implication that would lead to the abandonment of the multiplier model, which is based on the equilibrium between demand and supply. One may well ask, in what way is the Sraffian school’s rejection of the normal capacity utilization argument consistent with its theory of longperiod prices? After all, following Garegnani (1976), authors in this school take the classical view that actual prices gravitate around prices of production where the latter arise from the establishment of an inter-industry general rate of profit. However, it is denied that this situation coincides with the attainment of normal capacity utilization. Following Ciccone (1986, pp. 24–5), Palumbo and Trezzini (2003, p. 121) argue that, ‘unlike the attainment of adjustment between demand and capacity, the process of price gravitation in no way requires the whole system to be in a state of adjustment of all its variables.’ The rationale for this view is based on the argument that firms will often undertake certain kinds of investment that expand capacity, without taking demand considerations into account (ibid., p. 122). These types of investments are a response to technical change and intra-firm competition: Competition can lead entrepreneurs to take investment decisions that are not induced and justified by the expected expansion in demand. . . . In addition, however, competition is also the force that induces firms to change their methods of production and their products or to enter new sectors – in other words, to innovate – in order to exploit new profit opportunities. And it is the force that obliges the non-innovative firms to react and take their own investment decisions if they are to survive. Firms can make innovative investments even when faced with stagnating demand. In point of fact, they could even be said to have a greater stimulus to undertake (at least some kinds of) innovative investment when the demand for the products of their sectors is growing slowly or in decline, since in this case the ability to take innovative action aimed at cutting costs may well make the difference between survival and failure. (Palumbo and Trezzini, 2003, pp. 122–3)

36

Strategic competition, dynamics, and the role of the State

Quite simply then, competitive forces delink any relationship between demand and capacity. The above argument is problematic for a number of reasons. The whole objective behind cost-cutting is to expand market shares and to exploit new market shares. But are not these attempts ‘to exploit new profit opportunities’ consistent with the argument that demand rules the roost? Furthermore, is it conceivable that a firm, after investing in a new type of technology, will continue maintaining it if it finds that it has been unsuccessful in wresting market shares from other firms? There is also an internal inconsistency in their argument. If price- and cost-cutting are the basis on which firms seek to expand their market shares, then the capacity overutilization pressure faced by firms would not be sustainable. As argued in section 3 of this chapter, the threat of invasion by new firms would expand both aggregate demand and capacity and, provided there is no knife-edge instability, the overutilization problem would be eliminated. In other words, the pressure of competition which Palumbo and Trezzini admit as being real cannot be consistent with arbitrary rates of capacity utilization. Ciccone (1986) correctly argues that the relative rates of profits on new investments, as opposed to the existing capital stocks, determine the ebb and flow of capital between industrial sectors and thus the establishment of a general rate of profit. He concludes that since the newly created capacity from these new investments is the most suitable for expected levels of demand the establishment of prices of production ‘does not in fact seem to require the simultaneous gravitation of the effective utilization of capacity around its “normal” level – i.e., the level of utilization implicit in those prices’ (p. 24). Such an argument is, however, inconsistent with normal-cost pricing. Surely the calculation of a firm’s profit rate has to include the costs of all of its capital equipment that has not reached the scrapping margin? As Harrod (1952) observed, in the absence of sufficient sales it is unclear what business rationale there is for firms to maintain under- or non-utilized plant and equipment, especially since it is a cost to maintain this equipment. Furthermore, would not the accumulation of undesired excess capacity lead to a fall in prices and thus a capital outflow? Conversely, an undesired deficiency of capacity is likely to provoke a capital inflow. Either of these scenarios would lead to to-and-fro movements of each firm’s demand curve until u ≈ u*. It is important to point out that many heterodox authors do not generally distinguish between intended and unintended idle capacity (see Nell, 1985 or Ciccone, 1986), as Winston (1974) argued. There appears to be confusion between that kind of idle capacity which arises from a shortfall

The microfoundations of long-run growth

37

of demand and the type of idle capacity that a firm deliberately maintains to not risk losing market shares. This confusion is evident in Ciccone’s analysis. At first he discusses the rationale for reserve capacity as a deliberate competitive goal: Indeed, what is observed in reality is a utilization of capacity that varies from relatively high levels, corresponding to the peaks reached by production in boom periods, to relatively low levels, corresponding to times of recession. This simple ascertainment gives us reason to think that the size of capacity installed is commensurate with the relatively higher levels of demand that entrepreneurs expect to encounter, with a certain frequency, during the economic life of their plant. Its volume would therefore be considerably larger than the expected average levels of production that can be foreseen. The conclusion thus suggested by observation of the facts seems moreover to find an explanation in the need of the individual firm not to lose market shares when demand goes up – and so, ultimately in the pressure of competition. (Ciccone, 1986, p. 27)

However, in a footnote to the above paragraph Ciccone quotes Kalecki (1954, 1965) in order to make the claim that this reserve capacity is consistent with the excess capacity generated in a recession: The normal under-utilization of productive capacity referred to in the text seems precisely to be what Kalecki considers, in the following passage, as a characteristic proper to a capitalist economy: ‘A considerable proportion of capital equipment lies idle in the slump. Even on average the degrees of utilization throughout the business cycle will be substantially below the maximum reached during the boom (. . .). The reserve of capital equipment and the reserve army of unemployed are typical features of the capitalist economy at least throughout a considerable part of the cycle.’

The reader is reminded that this conflation of excess and reserve capacity stands in sharp contrast to Winston’s (1974) distinction between ex ante and ex post idle capacity or, in the same vein, Brunner’s (1952a) distinction between unexhausted economies of scale and the attainment of a practicable optimum. In the Sraffian school the contributions of Heinz Kurz are of particular importance since they describe in great detail the meaning of a normal rate of capacity utilization. Following Marris (1964), Winston (1974) and others, Kurz (1986, 1992) argues that planned idle capacity is a deliberate cost-minimization strategy because to operate beyond this level by, say, bringing in additional shifts would entail the payment of higher wages in order to induce workers to work beyond normal working hours. Since wage contracts are institutionally determined, it follows that to a large extent the cost-minimization range of production, or the normal rate of capacity utilization, chosen by firms is a function of institutional

38

Strategic competition, dynamics, and the role of the State

circumstances. In contrast to authors who are more inspired by Kalecki and Steindl, Kurz takes the position that the principle of effective demand cannot invalidate the usefulness of the traditional concept of ‘normal’ positions . . . the forces envisaged by the classical economists as pushing the system towards that centre of gravitation are still effective. . . . However, whereas in classical theory the process of gravitation around the (slowly moving) centre was assumed to follow a path sufficiently close to the one described by the centre itself, it can no longer be presumed that this is the case. Indeed, it cannot be precluded that deviations of the actual situation from the ‘normal’ one may become large, and remain so for a long period of time. (Kurz, 1986, p. 40)

For Kurz, however, the attainment of a normal rate of capacity utilization occurs only under conditions of what he calls free competition (Kurz, 1986, p. 43) in which there are no barriers to entry or exit. This, in his view, underpins the classical profit rate equalization dynamic which, as he argues, can only be true if all firms (both incumbents and potential new entrants) have access to the same technology; technology differentials would imply monopolistic or oligopolistic power (Kurz and Salvadori, 1995, Chapter 1). By implication, in this latter situation only would actual rates of capacity utilization differ persistently from normal (i.e. costminimizing) rates of capacity utilization. However, Kurz’s interpretation of Marx’s theory of competition is inconsistent with that of authors such as Shaikh (1982), Semmler (1984) and Botwinick (1993) whose analyses of Marxian competition assume neither readily available technology nor zero barriers to entry and exit. Because movements of fixed capital are involved one would expect unequal profit rates to exist at any given moment in time. However, the inexorable pressures of competition between industries lead to approximately equal profit rates only over a period of several years.33 Therefore the attainment of a normal rate of capacity utilization does not require the assumption of free competition. (b)

Kaleckian Perspectives

If the traditional notion of the normal rate of capacity utilization is determined by cost-minimizing criteria (which are themselves regulated to a large extent by factors, some of which are institutional, that are external to the firm), Kaleckian authors such as Lavoie (1995, 1996a, 2003) have taken the position that the normal or target of capacity utilization is entirely endogenous to the firm and thus even in the long run the firm faces no capacity constraints. Thus the dual paradoxes, the paradox of thrift

The microfoundations of long-run growth

39

and the paradox of costs,34 hold even in the long run. As in all Kaleckian models Lavoie assumes some version of imperfect competition. Lavoie (1996a) deals with two issues. First, he wants to show how in the basic Kaleckian model with excess capacity the dual paradoxes hold even in the long run. Second, he wants to rebut a common criticism made by Marxists and Sraffians who question the realism of the persistent excess capacity hypothesis, by showing how an endogenized normal rate of capacity utilization can resurrect the dual paradoxes in the long run. Basically, in this view firms are contented with any given level of the rate of capacity utilization.35 He first has to define what he means by the normal rate of capacity utilization. Commenting on the reality of idle capacity that firms maintain, Lavoie observes: It is assumed that each segment of a plant is operated at its optimal level, as defined by the engineers, under the standard requirements of cost-minimization. Companies have several plants, but expect that in normal conditions some of the segments will not be running. One may wonder why firms would continue to upkeep such plants. The reasons have to do with the presence of uncertainty, as was pointed out by Steindl (1952). Firms hold on to excess capacities to face an uncertain future, just as agents hold on to cash balances for liquidity purposes. These reserves of capacity allow firms to respond quickly to changes in demand and in its structure, and in addition they are deterrent to entry by new firms. The normal rate of capacity utilization is thus a convention, defined by the above considerations and by historical experience with regards to the rates of capacity that can be expected to be realized on average, as a consequence of macroeconomic conditions. (Lavoie, 1996a, p. 120, emphases added)

Firms are concerned with minimizing their overall costs, and it is certainly true that such a goal would be consistent with the maintenance of some reserve capacity. The question is how much slack is the firm likely to hold given the costs involved relative to expected demand in a world of uncertainty? If the firm were pursuing the principle of normal-cost pricing then it would be foolhardy for it to choose a price to the left of the minimum of the ATC curve, because of reasons discussed in section 3 of this chapter. On the other hand, if the firm’s pricing policies are based on a mark-up on a flat AVC curve then that precludes any need for costminimization. Thus contra the argument made by Lavoie (1996b) mark-up and full-cost pricing have very different implications for the analyses of competition and the optimal rate of capacity utilization. Furthermore, in the above passage Lavoie is quite correct in making the analogy between desired capacity and desired cash reserves. However, there is a confusion here between what people desire to hold as an ex ante decision and what they actually do hold ex post. If this distinction is not

40

Strategic competition, dynamics, and the role of the State

made, one would have to conclude that any level of cash reserves, determined by the vagaries of the market economy, would also be desired by agents. This is equivalent to saying that people have no portfolio choices. The confusion carries over into the implications for capacity utilization. Again, no one would deny that firms hold reserve capacity as a deliberate competitive strategy so as to absorb unexpected demand fluctuations. This simply means that, in contrast to neoclassical models, there is no unique point along which costs are minimized but a band whose width has to be sufficient to accommodate such fluctuations. It does not follow, however, that capitalist firms will be willing to tolerate any arbitrary amount of idle capital equipment lying around. They clearly do make a distinction between reserve and excess (redundant) capacity. Unless entrepreneurs anticipate correctly that aggregate demand will eventually pick up, there is no reason why in a world of Keynesian uncertainty firms would be willing, or even be able, to tolerate arbitrary amounts of redundant capital equipment forever in a highly competitive environment. And yet this appears to be the implication of the Kaleckian view.36 It is Lavoie’s intention to show that the dual paradoxes will hold even when the capacity utilization rate is constrained to be at its normal level. In order to do this he uses hysteresis effects in which ‘the long run or final value of a variable depends on the value of the variable in the past, by virtue of the influence of this past value on the current alleged exogenous variables’ (Setterfield, 1995, p. 14, cited from Lavoie, 1996a, p. 133). Let r* and u* be the normal rates of profit and capacity utilization, respectively, while r and u are the actual values of these two variables. Then any time that the actual rate of profit exceeds the normal rate firms raise the latter and conversely; if the actual rate of capacity utilization exceeds the normal rate firms raise the latter and conversely. This hysteresis effect is captured by the following two equations: r* 5 f(r 2 r*)

(3)

u* 5 s(u 2 u*)

(4)

where f and s are positive reaction coefficients. Basically, firms raise the mark-up rate so as to increase the normal rate of profit. Since aggregate demand raises the actual rate of profit via the actual capacity utilization rate, equation 3 shows that the normal rate of profit is itself a function of aggregate demand. Once this redefined notion of a normal rate of capacity utilization is used the question of cost-minimization does not arise. On the other hand, the traditional analysis of a normal rate of capacity utilization involves cost-minimization where key costs such as wages and shift

The microfoundations of long-run growth

41

premiums are institutionally determined (Kurz, 1986). It therefore follows that in the Kaleckian view capacity utilization is an entirely ‘free’ variable that is not tethered to any cost-related factors, many of which are shaped by institutional factors. The same criticism can be made against equation 3. When the firm increases its mark-up rate it can do so provided it is not threatened by other firms. But insulation from competitive forces is by its nature of a short-lived kind, since a firm confronts competitive pressures both from firms outside the industry and from existing firms within the industry (Harrod, 1952, p.144). This argument, one of the central conclusions of the OERG study, implies that when a firm raises its price it makes itself vulnerable to attack by other firms which can make inroads into its own markets. It might of course be objected that this kind of behavior is ruled out under imperfect competition; however, it is this very notion of competition that is challenged in section 3 of this chapter. Skott’s (1989) approach to the persistence of excess capacity is somewhat different.37 His point of departure is also imperfect competition with price-setting behavior and a downward-sloping demand curve faced by each firm. The rationale for the last feature of the model is that firms take into account competitive pressures when setting their own prices. However, Skott takes the position that in the long run firms will not tolerate large amounts of excess capacity or deficient capacity and investment will adjust accordingly to bring the capacity utilization rate in line with some desired rate u*. What determines u*? The key to Skott’s model is that whenever pure profits are positive (presumably during a boom) firms will build up excess capacity as a deterrent against potential new entrants into the sector. The existence of positive profits would raise the actual rate of capacity utilization and therefore provide a signal for new firms to enter the industry. Therefore, existing firms within the industry would lower their desired rates of capacity utilization u*, thus building up excess capacity. The deterrent effect would arise since such firms would be in a position to expand production and cut prices. Whenever pure profits are zero, the industry is in equilibrium in that there is no entrance or exit of firms. Skott’s model suffers from two problems. First, it is not obvious why the presence of excess capacity would be a deterrent to competition if a potential entrant selling in a different market can alter its production structure and enter with lower unit costs of production (Brunner, 1975b). Second, Skott’s model does not allow for the exit of firms if profits are negative. However, as argued above with regard to Lavoie’s model, if such a possibility exists then the scrapping of capacity within the industry would tend to eliminate excess capacity.

42

Strategic competition, dynamics, and the role of the State

Finally, some Post Keynesian authors such as Nell38 have argued that excess capacity is persistent because firms ‘mothball’ some plant and equipment during economic slowdowns. The rationale, according to him, is that an improvement of economic conditions would allow them to bring this machinery back into operation without losing market share. There are three problems with this argument. First, equipment is indeed ‘“mothballed” or held in reserve to meet unexpected surges in demand’.39 What is clearly at stake here is the state of business expectations. If firms’ ‘animal spirits’ decrease as a consequence of deepening economic uncertainty, mothballing becomes less of an option. On the other hand, firms are likely to mothball only if they expect the slowdown to be of a very temporary nature. Second, as McCombie and Thirlwall (1994, p. 132) argue, capital equipment is not retired because it ‘literally falls to bits’ but because it has become economically unviable. Such a situation typically arises with a piece of capital equipment when the expected profit stream from it is zero. Economic obsolescence of a piece of machinery arises from competition from newer and technically more advanced machinery and the fact that maintenance costs of older machines tend to rise. Moreover, the fall in the profit rate may be exacerbated if competition forces the price of the final output to fall. In sum, both the state of demand and the forces of competition will influence firms’ decision to maintain or discard capital equipment. Finally, there is an empirical issue. In her empirical work on capital discards,40 Powers (1988) showed that at both the two-digit SIC and the aggregate levels the US manufacturing sector’s discards decrease during business cycle expansions and increase during contractions. These important findings contradict the ‘mothballing’ argument in particular and, more generally, the persistent excess capacity thesis.

5.

CONCLUSION

By their very nature, oligopolistic market structures are said to exist because of the persistence of entry barriers. However, in this chapter an important though controversial claim is that, despite Andrews’s view that contemporary capitalism consists of what he calls competitive oligopolies (Lee, 1989), his and Brunner’s analyses of entry barriers undermine the theoretical arguments for the existence of oligopolistic markets. Brunner’s (1975b) article ‘Industrial Analysis Revisited’ in Studies in Pricing is particularly significant because of its critique of core aspects of Bain’s (1956) oligopoly theory (Tirole, 1988). An implication of this critique of entry barriers is the ebb and flow of capitals across different sectors, thereby

The microfoundations of long-run growth

43

leading to the approximate equalization of profit rates across these sectors, as argued by classical political economists. A further implication for the analysis of growth from the standpoint of both the OERG and classical political economy is that the pressure of competition forces firms to attain the normal rate of capacity utilization which is the only practicable optimum situation for a given firm. With long-run prices being determined by the conditions of production of the most efficient producers and barriers to entry susceptible to being scaled, the analytical framework of the OERG appears to provide a modern justification for the classical prices of production. It also leads to a theoretical justification for the empirical findings of several authors who do not find persistent profit rate differentials across industries.41 The key drive to minimize selling prices arises from the potential threat of invasion by firms outside an industry, a potential that can become an actual threat if such firms succeed in developing new technologies. In this connection normal-cost pricing is a strategic pricing policy by a firm in an attempt to remain a viable entity in an economic system enshrouded in uncertainty in which potential lower-cost firms could unexpectedly threaten its market share. This strategic competition under uncertainty, at the core of the OERG perspective, is very different from both perfect and imperfect competition. Once it is recognized that the capitalist firm is not an isolated Robinson Crusoe but is an institution that is socially embedded, it follows that the technological learning capacity will be shaped both by the firm’s own internal organization and by the level of development of the larger society, as that is going to determine whether or not it gets crucial inputs such as skilled labor or finance. Thus its ability to break into new industries will be facilitated by these institutional developments. In this sense it is unclear how Post Keynesian authors who emphasize the importance of institutions in shaping capitalist firms nonetheless claim that entry barriers are persistent. The important rediscovery by Hall and Hitch (1939) and Andrews (1949a, 1949b) of the principle of normal-cost pricing was met with a lot of resistance by the economics profession. The general view was to treat normal-cost pricing as being equivalent to mark-up pricing under monopolistic competition with marginal revenue equal to marginal cost (Lee and Irving-Lessmann, 1992). This tended to obfuscate the pathbreaking contribution to pricing theory made by Hall and Hitch as well as Andrews. As Lee and Irving-Lessmann observe, Andrews’s unwillingness to respond to his critics and his own professional difficulties in advocating a distinctly non-neoclassical pricing theory contributed to the conflation of monopoly mark-up with normal-cost pricing.

44

Strategic competition, dynamics, and the role of the State

It is quite interesting to note that the relationship between normalcost pricing and mark-up pricing was the subject of a relatively recent debate between Post Keynesian authors (Downward and Reynolds, 1996; Downward, Lavoie, and Reynolds, 1996; Lavoie, 1996b; Lee, 1996). The core of the dispute was Lee’s contention that normal-cost pricing represents a pricing policy that is distinctly different from mark-up pricing. The opposing view, articulated by Lavoie, Downward and Reynolds (henceforth LDR), was that there is no substantial difference between the two pricing policies since they both entail a mark-up on some measure of unit costs (Downward and Reynolds, 1996). LDR’s perspective is problematic for several reasons. First, if firm A’s price is based on a mark-up on its constant marginal cost and there is persistent excess capacity then that implies a level of output which will be on the downward-sloping part of the ATC curve. However, if firm B sets a price on the basis of its normal-cost level of output, then for a given mark-up A will have lower unit profits than B, which situates itself to its right (because A will have a higher AFC and thus a higher ATC). This in turn implies that B will have more internal funds to finance future investments.42 Second, as discussed in section 4 of this chapter, Kaleckian authors such as Lavoie have redefined the notion of a normal rate of capacity utilization. The traditional notion both in classical political economy and in the OERG view is that the normal rate of capacity utilization corresponds to its minimum-cost level of output. On other hand, as discussed above, in Lavoie the normal rate of capacity utilization is determined by demand, rising when the latter rises and falling and when the latter falls. This in turn implies that the firm would choose a level of output at any arbitrary point on the ATC curve, as that would be determined by demand. Costminimization plays no role here contra the views of Andrews, Harrod and Brunner. Third, it is true that, following a long line of authors in the Kaleckian tradition, Lavoie (1996b) is not proposing a naïve attained-and-held mark-up pricing policy, since he is very aware of the threat of competition from lower-priced firms. In fact, in order to highlight the apparent congruence between Kalecki and the OERG he cites Brunner (1952a, 1952b), who argues that any business, ‘in order to continue to be competitive with any product, must keep within a certain limit of price’ (Brunner, 1952b, p. 733), since otherwise a firm that has higher-than-normal costs ‘will not be able to get a price which will yield it normal net profits’ (Brunner, 1975a, p. 31). The question is, if Lavoie concedes the threat of lower-cost firms, which firms is he referring to? He cannot be referring to potential new entrants,

The microfoundations of long-run growth

45

since their threat is ruled out in the Kaleckian framework by the existence of persistent entry barriers. On the other hand, if the threat comes primarily from firms within the industry then that raises a problem for the Kaleckian perspective. Once allowance is made for different production conditions within an industry as well as their evolution through time, there is nothing to stop a given firm from eliminating its excess capacity in order to minimize production costs. But this would force all other firms to do likewise – thereby eliminating excess capacity. Given the distinction between excess and reserve capacity made in the strategic competition perspective, fixed investment by firms has to be modeled in terms of the deviation of the actual rate of capacity utilization (u) from the normal (or desired) rate (u*). This implies that a Harrodian investment function would make the capital accumulation rate a positive function of u − u*. Such a formulation ensures that firms add to capacity when u – u* . 0 and discard capacity when u – u* , 0. As argued above, this mechanism follows from the fact that firms seek to attain the normal (minimum-cost) rate of capacity utilization because of competitive pressures. On the other hand, Post Keynesian models (see for example Godley and Lavoie, 2007) typically make investment a positive function of the capacity utilization rate only, as firms appear to have no target (costminimization) rate. Given their perspective on competition, firms are under no pressure to expand market shares by cutting prices and costs. This in turn implies no pressure to achieve the practicable optimum rate of capacity utilization. A troubling implication of this form of the investment function is that, given a prior slump which causes excess capacity to rise, an increase in demand would stimulate fixed investment – so that firms would add to their capacity despite having excess capacity to begin with. Radically different long-run consequences follow by specifying the investment function in these two ways. If output is persistently below capacity then an increase in the propensity to consume (either private or public) and/or an increase in the wage share will increase effective demand, thereby raising the capacity utilization rate, investment and growth. This is why both the paradox of thrift and the paradox of costs hold in the long run in Post Keynesian models. On the other hand, in the extended Harrodian view, with output at capacity in the long run there will be a slowdown in the long-run growth rate if the consumption propensity or wage share rises (Garegnani, 1983, p. 23). From this, does it follow that, with capacity utilization rate at the normal level, government spending has to be cut as a matter of necessity in order to raise long-run growth? As discussed in Chapter 5, the answer is in the negative, as activist State policies of very different types

46

Strategic competition, dynamics, and the role of the State

suggest themselves once allowance is made for long-run growth at normal capacity when there is unemployment.

NOTES 1.

Andrews (1949a) uses the terms direct and indirect costs of production. Direct production costs involve direct labor (which consists of production workers who transform raw materials into finished products) and direct materials (which are used by direct labor to make the finished products). On the other hand, indirect production costs ‘include all workers not involved in the actual processes, the office staffs and all grades of management from the lowest grades of supervisors upwards, the machinery and equipment used in production, the premises in which production is carried out, and the land on which the premises stand’ (ibid., p. 86). We will use the more familiar terms ‘average variable cost’ and ‘average fixed cost’ to demarcate these two types of inputs, though it must be understood that in the current context these terms are used in the Andrewsian sense. Paying-out costs involve all payments that have to be met currently and include direct costs and paying-out indirect costs. Paying-out costs generate the continuous need for cash in order for the business to survive (Andrews, 1993, p. 210). 2. Andrews (1949a, p. 261) also mentions other factors such as increased maintenance and repair costs at higher rates of capacity utilization. Higher rates of capacity utilization, including nighttime or weekend operations, may raise other input costs also (Winston, 1974; Kurz and Salvadori, 1995). 3. Note that if labor productivity were, say, to rise during each shift the AVC curve will fall too. Other things being equal this would make the ATC curve fall more during each shift. 4. ‘What, then, was the effect of “competition”? In the main it seemed to be to induce firms to modify the margins for profits which could be added to direct costs and overheads so that approximately the same prices for similar products would rule within the “group” of competing producers. One common procedure was the setting of a price by a strong firm at its own full cost level, and the acceptance of this price by other firms in the “group”; another was the reaching of a price by what was in effect an agreement, though an unconscious one, in which all the firms in the group, acting on the same principle of “full cost”, sought independently to reach a similar result’ (Hall and Hitch, 1939, p. 19, emphasis added). 5. Marcuzzo (1996) discusses R.F. Kahn’s ‘discovery’ of the so-called inverted-L (L) cost curve. Marcuzzo however conflates average costs with prime costs where the former generally refer to the average total costs. On the other hand R.F. Kahn explicitly refers to prime costs: ‘It has already been indicated that a very common method of reducing output, notably in cotton-spinning and coal-mining, is to close down the whole plant some days and to work the whole plant for a full shift on other days. . . . When the maximum possible number of days is being worked each week, output cannot be increased further. The output is now equal to what may be called capacity output. At this point the prime cost curves [AVC, MC] which have hitherto been a horizontal straight line, move up vertically. The result is thus obtained that the prime cost curve (marginal or average) has the shape of a laterally inverted L . . . described as an L-shaped curve’ (Kahn, 1989, pp. 57 and 59, cited from Miller, 2000, p. 125). 6. Eiteman (1948, p. 903) observes that the payment of overtime will cause the ATC curve to shift up to a higher level without changing its shape. This situation is represented by Figure 2.2. 7. It must be noted though that it was Andrews who wrote about normal-cost pricing,

The microfoundations of long-run growth

8. 9.

10. 11.

12.

13.

14.

15.

16.

47

whereas Hall and Hitch referred to full-cost pricing. Normal-cost pricing corresponds to the minimum of the ATC curve. In his interpretation of Marshall, Andrews (1952, p. 182) argues ‘that the essential characteristic of Marshallian competition was not an infinite elasticity of demand for the individual firm but the rule of uniformity of price for identical products’. ‘. . . barriers to entry are normally low, far lower than it is customary to regard them as being. We have too narrow a concept of entry. New entry may come not only from new firms, which it is too easy to see as small and weak and thus uninfluential, but from existing firms in other industries who are looking for profitable outlets, or are anxious for diversification for its own sake (and in these cases barriers from economies of scale and from finance will be relatively unimportant). Such entry competition also comes from what Andrews has called cross-entry, from firms in the same industry, breaking across sub-markets’ (Brunner, 1975a, p. 23). One can for example think of Toyota, which went from the production of textiles to textile machinery to automotive parts and then finally to trucks and passenger cars all in a matter of a few decades (Mass and Robertson, 1996). Andrews went on to emphasize that even the producer of a brand new product, possibly protected by a patent, faces limits to its pricing policies by rivals because ‘[T]oo high a margin will give an incentive for other business men to get around the patent. . . . If the product is not patented, or on the expiry of the patent, the product will be considered by other business men and some of them may come in, probably cutting away a bit of the original producers’ market. To protect his position he will have to lower his price, i.e. granted that the specification of the article does not change he will have to reduce his gross profit-margin’ (Andrews, 1949b, p. 86). Edwards (1955, p. 111) is careful to distinguish between agriculture and industry, where in the former sector spatial location may confer certain advantages in production that cannot be replicated, whereas in industry locational factors in general do not adversely affect the ability of new entrants to replicate best-practice techniques. This is essentially the same argument made by Marx on the question of land rent in Capital, vol. III (Bina, 1985, 1989a). ‘. . . the most efficient firm will set the price (and the inefficient may hang on if they can); because no firm will tend (or continue) to quote lower than this firm, while it will itself remain subject to the threat of similarly efficient competition from without’ (Edwards, 1955, p. 114). This observation of Andrews and Brunner is reminiscent of Andrew Carnegie, who admitted that: ‘[I]n enormous establishments with five or ten millions of dollars of capital invested and with thousands of workers, it costs the manufacturer much less to run at a loss per ton or per yard than to check his production. . . . Twenty sources of expense are fixed charges, many of which stoppage would only increase. Therefore, the article is produced for months and, in some cases that I have known, for years, not only without profit or without interest on capital, but to the impairment of capital invested’ (cited from Botwinick, 1993, pp. 144–5). It is quite striking how the Andrews and Brunner argument mirrors that regarding barriers to exit made by authors such as Shaikh (1982), Semmler (1984), Bina (1985), Botwinick (1993) and others who base their model of competition on Marx. Numerous examples may be cited to illustrate this dynamic, for example sugar and sugar substitutes or meat and soy-based substitutes. In each such case the new competitor may initially have a low market share because the product it offers is not satisfactory enough (for example, it does not taste good). The market share could expand if the quality improves. It is remarkable to see the parallels in business history when one compares the initial dominance and decline of the British textile industry with the initial dominance and subsequent decline of the US automobile sector. In both cases the long wavelike growth and decline was due to the entry of smaller and more technologically competitive firms that increased their market shares incrementally while the older firms were stuck with

48

17.

18.

19. 20.

21. 22. 23.

24.

25.

26.

Strategic competition, dynamics, and the role of the State relatively antiquated technologies. See Lazonick (1981) for a discussion of the textile industry. Schwartz (2009) observes that Chrysler engineers’ attempts to learn from Fiat’s more sophisticated technology ‘is akin to an aging heavyweight boxer stepping into a gym where more agile bantamweight fighters train’. ‘Mr. Andrews has stressed the point that the “new entrant” may often, indeed usually, be an existing firm which is induced to take on a line of production hitherto new to it. Most firms produce a number of products. It is comparatively easy for an established firm, with its permanent cadre of management in existence, its buying and selling organisation and attachment of skilled and unskilled labour, to switch part of its organisation, which may only be a small part, to producing an article not before produced by it, and to do so on a scale quite adequate to secure the necessary cheapness of production’ (Harrod, 1952, p. 144). ‘The outstanding fact is the extreme precariousness of the basis of knowledge on which our estimates of prospective yield have to be made. Our knowledge of the factors which will govern the yield of an investment some years hence is usually very slight and often negligible. If we speak frankly, we have to admit that our basis of knowledge for estimating the yield ten years hence of a railway, a copper mine, a textile factory, the goodwill of a patent medicine, an Atlantic liner, a building in the City of London amounts to little and sometimes to nothing; or even five years hence. . . . We should not conclude from this that everything depends on waves of irrational psychology. On the contrary, the state of long-term expectation is often steady, and even when it is not, the other factors exert their compensating effects. We are merely reminding ourselves that human decisions affecting the future, whether personal or political or economic, cannot depend on strict mathematical expectation, since the basis for making such calculations does not exist’ (Keynes, 1953, pp. 149–50 and 162–3, emphasis added). See Harrod (1939a) in which he argues that the MR 5 MC policy is inconsistent with a long-period analysis with systemic uncertainty. ‘If the entrepreneur foresees the trend of events, which will in due course limit his profitable output to x − y units, why not plan to have a plant that will produce x − y units most cheaply, rather than encumber himself with excess capacity? To plan a plant for producing x units, while knowing that it will only be possible to maintain an output of x − y units, is surely to suffer from schizophrenia’ (Harrod, 1952, p. 149). See Shaikh (1989) and Chapter 5. May one even suggest that capitalist competition under Keynesian uncertainty forces individual entrepreneurs to be paranoid about the future? ‘The Marxist notion of competition defines a process, not a state. It describes an antagonistic and destructive process, not an equilibrium fantasy. For competition among capitals, it describes a war. To extend the analogy a bit further, the movement of capital from one industry to another corresponds to the determination of the terrain (site) of battle; the development and adoption of the weapons of war (the arms race); and the competition of one firm against another corresponds to the battle itself. . . . In all of this there can never be any guarantee for any individual capital that it will earn any profit at all, let alone the social average rate of profit’ (Shaikh, 1982, p. 77, emphasis added). Kalecki (1971b) modifies his earlier pricing model. In this later version the mark-up of price p on direct or prime costs u [(p − u)/u] is a function of the rivalry between oligopolists: [(p − u)/u] 5 f(p/p). This equation becomes p 5 u[1 1 f(p/p)], which can be shown to be related to equation 1 (Kriesler, 1988, p. 126, footnote 15). In this equation ∂p/∂f . 0 so that the higher is the average price of rival firms, the higher can the individual firm afford to set a high mark-up rate. Each player seeks to achieve the best strategic response, holding the responses of the opponent constant, and the resultant effect of this goal is the attainment of a Nash equilibrium, the notion of equilibrium most often employed in game theory. Thus in a Nash equilibrium each side in a two-person game ends up with the optimal choice. A payoff is defined as the net gain or loss an agent faces as a consequence of the actions of rivals. The payoff matrix represents all the possible combinations of payoffs accruing

The microfoundations of long-run growth

27.

28.

29.

30.

31.

32. 33.

34.

35.

49

to all the players. The key problem facing each player is to decide which strategy to choose, given that it does not know what other players will do, although it knows of all the possible payoffs. ‘. . . it is an assumption that players have the ability and time to collect and process any amount of information that we want them to, and that they have the skills to perceive all future contingencies’ (Dockner et al., 2000, p. 12). See also the excellent critical analysis of game theory by Hargreaves-Heap and Varoufakis (1995). Matters do not change dramatically when an attempt is made to dynamize game theory. The problem, as Foss (2000, p. 47) points out, is that if in standard game theory agents are equipped with hyper-rationality then in evolutionary game theory ‘one goes to the other extreme and portrays agents as following rigid rules, even if these turn out to be completely irrational’. Foss concludes: ‘The critical point here is both approaches – the standard and evolutionary game theory approaches – essentially imply that the entrepreneurial process of discovery becomes suppressed; in the standard approach there is no need for it, because, roughly, agents already know all that is worth discovering, and in the evolutionary approach, they are too stupid to discover anything anyway.’ ‘. . . virtually all detailed empirical inquiry into major technological advance has highlighted the inability of the actors involved to foresee the path of development, even in broad outline, or the major surprises that often occurred along the path. In contrast, the new models [i.e. neoclassical endogenous growth theory] assume perfect foresight or, if they admit less than that, that uncertainty about the future can be treated in terms of a well and correctly specified probability distribution of possible future events’ (Nelson, 1997, pp. 31–2). Solow (1985, p. 328) made a similar point when he argued that the economy is not a timeless system but is profoundly and erratically changed by concrete historical circumstances: ‘To express the matter more formally, much of what we observe cannot be treated as the realization of a stationary stochastic process without straining credulity. Moreover, all narrowly economic activity is embedded in a web of social institutions, customs, beliefs, and attitudes. Concrete outcomes are indubitably affected by these background factors, some of which may slowly and gradually, others erratically.’ Davidson (1982, p. 187) provides the following colorful description of the stationary vs non-stationary distinction: ‘If the process is stationary, whether Galileo dropped the stones from the Leaning Tower of Pisa in the sixteenth or twentieth century should not affect the rate at which the stones fall. If the process is nonstationary, the date at which the stones are dropped will affect their rate of fall.’ See the New York Times articles (Dickson, 1992; Holusha, 1994; Maynard, 2006) which describe the actual competitive pressures faced by businesses of all sizes. As these writers report, price- and cost-cutting are essential to the dynamics of competition. ‘It has been said that competition levels the rates of profit of the different spheres of production into an average rate of profit. . . . This occurs through the continual transfer of capital from one sphere to another, in which, for the moment, the profit happens to lie above average. . . .Yet with respect to each sphere of actual production – industry, agriculture, mining, etc. – the transfer of capital from one sector to another offers considerable difficulties, particularly on account of the existing fixed capital. Experience shows, moreover, that if a branch of industry, such as, say, the cotton industry, yields unusually high profits at one period, it makes very little profit, or even suffers losses, at another, so that in a certain cycle of years the average profit is much the same as in other branches’ (Marx, [1894] 1967, p. 208). With regard to the paradox of thrift, a lowering of the savings rate will raise long-run growth. Once the State is included so that the social savings rate equals the private savings rate plus the government budget surplus/GDP, then any reduction of the budget surplus/GDP (or increase in the budget deficit/GDP ratio) will also raise the long-run growth rate. Finally, according to the paradox of costs any increase in costs such as wage rates will also raise the long-run growth rate. Let I 5 investment, S 5 savings, Y 5 output, Y* 5 capacity, u 5 Y/Y*, s 5 savings

50

36.

