ARENA the Contribution of Joseph Schumpeer to Economics - Economic Development and Institutional...

This collection constitutes an examination of Schumpeter’s legacy that is widerthan most of those attempted up to now. As one of the key economists of thetwentieth century, Schumpeter’s theory is viewed in relation to his importantcontributions to areas as diverse as the history of economic analysis, economicmethodology and economic sociology, as well as to the theories of entrepreneur-ship, competition, innovation, business cycles, money, banking and finance.These wide-ranging contributions reveal Schumpeter’s adherence to a unifiedanalytical and methodological approach. Rather than evolutionary, thisapproach clearly forms part of the theoretical tradition in economics for whichinstitutions and institutional change are key aspects.

This timely book is an authoritative and original study into theSchumpeterian legacy and will be welcomed by historians of economic thought.It will be essential reading for economists interested in institutionalist, evolu-tionary and Austrian economics.

Richard Arena and Cécile Dangel-Hagnauer are both at the University ofNice - Sophia Antipolis, France.

The Contribution of Joseph Schumpeterto Economics

1 Economics as LiteratureWillie Henderson

2 Socialism and Marginalism inEconomics 1870–1930Edited by Ian Steedman

3 Hayek’s Political EconomyThe socio-economics of orderSteve Fleetwood

4 On the Origins of ClassicalEconomicsDistribution and value from WilliamPetty to Adam SmithTony Aspromourgos

5 The Economics of Joan RobinsonEdited by Maria Cristina Marcuzzo, LuigiPasinetti and Alesandro Roncaglia

6 The Evolutionist Economics ofLéon WalrasAlbert Jolink

7 Keynes and the ‘Classics’A study in language, epistemology andmistaken identitiesMichel Verdon

8 The History of Game Theory, Vol. 1From the beginnings to 1945Robert W. Dimand and Mary AnnDimand

9 The Economics of W.S. JevonsSandra Peart

10 Gandhi’s Economic ThoughtAjit K. Dasgupta

11 Equilibrium and EconomicTheoryEdited by Giovanni Caravale

12 Austrian Economics in DebateEdited by Willem Keizer, Bert Tieben andRudy van Zijp

13 Ancient Economic ThoughtEdited by B.B. Price

14 The Political Economy of SocialCredit and Guild SocialismFrances Hutchinson and Brian Burkitt

15 Economic CareersEconomics and economists in Britain1930–1970Keith Tribe

16 Understanding ‘Classical’EconomicsStudies in the long-period theoryHeinz Kurz and Neri Salvadori

17 History of EnvironmentalEconomic ThoughtE. Kula

18 Economic Thought in Communistand Post-Communist EuropeEdited by Hans-Jürgen Wagener

19 Studies in the History of FrenchPolitical EconomyFrom Bodin to WalrasEdited by Gilbert Faccarello

20 The Economics of John RaeEdited by O.F. Hamouda, C. Lee andD. Mair

Routledge Studies in the History of Economics

21 Keynes and the NeoclassicalSynthesisEinsteinian versus NewtonianmacroeconomicsTeodoro Dario Togati

22 Historical Perspectives onMacroeconomicsSixty years after the ‘General Theory’Edited by Philippe Fontaine and AlbertJolink

23 The Founding of InstitutionalEconomicsThe leisure class and sovereigntyEdited by Warren J. Samuels

24 Evolution of Austrian EconomicsFrom Menger to LachmannSandye Gloria

25 Marx’s Concept of Money: theGod of CommoditiesAnitra Nelson

26 The Economics of James SteuartEdited by Ramón Tortajada

27 The Development of Economics inEurope since 1945Edited by A.W. Bob Coats

28 The Canon in the History ofEconomicsCritical essaysEdited by Michalis Psalidopoulos

29 Money and GrowthSelected papers of Allyn Abbott YoungEdited by Perry G. Mehrling and Roger J.Sandilands

30 The Social Economics of Jean-Baptiste SayMarkets and virtueEvelyn L. Forget

31 The Foundations of Laissez-FaireThe economics of Pierre de BoisguilbertGilbert Faccarello

32 John Ruskin’s Political EconomyWillie Henderson

33 Contributions to the History ofEconomic ThoughtEssays in honour of R.D.C. BlackEdited by Antoin E. Murphy and RenéePrendergast

34 Towards an Unknown MarxCommentary on the manuscripts of1861–63Enrique Dussel

35 Economics and InterdisciplinaryExchangeEdited by Guido Erreygers

36 Economics as the Art of ThoughtEssays in memory of G.L.S. ShackleEdited by Stephen F. Frowen and PeterEarl

37 The Decline of RicardianEconomicsPolitics and economics in post-Ricardian theorySusan Pashkoff

38 Piero SraffaHis life, thought and cultural heritageAlessandro Roncaglia

39 Equilibrium and Disequilibrium inEconomic TheoryThe Marshall–Walras divideEdited by Michel de Vroey

40 The German Historical SchoolThe historical and ethical approach toeconomicsEdited by Yuichi Shionoya

41 Reflections on the ClassicalCanon in EconomicsEssays in honor of Samuel HollanderEdited by Sandra Peart and Evelyn Forget

42 Piero Sraffa’s Political EconomyA centenary estimateEdited by Terenzio Cozzi and RobertoMarchionatti

43 The Contribution of JosephSchumpeter to EconomicsEconomic development and institutional changeRichard Arena and Cécile Dangel-Hagnauer

44 On the Development of Long-runNeo-Classical TheoryTom Kompas

45 F.A. Hayek as a PoliticalEconomistEconomic analysis and valuesEdited by Jack Birner, Pierre Garrousteand Thierry Aimar

46 Pareto, Economics and SocietyThe mechanical analogyMichael McLure

47 Strategies and Programmes inCapital TheoryJack Birner

48 Economics Broadly ConsideredEssays in honor of Warren J. SamuelsEdited by Steven G. Medema, Jeff Biddleand John B. Davis

49 Physicians and Political EconomySix studies of the work of doctor-EconomistsEdited by Peter Groenewegen

50 The Spread of Political Economyand the Professionalisation ofEconomistsEconomic societies in Europe, Americaand Japan in the nineteenth centuryMassimo Augello and Marco Guidi

51 Historians of Economics &Economic ThoughtThe construction of disciplinarymemorySteven G. Medema and Warren J.Samuels

52 Competing Economic TheoriesEssays in memory of Giovanni CaravaleSergio Nisticò and Domenico Tosato

53 Economic Thought and Policy inLess Developed EuropeThe 19th centuryEdited by Michalis Psalidopoulos andMaria-Eugenia Almedia Mata

54 Family Fictions and Family FactsHarriet Martineau, Adolphe Queteletand the population question in england1798–1859Brian Cooper

Edited by Richard Arena andCécile Dangel-Hagnauer

London and New York

The Contribution of Joseph Schumpeter to EconomicsEconomic development andinstitutional change

First published 2002 by Routledge11 New Fetter Lane, London EC4P 4EE

Simultaneously published in the USA and Canadaby Routledge29 West 35th Street, New York, NY 10001

Routledge is an imprint of the Taylor & Francis Group

© 2002 Editorial material and selection, Richard Arena and CécileDangel-Hagnauer; the individual chapters, the contributors

All rights reserved. No part of this book may be reprinted or reproducedor utilised in any form or by any electronic, mechanical, or other means,now known or hereafter invented, including photocopying and recording,or in any information storage or retrieval system, without permission inwriting from the publishers.

British Library Cataloguing in Publication DataA catalogue record for this book is available from the British Library

Library of Congress Cataloging in Publication DataA catalog record for this book has been requested

ISBN 0–415–22824–7

This edition published in the Taylor and Francis e-Library, 2005.

“To purchase your own copy of this or any of Taylor & Francis or Routledge’scollection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.”

(Print Edition)

ISBN 0-203-99595-3 Master e-book ISBN

List of contributors ixPreface xiList of abbreviations xix

Introduction 1R I C H A R D A R E N A A N D C É C I L E D A N G E L - H A G N AU E R

PART IHistory of economic analysis 19

1 Schumpeter and the old Austrian school: interpretations andinfluences 21S A N D Y E G L O R I A - PA L E R M O

2 Schumpeter on Walras 40R I C H A R D A R E N A

3 Schumpeter on Marshall 66N AT H A L I E D U VA L

PART IIMethodology 87

4 On the boundaries between economic analysis and economicsociology 89A N D R É L E G R I S

5 Time and rationality in Schumpeter’s construct 106A N N E C H Â T E AU N E U F - M A L C L È S

Contents

PART IIIEconomic development and social change 125

6 Innovation and business cycles 127A G N È S F E S T R É

7 The long-term perspective: Schumpeter’s prediction of the end of capitalism 146O D I L E L A K O M S K I

PART IVEntrepreneurship and competition 165

8 Schumpeter on entrepreneurship 167R I C H A R D A R E N A A N D PAU L - M A R I E R O M A N I

9 Schumpeter on competition 184A L A I N R AY B AU T A N D F R A N C K S O S T H É

PART VMoney, banking and finance 201

10 Schumpeter on the institution of money 203C É C I L E D A N G E L - H A G N AU E R

11 Money, banking and dynamics: Schumpeter vs Hayek 221A G N È S F E S T R É

12 Financing economic activity: Schumpeter vs Keynes 241E R I C N A S I C A

Index 256

viii Contents

Richard Arena, LATAPSES-IODE (CNRS and University of Nice – SophiaAntipolis)

Anne Châteauneuf-Malclès, CREUSET, University of St-Etienne

Cécile Dangel-Hagnauer, LATAPSES-IODE (CNRS and University of Nice –Sophia Antipolis)

Nathalie Duval, French Ministry of the Economy and Public Finance

Agnès Festré, LATAPSES-IODE (CNRS and University of Nice – SophiaAntipolis)

Sandye Gloria-Palermo, CREUSET and University of Saint-Etienne

Odile Lakomski, CRIISEA, University of Picardi–Jules Verne, Amiens

André Legris, LATAPSES-IODE (CNRS and University of Nice – SophiaAntipolis)

Eric Nasica, LATAPSES-IODE (CNRS and University of Nice – SophiaAntipolis)

Alain Raybaut, LATAPSES-IODE (CNRS and University of Nice – SophiaAntipolis)

Paul-Marie Romani, LATAPSES-IODE (CNRS and University of Nice –Sophia Antipolis)

Franck Sosthé, LATAPSES-IODE (CNRS and University of Nice – SophiaAntipolis)

Contributors

More often than not, volumes such as this serve the purpose of trying to save aparticular author or debate from oblivion. This is most definitely not thepurpose of this book for the simple reason that Joseph A. Schumpeter has neverfallen into oblivion. Since his death in 1950, his work has never ceased toattract the attention of economists, even when focused on very different aspectsof his thought. The timeless nature and unremitting relevance of Schumpeter’sideas and analysis is directly related to their diversity encompassing, as theydo, the theory of competition, of entrepreneurship, of development and ofinnovation, monetary theory, the history of economic thought, economicmethodology and economic sociology – and even this list is by no meansexhaustive!

Such diversity has also encouraged varied and, at times, contradictory inter-pretations of Schumpeter’s work (Arena 1992). In the last two decades,however, this situation has undergone a gradual but pronounced change. Eventhough disagreements continue, there now exists a predominant, if not whollyundisputed, interpretation of Schumpeter’s contribution to economics. Theemergence of this interpretation is closely related to the rise and developmentof modern evolutionary economics. Originally perhaps most closely associatedwith the seminal contribution by Nelson and Winter (1982), evolutionaryeconomics is, in fact, frequently labelled ‘neo-Schumpeterian’, and its advocateswelcome Schumpeter’s legacy as the cornerstone of an alternative moderneconomic theory.

Put briefly, under this perspective Schumpeter is credited with havingconstructed an entrepreneurial theory of competition that, or so goes the argu-ment, has two distinctly evolutionary features. First, it is based on the idea thata process of natural selection among firms is a central feature of the marketmechanism. Second, at the level of the economic system as a whole, it impliesthe operation, with the passage of historical time, of a process of ‘creativedestruction’. This process is seen to ensure the elimination of some firmsthrough competition, as well as the creation of new ones managed byentrepreneurs who introduce new products, techniques, markets or forms oforganisation to the economy. ‘Creative destruction’ is then equated or

Preface

associated with biological evolution, implying an organic transformation of theeconomic system.

This interpretation is not wholly misleading. However, it is certainly reduc-tionist in that it ultimately restricts Schumpeter’s contribution to the analysis ofthe relation between competition and innovation in a market economy. Thisview suffers from two important shortcomings.

On the one hand, it underestimates Schumpeter’s contribution toeconomics. One main purpose of this book is to show that, while Schumpeter’sanalysis of competition and innovation certainly deserves a central place inmodern theory, his contributions to other subject areas are not only extensivebut also no less important. In order to argue this point, it is of course necessaryto take account of all his writings, including those less familiar to the majorityof economists, not least because, in some cases, English translations are still notavailable. What emerges from such an encompassing reading of his work is thatSchumpeter not only dedicated considerable time and effort to a wide range ofsubjects, producing a number of important and detailed contributions, oftencontaining original insights and results, but it is, moreover, difficult to maintainthat the theory of competition and innovation can be seen to be the solecentral and unifying theme of his work.

On the other hand, these wide-ranging contributions clearly revealSchumpeter’s adherence to a unified analytical and methodological approach.Far from being evolutionary, this approach clearly forms part of the institution-alist approach to economic theory. This is the second main insight emphasisedin the book and discussed in more detail in our introductory chapter.

For the moment, it suffices to clarify briefly what we mean by ‘institution-alist’. Essentially, our point is not that Schumpeter should be viewed as aninheritor of the German Historical School or as having worked in the traditionof American institutionalists. Rather, our meaning is a broader one: in our view,Schumpeter attempted to build a general theory of the relations between insti-tutional change and economic development. Such a theory required attentionto history, and Schumpeter complied with this requirement by developing aspecific and original approach that attributes to economic sociology a role ofintermediary between history and economic analysis. This theory also assignedan important role to the institutions of money and credit within a broaderframework of the analysis of the workings of a capitalist economy. It thereforerejected the device of an abstract barter economy in which money does notmatter. This is in direct contrast to a large part of modern evolutionaryeconomics that tends to underestimate the importance of money and credit infavour of a strong (or over-) emphasis on the role of technology. Furthermore,in the analysis of development, Schumpeter also gave pride of place to theconcepts of self-organisation and social leadership, thus outlining a mode ofrelations between men and society that has little in common with standardinterpretations of methodological individualism. Finally, Schumpeter’s workhighlighted the need for a theory of economic dynamics that differs substan-tially from the pure economic theory of exchange, focusing instead on the

xii Preface

actual impact of institutions on the workings of the economic system, includingsocial norms and behavioural rules.

This alternative institutionalist interpretation of Schumpeter’s contributionto economics runs through this book, albeit to varying degrees, naturallydepending on the specific focus of the different chapters. We begin with anintroduction that provides a more detailed discussion of the overarching themeof the book as it has just been outlined. The bulk of the book is organised infive parts, each taking up a specific topic in the following order: the history ofeconomic analysis; methodology; economic development and social change;entrepreneurship and competition and money, banking and finance.

Part I is devoted to Schumpeter as an historian of economic analysis, usingthe marginal revolution as an illustration. The three chapters that make up thispart examine how Schumpeter perceived and interpreted this revolution anddescribe the ways in which he drew inspiration for his own theoretical schemefrom early marginalist contributions. The first chapter is concerned withSchumpeter’s relationship with the ‘old’ Austrian school. As Sandye Gloria-Palermo reminds us, Schumpeter is the initiator of the historiographicapproach, predominant to this day, that considers Menger, Jevons and Walras asthe three protagonists of the so-called marginal revolution. This suggests thatSchumpeter appears to have underestimated the originality of the Austriantradition. Gloria-Palermo analyses how Schumpeter, the historian of economicanalysis, interpreted the contribution of each of the three founders of theAustrian tradition. She also identifies aspects of this tradition that Schumpeterintegrated into his own analytical work – namely, Menger’s analysis of the roleof institutions in the process of economic evolution, Böhm-Bawerk’s concept ofproduction as a time-consuming process, and Wieser’s analysis of the role playedby leaders in the process of economic change. More specifically, she attempts toexplain why Schumpeter tends to overlook these ingredients when he accountsfor the Austrian authors’ contribution, despite the use he makes of them inconstructing his own theory of economic development.

The second chapter deals with Schumpeter’s contribution to Walrasianhistoriography. Richard Arena shows that one of Schumpeter’s importantachievements was to rehabilitate Walras’s contribution to economic theory,especially in the English-speaking world. However, his interpretation paved theway for the standard interpretation of the Walrasian system endorsed by moderngeneral equilibrium theorists, whereas Arena provides evidence that this inter-pretation hardly qualifies as a faithful representation of Walras’s intellectualproject. He also shows that, notwithstanding the constant tribute he paid toWalras, Schumpeter also pointed to important limitations and shortcomings ofthe latter’s approach. This led him to consider the Walrasian system as aconstruct of limited validity and to try to develop a more general conception ofthe dynamics of capitalist market economies.

In the third chapter, Nathalie Duval examines how Schumpeter reconstructsMarshall’s contribution to what he considers ‘progress’ in economic thought, i.e.the building of static equilibrium theory. This led him to purge Marshall’s

Preface xiii

thought of what he considers as ‘non-scientific’, illustrative, empirical andhistorical digressions. This might seem paradoxical given that Schumpeterhimself paid considerable attention to history. In fact, he is especially critical ofwhat he considers a recurrent methodological bias in Marshall’s work, i.e. atendency to treat what he himself regards as the qualitatively differentphenomenon of economic development as a mere extension of static analysis byintroducing historical time into the framework of static analysis. Even so,Schumpeter was, of course, well aware of Marshall’s ‘evolution-mindedness’, butconsidered that Marshall’s understanding of dynamic analysis was flawed in thathe failed to understand that in order to deal with the dynamic aspects ofeconomic development a totally different theory is required.

Part II of the volume focuses on predominantly methodological issues. AndréLegris explores the boundaries between ‘pure’ economic theory and economicsociology. In Schumpeter’s framework, the former provides a self-containedtheoretical framework for the analysis of the mechanism leading to equilibrium– that is, of the economy of the circular flow. However, when Schumpeter turnshis attention to the process of development, he broadens his area of investiga-tion to what he calls economic sociology – that is, the analysis of institutions.Accordingly, he explores the economic sociology of innovations and examinesthe institutional environment that encourages the emergence of a specificactor-type, the entrepreneur. He emphasises, in particular, that credit and theinstitution of banking play a key role in allowing entrepreneurial aspirations tocome to fruition.