37. 38. 39. 40. 41. 42.

Strategic competition, dynamics, and the role of the State rate 5 S/Y, K 5 capital stock, and v 5 capital–capacity output 5 K/ Y*. Then beginning with I 5 S, I/K 5 GK 5 sY/K 5 su/v. Thus u 5 GKv/s. If the rate of capital accumulation is exogenous and v is constant, then u can take on any arbitrary value via the variations in s. On the other hand, if a standard Post Keynesian investment function is deployed (Godley and Lavoie, 2007) then, depending on the parameters of this function, a wide range of positive values of u can be obtained each of which would be acceptable by firms. The argument made here is a general one with respect to any kind of stockholding. Consider for example inventory stocks held by firms. Now firms clearly make a distinction between desired and actual inventory/sales ratios, with discrepancies showing up as changes in production plans. It is of interest to note that Flaschel and Skott (2006, p. 318, footnote 13) find Lavoie’s attempt to endogenize the normal rate of capacity utilization ‘unconvincing, Dutt’s (1997) attempt to provide a rationale for the adjustment process notwithstanding’. Discussion in the session on ‘Economic Policies in Latin America and the Caribbean: Critical and Alternative Perspectives’ at the Eastern Economic Association Annual Conference, Philadelphia, 2006. See www.oecd.org/dataoecd/13/37/2552593.doc, p. 31. Equipment is said to be discarded when it is removed from useful service. See Shaikh (2008) and the Appendix to the chapter. The relationship between a firm’s pricing policy, retained earnings and investment has been discussed by Post Keynesian authors such as Eichner (1980).

The microfoundations of long-run growth

51

APPENDIX Experimental work to test game theory has not produced results that conclusively prove the basic theoretical predictions of this framework (Hargreaves-Heap and Varoufakis, 1995, Chapter 8; Guala, 2006). More generally, as both Semmler (1984) and Botwinick (1993) carefully survey, there is very little empirical evidence to support the view that large-sized firms have monopoly power and can maintain entry barriers and persistently higher rates of profit. The classic US studies of inter-industrial profit rate differentials took place in the 1930s to the 1960s. These initial studies by Bain (1951), Stigler (1963), Weiss (1963) and Mann (1966) seemed to confirm the standard view that high levels of market concentration caused by high monopoly power produce high profit rates because of collusion between firms. However, as Semmler (1984, Chapter 4) argues, these studies had severe methodological problems, including the fact that observation periods were very small and critical concentration ratios were arbitrarily chosen. A study such as Bain (1951) which did show a significant positive relationship between concentration ratios and profit rates had extremely low correlation coefficients. One major problem with these studies is that they suffered from what one might call perspectivist illusion. That is, the studies were based on relatively small time slices when certain large-scale industrial sectors had above-average profit rates. However, in a number of studies, Brozen (1971a, 1971b, 1973) revised the early studies by Bain, Mann and Stigler by increasing the time period and found that the profit rates of these industries tend to converge to the average rate. On the basis of the empirical literature on differential profit rates Semmler (1984, pp. 127–8) draws the following conclusions: First, there does not seem to be overwhelming evidence that industrial concentration by itself leads to persistence of higher profit rates. Entry barriers seem to be a necessary condition for profit rate differentials. Second, entry barriers can turn into exit barriers, leading to profit rates for industries and firms below the average. . . . Third there are few studies which reveal unequivocally that firm size is the dominant variable for interfirm profitability. Higher profitability corresponding to firm size and larger market share in product lines may be the result of market power or of economies of scale and cost advantages, yet recent studies have shown that economies of scale and cost advantages influence the profitability of firms more than industrial concentration.

With regard to intra-industry competition a number of authors (Shaikh, 1982; Semmler, 1984; Botwinick, 1993) have challenged the view that prices are determined by monopoly power. Shaikh (1998) uses US input–output

52

Strategic competition, dynamics, and the role of the State

data to show that prices are regulated by direct and indirect labor times. In fact, outside the fictional world of perfect competition a number of technology vintages are likely to coexist provided they are all profitable simply because not all firms have the financial capacity to adopt the latest technologies, which typically entail larger outlays of fixed capital per unit output. This in turn implies that, since market competition tends to make selling prices roughly equal, a snapshot of an industry at any point in time will reveal an array of different profit rates because each firm has a different cost structure. This profit rate differential thus arises from the nature of technical change and not monopoly power. The empirical evidence provided here against imperfect competition applies as much to mining as it does to manufacturing industry. The contributions of Cyrus Bina (Bina, 1989a, 1989b; Bina and Vo, 2007) are particularly noteworthy, as he has shown that both the oil sector by itself and the aggregate fossil fuels sector (oil, coal and natural gas) are subject to highly competitive forces (in the classical/Marxian and strategic competition view). Finally, as Lee (1994) points out by citing a number of studies (Edwards, 1989; Fleischman and Parker, 1991), price-setting strategies have been used by firms since the time of Adam Smith. There is historical evidence that normal-cost pricing was practiced by firms by the middle of the nineteenth century (Lee and Irving-Lessman, 1992). In fact Adam Smith’s famous pin factory is an excellent illustration of firms in early capitalism seeking actively to raise productivity and mechanize the production process. This kind of behavior is inconsistent with the passive role of the capitalist-entrepreneur in neoclassical theory. Thus it is historically inaccurate to argue, as do many neoclassical and heterodox economists, that while late capitalism is characterized by monopolistic competition early capitalism was closer to the perfectly competitive model.

3. 1.

A review of the literature on growth INTRODUCTION

In this chapter both neoclassical and heterodox models of economic growth will be reviewed. We will begin with two variants of the neoclassical approach, the older Solow growth model and the new generation of neoclassical endogenous growth theory (NEGT)1 models. We will then proceed to evaluate Post Keynesian growth models which follow Kalecki in assuming some variant of imperfect competition. We will finally discuss Harrod’s model and a contemporary solution to his instability problem which constitutes the basis of the growth framework of this book.

2.

NEOCLASSICAL GROWTH MODELS

(a)

Solow Growth Model

Before the new contributions of P. Romer (1986) and Lucas (1988) the Solow model was the workhorse of neoclassical theorizing on the growth process, although it still remains very influential. The Solow model is based on the following production function: Y 5 F(K, HL)

(1)

where Y 5 output, K 5 capital input, L 5 labor input, and H is knowledge or the labor effectiveness. The following are the key assumptions of the Solow model (Snowdon and Vane, 2005): a.

The production function exhibits constant returns to scale so that multiplying the inputs by say x will also raise output by x: xY 5 F(xK, xL). This ensures that the production function can be written in intensive form y 5 F(k) where y 5 Y/L and k 5 K/L. b. The production function (1) exhibits diminishing returns to scale for all values of K and L (∂F/∂K . 0, ∂F/∂L . 0, ∂2F/∂K2 , 0, and ∂2F/∂L2 , 0). In y versus k space the partial derivative ∂y/∂k is the 53

54

Strategic competition, dynamics, and the role of the State

marginal product of capital (MPK). Thus, as the capital–labor ratio k rises, MPK rises at a diminishing rate so that in the limit as k S ` MPK S 0 and as k S 0 MPK S `. These are known as the Inada conditions. c. Full price flexibility ensures long-run equilibrium growth at full capacity and full employment. d. Wages, capital–labor and output–capital ratios are variable. e. Ex ante savings equal ex ante investment. The above assumptions imply that the economy is continuously at its full employment and full capacity level of output. Furthermore, it does not allow for disequilibrium dynamics. Let the Cobb–Douglas production function be Y 5 Ka(HL)1−a where 0 , a , #1 (Agénor, 2004). Let labor L and knowledge H grow at exogenous # rates (L/L 5 n and H/H 5 l), the depreciation rate 5 d, k 5 K/HL and y 5 Y/HL. Then it can be shown that2 y 5 f (k) 5 ka

(2)

Since investment equals savings, which is a constant fraction s of national income Y, the change in the capital stock (or net investment) is given by: # K 5 sY 2 dK

(3)

where d 5 depreciation rate. Differentiating k 5 K/HL, substituting # # L/L 5 n and H/H 5 l, and using equation 3 we get the rate of capital accumulation: # k 5 sy 2 (d 1 l 1 n) k

(4)

# In the steady state k 5 0, k 5 k, and y 5 y. Thus: sy 5 (d 1 l 1 n) k

(5)

Substituting this expression in equation 2 we get: 1

s k5a b 1 2a d1l1n

(6)

Since k 5 K/HL and y 5 Y/HL it follows that# in the # steady state both K and Y grow at the exogenous rate (l 1 n): K/K 5 Y/Y 5 (l 1 n) . In the Solow model flexible wages and the choice of technique ensure full employment (Foley and Michl, 1999, p. 139). Put simply, flexible wages

A review of the literature on growth

55

ensure convergence to a growth rate of Y equal to the population growth and the rate of technological progress, that is Harrod’s natural growth rate. A higher growth rate will be achieved only if either population growth or the rate of technological change increases. A higher savings rate will raise the level of output (or capital) per capita while temporarily raising the growth rate as the system moves from one steady-state level of output to another. On the other hand, an increase in the savings rate (and thus the investment/output ratio) has no long-run effect on growth quite simply because the latter is determined by the exogenous parameters n and l (Foley and Michl, 1999, p. 149). (b)

Neoclassical Endogenous Growth Theory

In the Solow model the effectiveness of labor H is taken as exogenous. One goal of NEGT is to endogenize H and counter the tendency of accumulable factors of production (i.e. basically the capital stock) from experiencing diminishing returns to scale (Dutt, 2003, p. 73). The other is to introduce a positive role for the savings (and investment) rate in the growth process, a link that is missing in the Solow model. Much of what follows draws on the excellent critical view of the NEGT literature by Cesaratto (1999). The production of knowledge plays an important part in NEGT models (P. Romer, 1994). The roots of this approach can be traced to Arrow’s (1962) ‘learning-by-doing’ model in which he endogenizes the technology variable H in the Solow model as follows: H 5 Kb

(7)

where 0 , b , 1 for the growth to be non-explosive. Here knowledge is treated as a pure public good because capital accumulation by individual firms generates technological knowledge that benefits all firms and raises their productivity. Arrow assumes perfect competition. Unfortunately, if this equation is substituted in the production function Y 5 AKa(HL)1−a # the resulting steady-state growth is Y/Y 5 n/1 2 b (Cesaratto, 1999, p. 780) and is thus still determined by exogenous parameters. Traditionally, neoclassical growth models treated knowledge as nonrivalrous and non-excludable, that is a pure public good. However, P. Romer (1990) argues that knowledge is also purposefully produced by profit-seeking firms so that while it is non-rivalrous it is a partially excludable production input. Consider the production function Y 5 F(A, K, L) where K and L are rival inputs and A is a non-rival input. If K and L are

56

Strategic competition, dynamics, and the role of the State

increased by a factor l then Y will increase by this amount: lY 5 F(A, lK, lL). This follows from the constant returns to scale assumption. On the other hand, if the non-rival input A is also increased by l then in such a situation output will increase by more than l and the production function will not be concave (P. Romer 1990, p. S76). Thus: lY , F(lA, lK, lL)

(8)

This latter situation is one of increasing returns to scale. Now in a standard perfectly competitive equilibrium the entire output is used up when each factor of production (i.e. rival input) is paid its marginal product. This is given by the Euler equation F(K, L) 5

0F # 0F # K1 L 5 profits 1 wages 0K 0L

(9)

However, if the non-rival input A is a productive input and were paid its marginal product then the firm would be making a loss: F(A, K, L) ,

0F # 0F # 0F # A1 K1 L 0A 0K 0L

(10)

Thus, as P. Romer (1990) argues, in such a situation the standard perfect competition model needs to be abandoned. Instead the presence of purposefully produced knowledge requires the need for imperfect competition and increasing returns to scale. It is for these reasons that the Romer (1990) model, by building on Dixit and Stiglitz (1977), deviates from previous NEGT models in assuming monopolistic competition à la Chamberlin. The producers or owners of knowledge need to have market power and earn monopoly rents so as to make it worth their while to engage in such types of activities. Based on what he calls a neo-Schumpeterian framework,3 P. Romer (1990) models technological change which arises from the outputs of profit-making monopolistic firms. Growth is fueled by the production of different types of intermediate inputs which raise labor and capital productivity (Aghion and Howitt, 1998). The model consists of three sectors: a final output sector which is perfectly competitive, a capital (intermediate) goods sector which is monopolistically competitive, and a research sector that is also perfectly competitive. Final output Y is produced with the help of four types of inputs (capital, unskilled labor, human capital and technology) and is represented by a Cobb–Douglas production function which is homogeneous of degree one: Y 5 HYaLbAx1−a−b

(11)

A review of the literature on growth

57

where HY 5 human capital needed to produce the final output, L 5 labor, A 5 knowledge and x is the equilibrium value of intermediate inputs corresponding to their monopoly price, which is obtained via a mark-up on marginal costs. Total human capital HK and labor L are fixed, with part of human capital HY being used to produce final goods and another part HA employed to do research (HK 5 HY 1 HA). Technology A evolves according to the equation: # A 5 dHA (12) A where d is a productivity parameter and HA is the human capital employed in conducting research. The production of knowledge affects final output both by creating new designs and by increasing the total stock of knowledge, which increases human capital productivity in the research sector. As in mark-up pricing models, the monopolist’s profit maximization is obtained by setting marginal revenue equals marginal cost so that markup depends on the elasticity of the demand curve.4 Finally, in the steady state the growth rate of output equals the growth rate of the capital stock (so that the capital–output ratio is constant), which in turn equals the growth rate of technology: # # # Y K A 5 5 5 dHA (13) Y K A The steady-state growth rate g is: g5

dHK 2 Lr sL 1 1

(14)

where L 5 [1/(1 − a − b)][a/ a 1 b], s is a parameter in the utility function, and r is the rate of time preference. Thus a community’s preferences between current and future consumption (savings) affect the growth rate. In this equation the mark-up 5 [1/(1 − a − b)]. It must be recalled that the rationale for monopolistic competition in this model is that firms will be induced to produce new knowledge if they earn monopoly profits. This generation of knowledge will in turn bring about growth. NEGT authors claim that this link between monopoly profits, innovations and growth is a Schumpeterian aspect of their model.5 The problem with equation 14 is that the relationship between monopoly profits and growth is ambiguous at best. For a higher mark-up and thus higher monopoly profits (provided the monopolist still operates in the elastic portion of the demand curve), caused by a higher a, the parameter L will be higher but the growth rate will be lower. It is not clear how this

58

Strategic competition, dynamics, and the role of the State

result, which is popular in stagnationist models, can be rationalized on the basis of a framework that assumes Say’s law. On the other hand, if the mark-up is raised via an increase in b then this will increase the growth rate only if (a 1 b) . ½.6 It is unclear what the economic meaning is of these relationships between a and b, monopolist profits and the growth rate or how they relate to Schumpeter’s framework. The so-called AK models of growth (Rebelo, 1991; Barro and Sala-iMartin, 1995, Chapter 1; D. Romer, 1996, p. 104) aggregate physical and human capital into one overall measure of the capital stock. The goal of AK models is to eliminate all non-produced factors of production (such as labor) which could cause produced factors of production (such as capital) to encounter diminishing returns.7 Thus all production inputs are treated as reproducible whose accumulation depends on savings.8 The production function is linear so that: Y 5 AK

(15)

where A represents technology. This linear production function does not satisfy the Inada conditions. Written in growth form equation 15 becomes: # # Y K I I/Y s* 5 5 5 5 (16) Y K K K/Y C where s* 5 social savings rate and C 5 capital–capacity ratio. Hence a higher social savings rate raises output growth. As equation 16 is derived from the neoclassical AK model, Cesaratto (1999) calls the former a pseudo-Harrod–Domar equation. The similarity to the Harrod–Domar equation is however superficial. While Harrod (and Domar) followed Keynes in assuming the persistence of long-run involuntary unemployment, AK models essentially conjure away the role of labor in the production function. Thus the question of involuntary unemployment does not even arise in AK models. Finally, based on Uzawa (1965), another group of models endogenizes labor productivity by incorporating investment in education, R&D and so on, all of which are determined by private savings. The Lucas (1988) model is an extension of Uzawa. The model is based on the production function Y 5 Ka(uHL)1−a where u is the proportion of labor time used to produce ordinary goods (say corn) and H evolves according to the following law of motion: # H 5 Hxd [ 1 2 u ]

(17)

A review of the literature on growth

59

where (1 − u) is the share of labor time employed to produce efficiencyenhancing knowledge. Lucas assumes x 5 1 so that equation 17 becomes # H/H 5 d [ 1 2 u ] . That is, productivity grows exogenously as in the Solow model. The only additional twist provided by Lucas is that the term (1 − u) reflects society’s savings decision so that the balanced growth path is a function of both the rate of time preference (i.e. thriftiness) and the productivity parameter d. Despite their highly influential status, it is unclear what the relevance of NEGT is in the analysis of real-world problems or in policy formulation. Consider first the issue of stability. As P. Romer (1994) observes, all models of endogenous growth require the evolution of a technology # parameter, call it Z, to be modeled as Z 5 kZf (k . 0) so that # Z 5 kZf 21 (18) Z Non-explosive growth will be possible only if f 5 1. For example, in the Lucas and Romer models if the productivity parameters x and d, respectively, are greater than or less than 1 growth will be knife-edge unstable. This problem of NEGT has been noted by a number of authors (Cesaratto, 1999). Solow (1994) observes that, in their attempt to eliminate diminishing returns to capital, NEGT models run into the problem of knife-edge instability when there are increasing returns to capital. Only constant returns to capital yield stable growth. However, as Solow (ibid., p. 51) points out: This branch of the new growth theory seems unpromising to me on straight theoretical grounds. If it found support in empirical material, one would have to reconsider and perhaps try to find some convincing reason why Nature has no choice but to present us with constant returns to capital. On the whole, however, the empirical evidence appears to be less than not strong; if anything, it goes the other way.

Of course various ad hoc assumptions or mathematical tricks can always be used to contain the explosive nature of increasing returns to capital (P. Romer, 1986, 1994). It is unclear, however, what the economic meaning is of such mathematical formulations. The second issue deals with the question of the labor market. The labor market is assumed to be in continuous full employment or, as in the case of AK models, labor does not explicitly appear as a factor of production because physical and human capital are combined into an aggregate measure of the capital stock. Implicitly then the warranted and natural growth rates are in continuous equality. How can such a framework be used in a world afflicted by mass (involuntary) unemployment?

60

Strategic competition, dynamics, and the role of the State

Third, beginning with the capital controversy (Cohen and Harcourt, 2003), the production function methodology has been critiqued at a theoretical and empirical level by a large number of authors.9 One problem, as Felipe and Fisher (2003, p. 208) argue, ‘is that the conditions under which a well-behaved production function can be derived from micro production functions are so stringent that it is difficult to believe that actual economies satisfy them’. The further problem is that the production function does not represent any fundamental technological link between output, capital and labor but rather is the result of the algebraic manipulation of a national income identity in which net output equals wages plus profits. In fact, as Shaikh (1974) proves, the production function can be utilized statistically to model a bogus economy. On the basis of their considerable investigation of the properties of production functions, authors such as Felipe, Fisher, McCombie and Shaikh have shown that an aggregate production function has no theoretical or empirical meaning. Moreover, as Fisher summarized the issue, the conditions that need to be satisfied in order to derive an aggregate production function have no relevance to a real-world economy, which is precisely why he dismisses it as a ‘fairy tale’ (Fisher, 2005). Perhaps Robert Solow summarizes best the current status of the production function framework: The current state of play with respect to the estimation and use of aggregate production functions is best described as Determined Ambivalence. We all do it and we all do it with a bad conscience. . . . One or more aggregate production functions is an essential part of every complete macroeconometric model. . . . It seems inevitable. . . . There seems no practical alternative. . . . [Yet, n]obody thinks there is such a thing as a ‘true’ aggregate production function. Using an estimate of a relation that does not exist is bound to make one uncomfortable. (Solow, 1987, p. 15, cited from Shaikh, 2005, p. 447)

Nonetheless, practitioners of this framework claim that it ‘works’ at a practical level since it produces the ‘correct’ empirical results. It is also striking that neoclassical authors such as Greiner, Semmler and Gong (2004) who use the NEGT provide no rebuttal of the critical literature on production functions. What, one may ask, is one to make of their public policy analysis if their framework is based on a fiction? Fourth, the fictional quality of NEGT also includes its use of utility functions which, after all, are based on the axioms of choice. As is suggested by Keen (2001), unrealistic mathematical and behavioral assumptions are needed to construct utility functions and, once again, one may question the relevance of NEGT to the analysis of real-world problems, given these dubious microfoundations. The reader is also reminded of Sen’s (1977)

A review of the literature on growth

61

excellent critical analysis of the theory of the rational individual which is at the core of neoclassical theory. Fifth, at the core of NEGT models is the Robinson Crusoe-like utilitymaximizing representative agent ‘so that society’s economic choices are those of a single optimizer, whose tastes and opportunity sets are just microcosmic versions of those of the whole society’ (Haliassos and Tobin, 1990, p. 908). It is unclear how real-life societies with their deep inequalities of power and wealth can be modeled by this representative agent. Continuing further in their critique of the representative agent methodology, Haliassos and Tobin observe: ‘[I]t is hard to see why Robinson Crusoe has a government or why that projection of himself has different objectives from those of private citizen Crusoe’ (ibid., p. 908). Thus it is not clear why one needs to study the impact of government activities in such models. Moreover, the representative agent methodology implies that ‘[T]he problems of coordination emphasized by Keynes and other macroeconomists – between investors and savers, borrowers and lenders, capitalists and workers – are finessed’ (ibid., pp. 908–9). Quite simply, there is no room for endogenously generated disequilibria (or instability) in neoclassical models, a key difference between NEGT and the extended Harrodian framework (see Chapter 4). Sixth, NEGT models are based both on innovations and on rational expectations. However, since innovations involve the creation of something new, can one really make them determinate processes to which rational expectations apply (Setterfield, 1994)? After all, if all innovations are rationally expected, in what sense can they be considered to be new? In other words, how can one call them innovations? Seventh, a particularly problematic aspect of the NEGT framework arises from its underlying microfoundation, which is based on either perfect or monopolistically competitive firms. As discussed in Chapter 2, the only way that neoclassical theory attempts to explain real-world firm behavior is to make them operate in imperfectly competitive markets. If growth is driven by innovating and aggressive profit-seeking firms then that raises two questions. How does NEGT model technological change? How is profit-seeking behavior consistent with monopolistic competition? On the first question, Nelson (1997) has criticized NEGT for its attempt to model technological change in a general equilibrium context in which rational expectations are assumed. Following other historians of technology Nelson argues that technological change takes place fundamentally under uncertainty and so cannot be forecast either using perfect foresight or in terms of some stochastic process. This is not surprising since, following Moses Abramovitz, Nelson points out that technological change

62

Strategic competition, dynamics, and the role of the State

is shaped by a complex combination of economic, political and cultural factors as well as financial and educational institutions. Nelson argues that these broad factors shape the organization and management of firms in different national contexts and thus their competitive performances (Chandler, 1990; Lazonick, 1990). As argued in the previous chapter, one of the central findings of the OERG was that entrepreneurs take a long-period perspective in deciding on their price, given that their fixed costs last several years. If future competition is enshrouded in uncertainty because of the nature of technological change and competition then, as Harrod (1939a; 1952) argued, for the individual firm the MR 5 MC calculation is not realistic. At best the profit maximization calculation in Romer’s model (or indeed any other model whose basis is monopolistic competition) yields the short-run equilibrium price and output. If that is true how can it be consistent with a long-run growth model? Indeed if the forces shaping long-run competition are rife with uncertainty then the ability of the representative agent to optimize his or her utility comes into doubt. However, the representative agent framework underpinning these models creates an additional problem. Despite their vaunted claims regarding Schumpeterian microfoundations, it is unclear what the precise role of the firm is in NEGT models.10 There are two reasons for raising this question. For one thing, unlike heterodox growth models where business investment is explicitly tied to output, NEGT relies on production functions (with all their problems as discussed above) and the representative agent who maximizes his utility on the basis of consumption decisions rather than the pursuit of profits. On the other hand, models of the firm from classical political economy through Schumpeter and the Austrian school to the OERG have emphasized the profits–investment nexus as the differentia specifica of the capitalist firm. Furthermore, in neo-Schumpeterian models innovations are tied to a monopolistically competitive sector which has persistently above-normal profits (although individual firms within it may grow or die). Now what happens to innovations if the pressure of competition forces incumbent firms to lower selling prices and thereby eliminate the monopoly profits as the OERG authors argued? Does it follow that the rate of innovations will die down and firms will attenuate their aggressive and expansionary strategies? From the neoclassical perspective they would, because this limiting case is the perfectly competitive scenario which would cause them to become passive price-taking entities. While perhaps neoclassical authors would have no problems with perfect competition, one may question its relevance to the real world. As Andrews (1993, pp. 328–9) put it:

A review of the literature on growth

63

Business men in manufacture and distribution whose own thinking dwells on the continuous attempt to displace rivals forced on them under pain of themselves losing ground to competitors are often surprised if they happen to pick up economics text-books to find that in perfect competition, the hypothetical condition which is the quintessence of competition as the economist sees it, there is no mention of this, to them, major aspect of the competitive struggle. Reading on, they may be still more bewildered to discover analyses in which some of their chief competitive weapons do appear, but are then described as characteristics of ‘imperfect’ or ‘monopolistic’ competition!

On the other hand, as discussed in Chapter 2, in the OERG approach (as in classical political economy), the attainment of normal-cost pricing would by no means attenuate the competitive struggle.

3.

HETERODOX GROWTH MODELS

One important way of distinguishing modern heterodox growth models from neoclassical ones is the fact that the former reject both Say’s law and the proposition that flexible wages produce full employment. However, despite these core common features there are considerable differences among heterodox models, especially with regard to assumptions regarding the nature of the long run and the effect of the savings rate on growth. In classical political economy there were three broad approaches to the theory of growth. The underconsumptionist (Bleaney, 1976; Shaikh, 1978) school took the view that capitalism persistently suffers from excess capacity because of a lack of a sufficient degree of internally generated demand. The underconsumptionists claimed to have uncovered the following contradiction. Imagine a simple corn-based economy. From the annual gross output a portion (i.e. replacement investment) is deducted to be used to replace the corn seedlings used up so as to continue production at the same level for the next year. Once wages are deducted (in corn units) from the net output the surplus (profits) remaining accrues to the capitalist class. Now, reasoned the underconsumptionists, if capitalists save part of profits then that will constitute a reduction in aggregate demand. In this event only part of the surplus corn will be bought up, while the rest will be unsold, thereby leading to a demand gap. On the other hand, if capitalists were to consume the entire surplus then the demand gap would be zero, but there would be no investment. Of course, the surplus saved could be invested. However, for the underconsumptionists it was consumption demand that rules the roost but, since wages constitute only a part of net output, consumption demand would always be insufficient to buy up the entire output. Underconsumptionists of all political stripes came to this

64

Strategic competition, dynamics, and the role of the State

same fundamental conclusion regarding capitalism and suggested different mechanisms via which such a basically stagnant economic system could grow. All these mechanisms entailed the injection of demand from outside the pure private sector capitalist economy, thereby filling the demand gap. For example, Malthus’s position was that the landlord class had a positive economic role to play because of its high consumption demand, especially for luxury goods. Rosa Luxemburg argued that the additional demand would come from the periphery, which would pit capitalists from the center against each other, thereby eventually propelling European states to intervene militarily in the periphery. Thus capitalism’s stagnationist tendencies could only be resolved via imperialism and wars. If the underconsumptionists argued that aggregate supply would always outpace aggregate demand, J.-B. Say and later David Ricardo took the position that supply would always create its own demand, as the process of production necessarily involves the disbursement of expenditures in order to pay for circulating and fixed capital inputs. These expenditures are the incomes of the suppliers of these inputs and thus constitute the source of demand for the output. Thus, for Say and Ricardo, aggregate demand could never be different from aggregate supply. Since profit rate equalization entailed the attainment of a normal rate of capacity utilization, it follows that Say’s law made the expansion of capacity output generate the demand for it. Of course, in this original version of ‘supply-side’ economics this balance was not consistent with the full employment of labor (Mongiovi, 1990).11 Marx’s framework constituted the third approach to accumulation, in which both the underconsumptionism and Say’s law were rejected. In the schemes of reproduction, Marx showed that the expenditures of capitalist households and firms generated the consumption and investment demand needed to equal the output produced by firms (Shaikh, 1978). However, while in the schemes of reproduction Marx demonstrated numerically that a sufficient level of demand would be endogenously generated, he also argued elsewhere that in practice a number of factors would prevent the continuous equality between demand and supply. First, because production takes time and investment decisions made by firms are initiated on the basis of expected future sales, over- and under-shooting of aggregate demand and supply are likely to be the norm. Given the decentralized nature of market transactions (‘anarchy of capitalist production’) Marx was very aware of coordination problems between buyers and sellers12 so that the equalization of aggregate demand and supply would happen only in an approximate sense over some period of time.13 Second, since money is endogenous in Marx’s framework (Itoh and Lapavitsas, 1999), should firms or households choose to hold on to money

A review of the literature on growth

65

rather than spend it (hoarding motive) at the end of a production cycle there would be insufficient demand.14 Marx did more than merely show the existence of a sufficient degree of internally generated demand to meet the output. He also showed the existence of a growth path whose rate would increase if the capitalist class saved a greater proportion of the surplus value produced.15 In this framework, an increase in the proportion of non-production or social consumption activities (such as government consumption expenditures) would slow down growth (Shaikh and Tonak, 1994). Eltis (1979) also makes a very similar argument in a broader classical framework. If implicit in the writings of Marx and the classical authors such as the Physiocrats was some notion of endogenously generated cyclical growth, it was Harrod and Domar who first attempted to formalize such a growth model. A number of authors such as Robinson (1960), Blaug (1978) and Eltis (1998) have commented on the close parallels between the schemes of reproduction and Harrod–Domar’s growth model. Unlike Keynes or Kalecki, Harrod–Domar did not consider investment to be equivalent to exports or government spending, since they recognized that investment has a dual effect: it increases both demand and capacity. Because investment in the Keynesian system is merely an instrument for generating income, the system does not take into account the extremely essential, elementary and well-known fact that investment also increases capacity. This dual character of the investment process . . . provides us with both sides of the equation. (Domar 1946, reprinted in Sen, 1970, pp. 67–8)

Since both demand and capacity are endogenously generated, a central concern of Harrod–Domar was to identify the conditions that would bring these two variables into approximate equality along what Harrod called the warranted growth path. From the standpoint in particular of the modern history of growth theory, this insight of Harrod and Domar was crucial. First, it pointed the way to an endogenously generated growth path driven by investment in which the capacity utilization rate is at some normal level, that is at the minimum-cost level on the average total cost curve. This is a revival of the classical-Marxian view. It must be emphasized that this approximate equalization between demand and capacity does not imply the validity of Say’s law, since in the latter demand passively adjusts to capacity output, which under general equilibrium is consistent with the full employment level of output. On the other hand, in the Harrod–Domar perspective both demand and capacity vary because of the dual effect of investment. Second, the warranted growth path is consistent with varying degrees of unemployment, since there is no assumption that labor market ‘flexibility’

66

Strategic competition, dynamics, and the role of the State

will lead to full employment. Further, and equally significantly, Harrod in particular was concerned with modeling endogenously generated instability arising from the mutual interactions of capacity and demand. Thus in his modeling of long-run growth Harrod deviated both from the underconsumptionist and from the Say’s law perspectives. The problem, as Harrod found, is that this interaction between demand and capacity leads to a runaway instability process. For example, if there is excess capacity (demand less than capacity) firms reduce investment, which further reduces demand and so on. The opposite happens when demand exceeds capacity. The warranted growth path is thus seemingly impossible to attain.16 The knife-edge instability characteristic of the warranted growth path has more or less led to the disappearance of this Weltanschauung from the growth literature. It has generated a number of divergent trajectories. In the neoclassical perspective, Say’s law, rational expectations and labor market flexibility ensure a long-run growth path that is consistent with the full employment of labor and the full utilization of the capital stock. Instability or cycles along the growth path are modeled as exogenous and random shocks. Post Keynesian growth models (see for example Lavoie and Godley, 2001–02; Godley and Lavoie, 2007) generally deal with growth without distinguishing between trends and cycles. The growth path of interest in this literature is consistent with an arbitrary rate of capacity utilization. Post Keynesian authors who defend the persistent excess capacity thesis do so on the ground that, in contrast to the case in neoclassical models, unemployment can be a long-run problem in capitalism. However, this implicitly assumes acceptance of the neoclassical view that the elimination of excess capacity will also automatically lead to the elimination of unemployment, an argument rejected by Harrod. It may be noted that Kalecki (1971a) did distinguish between trend and cycles but treated the underlying trend as exogenously determined and focused attention on the cycles around it. As with neoclassical models, the Post Keynesian ones assume continuous equality between aggregate demand and supply, although, in sharp contrast to the former, the causality runs from aggregate demand to supply. Finally, there has been in the last two decades a revival of the Harrodian warranted growth framework in which a solution to the famous knifeedge problem has been proposed (Shaikh, 1989, 1991, 1992). Central to this solution is the distinction between the different effects of circulating and fixed investment. Following Ricardo, Marx and Leontief, circulating investment (in raw materials and labor) adds to output while, following Harrod–Domar, fixed investment adds to capacity. Circulating investment

A review of the literature on growth

67

has two implications. First, because production takes time, inputs need to be first purchased before the initiation of the production process (Gilibert, 1987, p. 260; Shaikh, 2009). Thus ex ante production plans take place on the basis of expected demand, thereby allowing for the possibility of disequilibria between aggregate demand and supply. This distinguishes this model from both neoclassical and Post Keynesian models, whose point of departure is the equality between demand and supply. Second, as shown in Chapter 4, the mutual interactions of circulating and fixed investments allow for stable fluctuations in output and capacity around the normal rate of capacity utilization. Thus in this extended Harrodian model, we find a third approach to the relationship between supply and demand and demand and capacity that harks back to Marx. In the first approach, which is the basis of neoclassical theory, demand adjusts itself passively to (actual and potential) supply at full employment and full capacity utilization. On the other hand, on the basis of Keynes’s Principle of Effective Demand, we find the opposite in Post Keynesian models, since demand is the driving variable to which supply passively adjusts itself. Since it is exogenous demand which rules the roost in Keynesian theory, this equilibration process can be consistent with arbitrary levels of employment and capacity utilization. Finally, in the extended Harrod framework, disequilibria between demand and supply and demand and capacity jointly correct themselves in an approximate sense at different adjustment speeds. Neither demand nor supply rules the roost in this equilibration process. In a paper which Kalecki delivered at UNESCO in 1968 at the centenary of the publication of Capital, he argued that investment is exogenous and thus is regulated by institutional and political factors (Halevi and Taouil, 2002). Treating investment as exogenous implies that in the absence of externally motivated stimulating factors capitalism would stagnate. This view, which harked back to the old underconsumptionist approach, was further reinforced by Steindl (1952, p. 2) and Baran and Sweezy (1968), who argued that advanced capitalism was characterized by large firms that had monopoly power and thus maintained huge profit mark-ups. An increase in the mark-up would lower the wage share in national income and, by reducing consumer demand, would reinforce the economic system’s stagnationist tendencies. This is also an argument that one finds in the writings of contemporary structuralist authors (Taylor, 1985). What one might call the second generation of neo-Kaleckian growth models maintained this basic Post Keynesian Weltanschauung but attempted to explicitly model investment. The attempt of this class of models is to show that different parameters can yield either stagnationist

68

Strategic competition, dynamics, and the role of the State

or exhilarationist outcomes.17 Consider the following linear investment function,18 which is common in the Post Keynesian literature (see for example Godley and Lavoie, 2007): gK 5

I 5 f0 1 f1r 1 f2u K

(19)

In this equation the intercept f0 . 0 represents ‘animal spirits’ and I and K are fixed investment and capital stock respectively. Let Y 5 output, u 5 capacity utilization rate 5 Y/K, r 5 profit rate 5 pu (p 5 profit share in valued added), and f1 and f2 be positive reaction coefficients. An increase in the wage share will raise u but lower p, and the equilibrium (when the savings rate 5 investment rate) profit rate and growth rate will increase.19 Thus what is good for workers is also good for capitalists. However, Blecker (2002) shows that this situation holds under very simplifying conditions, including the above functional form of the investment function as well as the assumptions of no savings out of wages, no taxes and no foreign trade. Following Marglin and Bhaduri (1990), Blecker (2002) argues that there is an ambiguity built into the above linear function. Now the sign of the partial derivative ∂gK/∂u 5 f2 . 0 will hold only if r is a constant. This in turn implies that from r 5 pu an increase in u can only come about if p falls by the same amount.20 Therefore, use of this linear function implies that despite the fall in profitability (p) when sales and capacity utilization rate (u) rise firms nonetheless increase investment. One would expect a rise in sales and capacity utilization to make investment rise, while a fall in profitability will make it fall. Thus they conclude that the sign of f2 is ambiguous.21 Instead Marglin and Bhaduri propose a more general investment function: gK 5 f (p, u)

(20)

where fπ . 0 and fu . 0 and p is given by the degree of monopoly power. According to Marglin and Bhaduri, these partials lead to a more plausible analysis of investment, since they suggest that investment will increase if u rises, given p. The savings function is: S sp 5 5 sr 5 spu K K

(21)

Equilibrium between savings and investment implies: spu 5 f (p, u)

(22)

A review of the literature on growth

69

Totally differentiating equation 22 gives: 2 (su 2 fp) du 5 dp sp 2 fu

(23)

Stagnationist (du/dp , 0) or exhilarationist (du/dp . 0) regimes will result as a consequence of the relative magnitudes of the right-hand variables, although the condition sp 5 fu is ruled out. However, Blecker argues that exhilarationism requires that the upward distribution of income will have to raise investment demand by an amount which will more than compensate for the depressed consumption demand, thereby ensuring that aggregate demand rises. Blecker argues that such a condition is unlikely to be met in practice (2002, p. 136). Finally, Blecker shows that if an investment function of the form: gk 5 f0 1 f1pv 1 f2u

(24)

is used, both stagnationist and exhilarationist outcomes may be produced once taxes out of savings and foreign trade are included in the model (2002, pp. 140–41). In this equation v 5 capacity–capital ratio and p 5 profits after taxes. The equation is also immune to the Marglin and Bhaduri criticism, since an increase in u can occur holding p and v (and not r) constant where r 5 pv. As with other neo-Kaleckian models (see for example Lavoie, 2002; Flaschel and Skott, 2006), the properties of the above model are ultimately determined by parameters whose relative magnitudes affect fundamental growth properties. Post Keynesians would on the other hand argue that it is this flexibility that makes their framework adaptable to different circumstances. While flexibility may be a positive feature, it can also be argued that in this Weltanschauung capitalism does not have any basic structural features or laws of motion. Radically different growth regimes are ultimately determined by different parameter values. Significantly, unlike the case for the extended Harrodian growth model, it is unclear what the microeconomic foundations are of these different growth regimes. In other words, since the engine of the capitalist economy is the business enterprise what would prompt oligopolistic or monopolistic firms to behave in a stagnationist manner when macroeconomic parameters have one set of relative values and behave in an exhilarationist manner when they have another? Whatever the functional form, all Post Keynesian investment functions include an ‘animal spirits’ term and a term representing the capacity utilization rate. Both of these terms ensure a certain degree of arbitrariness in the relationship between output and capacity.22 It is the presence of these two

70

Strategic competition, dynamics, and the role of the State

terms that makes the ‘anything goes’ principle possible, that is the capacity utilization rate is a ‘free’ variable rather than structurally constrained. Post Keynesian authors would argue that the persistence of excess capacity is one of the strengths of their framework, as it not only keeps open the role of demand but also enables them to eschew steady-state modeling à la neoclassical economics. From the Post Keynesian perspective there is no tendency for the economy to equilibrate, since ‘the long-run trend is but a slowly changing component of a chain of short-period situations; it has no independent identity’ (Kalecki, 1971a, p. 165, cited from Setterfield, 2002, p. 226). Setterfield (2002, p. 226) summarizes this Weltanschauung: The fact that the long-run trend (read ‘equilibrium’) has no identity independent of the series of short-run outcomes of which it is ultimately just an average means that it (long-run equilibrium) is not something ‘out there’, acting as a centre of gravity towards which the economy inexorably ‘tends’. Nor is it a state that the economy can ever usefully be described as having ‘got into’.