In Chapter 5, Anne Châteauneuf-Malclès deals with the question of timeand rationality in Schumpeter’s writings. She begins by reminding us that,although he never addressed the question of time directly, it was of pivotalimportance to Schumpeter’s work, economic change having been his mainconcern. Two models of time underlie his work. Retrospective time is character-istic of the circular flow, meaning that it is past-oriented. By emphasising thatthe passage of time creates permanent conditions which are favourable to theemergence of routines, the author shows how Schumpeter transforms the time-less Walrasian model into the model of the circular flow. Prospective time is thedistinctive feature of development where future-oriented entrepreneurs havethe skills needed to confront the discontinuities and uncertainty generated byinnovative activity. Interestingly, the two forms of temporality generatedifferent forms of institutions. Retrospective time leads to the emergence ofwell-functioning networks that encourage coordination and reduce uncertainty,while the appearance of credit institutions is directly related to prospectivetime.

Part III of the book is concerned with Schumpeter’s important contributionsto the theory of economic development and social change. Agnès Festré high-lights the originality of Schumpeter’s conception of the business cycle, tracing itback to his unconventional conception of dynamics. First, dynamics is opposedto statics as described by the ‘pure’ circular flow model. Second, dynamicsexcludes growth factors such as saving or increases in population, which

xiv Preface

Schumpeter also accounts for within the framework of the circular flow. Rather,for Schumpeter, dynamics is synonymous with development, of which innova-tion is the only cause. Having reminded us of these basic distinctions, Festréanalyses the mechanisms and phases of the business cycle. She examines therole of credit, the dissemination of innovations, the behaviour of prices, theimportance of sectoral as opposed to aggregate features of the dynamic process.Festré also points to the weaknesses of Schumpeter’s theory of business cycles asperceived by a number of economists. Whether or not well-founded, these criti-cisms have stimulated modern research on industrial innovation and R&D,thereby complying with Schumpeter’s desire to see his ideas scrutinised in thelight of further evidence.

The chapter by Odile Lakomski discusses Schumpeter’s view of the long-term perspectives of capitalism. Like Marx, Schumpeter was convinced of theinevitability of an eventual ‘march into socialism’ However, Schumpeter didnot believe that the threat to the capitalist system was to be found in economicinstability and the development of crises. Quite the opposite, he consideredthat capitalism becomes ever more stable, from an economic point of view, as aresult of both increasing industrial concentration and the more bureaucraticorganisation of innovation-related activities. Ironically, it is precisely economicsuccess that leads to socialism in that this success produces a gradual transfor-mation of cultural and political values, of the role of the bourgeoisie within thepolitical system and, more generally, of the institutions characterising capi-talism. Thus, Schumpeter provides an original view of what may be in store forcapitalism by combining the analysis of institutional change with the theory ofeconomic development.

Entrepreneurship and competition are the focus of attention in Part IV of the volume. It begins with a chapter on Schumpeter’s theory of entrepreneurship, in which Richard Arena and Paul-Marie Romani showhow Schumpeter applies his methodological approach, consisting in com-bining economic theory with history and economic sociology, toentrepreneurship. For Schumpeter, entrepreneurship is a specific form ofsocial leadership. Entrepreneurs are social leaders whose function is to carryout innovations. They are thus the instigators of economic change, whichexplains why their function is only transitory – a claim Schumpeter substanti-ates by making use of both historical analysis and economic theory.Schumpeter’s historical analysis of entrepreneurship puts specific emphasis oninstitutional and social change, that is, on the rise and decline of institutions,sets of values and forms of organisation. Arena and Romani also discuss theeconomic aspects of entrepreneurial rationality, thereby providing an explana-tion of Schumpeter’s scepticism as regards the Walrasian conception of theentrepreneur. The stance the authors take in this chapter contrasts with themore conventional interpretation, especially within evolutionary economics,according to which Schumpeter stresses entrepreneurial competition as aprocess of natural selection. Instead, the authors suggest that the methodologyunderlying Schumpeter’s analysis of economic dynamics, by relating economic

Preface xv

development with institutional change, enables him to account for change asan endogenous process.

Schumpeter believed in the dynamic efficiency of monopolistic structures, asAlain Raybaut and Franck Sosthé remind us in their chapter on Schumpeterand competition. However, the authors emphasise that Schumpeter’s theory ofcompetition is also concerned with innovation, prices and the formation ofmarket equilibria. First recalling Schumpeter’s assessment of how competition istreated in the history of economic analysis before 1914, Raybaut and Sosthéthen proceed to study Schumpeter’s conception of competition both within thecircular flow and in the context of development. When analysing the circularflow, Schumpeter highlights the role played by routines and adaptive behaviourin order to explain how competition leads to equilibrium. Such routines aredisturbed when economic development enters the picture. As new combina-tions emerge alongside older ones as a result of innovation, prices for newproducts are determined according to the principle of monopoly pricing and theuneven movement of the economic system makes calculations based on experi-ence impossible. Schumpeter, however, believed that the economy willeventually settle in the neighbourhood of a new competitive equilibrium. Whatthen prevails is some kind of hybrid competition. On these grounds,Schumpeter views pure competition and pure monopoly as limiting cases,although his appraisal of the emerging theory of imperfect competition is ratherreserved.

The fifth and final part of the book is devoted to Schumpeter’s theory ofmoney and credit. The first chapter deals with Schumpeter’s conception ofmoney as an institution that is explained in his treatise on money and banking,first published posthumously in German in 1970. In this chapter, CécileDangel-Hagnauer suggests that the theme Schumpeter develops in this treatiseis one that is located at the frontier of economic theory, economic sociologyand history. Thus, because Schumpeter considers money as an institution, hebegins by constructing a stylised institutional framework within which heembeds the concept of money. Schumpeter’s point of departure is, therefore,economic sociology. But money is also the focus of economic theory. The bridgeSchumpeter builds between economic theory and economic sociology is whatallows him to explain the ‘essence’ of money. This he accomplishes quitestraightforwardly when he considers the working of the institution of money inthe economy of the circular flow. Difficulties crop up, however, when he turnsto the analysis of money within the context of economic development. Theproblems that arise are dealt with by analysing the actual working of and thehistorical forms taken by the institution of money when the economy is subjectto change. It is also at this point that credit enters the picture, providing themonetary system with some degree of flexibility.

In the following chapter, Agnès Festré highlights the common Wicksellianorigin of Schumpeter’s and Hayek’s conceptions of money and banking. Thiscontrasts with the differences in these authors’ approaches to the mechanisms

xvi Preface

underlying economic dynamics. In Hayek’s analysis, the inability of the bankingsector to ensure that the natural and real rates of interest coincide leads to aredistribution of purchasing power from consumers to producers and, thus, toforced saving. The resulting excessive investment expenditures during theupswing are detrimental to the economy in that they prompt it to deviate fromits natural tendency towards equilibrium, resulting in economic crisis and theeventual return to equilibrium. In Schumpeter’s approach, the increased activityin the investment goods sector during the upswing is triggered by a spurt ofinnovations financed through new bank loans. The ensuing redistribution ofpurchasing power from traditional producers to innovators is another instanceof forced saving, albeit of a different nature. Banking institutions here are essen-tial players in the development process. The instability to which the actions ofbanking institutions contribute through the business cycle is thus, forSchumpeter, the very vehicle of progress.

The chapter by Eric Nasica contains a comparative study of Schumpeter’sand Keynes’s views on the financing of economic activity. Nasica points to thesimilarities in the two authors’ approach to money and finance. First, in bothKeynes’s ‘co-operative’ economy and in Schumpeter’s circular flow – that is, intheir respective static models – money is essentially a technical device. Second,both Keynes and Schumpeter introduce a more complex concept of money: theformer when he deals with the ‘entrepreneurial’ economy in which instabilityand fluctuations tend to develop; the latter when he introduces innovation anddynamics in the form of economic development. Third, their analyses leadthem to reject the quantity theory of money. However, their approaches alsoexhibit important differences. Whereas Keynes places strong emphasis on therole played by financial markets and the long-term interest rate, Schumpeterinsists on the role of credit and banking and considers the short-term interestrate to be the more relevant variable for the analysis of the working of bankingand financial markets. While preceding Keynes’s contribution by more thantwenty years, Schumpeter’s theory also contains a more in-depth analysis of theworking of banking institutions. For Schumpeter, financial institutions are typi-cally entrepreneurial organisations insofar as they are constantly striving toinnovate and to increase profits.

This preface would not be complete without our thanks to MichelRainelli, Françoise De Bandt, Sylvain André, Muriel Destailleur, ElisabethDuruisseau, Pierre Goursaud, Muriel Mathéry, Martine Naulet and KatiaRolland for their invaluable help in the making of this book. We would alsolike to thank Stephanie Blankenburg for the work she put into the initialversions of the articles that make up this volume. Her interaction with theauthors contributed not only to an improvement in the style of the articles,but also to the clarification of many passages. Her help was particularly crucialin the translation of quotes from Schumpeter’s 1908 and 1970 contributionsfor which no English translation from the original German exists as yet.

Preface xvii

Bibliography

Arena, R. (1992) ‘Schumpeter after Walras: “economie pure” or “stylised facts”?’, in T.Lowry (ed.) Perspectives on the History of Economic Thought, Vol. VIII, Aldershot:Edward Elgar.

Nelson, R. and Winter, S., (1982) An Evolutionary Theory of Economic Change,Cambridge: Cambridge University Press.

xviii Preface

BC Business CyclesCSD Capitalism, Socialism and DemocracyDW Das Wesen und der Hauptinhalt der theoretischen

NationalökonomieHEA History of Economic AnalysisTED The Theory of Economic Development

List of abbreviations

Today, most scholars share the view that J.A. Schumpeter was not only a greateconomist but also one of the main founders of modern economic evolutionarytheory.1 In this view, Schumpeter’s approach to economic change, as it is gener-ally understood, led him to build a theory in which innovations emerge as theresult of a process of selective competition. This contrasts with the traditionaland static conception whereby competition constitutes an end-state.