On the other hand, as with neoclassical models, all Post Keynesian models begin with short-run equilibrium between aggregate demand and supply. Far from eschewing equilibrium, the Post Keynesian models are in fact a hybrid of short-run equilibrium and long-run disequilibrium. At a microfoundational level it is unclear why firms’ investment decisions should ensure continuous equality between supply and demand but continuous inequality between supply and capacity. Equally troubling, the presence of the capacity utilization term in the investment function suggests that, starting from any arbitrarily high amount of excess (or redundant) capacity, firms will be willing to add to capacity despite a considerable amount of slack capacity to begin with. Again, the microfoundations of this implication are obscure. Equally important, as discussed in Chapter 2 the warranted path corresponds to the lowest-cost range of output on cost curve, called the normal rate of capacity utilization. This minimum-cost level is itself to a large extent a function of institutional and political factors such as premiums needed to pay for overtime (Foss, 1963; Kurz, 1992), the nature of collective bargaining which would determine wages, State provisioning and subsidization of certain inputs and so on. In short, the firm is a passive cost-taker for a wide range of inputs because their prices are exogenous to the firm. In the classical and Harrodian tradition, firms attempt to adjust capacity to bring it into line with demand, given costs.23 This is equivalent to cost-minimization.24 It is also equivalent to ensuring that fixed investment responds to the gap between the actual (u) and (desired) normal (un) rates of capacity utilization.25 On the other hand, the presence of the capacity utilization term in Post Keynesian investment functions implies

A review of the literature on growth

71

that firms will be satisfied with any level of this variable, as determined by exogenous factors including fiscal policy.26 Cost-minimization from the threat of incumbent firms or from new entrants appears to have no place in this framework.27 Authors inspired by Sraffa have sought to graft Keynesian macroeconomics on to classical microeconomic foundations. In contrast to the Kaleckian perspective, the position taken by some Sraffian authors is that the economy’s long-period position is characterized by the convergence of output and capacity. For example, Serrano (1995) derives what he calls a Sraffian supermultiplier in order to demonstrate how via autonomous expenditures long-run effective demand regulates production capacity. Recognizing that investment expands both capacity and demand, Serrano derives two sets of equations. The first is the multiplier: Y5

Z1I 1 2 wl

(25)

where Y 5 output, Z 5 autonomous demand, I 5 investment, w 5 real wage per worker, l 5 labor input coefficient and wl 5 wage share. Investment at time t will be determined by expected effective demand (Dt+1) and the capital–output ratio: I 5 vDt 11

(26)

If the level of investment at time t creates a level of capacity output in t11 then the current ratio of investment to capacity will generate a particular growth rate of capacity in the next period. The growth rate of capacity will have to keep pace with the growth of effective demand. Let gt+1 5 desired growth rate of capacity 5 (Y *t11 2 Y*) / (Y*) 5 expected growth rate of demand. Then it can be shown that: I 5 v (1 1 gt11) Y*

(27)

This equation states that the investment to capacity ratio is a function of the growth of expected demand and the capital–output ratio. Assuming that Y 5 Y* equations 25 and 27 can be combined to yield the Sraffian supermultiplier: Y* 5

Z [ 1 2 wl 2 v (1 1 gt11) ]

(28)

Thus capacity output is related to autonomous demand via the supermultiplier. The problem with Serrano’s model is that he assumes that output is at capacity. In other words, equation 28 is true because Serrano assumes

72

Strategic competition, dynamics, and the role of the State

that aggregate demand and aggregate capacity output are equal. The process by which capacity and output adjust relative to each other is not discussed by Serrano. This process was central to Harrod’s analysis and, of course, produced the knife-edge problem, which Harrod (1959) sought to solve and to which Shaikh (1989) provided a solution. On the other hand many Sraffians take the position that output is likely to be below capacity even in the long run (Palumbo and Trezzini, 2003). Nell (1997), who draws on both Kalecki and Sraffa, rules out the possibility of output and capacity cycling around each other because of the knifeedge property of Harrod’s model. Thus, as Lavoie (2003) correctly points out, there is a convergence between Kaleckians and most Sraffians with regard to the analyses of the long run. With output more or less persistently below normal capacity exogenous injections of demand, for example from government spending, will also raise output growth. As discussed in Chapter 2, Harrod’s argument was that competition will force firms to eliminate excess capacity in order to minimize selling prices and costs. This will happen both because incumbent firms within an industry will defensively seek to guard their market shares and because industry demand curves will shift because of new capital flows. Given the uncertainties of a market economy, one would naturally not expect a smooth convergence between output and capacity, as in the neoKeynesian view. The reason is that such an equilibrium outcome is inconsistent with the fact that firms adjust capacity to demand under conditions of uncertainty, as described by the OERG. Given the long-term and relatively inflexible nature of fixed capital, it would be quite normal to expect persistent over- and under-shooting of capacity and output. Provided that the disequilibrium dynamics are stable, the equalization between actual and potential output (as with demand and supply) will happen only in an average sense over the course of several cycles. This is a far cry from both the fully adjusted position of the neo-Keynesian perspective and the persistent disequilibrium perspective of the Kaleckian and Sraffian traditions.

4.

CONCLUSION

In the analysis of economic growth three questions are of central importance. First, is the capitalist economy capable of generating growth endogenously and ensuring the creation of a sufficient degree of effective demand so as to realize output at capacity? Second, is growth smooth or is it punctuated by crises and instability? Third, is the system left to its own devices capable of generating growth at full employment?

A review of the literature on growth

73

With regard to the first question, in Marx, Harrod and NEGT growth is endogenous, while in Solow it is exogenous. Marx’s schemes of reproduction and the Harrodian warranted growth path both show the possibility of an endogenously generated growth path. However, as shown in the extended Harrodian model in the next chapter, the creation of a sufficient degree of internally generated effective demand should not be conflated with Say’s law, since both demand and capacity vary and interact with each other to attain rough equality (if the growth path is stable). On the other hand, NEGT assumes Say’s law, in which demand adjusts passively to capacity. Further, unlike NEGT models, in Harrod and Marx both growth and cyclical instability are endogenous. Because they reject the view that laissez faire capitalism automatically produces full employment, models inspired by Kalecki and Steindl are demand-constrained so that output is persistently below capacity. For Post Keynesian economists, output can grow only at a rate equal to potential output (capacity) at full employment (Palley, 2002, pp. 25–6), a viewpoint that sharply differs from the Harrod–Domar perspective. While the Sraffian tradition emphasizes the centrality of normal capacity utilization because of its commitment to classical price theory, in its analysis of growth it assumes either that output is at capacity (as with Serrano) or that, for all practical purposes, the convergence of output to capacity takes a very long time (as with Kurz and Trezzini and Palumbo). Post Keynesian growth models want to maintain the centrality of the role of exogenous demand injections and exogenous changes in institutions (which affects the ‘animal spirits’ term in Post Keynesian investment functions). But in order to do so capacity utilization has to be able to freely adjust to whatever level is determined by demand. Thus the growth rate is essentially regulated by exogenous demand and institutions. Coming from the opposite perspective, growth in the Solow model is also exogenous, determined by parameters representing population growth and technological progress. With regard to the second question, in Marx and Harrod booms and depressions are intrinsic features of the growth process. On the other hand, in neoclassical theory crises are likely to occur only because of either exogenous shocks or ‘inappropriate’ policy interventions in the market. On the other hand, in the Post Keynesian perspective, while short-run business cycles are largely endogenous, generally generated by Minsky-type debt dynamics (Wolfson, 1994), the nature of long-run cycles or growth is essentially determined by exogenous factors. Unless ‘appropriate’ policies and institutions are in place to jump-start effective demand and/or raise ‘animal spirits’, economic stagnation will result (see for example Flaschel and Skott, 2006). In this framework, exogenous changes in monopoly

74

Strategic competition, dynamics, and the role of the State

power enhance the economic system’s stagnationist tendencies by the creation of oligopolies and the suppression of effective demand via increases in the profit share (Mott, 2002). Finally, in contrast to the neoclassical perspective, all the heterodox perspectives are united in the view that laissez faire capitalism has no intrinsic forces that will automatically propel the system to full employment. Clearly heterodox perspectives emphasize the role of the State in bringing the long-run growth rate closer to the natural (full employment) growth rate, although in the Marxian view the limit to such policies is determined by profitability. In the Keynesian tradition there are major differences on this question between the Harrodian and the Kaleckian wings, with the latter advocating continuous increases in budget deficits to attain this aim. On the other hand, as discussed in Chapter 5, the role of State policies is more nuanced in the Harrodian perspective.

NOTES 1. 2. 3. 4. 5. 6. 7.

8.

9. 10.

I am grateful to Mark Setterfield for proposing the acronym NEGT, which implies that there are also non-neoclassical endogenous growth models such as the Harrod–Domar model. Since Y 5 Ka (HL) 1 2a so that y 5 Y/HL 5 ka. See Kurz (2008) for a non-neoclassical interpretation of Schumpeter. See also Schumpeter (1928), in which he discusses the unstable nature of capitalism, an analysis that is inconsistent with neoclassical general equilibrium theory. Dutt (2003) argues that Kalecki (1971a) had an analysis of the determinants of the mark-up which was different from the MR 5 MC approach favored by neoclassical theory. It may be pertinent to ask how a general equilibrium model of steady-state growth can be constructed on the basis of a framework at whose core is the notion of ‘creative destruction’ and the dynamic disequilibrium nature of capitalist growth (Fine, 2000). This is because 0L/0b 5 { a/ (1 2 a 2 b) } { (2 (a 1 b) 2 1) / ( (1 2 a 2 b) (a 1 b) ) } . The sign of this partial will depend on the sign of 2(a 1 b) − 1. ‘. . . all that is required to assure the flexibility of perpetual growth is the existence of a “core” of capital goods that is produced with constant returns technologies and without the direct or indirect use of non-reproducible factors’ (Rebelo, 1991, p. 502, cited from Cesaratto, 1999, p. 784). For example, in the P. Romer (1986) model, in which perfect competition is assumed, knowledge production by individual firms ‘leaks out’ to the larger economy and so expands the knowledge stock of all firms. In equation 7, H 5 K (assuming that b 5 1). If L is constant (say L 5 1) and at the full employment level then substituting H 5 K in the production function Y 5 AKa(HL)1−a gives Y 5 AK (Cesaratto, 1999; Dutt, 2003). See for example Shaikh (1974, 1980), Felipe and Fisher (2003), and Felipe and McCombie (2005a, 2005b). As Michl (2000, p. 188) pithily put it: ‘It is hard to see how a model in which each individual is simultaneously assumed to function as capitalist, worker, inventor and monopolist can faithfully represent any real society, even as a first approximation, for it sounds more like Marx’s idyllic description of life under communism.’

A review of the literature on growth 11.

12. 13. 14. 15. 16. 17.

18. 19.

20. 21. 22.

23. 24. 25. 26.

27.

75

On the other hand, as Mongiovi (1990, pp. 72–3) argues with regard to Say’s law: ‘Throughout the classical literature, a decision to save is presumed to entail a decision to add to one’s capital stock; the portion of national income not spent on consumption goods is by assumption used to expand the economy’s productive capacity, so that aggregate demand is almost instantaneously made equal to the value of output.’ In this regard Marx and the classical school had anticipated the Stockholm school’s later distinction between ex ante and ex post variables. This period can be formally shown to correspond to the three- to five-year inventory cycle. See Chapter 6. Note that in Marx’s schemes of reproduction there is no bank credit. Rather the endogeneity of money arises from his theory of value. See Itoh and Lapavitsas (1999). Marx never showed the process which would bring an increase in the savings rate into equality with a higher rate of investment. See Shaikh (1989) and Chapter 4 for a formalization of this disequilibrium adjustment process. Blaug (1978, p. 264) and Eltis (1998, p. 25) have commented on the knife-edge character of Marx’s equilibrium growth path. Note that stagnationism does not imply that the economy is necessarily always stagnant. A wage-led regime is considered stagnationist in that an increase in the wage share raises aggregate demand while an increase in the profit share depresses it. On the other hand, in an exhilarationist regime an increase in the profit share raises aggregate demand. This section on neo-Kaleckian investment functions is based on Blecker’s (2002) excellent paper. In this model the price is a mark-up f (. 1) on direct labor costs where f is determined by the degree of monopoly power. If a 5 labor coefficient in hours/output and W 5 money wage rate then aW 5 unit labor cost. The price p 5 faW. Abstracting from raw materials the price equals valued added. Then the profit share in value added 5 p 5 (p − aW)/p 5 (f − 1)/ f. If Y 5 output and u 5 proxy for the capacity utilization rate 5 Y/K then the profit rate r 5 (p − aW)Y/pK 5 [(f − 1)/ f]u 5 pu. Let I 5 investment and S 5 savings so that the investment rate gK 5 I/K and the savings rate gS 5 S/K 5 sr (all savings come out of profits). Combining r 5 pu and gS 5 sr and setting gS 5 sr 5 gK 5 I/K 5 f0 1 f1r 1 f2u yields spu 5 f0 1 f1pu 1 f2u. This gives the solution u 5 f0/[(s − f1) p − f2]. It can be shown that du/dp , 0. Thus, given the mark-up rate f, an increase in the wage share (which will lower the profit share p) will raise u and the equilibrium rate of accumulation gK. See Blecker (2002, pp. 132–4). If dr/r 5 0 then dp/p 5 2du/u On the other hand from Harrod’s perspective, given the price, a drop in sales will lead to higher average total costs and thus a decrease in the profit margin. The appearance of excess capacity would in turn produce a drop in investment. Indeed the Kalecki–Steindl assumption regarding the long-run persistence of excess capacity is reminiscent of the old underconsumptionist perspective. It is unclear whether this capacity is desired or undesired and if it is the latter why firms would not eventually jettison it if there is a fall in ‘animal spirits’. See Chapter 2. See Chapter 2, in which I show how one can use Harrod’s perspective to eliminate overor underutilization of capacity so as to attain the minimum portion on the average total cost curve. Given the largely exogenous nature of costs, it is therefore puzzling to see that Lavoie (1996) endogenizes the normal rate of capacity utilization. The only way to interpret this is to assume that cost-minimization by firms plays no role in his model. See equation A2.23 in Appendix 2 of Chapter 4. This might explain why many contemporary Post Keynesian authors (see for example Wray, 1998) place enormous weight on the ability of the capitalist State to regulate output and employment. See Moudud (2006) and Chapter 6 for a critique of this view of the State. Steindl does discuss cost-cutting behavior by firms. However, he argues that price- and

76

Strategic competition, dynamics, and the role of the State cost-cutting behavior only exists as long as there are firms with cost differentials. Eventually, the process of competition weeds out high-cost firms, and all remaining firms have similar cost structures. The incentive to cut prices is eliminated and, in this state of oligopolistic competition, firms raise prices relative to costs, since their monopoly power increases. The remaining oligopolies build up excess capacity and lower investment spending. The combination of excess capacity with a rising mark-up rate brings about economic stagnation (Mott, 2002).

4. 1.

A model of disequilibrium dynamics INTRODUCTION

In this chapter a model of growth and cycles, which is anchored in a stockflow consistent (SFC) framework, will be derived. Along the lines pioneered by Richard Stone and Wynne Godley, the distinguishing feature of the SFC framework is that all flows from each sector of the economy (households, businesses and the government) are explicitly related to each other in a social accounting matrix (SAM) as well as the corresponding balance sheets. The SAM is a flow matrix and is linked to each sector’s stocks of assets and liabilities. While the SFC framework has in recent years gained some currency in Post Keynesian circles, the version deployed in this chapter makes a distinction between ex ante and ex post, while that used by Post Keynesian authors is cast entirely in ex post terms. One important implication of an ex post SAM is that the money supply and money demand are always equal to each other, whereas they are not necessarily so in the ex ante case. Furthermore, the ex ante SAM allows for disequilibria between ex ante savings and investment to be modeled. Section 2 discusses the key features of an ex ante SAM and shows how it forms the basis of the cyclical growth model. In this section I will also examine the Post Keynesian view that money supply and money demand have to be equal in a correctly specified SFC model. Finally, section 2 will investigate the effect of a fall in the social savings rate, caused by an increase in the budget deficit, in the context of the growth model.

2.

DISEQUILIBRIUM DYNAMICS IN AN SFC CONTEXT

(a)

Core Features and Assumptions

A central feature of the extended Harrodian growth model is that it does not assume equilibrium between aggregate demand and supply and output and capacity, respectively. The crucial issue that arises is the stability of

77

78

Strategic competition, dynamics, and the role of the State

these imbalances. Clearly, a growth framework in which the short-run path is knife-edge unstable cannot possibly reach a long-run growth path that is stable. Thus, in order to study the properties of the warranted growth path, one must first analyse the processes via which the economic system gets to it. That is, if the short-run growth rate rises above the trend a stabilizing mechanism has to ensure that the former eventually slows down; conversely, if the short-run growth rate falls below the trend it has to eventually increase. Shaikh (1989, 1992) has proposed the negative feedback effects of finance charges and business costs, respectively, as such stabilizing mechanisms. The model that deals with finance is particularly important, as it can be elaborated into an SFC model as shown below. The model derived here has the following core features. First, it explicitly distinguishes between the ex ante plans and expectations of different sectors and ex post outcomes. This is a necessary feature of a dynamic economic system in which there is pervasive Keynesian uncertainty and individuals in each sector make expenditure plans on the basis of actual and expected revenues and loans from other sectors. An important implication of the ex ante framework is that it allows for discrepancies to arise between firms’ planned output on the basis of their sales expectations and customers’ planned demand. Furthermore, since planned capacity by firms will be installed on the basis of expected demand, there also will be persistent fluctuations between demand and capacity. Second, the ex ante–ex post distinction allows for the modeling of different types of economic cycles which are characterized as disequilibrium adjustment processes as firms adjust their planned investments each period in reaction to discrepancies between actual and expected demand. At the core of these adjustment processes is the distinction between fast and slow variables (Shaikh, 1989, p. 72). Typically, production adjustments to supply–demand imbalances occur over a time period that is more rapid than adjustments to capacity since the latter entail changes in the capital stock. This corresponds to the short-run–long-run distinction which is common to all macroeconomic theories. Demand–supply imbalances in the markets for goods and services induce firms to adjust their inventories toward their target levels and adjust their production plans. These production plans entail planned outlays of circulating investment (raw materials and wages) and fixed investment on the basis of expected sales. Such adjustments will make demand and supply (and thus investment and savings) equalize over some period of time, albeit in a highly turbulent manner. However, fixed investment, which adds to aggregate demand, also expands capacity, as Harrod and Domar emphasized. Firms will thus also need to adjust output and capacity so as not to be pushed too far away from their normal or cost-minimizing rate of

A model of disequilibrium dynamics

79

capacity utilization. As discussed in Chapter 2, given the state of effective demand, competitive pressures force firms to adjust production capacity in order to attain their normal or target rate of capacity utilization. The adjustment of capacity to output is necessarily a slower process than that of output to demand, since the former operates on the average time path attained via the approximate equalization of demand and supply. Financial market disequilibria adjust over a time span that is faster than for disequilibria in the real sector because both the demand and the supply sides adjust more quickly. With regard to the supply side, financial assets and liabilities do not entail any production and are supplied virtually on demand to those willing to purchase assets or creditworthy enough to borrow. On the demand side, fluctuations in demand for financial assets are likely to be more rapid than those for investment goods, because financial market innovations (such as repurchase agreements or re-financing possibilities) allow holders of assets or liabilities to alter the composition of their portfolio relatively easily. For example, Post Keynesian authors have emphasized the fact that banks are never reserve-constrained in supplying loans because of the relative ease with which they can alter their demands for and supplies of assets and liabilities. Implicitly, this assumes that the ability of banks to attract reserves by selling some type of bond or deposit is a relatively easy process because the demand for such assets responds quickly to the interest rate offered. On the other hand the decision of a firm to replace its fixed capital stock operates over a much longer time period determined by the depreciation rates of component parts. Given substantial reserves of unemployed workers, the supply of labor also adjusts relatively quickly to the demand for it. The standpoint taken here is that the expected supply of labor by workers will adjust relatively quickly to the planned demand for it by firms. From this standpoint, the approximate equalization of the supply and demand for labor can be consistent with varying amounts of involuntary unemployment. On the other hand, in the neoclassical perspective labor market equilibrium arises from the equalization of the planned demand for labor by firms with the planned supply of labor by workers, given the latter’s labor–leisure trade-off. This equality ensures zero involuntary unemployment in the neoclassical framework. It should be emphasized that these disequilibrium adjustment processes do not lead to equilibria as that word is conventionally understood but rather to persistent cycles around trends that are generally time-varying. Third, unlike the case in much of the growth literature, the effects of three types of investment are treated separately. Any desired change in the level of output by firms will require corresponding changes in the inputs of raw materials and labor power. This relationship, familiar to Marx, Ricardo and Leontief, implies that circulating investment will lead to

80

Strategic competition, dynamics, and the role of the State

changes in the level of production. Further, following Harrod and Domar, net fixed investments add to capacity. Finally, investment in finished goods represents desired changes in final goods inventories. Fourth, the model has as its point of departure an aggregate budget constraint, which is the sum of sectoral budget restraints. Clower (1965) argues that the significance of sectoral budget constraints is that, by relating planned uses of funds to expected sources of funds, each sector’s decisions are financially consistent. These restraints are financial restrictions which ensure that every sector’s finance requirements equal the excess of its planned expenditures over its revenue flows. On the other hand, Post Keynesian SFC accounts are in terms of ex post accounting identities (Dos Santos, 2006; Godley and Lavoie, 2007). Following the Post Keynesian literature, bank credit is produced on demand to creditworthy firms. Injections of credit increase the money supply, while the payback of loans reduces the money supply. The total money supply is determined by bank credit and the portion of the budget deficit that is money-financed. It will be assumed that the government monetizes a fixed part of the budget deficit while bond-financing the remaining part of it. It would be coincidental if the planned increase in the money supply, created as mentioned above, should equal the planned demand for money held by the private sector, where the latter can be interpreted as the private sector’s liquidity preference (Peterson and Estenson, 1996, Chapter 11). In this ‘buffer stock’ (Rabin, 1979, 1993; Coghlan, 1981; Goodhart, 1984, 1989; Yeager, 1986; Dow and Dow, 1989; Laidler, 1990, 1993) or inventory approach to the stock of money holding (Baumol, 1952) the private sector seeks to maintain some desired stock of money relative to its income, and any deviation of actual stockholdings from desired stockholdings corresponds to a dynamic adjustment process. The ex ante SAM can thus be used to understand how disequilibrium money (Goodhart, 1984) comes about. It will be shown below that a flow matrix written in ex ante terms necessarily involves inequality between money supply and money demand as long as the market for goods and services is in disequilibrium. However, in both the ‘Keynesian’ neoclassical synthesis and new classical economics such disequilibria are annulled instantaneously. Interest rate flexibility in the neoclassical synthesis case and wage and price flexibility in the new classical case ensure continuous market-clearing. On the other hand, given the relationship between goods market and monetary disequilibria derived below shocks to the money stock are likely to dissipate over the time frame needed for excess demand in the real sector to approximately equal zero. Fifth, and finally, the model abstracts from price changes.

A model of disequilibrium dynamics

(b)

81

A Social Accounting Matrix in a Closed Economy

Table 4.1 is a SAM cast in ex ante terms which relates all flows and changes in stocks to each other in a consistent manner. This in turn implies that the planned savings of each sector are equal to its planned change in net worth. Planned savings equal the balance on the current account and constitute the excess of expected income over planned expenditures. The planned change in net worth, obtained from the capital account, is the change in planned (or expected) assets minus the change in planned (or expected) liabilities. The fact that planned savings 5 planned change in net worth implies that for each sector the sum of the current and capital accounts has to equal zero. The negative signs refer to money outflows from each sector and positive signs refer to money inflows into each sector. Furthermore, the symbol = represents the planned or expected change in a magnitude, that is the difference between current planned or expected levels and the initial level (e.g. =Lp 5 Lp − L−1). Moreover, all price changes including capital gains and losses are ignored at this level of abstraction. These issues, while highly important in themselves, do not in any way change the basic dynamics of the extended Harrodian growth model in Shaikh (1989). Finally, the subscripts ‘h’, ‘nf’, ‘b’ and ‘g’ stand for households, nonfinancial businesses (including State-owned enterprises), banks and government respectively. We will consider each sector’s budget restraint which relates its planned uses of funds to its expected sources of funds. Household sector Sources of funds 5 Uses of funds (We 1 DIVhe 1 IDhe 1 IBChe 1 IBGhe) 1 =Lhp 5 Ch 1 Thp 1 ILhp 1 =Hhp 1 =Dhp 1 =BChp 1 =BGhp 1 =EQhp (1) p

The first parenthesis on the LHS represents total expected revenue from expected wages (We), expected dividends (DIVhe), expected interest earnings from bank deposits (IDhe), and expected interest earnings from corporate bonds (IBChe) and government bonds (IBGhe), and =Lhp 5 planned addition to bank loans. On the RHS, planned consumption 5 Chp, planned (direct and indirect) tax payment 5 Thp, planned interest payment on bank loans (5 planned finance charges 5 interest payments 1 principal due) 5 ILhp, planned addition to cash stocks 5 =Hhp, planned addition to bank deposit stocks 5 =Dhp, planned addition to corporate bonds 5 =BChp,

82

Strategic competition, dynamics, and the role of the State

planned addition to government bonds 5 =BGhp, and planned additions to equity 5 =EQhp. Finally, planned savings are Shp: Shp 5 (We 1 DIVhe 1 IDhe 1 IBChe 1 IBGhe) − (Chp 1 Thp 1 ILhp) (2a) The change in planned net worth =NWhp is: =NWhp 5 (=Hhp 1 =Dhp 1 =BChp 1 =BGhp 1 =EQhp) − =Lhp

(2b)

Non-financial businesses (includes State-owned enterprises) Sources of funds 5 Uses of funds (Cnfe 1 Infe) 1 =INVnfe 1 IDnfe 1 IBGnfe 1 =EQnfe 1 =BCnfe 1 =Lnfp 5 Wnfp 1 Tnfp 1 ILnfp 1 IBCnfp 1 DIVnfp 1 Ifnfp 1 =INVnfp 1 =Hnfp1 =Dnfp 1 =BGnfp (3) In this equation the first parenthesis on the LHS represents total expected net sales Se where Cnfe 5 expected sales of firms to households and the government, Infe 5 expected sales of investment goods by one set of firms to another group of firms so that expected net sales 5 (Cnfe 1 Infe), =INVnfe 5 expected change in inventories (so that expected output Ye 5 Cnfe 1 Infe 1 =INVnfe 1), IDnfe 5 expected interest earnings from bank deposits, IBGnfe 5 expected interest earnings from government bonds, =EQnfe 5 expected sale of equity, =BCnfe 5 expected sale of corporate bonds to households, and =Lnfp 5 planned increase in bank loans. On the RHS, Wnfp 5 planned wage payments, Tnfp 5 planned tax payment,2 ILnfp 5 planned interest payment on bank loans (5 planned finance charges 5 interest payments 1 principal due), IBCnfp 5 planned interest payments on corporate bonds, DIVnfp 5 planned dividend payment, Ifnfp 5 planned fixed investment, =INVnfp 5 planned addition to inventories (in raw materials, work-in-progress and finished goods), =Hnfp 5 planned addition to cash, =Dnfp 5 planned addition to bank deposits, and =BGnfp 5 planned addition to government bonds. Planned savings Snfp are: Snfp 5 [(Ce 1 Infe) 1 =INVnfe 1 IDnfe 1 IBGnfe)] − (Wnfp 1 Tnfp 1 ILnfp 1 IBCnfp 1 DIVnfp)

(4a)

Note that sales are defined as net of the equivalent amount of intermediate inputs used up so that planned additions to inventory stocks include additions to raw material stocks. Thus for the non-financial sector planned profits net of taxes 5 PNnfp 5 Snfp 1 DIVnfp are:

A model of disequilibrium dynamics

PNnfp 5 [(Ce 1 Infe) 1 =INVnfe 1 IDnfe 1 IBGnfe)] − (Wnfp 1 Tnfp 1 ILnfp 1 IBCnfp)

83

(4b)

The change in planned net worth =NWnfp is: =NWnfp 5 (Ifnfp 1 =INVnfp 1 =Hnfp1 =Dnfp 1 =BGnfp) − (=Lnfp 1 =EQnfe 1 =BCnfe)

(4c)

Banks Sources of funds 5 Uses of funds =Dbe 1 ILbe 1 IBGbe 1 =BCbe 1 =EQbe 1 =BRbp 5 Ifbp 1 =INVbp 1 DIVbp 1 IBCbp 1 Tbp 1 Wbp 1 IDbp 1 IBRbp 1 =BGbp 1 =Hbp 1 =Lbe

(5)

On the LHS, =Dbe 5 expected supply of deposits, ILbe 5 expected interest earnings on bank loans, IBGbe 5 expected interest earnings on government bonds, =BCbe 5 expected sale of corporate bonds to households, =EQbe 5 expected sale of equity to households, and =BRbp 5 planned borrowed reserves from central bank. On the RHS, Ifbp 5 planned fixed investment, =INVbp 5 planned additions to inventories (office supplies, etc.), DIVbp 5 planned dividend payments, IBCbp 5 planned interest payments on corporate bonds to households, Tbp 5 planned tax payments, Wbp 5 planned wage payments, IDbp 5 planned interest payments on bank deposits, IBRbp 5 planned interest payments on borrowed reserves, =BGbp 5 planned purchase of government bonds, =Hbp 5 planned additions to reserves, and =Lbe 5 expected loan supply to households and non-financial firms. Finally, the planned savings of the banking sector Sbp are given by: Sbp 5 (ILbe 1 IBGbe) − (DIVbp 1 Tbp 1 Wbp 1 IDbp 1 IBCbp 1 IBRbp) (6a) Planned bank profits net of taxes PNbp 5 Sbp 1 DIVbp are: PNbp 5 (ILbe 1 IBGbe) − (Tbp 1 Wbp 1 IDbp 1 IBCbp 1 IBRbp)

(6b)

The change in planned net worth =NWbp is: =NWbp 5 (Ifbp 1 =INVbp 1 =BGbp 1 =Lbe 1 =Hbp) − (=Dbe 1 =BCbe 1 =EQbe 1 =BRbp)

(6c)

84

Strategic competition, dynamics, and the role of the State

Government Both the Treasury and the central bank have been merged. Sources of funds 5 Uses of funds Tge 1 IBRge 1 =Hge 1 =BGge 5 Cg 1 Ifgp 1 =INVgp 1 =BRge 1 IBGgp p

(7)

On the LHS, Tge 5 expected taxes from all other sectors (includes expected dividends from the surpluses of State-owned enterprises), IBRge 5 expected interest earnings from borrowed reserves, =Hge 5 expected additions to high-powered money, =BGge 5 expected sales of bonds to all sectors by the Treasury. On the RHS, Cgp 5 planned purchase of goods and services,3 Ifgp 5 planned purchase of equipment and structures from the private sector, =INVgp 5 planned additions to inventories, =BRge 5 expected discount window loans, and IBGgp 5 planned interest payment on government bonds. Finally, planned government savings Sgp are: Sgp 5 (Tge 1 IBRge − IBGgp) − Cgp

(8a)

Note that this definition of government savings, which excludes public investment expenditures, follows the approach taken by Musgrave (Musgrave and Musgrave, 1973, pp. 489–90), Keynes and others (see Chapter 5 for a discussion of capital budgeting and growth). The corresponding change in planned net worth =NWgp is: =NWgp 5 [=BRge 1 (Ifgp 1 =INVgp) − =Hge − =BGge]

(8b)

To understand the construction of the ex ante SAM the following points need to be remembered. First, the items in each column add up to zero because the sum of the current and capital accounts constitute the budget restraint for each sector (equations 1, 3, 5 and 7). As discussed above, this is equivalent to saying that for each sector planned savings 5 change in planned net worth. It follows therefore that the row sum of the last row has to equal zero. However, each row sum, which represents each sector’s transactions with other sectors, may not necessarily be equal to zero in an uncertain world in which every sector plans certain expenditure or expects certain receipts that may not equal other sectors’ plans and expectations. For example, firms plan on producing an output on the basis of their expected sales that may not be equal to the planned demand of customers. Imbalances between planned production and expected sales are likely to be the norm in a world of Keynesian uncertainty in which production