In this book we argue that this conventional interpretation of Schumpeter isboth reductionist and misleading. It is certainly reductionist in that it tends tolimit Schumpeter’s contribution to economics to what we will argue is but oneaspect of his much wider analytical construct. It is also misleading since, farfrom looking favourably on evolutionary approaches, Schumpeter alwaysregarded the possibility of borrowing from biology with scepticism. It is ourbelief that an evolutionary interpretation of Schumpeter’s contribution not onlyminimises its profound originality, but also misrepresents it. To prove our point,we will start from the fundamental distinction that Schumpeter makes betweenthe various ‘Techniques of economic analysis’ in his History of Economic Analysis(HEA).

The essential importance of economic sociology amongstSchumpeter’s techniques of economic analysis

It is worthwhile to recall Schumpeter’s characterisation of the role of economicsociology as a complementary technique, alongside the three techniques ofeconomic analysis he lists at the beginning of the HEA, namely, history, statis-tics and ‘theory’:2

The schemata of economic theory derive the institutional frameworkswithin which they are supposed to function from economic history, whichalone can tell us what sort of society it was, or is, to which the theoreticalschemata are to apply. Yet, it is not only economic history that renders thisservice to economic theory. It is easy to see that when we introduce theinstitution of private property or of free contracting or else a greater orsmaller amount of government regulation, we are introducing social facts

Introduction

Richard Arena and Cécile Dangel-Hagnauer

that are not simply economic history but are a sort of generalized or typifiedor stylized economic history. And this applies still more to the generalforms of human behaviour which we assume either in general or for certainsocial situations but not for others … To use a felicitous phrase: economicanalysis deals with the question how people behave at any time and whatthe economic effects are they produce by so behaving; economic sociologydeals with the questions how they came to behave as they do. If we definehuman behaviour widely enough so that it includes not only actions andmotives and propensities but also the social institutions that are relevant toeconomic behaviour such as government, property inheritance, contract,and so on, that phrase really tells us what we need.

(Schumpeter 1954: 20–1, emphasis added)

In this passage, Schumpeter explains the relationship between economic anal-ysis and economic sociology. To get the full picture, however, it is necessary tocomplement this statement with Schumpeter’s remarks on this question in DasWesen und der Hauptinhalt der theoretischen Nationalökonomie (DW) as well as inhis sociological writings.3 These texts do, in effect, add considerable substanceto Schumpeter’s statement in the above passage. Careful reading reveals thatSchumpeter regarded the ‘science of organisation’ as part of economic sociology(Schumpeter 1908: 133). Thus economic sociology includes:

the science of state forms but also the science of the forms of law and of theremaining social relations and structures and, thirdly, the science ofeconomic organisation as such: on the one hand, the division of labour andon the other hand, the formation of cartels, of labour associations, etc.

(ibid.)

What Schumpeter asserts in the HEA is that, for the economist, historyprovides the raw material that consists of empirical sets of diverse institutionsand forms of organisation. However, this raw material requires further work inorder to produce the assumptions that are made when the economist sets out tobuild an economic theory. First, economic sociology must ‘generalise’, ‘typify’and ‘stylise’ the empirical forms of institutions and organisations so as to trans-form the historical set from which they are drawn into a more abstract set ofideal types on which the economist can then build his analytical assumptions.4

Second, referring to the example of fiscal sociology, Schumpeter (1918 [1991]:177 fn. 18) emphasises that the historical order according to which institutionsand organisational forms emerge, develop and decline must not be confused withthe analytical process by which sociology provides a logical explanation of thesechanges. Seen thus, historical chronology is partially arbitrary whereaseconomic sociology must respect the necessary requirements of consistent anal-ysis. Third, economic sociology must extract from history what is strictlyeconomic, and this obviously presupposes a relative autonomy or ‘self-contain-ment’ of the economic sphere (Schumpeter 1908: 135). Schumpeter provides

2 Introduction

an illuminating example of this when he argues that the legal aspects of theinstitution of credit (namely, the strict requirement for the borrower to repaythe lender) must be distinguished from – but also determine – its economicaspects (namely, the implications of this requirement for the expectations andeconomic behaviour of agents) (Schumpeter 1917/18 [1956]: 155–7). Finally, ifwe consider economic activity from the point of view of dynamics, it can alsocreate feedback effects on institutions, organisation or law. A case in point isSchumpeter’s analysis of taxation and, more specifically, his emphasis on thelimitations of fiscal impositions arising from the need not to squeeze profitsbeyond a certain point (Schumpeter 1918 [1991]: 112).

Economic sociology, according to Schumpeter, can thus be defined as thescience of the emergence, maintenance and decline of societal institutions andforms of organisation that influence economic behaviour.

Social motives and social classes

For Schumpeter, human motives are never strictly individual. Rather, they arealways embedded in a social context and related to the historical circumstancesunder which they have emerged. From this point of view, two main concepts areessential.

On the one hand, following Wieser’s conception of economic sociology,Schumpeter argues that, whatever the social environment, men are alwaysdivided into two categories: leaders and followers.5 It should, however, be notedthat Schumpeter does not regard leaders as superior or ‘great men’ (Schumpeter1927 [1951]: 216). They are not in possession of special intellectual qualitiesthat would lead them to play a pre-eminent social role. Nor do they have aconscious concept of social optimality that they would strive to put into practise(ibid.). Rather, ‘[w]e are content to say that social leadership means to decide, tocommand, to prevail, to advance. As such it is a special function, always clearlydiscernible in the actions of the individual and within the social whole’ (ibid.:217).

Leaders’ motives are related to their ‘instinctive urge to domination’(Schumpeter 1919 [1951]: 15), an ‘excess of energy’ (ibid.: 34) or ‘activity urgesspringing from capacities and inclinations that had once been crucial tosurvival, though they had now outlived their usefulness’ (ibid.: 44). These ‘urges’(or this Trieb, ibid.: 83) are defined by Schumpeter as human inclinations thathave more to do with ‘instinct’ than with reason (ibid.: 83–4). They involvecreativity and entail permanent changes to the sphere in which they appear (bethis the arts, science, economic activity, etc.). Shionoya neatly summarises thiswhen he writes that

In the first place, creative activity cannot be predicted by applying theordinary rules of inference from pre-existing facts. It is so unique that themechanism of the modus operandi must be examined on a case-by-casebasis. Second, creative activity shapes the whole course of subsequent

Introduction 3

events and their long-term outcome, and causes discontinuity frompreceding situations. Third, creative activity is an enigma of human beingsand has something to do with the distribution of talent and therefore withthe phenomenon of leadership.

(Shionoya 1997: 175)

Like Wieser, Schumpeter regards followers as playing a more passive role in thatthey are the mere recipients of leaders’ decisions, acting to diffuse them. Theycan reinforce these decisions and contribute to their social generalisationthrough the adoption of imitative behaviour or the manifestation of trust. Butthey can also resist them, slowing down the process of diffusion or sometimeseven preventing the mechanisms of social diffusion from working.

On the other hand, however, leadership is not independent from the socialcontext in which it appears. Schumpeter strongly stresses this aspect of socialbehaviour. First, the Trieb or ‘urge’ provides only part of the social explanationof leaders’ motives. Referring to warlike societies, Schumpeter argues that

[t]he explanation lies, instead, in the vital needs of situations that moldedpeoples and classes into warriors – if they wanted to avoid extinction – andin the fact that psychological dispositions and social structures acquired inthe dim past in such situations, once firmly established, tend to maintainthemselves and to continue in effect long after they have lost theirmeaning and their life-preserving function.

(Schumpeter 1919 [1951]: 83–4)

Second, the social scientist must also take account of what the ‘subsidiaryfactors that facilitate the survival of such dispositions and structures’ (ibid.)could be. That is, s/he must pay attention to the interests of social classes and ofthose individuals whose interests are being served by maintaining a state of war.In other words, the second concept that needs to be introduced at this stage ofour discussion is the concept of social class. For Schumpeter, a social class isdefined as a set of individuals who, in a specific social context, are able toperform a given and specific social function:

The ultimate foundation on which the class phenomenon rests consists ofindividual differences in aptitude. What is meant is not differences in anabsolute sense, but differences in aptitude with respect to those functionswhich the environment makes ‘socially necessary’ – in our sense – at anygiven time; and with respect to leadership, along lines that are in keepingwith those functions.

(Schumpeter 1927 [1951]: 210)

Schumpeter, therefore, does not seem to think that it is possible to define socialclasses from either a purely individualistic or holistic methodological point ofview:

4 Introduction

We cannot help those who are unable to see that the individual is a socialfact, the psychological an objective fact, who cannot give up toying with theempty contrasts of the individual vs. the social, the subjective vs. theobjective.

(ibid.: 211, emphasis in the original)

This characterisation of social classes explains why class interest exists as suchand why the fact that an individual belongs to a given social class influencesthis individual in a way that does not solely depend on his or her own free willbut also on what Wieser called social ‘compulsory forces’. As Shionoya haspointed out:

At the outset of his 1927 article on social classes, Schumpeter noted thatthe concept of classes he was going to use related to historical and socialentities, not to a conceptual artifact like landowners and workersconstructed in economic theory. A class is more than a mere aggregate ofits members and has its own peculiar life and characteristic spirit. Thisspecification clearly means that a theory of social classes is a sociologydealing with institutional and environmental conditions that circumscribethe behavior and thought of individuals and that methodological individu-alism does not hold in this discipline.