85 −Wnfp −DIVnfp +IDnfe +IBGnfe

+IDhe

+IBGhe

Interest on deposits Interest on government bonds

+Cnfe +Infe + =INVnfe

Current account

+We +DIVhe

Capital account −Ifnf − =INVnfp

p

Capital account

Non-financial firms (private and public)

Wages Dividends

Memo: expected output Memo: excess demand

−Chp

Current account

Households

Ex ante social accounting matrix

Consumption Investment

Table 4.1

+IBGbe

−IDbp

−Wbp −DIVbp

Current account −Ifb − =INVbp

p

Capital account

Banks

−IBGgp

−Cgp

Current account −I − =INVgp

p fg

Capital account

Government

(IBGhe + IBGnfe + IBGbe) − IBGgp

Cnfe − (Chp + Cgp) (Infe + =INVnfe) − (Ifnfp + Ifbp + Ifgp + =INVnfp + =INVbp + =INVgp) Ye = Cnfe + Infe + =INVnfe E = −(Ye – Yp) = −[(Cnfe + Infe + =INVnfe) − (Chp + Cgp + Ifnfp + Ifbp + Ifgp + =INVnfp + =INVbp + =INVgp)] We – (Wnfp + Wbp) DIVhe – (DIVnfp + DIVbp) (IDhe + IDnfe) − IDbp

Ex ante disequilibria

86

Current account

Current account Shp subtotal: planned savings Change in highpowered money Change in borrowed reserves Change in deposits −=Hnfp

−=Dnfp

−=Dhp

Snfp

−Tnfp Sbp

−Tfp

+ILbe −IBRbp

−ILnfp

Current account −IBCbp

Capital account

−IBCnfp

Current account

Non-financial firms (private and public)

−=Hhp

Capital account

Households

(continued)

Interest on +IBChe corporate bonds Interest on loans −ILhp Interest on borrowed reserves Taxes −Thp

Table 4.1

+=Dbe

−=BRge

+=BRbp

Capital account

+=Hge

Sgp

+Tge

+IBRge

Current account

Government

−=Hbp

Capital account

Banks

=Dbe – (=Dhp + =Dnfp)

=Hge – (=Hhp + =Hnfp + =Hbp) =BRbp − =BRge

Tge − (Thp + Tnfp + Tbp)

IBChe − (IBCnfp + IBCbp) ILbe – (ILhp + ILnfp) IBRge − IBRbp

Ex ante disequilibria

87

+=BCnfe −=BGnfp

+=EQnfe

−=BChp

−=BGhp

−=EQhp

+=EQbe

−=BGbp

+=BCbe +=BGge

Change in bank +=Lhp +=Lnfp −=Lbe loans Capital account −=NWhp −=NWnfp −=NWbp −=NWgp subtotal: −=net worth Column sums Shp − =NWhp = 0 Snfp − =NWnfp = 0 Sbp − =NWbp = 0 Sgp − =NWgp = 0

Change in private bonds Change in government bonds Change in equity

0

(=EQnfe + =EQbe) − =EQhp (=Lhp + =Lnfp) − =Lbe

(=BCnfe + =BCbe) − =BChp =BGge – (=BGhp + =BGnfp + =BGbp)

88

Strategic competition, dynamics, and the role of the State

takes time. On the other hand, Godley, Taylor and other authors in the SAM literature write the SAM entirely in ex post terms so that the final column consists of zeros. Second, business investment appears both as expected net sales of investment goods in the current account (1Ifnfe) and as planned expenditures in the capital account (−Ifnfp). In general these will not be equal, since sellers of a particular type of investment good are unlikely to be buyers of the same kind of investment good. Moreover, planned additions to business inventories include additions to stocks of raw materials, work-in-progress (semi-finished goods) and finished goods. Shaikh (2009) shows that in national income accounts circulating investment (investment in labor power and raw materials) is exactly equal to changes in the inventory stocks of raw materials and work-in-progress. Third, government planned purchases of fixed assets (equipment and structures) and planned inventory investment are part of its capital account. Planned investment in State-owned enterprises is counted as part of the business sector. Fourth, the SAM can be used to derive the aggregate budget restraint (ABR) for the entire economy. Two versions of the ABR can be derived. Let the planned social savings be S*p (5 Shp 1 Snfp 1 Sbp 1 Sgp) and aggregate planned investment demand Ip (5 Infp 1 =INVnfp 1 Ibp 1 =INVbp 1 Igp 1 =INVgp). Note that in this definition of aggregate investment only one part (Infp 1 =INVnfp) contributes to aggregate output (including finished goods inventories) and capacity, while the other two are purchases of investment goods by banks and the government produced in the nonfinancial business sector. In Appendix 1 to this chapter, it is shown that the excess demand in the goods and services sector, the actual money supply (Ms) and planned money demand (Mdp) are related as follows: E 5 Ip − S*P 5 Ms − Mdp

(9)

In this equation E 5 excess demand in the market for goods and services 5 planned demand by households, non-financial firms, financial firms and the government (Yp) − expected demand (sales) by non-financial firms (Ye) 5 Yp − Ye. Ip 5 planned investment by all sectors, which includes planned additions to inventories. The Post Keynesian literature is unified in the view that the money supply is endogenously generated and is regulated by the rate of economic activity. Money creation is intimately linked to the injection of credit because of the key proposition that ‘loans create deposits’. There is, however, less agreement about the role that the demand for money

A model of disequilibrium dynamics

89

plays. Some Post Keynesians either deny the existence of an independent demand for money (Moore, 1997) or argue that it is economically meaningless to distinguish between money supply and money demand (Godley, 1999; Lavoie and Godley, 2001–02). Another strand of the Post Keynesian literature takes the position that there is no guarantee that the growing supply of deposits created by the borrowing activities of one segment of the economy will necessary ensure that all segments of society willingly hold the deposits at all points in time (Howells, 2001). There could thus be a positive or negative excess demand for money. Authors who take this view have suggested a variety of mechanisms that will bring about a rapid equalization between the supply and demand for money. Beginning with their 1981 critique of monetarist theory and policy, Kaldor and Trevithick sought to show that any imbalance between the supply and demand for money would virtually automatically be extinguished so that there could never be a situation of ‘too much money chasing too few goods’. The Kaldor–Trevithick view was that any excess money supply would automatically lead to the repayment of bank loans: Unlike commodity money, credit money comes into existence as a result of borrowing from the banks . . . and it is extinguished as a result of repayment of bank debt (which happens automatically under a system when an excess of receipts over outlays is directly applied to a reduction of outstanding overdrafts). Hence in a credit money economy, unlike with commodity money, the outstanding ‘money stock’ can never be in excess of the amount which individuals wish to hold. (Kaldor and Trevithick, 1981, p. 7, cited from Howells, 1995, p. 93)

Arestis and Howells (1996) criticize the Kaldor–Trevithick mechanism everybody is unlikely to have an overdraft. Instead, they propose an alternative mechanism that relies on relative variations in interest rates. When deposit holders hold more money than they desire, they will shift their funds to the bond market. This will raise the prices of bonds and lower their interest rates. Consequently, firms will reduce their demand for bank credit and increase the sale of corporate bonds as an alternative source of finance. This will lower bank credit supply and thus the excess money supply, while the increase in the interest rate raises money demand. Thus, the relative variations of the interest rates close the initial imbalance. Beginning with the theory of monetary circuit (Graziani, 1990; Lavoie, 1992), Lavoie (1999) uses the Kaldor–Trevithick reflux mechanism to show how monetary imbalances are extinguished. One key element of the theory of monetary circuit is that household demand for bank deposits and the outstanding stock of bank loans and advances are not independent variables. Lavoie begins his discussion by recognizing that, when firms pay out

90

Strategic competition, dynamics, and the role of the State

wages and dividends, ‘they always wind up with some undesired money balances’ (1999, p. 106). As a consequence of these undesired money balances, households will purchase goods and services and will offer what Moore (1988) calls convenience lending to firms. With the money from their higher sales, firms reduce their bank borrowings. Thus: As households get rid of undesired deposits, advances will be paid back, so that they decrease pari passu with the decreased stock of money held by households. The gross flow of advances and hence the flow of money depends on the demand for credit of firms, but the stock of advances and of money deposits depends in addition on the portfolio behaviour of all agents. (Lavoie, 1999, p. 106, emphasis in the original)

In other words, by including a price mechanism, the Arestis–Howell mechanism is a variant of the reflux mechanism. Lavoie points out that lines of credit held by households also provide the means by which any excess money supply is automatically eliminated. When an employer makes a deposit into the account, it is instantaneously used to reduce any overdrafts so that the money supply falls. Thus far the discussion has assumed a pure private sector closed economy. However, matters are no different once monetary injections from budget deficits or the balance of payments are included. In a manner that echoes the arguments of Le Bourva (1992), Arestis summarizes the situation as follows: The endogeneity character of the money supply implies that there can never be ‘an excess supply of money’. The recipients of such an ‘excess’ would use it to diminish their liabilities so that the ‘excess’ is extinguished as a result of the repayment of bank debts. This argument explains the post-Keynesian contention that government deficits and favourable balance of payments have no direct effects on the creation of money. For any money thus created is completely compensated by an equivalent reduction in credit money. (Arestis, 1988, p. 65, cited from Lavoie, 1999, p. 108)

Thus, as Lavoie (1999) concludes, the law of reflux applies to injections of money from such additional sources also. The Post Keynesian literature surveyed thus far assumes that financial market adjustments will ensure the seamless and rapid disappearance of any money supply–money demand imbalance. However in her comments on the Kaldor–Trevithick mechanism, Victoria Chick (1992) argues that the adjustment process may be more complicated: However, the money might not fall into the hands of those with overdrafts in the first instance. The money may be spent and push up prices after which the

A model of disequilibrium dynamics

91

money may be willingly held; or the expenditure could increase profits and the profits used either to finance [expenditure] or to repay bank debt. The latter is only one possibility. (Chick, 1992, p. 205, cited from Howells, 1995, p. 94, emphasis added)

On the other hand, the adjustment process is virtually automatic in the models of the other Post Keynesian authors. Yet clearly the various versions of the compensation principle to which all these authors subscribe require that the fall in bank credit has to be exactly equal and opposite to the initial increase in the money supply. In the Arestis–Howells model, for example, the elasticity of substitution between credit and bond finance has to be just right so as to bring about the necessary decrease in bank credit in order to compensate for the initial excess money supply. One may well ask, what factors could bring about such an exact compensation? Further, Lavoie (1999) argues that part of the excess money held by households would be spent on goods and services and once spent would end up as the deposits in the business sector, which would use it to pay down its debt to the banking sector. While this latter argument is correct, in reality it ignores the actual evolution of time: until aggregate demand and supply balance out there is no reason why the additional money balances would more or less instantaneously reduce debt. Suppose, for example, firms had been overoptimistic ex ante such that their expected sales exceeded the planned demand for goods by households. In such a case the ex post confrontation of demand and supply will produce undesired inventory accumulation and thus not enough cash flow for firms to pay down their debt in its entirety. The point taken in this book is that production takes time, planned production occurs on the basis of expected sales under pervasive Keynesian uncertainty, and the relationship E 5 Ms − Mdp follows from an ex ante SFC system of accounts. The first two factors imply that demand and supply will persistently over- and under-shoot, bringing about adjustments by firms that take time. Thus excess demand E will be different from zero at any given moment in time and will be zero only in an approximate sense over a period of years. The third factor implies that, during the time that E ≠ 0, Ms and Mdp will not be equal to one another, while the equalization of Ms and Mdp will occur only when E ≈ 0. The Post Keynesian claim that the new money injected from a budget deficit is more-or-less instantaneously extinguished through financial market channels is contradictory because it ignores the real effects of monetary demand injections – a central plank of Keynesian theory of all types. If monetary demand does stimulate production then, as discussed above, it will take time for the excess money supply to extinguish itself because of disequilibria between demand and supply in the real sector.

92

Strategic competition, dynamics, and the role of the State

On the other hand, from the Godley–Lavoie (Lavoie and Godley, 2001– 02; Godley and Lavoie, 2007) standpoint, in a modern banking system there is no economic meaning in distinguishing between the supply and demand for money, since all stocks and flows are linked to each other in a consistent manner. However, this result is true only if the SAM is written in ex post terms, since in that case E 5 0 and from equation 9 the condition Ms 5 Mdp will always be true. Implicitly, production requires no time and output instantaneously meets demand. Thus the ex ante and ex post versions of the SAM have very different macroeconomic implications. (c)

Disequilibrium Dynamics in a Model of Cyclical Growth

In this section an ultra-simple model of cyclical growth is derived. It is concerned only with the modeling of the investment behavior of non-financial firms, in particular the ways in which investment reacts to credit flows and debt and makes output and capacity cycle around each other. Because of the model’s growth context, all variables are deflated by another trended variable (such as output), since our concern is with changes in the level of each variable relative to the trend. The level of each variable is in upper case (e.g. E), while its corresponding fraction relative to output is in lower case (e.g. e). Because of the dynamic framework, changes in all variables are modeled relative to the trend of output. The dynamics involve modeling the disequilibrium between aggregate demand and supply as well as that between the output path thus traced out and the path of capacity. These imbalances, represented by excess demand (e) and the rate of capacity utilization (u), will affect circulating and fixed investment in different ways. The basic dynamics are as follows. Suppose the exogenously given social savings rate (social savings as a proportion of output) drops. This will make the aggregate investment share exceed the savings rate (e . 0), and the first response of firms will be to expand output by increasing circulating investment. However, in order to do so they will need to borrow from banks. Thus the excess demand will be fueled by bank credit. Excess demand E is the gap between aggregate planned investment and planned social savings.4 From the SAM planned social savings S*p 5 (Shp 1 Snfp 1 Sbp 1 Sgp) and aggregate planned investment demand Ip 5 (Ifnfp 1 =INVnfp) 1 (Ifbp 1 =INVbp) 1 (Ifgp 1 =INVgp). E 5 Ip − S*p

(10)

In total planned investment Ip, only one part (Ifnfp 1 =INVnfp) contributes to aggregate output and capacity, while the other two are purchases of

A model of disequilibrium dynamics

93

investment goods by banks and the government produced in the nonfinancial business sector (which includes State-owned enterprises). For non-financial firms the term (Ifnfp 1 =INVnfp) 5 Ifnfp 1 Icnfp 1 Ivnfp where Ifnfp 5 planned fixed investment on new plant and equipment, Icnfp 5 planned circulating investment, and Ivnfp 5 planned finished goods inventory investment.5 Thus equation 10 can be rewritten as: E 5 (Ifnfp 1 =INVnfp) 1 (Ifbp 1 =INVbp) 1 (Ifgp 1 =INVgp) − S*p 5 (Ifnfp 1 Icnfp 1 Ivnfp) 1 Inpp − S*p (11a) Here Inpp 5 (Ifbp 1 =INVbp) 1 (Ifgp 1 =INVgp) 5 purchase of investment goods by banks and the government although both sets of planned expenditures involve no further expansion of output. Hence they are forms of non-production expenditures, represented by the subscript ‘np’. Equation 11a can alternatively be written as: E 5 (Ifnfp 1 Icnfp 1 Ivnfp) − S**p

(11b)

The new variable S**p (5 S*p − Inpp) is the portion of the social savings absorbed by non-production expenditures of the type that entail the purchase of investment goods by banks and the government. When E 5 0, firms’ actual inventory stocks will equal their desired inventory stocks. However, if demand and supply are not equal then there will be a discrepancy between actual and desired inventory stocks. Since actual inventory stocks 5 desired inventory stocks 1 undesired inventory stocks, it follows that E 5 −(change in undesired inventory stocks). This ensures that during upswings when E . 0 actual inventories fall below their desired levels, while in downswings when E , 0 actual inventories rise above desired levels. Dividing equation 11a through by net output Y we get: e 5 acnf 1 avnf 1 afnf 1 anp − s* 5 acnf 1 avnf 1 afnf − s**

(12)

where e 5 E/Y, acnf 5 Icnf/Y, avnf 5 Ivnf/Y, afnf 5 Ifnf/Y, anp 5 Inp/Y, s* 5 S*/Y, and s** 5 s* − anp. We have specified three types of investments: fixed investment afnf, circulating investment acnf, and investment in finished goods avnf. Following Harrod and Domar, fixed investment adds to capacity by expanding the fixed capital stock Kf (Ifnf 5 DKfnf) while, following Leontief and the classical tradition, circulating investment adds to output (Icnf 5 DKcnf). These relationships are respectively represented by m (the circulating capital– output ratio) and k (the fixed capital–output capital ratio):

94

Strategic competition, dynamics, and the role of the State

m5

k5

Kcnf Y Kfnf Y*

(13)

(14)

If m and k are parameters then we get the growth rates of output and capacity, respectively: 1 DKcnf 1 Icnf 1 DY 5 5 5 acnf (15) m Y m Y m Y DY* 1 DKfnf 1 If 1 5 5 5 afnf k Y* k Y* k Y*

(16)

The ratio between output and capacity is the rate of capacity utilization: u5

Y Y*

(17)

The disequilibrium variables, u and E, correspond to realized, short-run trend and long-run trend (or normal) profits, which are respectively PR, P and Pn (all of which are gross of finance charges): PR 5 P 1 E and P 5 Pnu

(18)

Starting with aggregate demand 5 aggregate supply (e 5 0) suppose that e . 0. Since realized profits PR/Y 5 P/Y 1 e, when e . 0 realized profits will rise relative to the short-run trend of output, so that PR/Y . P/Y. It follows that firms will be stimulated to increase circulating investment (and thus output) whenever PR/Y . P/Y and to reduce it when PR/Y , P/Y. The positive excess demand implies the excess of planned investment share over the social savings rate, which in turn produces an injection of bank credit. However, the injection of credit leads to the growth of finance charges. It will be assumed that fresh bank loans (Lb) at the end of time period (t−1) will entail finance charges (payment of principal 1 interest rate on the loan) in period t. The over-accumulation of debt will slow down the cyclical (disequilibrium) growth rate and bring about the recessionary phase when excess demand falls, eventually becoming negative (e , 0). The interest rate on bank loans moves procyclically because of variations in credit market pressures during upswings and downswings. These disequilibrium adjustments happen relatively fast during a time period when the fixed investment share afnf is stable. Thus capacity grows at some rate arbitrarily different from output. As discussed in Chapter 2,

A model of disequilibrium dynamics e lb i

0.15 0.15 0.1

95

0.075

0.1

0.0725

0.05 0.05 0

0.0775

0.07

0

–0.05 –0.05

0.0675

–0.1 –0.1

0.065

–0.15 –0.15

0.0625 0.06

–0.2 –0.2 0

5

10

15

20

time {s}

Note: Excess demand and debt are the first and second on the left scale, respectively. Interest rate is on the right scale.

Figure 4.1

Fast adjustment process involving excess demand (e), business debt (lb) and the interest rate (i)

persistent over- or underutilization of capacity (relative to the practicable optimal rate of capacity utilization) is not sustainable for competitive reasons. Thus firms will adjust their fixed investment to bring capacity into rough equality with output (demand). This latter adjustment mechanism is at the core of the slow adjustment process and is the basis of Harrod’s treatment of investment and instability, as discussed below. However, in the current context the instability is not of the knife-edge kind. Figure 4.1 shows the cyclical relationship between excess demand, debt and the interest rate. The accumulation of debt effectively stabilizes the short-run disequilibrium cycles, thereby ensuring the rough equality of aggregate demand and supply over a period of several years. As can be seen from equations 15 and 16, the circulating and fixed investment shares trace out paths of output and capacity, respectively. This path of output arises as a result of the approximate equalization of demand and supply. However, this output path is likely to be arbitrarily below or above the path of capacity traced out, since the fixed investment share is constant in the short run. Figure 4.2 shows the paths of the logs of realized output YR (5 Y 1 E) and actual output Y, with the former cycling around the latter. Note that the growth path of actual output is arbitrarily related to the growth of capacity in the short run or fast adjustment process. However, firms will eventually adjust their fixed investment shares when confronted with the over- or underutilization of capacity so as to attain the normal rate of capacity utilization. This is the slow adjustment process

96

Strategic competition, dynamics, and the role of the State

3.4 3.2 3 2.8 2.6 2.4 lnY lnYR

2.2 2 0

Figure 4.2

5

10 time {s}

15

20

Natural logarithms of short-run trend of output (Y) and realized output (YR)

in which both circulating and fixed investment will adjust to disequilibria. For example, in the event that u 5 Y/Y* . 1 firms will increase the fixed investment share, which in turn will raise capacity growth (equation 16). Since e ≈ 0 over a series of short-run cycles, equation 12 becomes: acnf 1 avnf 1 afnf − s** 5 0

(19)

Given avnf and s**, an increase in the fixed investment share afnf necessarily implies a relative reduction in the circulating investment share, which will in turn slow output growth (equation 15). Thus starting with u . 1 these twin movements of output and capacity will make u move back toward u 5 1. This will ensure stable convergence between output and capacity (albeit with under- and over-shooting) and the establishment of the warranted path. We next investigate the effect of a decrease in the social savings rate, possibly caused by an increase in the budget deficit. This situation is illustrated in Figure 4.3, which plots the growth rates of output (GY) and capacity (GY*), the capacity utilization rate (u) and the social savings rate (s**). To understand the above dynamic it must be realized that excess demand is approximately zero across a series of short-run cycles (equation 19). This implies that a fall in the social savings rate will eventually have to lower the circulating investment share (given avnf and an initial value of afnf stable around some initial value of u, which for convenience we take as u ≈ 1). This in turn will lead to a slowdown of the growth of output relative to the growth of capacity, thereby making the capacity utilization rate fall

A model of disequilibrium dynamics

97 GY GY* u s**

0.16 0.16

0.298889 1.2

0.14 0.14

0.287778 1.15

0.12 0.12

0.276667 1.1

0.1

0.265556 1.05

0.1

0.08 0.08

0.254444 1

0.06 0.06

0.243333 0.95

0.04 0.04

0.232222 0.9

0.02 0.02

0.221111 0.85

0

0.21

0 0

20

40

60

80

100

120

0.8

140

time {s}

Note: GY and GY* are on the left scale while u and s** are the first and second columns, respectively, on the right scale.

Figure 4.3

The effect of a fall in the adjusted social savings rate (s**) on actual output growth (GY), capacity growth (GY*) and the capacity utilization rate (u)

below unity. The fixed investment share will fall as a consequence because of the appearance of excess capacity. Thus the growth of capacity will also fall. Now given that s** has reached a lower steady level and afnf falls because u , 1, there is room for acnf to rise with excess demand approximately equal to zero (equation 19). Thus as a consequence of the initial fall in s** output accelerates with a lag (i.e. GY rises), thereby in turn raising u, a kind of dynamic Keynesian effect. The increase in u will be more pronounced if the response of afnf is relatively slower. This situation would imply that output growth rises faster relative to capacity growth (GY*). Thus in this situation a fall in the social savings rate will bring about an increase in the capacity utilization rate in the next period while ultimately lowering the warranted growth rate.6 As the system settles along its new steady-state values, u will again be approximately equal to unity. Appendix 2 derives the equations for both adjustment processes and shows that the equation for the warranted growth rate is: Gw 5

a

Kfnf Y*

1

s** Kcnf Y*

1

Kvnf Y*

b

5

s* 2 anp s** 5 C C

(20)

where C 5 C 5 ((Kfnf) / (Y*) 1 (Kcnf) / (Y*) 1 (Kvnf) / (Y*)) 5 total capital stock–capacity ratio 5 (fixed capital stock 1 circulating capital stock 1

98

Strategic competition, dynamics, and the role of the State

finished goods inventory stocks)/capacity ratio, s* 5 social savings rate 5 private savings rate 1 government savings rate, and s** 5 social savings rate adjusted for purchases of investment goods by the banking and government sectors. The lower s** thus lowers Gw. Chapter 5 investigates policies to raise the warranted growth rate.

3.

CONCLUSION

A careful analysis of the extended Harrodian growth model should dispel the notion that Say’s law is assumed simply because the social savings rate has a positive impact on the warranted growth rate. The establishment of the short-run trend growth path is established via the joint interactions of supply and demand. Further, as the equations for the slow adjustment process show, both output (i.e. demand) and capacity adjust to discrepancies between themselves (see Appendix 2 to this chapter) in order to establish the warranted growth rate. In contrast, in NEGT models the actual growth path is determined by the growth of capacity, while the latter is in turn determined by a vector of supply-side variables (Roberts and Setterfield, 2007). Aggregate demand generally does not occupy a central place in these models, and full employment at the NAIRU level is assumed, although some allowance is made for transitional technological unemployment because of the time it takes workers who were laid off from a plant with an old technology to move to another one with a new technology (Aghion and Howitt, 1998). On the other hand, in Post Keynesian models aggregate demand rules the roost by determining output and employment growth. In the extended Harrodian framework the social savings rate exerts its positive effect on the warranted path despite the fact that bank credit is endogenous and the real and financial sectors are consistently integrated in an SFC framework. The SFC framework shows that each sector’s savings are exactly equal to its change of net worth. Any financial disruptions will therefore affect the social savings rate and thus the growth rate. A scenario may be constructed which would be relevant to the current economic crisis. Suppose that the household sector’s borrowings from the banking sector have been increasing so that its liabilities increase faster than its assets (both change relative to output). This would lead to a fall in its savings rate. If at some point household bankruptcies increase because household debt rises faster than income, then the banking sector’s profit flows and thus savings rate will collapse. This drop in the social savings rate will thus lower the warranted growth rate.

A model of disequilibrium dynamics

99

NOTES 1. The expression Ye 5 Cnfe 1 Infe 1 =INVnfe is equivalent to equation A1 in Godley (1999, p. 404). 2. As is the practice, it is assumed that dividends from State-owned enterprises (SOEs) are paid to the Treasury. Since, following US national income and product accounts (NIPA) convention, SOEs are treated as part of the business sector their dividend payments are treated as a form of tax payment to the Treasury so as to maintain notational simplicity. 3. This category would include wages paid to government workers since those constitute a payment for their services. 4. For expositional purposes all superscripts indicating planned and expected variables have been suppressed. It must be remembered that all the variables are ex ante ones. 5. See Shaikh (2009) for a discussion of the treatment of circulating investment in national accounts. See also Shaikh (1989, pp. 66–7), which shows how fixed, circulating and finished goods inventory investment are obtained from the expression for excess demand. 6. These opposite short-run–long-run effects parallel those of Duménil and Lévy (1999).

100

Strategic competition, dynamics, and the role of the State

APPENDIX 1 The SAM can be used to derive expressions for the aggregate money supply and money demand. With regard to this derivation the reader’s attention is drawn to the argument made in the chapter that financial market disequilibria clear very rapidly because no production is involved. Thus in the ex ante SAM the differences between the planned demands and expected supplies of high-powered money, borrowed reserves, deposits, bonds, equity and loans can be taken to be zero. The planned change in the money supply (=Msp) occurs via the expected injection of bank credit into the private and public sectors plus the net injection of high-powered money (=Hnetp) by the fiscal policies of the State. Note that, equivalently, the private component of the planned money supply increase arises from the planned demand for loans by households (=Lhp) and non-financial firms (=Lnfp). Thus the expected loan supply =Lbe 5 =Lhp 1 =Lnfp. The expected supply of credit by banks to the private sector (=Lbe) as well as the planned purchases of government bonds by banks (=BGbp) originates in banks’ excess reserves. The total expected supply of bank credit =Ltotale and high-powered money =Hnetp is given by: =Ltotale 5 =Lbe 1 =BGbp 5 (=Lhp 1 =Lnfp) 1 =BGbp

(A1.1)

=Hnetp 5 =Hge − =BRbp

(A1.2)

Another way of interpreting equation A1.2 is that the total expected additions to high-powered money (=Hge) occur via injections from a fiscal deficit (=Hnetp) plus the planned demand for borrowed reserves by commercial banks from the central bank (=BRbp). It will be assumed that the expected supply of borrowed reserves by the central bank (=BRge) is demand-driven so that =BRge 5 =BRbp. That is, the central bank is treated as a lender of last resort that provides discount window loans on demand in order to maintain stability of the financial system. This is consistent with the Post Keynesian view of central banking. The planned change in the money supply is given by adding equations A1.1 and A1.2: =Msp 5 =Ltotale 1 =Hnetp 5 (=Lbe 1 =BGbp) 1 (=Hge − =BRbp) 5 (=Lhp 1 =Lnfp 1 =BGbp) 1 (=Hge − =BRge) (A1.3) Banks’ expected supply of deposits is largely determined by expected loan supply (loans create deposits, as the Post Keynesian tradition has

A model of disequilibrium dynamics

101

emphasized). Since the money supply is by definition equal to cash plus deposit holdings, the above planned money supply exists in the form of the expected supply of deposits by banks (=Dbe) to households and nonfinancial firms and the planned cash holdings of households and nonfinancial firms (=Hhp 1 =Hnfp): =Msp ; =Hhp 1 =Hnfp 1 =Dbe

(A1.4)

By definition planned high-powered money exists in the form of planned cash holdings of households and non-financial firms (=Hhp 1 =Hnfp) and the reserves of the banking sector (=Hbp). This will be equal to the expected injection of high-powered money (=Hge): =Hge 5 =Hhp 1 =Hnfp 1 =Hbp

(A1.5)

Combining equations A1.3, A1.4 and A1.5 we get: =Dbe 5 =Lbe 1 =BGbp 1 (=Hbp − =BRbp) 5 (=Lhp 1 =Lnfp 1 =BGbp) 1 (=Hbp − =BRge)

(A1.6)

ABR Version 1 Since the SAM implies that planned social savings S*p 5 change in planned aggregate net worth, it follows that S*p 5 planned aggregate investment 1 change in planned net financial asset accumulation. From the SAM’s capital account sub-total the change in planned aggregate net worth 5 =NWp 5 =NWhp 1 =NWnfp 1 =NWbp 1 =NWgp. From the SAM: S*p − =NWp 5 S*p − (=NWhp 1 =NWnfp 1 =NWbp 1 =NWgp) 5 0 (A1.7a) Therefore: S*p − [(Ifnfp 1 =INVnfp) 1 (Ifbp 1 =INVbp) 1 (Ifgp 1 =INVgp)] 1 [=Hge − (=Hhp 1 =Hnfp 1 =Hbp)] 1 (=BRbp − =BRge) 1 [=Dbe − (=Dhp 1 =Dnfp)] 1 [{(=BCnfe 1 =BCbe) − =BChp} 1 {=BGge − (=BGhp 1 =BGnfp 1 =BGbp)}] 1 [(=EQnfe 1 =EQbe) − =EQhp] 1 [(=Lhp 1 =Lnfp) − =Lbe] 5 0 (A1.7b) Let Ip 5 aggregate planned investment by all sectors 5 [(Ifnfp 1 =INVnfp) 1 (Ifbp 1 =INVbp) 1 (Ifgp 1 =INVgp)]. Then equation A1.7b can be rewritten as:

102

Strategic competition, dynamics, and the role of the State

S*p − Ip 1 [=Hge − (=Hhp 1 =Hnfp 1 =Hbp)] 1 (=BRbp − =BRge) 1 [=Dbe − (=Dhp 1 =Dnfp)] 1 [{(=BCnfe 1 =BCbe) − =BChp} 1 {=BGge − (=BGhp 1 =BGnfp 1 =BGbp)}] 1 [(=EQnfe 1 =EQbe) − =EQhp] 1 [(=Lhp 1 =Lnfp) − =Lbe] 5 0 (A1.7c) Substituting =Dbe 5 =Lbe 1 =BGbp 1 (=Hbp − =BRbp) from equation A1.6 into A1.7c we get: S*p − Ip 1 [(=Hge − =BRge) 1 =BGbp 1 (=Lhp 1 =Lnfp)] − [(=Hhp 1 =Hnfp) 1 (=Dhp 1 =Dnfp)] 1 [{(=BCnfe 1 =BCbe) − =BChp} 1 {=BGge − (=BGhp 1 =BGnfp 1 =BGbp)}] 1 [(=EQnfe 1 =EQbe) − =EQhp] 5 0

(A1.8)

Following equation A1.3 the term in the first brackets [(=Hge − =BRge) 1 (=Lhp 1 =Lnfp) 1 =BGbp] represents the planned change in the aggregate money supply =Msp. On the other hand, the second term in the brackets [(=Hhp 1 =Hnfp) 1 (=Dhp 1 =Dnfp)] represents the change in desired money holdings of the non-bank private sector (households and non-financial firms) =Mdp. Thus equation A1.8 becomes: S*p − Ip 1 =Msp − =Mdp 1 [{(=BCnfe 1 =BCfe) − =BChp} 1 {=BGge − (=BGhp 1 =BGnfp 1 =BGfp)}] 1 [(=EQnfe 1 =EQfe) − =EQhp] 5 0 (A1.9) Since the symbol for a variable X is its desired change relative to the beginning of a period (=X 5 Xt − Xt−1), the above equation can also be written in terms of stocks: (S*p − Ip) 1 (Msp − Mdp) 1 [{(BCnfe 1 BCfe) − BChp} 1 {BGge − (BGhp 1 BGnfp 1 BGfp)}] 1 [(EQnfe 1 EQfe) − EQhp] 5 0 (A1.10) The above equation can be rewritten as: (Ip − S*p) 1 [{BChp − (BCnfe 1 BCfe)}1 {BGhp − (BGnfe 1 BGfe)}] 1 [(EQnfe 1 EQfe) − EQhp] 5 Msp − Mdp (A1.11a) Assuming that financial market disequilibria clear very rapidly (since no production is involved), equation A1.11a reduces to:

A model of disequilibrium dynamics

(Ip − S*p) 5 Msp − Mdp

103

(A1.11b)

Let the State actually increase the money supply by the amount it desires in order to finance the budget deficit. Further, if credit is demand-determined, then the planned demand for credit will produce an actual increase in the endogenous component of the money supply. These two propositions imply that the planned money supply Msp 5 actual money Ms. Then equation A1.11b can be rewritten as: (Ip − S*p) 5 Ms − Mdp

(A1.12)

ABR Version 2 Alternatively, since all the row sums (‘ex ante disequilibria’) must add to zero we can also write an alternative equation for the ABR. Let E 5 excess demand in the market for goods and services 5 planned demand by customers − expected demand (sales) by non-financial firms 5 Yp − Ye. The column sum is given by: −E 1 [We − (Wnfp 1 Wbp)] 1 [DIVhe − (DIVnfp 1 DIVbp)] 1 [(IDhe 1 IDnfe) − IDbp] 1 [(IBGhe 1 IBGnfe 1 IBGbe) − IBGgp] 1 [IBChe − (IBCnfp 1 IBCbp)] 1 [ILbe − (ILhp 1 ILnfp)] 1 (IBRge − IBRbp) 1 [Tge − (Thp 1 Tnfp 1 Tbp)] 1 [=Hge − (=Hhp 1 =Hnfp 1 =Hbp)] 1 (=BRbp − =BRge) 1 [=Dbe − (=Dhp 1 =Dnfp)] 1 [(=BCnfe 1 =BCbe) − =BChp] 1 [=BGge − (=BGhp 1 =BGnfp 1 =BGbp)] 1 [(=EQnfe 1 =EQbe) − =EQhp] 1 [(=Lhp 1 =Lnfp) − =Lbe] 5 0 (A1.13) Substituting for =Dbe 5 =Lbe 1 =BGbp 1 (=Hbp − =BRbp) from A1.6 into equation A1.13 we get: −E 1 [We − (Wnfp 1 Wbp)] 1 [DIVhe − (DIVnfp 1 DIVbp)] 1 [(IDhe 1 IDnfe) − IDbp] 1 [(IBGhe 1 IBGnfe 1 IBGbe) − IBGgp] 1 [IBChe − (IBCnfp 1 IBCbp)] 1 [ILbe − (ILhp 1 ILnfp)] 1 (IBRge − IBRbp) 1 [Tge − (Thp 1 Tnfp 1 Tbp)] 1 [(=Hge − =BRge) 1 =BGbp (=Lhp 1 =Lnfp)] − (=Hhp 1 =Hnfp) − (=Dhp 1 =Dnfp) e 1 [(=BCnf 1 =BCbe) −=BChp] 1 [=BGge − (=BGhp 1 =BGnfp 1 =BGbp)] 1 [(=EQnfe 1 =EQbe) − =EQhp] 5 0 (A1.14) =Mdp 5 desired additions to money stocks of the non-bank private sector (i.e. desired change in money demand) 5 (=Hhp 1 =Hnfp) 1 (=Dhp 1

104

Strategic competition, dynamics, and the role of the State

=Dnfp) and, given equation A1.3 for the desired change in money supply (=Msp), equation A1.14 can be rewritten as: −E 1 (=Msp − =Mdp) 1 [We − (Wnfp 1 Wbp)] 1 [DIVhe − (DIVnfp 1 DIVbp)] 1 [(IDhe 1 IDnfe) − IDbp] 1 [(IBGhe 1 IBGnfe 1 IBGbe) − IBGgp] 1 [IBChe − (IBCnfp 1 IBCbp)] 1 [ILbe − (ILhp 1 ILnfp)] 1 (IBRge − IBRbp) e 1 [Tg − (Thp 1 Tnfp 1 Tbp)] 1 [(=BCnfe 1 =BCbe) − =BChp] 1 [=BGge − (=BGhp 1 =BGnfp 1 =BGbp)] 1 [(=EQnfe 1 =EQbe) − =EQhp] 5 0 (A1.15) Let us assume that, in the short run, interest, dividend, wage, tax, transfer and subsidy payments are determined in advance, primarily because these variables are to a large extent determined institutionally or contractually. Note that at a given wage rate w the expected wage payments by workers’ households We are wLe, and planned wage payments by firms Wp are wLp, where Le 5 expected supply of labor while Lp 5 planned labor demand. Then the condition Wp 5 We implies that Le 5 Lp. Authors in the classical and Keynesian traditions would interpret this equality by arguing that the expected supply of labor is determined by its planned demand by firms, thereby allowing for varying degrees of unemployment. Thus approximately (a) planned interest payments are equal to expected interest payments, (b) planned dividend payments are equal to expected dividend receipts, (c) expected wage receipts are equal to planned wage payments, and (d) planned tax payments are equal to expected tax receipts. Then equation A1.15 reduces to: −E 1 (=Msp − =Mdp) 1 [(=BCnfe 1 =BCbe) − =BChp] 1 [=BGge − (=BGhp 1 =BGnfp 1 =BGbp)] 1 [(=EQnfe 1 =EQbe) − =EQhp] 5 0

(A1.16)

or: E 1 [=BChp − (=BCnfe 1 =BCbe)] 1 [(=BGhp 1 =BGnfp 1 =BGbp) − =BGge] 1 [=EQhp − (=EQnfe 1 =EQbe)] 5 (=Msp − =Mdp) (A1.17) The LHS of equation A1.14 is the sum of the excess demands in three markets: goods and services, bonds and equity. Now, again, if financial market disequilibria adjust quickly relative to those in the real sector then equation A1.17 becomes: E 5 =Msp − =Mdp

(A1.18)

A model of disequilibrium dynamics

105

As above, this in turn implies that: E 5 Msp − Mdp

(A1.19)

Again, for reasons discussed above, if planned money supply (Msp) 5 actual money supply (Ms) then: E 5 Ms − Mdp

(A1.20)

Thus equations A1.12 and A1.20 jointly imply that: E 5 (Ip − S*p) 5 Ms − Mdp

(A1.21)

As long as the disequilibrium between aggregate demand and supply in the goods market persists, the actual money supply will differ from desired money holdings, a result which follows from the ex ante SAM and the assumption that financial market disequilibria clear very rapidly compared to the case in the real sector. Appendix 2 discusses the dynamics of excess demand in the real sector.

106

Strategic competition, dynamics, and the role of the State

APPENDIX 2 The fast and slow adjustment processes discussed below are a restatement of the model of cyclical growth in Shaikh (1989). The fast adjustment process It will be assumed that firms maintain a fixed proportion of their finished goods inventory stocks Kv relative to the stock of circulating capital Kcnf: Kvnf 5 v Kcnf. It follows therefore that: avnf 5 vacnf

(A2.1)

Repeating equation 12: e 5 acnf 1 avnf 1 afnf 1 anp – s* 5 acnf 1 avnf 1 afnf − (s* − anp) 5 acnf 1 avnf 1 afnf − s** (A2.2a) Combining equation A2.1 with A2.2a we get: e 5 (1 1 v)acnf 1 afnf − s**

(A2.2b)

We next need to specify the determinants of acnf and afnf. In the fast adjustment process the fixed investment share afnf is taken as given, so that it becomes a variable in the slow adjustment process. It may be recalled that the framework models two disequilibrium processes that have different speeds of adjustment. The disequilibria are respectively modeled by excess demand e and the capacity utilization rate u. These disequilibrium variables correspond to realized, short-run trend and long-run trend (or normal) profits, which are respectively PR, P and Pn (all of which are gross of finance charges): PR 5 P 1 E and P 5 Pnu

(A2.3)

As discussed in Chapter 2 the long-run trend corresponds to the situation when e 5 0 and u 5 1. Starting with aggregate demand 5 aggregate supply (e 5 0) and output 5 capacity (u 5 1), suppose that e . 0 and u . 1. Since realized profits PR/Y 5 P/Y 1 e, it follows that firms will be stimulated to increase production whenever PR/Y . P/Y and to reduce it when PR/Y , P/Y. Given the excess of planned investment over planned savings the finance gap has to be financed by higher levels of bank borrowing. It will be assumed that fresh bank loans (Lb) at the end of time period t−1 will entail finance charges (payment of principal 1 interest rate on the loan)

A model of disequilibrium dynamics

107

in period t. It is assumed that all loans are paid back within one period. Finance charges Ft at time t are thus: Ft 5 (1 1 i) Lbt21

(A2.4)

Internally available finance in every period, Zt, is realized profits net of finance charges: Zt 5 (Realized Profits in t−1) − (Finance Charges in t) (P 1 E) t 21 2 (1 1 i) Lbt21

(A2.5)

where i 5 interest rate. Firms are assumed to increase circulating investment (relative to the trend) whenever the realized profits net of finance charges exceed the short-run trend of profits: Icnf Icnf a b 2 a b 5 h1 [ { (P 1 E) t21 2 (1 1 i) Lbt21 2 Pt21 } /Pt21 ] P t P t 21 5 h1 [ Et 21 2 (1 1 i) Lbt21 ] /Pt 21

(A2.6)

Let the share of profits in output P/Y 5 g be a constant. Then the above equation becomes (in continuous terms): # acnf 5 h2 [ e 2 (1 1 i) lb ]

(A2.7)

where lb 5 Lb/Y and the dot over acnf represents the first derivative with respect to time. Since in equation A2.2b v and s* are parameters and afnf has a constant # # value afnf , in the short run it follows that e 5 (1 1 v) acnf and equation A2.7 becomes: # e 5 h3 [ e 2 (1 1 i) lb ]

(A2.8)

where h3 5 h2(1 1 v). It was shown in equation 11b that excess demand in the goods/services market is given by: E 5 (Ifnfp 1 Icnfp 1 Ivnfp) − S**p

(A2.9)

Following the endogenous money approach bank credit fuels the excess of investment over savings. Shaikh’s (1989) variant to this argument is that this injection of credit (=Lbt) has to be net of finance charges Ft (equation A2.4):

108

Strategic competition, dynamics, and the role of the State

Et 5 =Lbt − Ft 5 =Lbt − (1 1 i)Lbt−1

(A2.10)

Thus Lbt 5 Et 1 (1 1 i)Lbt−1. Since Lbt 5 Lbt−1 1 =Lbt and Et 5 Et−1 1 =Et in continuous terms equation A2.10 becomes: # # Lb 5 (1 1 i) Lb 1 E 1 E

(A2.11)

After equation of the # dividing # # A2.11 by Y# and # making substitutions # form lb 5 (Lb) / (Y) 2 (Y) / (Y) lb and e 5 (E) / (Y) 2 (Y) / (Y) e, equation A2.11 can be rewritten as: # # # Y lb 5 E 1 ilb 1 e 1 (e 2 lb) (A2.12) Y # In continuous terms equation 15 is Y /Y 5 (1/m) acnf . Equation A2.2b can be rewritten as acnf 5 (e 1 s** 2 afnf)/(1 1 v). Then equation A2.12 can be rewritten as: (e 1 s** 2 afnf) # # (e 2 lb) lb 5 e 1 ilb 1 e 1 m (1 1 v)

(A2.13)

Let mr 5 1/m (1 1 v) so that: # # lb 5 e 1 ilb 1 e 1 mr (e 1 s** 2 afnf) (e 2 lb)

(A2.14)

Let d 5 s** 2 afnf. Equation A2.14 can be rewritten as: # # lb 5 e 1 (1 1 i) e 1 (mrd 2 i) (e 2 lb) 1 mre (e 2 lb)

(A2.15)

The core differential equation system representing the fast adjustment process is: # e 5 h3 [ e 2 (1 1 i) lb ] # # lb 5 e 1 (1 1 i) e 1 (mrd 2 i) (e 2 lb) 1 mre (e 2 lb)

(A2.8) (A2.15)

Finally it will be assumed that the interest rate on bank loans i varies procyclically with the demand for credit: # # i 5 h4lb where h4 . 0. Let z 5 e − lb. Equations A2.8 and A2.15 become:

(A2.16)

A model of disequilibrium dynamics

109

# e 5 2h3ie 1 h (1 1 i) z

(A2.17)

# z 5 2 (1 1 i) e 2 (mrd 2 i) z 2 mrez

(A2.18)

One way to provide a tractable solution to the three-dimensional system of equations (A2.8, A2.15 and A2.16) is to assume that suitable monetary policies stabilize the interest rate over the short-run cycle. Then the system of equations reduces to a two-dimensional one comprising A2.17 and A2.18. This system of equations has the following critical points: a. e 5 z 5 0 b. z 5 21 2 (1 1 mrd)i/(1 1 i), mre 5 (1 2 i)/I 2 (1 1 mrd) The Jacobian of the above system is: c

2hi 2 (1 1 i) 2 mrz

h (1 1 i) d 2 (mrd 2 i) 2 mre

If the Jacobian is linearized around the second critical point (b) the determinant is negative. However, linearization around the first critical point (a) gives a positive determinant and a negative trace. Around e 5 z 5 0, the trace 5 −[hi 1 (mrd − i)] is negative, as h, i, mr and d are positive. Given the signs of these parameters a sufficient condition for local stability is mrd ≥ i. The stability condition mrd ≥ i has an economic meaning. From equation 2.2b when e 5 0: (1 1 v)acnf 1 afnf 2 s** 5 0

(A2.19)

Since d 5 s** 2 afnf then acnf 5 d/(1 1 v). Equation 15 in continuous terms is: # Y 1 5 acnf (A2.20) m Y Since mr 5 1/m (1 1 v) then: # Y d 5 5 mrd (A2.21) Y m (1 1 v) # Thus stability requires that Y/Y ≥ i. Now# if profi # ts P are a constant proportion of output Y then it follows that P /P 5 Y /Y and the stability con# dition is equivalently P/P ≥ i, i.e. the growth rate of the mass of profits has to be greater than or equal to the interest rate.

110

Strategic competition, dynamics, and the role of the State

The slow adjustment process In the slow adjustment process excess demand is approximately zero. From equation A2.2b: (1 1 v)acnf 1 afnf 2 s** 5 0

(A2.19)

Thus: acnf 5

s** 2 afnf 11v

(A2.22)

Firms attempt to bring capacity into line with output by adjusting their capital stock: # afnf 5 h5 (u 2 u*) 5 h5 (u 2 1) (A2.23) afnf That is: # afnf 5 h5 (u 2 1) afnf

(A2.23)

In this equation the normal (or target) rate of capacity utilization u* has been set equal to 1 so that u 5 u* 5 1 when output (Y) 5 capacity (Y*). This equation states that whenever there is excess capacity (u , 1) firms reduce fixed investment because they desire to decrease capacity; conversely they desire to increase fixed investment and capacity when u . 1. This equation is at the heart of the Harrodian view that firms adjust fixed investment in response to demand so as to bring capacity into line with output (i.e. attain the practicable optimum in the Andrews/Brunner sense). It should be contrasted with Post Keynesian fixed investment functions in which the rate of fixed capital accumulation is a function of the level of capacity utilization. This leads to the implication that firms add to their production capacity even if they have considerable degrees of excess capacity to begin with, a response which is difficult to justify on empirical grounds. With regard to equation A2.23 it should be noted that when u 5 1 the condition that the growth rate of the fixed investment share is zero is consistent with a growing economy (i.e. trended variables) in which the level of fixed investment would be growing at the same rate as the level of output. This specification was at the heart of a debate between Harrod and Keynes regarding dynamic versus static specifications of the economy in which the former insisted on specifying all variables in growth terms, much to the latter’s mystification (Kregel, 1980). The capacity utilization rate u 5 Y/Y* where Y* 5 capacity. The fixed

A model of disequilibrium dynamics

111

capital–capacity ratio is k 5 Kfnf /Y* (equation 14). Given these relationships and equation A2.20 it follows that: # # # # Kfnf Ifnf Ifnf Y* Y u Y* 1 1 1 Y 5 acnf 2 5 acnf 2 5 acnf 2 5 2 u m m m Y Y* Kfnf Kfnf Y Kfnf Y* 5

(s** 2 af) afnf u afnf u 1 5 2 acnf 2 m k k m (1 1 v)

(A2.24)

Since mr 5 1/m (1 1 v) , equation A2.24 becomes: afnf u2 # u 5 mr (s** 2 afnf) u 2 k

(A2.25)

The slow adjustment process is represented by: afnf u2 # u 5 mr (s** 2 afnf) u 2 k # afnf 5 h5 (u 2 1) afnf

(A2.25) (A2.23)

Since u changes with output and fixed investment adds to capacity, the equations A2.25 and A2.23 capture the mutual interactions of output and capacity. This system has the following critical points: a. u 5 1, afnf 5 mrs**k/mrk 1 1 b. u 5 0, afnf 5 0 The Jacobian of the system A2.23 and A2.25 is: £

e mr (s** 2 afnf) 2 h5afnf

2afnf u k

f

e 2mru 2

u2 f k §

h5u 2 h5

If the Jacobian is linearized around the second critical point (b) the determinant (5 2mrs**h5) is negative and so this critical point is unstable. On the other hand, around the first critical point (a) the trace 5 (mrs**) / (mrk 1 1) , 0 and the determinant 5 h5mrs** . 0. Thus the system is stable around the critical point u 5 1, afnf 5 (mrs**k) / (mrk 1 1) , i.e. the warranted growth path. An expression for the warranted growth rate can be derived by combining this expression for the fixed investment share with equations A2.20 and A2.22. Along the warranted growth rate (Gw) the growth of output equals the growth of capacity: # # Y Y* 1 1 s** 2 afnf s** 1 5 5 Gw 5 acnf 5 c 2 a d 5 m m 11v Y Y* m (1 1 v) m (1 1 v) fnf

112

Strategic competition, dynamics, and the role of the State

5 mrs** 2 mr c

mrs**k mrs** 5 d5 mrk 1 1 mrk 1 1

s**

s** 5 1 k 1 m (1 1 v) k1 (A2.26) mr In this expression mv 5 (Kcnf) / (Y) v 5 Kvnf /Y) is the stock of finished goods inventory/output ratio. Thus the equation for the warranted growth rate is: s* 2 anp s** 5 (A2.27) Gw 5 Kfnf Kcnf Kvnf Kfnf Kcnf Kvnf 1 1 1 1 a b a b Y* Y* Y* Y* Y* Y* This equation is virtually identical to Harrod’s original equation for the warranted growth rate. In this equation it must be remembered that s* 5 private (household 1 non-financial firms 1 banks) savings rate 1 government savings rate, while anp is the purchase of investment goods by banks and the government. The interactions of fixed and circulating investments provide the mechanism to stabilize the warranted growth rate. Now suppose that u . 1. From equation A2.23, afnf will start to rise so that capacity will grow faster. On the other hand, given v and s** in equation A2.19, the equality between demand and supply will require acnf to fall, leading to a deceleration of actual output. The consequence will be that the capacity utilization will tend to move back toward its normal level and converge to it in a cyclical manner with constant over- and undershooting. In passing, it may be noted that in his article ‘Supplement on Dynamic Theory’ (1952, p. 289) Harrod has an equation for the warranted growth rate which is virtually identical to equation A2.27 in that circulating and fixed investments are explicitly distinguished. But in order to derive this equation from the condition that investment 5 savings Harrod needed to have implicitly assumed the links between circulating investment and output and fixed investment and capacity, respectively, i.e. equations 13 and 14. Thus, while Shaikh (1989) was the first one to show formally the stability of the warranted path, there was implicit to Harrod’s own framework a solution to his famous knife-edge problem. Finally, note that the extended-Harrodian model above is not based on the accelerator as in Harrod’s original model. In this connection it is important to point out that in Harrod’s model investment is not regulated by profitability relative to the interest rate, as in Keynes and Marx. However, in his later work (Harrod, 1969, ch. 8) Harrod’s treatment of investment had shifted with profits having a positive effect on investment (ibid., p. 186) and interest having a negative effective because of the higher cost of finance (ibid., p. 194).

5. 1.

Warranted growth and the role of the State INTRODUCTION

Conventional discussions of Harrod’s (and Domar’s) growth framework generally emphasize their Keynesian features and tend to focus on the modeling of instability. While these aspects of the Harrod–Domar model are important, this secondary literature generally does not deal with the policy aspects of these authors’ framework, especially the fact that the so-called paradox of thrift is overturned in the long run, although this is a feature of their models which both authors emphasized. As shown analytically in the previous chapter, Harrod’s position was that a change in the savings rate has opposite effects on the cyclical and trend components of growth. It is worth quoting at length from an undated letter that Harrod wrote to Joan Robinson regarding the relationship between savings, investment and growth (Besomi, 2006, p. 29): Your letter continues to ignore the vital distinction between movements in actual growth and movements in warranted growth – both being quite different from natural growth, which is the essence of my theory. An increase at a point of time in the ‘desire of firms to accumulate’ is a depressant of actual growth. This is Keynes, and I remain a Keynesian in this respect. An increase in the desire of firms to accumulate, to the extent that this is not ephemeral and shortly to be reversed, raises the warranted rate. This is neither Keynesian nor anti-Keynesian, because it is a dynamic principle, and there is no dynamics in Keynes. I explained that there is no dynamics in Keynes in my lecture to the Econometric Society (later published in Econometrica), in 1936. I suppose that what you call the ‘desire of firms to accumulate’ is what I would call ‘savings by firms’. This is of course something quite different from the amount of net capital formation that firms authorise. The latter may be greater or less than their saving. It may be greater because firms may finance part of their net capital formation by raising money from the public by the issue of bonds or shares; thereby they mobilise for their own purposes the savings of such people. Alternatively, they may authorize an amount of capital formation which is less than what they save; in that case they may hold the difference as a cash balance or add to their portfolio of securities.

In his final book, Economic Dynamics (1973), Harrod calls the opposite trend-cycle effects of a change in the social savings rate the central paradox 113

114

Strategic competition, dynamics, and the role of the State

(ibid., p. 102) of expansionary policies. In the chapter entitled ‘Problems and Conflicts’ in this book, he first informs the reader that ‘In the determination of the warranted rate the amount of saving that people want to make is, so to say, the lord and master’ (ibid., p. 100) and then devotes the bulk of this chapter to the analysis of different policy scenarios that lead to his central paradox. As with the following chapter (‘Foreign Trade’), Harrod makes it clear that in the case of undersaving countries (those in which warranted growth rates are below the natural or full employment growth rate) public saving has to supplement private saving in order to hoist up the warranted rate. In the previous chapter it was shown that Harrod’s fundamental insights regarding growth and cycles could be formalized in a non-linear model of cyclical growth in which cycles are treated as disequilibrium growth paths (between investment and savings and output and capacity, respectively). Since the trend and cycles are intrinsically interconnected, then, provided the disequilibria are not knife-edge unstable, a decrease in the savings rate will create a cyclical stimulus while eventually leading to a slowdown of the trend rate of growth and thus ultimately of the warranted growth path. It is worth emphasizing that stability is the key issue here. If such cycles are stable then the trend investment share will be regulated by the savings rate; otherwise there will be a runaway process.1 A central, though controversial, contention of this book is that the role of the social savings rate does not imply either an acceptance of Say’s law or the claim that the positive role of savings applies only to early capitalism with an undeveloped banking system (Chick, 1983, pp. 185–6). As shown in the previous chapter, not only is the social savings rate derived from a full articulated stock-flow system with endogenous credit, but disequilibria in both the fast and the slow adjustment processes involve bi-directional interactions between supply and demand as well as output and capacity, respectively. With regard to taxation policies, Harrod in Economic Dynamics argues that taxation policy should vary depending on whether Gw , Gn or Gw . Gn – but in a manner which may not seem obvious at first glance: The vital time to apply reflation in an economy of excess savings is when it has reached the upper limit of the boom and is on the full employment ceiling. The Government should start running sufficient Budget deficits by reducing taxes to offset the excess savings by persons and companies. . . . There is a paradox involved here, to which there may be mental resistance. The foot should be put on the accelerator when unemployment is at its minimum level. The view of the ‘man-in-the-street’ probably is that the foot should be put on the accelerator, when the economy is in recession and unemployment is increasing, a view that seems natural and plausible. But it is wrong. The foot should be put

Warranted growth and the role of the State

115

on the accelerator when unemployment is still at a minimum. (Harrod, 1973, pp. 105–106)

The idea is that in order to reduce inflationary pressures when Gw . Gn the higher budget deficit will tend to lower Gw and thereby reduce these pressures. On the other hand, in the event that Gw , Gn Harrod argues: In the case of undersaving countries (i.e. warranted growth less than natural growth) . . . it is desirable that private saving should be supplemented by government saving. The latter can be achieved only through reducing, by extra taxation, the purchasing power in the hands of citizens. (Harrod, 1973, p. 136, emphasis added)

This policy would presumably work if the higher tax rate raised the social savings rate, although it is unclear from Harrod’s analysis how this result will ensue. Harrod also introduces the role of public investment in the situation when Gw , Gn, an aspect of his policy framework that he shared with other authors in the Keynesian tradition.2 In discussing the situation in which Gw . Gn Harrod argues that in such situations in which private saving is insufficient to give a warranted rate of growth equal to what the economy is capable of, it should be supplemented by official saving and official investment of like amount. A mere Budget surplus will not cause countries in these countries to move in the right way. A parallel increase of investment is also required. In the foregoing, I have referred to this investment as investment by the official authorities. (ibid., p. 115)

In fact he suggests that the government should raise taxes to run a surplus and ‘use its surplus to make investments on its own account’ (ibid., p. 137). This policy would appear to be similar to the ones proposed by Keynes regarding the relevance of the composition of government spending. Some authors (Smithin, 1989; Kregel, 1993; Brown-Collier and Collier, 1995) have argued that Keynes was not ‘a wild-eyed deficit spender’ (Kregel, 1993, p. 436) but in fact advocated the separation of an ‘ordinary’ (current) budget from a capital budget. The surplus on the latter would be used to finance public capital formation and/or pay down the public debt. This important link role of capital budgeting was also emphasized by Musgrave and Musgrave (1973). Harrod does not elaborate on his tax-cum-public-investment policies, although they do raise some important questions. For example, which sector(s) should bear the burden of higher taxes, given that Harrod notes that higher taxes on the business sector might ‘diminish the incentive of industry to invest’ (1973, p. 137)? Furthermore, it is unclear how

116

Strategic competition, dynamics, and the role of the State

public investment is likely to raise the warranted growth rate. It must be stressed here that, outside a warranted growth rate context, the composition of government spending is of no relevance, since any type of public expenditures will raise the growth rate (Godley and Lavoie, 2007).

2.

HARROD’S POLICY INSIGHTS: SOLUTIONS TO AMBIGUITIES AND CONTRADICTIONS

2.1

Taxation Policies

In this subsection I will abstract from issues regarding the composition of public expenditures. Consider the warranted growth rate equation derived in Chapter 4: s* 2 anp s** Gw 5 5 (1) m1v1k C where s* 5 social savings rate, s** 5 social savings rate adjusted for purchase of investment goods by banks, C 5 (Kfnf) / (Y*) 1 (1 1 v) Kcnf*Y* 5 total capital–capacity ratio, Kfnf 5 fixed capital stock, Kcnf 5 circulating capital stock, v 5 finished goods/output ratio, m 5 Kcnf/Y*, and k 5 Kfnf/Y*. To begin with let us ignore the sectoral tax rates and assume for simplicity that the aggregate private sector (households, non-financial firms and banks) pays a tax rate q, so that the social saving rate s* is: s* 5 sp (1 2 q) 1 q 2 g

(2)

where sp 5 pre-tax private saving rate and g 5 government/output ratio. Thus the expression for the warranted growth rate is: sp (1 2 q) 1 q 2 g 2 anp Gw 5 (3) C In the event of a balanced budget (q 5 g) 0Gw/0q 5 2sp , 0. In this situation the higher tax rate reduces private savings, and the balanced budget condition ensures that the destruction of private savings is accompanied by an increase in the government consumption rate that equals the tax rate increase. Thus the leakage rate falls, leading to a fall in the warranted growth rate. However, with an unbalanced budget (q ≠ g) 0Gw /0 q 5 1 2 sp . 0. Thus a higher aggregate tax rate will raise the warranted growth rate only in the event of an unbalanced budget. In this situation an increase in q destroys private savings, but this corresponds to

Warranted growth and the role of the State

117

an increase in government savings (in the event of an initial surplus) or a decrease in government dissavings (in the event of an initial deficit). Thus the leakage rate rises. One may thus interpret Harrod’s tax policy as one which would have the effect of raising Gw only if the budget is unbalanced. Furthermore, higher values of q could be accompanied by a relatively slower increase of the government spending share g, which would have the effect of raising the warranted growth rate. Such a policy is an important one if the goal is to increase various types of social expenditures to strengthen the social safety net as well as public investments. It should be contrasted with neo-liberal policies that generally involve tax cuts followed by cuts in government spending. The issue of social expenditures to target poverty suggests that the household sector in the SAM in Chapter 4 needs to be disaggregated in order to separate out working-class from capitalist-class households, a distinction made by many authors, such as Kaldor and Pasinetti, in the treatment of growth. Such a class-based distinction of household income allows for an identification of the distributional consequences of taxation policies. We will start the analysis by aggregating the different types of income represented in the SAM in Chapter 4 (to simplify the notation we will ignore the distinction between planned and expected variables). Total interest on deposits and government bonds (IDh and IBGh respectively) is earned by both capitalist- and working-class households. Interest on corporate bonds and dividends are earned only by capitalist-class households. Moreover, both types of households make interest payments on their bank loans. Finally, by definition, wages are earned by workingclass households. Thus pre-tax income for each type of household (represented by the subscripts ‘w’ and ‘c’ respectively) includes net interest earnings: Working-class household income 5 Ywh 5 W 1 IDwh 1 IBGwh − ILwh (W 5 wages, IDwh 5 interest on bank deposits, IBGch 5 interest on government bonds, and ILwh 5 interest payments on bank loans; typically, in developing countries IDwh, IBGwh and possibly ILwh will be zero). ii. Capitalist-class household income 5 Ych 5 DIV 1 IDch 1 IBCch 1 IBGch − ILch (DIV 5 dividends, IDch 5 interest on bank deposits, IBCch 5 interest on corporate bonds, IBGch 5 interest on government bonds, and ILch 5 interest payments on bank loans). If INTch 5 IDch 1 IBCch 1 IBGch − ILch 5 net interest earned by capitalist-class households then Ych 5 DIV 1 INTch.

i.

118

Strategic competition, dynamics, and the role of the State

iii. The pre-tax profit flows of financial and non-financial firms are combined and represented by P. iv. Capitalist- and working-class households pay a consumption tax (indirect tax) in addition to the direct taxes on their respective incomes. In order to investigate the impact of different types of tax rates on growth we begin with a tax function that is somewhat inspired by Pasinetti (1989) but has the following additional features: (a) dividends net of aftertax profits, PN, accrue only to capitalist-class households: a part of profits is retained within (non-financial and banking) firms, Sf, while the rest is disbursed as dividend payments to capitalist-class households, that is PN 5 Sf 1 DIV; and (b) government spending G is some fraction g of output Y: G 5 gY.3 Pasinetti’s taxation equation includes both direct and indirect taxes on working-class and capitalist-class households. The direct tax rates are tw, tc and tp, while ti (0 ≤ ti , 1) is a proportional (direct) tax on all consumption expenditures of households. The taxation function T is given by:4 T 5 tpP 1 tcYch 1 twYwh 1 ti[(1 − sc)(1 − tc)Ych 1 (1 − sw)(1 − tw)Ywh] (4) Here tw, tc and tp are taxes on working-class households, capitalist-class households and firms respectively, while P is profits gross of taxes. As shown in Appendix 1 to this chapter, the social savings rate is now: s* 5 [r(1 − tp) 1 sc(1 − tc)d(1 − tp) 1 tp 1 tcd(1 − tp) 1 ti(1 − sc)(1 − tc)d(1 − tp)]a 1 [sc(1 − tc) 1 tc 1 ti(1 − sc)(1 − tc)]x 1 [sw(1 − tw) 1 tw 1 ti(1 − sw)(1 − tw)]g − g

(5)

In this equation r 5 retained earnings rate, d 5 dividend payout rate, a 5 share of before-tax profits 5 P/Y, x 5 share of net interest earned by capitalist households INTch/Y, g 5 share of working-class household income 5 Ywh/Y, and g 5 G/Y. Thus dividends DIV 5 d(1 − tp)P. Therefore b 5 share of capitalist-class household income Ych/Y 5 DIV/Y 1 INTch/Y 5 d(1 − tp)P/Y 1 INTch/Y 5 d(1 − tp)a 1 x. As shown in Appendix 1 the taxation-adjusted equation for the warranted growth rate (equations 1 and 5) shows that (a) an increase in any of the four types of tax rates will raise Gw,5 (b) an increase in the retained earnings rate will raise Gw, a result which is consistent with Harrod’s argument,6 and (c) an increase in the government spending share will lower Gw. It may be objected, as Harrod pointed out, that an increase in business taxes may reduce the motivation to invest. Thus one way of making sense of Harrod’s tax policy proposal is with regard to the composition of taxes:

Warranted growth and the role of the State

119

the aggregate tax rate q (5 T/Y) could be raised by keeping the business tax rate tp constant and raising household direct and indirect tax rates (tc and ti, respectively). Distributional concerns and poverty reduction goals would be satisfied by imposing different consumption taxes on working-class and capitalistclass households. While perhaps difficult politically at times, the aggregate tax rate could be raised by increasing direct and indirect taxes on capitalist-class households while keeping constant or lowering business taxes, as argued by Kaldor (1956). Dividend payout rates may at least partially be responsive to business– personal tax rate differentials (Damodaran, 2003) where the personal tax rates refer to those borne by capitalist-class households, which are the ones that own equity. Feldstein (1974) reports that, since the introduction of the US Social Security Act of 1937, increased marginal personal tax rates have induced firms to decrease dividend payout rates. He finds similar evidence for British firms (Feldstein, 1970). Pechman (1987) also finds that for the period 1929–86 higher personal income tax rates lowered dividend payout rates.7 In terms of the above model, these findings would suggest that a tax policy in which the business tax rate (tp) is kept constant while the capitalist-class household tax rate (tc) is raised would increase r and thus Gw. Finally, while the above tax function is quite parsimonious, in reality of course government tax revenues arise from many sources. For example, low business taxes could be combined with higher taxes elsewhere and/or the closure of tax loopholes, which include elimination of the deduction of mortgage interest on the personal income taxes of wealthy households and deductions on fringe benefits (Pechman, 1989; Gramlich, 1992; Pechman and McPherson, 1992). Pechman, for example, uses a broad definition of income, which includes non-wage income (such as capital gains and gratuitous receipts such as transfer payments, gifts and inheritances), and observes that tax loopholes on these lead to quite considerable loss of government revenue. 2.2

The Role of Public Investment

As with other authors in the early Keynesian tradition Harrod emphasized the importance of public investment in raising output and employment. Yet it is unclear how an increase in public investment will raise the warranted growth rate. For the Post Keynesian tradition, which deals with non-warranted growth rates, an increase in public investment will raise the growth rate by stimulating demand. However, if excess capacity is persistent then any type of public expenditure will have a positive effect.

120

Strategic competition, dynamics, and the role of the State

This would explain why in contemporary Post Keynesian models (see for example Godley and Lavoie, 2007) public investment plays no special role in the treatment of fiscal policy. In order to fully comprehend the implications of Harrod’s public investment proposals it is instructive to digress a little in order to understand the effects of autonomous investments in general on the warranted growth path. For Harrod autonomous investment is ‘conceived to be quite independent both of the current level of income and its current rate of growth’ (Harrod, 1970, pp. 57–8). Thus an increase in such investment will have only one effect: It may be noticed that the larger the volume of outlay which will be sustained independently of the current rate of growth, the smaller is the warranted rate of growth. A larger part of savings being absorbed in such outlay, there will be a smaller part to be looked after by the acceleration principle. (Harrod, 1970, p. 58)

Harrod shows this result to be true from the following equation: Gw 5

s 2 i aut C

(6)

where iaut is autonomous investment. In its definition of public investment the US national income and product accounts (NIPA) include direct production activities by government enterprises as well as the purchase of equipment, software and structures by the government from the private sector. From the warranted growth rate standpoint these two types of investments will have opposite effects. Economically, the government’s decision to purchase the services of, say, a private construction company to build a road is no different from its purchase of food to feed soldiers. The fact that the output supplied and ultimately consumed in the one case is more durable than the other does not take away from the fact that both are purchases from the private sector and will thus entail a reduction of the social savings rate. On the other hand, production by State-owned enterprises (SOEs) will expand output, since their actions cannot be distinguished economically from those of private firms. We must thus split up total government expenditures G into three components: government consumption expenditures (Cg), government purchases of equipment and structures from the private sector (Igaut), and government production activities by SOEs (Igp). The last component is treated as part of the business sector and comprises circulating, fixed and finished goods investments. Appendix 2 to this chapter derives the equations for the warranted growth rate and the social savings rate. The expression for the warranted growth rate is:

Warranted growth and the role of the State

Gw 5

s* 2 (Ib/Y) 2 (Iaut g /Y) C

121

(7)

The terms Ib/Y and Igaut/Y are the purchases of investment goods from non-financial firms by banks and the government, respectively. The social savings rate s* is given by: s* 5 [r(1 − tp) 1 sc(1 − tc)d(1 − tp) 1 tp 1 tcd(1 − tp) 1 ti(1 − sc)(1 − tc)d(1 − tp)]a 1 [sc(1 − tc) 1 tc 1 ti(1 − sc)(1 − tc)]x 1 [sw(1 − tw) 1 tw 1 ti(1 − sw)(1 − tw)]g − cg (8) In equation 7, C 5 total capital stock–capacity ratio 5 (fixed capital stock 1 circulating capital stock 1 finished goods inventory stocks)/capacity for aggregate private and government firms. In the expression for the social savings rate the public sector savings rate is the aggregate tax rate q minus the public consumption share cg (Musgrave and Musgrave, 1973). As discussed by Keynes and Harrod, the relevance of capital budgeting can be seen by considering equations 7 and 8. Say there is a budget deficit and the government maintains a steady public debt/output ratio. Given cg (5 Cg/Y), a higher tax rate q would allow both forms of public investment to accelerate relative to the output trend. If Igaut/Y increases by less than the increase in q in equation 7, s* would rise faster than Igaut/Y: there would thus be an expansion in public sector production and therefore an increase in the warranted growth rate Gw. This way of analyzing the capital budgeting–growth nexus makes SOE production potentially complement private sector production. In a developing-country context in particular, the output of the SOE sector may provide crucial inputs into private sector production, which at an initial stage may be involved in low-value-added or light industry production. Public sector firms may steer the economy toward heavy or highvalue-added production by buying and selling crucial inputs from and to the private sector.8 If successful, such a strategy could alter the economy’s long-run growth trajectory.9 A high tax rate could finance such production activities as well as expand anti-poverty programs.10 Finally, if a country’s capitalist class sees such policies as being in its long-term interests then that would induce private firms to raise the retained earnings rate, which would further raise the warranted growth rate. We will conclude this section on the fiscal policy–growth nexus by briefly reviewing the current literature on this topic. The most comprehensive treatment of this issue from the Post Keynesian position is in Godley and Lavoie (2007, Chapter 11), in which, because of the persistence of excess capacity, a higher budget deficit raises the growth rate and the rate

122

Strategic competition, dynamics, and the role of the State

of capital accumulation.11 In this framework, the composition of government spending is of no direct relevance, since any kind of State expenditure will raise growth as long as excess capacity is persistent. From the neoclassical perspective, it must be remembered that changes in the social savings rate have no effect on growth in the Solow model but could affect it in the same direction in endogenous growth theory (NEGT). In the Ricardian Equivalence view (Barro, 1974, 1989), if individuals have rational foresight they will realize that the government will eventually have to pay for today’s permanent higher deficit by raising taxes in the future. Thus, in this rational expectations perspective, households will raise their own savings rate by an amount equal to the government dissavings rate. The consequence will be no change in the social savings rate and growth rate. In the neoclassical framework the composition of government expenditures does matter. For example, in the Barro (1990) model an increase in the tax rate in the context of a balanced budget has two effects. On the one hand it has a negative effect on growth by lowering the marginal product of capital; on the other, it raises growth, since the flow of government services is assumed to have a positive effect on private investment. At relatively low tax rates the positive effect dominates so that up to some limit raising the tax rate raises the growth rate. On the other hand, at relatively higher tax rates the negative effect dominates so that increasing tax rates lowers growth. Finally, in the context of the so-called AK model of endogenous growth of Rebelo (1991) an increase in the tax rate lowers the growth rate permanently.