(Shionoya 1997: 226)

What then is the relationship between leadership and social classes? It is clearthat, for Schumpeter, these two ideal typical concepts must be carefully distin-guished. In a market economy, for instance, leaders – that is, entrepreneurs – donot form a social class (Schumpeter 1912 [1934]: 78 and 1939: 104). Althoughthey exert a strong influence on social order through their innovative role, thuscontributing to the evolution of social structure, this does not imply that ‘theentrepreneurial function will lead to certain class positions for the successfulentrepreneur and his family’ (ibid.: 78). Moreover, the entrepreneurial functioncannot be inherited (ibid.: 79). Finally, leaders use the social structure toachieve their ends. For instance, in ancient Egypt kings used the military aris-tocracy to organise society according to their own objectives (Schumpeter 1919[1951]: 165).

Institutions and forms of organisation

Entrepreneurs are the economic leaders of the market economy. This represents‘a fundamental truth of the sociology of industrial society’ (Schumpeter 1939:96) in that entrepreneurs create the ‘institutional patterns’ of economic devel-opment. The excess energy that characterised the leaders of ancient societiesbased on aristocratic hierarchies and military objectives now turns into whatSchumpeter calls ‘energetic’ – as opposed to ‘hedonistic’ – rationality or egoism

Introduction 5

in DW as well as in the first German edition of The Theory of EconomicDevelopment (TED). 6 In modern societies,

[t]here is much less excess energy to be vented in war and conquest than inany precapitalist society. What excess energy there is flows largely intoindustry itself, accounts for its shining figures – the type of the captain ofindustry … In a purely capitalist world, what was once energy for warbecomes simply energy for labor of every kind.

(Schumpeter 1919 [1951]: 90)

In market economies, excess energy is channelled into the introduction of inno-vations, such as new products or new productive techniques. These innovationsdo not result from exogenous shocks or endogenous mechanisms of technologycreation generated by firm managers or owners. Rather, they are introduced bywhat Schumpeter called ‘New Men’ (Schumpeter 1939: 96). In other words,they presuppose the emergence of leaders who use their excess energy topromote the transition from the circular flow to economic development.Therefore, innovations and economic development appear to be the naturalconsequences of the particular new form of leadership that prevails in a marketeconomy.

However, innovations do not last forever. Gradually, they are diffusedthroughout the economic system and transformed into routines or ‘habitualeconomic methods’ (Schumpeter (1912 [1934]: 8). As they come to prevail,these individual routines and the resulting network of social rules or normseventually produce the ‘institutional patterns’ that pervade the markets andinfluence the internal organisation of the firm.

Money and credit form the second fundamental institution to be found inmarket economies. To carry out new technical combinations firms must invest,and this investment must, in turn, be financed:

Another [problem] exists for us: the problem of detaching productivemeans (already employed somewhere) from the circular flow and allottingthem to new combinations. This is done by credit, by means of which onewho wishes to carry out new combinations outbids the producers in thecircular flow in the market for the required means of production. Andalthough the meaning and object of this process lies in a movement ofgoods from their old towards new employments, it cannot be describedentirely in terms of goods without overlooking something essential, whichhappens in the sphere of money and credit and upon which depends theexplanation of important phenomena in the capitalist form of economicorganisation, in contrast to other types.

(Ibid.: (1912 [1934]: 71)

According to Schumpeter, money is an institutional device and a logical prereq-uisite of the market economy. This is why Schumpeter was so insistent on the

6 Introduction

idea that money could not be seen as a particular good or commodity. Instead,‘the monetary circulation is, in its nature and main function in the marketeconomy … nothing but a [social] clearing system’ (Schumpeter 1917/18[1956]: 155). This does not mean, however, that money is a creation of the stateor of law as, for instance, Knapp had argued:

[M]oney is as little and in no other sense a creature of the law than is anyother social institution such as marriage or private property. The compar-ison is instructive. … [T]he essential nature of marriage relations explainsthe legal provisions which regulate them, but the legal provisions do notexplain the essential nature and causes of marriage relations. Similarly,money transactions are regulated or shaped by the legal system, but as anobject of regulation they retain a separate existence apart from the legalsystem itself and can be explained only by their own nature or by the innernecessities of the market economy.

(ibid.: 160–1)

Money is thus analysed by Schumpeter as a ‘claim ticket’ and ‘receipt voucher’recognised by every agent in the economy as socially valuable. In this sense,Schumpeter’s analysis of the existence of money provides us with anotherexample of how he builds a ‘bridge’ between economic sociology and economytheory, or between money as an institution and money as the basis of incomecirculation.

Financial markets are another important institution of capitalist economies.As we know, Schumpeter did not consider financial markets as fundamentallyspeculative.7 For him, they participated, together with banks, in the process oftransforming the financing of innovation into more permanent funding. Thus,Schumpeter did not ascribe to financial markets the role that Keynes assignedto them.8 According to Schumpeter, financial markets are neither asautonomous nor as predominant as they are in Keynes’s approach. However,they play a key role in the transformation of saving into investment. This iswhy Schumpeter characterises them as the ‘heart, although [… not …] thebrain’ of capitalist economies (Schumpeter 1939: 127).

The purpose of economic sociology is not only to define the main institu-tional patterns of capitalism but also to analyse its prevailing forms oforganisation. A very good example is provided by the Schumpeterian approachto firms and competition.9 For Schumpeter, the analysis of market forms is anobjective not only of economic theory but also of the ‘science of organisation’and, therefore, of economic sociology. This, of course, explains why the analysisof the evolution of forms of productive organisation received such considerableattention in his writings, be it in the context of his discussion of entrepreneur-ship, of capitalism’s tendency to ‘trustification’, or of their respective impact oninnovations. From this point of view, a significant example of Schumpeter’sapproach is contained in Business Cycles (BC). Chapter 3 of the first volume(Schumpeter 1939: 72–123) is devoted to the analysis of ‘How the economic

Introduction 7

system generates evolution’. In this chapter, Schumpeter formulates his ‘Theoryof innovation’ (ibid.: 87–102). Apart from defining the notion of innovation,this theory of innovation – containing the essence of what Schumpeter calls‘the sociology of industrial society’ – explains the emergence of innovations(ibid.: 96). Far from explaining innovations in terms of some kind of stochasticprocess or as the result of a purely economic transition from old to new produc-tion functions, Schumpeter locates them in economic sociology. This isprecisely what he means when he notes that ‘innovations are always associatedwith the rise to leadership of New Men’ (ibid.). He justifies this view byinvoking a methodological argument that directly reflects his interpretation ofthe relation between economic theory and economic sociology:

The main reason for introducing this assumption [the assumption of therelation between ‘innovations’ and ‘New Men’] into a purely economicargument not primarily concerned with the structure of society is that itprovides the rationale for the preceding assumption.

(ibid.)

More precisely, the emergence of entrepreneurs or the transition from ‘competi-tive’ to ‘trustified capitalism’ is described as a change in the forms of organisation.This change is the result of the emergence of new men or new leaders who,through their innovative activity, generate ‘a process subject to institutionalchange’. Changes in the forms of organisation are therefore primarily sociologicalrather than economic in nature. Entrepreneurs are the new leaders who replacethe owners in the circular flow and, in particular, the old leaders. At some pointin historical time, and as the result of organisational change, the managers ofgiant firms become the ‘new’ leaders, replacing individual entrepreneurs whohave become ‘old’ leader-types. Here again, the sociological distinction betweenleaders and followers appears to be the key to organisational transformations.

The specific interest that Schumpeter took in the study of institutions andforms of organisation also sheds some light on his view of economic rationality.Contrary to conventional economic analysis as well as modern neo-institutionaltheories, Schumpeter did not rely only on the assumption of utility maximisation.Instead, he held the view that human motives were diverse and multi-faceted.Schumpeter was, thus, not a methodological individualist. For him, individualbehaviour was not independent of institutions and social forms of organisation.Entrepreneurial competition, for instance, cannot exist in a warlike society. Butindividual behaviour – whether ‘hedonistic’ or ‘energetic’ – can, through socialinteraction, prompt the emergence of new social rules or institutions thatappear a posteriori, that is, independently from a priori existing individual wills.

History and institutional change

The relation that Schumpeter established between history, economic sociologyand economic theory was instrumental in shaping his approach to institutional

8 Introduction

change. From this perspective, as we have already stressed, the existence of botha specific social structure and the social phenomenon of leadership play a funda-mental role.

The existence of a social structure can best be described as the organisationof society in social groups to which individuals belong or aspire to belong. This,in turn, influences, for example, the way in which they make their choices:

[I]t is society that shapes the particular desires we observe; … wants mustbe taken with reference to the group which the individual thinks of whendeciding his course of action – the family or any other group, smaller orlarger than the family; … the field of individual choice is always, though invery different ways and to very different degrees, fenced in by social habitsor conventions and the like.

(Schumpeter 1912 [1934]: 91)

This social structure also provides the social framework within which institu-tional changes are embedded. A good example of the way in which Schumpeteremploys this framework is provided by his analysis of what happens when the‘New Men’ become entrepreneurs. On the one hand, certain institutions mustalready exist as a matter of ‘logical priority’ (Schumpeter 1939: 114) to renderthe emergence of the entrepreneur feasible. Thus, the existence of a bankingsystem based on credit allows entrepreneurs to employ new means of productionwithout these having to be transferred a priori from existing industries to inno-vative ones (ibid.: 114). On the other hand, these institutions are not simplyforms of social organisation. They also take the form of new behavioural rules,what Schumpeter called ‘the attitudes of the public mind’ (Schumpeter 1950:135).