3.

CONCLUSION

Following Keynes’s proposals many authors in the broad Keynesian tradition have discussed the importance of capital budgeting. This chapter has applied the relationship between capital budgeting and growth in the context of the warranted growth rate. Following an extension of Harrod’s (1973) policy proposals it has been shown how a tax-cum-publicinvestment policy, entailing a surplus on the current budget to finance public investments, can be used to raise the warranted growth rate. It has been shown that policymakers may have some flexibility with regard to taxation policy in order to raise the social savings rate. Further, taxation policy could be combined with a public investment strategy such that direct production activities by the State raise Gw. There may be additional positive effects if the presence of SOEs raises private sector ‘animal spirits’

Warranted growth and the role of the State

123

and induces the latter to raise the retained earnings rate. On a cautionary note, as with any type of interventionist State policies, the last-mentioned effect will depend on how the capitalist class views SOEs. Readers who are familiar with Kaldor’s policies will find much in the above proposals that is reminiscent of those made by him. Commenting on the development strategies in Chile, Kaldor observed: [The] high propensity to consume of the capitalist class [can be found in the fact that they] appear to have spent on personal consumption more than two-thirds of their gross income, or three-quarters of their net income after tax. In comparison to other countries the luxury consumption of the property-owning classes appears to take up an altogether disproportionate share of national resources, part of which would be automatically released for investment purposes if a more efficient system of progressive taxation were introduced and/or if effective measures were taken to encourage the retention of profits by enterprises. (Kaldor, 1956, p. 266, cited from Palma and Marcel, 1989, p. 250, emphasis added)

As Palma and Marcel (1989, p. 252) conclude: Kaldor’s main proposition from this point of view was that a developing country like Chile does generate a surplus large enough to sustain a level of investment needed for a fast rate of growth and high levels of employment. Nevertheless, too large a proportion of that surplus was wasted in luxury consumption by the high-income groups. . . . Government intervention, particularly through taxation and an effective investment policy by the public sector, was the most appropriate way to achieve a dynamic equilibrium. In other words, what Kaldor proposed were institutional changes that would make the Chilean public sector both a high-saving (through better taxation) and a high-investing affair.

In sum the extended Harrodian model shows that the State needs to play a central role if the goal is to raise long-run growth and employment. Thus the extended Harrodian framework is similar to the broad Post Keynesian one in that both emphasize the role of the State to raise output and employment. The Post Keynesian literature, however, emphasizes the role of budget deficits in achieving these outcomes. While higher deficits will certainly provide a short-run (cyclical) stimulus, as the previous chapter showed, the trend growth path will be partially regulated by the social savings rate in Harrod. Thus a higher budget deficit will not raise the warranted growth rate, and fiscal policies of the type investigated in this chapter would have to be implemented. Richard Goodwin summarized the policy issues well: The prime aim of Keynes was to persuade economists that demand, not supply, determined output and employment . . . in that he was dramatically successful. But when one considers more closely the possible dynamical consequences,

124

Strategic competition, dynamics, and the role of the State

problems arise. From any continuing alteration in public spending and taxing there will occur first acceleration, then to be followed by deceleration. . . . Harrod, who had followed the development of The General Theory, saw clearly that its basic shortcoming lay in the dynamical problem. . . . To lift employment to any desired degree of fullness and maintain it, then requires a very ambitious, dynamically variable policy. (Goodwin, 1997, pp. 162–3, emphasis added)

NOTES 1.

2.

3.

4.

5.

The stability of both types of cycles is crucial. The savings–investment imbalance needs to be stable to establish a short-run trend growth path, and it is the interaction between this trend growth path and the growth of capacity which is necessary for the warranted growth path to be established. Thus the starting point of this framework is the savings– investment disequilibrium. On the other hand, following Keynes in The General Theory, the Post Keynesian tradition begins by assuming that savings equal investment. Keynes came out in favor of a ‘comprehensive socialisation of investment’ (Keynes, 1953, p. 378), since for him it was not just a question of any type of demand injection but rather one that was productive: ‘I have been advocating government expenditure without much reference to the purpose to which the money is devoted. The predominant issue, as I look at the matter, is to get the money spent . . . [b]ut productive and socially useful expenditure is naturally to be preferred to unproductive expenditure’ (Keynes, 1982, cited from Smithin, 1989, p. 226, emphasis added). Domar (1957, p. 60) also emphasizes that public spending be preferably of the productive kind. In Pasinetti’s original model, a portion of profits also accrued to workers. Pasinetti’s original model also began with a government consumption function, G 5 (1 − sT)T, where sT is the government’s savings propensity and can be greater than or equal to or less than zero. Given sT, this, however, makes the government spending ratio, g, change endogenously in the same direction when the tax rate changes. The goal of the current chapter is to explore the possibility of changing g independently of the tax rate so that policymakers have two policy instruments to raise the warranted growth rate. Thus, the function G 5 gY is used. The government spending ratio is implicitly or explicitly treated as an exogenous parameter in all the early Keynesian growth models, such as those of Harrod, Domar and Robinson (Sen, 1970). Note that Pasinetti includes an indirect tax rate on government consumption expenditures. This would involve the payment of taxes to the Treasury by various government agencies that purchase goods and services from the private sector. In terms of the SAM this would involve an internal transfer within the government sector and not an additional injection of revenue into the aggregate government sector from other sectors. Therefore the taxation function used here (equation 4) does not have an indirect tax on government expenditures. The equivalence of the impact of different types of taxation was noted by R.F. Kahn in his analysis of Joan Robinson’s growth model: ‘There is a practical reason why it is important to consider the implication of the degree of thriftiness. In the usual models of economic growth, such as Mrs. Robinson’s golden age, the State is left out of the picture and in particular there is no room for saving by the State. Now, an economy might be developing as a golden age but, thriftiness being low, the methods of production are often so primitive, as a result of the scarcity of capital, that the word “golden” is a mockery. The practical question which then arises is whether the economy would not be in a happier condition if the State were contributing to saving out of the proceeds of higher taxes. For the purpose of the analysis one can regard such State saving as assimilated into the simpler model which disregards it. For in so far as it is financed by taxation of profits its effects are the same as those of a higher capitalists’ savings coefficient; in so

Warranted growth and the role of the State

6.

7. 8. 9.

10.

11.

125

far as it is financed by indirect taxation, it is equivalent in its effects to those of higher savings coefficients for both capitalists and wage-earners, to the extent of their consumption of taxed commodities’ (Kahn, 1970, p. 148, emphases added). This result is also true at the highest level of abstraction when there is no State. Then profits P 5 retained earnings RE 1 dividends DIV where RE 5 rP, DIV 5 dP, r 5 retained earnings rate, and d 5 dividend payout rate. Then, ignoring working-class household savings, the social savings rate s* 5 capitalist household savings rate 1 business savings rate 5 r(P/Y) 1 sc(P/Y) − sc r(P/Y) 5 ra 1 sca − scra. Therefore ∂s*/∂r 5 (1 − sc)a . 0, which will raise Gw. On the other hand, since the dividend payout d 5 1 − r, s* 5 a − da 1 scda, so that ∂s*/∂d 5 −(1 − sc)a , 0, which will lower Gw. The higher retention rates were further stimulated by investment tax credits and generous depreciation allowances. See also Pechman (1989) for further details on taxation policy. The provisioning of lower-cost materials inputs would lower average variables costs and thus possibly the coefficient m in equation A2.6. Other things being equal, this would raise the warranted growth rate. This argument is reminiscent of that made by Wade (1990, pp. 110–11), who observed that: ‘In many sectors, public enterprises have been used as the chosen instrument for a big push. This is true for the early years of fuels, chemicals, mining, metals, fertilizers, and food processing; but even in sectors where public enterprises did not dominate, such as textiles and plastics, the state aggressively led private producers in the early years. Later, during the late 1950s and 1960s, public enterprises accounted for a large part of total investment in synthetic fibers, metals, shipbuilding, and other industries. . . . public enterprises have often played a central role in creating new capacities’. For a thorough investigation of the efficiency of state-owned enterprises in developing countries see Chang and Singh (1993, 1997). Musgrave and Musgrave (1973, p. 732) propose higher excise taxes on the purchases of wealthy households (luxury consumption), and progressive property taxes. The surplus thus generated on the current account budget would be used to finance public investment projects. Also see Nurkse (1967), who discusses a similar link between public savings and investment. In contemporary heterodox economics, Seccareccia (1995, p. 75) makes a similar policy proposal regarding the link between luxury consumption and public investment. The inexplicable issue with their model is that this stimulus leads to a fall in the capacity utilization rate (Godley and Lavoie, 2007, p. 412, footnote 20). It is unclear what the economic meaning is of this result or of the outcome that a higher rate of capital accumulation is accompanied by a fall in the capacity utilization rate. This latter result contradicts their investment function, in which they adopt the standard Kaleckian argument that the capital accumulation rate is a positive function of the capacity utilization rate (ibid., p. 385, equation 11.7).

126

Strategic competition, dynamics, and the role of the State

APPENDIX 1 This appendix derives the social savings rate. For simplicity we will merge the savings of non-financial firms and banks into one variable, Sf, and assume that this aggregate firm sector pays a direct tax rate tp out of its profits P. We start with the tax function: T 5 tpP 1 tcYch 1 twYwh 1 ti[(1 − sc)(1 − tc)Ych 1 (1 − sw)(1 − tw)Ywh] (A1.1a) where P 5 profits gross of taxes of financial and non-financial firms, Ych 5 capitalist-class household income, Ywh 5 working-class household income, tp 5 taxes on business profits, tc 5 taxes on capitalist-class households, tw 5 taxes on working-class households, and ti 5 indirect taxes on the purchases of capitalist- and working-class households. Divide equation A1.1a through by Y and let a 5 P/Y, b 5 Ych/Y, g 5 Ywh/Y, and x 5 share of net interest earned by capitalist-class households INTch/Y where INTch 5 net interest earned. Let d 5 dividend payout rate and R 5 business retained earnings rate. Then DIV 5 d(1 − tp)P and b 5 Ych/Y 5 DIV/Y 1 INTch/Y 5 d(1 − tp) P/Y 1 INTch/Y 5 d(1 − tp)a 1 x. The tax rate q 5 T/Y is given by: q 5 tpa 1 tcb 1 tw g 1 ti[(1 − sc)(1 − tc)b 1 (1 − sw)(1 − tw)g] 5 tpa 1 tc[d(1 − tp)a 1 x] 1 tw g 1 ti[(1 − sc)(1 − tc)b 1 (1 − sw)(1 − tw)g] (A1.1b) where ∂q/∂tw . 0, ∂q/∂tP . 0, and ∂q/∂ti . 0. Beginning with the national income identity, Y 5 C 1 I 1 G, we subtract total taxes, T, and rearrange to give: (Y − T − C) 1 (T − G) 5 I

(A1.2)

Sp 1 Sg 5 I

(A1.3)

so that:

where Sp 5 private savings 5 savings of capitalist- and working-class households and of firms and Sg 5 government savings. Let Swh 5 workers’ household saving, Sch 5 capitalist household saving, and Sf 5 business savings so that Sp 5 Swh 1 Sch 1 Sf. Thus: Sf 1 Sch 1 Swh 1 Sg 5 S* 5 I

(A1.4)

Warranted growth and the role of the State

127

where S* 5 social savings. Let P 5 profits gross of taxes, PN 5 profits net of taxes, and tp 5 business tax rate so that PN 5 P − tpP. Then: Sf 5 rPN 5 r(1 − tp)P

(A1.5)

Since tc 5 capitalist household tax rate and Ych 5 DIV 1 INTch: Sch 5 sc(1 − tc)Ych 5 sc(1 − tc)DIV 1 sc(1 − tc)INTch

(A1.6)

If tw 5 working-class household tax rate then: Swh 5 sw(1 − tw)Ywh

(A1.7)

Finally, government savings Sg is: Sg 5 T − G

(A1.8)

Thus the social savings S* is: S* 5 r(1 − tp)P 1 sc(1 − tc)DIV 1 sc(1 − tc) INTch 1 sw(1 − tw) Ywh 1 (T − G) 5 r(1 − tp)P 1 sc(1 − tc)d(1 − tp)P 1 sc(1 − tc) INTch 1 sw(1 − tw)Ywh 1 (T − G) (A1.9) Substituting equation A1.1a into A1.9: S* 5 r(1 − tp)P 1 sc(1 − tc)d(1 − tp)P 1 sc(1 − tc) INTch 1 sw(1 − tw)Ywh 1 (T − G) 5 r(1 − tp)P 1 sc(1 − tc)d(1 − tp)P 1 sc(1 − tc) INTch 1 sw(1 − tw)Ywh 1 tpP 1 tcYch 1 twYwh 1 ti[(1 − sc)(1 − tc)Ych 1 (1 − sw)(1 − tw)Ywh] − G (A1.10) Rearranging equation A1.10: S* 5 [r(1 − tp) 1 sc(1 − tc)d(1 − tp) 1 tp 1 tcd(1 − tp) 1 ti(1 − sc)(1 − tc)d(1 − tp)]P 1 [sc(1 − tc) 1 tc 1 ti(1 − sc)(1 − tc)]INTch 1 [sw(1 − tw) 1 tw 1 ti(1 − sw)(1 − tw)]Ywh − G (A1.11) Dividing through by Y we get the following expression for the social savings rate:

128

Strategic competition, dynamics, and the role of the State

s* 5 S*/Y 5 [r(1 − tp) 1 sc(1 − tc)d(1 − tp) 1 tp 1 tcd(1 − tp) 1 ti(1 − sc)(1 − tc)d(1 − tp)]a 1 [sc(1 − tc) 1 tc 1 ti(1 − sc)(1 − tc)] x 1 [sw(1 − tw) 1 tw 1 ti(1 − sw)(1 − tw)]g − g (A1.12) As discussed in Chapter 4, in total investment I only that part carried out by non-financial firms contributes to aggregate output and capacity, while the other two are purchases of investment goods by banks and the government produced in the non-financial business sector (which includes State-owned enterprises). For non-financial firms, total investment 5 Inf 5 Ifnf 1 Icnf 1 Ivnf where Ifnf 5 gross fixed investment on new plant and equipment, Icnf 5 circulating investment and Ivnf 5 final goods inventory investment. Ignoring public investment for now (see Appendix 2) then if Ib is investment by the banking sector (5 purchase of fixed capital goods by banks from the non-financial business sector) the investment 5 savings relationship is: Ifnf 1 Icnf 1 Ivnf 1 Ib − S* 5 0

(A1.13)

Since Ib is a form of non-production expenditure it absorbs part of the savings (or social surplus). Thus equation A1.13 can equivalently be written as: Ifnf 1 Icnf 1 Ivnf 5 S* − Ib

(A1.14)

Let Icnf 5 circulating investment 5 mDY, Ifnf 5 fixed investment 5 kDY*, and Ivnf 5 vIcnf (m 5 circulating capital–output ratio, k 5 fixed capital–output ratio, and v 5 finished goods inventory to materials inventory ratio). Since along the warranted growth path DY/Y 5 DY*/Y* it follows that, using equation A1.14, the warranted growth rate is given by: Gw 5

s* 2 Ib /Y m1k1v

(A1.15a)

where Gw 5 DY/Y 5 DY*/ Y* and s* is given by equation A1.12. Let C 5 (m 1 k 1 v) 5 total capital–output ratio. Then: Gw 5

s* 2 Ib /Y C

(A1.15b)

Equation A1.15b has the following properties: ∂Gw/∂tp 5 [ad(1 − tc)(1 − ti)(1 − sc)]/C . 0

(A1.16)

∂Gw/∂tc 5 [d(1 − tp) (1 − sc)(1 − ti)a 1 (1 − sc)(1 − ti)x]/C . 0 (A1.17)

Warranted growth and the role of the State

129

∂Gw/∂tw 5 (1 − sw)(1 − ti)g/C . 0

(A1.18)

∂Gw/∂ti 5 [(1 − sc)(1 − tc)d(1 − tp)a 1 (1 − sc)(1 − tc)x 1 (1 − sw)(1 − tw) g]/C . 0

(A1.19)

∂Gw/∂g 5 −1/C , 0

(A1.20)

∂Gw/∂r 5 a(1 − tp)(1 − tc)(1 − ti)(1 − sc)/C . 0

(A1.21)

∂Gw/∂d 5 −a(1 − tp)(1 − tc)(1 − ti)(1 − sc)/C , 0

(A1.22)

Thus increases in all the tax rates and the retained earnings rate raise Gw, while an increase in the government spending share lowers Gw.

130

Strategic competition, dynamics, and the role of the State

APPENDIX 2 In this appendix the role of public investment in State-owned enterprises is introduced, which parallels Sardoni and Palazzi’s (2000) treatment of this issue in the context of Domar’s growth model. Beginning with the savings–investment equality social savings s* equal to all investment in the public and private sectors: [(Icnfp 1 Icnfg) 1 (Ifnfp 1 Ifnfg) 1 (Ivnfp 1 Ivnfg) 1 Ib 1 Igaut] 5 S*

(A2.1)

where Icnfp and Icnfg 5 circulating investment among private (subscript ‘p’) and government (subscript ‘g’) non-financial firms respectively, and Ifnfp and Ifnfg 5 fixed investment in the two sectors respectively, Ivnfp and Ivnfg 5 finished goods inventories in the two sectors respectively, Ib 5 banking sector’s purchases of equipment and structures from non-financial firms, and Igaut 5 government purchases of equipment and structures from nonfinancial firms. Paralleling equation A1.9 the social savings S* is: S* 5 r(1 − tp)P 1 sc(1 − tc)d(1 − tp)P 1 sc(1 − tc) INTch 1 sw(1 − tw)Ywh 1 (T − Cg)

(A2.2)

where Cg 5 cgY 5 public consumption. The tax function remains the same: T 5 tpP 1 tcYch 1 twYwh 1 ti[(1 − sc)(1 − tc)Ych 1 (1 − sw)(1 − tw)Ywh] (A2.3) Then the social savings rate is: s* 5 S*/Y 5 [r(1 − tp) 1 sc(1 − tc)d(1 − tp) 1 tp 1 tcd(1 − tp) 1 ti(1 − sc)(1 − tc)d(1 − tp)]a 1 [sc(1 − tc) 1 tc 1 ti(1 − sc)(1 − tc)]x 1 [sw(1 − tw) 1 tw 1 ti(1 − sw)(1 − tw)]g − cg (A2.4) We will define the following new variables: Kcnfp 5 circulating capital stock in private non-financial firms, Kcnfg 5 circulating capital stock in government non-financial firms, Kfnfp 5 fixed capital stock in private non-financial firms, Kfnfg 5 fixed capital stock in government non-financial firms, Kvnfp 5 finished goods inventory stock in private non-financial firms, Kvnfg 5 finished goods inventory stock in government non-financial firms, Ivnfp 5 finished goods inventory investment in private non-financial firms, Ivnfg 5 finished goods inventory

Warranted growth and the role of the State

131

investment in government non-financial firms, Yp 5 output produced by the private sector, Yg 5 output produced by the government sector, Y*p 5 capacity produced by the private sector, and Y*g 5 capacity produced by the government sector. Then for the aggregate private– public production sector the circulating capital stock to output ratio m 5 (Kcnfp 1 Kcnfg) / (Yp 1 Yg) , the fixed capital stock to capacity ratio k 5 (Kfnfp 1 Kfnfp) / (Y*p 1 Y*g) , and the finished goods inventory stocks to circulating capital stock ratio v 5 (Kvnfp 1 Kvnfg) / (Kcnfp 1 Kcnfg) . If m, k and v are exogenous parameters then m 5 (Icnfp 1 Icnfg) / (DYp 1 DYg) , k 5 (Ifnfp 1 Ifnfg) / (DY*p 1 DY*g) and v 5 (Ivnfp 1 Ivnfg) / (Icnfp 1 Icnfg) (where DKcnfp 5 Icnfp, DKcnfg 5 Icnfg, DKfnfp 5 Ifnfp, DKfnfg 5 Ifnfg, DKvnfp 5 Ivnfp, and DKvnfg 5 Ivnfg). If Y 5 Yp 1 Yg 5 aggregate output from the two sectors and Y* 5 Y*p 1 Y*g 5 aggregate capacity from the two sectors and Y 5 Y* then, after substituting the expressions for m, k and v into equation A2.1, the equation for the warranted growth rate Gw is: s* 2 (Ib/Y) 2 (Iaut g /Y) Gw 5 (A2.5) C In this equation, s* is given by equation A2.4 and C 5 m 1 k 1 v 5 total capital stock–capacity ratio 5 (fixed capital stock 1 circulating capital stock 1 finished goods inventory stocks)/capacity for aggregate private and government firms.

6.

Conclusion: the relevance of microfoundations and politics

This book traces its roots to my dissertation at the New School for Social Research completed just over a decade ago. The world was a very different place in those days. The global economy was booming and free market triumphalism proclaimed an indefinitely rosy future in which the rising global tide would lift all boats. At the time dissenting economists like myself, who were deeply skeptical of free market policies worldwide, were dismissed as nay-sayers. Flash forward a decade and the world has become a very different place, with a large section of humanity confronting the twin specters of mass unemployment and poverty in what is increasingly seen as a crisis that has the potential to parallel the Great Depression of the 1930s. This has been accompanied by growing political and social instability in many developing countries in particular as well the alarming growth of violent social movements.1 While it would be too much of a stretch to reduce all form of political/social instability to economic crises, it would be a serious mistake to ignore the corrosive effects that long-term unemployment have on societies. The optimist in me feels that the current crisis could potentially provide the context in which critical-minded scholars will raise basic questions again about economic theory, the mode of functioning of capitalism itself, and the nature of progressive economic policies in the context of capitalism.2 That is exactly what the Great Depression of the 1930s accomplished. One does not of course have to accept capitalism as inevitable. However, the pragmatist in me asks: given the actual reality of capitalism, what is the scope for alternative economic and social policies within its context to reduce unemployment and poverty? It is the investigation of these basic questions and their larger policy implications that have motivated the writing of this book. When the New Right came triumphantly into power in the US and the UK 25 years ago its battle cry was ‘There is no alternative’ or TINA, a slogan attributed to Margaret Thatcher. Until recently, the neo-liberal tidal wave virtually eclipsed dissenting perspectives. With the fabric of neo-liberalism now in tatters, many economists have become increasingly 132

Conclusion

133

supportive of proactive State policies. This revival of the need for a developmental state has produced a prodigious literature, which has included dissenting neoclassical economists such as Dani Rodrik, Joseph Stiglitz and Paul Krugman, as well as heterodox authors who have been inspired by Steindl and Kalecki. The current book is a contribution to this literature in which the analysis of fiscal policy – one of the core policies of the developmental state – is situated in a context that differs fundamentally from both neoclassical and Post Keynesian growth models. In a fundamental sense, all growth frameworks that emphasize the role of the State in the development process hark back to the classical growth and development authors such as Schumpeter (1934), Rosenstein-Rodan (1943, 1961), Gerschenkron (1962) and Nurkse (1967). As Taylor (2004, pp. 361–2) summarizes: Schumpeter’s views about the disequilibrium nature of the development process are valid, and provide an underlying framework which can be applied widely. One can postulate conditions under which development will be increasingly rapid, capital-intensive, and reliant upon a greater role of the State. In nineteenth-century Europe, for example, greater relative ‘backwardness’ called forth more dramatic transitions (Gerschenkron, 1962). Economies of scale are important. As already noted, coordinated investment across many sectors in a Big Push may be required to give balanced output and demand expansion to take advantage of decreasing average costs economywide. Building upon Schumpeter’s metaphor, the idea was to get the economy from a ‘vicious’ to a ‘virtuous’ circle for growth. The investment must be planned, since pervasive market failures such as decreasing costs and imperfect tradeability mean that price-driven, decentralized investment decisions will not be socially optimal.

Implicitly or explicitly Schumpeter’s contributions constitute the underpinnings of both heterodox (see for example Reinert, 1999) as well as NEGT models. The power of Schumpeter’s framework is the role of profit-seeking behavior by firms, a goal that leads to persistent innovations and a competitive struggle between firms that produces a process of ‘creative destruction’ as weaker firms are weeded out by technologically stronger ones. Following Chamberlin (1951), the conventional wisdom appears to be that Schumpeter’s theory can be seen as a dynamized version of the theory of imperfect competition.3 It is this Schumpeterian vision of the firm that underpins calls for proactive State policies both by heterodox economists such as Chang (2004) and Reinert (1999) and by dissident neoclassical ones such Stiglitz, Krugman and Rodrik. Significantly, for the latter group of authors the rejection of neo-liberalism does not imply the rejection of neoclassical theory. Advocating the need for an alternative set of policies, Rodrik

134

Strategic competition, dynamics, and the role of the State

cautions the reader that this does not imply the need to jettison neoclassical theory, which according to him embodies a set of universal principles: ‘Critics of neoliberalism should not oppose mainstream economics – only its misuse. . . . Neo-liberalism is to neo-classical economics as astrology is to astronomy. In both cases, it takes a lot of blind faith to go from one to the other.’ 4 Dissenters like Rodrik have carved out an intellectual space within mainstream economics by rejecting the ‘one-size-fits-all’ policy advocated by proponents of laissez faire. For Rodrik the key to growth and development is an eclectic blend of policies, combining both markets and a proactive State, tailored to specific local conditions. What is quite significant about Rodrik, as with other dissident neoclassical economists, is that he is able to derive many policy recipes from a single theoretical framework, neoclassical economics. The lynchpin of the very sensible industrial policies proposed by Rodrik (2004) is the existence of market failure, in the form of either monopolistic or oligopolistic competition. For Rodrik the existence of pervasive market failures provides the need for some types of ex ante proactive State policies. The two types of market failures he emphasizes are information externalities, which arise from firms’ lack of sufficient knowledge of an economy’s cost structure, and coordination externalities, which arise from economies of scale. Basically, in the latter case individual firms may not spontaneously invest in large-scale projects because they have no guarantee that other firms will buy their outputs or supply their inputs. Given the interdependencies of industrial production the consequence may be no investment at all under laissez faire. There is no question that firms face such uncertainties, but the question is, what exactly does Rodrik mean by uncertainty when he writes about the need for ex ante proactive policies and what are the implications? Following Keynes, Post Keynesians have emphasized that the economic system is fundamentally non-ergodic (Davidson, 1991). For example, technological change is non-ergodic (Rosenberg, 1994). That is, future innovations and competitive strategies cannot be predicted in a probabilistic sense. Such a view of uncertainty, advocated by Keynes, is however inconsistent with neoclassical general equilibrium theory, with which Rodrik aligns himself. As argued in this book, it is also not consistent with imperfect competition. Following the OERG, the central claim made in Chapter 2 is that in a world characterized by this type of uncertainty it would be suicidal for firms to maintain a high-price and high-cost strategy, as that could subject them to attack either by existing firms that are currently less competitive or by potential new entrants. Depending on its depth and duration a cyclical downturn may subject firms with huge sunk costs, which they cannot

Conclusion

135

get rid of quickly, to large losses. These barriers to exit may make them even more vulnerable to new entrants. As discussed in Chapter 2, even capital-intensive industries can be highly porous fortresses, and potential new entrants, which may be much smaller firms, may have access to finance (for example, State subsidies) that enables them to face losses in a new market dominated by powerful firms. However, the availability of financing can allow them the time to establish a beachhead and perhaps eventually cut into the market shares of established firms. This is the point that Bryce and Dyer (2007) make in their Harvard Business Review article in which they discuss the strategies used by new entrants to break into markets dominated by other firms. Now, if large firms are subject to low-price and low-cost competition, possibly from much smaller firms, how can one still use the notion of market power? After all, the essence of the market power argument is that a smaller number of large-sized firms can erect barriers to entry and reward themselves with high prices and profits, subject to their own rivalries. Put simply, how can either monopolistic or oligopolistic competition be relevant to our understanding of real-world competition? The irrelevance of imperfect competition is also obvious given the historical reality that many smaller firms from developing countries have ‘made it into the big league’, usually by successfully applying policies that Rodrik himself proposes. Now, since he is proposing these policies so as to make it possible for small developing-country firms ‘to make it into the big league’ then presumably Rodrik must be implicitly assuming that developed-country firms could see their market shares erode. The question is, isn’t such a possibility ruled out by oligopolistic competition? On the other hand, if oligopolistic sectors could experience a decline in their hegemony because of incursions by new entrants then how can they be assumed to have market power over any longer time period? The point is that there is an internal inconsistency in advocating activist industrial policies and in rationalizing such policies on the supposed existence of oligopolistic competition. After all, the goal of such policies is potentially to make the LDC firms enter the oligopolistic industries. If it is admitted that in fact oligopolistic sectors could be entered by suitably competitive firms then what is the relevance of using this notion of competition? I would argue that this latter, and more realistic, possibility requires the microfoundations of such policies to be the very different theory of competition discussed in Chapter 2. Neoclassical authors such as Rodrik face another problem. As Bhagwati (1998) argues, making a case for an interventionist State because of the existence of market failure is not a strong one, because if the market failure is fixed via ‘suitable’ policies then the case for laissez faire can be

136

Strategic competition, dynamics, and the role of the State

reestablished. Within the logic of neoclassical theory, from which Rodrik does not desire to depart, this is an entirely consistent argument. It may be remembered that neoclassical authors do not deny the economic relevance of perfect competition. As Krugman and Wells (2006, p. 365) point out, ‘much of what we learn from the study of perfectly competitive markets – about costs, entry and exit, and efficiency – remains valid despite the fact that many industries are not perfectly competitive’. In this connection it is pertinent to ask neoclassical authors who analyze public policies in the context of NEGT the following question: if their goal is to raise growth in order to lower unemployment would not a more straightforward policy be to deregulate the labor market? As in Solow (1956), this would ensure that the warranted growth rate converges to the natural growth rate. Of course, heterodox economists would reject such a labor market model, in which case proactive State policies would automatically become the basis of policies to reduce unemployment. On the other hand, my view is that the case for proactive State policies can be made only on the basis of a theory of competition that is neither perfect competition nor imperfect competition but captures the essence of how real-world firms, both big and small, do behave. Sensible economic policies have to go beyond the neoclassical false dichotomy regarding market structure. To the extent the concern is with long-run involuntary unemployment it follows that such policies can be situated only in a non-neoclassical context. This book has made two new contributions. First, it has provided a new interpretation of the microeconomic analyses of the OERG and has suggested that the theory of competition developed in particular by P.W.S. Andrews, Elizabeth Brunner, H.R. Edwards and others has much more in common with classical political economy than with orthodox theories of competition. One core aspect of this new interpretation, which is based on the OERG’s analysis, is that barriers to entry are not sustainable under the onslaught of ongoing competition. Significantly, the strategic competition perspective is also consistent with actual business practice and history.5 Thus the position taken in this book is that the theories of monopolistic and oligopolistic competition are not realistic descriptions of market behavior. It must be emphasized that my rejection of oligopolistic competition follows from the analytical content of the OERG authors rather than their stated conclusions. In this regard the theory of strategic competition differs from the evolutionary model of competition of Nelson and Winter (1982), as the latter does not repudiate oligopolistic competition. Second, the nature of fiscal policy to raise the warranted growth rate has been elaborated. In this regard this book deals with an important question in the growth literature. A key issue confronting all growth frameworks concerns the nature of the forces, if any, which would make the warranted

Conclusion

137

growth rate (Gw) converge to the natural growth (Gn). While Kaldor showed that Gw could be made equal to Gn via changes in the distribution of income, in Solow this equality arises via a flexible capital–output ratio (Sen, 1970). Following Harrod, in the extended Harrodian model the equality of these two growth rates is accomplished via different types of State policies. These policies of raising taxes on wealthy households and/ or on luxury consumption and of increasing public investment in Stateowned enterprises are in stark contrast to neo-liberal policies. In the extended Harrodian framework the pressure of competition forces each firm to its practicable optimal rate of capacity utilization, which could be consistent with considerable degrees of idle capacity and involuntary unemployment even though output would be at the target level of capacity. However, with output approximately equal to capacity, the social savings rate becomes an important, though not sole, regulator of long-run growth. A stock-flow consistent treatment of flows and stocks ensures that the social savings rate is the nexus between income and expenditure flows and their balance sheet stock counterparts. The presence of endogenous bank credit does not rule out the regulating role of the social savings rate. A further controversial claim made in this book is that the positive impact of the social savings rate does not arise because Say’s law is assumed but because excess capacity is eliminated. In Say’s law demand responds to capacity output, whereas in the extended Harrodian view both output (demand) and capacity respond to each other in the establishment of the warranted growth rate. In fact, as Harrod himself showed, endogenous instability arises because of the demand- and capacity-generating impact of investment leading to their mutual interaction. One way to raise the aggregate tax rate would be to increase direct taxes on the incomes of elite households and indirect taxes on their consumption, a policy proposal that harks back to Kaldor (Palma and Marcel, 1989). Such policies could be accompanied by taxation policies to raise the business retained earnings rate, an outcome that would increase the warranted growth rate. These measures could promote a growth-withequity development strategy, which has been a feature of all successful developmental states (Leftwich, 2008). Thus, in this extended Harrodian framework, taxation policy, to fund a social safety net and public investments, becomes a crucial policy lever. Such a tax-financed public investment policy could be the basis of a longterm nation-building development strategy, much as the New Deal’s public investment program accomplished in the United States (Leigninger, 2007). It is hoped that this will be at the core of an alternative post-neoliberal policy program.

138

Strategic competition, dynamics, and the role of the State

In this regard, the book attempts to deal with the central economic and social problems of our times: mass unemployment, poverty, and in the Third World, underdevelopment. There is therefore no assumption that the raising of growth by itself would resolve such problems, a cautionary note which echoes the arguments made by Stiglitz and Sen and by authors involved in the UNDP’s Human Development Report.6 The types of fiscal policies proposed in this book are more complicated than the straightforward policy of raising budget deficits as in the employer-of-last-resort (ELR) proposal made by authors such as Wray (1998) and others. In the latter the State’s expansionary policies essentially entail the printing of money and so government spending is not in any way anchored to its taxation revenues. On the other hand, capital budgeting requires not only the appropriate types of public investments but also the capacity of the State to raise the necessary taxes in order to maintain a current account surplus. The reader is cautioned against a naïve mathematical relationship between economic theory and policy. A mathematical model may reveal, at a fairly abstract level, the persistent underlying economic forces that regulate growth and the impact of public policies. If constructed in a way that typifies rather than idealizes the capitalist economy, thereby highlighting the causal links between variables, it may even do a good job in revealing the economic structure–policy nexus. On the other hand, in the real world variables can often move in unexpected ways relative to one another, thereby counteracting policymakers’ intended goals. To use a metaphor, if a well-formed model is like a good ocean map that helps guide a ship, in the actual operation of the ship the captain would be ill advised to consider it an infallible guide because of unexpected storms or hidden underwater reefs. Therefore, one should avoid adopting the simplistic view that effective public policies can be mechanically applied on the basis of a mathematical model. Exercises on short-run growth carried out by official agencies using Harrod’s model (Easterly, 1999) are thus deeply problematic. Quite aside from the fact that Harrod’s concern was with both short- and long-run growth, the problem with naïve technocratic policies to raise the growth rate is that they may be countered by endogenously generated instability and crises over which the State may have no control. It is the modeling of endogenous growth and instability that makes Harrod’s framework intellectually powerful and relevant and therefore difficult to appreciate from a general equilibrium framework (Besomi, 2001).7 A country’s taxation strategy is often a window into its class structure, highlighting the power that different social classes exercise at the level of the State. While the majority of the citizens in a democratic developmental

Conclusion

139

state may vote in a government whose mandate is to raise taxes on the wealthy minority, it does not follow that the State will be able to do so either easily or at all. The wealthy may oppose such taxes in various ways or slow down investment spending, which would adversely affect growth and employment. One can see here the limitations to the types of policies discussed in Chapter 5. The ruling classes may simply go on an investment strike if faced with higher taxes that lower their living standards and SOEs that could potentially outcompete private firms. On the other hand, the State may be able to implement such policies if the ruling classes see them as being either not inimical to their long-term interests or beneficial to them. For example, business activity clearly benefits from better infrastructure and a healthier and well-educated workforce.8 These issues suggest that economic policymaking cannot be divorced from questions of power and politics. The reason is quite simply that all economic development policies are shaped by political factors in the same way that the latter are affected by the former (Leftwich, 2008). Neoclassical theory conceptualizes the economy as one in which there are utility-maximizing representative agents. The only difference between agents lies in their ownership of different types of endowments – land, labor or capital – although these different types of ownership do not translate into differences in social power. Heterodox economics, rooted broadly in the classical political economy of Smith, Ricardo and Marx or the institutionalism of Veblen, rejects this world view. Once inequalities of power are made central to economic analysis then, as Leftwich (2008, p. 10) puts it: (a) When people change the way they use, produce and distribute resources, they also change their (social and political) relations – relations of power – with each other; and (b) When people change their political and social (power) relations with each other, they usually change the way they use, produce and distribute resources.