The social phenomenon of leadership, too, is instrumental in providing anexplanation of institutional change. Social leadership based on ‘energetic’ effortis, in fact, the main source of such change:

Certain other things, such as, for example, the element of ‘effort’, couldperhaps be even more useful for an ‘energetic’ theory of economics thatwould have something to say about economic development. Moreover,changes in human nature, social organisation, etc., often have economiccauses. Finally, even a development that cannot be explained in economicterms often has economic consequences so that we might well have some-thing to contribute to its clarification.

(Schumpeter 1908: 621)

Here, Schumpeter highlights a major characteristic of leaders’ behaviour.Leaders do not follow the logic of prevailing rules, namely to minimise theirefforts in order to reach a given objective. Quite the contrary, they invent newrules to reach new objectives. This invention – or, more precisely, this ‘innova-tion’ – requires effort. Leaders are able to produce this effort because they

Introduction 9

possess an excess of energy that is obviously useless when individual behaviourrelies on routines and is purely ‘hedonistic’. The success of an innovation doesnot, however, depend on its intrinsic characteristics but on it being accepted bythe community of followers or imitators. This process is based on self-organisa-tion and self-reinforcement, and can be found already in Menger as well as inWieser.10 It is based on the phenomenon of social imitation, which is whySchumpeter considers that its analysis falls within the realm of economic soci-ology. As already noted, economic sociology plays here the role of a conceptual‘bridge’ between history and economic theory. For instance, when defining thefunction of the entrepreneur in BC, Schumpeter writes:

For actions which consist in carrying out innovations we reserve the termEnterprise; the individuals who carry them out we call Entrepreneurs. Thisterminological decision is based on a historical fact and a theoretical propo-sition, namely, that carrying out innovations is the only function which isfundamental in history and essential in theory to the type usually desig-nated by that term.


And later, to explain that entrepreneurs’ ‘genealogies display most variedorigins’, he asserts that ‘economic theory and sociology should combine toaccount for their institutional patterns’ (ibid.: 104).

At a given point of historical time, leaders introduce new institutions thatare more adequate to new objectives and followers accept them (or not),thereby turning them (or not) into innovations. This self-organising process isof great general importance to Schumpeter’s explanation of institutionalchange. Hence, the emergence of entrepreneurs as well as of large firms is essen-tially described as the result of leadership. Moreover, the explanation extends tobanks that are simply a ‘new kind of firm’: ‘They are nothing but establishmentsfor the manufacture of means of payment’ (ibid.: 112). Thus, for Schumpeter,banks, like firms, are the result of a process of self-organisation.

Another aspect of Schumpeter’s approach to institutional change derivesfrom the fact that such change takes time. This is mainly a consequence of indi-vidual agents’ resistance and propensity to routine: ‘Everyone knows, of course,that to do something new is very much more difficult than to do something thatbelongs to the realm of routine, and that the two tasks differ qualitatively andnot only in degree’ (ibid.: 99). Schumpeter highlights three ‘classes’ of reasons toexplain such institutional inertia. First, innovations often face an environmentcharacterised by resistance. This can come in very different forms, from ‘disap-proval’ to ‘aggression’ (ibid.: 100). Second, this environment can be welladapted to routine and is, therefore, a priori not prepared to welcome innova-tion. The third ‘class’ of reasons is related to the attitude of individual agentstowards uncertainty: as Schumpeter notes, ‘most people feel an inhibition whenthe possibility of treading a new path offers itself ’ (ibid.).

The notion of resistance to institutional change is particularly important for

10 Introduction

Schumpeter’s approach to capitalist market economies. On the one hand, itexplains why innovations may spread only slowly among entrepreneurs. On theother hand, it also explains the rise of large firms: according to Schumpeter,‘Trustified Capitalism’ is characterised by a tendency for entrepreneurial leader-ship to disappear.

This social function is already losing importance and is bound to lose it atan accelerating rate in the future even if the economic process itself ofwhich entrepreneurship was the prime mover went on unabated. For … itis much easier now than it has been in the past to do things that lie outsidefamiliar routine – innovation itself being reduced to routine.


Obviously, once innovation itself is routinised, resistance to it lessens or evendisappears.

Institutional inertia is, however, not specific to market economies. It ispresent in any type of society. An example is Schumpeter’s analysis of the mili-tarisation of the ancient Egyptian society under the ‘ “New” Empire’. Thisexample is particularly interesting for our purpose since it demonstrates how anexternal event – the war of liberation from the Hyksos – led to the emergenceof a class of professional soldiers. However, having come into existence, thisclass contributed to the emergence and maintenance of a new social and polit-ical organisation based on the centralisation of power under a militaryaristocracy. As Schumpeter noted: ‘Created by wars that required it, the machinenow created the wars it required’ (Schumpeter 1919 [1951]: 33, emphasis in theoriginal). Taken together, these characteristics of Schumpeter’s conception ofinstitutional change permit us to consider what is generally considered theevolutionary element or even nature of his approach.

Schumpeter and evolutionary explanations

Far from confirming the idea that Schumpeter adhered to an evolutionaryunderstanding of economics, a careful investigation of his writings reveals amarked scepticism as regards evolutionary explanations. First, Schumpeterconsiders that this type of explanation is both unscientific and old-fashioned:

[T]he evolutionary idea is now discredited in our field, especially withhistorians and ethnologists, for still another reason. To the reproach ofunscientific and extra-scientific mysticism that now surrounds ‘evolu-tionary’ ideas, is added that of dilettantism. With all the hastygeneralisations in which the word ‘evolution’ plays a part, many of us havelost patience.

We must get away from such things.(Schumpeter 1912 [1934]: 57–8)

Introduction 11

Second, he makes little secret of his hostility towards biological analogies. InDW, he refers to Marshall’s attempt at making use of such analogies, notingthat this did not produce the result one could reasonably have anticipated butrather created ‘the danger of confusion’ (Schumpeter 1908: 538). More gener-ally, Schumpeter was sceptical about any kind of conceptual imports frombiology into economic analysis. In his view, the point was that, even if onewere to assume that all human activity could ultimately be given a biologicalexplanation, economics remained a self-sufficient and independent disciplinebecause ‘the treatment of economic phenomena for themselves, withoutentering into the analysis of their innermost nature, offers more than such ananalysis’, that is, of the ultimately biological explanation of all human action(ibid.).

Third, evolutionary explanations are rather useless since they are toogeneral: ‘As we have stressed constantly, analogies and generalities lead tonothing, only detailed analysis can produce worthwhile results’ (ibid.: 93).

Finally, Schumpeter also criticises the ‘causal-genetic’ explanations of hisAustrian masters, suggesting that they provide a striking example of thedanger of ‘evolutionary’ reasoning: ‘[T]he historical beginnings of aphenomenon by no means always show it in its simplest and purest form, sothat an attempt to get at the essential nature of the problem by genetic treat-ment may easily be misleading’ (Schumpeter 1917/18 [1956]: 157). This is thereason why he preferred functional explanations, generally employed by math-ematical economists, to causal ones (Schumpeter 1908: 47).

Therefore, the analysis of economic development should be based neither onbiological analogies nor on evolutionary explanations but on history. This viewis clearly stated in the beginning of Chapter 2 of the TED. In order fully tounderstand Schumpeter’s position, the best point of departure is undoubtedly afootnote in which Schumpeter specifies his own definition of economic devel-opment:

In the first edition of this book, I called it ‘dynamics’. But it is preferableto avoid this expression here, since it so easily leads us astray because ofthe associations which attach themselves to its various meanings. Better,then, to say simply what we mean: economic life changes; it changes partlybecause of changes in the data, to which it tends to adapt itself. But this isnot the only kind of economic change; there is another which is notaccounted for by influence on the data from without, but which arisesfrom within the system, and this kind of change is the cause of so manyimportant economic phenomena that it seems worth while to build atheory for it, and in order to do so, to isolate it from all the other factors ofchange. The author begs to add another more exact definition, which he isin the habit of using: what we are about to consider is that kind of changearising from within the system which so displaces its equilibrium point that thenew one cannot be reached from the old one by infinitesimal steps. Add success-

12 Introduction

ively as many mail coaches as you please you will never get a railwaythereby.

(Schumpeter (1912 [1934]: 64 fn. 1, emphasis in the original)

This passage requires a number of clarifications. First, economic developmentexcludes balanced growth. In several places throughout his writings,Schumpeter explicitly states that the pure theory of the circular flow is suffi-cient to explain smooth or balanced growth.

Economic development also excludes the study of disturbances that cancreate transitional and local instability but that do not prevent the fundamentaltendency towards equilibrium from working.

Economic development cannot therefore be reduced to the economic effectsof a shock to ‘data’. It necessarily implies an endogenous change. It supposes‘such changes in economic life as are not forced upon it from without but ariseby its own initiative, from within’ (ibid.: 63).

Economic development requires the construction of a specific economictheory, one that substantially differs from the theory of the circular flow,namely, from what, since Walras, economists have called ‘pure economics’.