The challenge for a new developmental state is to recognize the ‘messiness’ involved in the intermingling of economics and politics. In this regard there has to be an important departure from the traditional technocratic approach to policymaking by both leftist and rightist governments. If for the latter the omnipotent self-regulating market was to be the panacea for a country’s economic problems, for the former it was the equally omnipotent central planner who was to do the trick. In their own ways, both extremes relied on complex mathematical models to justify their policies. Not surprisingly, both failed in many significant ways because of the messy interaction between economics and politics and ways in which economic

140

Strategic competition, dynamics, and the role of the State

and political power satisfy the interests of self-serving elites whatever the latter’s ideological orientation. In sum, the rationale for a developmental state has to be rooted not only in an appropriate model of economic growth but also in one which is based on an understanding of the political economy of the State. There are two major issues here. The first issue is that the State has to have the capacity to effectuate an economic transformation strategy. Proximately, State capacity will be determined by the quality of the civil service and its remuneration. However, the quality of the civil service will in turn be a function of the overall level of social development of the society, including the quality of its healthcare and higher educational systems. Clearly, a well-educated and healthy population is more likely to generate competent State managers than if it is undernourished and ill educated. Furthermore, as Kohli (2004) shows, State formation is itself a function of specific historical processes. In particular, State formation in ex-colonies will be shaped by the nature of the interaction between former colonial powers and their colonies. However, aside from historical factors, access to taxation revenue will have a significant effect on the ability of the State to create an appropriate level of social development. Class conflict becomes important in determining which social classes can be made to pay taxes by the larger society in a democratically based country. Finally, class becomes an important factor in determining whether or not a developmental State is efficient. Chibber (2003) has pointed out that State intervention is strongly shaped by the capitalist class’s preferences. On the basis of a number of country studies, notably India and South Korea, Chibber argues that State managers cannot unilaterally coerce the economically dominant classes to carry out socially useful activities, since it is the profit motive of these classes that is the basis of business investment in a capitalist economy. For example, in the case of South Korea the capitalist class was aware that in order to penetrate foreign markets it needed to be internationally competitive and thus required extensive State support. On the other hand, India traditionally pursued a narrow import substitution industrialization strategy. With a large, well-protected home market Indian firms had little incentive to improve their competitiveness and, as Chibber argues, opposed interventionist policies in various ways. One consequence of these different developmental paths was that in the case of South Korea the State was able to develop efficient planning institutions such as the Economic Planning Board (like the Ministry of Trade and Industry in Japan), whereas corresponding agencies in India were much less efficient. The neo-liberal onslaught against the role of the State in social and industrial development was in the final instance grounded in deeply

Conclusion

141

ideological assumptions about the capitalist economy, in particular the theories of perfect and imperfect competition. While the Post Keynesian tradition has provided a very powerful critique and alternative to the neoclassical tradition, it suffers from some problematic assumptions of its own about business behavior. It is hoped that the alternative framework of analysis developed in this book will provide a fresh point of departure in the investigation of the nature of the capitalist economy and the role of progressive economic policies.

NOTES 1. See ‘Global Recession Making World a More Violent Place’ by Peter Griffiths, June 2, 2009 (http://abcnews.go.com/US/wireStory?id57732013). The analysis was carried out on the basis of what is called the Global Peace Index compiled by The Economist magazine. 2. I use the term ‘progressive economic policies’ to represent a wide range of targeted State policies to reduce long-run unemployment and poverty. As with all heterodox perspectives, my policy framework begins with the rejection of the neoclassical view that laissez faire will resolve these problems. 3. Note that in this book no position will be taken regarding this claim, as it will be investigated in a future paper. 4. http://www.hks.harvard.edu/fs/drodrik/After%20Neoliberalism.pdf 5. See for example Kim and Mauborgne (2005). 6. See Goodman (2009). 7. Harrod was also aware that growing capital-intensity would drive the warranted growth rate farther away from the natural growth rate, as in Marx (Shaikh, 1978). ‘A continuing increase in the required capital output ratio (Cr) is conjugated with a deceleration in the warranted growth rate. Rather a paradox! Capital intensity increasing, and, by consequence, growth rate declining. Yet the basic equation shows that’ (Harrod, 1973, p. 30). The interesting issue, for future research, is the extent to which the State can modulate such a crisis. 8. It is no surprise that naïve finance-gap type models in which inflows of foreign aid or finance are supposed to augment the domestic social savings rate and, by raising the investment rate, boost growth in the Harrod model have yielded poor empirical results (Easterly, 1999). If it were only true that the capitalist class would so passively increase its planned investment in response to policy prodding! What, after all, is to guarantee that foreign financial flows will go into long-term investment projects in countries faced by enormous degrees of uncertainty? What is to prevent the capitalist class from using this additional money to make economically and socially unimportant short-term investments (e.g. luxury car showrooms) and/or increase luxury expenditures?

References Agénor, Pierre-Richard (2004), The Economics of Adjustment and Growth, Cambridge, MA: Harvard University Press. Aghion, Pierre and Peter Howitt (1998), Endogenous Growth Theory, Cambridge, MA: MIT Press. Amsden, Alice (1989), Asia’s Next Giant: South Korea and Late Industrialization, New York: Oxford University Press. Andrews, P.W.S. (1949a), Manufacturing Business, London: Macmillan Press. Andrews, P.W.S. (1949b), ‘A Reconsideration of the Theory of the Individual Business’, Oxford Economic Papers, 1 (1), pp. 54–89. Andrews, P.W.S. (1951), ‘Industrial Analysis in Economics – With Especial Reference to Marshallian Doctrine’, in T. Wilson and P.W.S. Andrews (eds), Oxford Studies in the Price Mechanism, Oxford: Clarendon Press, pp. 139–72. Andrews, P.W.S. (1952), ‘The Netherlands Lectures’, paper presented at the University of Groningen. Reprinted in Fred S. Lee and Peter S. Earl (eds) (1993), The Economics of Competitive Enterprise: Selected Essays of P.W.S. Andrews, Aldershot, UK and Brookfield, VT, USA: Edward Elgar Publishing, pp. 175–232. Andrews, P.W.S. and Elizabeth Brunner (eds) (1975), Studies in Pricing, London: Macmillan. Andrews, P.W.S. (1993), ‘Competition in the Modern Economy’, in Fred S. Lee and Peter S. Earl (eds), The Economics of Competitive Enterprise: Selected Essays of P.W.S. Andrews, Aldershot, UK and Brookfield, VT, USA: Edward Elgar Publishing, pp. 323–62. Arestis, Philip (1988), ‘Post Keynesian Theory of Money, Credit, and Finance’, in P. Arestis (ed.), Post Keynesian Monetary Economics: New Approaches to Financial Modelling, Aldershot, UK and Brookfield, VT, USA: Edward Elgar Publishing, pp. 41–71. Arestis, Philip and Peter Howells (1996), ‘Theoretical Reflections on Endogenous Money: The Problem with “Convenience Lending”’, Journal of Post Keynesian Economics, 20, pp. 539–51. Arrow, K.J. (1962), ‘The Economic Implications of Learning by Doing’, Review of Economic Studies, 29, pp. 155–73. Aumann, R.J. (1987), ‘Game Theory’, in John Eatwell, Murray Milgate 142

References

143

and Peter Newman (eds), The New Palgrave: A Dictionary of Economics, London: Macmillan Press, pp. 460–79. Bain, J.S. (1951), ‘Relation of Profit Rate to Industry Concentration, American Manufacturing 1936–40’, Quarterly Journal of Economics, 65 (3), August, pp. 293–324. Bain, J.S. (1956), Barriers to New Competition, Cambridge, MA: Harvard University Press. Baran, Paul and Paul Sweezy (1968), Monopoly Capitalism, New York: Monthly Review Press. Barro, R.J. (1974), ‘Are Government Bonds Net Wealth?’, Journal of Political Economy, 82, November/December, pp. 1095–1117. Barro, R.J. (1989), ‘The Ricardian Approach to Budget Deficits’, Journal of Economic Perspectives, 3, Spring, pp. 37–54. Barro, R.J. (1990), ‘Government Spending in a Simple Model of Endogenous Growth’, Journal of Political Economy, 98, October, pp. s103–s125. Barro, R.J. and X. Sala-i-Martin (1995), Economic Growth, New York: McGraw-Hill. Baumol, W.J. (1952), ‘The Transactions Demand for Cash: An Inventory Theoretic Approach’, Quarterly Journal of Economics, 66 (4), pp. 545–56. Besomi, Daniele (2001), ‘Harrod’s Dynamics and the Theory of Growth: the Story of a Mistaken Attribution’, Cambridge Journal of Economics, 25, pp. 79–96. Besomi, Daniele (2006), ‘R.F. Harrod and J.V. Robinson: Correspondence on Growth Theory, 1965–1970’, History of Economics Review, 44, Summer, pp. 16–31. Bhagwati, Jagdish (1998), ‘Free Trade: Why the AFL-CIO, the Sierra Club, and Congressman Gephardt Should Like It’, remarks made at a panel discussion on September 18 on the occasion of the award of the Seidman Distinguished Award in Political Economy, Columbia University (http://www.columbia.edu/~jb38/index_papers.html). Bina, Cyrus (1985), The Economics of the Oil Crisis, New York: St Martin’s Press. Bina, Cyrus (1989a), ‘Some Controversies in the Development of Rent Theory: The Nature of Oil Rent’, Capital & Class, 39, Winter, pp. 82–112. Bina, Cyrus (1989b), ‘Competition, Control, and Price Formation in the International Energy Industry’, Energy Economics, 11 (3), pp. 162–8. Bina, Cyrus (2006), ‘The Globalization of Oil: A Prelude to a Critique of Political Economy’, International Journal of Political Economy, 35 (2), pp. 4–34.

144

Strategic competition, dynamics, and the role of the State

Bina, Cyrus and Minh Vo (2007), ‘OPEC in the Epoch of Globalization: An Event Study of Global Oil Prices’, Global Economy Journal, 7 (1), pp. 1–49. Blaug, M. (1978), Economic Theory in Retrospect, 3rd edn, Cambridge: Cambridge University Press. Bleaney, Michael (1976), Underconsumption Theories: A History and Critical Analysis, New York: International Publishers. Blecker, Robert (2002), ‘Distribution, Demand, and Growth in NeoKaleckian Macro-Models’, in Mark Setterfield (ed.), The Economics of Demand-Led Growth, Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing. Botwinick, Howard (1993), Persistent Inequalities, Princeton, NJ: Princeton University Press. Brander, James A. (1986), ‘Rationales for Strategic Trade and Industrial Policy’, in Paul R. Krugman (ed.), Strategic Trade Policy and the New International Economics, Cambridge, MA: MIT Press. Brown-Collier, E. and B.E. Collier (1995), ‘What Keynes Really Said about Deficit Spending’, Journal of Post Keynesian Economics, 17 (3), pp. 341–55. Brozen, Y. (1971a), ‘Bain’s Concentration and Rates of Return Revisited’, Journal of Law and Economics, 14, October, pp. 351–69. Brozen, Y. (1971b), ‘The Persistence of “High Rates of Return” in High Stable Concentration Industries’, Journal of Law and Economics, 14, October, pp. 501–12. Brozen, Y. (1973), ‘Concentration and Profits – Does Concentration Matter?’, in J. Fred Weston and Stanley J. Ornstein (eds), The Impact of Large Firms on the U.S. Economy, Lexington, MA: D.C. Heath and Company. Brunner, Elizabeth (1952a), ‘Competition and the Theory of the Firm, Part I’, Economia Internazionale, 5, pp. 509–22. Brunner, Elizabeth (1952b), ‘Competition and the Theory of the Firm, Part II’, Economia Internazionale, 5, pp. 727–44. Brunner, Elizabeth (1975a), ‘Competitive Prices, Normal Costs and Industrial Stability’, in P.W.S. Andrews and Elizabeth Brunner (eds), Studies in Pricing, London: Macmillan, pp. 18–34. Brunner, Elizabeth (1975b), ‘Industrial Analysis Revisited’, in P.W.S. Andrews and Elizabeth Brunner (eds), Studies in Pricing, London: Macmillan, pp. 35–46. Bryce, David J. and Jeffrey H. Dyer (2007), ‘Strategies to Crack Open Well-Guarded Markets’, Harvard Business Review, 85 (5), pp. 84–92. Cesaratto, Sergio (1999), ‘Savings and Economic Growth in Neoclassical Theory’, Cambridge Journal of Economics, 23 (6), November, pp. 771–93.

References

145

Chamberlin, Edward H. (1951), ‘Monopolistic Competition Revisited’, Economica, 18 (72), pp. 343–62. Chandler, Alfred (1990), Scale and Scope: The Dynamics of Industrial Capitalism, Cambridge, MA: Harvard University Press. Chang, Ha-Joon (2002), Kicking Away the Ladder, London: Anthem Press. Chang, Ha-Joon (2004), Globalisation, Economic Development and the Role of the State, London: Zed Books. Chang, Ha-Joon and Ajit Singh (1993), ‘Public Enterprises in Developing Countries and Economic Efficiency: A Critical Examination of Analytical, Empirical and Policy Issues’, UNCTAD Review, 4, pp. 1–94. Chang, Ha-Joon and Ajit Singh (1997), ‘Can Large Firms Be Run Efficiently without Being Bureaucratic?’, Journal of International Development, 9 (6), pp. 865–75. Chibber, Vivek (2003), Locked in Place: State-Building and Late Industrialization in India, Princeton, NJ: Princeton University Press. Chick, Victoria (1983), Macroeconomics after Keynes: A Reconsideration of The General Theory, Cambridge, MA: MIT Press. Chick, Victoria (1992), ‘The Evolution of the Banking System’, in P. Arestis and S.C. Dow (eds), On Money, Method and Keynes: Selected Essays, London: Macmillan. Ciccone, R. (1986), ‘Accumulation and Capacity Utilization: Some Critical Considerations on Joan Robinson’s Theory of Distribution’, Political Economy, II, pp. 17–36. Clifton, James A. (1977), ‘Competition and the Evolution of the Capitalist Mode of Production’, Cambridge Journal of Economics, 1, pp. 137–51. Clifton, James A. (1983), ‘Administered Prices in the Context of Capitalist Development’, Contributions to Political Economy, 2, pp. 23–38. Clower, R.W. (1965), ‘The Keynesian Counter-Revolution: A Theoretical Appraisal’, in F.H. Hahn and F.P.R. Brechling (eds), The Theory of Interest Rates, London: Macmillan. Coghlan, R. (1981), Money, Credit, and the Economy, London: George Allen and Unwin. Cohen, Avi and G.C. Harcourt (2003), ‘Whatever Happened to the Cambridge Capital Theory Controversies?’, Journal of Economic Perspectives, 17 (1), March, pp. 199–216. Cypher, James M. and James L. Dietz (2004), The Process of Economic Development, London and New York: Routledge. Daito, Eisuke (2000), ‘Automation and the Organization of Production in the Japanese Automobile Industry: Nissan and Toyota in the 1950s’, Enterprise & Society, 1, March, pp. 139–78. Damodaran, Aswath (2003), ‘Dividends and Taxes: An Analysis of the

146

Strategic competition, dynamics, and the role of the State

Bush Dividend Tax Plan’, New York University, Stern School of Business, mimeo, pp. 2–24 (http://pages.stern.nyu.edu/~adamodar/). Davidson, Paul (1982), ‘Rational Expectations: A Fallacious Foundation for Studying Crucial Decision-Making Processes’, Journal of Post Keynesian Economics, 5 (2), pp. 182–98. Davidson, Paul (1991), ‘Is Probability Theory Relevant for Uncertainty? A Post Keynesian Perspective’, Journal of Economic Perspectives, 5 (1), Winter, pp. 129–43. Davidson, Paul and J. Kregel (1994), Employment, Growth and Finance: Economic Reality and Economic Growth, Aldershot, UK and Brookfield, VT, USA: Edward Elgar Publishing. Dickson, Martin (1992), ‘A Little Lesson for Big Steel: Mini-Mills and Imports Have Forced U.S. Companies to Reform’, New York Times, August 7. Dixit, Avinash and Joseph Stiglitz (1977), ‘Monopolistic Competition and Optimum Product Diversity’, American Economic Review, 67 (3), pp. 297–308. Dockner, Engelbert J., Steffen Jorgensen, Ngo Van Long and Gerhard Sorger (2000), Differential Games in Economics and Management Science, Cambridge: Cambridge University Press. Domar, Evsey (1946), ‘Capital Expansion, Rate of Growth, and Employment’, Econometrica, 14, pp. 137–47. Domar, Evsey (1957), Essays in the Theory of Economic Growth, New York: Oxford University Press. Dos Santos, Claudio (2006), ‘Keynesian Theorising during Hard Times: Stock-Flow Consistent Models as an Unexplored “Frontier” of Keynesian Macroeconomics’, Cambridge Journal of Economics, 30, pp. 541–65. Dow, A.C. and S.C. Dow (1989), ‘Endogenous Money Creation and Idle Balances’, in J. Pheby (ed.), New Directions in Post Keynesian Economics, Aldershot, UK and Brookfield, VT, USA: Edward Elgar Publishing. Downward, Paul and Peter Reynolds (1996), ‘Alternative Perspectives on Post-Keynesian Price Theory’, Review of Political Economy, 8 (1), pp. 67–78. Downward, Paul, Marc Lavoie and Peter Reynolds (1996), ‘Realism, Simulations, and Post-Keynesian Pricing Models: A Response to Lee’, Review of Political Economy, 8 (4), pp. 427–32. Duménil, G. and D. Lévy (1999), ‘Being Keynesian in the Short Term and Classical in the Long Term’, Manchester School, 67 (6), pp. 684–716. Dutt, A.K. (1997), ‘Equilibrium, Path Dependence, and Hysteresis in Post-Keynesian Models’, in P. Arestis, G. Palma and M. Sawyer

References

147

(eds), Capital Controversy, Post Keynesian Economics, and the History of Economic Thought: Essays in Honour of Geoff Harcourt, London: Routledge. Dutt, Amitava (2003), ‘New Growth Theory, Effective Demand, and PostKeynesian Dynamics’, in Neri Salvadori (ed.), Old and New Growth Theories, Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing. Easterly, William (1999), ‘The Ghost of Financing Gap’, Journal of Development Economics, 60 (2), pp. 423–38. Edwards, H.R. (1955), ‘Price Formation in Manufacturing Industry and Excess Capacity’, Oxford Economic Papers, 7 (1), pp. 94–118. Edwards, J.R. (1989), ‘Industrial Cost Accounting Developments in Britain to 1830: A Review Article’, Accounting and Business Research, 19, pp. 305–17. Eichner, Alfred (1976), The Megacorp and Oligopoly: Micro Foundations of Macro Dynamics, New York: Cambridge University Press. Eichner, Alfred (1980), ‘A General Model of Investment and Pricing’, in E.J. Nell (ed.), Growth, Profits, and Property: Essays in the Revival of Political Economy, Cambridge: Cambridge University Press, pp. 118–33. Eiteman, Wilford J. (1948), ‘The Least Cost Point, Capacity, and Marginal Analysis: A Rejoinder’, American Economic Review, 38 (5), pp. 899–904. Eiteman, Wilford J. and Glenn E. Guthrie (1952), ‘The Shape of the Average Cost Curve’, American Economic Review, 42 (5), pp. 832–8. Eltis, Walter (1979), ‘How Rapid Public Sector Growth Can Undermine the Growth of the National Product’, in Wilfred Beckerman (ed.), Slow Growth in Britain: Causes and Consequences, Oxford: Oxford University Press, pp. 118–39. Eltis, Walter (1998), ‘The Harrod–Domar Equation from Quesnay to Marx to Harrod and Domar’, in Giorgio Rampa, Luciana Stella and A.P. Thirlwall (eds), Economic Dynamics, Trade and Growth, New York: St Martin’s Press. Feldstein, Martin (1970), ‘Corporate Taxation and Dividend Behavior’, Review of Economic Studies, 37 (1), pp. 57–72. Feldstein, Martin (1974), ‘Social Security, Induced Retirement and Aggregate Capital Accumulation’, Journal of Political Economy, 82 (5), pp. 905–26. Felipe, Jesus and Franklin M. Fisher (2003), ‘Aggregation in Production Functions: What Applied Economists Should Know’, Metroeconomica, 54 (2 & 3), pp. 208–62. Felipe, Jesus and J.S.L. McCombie (2005a), ‘How Sound Are the

148

Strategic competition, dynamics, and the role of the State

Foundations of the Aggregate Production Function?’, Eastern Economic Journal, 31 (3), Summer, pp. 467–88. Felipe, Jesus and J.S.L. McCombie (2005b), ‘Why Are Some Countries Richer than Others? A Skeptical View of the Mankiw-Romer-Weil’s Test of the Neoclassical Growth Model’, Metroeconomica, 56 (3), pp. 360–92. Felipe, Jesus and J.S.L. McCombie (2006), ‘The Tyranny of the Identity: Growth Accounting Revisited’, International Review of Applied Economics, 20 (3), pp. 283–99. Felipe, Jesus and J.S.L. McCombie (2007), ‘Is a Theory of Total Factor Productivity Really Needed?’, Metroeconomica, 58 (1), pp. 195–229. Fine, Ben (2000), ‘Endogenous Growth Theory: A Critical Assessment’, Cambridge Journal of Economics, 24 (2), March, pp. 245–65. Fisher, Franklin (2005), ‘Aggregate Production Functions – A Pervasive, But Unpersuasive, Fairy Tale’, Eastern Economic Journal, 31 (3), pp. 489–91. Flaschel, Peter and Peter Skott (2006), ‘Steindlian Models of Growth and Stagnation’, Metroeconomica, 57 (3), pp. 303–38. Fleischman, K. and D. Parker (1991), ‘British Entrepreneurs and PreIndustrial Revolution Evidence of Cost Management’, Accounting Review, 66, pp. 361–75. Foley, Duncan K. and Thomas R. Michl (1999), Growth and Distribution, Cambridge, MA: Harvard University Press. Foss, Murray F. (1963), ‘The Utilization of Capital Equipment: The Postwar Period Compared with Prewar’, Survey of Current Business, 43, pp. 8–16. Foss, Nicolai (2000), ‘Austrian Economics and Game Theory: A Stocktaking and an Evaluation’, Review of Austrian Economics, 13 (1), pp. 41–58. Garegnani, P. (1976), ‘On a Change in the Notion of Equilibrium in Recent Work on Value and Distribution’, in M. Brown, K. Sato and P. Zarembka (eds), Essays in Modern Capital Theory, Amsterdam: North Holland. Garegnani, P. (1983), ‘Notes on Consumption, Investment, and Effective Demand’, in John Eatwell and Murray Milgate (eds), Keynes’s Economics and the Theory of Value and Distribution, London: Duckworth, pp. 21–69. Garegnani, P. (1992), ‘Some Notes for an Analysis of Accumulation’, in Joseph Halevi, David Laibman and Edward J. Nell (eds), Beyond the Steady State, New York: St. Martin’s Press, pp. 73–95. Gerschenkron, Alexander (1962), Economic Backwardness in Historical Perspective, Cambridge, MA: Harvard University Press.

References

149

Gilibert, Giorgio (1987), ‘Production: Classical Theories’, in John Eatwell, Murray Milgate and Peter Newman (eds), The New Palgrave: A Dictionary of Economics, London: Macmillan Press, pp. 990–92. Godley, Wynne (1999), ‘Money and Credit in a Keynesian Model of Income Determination’, Cambridge Journal of Economics, 23, pp. 393–411. Godley, Wynne and Marc Lavoie (2007), Monetary Economics: An Integrated Approach to Credit, Money, Income, Production and Wealth, Basingstoke: Palgrave Macmillan. Goodhart, C.A.E. (1984), Monetary Theory and Practice: The UK Experience, London: Macmillan. Goodhart, C.A.E. (1989), ‘Has Moore Become Too Horizontal?’, Journal of Post Keynesian Economics, 12 (1), pp. 29–34. Goodman, Peter (2009), ‘Emphasis on Growth Is Called Misguided’, New York Times, September 23. Goodwin, R. (1997), ‘Keynes and Dynamics’, in G.C. Harcourt and P.A. Riach (eds), A ‘Second Edition’ of The General Theory, New York: Routledge Press, pp. 162–3. Gramlich, Edward (1992), ‘Social Needs and Tight Budgets: Why Not Try Economic Efficiency?’, in Joseph Pechman and M.S. McPherson (eds), Fulfilling America’s Promise: Social Policies in the 1990s, Ithaca, NY: Cornell University Press. Graziani, A. (1990), ‘The Theory of the Monetary Circuit’, Économies et Sociétés, 24 (6), June, pp. 7–36. Greiner, A., W. Semmler and G. Gong (2004), The Forces of Economic Growth: A Time Series Perspective, Princeton, NJ: Princeton University Press. Guala, Francesco (2006), ‘Has Game Theory Been Refuted?’, Journal of Philosophy, 103, pp. 239–63. Hahn, Frank (1970), ‘Some Adjustment Problems’, Econometrica, 38 (1), pp. 1–17. Hahn, Frank (1973), On the Notion of Equilibrium in Economics, Cambridge: Cambridge University Press. Halevi, Joseph and Rédouane Taouil (2002), ‘The Exogeneity of Investment: From Systemic Laws of Accumulation and Growth to Effective Demand Conditions’, in Mark Setterfield (ed.), The Economics of Demand-Led Growth, Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing, pp. 41–54. Haliassos, Michael and James Tobin (1990), ‘The Macroeconomics of Government Finance’, in B.J. Friedman and F. Hahn (eds), Handbook of Monetary Economics, vols 1 and 2, Handbooks in Economics, no. 8, Amsterdam, Oxford and Tokyo: North Holland.

150

Strategic competition, dynamics, and the role of the State

Hall, R.L. and C.J. Hitch (1939), ‘Price Theory and Business Behavior’, Oxford Economic Papers, 2, May, pp. 12–45. Hargreaves-Heap, Shaun and Yanis Varoufakis (1995), Game Theory: A Critical Introduction, 1st edn, New York: Routledge. Harrod, Roy F. (1939a), ‘Price and Cost in Entrepreneurs’ Policy’, Oxford Economic Papers, 2, pp. 1–11. Harrod, Roy F. (1939b), ‘An Essay in Dynamic Theory’, Economic Journal, 49, pp. 14–33. Harrod, Roy F. (1952), Economic Essays, New York: Harcourt Brace Jovanovich. Harrod, R.F. (1959), ‘Domar and Dynamic Economics’, Economic Journal, 69 (275), pp. 451–64. Harrod, R.F. (1969), Money, New York, NY: MacMillan St. Martin’s Press. Harrod, R.F. (1970), ‘Dynamic Theory’, reprinted in A. Sen (ed.), Growth Economics, Harmondsworth: Penguin Books, pp. 43–64. Harrod, Roy F. (1973), Economic Dynamics, London: Macmillan. Hicks, John R. (1979), Causality in Economics, New York: Basic Books. Holusha, John (1994), ‘Why American Steel Is Big Again’, New York Times, July 21. Howells, Peter (1995), ‘The Demand for Endogenous Money’, Journal of Post Keynesian Economics, 18 (1), pp. 89–106. Howells, Peter (2001), ‘The Endogeneity of Money’, in Philip Arestis and Malcolm Sawyer (eds), Money, Finance, and Capitalist Development, Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing, pp. 134–78. Itoh, M. and Costas Lapavitsas (1999), Political Economy of Money and Finance, New York: Macmillan. Kahn, R.F. (1970), ‘Analysis of Growth’, reprinted in A. Sen (ed.), Growth Economics, Harmondsworth: Penguin Books, pp. 141–57. Kahn, R. (1989), The Economics of the Short Period, New York: St Martin’s Press. Kaldor, Nicholas (1935), ‘Market Imperfection and Excess Capacity’, in F. Targetti and A.P. Thirlwall (eds), The Essential Kaldor, New York: Holmes & Meier, pp. 65–83. Kaldor, Nicholas (1956), ‘Economic Problems of Chile’, ECLAC mimeo. Also in Kaldor (1964). Kaldor, Nicholas (1964), Essays on Economic Policy, vol. II, London: Duckworth. Kaldor, N. and Trevithick, J.A. (1981), ‘A Keynesian Perspective on Money’, Lloyds Bank Review, 139, pp. 1–19.

References

151

Kalecki, M. (1954), Theory of Economic Dynamics, London: George Allen & Unwin, p.131. Kalecki, M. (1965), Theory of Economic Dynamics, 2nd edn, London: Unwin University Books. Kalecki, M. (1971a), Selected Essays on the Dynamics of the Capitalist Economy, 1933–1970, Cambridge: Cambridge University Press. Kalecki, M. (1971b), ‘The Class Struggle and the Distribution of National Income’, Kyklos, 24 (1), pp. 1–8. Keen, Steve (2001), Debunking Economics: The Naked Emperor of the Social Sciences, London: Zed Books. Keynes, J.M. (1937), ‘The General Theory of Employment’, Quarterly Journal of Economics, 51 (2), pp. 209–23. Keynes, J.M. (1939), ‘Professor Tinbergen’s Method’, Economic Journal, 49 (195), pp. 558–68. Keynes, J.M. (1953), The General Theory of Employment, Interest, and Money, New York: Harcourt Brace Jovanovich. Keynes, J.M. (1982), Collected Writings, vol. XXI, London: Macmillan. Kim, Linsu (1997), Imitation to Innovation, Boston, MA: Harvard Business School Press. Kim, W. Chan and Mauborgne, Renée (2005), Blue Ocean Strategy, Boston, MA: Harvard University Press. Knight, Frank (1946), ‘Immutable Law in Economics: Its Reality and Limitations’, American Economic Review, 36 (2), pp. 93–111. Kohli, Atul (2004), State-Directed Development: Political Power and Industrialization in the Global Periphery, New York: Cambridge University Press. Kregel, J.A. (1980), ‘Economic Dynamics and the Theory of Steady Growth: An Historical Essay on Harrod’s “Knife-Edge”’, History of Political Economy, 12 (1), pp. 97–123. Kregel, J.A. (1993), ‘Keynesian Stabilization Policy and Post War Economic Performance’, in A. Szirmai, B. van Ark and D. Pilat (eds), Explaining Economic Growth: Essays in Honour of Angus Maddison, Amsterdam: North Holland, pp. 429–45. Kriesler, Peter (1988), ‘Kalecki’s Pricing Theory Revisited’, Journal of Post Keynesian Economics, XI (1), pp. 108–30. Krugman, Paul (1987), ‘Is Free Trade Passé?’, Journal of Economic Perspectives, 1 (2), pp. 131–44. Krugman, Paul and Robin Wells (2005), Microeconomics, New York: Worth Publishers. Kurz, H. (1986), ‘Normal Positions and Capital Utilization’, Political Economy: Studies in the Surplus Approach, 2 (1), pp. 37–54. Kurz, Heinz (1992), ‘Accumulation, Effective Demand and Income

152

Strategic competition, dynamics, and the role of the State

Distribution’, in Joseph Halevi, David Laibman and Edward J. Nell (eds), Beyond the Steady State, New York: St Martin’s Press, pp. 73–95. Kurz, Heinz (2008), ‘Innovation and Profits: Schumpeter and the Classical Heritage’, Journal of Economic Behavior and Organization, 67, pp. 263–78. Kurz, Heinz and Neri Salvadori (1995), Theory of Production: A LongPeriod Analysis, Cambridge, Melbourne and New York: Cambridge University Press. Laidler, D. (1990), Taking Money Seriously, Cambridge, MA: MIT Press. Laidler, D. (1993), The Demand for Money, New York: Harper Collins Publishers. Lavoie, Marc (1992), Foundations of Post-Keynesian Economic Analysis, Aldershot, UK and Brookfield, VT, USA: Edward Elgar Publishing. Lavoie, Marc (1995), ‘The Kaleckian Model of Growth and Distribution and Neo-Ricardian and Neo-Marxian Critiques’, Cambridge Journal of Economics, 19 (6), pp. 789–818. Lavoie, Marc (1996a), ‘Traverse, Hysteresis, and Normal Rates of Capacity Utilization in Kaleckian Models of Growth and Distribution’, Review of Radical Political Economics, 28 (4), pp. 113–47. Lavoie, Marc (1996b), ‘Mark-Up Pricing versus Normal Cost Pricing in Post-Keynesian Models’, Review of Political Economy, 8 (1), pp. 57–66. Lavoie, Marc (1999), ‘The Credit-Led Supply of Deposits and the Demand for Money: Kaldor’s Reflux Mechanism as Previously Endorsed by Joan Robinson’, Cambridge Journal of Economics, 23 (1), pp. 103–13. Lavoie, Marc (2002), ‘The Kaleckian Growth Model with Target Return Pricing and Conflict Inflation’, in Mark Setterfield (ed.), The Economics of Demand-Led Growth, Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing, pp. 172–88. Lavoie, Marc (2003), ‘Kaleckian Effective Demand and Sraffian Normal Prices: Towards a Reconciliation’, Review of Political Economy, 15 (1), pp. 53–74. Lavoie, Marc and Wynne Godley (2001–02), ‘Kaleckian Models of Growth in a Coherent Stock-Flow Monetary Framework: A Kaldorian View’, Journal of Post Keynesian Economics, 24 (2), Winter, pp. 277–312. Lazonick, William (1981), ‘Competition, Specialization, and Industrial Decline’, Journal of Economic History, XLI (1), pp. 31–8. Lazonick, William (1990), Competitive Advantage on the Shop Floor, Cambridge, MA: Harvard University Press. Lazonick, William (2004), ‘Indigenous Innovation and Economic Development: Lessons from China’s Leap into the Information Age’, Industry and Innovation, 11 (4), pp. 273–97.