Capital formation, interest on capital, entrepreneurial profit and crises –all these are phenomena in the face of which pure economics at presentfails. Nevertheless, these phenomena will have to be regarded, for betterof for worse, as ‘economic’ or even (in another sense, perhaps one thatwill emerge in the future) ‘pure economic’ phenomena; the contrarywould not be acceptable to anyone. Therefore, they have to be tackledsomehow, and to the extent that – as seems to us to be evidently the case– they fall outside the realm of any other discipline, they can be conve-niently grouped together – perhaps including some other problems – andthis group of questions can be given one single name, to call it thus,‘dynamics’.

(Schumpeter 1908: 615–16, footnote omitted)

The study of economic dynamics therefore requires the use of preciselythe methodological approach we have outlined above – namely, a combinationof the techniques of economic analysis with the techniques of other socialdisciplines. Schumpeter actually continues the above passage with thefollowing remark: ‘However, this by no means implies … that this dynamicsrepresents a system as does statics, and that, as is the case with statics, it wouldhave a unified method at its disposal as well as interrelated results’ (ibid.: 616).

From the very start, therefore, this theory of dynamics requires importantconceptual changes to the economist’s toolbox and, in particular, the introduc-tion of a new type of economic agent – the entrepreneur – as well as a new typeof rationality – energetic rationality. The theory of economic dynamics doesnot thus involve any kind of ‘evolutionary’ belief or assumption. Instead, what

Introduction 13

it implies is a new approach, based on the combination of history witheconomic theory. Thus, Schumpeter believes it is important to consider ‘histor-ical’ and ‘descriptive facts’ (ibid.). The reason for this is provided bySchumpeter in his HEA:

[T]he subject matter of economics is essentially a unique process in historictime. Nobody can hope to understand the economic phenomena of any,including the present, epoch who has not an adequate command of histor-ical facts and an adequate amount of historical sense or of what may bedescribed as historical experience.

(Schumpeter 1954: 12–13, emphasis in the original)

However, building a theory of economic dynamics also implies changes withinthe conceptual foundations of economic analysis. For instance, one will need toexplain structural changes, such as those made possible by the generalisation ofcredit or associated with the changes in of income distribution (Schumpeter1908: 619).

Economic sociology provides the missing element in the edifice of economicdynamics, which is why Schumpeter assigned to it the status of ‘technique’, aswe pointed out at the beginning of this chapter. It is from this perspective thatwe must interpret what Schumpeter meant when he noted that ‘economicdevelopment is so far simply the object of economic history, which in turn ismerely a part of universal history, only separated from the rest for purposes ofexposition’ (Schumpeter 1939: 583).

Finally, the analysis of institutional and economic changes – whose interac-tion is typical of economic development – can only be achieved by the jointreliance on history, economic sociology and economic theory. Consequently,the study of dynamics cannot be reduced to the analysis of the conditions ofconvergence towards some predetermined ‘dynamic’ or ‘long-run equilibrium’.Economic dynamics is a much more complex phenomenon. It encompasses atleast three types of situations: situations in which the economic systemapproaches a state of ‘ideal equilibrium’ and appears to be moving towards it;situations in which ‘equilibrium points’ do not exist and are replaced by‘neighborhoods of equilibrium’, namely ‘ranges within which the system as awhole is more nearly in equilibrium than it is outside of them’; and situationsin which structural change is so strong that ‘there is no equilibrium at all’(ibid.: 70–1).

Concluding remarks

Schumpeter’s contribution to economics is often reduced to his theory ofentrepreneurial competition. Modern evolutionary economists, in particular,have contributed greatly to this interpretation. Generally speaking, what isoften argued is that the evolutionary nature of Schumpeter’s theory derives fromhis conception of ‘creative destruction’. According to this view, the

14 Introduction

Schumpeterian system does, in fact, provide a typical evolutionary framework inwhich competition acts as a Darwinian process of elimination amongstentrepreneurs, selecting the fittest and thus shaping the future characteristics ofthe economy. This interpretation is, for example, defended in the seminalcontribution by Dosi, Freeman, Nelson, Silverberg and Soete (1988).11, 12 Aswe have argued, this perspective is not convincing. Rather than being rooted inan evolutionary framework based on biological analogies, Schumpeter’s theoryof economic development is intimately linked to history. One major implicationof this link is that social determinism is absent from Schumpeter’s view ofdynamics. Instead, economic development appears to be no more than simply aset of consequences of the structural change that permanently affects theeconomic system in historical time:

The term evolution may be used in a wider and in a narrower sense. In thewider sense it comprises all the phenomena that make an economic processnon-stationary. In the narrower sense it comprises these phenomena minusthose that may be described in terms of continuous variation of rates withinan unchanging framework of institutions, tastes, or technological horizonsand will be included in the concept of growth.


That the theory of economic development falls within the realm of historicalanalysis does not, however, imply that economic dynamics is chiefly an histor-ical phenomenon. Rather, what is meant is that instead of importing biologicalconcepts, analogies and metaphors, the study of economic dynamics, ifconducted from a historical perspective, provides ‘Explanatory Hypotheses’(ibid.: 14) that economic sociology can then help to develop into assumptionsuseful to the economist.

Consequently, Schumpeter’s approach appears to be far more than a meretheory of competition based on selective mechanisms. Put briefly, whatSchumpeter provides is a general theory of economic development that is builtin essential ways on ingredients taken or borrowed from the theory of social,institutional and organisational change. This interpretation is not incompatiblewith the view Shionoya developed in his thought-provoking 1997 contribution,notably in the stimulating chapters on Schumpeter’s economic methodologyand contribution to economic sociology. These themes also receive ampleattention in the present volume, namely in the chapters on the boundariesbetween economic analysis and economic sociology, on time and rationality inSchumpeter’s construct, on entrepreneurship and on the long-term perspectivesof capitalism. However, our interest is broader in that we also focus onSchumpeter’s essential contributions to both the history of economic analysisand economic analysis as such, for example, in the chapters dealing with thetheory of competition and the theory of money and banking.

In our view, Schumpeter’s theory of competition depicts a process of self-organisation and self-reinforcement rather than a process of elimination, as is

Introduction 15

often maintained by modern evolutionary authors. Contrary to this latter inter-pretation, we argue that Schumpeter’s theory of competition draws extensivelyon the more general conception of social leadership he borrowed from theAustrian sociology of capitalism. Entrepreneurs play the role of leaders in theinnovative process, whereas incumbent firm owners either become followers,thereby contributing to the diffusion of technical progress in the economicsystem, or resist following in the footsteps of the entrepreneurial leaders. Whennumerous and influential, their resistance will be effective and they may be ableto hold on to their own techniques of production. Otherwise they will simplydisappear. Therefore, innovation cannot be explained within the conventionalframework assumed by standard theories of the choice of techniques and/ortechnical progress. Instead, it requires an analysis conducted within the realm ofeconomic sociology so as to account for the emergence of ‘new men’ and newattitudes of social leadership.

In conclusion, it is clear that Schumpeter’s approach had little in commonwith what today is generally seen to constitute evolutionary approaches toeconomics, except insofar as it provides space for the use of concepts such asself-organisation or institutional inertia. Schumpeter himself confirms this:

We notice the attempts that were made to apply Darwinian concepts ofstruggle for Existence and Survival of the Fittest to the facts of industrialand professional lie in capitalist society … It may be … that certain aspectsof the individual-enterprise system are correctly described as a struggle forexistence, and that a concept of survival of the fittest in this struggle canbe defined in a non-tautological manner. But if this be so, then theseaspects would have to be analysed with reference to economic facts aloneand no appeal to biology would be of the slightest use; vice versa, any opin-ions that biologists may entertain on the subject would be ruled out aslaymen’s talk.

(ibid.: 789)

Hence, there can be little doubt that Schumpeter did not interpret his theory ofcompetition in terms of a Darwinian evolutionary process. Obviously, this doesnot come as a surprise to us and, interestingly, Hodgson (1994: 146) highlightsthis same passage to underline his own scepticism of evolutionary interpreta-tions of Schumpeter.

Schumpeter thus appears to be an institutionalist rather than an evolu-tionary economist, the main objectives of his efforts at building a new theorybeing the following:

i to build a bridge between economic theory and history, the analysis of insti-tutions providing the intermediary or conduit;

ii to account for the process of emergence of institutions;iii to analyse the effects of institutions on economic activity;iv to establish a link between institutional and economic change.

16 Introduction

In other words, Schumpeter took institutions seriously and tried to show thatanalysing their role is crucial to the understanding of economic dynamics. Thisis why his message is still surprisingly modern.

Notes1 See, for instance, Nelson and Winter 1982; Iwai 1984a and 1984b; Dosi, Freeman,

Nelson, Silverberg and Soete (eds) 1988; Aghion and Howitt 1998).2 Schumpeter 1954: 12. It is also worth noting that Schumpeter wrote: ‘What distin-

guishes the “scientific” economist … is a command of techniques that we class underthree heads: history, statistics and “theory”.’ In brackets, Elizabeth B. Schumpeteradded: ‘Later in this chapter, J.A.S. added … a fourth fundamental field, EconomicSociology’.

3 Schumpeter 1908, 1917/1918 [1956], 1918 [1991], 1918/1919 [1951], 1926, 1927[1951], 1928 and 1950.

4 Shionoya has convincingly shown that ‘in spite of Schumpeter’s harsh criticism andpartial misunderstanding of Weber’s methodology, the substance of their methodolo-gies appear to be the same’ (Shionoya 1997: 216). Therefore, he treated, in fact,institutions and forms of organisation as ideal types (ibid.: 220).

5 See Arena and Gloria-Palermo (2001); Arena (forthcoming).6 ‘Hedonistic’ rationality is the rationality of followers who prefer to minimise their

efforts to attain their ends and, therefore, to rely on routinised modes of behaviour.See Arena and Romani (Chapter 8) for a more in-depth analysis of the distinctionbetween hedonistic and energetic rationality.