References

153

Lazonick, William (2008), ‘Entrepreneurial Ventures and the Developmental State: Lessons from the Advanced Economies’, World Institute of Development Economics Research, dp2008/01, pp. 1–44 (available at http://www.wider.unu.edu/publications/working-papers/ discussion-papers/2008/en_GB/dp2008-01). Le Bourva, J. (1992), ‘Money Creation and Credit Multipliers’, Review of Political Economy, 4 (4), pp. 447–66. Lee, Frederic S. (1989), Post Keynesian Price Theory, Cambridge: Cambridge University Press. Lee, Fred S. (1994), ‘From Post-Keynesian to Historical Price Theory, Part I: Facts, Theory and Empirically Grounded Pricing Model’, Review of Political Economy, 6 (3), pp. 303–36. Lee, Fred S. (1996), ‘Price, the Pricing Model, and Post-Keynesian Price Theory’, Review of Political Economy, 8 (1), pp. 87–99. Lee, Fred S. and J. Irving-Lessmann (1992), ‘The Fate of an Errant Hypothesis: The Doctrine of Normal-Cost Prices’, History of Political Economy, 24 (2), pp. 273–356. Leftwich, Adrian (2008), ‘Developmental States, Effective States, and Poverty Reduction: The Primacy of Politics’, United Nations Research Institute for Social Development, pp. 1–28 (http://www.unrisd.org/ unrisd/website/projects.nsf/89d2a44e5722c4f480256b560052d8ad/68c4 0b61b9f737f6c125743900508c69/$FILE/Leftwichweb.pdf). Leigninger, Robert D. (2007), Long-Range Public Investment: The Forgotten Legacy of the New Deal, Columbia, S.C.: University of South Carolina Press. Lerner, Abba P. (1934), ‘The Concept of Monopoly and the Measurement of Monopoly Power’, Review of Economic Studies, 1 (3), pp. 157–73. Lim, David (1976), ‘On the Measurement of Capital Utilization in Less Developed Countries’, Oxford Economic Papers, 28 (1), pp. 149–59. Lucas, R.E. (1988), ‘On the Mechanics of Economic Development’, Journal of Monetary Economics, 22, pp. 3–42. McCombie, J.S.L. and A.P. Thirlwall (1994), Economic Growth and the Balance-of-Payments Constraint, Basingstoke: Macmillan. McNulty, Paul (1967), ‘A Note on the History of Perfect Competition’, Journal of Political Economy, 75 (4), August, pp. 395–9. Mann, M. (1966), ‘Seller Concentration, Barriers to Entry and Rates of Return in Thirty Industries, 1950–60’, Review of Economics and Statistics, 48 (2), August, pp. 296–307. Marcuzzo, Maria Christina (1996), ‘Alternative Microeconomic Foundations for Macroeconomics’, Review of Political Economy, 8 (1), pp. 7–22. Marglin, S.A. and Amit Bhaduri (1990), ‘Profit Squeeze and Keynesian

154

Strategic competition, dynamics, and the role of the State

Theory’, in S.A. Marglin and J. Schor (eds), The Golden Age of Capitalism, Oxford: Oxford University Press, pp. 153–86. Marris, R. (1964), The Economics of Capital Utilisation, Cambridge: Cambridge University Press. Marx, Karl (1954), Capital, vol. I, London: Lawrence & Wishart. Marx, Karl (1956), Capital, vol. II, London: Lawrence & Wishart. Marx, Karl ([1894] 1967), Capital, vol. III, New York: International Publishers. Mass, William and Andrew Robertson (1996), ‘From Textiles to Automobiles: Mechanical and Organizational Innovation in Toyoda Enterprises, 1895– 1933’, Business and Economic History, 25 (2), pp. 1–37. Mathews, John A. and Dong-Sung Cho (2000), Tiger Technology: The Creation of a Semiconductor Industry in East Asia, New York: Cambridge University Press. Maynard, M. (2006), ‘World Automakers Gather in Paris, Looking for a Lift’, New York Times, September 27. Michl, Tom (2000), ‘Notes on the New Endogenous Growth Theory’, Metroeconomica, 51 (2), pp. 182–90. Milanovic, Branko (2003), ‘The Two Faces of Globalization: Against Globalization as We Know It’, World Development, 31 (4), pp. 667–83. Miller, Richard A. (2000), ‘Ten Cheaper Spades: Production Theory and Cost Curves in the Short Run’, Journal of Economic Education, 31 (2), pp. 119–30. Mongiovi, Gary (1990), ‘Notes on Say’s Law, Classical Economics, and the Theory of Effective Demand’, Contributions to Political Economy, 9, pp. 69–82. Moore, Basil (1988), Horizontalists versus Verticalists: The Macroeconomics of Credit Money, Cambridge: Cambridge University Press. Moore, Basil (1997), ‘Reconciliation of the Supply and Demand for Endogenous Money’, Journal of Post Keynesian Economics, 19 (3), Spring, pp. 423–8. Mott, Tracy (2002), ‘Longer-Run Aspects of Kaleckian Macroeconomics’, in Mark Setterfield (ed.), The Economics of Demand-Led Growth, Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing, pp. 153–71. Moudud, Jamee (2006), ‘How State Policies Can Raise Economic Growth’, Challenge, March–April, pp. 33–51. Musgrave, R.A. and P.B. Musgrave (1973), Public Finance in Theory and Practice, New York: McGraw-Hill. Nell, E.J. (1985), ‘Jean Baptiste Marglin: A Comment on “Growth, Distribution, and Inflation”’, Cambridge Journal of Economics, 9, pp. 173–8.

References

155

Nell, E.J. (1997), ‘Understanding Effective Demand’, in P. Arestis, G. Palma and M. Sawyer (eds), Markets, Unemployment, and Economic Policy: Essays in Honour of Geoff Harcourt, vol. 2, London and New York: Routledge. Nelson, Richard R. (1997), ‘How New Is New Growth Theory?’, Challenge, 40 (5), pp. 29–58. Nelson, Richard R. and Sidney G. Winter (1982), An Evolutionary Theory of Economic Change, Cambridge, MA: Harvard University Press. Nightingale, John (1978), ‘On the Definition of Industry and Market’, Journal of Industrial Economics, 27 (1), pp. 31–40. Nurkse, R. (1967), Problems of Capital Formation in Underdeveloped Countries, New York: Oxford University Press. Palley, Thomas I. (2002), ‘Keynesian Macroeconomics and the Theory of Economic Growth: Putting Aggregate Demand Back in the Picture’, in Mark Setterfield (ed.), The Economics of Demand-Led Growth, Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing, pp. 19–40. Palma, Gabriel and Mario Marcel (1989), ‘Kaldor on the “Discreet Charm” of the Chilean Bourgeoisie’, Cambridge Journal of Economics, 13 (1), pp. 245–72. Palumbo, Antonella and Attilio Trezzini (2003), ‘Growth without Normal Capacity Utilization’, European Journal of the History of Economic Thought, 10 (1), Spring, pp. 109–35. Pasinetti, Luigi (1989), ‘Ricardian Debt/Taxation Equivalence in the Kaldor Theory of Profits and Income Distribution’, Cambridge Journal of Economics, 13, pp. 25–36. Paul, M.E. (1954), ‘Notes on Excess Capacity’, Oxford Economic Papers, 6 (1), pp. 33–40. Pechman, Joseph (1987), Federal Tax Policy, Washington, DC: Brookings Institution. Pechman, Joseph R. (1989), Tax Reform, the Rich and the Poor, Washington, DC: Brookings Institution. Pechman, Joseph and Michael S. McPherson (eds) (1992), Fulfilling America’s Promise: Social Policies in the 1990s, Ithaca, NY and London: Cornell University Press. Peterson, Wallace C. and Paul S. Estenson (1996), Income, Employment, and Economic Growth, New York: W.W. Norton. Powers, Susan (1988), ‘The Role of Capital Discards in Multifactor Productivity Measurement’, Monthly Labor Review, June, pp. 27–35. Rabin, A.A. (1979), ‘A Note on the Link between Balance of Payments Disequilibrium and the Excess Demand for Money’, Southern Economic Journal, 45, April, pp. 1233–7.

156

Strategic competition, dynamics, and the role of the State

Rabin, A.A. (1993), ‘A Clarification of the Excess Demand for or Excess Supply of Money’, Economic Inquiry, XXXI, July, pp. 448–55. Rebelo, S. (1991), ‘Long-Run Policy Analysis and Long-Run Growth’, Journal of Political Economy, 99, pp. 500–521. Reinert, Erik S. (1999), ‘The Role of the State in Economic Growth’, Journal of Economic Studies, 26 (4/5), pp. 268–326. Roberts, Mark and Mark Setterfield (2007), ‘What Is Endogenous Growth?’, in P. Arestis, M. Baddeley and J.S.L. McCombie (eds), Understanding Economic Growth: New Directions in Theory and Policy, Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing. Robinson, Joan (1956), The Accumulation of Capital, London: Macmillan. Robinson, Joan (1960), ‘Marx, Marshall, and Keynes’, in Collected Economic Papers, vol. 2, Oxford: Basil Blackwell, pp. 1–17. Robinson, Joan (1962), Essays in the Theory of Economic Growth, New York: St Martin’s Press. Rodrik, Dani (2004), ‘Industrial Policy for the Twenty-First Century’, John F. Kennedy School of Government (available at http://www.hks. harvard.edu/fs/drodrik/UNIDOSep.pdf. Romer, D. (1996), Advanced Macroeconomics, New York: McGraw-Hill. Romer, Paul (1986), ‘Increasing Returns and Long Run Growth’, Journal of Political Economy, 94, pp. 1002–37. Romer, Paul (1990), ‘Endogenous Technical Change’, Journal of Political Economy, 98, pp. S71–S102. Romer, Paul (1994), ‘The Origins of Endogenous Growth’, Journal of Economic Perspectives, 8, pp. 3–22. Rosenberg, Nathan (1994), Exploring the Black Box: Technology, Economics, and History, Cambridge: Cambridge University Press. Rosenstein-Rodan, Paul N. (1943), ‘Problems of Industrialization of Eastern and South-Eastern Europe’, Economic Journal, 53, June– September, pp. 202–11. Rosenstein-Rodan, Paul N. (1961), ‘Notes on the Theory of the Big Push’, in H.S. Ellis and H.C. Wallich (eds), Economic Development for Latin America, New York: St Martin’s Press, pp. 57–67. Sardoni, Claudio and Paolo Palazzi (2000), ‘Public Investment and Growth’, in H. Bougrine (ed.), The Economics of Public Spending, Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing. Sawyer, Malcolm (1985), The Economics of Michal Kalecki, Armonk, NY: M.E. Sharpe. Schwartz, Nelson D. (2009), ‘To Shrink a US Car, Chrysler Goes to Poland’, New York Times, July 18.

References

157

Schumpeter, Joseph A. (1928), ‘The Instability of Capitalism’, Economic Journal, 38 (151), pp. 361–86. Schumpeter, Joseph (1934), The Theory of Economic Development, Cambridge, MA: Harvard University Press. Seccareccia, M. (1995), ‘Keynesianism and Public Investment: A LeftKeynesian Perspective on Government Expenditures and Debt’, Studies in Political Economy, 46, Spring, pp. 43–76. Semmler, Willi (1984), Competition, Monopoly, and Differential Profit Rates, New York: Columbia University Press. Sen, A. (ed.) (1970), Growth Economics, Harmondsworth: Penguin Books. Sen, A. (1977), ‘Rational Fools: A Critique of the Behavioral Foundations of Economic Theory’, Philosophy and Public Affairs, 6 (4), pp. 317–44. Serrano, Franklin (1995), ‘Long Period Effective Demand and the Sraffian Supermultiplier’, Contributions to Political Economy, 14, pp. 67–90. Setterfield, Mark (1994), ‘Recent Developments in Growth Theory: A PostKeynesian View’, in Paul Davidson and J.A. Kregel (eds), Employment, Growth, and Finance, Aldershot, UK and Brookfield, VT, USA: Edward Elgar Publishing, pp. 112–27. Setterfield, Mark (1995), ‘Historical Time and Economic Theory’, Review of Political Economy, 7 (1), pp. 1–27. Setterfield, Mark (2002), ‘A Model of Kaldorian Traverse: Cumulative Causation, Structural Change and Evolutionary Hysteresis’, in Mark Setterfield (ed.), The Economics of Demand-Led Growth, Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing, pp. 215–33. Shaikh, A. (1974), ‘Laws of Production and Laws of Algebra: The Humbug Production Function’, Review of Economics and Statistics, LXI (1), February, pp. 115–20. Shaikh, A. (1978), ‘An Introduction to the History of Crisis Theories’, in Crisis Reader Editorial Collective (ed.), US Capitalism in Crisis, New York: URPE (http://homepage.newschool.edu/%7EAShaikh/). Shaikh, A. (1980), ‘Laws of Production and Laws of Algebra: Humbug II’, in E.J. Nell (ed.), Growth, Profits and Property: Essays in the Revival of Political Economy, Cambridge: Cambridge University Press, pp. 80–95. Shaikh, A. (1982), ‘Neo-Ricardian Economics: A Wealth of Algebra, A Poverty of Theory’, Review of Radical Political Economics, 14 (2), pp. 67–83. Shaikh, A. (1989), ‘Accumulation, Finance, and Effective Demand in Marx, Keynes, and Kalecki’, in Willi Semmler (ed.), Financial Dynamics and Business Cycles: New Perspectives, Armonk, NY: M.E. Sharpe, pp. 65–86. Shaikh, A. (1991), ‘Wandering around the Warranted Path: Dynamic

158

Strategic competition, dynamics, and the role of the State

Nonlinear Solutions to the Harrodian Knife-Edge’, in E.J. Nell and W. Semmler (eds), Kaldor and Mainstream Economics: Confrontation or Convergence?, London: Macmillan. Shaikh, Anwar (1992), ‘A Dynamic Approach to the Theory of Effective Demand’, in D. Papadimitriou (ed.), Profits, Deficits, and Instability, London: Macmillan, pp. 271–94. Shaikh, Anwar (1998), ‘The Empirical Strength of the Labor Theory of Value’, in R. Bellofiore (ed.), Conference Proceedings of Marxian Economics: A Centenary Approach, London: Macmillan, pp. 225–51. Shaikh, Anwar (2005), ‘Nonlinear Dynamics and Pseudoproduction Functions’, Eastern Economic Journal , 31 (3), pp. 447–66. Shaikh, Anwar (2008), ‘Competition and Industrial Rates of Return’, in Philip Arestis and John Eatwell (eds), Essays in Honour of Ajit Singh, New York: Palgrave Macmillan, pp. 167–94. Shaikh, Anwar (2009), ‘Classical Accounting and National Income Accounts’, mimeo, New School for Social Research. Shaikh, Anwar and E.A. Tonak (1994), Measuring the Wealth of Nations, Cambridge: Cambridge University Press. Shapiro, M. (1989), ‘Assessing the Federal Reserve’s Measures of Capacity and Utilization’, Brookings Papers on Economic Activity, 1, pp. 181–241. Skott, Peter (1989), Conflict and Effective Demand in Economic Growth, Cambridge: Cambridge University Press. Smith, Adam (1937), The Wealth of Nations, New York: Modern Library. Smithin, J.N. (1989), ‘The Composition of Government Expenditures and the Effectiveness of Fiscal Policy’, in J. Pheby (ed.), New Directions in Post Keynesian Economics, Aldershot, UK and Brookfield, VT, USA: Edward Elgar Publishing, pp. 209–27. Snowdon, Brian and Howard R. Vane (2005), Modern Macroeconomics: Its Origins, Development, and Current State, Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing. Solow, R. (1956), ‘A Contribution to the Theory of Economic Growth’, Quarterly Journal of Economics, 70, pp. 65–94. Solow, R. (1985), ‘Economic History and Economics’, American Economic Review Papers and Proceedings, 75 (2), pp. 328–31. Solow, R. (1987), ‘Second Thoughts on Growth Theory’, in A. Steinherr and D. Weiserbs (eds), Employment and Growth: Issues for the 1980s, Dordrecht: Martinus Nijhoff Publishers. Solow, R. (1994), ‘Perspectives on Growth Theory’, Journal of Economic Perspectives, 8 (1), pp. 45–54. Steindl, Josef (1952), Maturity and Stagnation in American Capitalism, Oxford: Basil Blackwell.

References

159

Stigler, G.J. (1957), ‘Perfect Competition, Historically Contemplated’, Journal of Political Economy, LXV (1), pp. 1–17. Stigler, G.J. (1963), Capital and Rates of Return in Manufacturing Industries, Princeton, NJ: Princeton University Press. Stiglitz, Joseph E. (2002), Globalization and Its Discontents, New York: W.W. Norton. Streeten, Paul (1955), ‘Two Comments on the Articles by Mrs. Paul and Professor Hicks’, Oxford Economic Papers, 7 (3), October, pp. 259–64. Taylor, Lance (1985), ‘A Stagnationist Model of Economic Growth’, Cambridge Journal of Economics, 9, pp. 383–403. Taylor, Lance (2004), Reconstructing Macroeconomics: Structuralist Macroeconomics and Critiques of the Mainstream, Cambridge, MA: Harvard University Press. Tirole, Jean (1988), The Theory of Industrial Organization, New Delhi: Prentice-Hall. Trezzini, Attilio (1995), ‘Capacity Utilisation in the Long Run and the Autonomous Components of Aggregate Demand’, Contributions to Political Economy, 14, pp. 33–66. Tsoulfidis, Lefteris and Persefoni Tsaliki (2005), ‘Marxian Theory of Competition and the Concept of Regulating Capital: Evidence from Greek Manufacturing’, Review of Radical Political Economics, 37 (1), pp. 5–22. Uzawa, H. (1965), ‘Optimum Technical Change in an Aggregative Model of Economic Growth’, International Economic Review, 6, pp. 18–31. Varian, Hal R. (2006), Intermediate Microeconomics: A Modern Approach, New York: W.W. Norton. Wade, Robert (1990), Governing the Market, Princeton, NJ: Princeton University Press. Weintraub, Sidney (1955), ‘Revised Doctrines of Imperfect Competition’, American Economic Review, 45 (2), Papers and Proceedings of the Sixty-Seventh Annual Meeting of the American Economic Association, pp. 463–79. Weiss, L.W. (1963), ‘Average Concentration Ratios and Industrial Performance’, Journal of Industrial Economics, 75, July, pp. 237–54. Winston, Gordon C. (1974), ‘The Theory of Capacity Utilization and Idleness’, Journal of Economic Literature, 12 (4), pp. 1301–20. Wolfson, Martin (1994), Financial Crises: Understanding the Postwar Experience, Armonk, NY: M.E. Sharpe. Wray, L.R. (1990), Money and Credit in Capitalist Economies: The Endogenous Money Approach, Aldershot, UK and Brookfield, VT, USA: Edward Elgar Publishing.

160

Strategic competition, dynamics, and the role of the State

Wray, L.R. (1998), Understanding Modern Money, Cheltenham, UK and Lyme, NH, USA: Edward Elgar Publishing. Yeager, L.B. (1986), ‘The Significance of Monetary Disequilibrium’, Cato Journal, Fall, pp. 369–99. Zamagni, Stefano (1987), Microeconomic Theory, New York: Basil Blackwell.

Index Abramovitz, M. 61 Agénor, Pierre-Richard 54 Aghion, Pierre 56, 98 Amsden, Alice 21, 30 Andrews, P.W.S. (and) 4, 6, 7, 10, 12, 17–18, 19–20, 22, 23, 24–5, 31, 32, 42, 43, 44, 46–7, 62–3, 110, 136 see also Brunner, Elizabeth critique of entry barriers 7 redefinition of an industry 21 theory of competition 17, 20 threat of new entrants 21 Arestis, Philip 89, 90 Aumann, R.J. 28 Austrian school 62 average fixed costs (AFC) curve 12–14, 26 average total cost (ATC) curve 12–14, 19, 39, 44 average variable cost (AVC) curve 12–14, 24, 26, 39 Bain, J.S. 19, 31, 51 and oligopoly theory 42 Baran, Paul 67 Barro, R.J. 58, 122 Baumol, W.J. 80 Bertrand equilibrium 27 Besomi, Daniele 8, 113, 138 Bhaduri, Amit 68–9 Bhagwati, Jagdish 135 Bina, Cyrus 7, 16, 20, 47, 52 and regulating capital 20 Blaug, M. 65, 75 Bleaney, Michael 63 Blecker, Robert 68–9, 75 Botwinick, Howard 7, 16, 18, 19, 26, 38, 47, 51 Brander, James A. 29 Brown-Collier, E. 115

Brunner, Elizabeth (and) 4, 10, 14, 15, 18, 19, 20–21, 24, 32, 37, 41, 42, 44, 47, 110, 136 critique of entry barriers 7 redefinition of an industry 21 Bryce, David J. 31, 135 capability utilization 15–32, 40 see also excess capacity and idle industrial capacity Capital 67 capital equipment/discards 42 Cesaratto, Sergio 55, 58, 59, 74 Chamberlin, Edward H. 56, 133 Chandler, Alfred 62 Chang, Ha-Joon 21, 30, 133 Chibber, Vivek 30, 140 Chick, Victoria 90–91, 114 Cho, Dong-Sung 21, 32 Ciccone, R. 35–7 classical authors 104 classical political economy 43 Clifton, James A. 7, 16, 23 Clower, R.W. 80 Cobb–Douglas production function 54, 56 Coghlan, R. 80 Cohen, Avi 60 Collier, B.E. 115 competition see also competition theories and Bertrand equilibrium 27 and capability utilization 10, 15–32, 137 and competitive mark-ups 7 evolutionary model of 136 and game theory 27–30 imperfect 134 intra-industrial 27–8 low-price/low-cost 135 oligopolistic 16–17, 27–9, 135

161

162

Strategic competition, dynamics, and the role of the State

and payoff matrix 29 perfect/monopolistic 16–17, 25 competition theories (of) Andrews and Brunner 17–20 classical 17 and implications for capability utilization 15–32 Kalecki 17, 25 Marshall 17 Marx 10, 16–18, 38 neoclassical 17, 30 Cypher, James M. 1 Daito, Eisuke 31 Damodaran, Aswath 119 Davidson, Paul 22, 30, 31, 49, 134 definition(s) of income 119 public investment (NIPA) 120 and redefinition of an industry (Brunner) 21 Dietz, James L. 1 disequilibrium dynamics in SFC context (and) 77–98 see also figures; stock-flow consistent (SFC) framework and tables appendix 1 100–5 ABR Version 1 101–3 ABR Version 2 103–5 appendix 2 106–12 fast adjustment process 106–9 slow adjustment process 109–12 core features and assumptions 77–80 in a model of cyclical growth 92–8 social accounting matrix in a closed economy 81–92 Dixit, Avinash 56 Dockner, Engelbert J. 28, 49 Domar, Evsey 33, 58, 65, 78, 80, 93, 124 see also growth models Dos Santos, Claudio 80 Dow, A.C. 80 Dow, S.C. 80 Downward, Paul 44 Duménil, G. 6, 99 Dutt, A.K. 55, 74 Dyer, Jeffrey H. 31, 135

Easterly, William 138, 141 Economic Dynamics 6, 113–14 Edwards, H.R. 19, 24, 47, 136 Effective Demand, Principle of 3 Eichner, Alfred 25–6, 50 Eiteman, Wilford J. 15, 46 elasticity 11 see also Marx, Karl Eltis, Walter 5, 65, 75 employer-of-last-resort (ELR) proposal 138 endogenous growth theory, neoclassical (NEGT) 55–63, 122 entry barriers 27 Andrews/Brunner critique of 7 equilibrium, general 3 Estenson, Paul S. 80 Euler equation 56 ex ante vs ex post idle capacity 10–15 ex ante–ex post distinction 8, 77–8 ex post accounting identities 80 excess capacity 32–42, 119–20 Kaleckian perspectives on 38–42 persistence of 41 Skott model of 41 Sraffian perspectives on 32–8 excess/reserve capacity and investment, distinction between 4–5 Feldstein, Martin 119 Felipe, Jesus 60, 74 figures average variable and total costs with different shift premiums 14 effect of fall in adjusted social savings rate 97 fast adjustment process involving excess demand, business debt and interest rate 95 natural logarithms of short-run trend of output (Y) and realized output (YR) 96 unit cost curves with no shift premiums 13 unit cost curves with shift premiums 14 Fisher, Franklin M. 60, 74 fixed investment, dual nature of 33 Flaschel, Peter 69, 73 Foley, Duncan K. 3, 54–5 foresight, imperfect 34

Index foresight, perfect 34 Foss, Murray F. 11, 12, 27, 49, 70 Friedman, Milton 1 Fukuyama’s end of history thesis 1 game theory 27–30 as assumption of common knowledge 28 assumption about consistently aligned beliefs 28 Garegnani, P. 5, 32, 33, 35, 45 General Theory, The 30, 124 see also Keynes, J.M. Gerschenkron, Alexander 133 Gilibert, Giorgio 67 Globalization and Its Discontents 2 Godley, Wynne 3, 8, 45, 50, 66, 68, 77, 80, 88, 89, 92, 99, 116, 120, 121, 125 Gong, G. 3, 60 Goodhart, C.A.E. 80 Goodwin, Richard 123–4 Gramlich, Edward 119 Graziani, A. 89 Greiner, A. 3, 60 growth frameworks 136–7 growth models 3–4, 71–2 see also Harrod, Sir Roy; literature and models Domar’s 113, 130 extended Harrodian 7–8, 61, 67, 69, 77, 98, 112, 123 Harrod–Domar 65–6 heterodox 63–72 Kaleckian 10 neo-Kaleckian 67–8, 69 Lavoie 40–41 Lavoie–Godley and Taylor 8 Lucas 58–9 neoclassical 53–63, 67 neoclassical endogenous growth theory (NEGT) 55–63 Solow 53–5 neo-Schumpeterian neoclassical endogenous 3, 10 non-Keynesian 33 post-Keynesian 6, 45, 53, 66–7, 70, 73, 98, 133 Romer 59, 62 Uzawa 58

163

growth theory, approaches to 63–5 see also Marx aggregate demand and supply 64 and Say’s law 64 underconsumptionist 63–4 Guala, Francesco 51 Guthrie, Glenn E. 15 Hahn, Frank 16 Halevi, Joseph 67 Haliassos, Michael 61 Hall, R.L. 23, 43, 47 Harcourt, G.C. 60 Hargreaves-Heap, Shaun 51 Harrod, Sir Roy (and) 4–9, 10, 17, 19, 21–5, 28, 32, 33, 34–6, 41, 44, 48, 58, 62, 65–6, 72, 73, 78, 80, 93, 95, 110, 112, 113–16, 122, 124, 137, 141 Economic Dynamics 6, 113–14 extended growth model 7–8, 61, 67, 72, 73, 98, 112, 113, 123, 137, 138 instability problem 5 and investment 95 investment function/view 45 letter to Joan Robinson 113 perspective 75 public investment 115, 119 ‘Supplement on Dynamic Theory’ 112 taxation policies 114–15, 116–19 taxation-cum-public-investment policies 9, 115–16 treatment of investment 112 Weltanschauung 6 Harrod–Domar equation/pseudo-Harrod–Domar equation 58 growth framework, Keynesian features of 113 perspective 73 Harrodian(s) 73 view of firms adjusting fixed investment 110 warranted growth framework 66–7 Harvard Business Review 31, 135 Hayek, Friedrich 1 heterodox growth models 63–72 Hicks, John R. 31 Hitch, C.J. 23, 43, 47

164

Strategic competition, dynamics, and the role of the State

Howells, Peter 89, 91 Howitt, Peter 56, 98 Human Development Report (UNDP) 138 idle industrial capacity 10–15 intended/ex ante 11–12, 15, 36–7 and Keynesian demand-stimulation policies 4 reasons for 11–12 undesired/redundant and planned/ reserve 4 unintended/ex post 11, 15, 36–7 ‘Industrial Analysis Revisited’ 19, 20, 42 International Monetary Fund (IMF) 1 Irving-Lessmann, J. 43, 52 Itoh, M. 4, 64, 75 Kaldor, Nicholas 33, 89, 117, 119, 123, 137 and implausibility of long-run excess capacity 21 Kaldor–Trevithick reflux mechanism 89–91 Arestis–Howell mechanism as variant of 90, 91 Kalecki, M. 3, 4, 11, 24–5, 37, 38, 48, 65, 66, 67, 70, 72, 73, 133 authors in tradition of 44 and theory of competition 17 Kalecki–Steindl assumption 75 Kaleckian(s) 72, 73 argument 125 framework and persistent entry barriers 45 growth models 10 investment theory 5 literature 5–6 mark-up model 25 Keen, Steve 60 Keynes, J.M. 3, 4, 6, 7, 9, 30–31, 48, 61, 65, 110, 112, 113, 115, 121, 122, 124, 134 and Principle of Effective Demand 33, 67 and role of uncertainty 22, 30–31, 134 Keynesian(s) authors 104, 115, 119, 122 dynamic effect 97

macroeconomics 71 neoclassical synthesis 80 New 2 scenario 11 of shortfall of demand 15 theory 91 tradition 6, 74, 119 uncertainty 4, 40, 78, 84, 91 Keynesian-type and State-led development policies 1 Kim, Linsu 21, 32 Knight, Frank 7, 16 Kohli, Atul 140 Kregel, J.A. 110, 115 Krugman, Paul 29, 133, 136 Kurz, Heinz 11, 12, 37–8, 41, 46, 70, 73, 74 Laidler, D. 80 Lapavitsas, Costas 4, 64, 75 Lavoie, Marc 3, 5, 9, 38–40, 44–5, 50, 66, 68, 69, 72, 80, 89–91, 92, 116, 120, 121, 125 model 8, 40–41 Lazonick, William 30, 32, 48, 62 Le Bourva, J. 90 Lee, Frederic S. 42, 43, 44, 52 Leftwich, Adrian 137, 139 Leigninger, Robert D. 137 Leontief, W.W. 66, 79, 93 and classical tradition 93 Lerner, Abba P. 24 Lévy, D. 6, 99 Lim, David 11 literature classical/Marxist 19 on fiscal policy–growth nexus 121–2 on growth 53–76, 136 heterodox growth models 63–72 neoclassical growth models 53–63 late industrialization 21 Post Keynesian 68, 70 long-run growth see microfoundations of long-run growth lower-cost firms, threat of 44–5 Lucas, R.E. 53, 58–9 Luxemburg, R. 64 McCombie, J.S.L. 42, 60, 74 McNulty, Paul 7, 16, 26–7

Index McPherson, Michael S. 119 Malthus, Thomas 64 Marcel, Mario 123, 137 marginal cost (MC) curve 22, 24, 26 marginal revenue (MR) curve 22 Marglin, S.A. 68–9 mark-up and full-cost pricing 39 Marris, R. 11, 37 Marx, Karl (and) 9, 15, 47, 49, 64–5, 66, 67, 73, 74, 75, 79, 112, 139 analysis of collusion 26 analysis of competition 24 competition theory 10, 16–18, 38 elasticity 11 Marx’s framework 64–5 Marxian literature 16, 19 Marxists 39 Mass, William 31, 47 Mathews, John A. 21, 32 Michl, Thomas R. 3, 54–5, 74 microfoundations of long-run growth (and) 10–52 appendix 51–2 competition theories: implications for capacity utilization 15–32 see also main entry elasticity (Marx) 11 ex ante vs ex post idle capacity 10–15 persistence of excess capacity, Sraffian and Kaleckian approaches to 32–42 reasons for idle industrial capacity 11–12 microfoundations and politics, relevance of 132–41 Milanovic, Branko 1 Mill, John Stuart 16 Minsky, H.P. 3 and Minsky-type debt dynamics 73 models see also growth models AK 59 of cyclical growth see disequilibrium dynamics in SFC context disequilibrium dynamics 77–112 endogenous growth (Rebelo) 122 evolutionary model of competition 136 excess capacity (Skott) 41 investment-constrained 3 IS-LM 4

165

Kaleckian with excess capacity 39 Kaleckian mark-up 25 of monopolistic competition (Chamberlin–Robinson) 25, 28, 32 of oligopolistic competition 25–6 post-Keynesian 98, 120 pricing 24 Romer 62 Solow 122 monetary circuit theory 89–90 Mongiovi, Gary 64, 75 Moore, Basil 89, 90 Mott, Tracy 74, 75 Musgrave, P.B. 115, 121, 125 Musgrave, R.A. 115, 121, 125 Nash equilibrium 29, 48 Nell, E.J. 36, 42, 72 Nelson, Richard R. 30, 49, 61–2, 136 neoclassical authors 135–6 endogenous growth theory (NEGT) 2–3, 55–63, 73, 136 general equilibrium theory 134 theory 22, 30, 136 neo-liberalism 132, 133–4 and IMF, critique of 1–2 neo-Schumpeterian framework 56 models 62 new classical economics 80 New Deal: public investment program 137 new entrants, threat of 21, 44–5, 135 New Right, the 1, 132–3 New School for Social Research 132 Nightingale, John 20 normal-cost pricing, principle of 43 normal-cost pricing and mark-up pricing, relationship between 44 Nurkse, R. 133 Oxford Economists’ Research Group (OERG) 4, 6, 7, 22–3, 41, 43, 44, 62–3, 72, 134, 136 analyses of competition/firm behavior 7–8, 15–17 microeconomic analyses of the 136 survey of entrepreneurs 22

166

Strategic competition, dynamics, and the role of the State

Palazzi, Paolo 130 Palley, Thomas I. 73 Palma, Gabriel 123, 137 Palumbo, Antonella 32, 33–6, 72, 73 Pasinetti, Luigi 117, 124 Paul, M.E. 23, 24 payoff definition of 48–9 matrix 29 Pechman, Joseph R. 119, 125 Peterson, Wallace C. 80 post-Keynesian(s) 69, 134 authors/literature 43, 44, 70, 77, 79, 80, 90, 91, 123 framework 123 growth models 6, 45, 53, 66–7, 70, 73, 98, 133 investment functions 70–71, 110 models 98, 120 perspective 73 position 121–2 and SFC framework 77 theory 24 tradition 119, 141 Powers, Susan 42 price- and cost-cutting 23 price-setting strategies 52 pricing models 24 pricing/pricing theory full or normal cost 23 marginalist 23 normal-cost 4 production functions 3 profit maximization (marginal revenue MR = marginal costs MC) 22–3, 24 public investment 138 in State-owned enterprises 130–31 Rabin, A.A. 80 rational expectations 3 Reagan, Ronald 1 Rebelo, S. 58 and AK model of endogenous growth 122 regulating capital 20 Reinert, Erik S. 133 reserve capacity 4 Reynolds, Peter 44

Ricardian Equivalence view 122 Ricardo, D. 64, 66, 79, 139 Roberts, Mark 98 Robertson, Andrew 31, 47 Robinson, Joan 33, 65, 113, 124–5 and growth model 124 Rodrik, Dani 133–6 Romer, P. 10, 53, 56, 58, 59, 74 Romer’s model 62 Rosenberg, Nathan 30, 134 Rosenstein-Rodan, Paul N. 133 Sala-i-Martin, X. 58 Salvadori, Neri 38, 46 Sardoni, Claudio 130 Sawyer, Malcolm 11, 24, 25 Say, J.-B. 64 Say’s law 3, 6, 9, 58, 63, 64, 65–6, 73, 98, 114, 137 Schumpeter, J. 133 Schumpeterian aspect of Brunner/Andrews framework 20–21 framework 133 microfoundations 62 NEGT model 57 vision of the firm 133 Schwartz, Nelson D. 48 Seccareccia, M. 125 Semmler, Willi 3, 7, 16, 18, 19, 38, 47, 51, 60 Sen, A. 60–61, 65, 138 Serrano, Franklin 71–2 Setterfield, Mark 40, 62, 70, 74, 98 Shaikh, Anwar 7, 8, 16, 18, 19, 20, 38, 47, 48, 51–2, 60, 63, 64, 65, 66, 67, 72, 74, 75, 78, 81, 88, 99, 106–7, 112 and extended Harrodian cyclical growth model 7 model 8, 9 Shapiro, M. 12 Skott, Peter 69, 73 excess capacity model 41 Smith, Adam 16, 24, 26–7, 52, 139 Smithin, J.N. 115 social accounting matrix (SAM) 100–103, 117 in a closed economy 81–92 ex ante 8–9, 77–98, 105

Index ex ante–ex post 78 ex post 77, 80, 88 social and industrial development, role of State in 140 social savings rate 126–9, 137 Solow, R. 3, 49, 59, 60, 73, 136 Solow model 122 Sraffa, Piero 71, 72 Sraffian(s) 39, 71–2 framework 33 multiplier 71 rejection of normal capacity utilization 35 state-owned enterprises (SOEs) 120–23, 139 Steindl, Josef 4, 38, 67, 73, 75, 133 Stigler, G.J. 16 Stiglitz, Joseph E. 2, 56, 133, 138 and critique of neo-liberalism and IMF 1–2 stock-flow consistent (SFC) framework 3–4, 98 ex ante–ex post 77–8 and extended Harrodian model 8, 77–8 Stone, Richard 3, 77 Streeten, Paul 23 Studies in Pricing 19, 20, 42 and ‘Industrial Analysis Revisited’ 20 supply and demand, bi-directional interactions between 114 survey (of/on) British firms during inter-war period (OERG) 12 entrepreneurs and short-run marginal revenue curve 22 Sweezy, Paul 67 table ex ante social accounting matrix 85–7 Taouil, Rédouane 67 taxation policies 114–15, 116–19, 137 taxation strategy 138 Taylor, Lance 3, 4, 8, 29, 67, 88, 133 technological change as non-ergodic 134

167

Thatcher, Margaret 1 and TINA 132 Thirlwall, A.P. 42 Tirole, Jean 42 Tobin, James 61 Trevithick, J.A. 89 Trezzini, Attilio 32, 33–6, 72, 73 Tsaliki, Persefoni 20 Tsoulfidis, Lefteris 20 uncertainty 30–31, 134 underconsumption/ underconsumptionists 63–4 undersaving countries 114 UNESCO 67 US Social Security Act (1937) 119 Uzawa, H. 58 Varian, Hal R. 27 Varoufakis, Yanis 51 Veblen, T. 139 Wade, Robert 125 warranted growth and role of the State 113–31 see also Harrod, Sir Roy appendix 1: social savings rate 126–9 appendix 2: public investment in State-owned enterprises 130–31 Harrod’s policy insights: solutions to ambiguities and contradictions 116–22 and taxation policies 114–15 and undersaving countries 114 warranted growth rate/natural growth rate 136–7 Weintraub, Sidney 23 Wells, Robin 29, 136 Weltanschauung 6, 66, 67, 69 Winston, Gordon C. 4, 5, 11, 12, 13, 14, 15, 36, 37 Winter, Sidney G. 136 Wolfson, Martin 73 World Bank 1–2 Wray, L.R. 4, 75, 138 Yeager, L.B. 80 Zamagni, Stefano 31