7 See Nasica’s contribution to this book. See also Arena 1992.8 See again Nasica’s contribution to this book. See also Arena 1985.9 This theme is studied in more detail later in this volume by Raybaut and Sosthé and

by Arena and Romani.10 See Arena and Gloria-Palermo (2001).11 See especially the article by Clark and Juma.12 Nelson and Winter (1982: 39) were much more cautious: ‘although Schumpeter had

some harsh words for loose invocations of evolutionary ideas in the analysis ofeconomic development (1934, pp. 57–8), we believe that he would have acceptedour evolutionary models as an appropriate vehicle for the explication of his ideas’.

Bibliography

Aghion, P. and Howitt, P. (1998) Endogenous Growth Theory, Cambridge: MIT Press.Arena, R. (1985) ‘Circulations, revenu et capital: Théorie monétaire et tradition non

quantitative’ in R. Arena, A. Graziani and J. Kregel (eds) Production, Circulation etMonnaie, Paris: Presses Universitaires de France.

—— (1992) ‘Schumpeter after Walras: “economie pure” or “stylised facts”?’ in T. Lowry(ed.) Perspectives on the History of Economic Thought, Vol. VIII, Aldershot: EdwardElgar.

—— (forthcoming) ‘Wieser et la sociologie économique: une interprétation’, Cahiersd’Economie Politique.

Arena, R. and Gloria-Palermo, S. (2001) ‘Evolutionary themes in the Austrian tradition:Menger, Wieser and Schumpeter on institutions and rationality’ in P. Garrouste andS. Ioannides (eds) Evolution and Path Dependence in Economic Ideas, Aldershot:Edward Elgar.

Dosi, G., Freeman, C., Nelson, R., Silverberg, G. and Soete, L. (eds) (1988) TechnicalChange and Economic Theory, London: Pinter.

Introduction 17

Hodgson, G. (1994) Economics and Evolution – Bringing Life Back into Economics,Cambridge: Polity Press.

Iwai, K. (1984a) ‘Schumpeterian dynamics, Part I: An evolutionary model of innovationand imitation’, Journal of Economic Behavior and Organisation, vol. 5 (2), June:159–60.

—— (1984b) ‘Schumpeterian dynamics, Part II: Technological progress, firm growth and“economic selection” ’, Journal of Economic Behavior and Organisation, vol. 5 (3–4),September–December: 321–51.

Nelson, R. and Winter, S. (1982) An Evolutionary Theory of Economic Change,Cambridge: Cambridge University Press.

Schumpeter, J.A. (1908) Das Wesen und der Hauptinhalt der theoretischen Nation-alökonomie, Munich and Leipzig: Dunker und Humblot.

—— (1912) Theorie der wirtschaftlichen Entwicklung, Leipzig: Duncker und Humblot.Preface dated, Vienna, July 1911. English translation of 2nd edn as The Theory ofEconomic Development: An Inquiry into Profits, Capital, Credit, Interest, and the Busi-ness Cycle, Harvard University Press: Cambridge, Mass., 1934.

—— (1917/1918) ‘Das Sozialproduct und die Rechenpfennige. Glossen und Beiträge zurGeldtheorie von heute’, Archiv für Sozialwissenschaft und Sozialpolitik, Vol. 44 (A.W.Marget (trans) ‘Money and the Social Product’, International Economic Papers, vol. 6,1956).

—— (1918) ‘Die Krise des Steuerstaates’, Zeitfragen aus dem Gebiet der Soziologie, 4.Translated into English as ‘The crisis of the tax state’, International Economic Papers,vol. 4, 1954. Reprinted in J.A. Schumpeter (1991).

—— (1919) ‘Zur Soziologie der Imperialismen’, Archiv für Sozialwissenschaft undSozialpolitik, 46: 1–39; 275–310. Translated into English as ‘The sociology of imperi-alisms’ in J.A. Schumpeter (1951). Reprinted in J.A. Schumpeter (1991).

—— (1926) ‘Gustav v. Schmoller und die Probleme von Heute’, Schmollers Jahrbuch füerGesetzgebung, Verwaltung und Volkswirtschaft im Deutschen Reiche, 50, I: 337–88.

—— (1927) ‘Die sozialen Klassen im ethnisch homogenen Milieu’, Archiv für Sozialwis-senschaft und Sozialpolitik, vol. 57: 1–67. Translated into English as ‘Social classes inan ethnically homogenous environment’ in J.A. Schumpeter (1951). Reprinted inJ.A. Schumpeter (1991).

—— (1928) ‘Der Unternehmer in der Volkswirtschaft von heute’, in B. Harms (ed.)Strukturwandlungen der Deutschen Vereinigung für Staatswissenschaflichte Forschung,Berlin: Verlag von Reimar Hobbing.

—— (1939) Business Cycles: A Theoretical, Historical, and Statistical Analysis of the Capi-talist Process, 2 vols, New York: McGraw-Hill.

—— (1950) Capitalism, Socialism and Democracy, 3rd edn, London: Allen & Unwin; 1stedn, New York: Harper, 1942. Reprinted (with a new introduction by R. Swedberg),London: Routledge, 1992.

—— (1951) Imperialism and Social Classes, Oxford: Basil Blackwell.—— (1954) History of Economic Analysis, London: Allen & Unwin. Reprinted, London:

Routledge, 1994.—— (1991) The Economics and Sociology of Capitalism (ed. R. Swedberg), Princeton:

Princeton University Press.Shionoya, Y. (1997) Schumpeter and the Idea of Social Science: A Metatheoretical Study,

Cambridge: Cambridge University Press.

18 Introduction

Part I

History of economicanalysis

Introduction

Schumpeter is at the root of the ongoing historiographic practice of classifyingMenger along with Jevons and Walras as the three protagonists of the so-calledmarginalist revolution. All three authors are presented as the independentdiscoverers of very much the same marginal utility principle.1 To Schumpeter,the marginalist revolution boils down to the simultaneous and independentrediscovery of this principle. However, he acknowledges that it is not the ideaof marginal utility per se which is revolutionary. Indeed, this idea had alreadybeen formulated by Gossen earlier on, and, when expressed in basic terms,seems to flow from mere introspection, revealing little more than a triviality. Infact, the label ‘revolutionary’ is bestowed on these three economists not somuch for their exposition of the idea that the evaluation of goods depends onthe intensity of individual needs, but rather for the radical change in the orien-tation of economic theory this principle entails. It is a ‘marginalist revolution’in the specific sense that it proposed totally new foundations of economicthought to replace those of the classical tradition. In this sense, Menger’scontribution is undoubtedly as path-breaking as those of Walras and Jevons.2

What is essential to Schumpeter is the acknowledgement of the marginal utilityprinciple as the foundation of all economic analysis and of human needs as theprimary driving force of economic mechanisms.

In the 1970s, two articles were to mark the beginning of an alternative inter-pretation of the marginalist take-off. Streissler (1972) explicitly asks ‘to whatextent was the Austrian school marginalist?’, and Jaffé (1976) proposes a ‘de-homogenisation’ of Walras, Jevons and Menger. Jaffé insists that the applicationof a single, unifying label to all three authors leads one to overlook the differ-ences between the distinct traditions founded by Walras, Menger and, to alesser extent, Jevons. Streissler’s objective is to show that, in many respects, theAustrian school cannot be considered representative of the marginalist tradi-tion. He thus emphasises Menger’s originality.

There is no doubt that Walras’s and Menger’s work has followed differentanalytical paths and that, today, the Austrian tradition can legitimately claim toconstitute a competitive and independent stream of thought. Therefore thequestion arises of how it was possible for Schumpeter, himself extensively

1 Schumpeter and the oldAustrian schoolInterpretations and influences

Sandye Gloria-Palermo

trained in the Austrian tradition, to have failed to mention the crucial differ-ences separating Menger and Walras and to have subsumed them under a singleindistinct movement.

The fact that Schumpeter underrates the differences between Menger andthe marginalists might lead one to think that he fails to discern any element oforiginality in the early Austrian tradition. This interpretation, however, ceasesto be convincing if one takes into consideration the gist of Schumpeter’s ownanalytical work on which Austrian originality has clearly left its mark. Certainaspects of the Austrian tradition were subsequently taken up by Schumpeter,and it is precisely these aspects which account for his own originality as a theo-rist. Thus, Schumpeter develops a dynamic conception of the role of theentrepreneur, breaking with the static framework of general equilibrium –which he so ardently admired – and providing a theory of economic evolutionin which change originates from within the system.

We are then faced with a striking paradox: as a historian of thoughtSchumpeter provides an interpretation of early Austrian thought according towhich its early exponents are mere marginalists, revealing little originality vis-à-vis Walrasian and Jevonian logic. As a theorist, though, he makes use ofprecisely those elements which constitute Austrian original thought, and whichhe chose to ignore earlier on, to develop his theory of economic evolution. It isby solving this paradox that we intend to shed some light on the nature of therelationship between Schumpeter and the early Austrian tradition.

In order to do so, we first need to go back to Schumpeter’s analytical distinc-tion between the circular flow and economic evolution. This distinctionessentially emphasises the different roles played by consumers and entrepreneursrespectively in these two analytical frameworks. As will become clear, despitecertain embryonic ideas to the contrary, Menger, Wieser and Böhm-Bawerkeffectively position themselves in the circular flow logic, thus justifying theSchumpeterian view of the history of thought. As a theorist, Schumpeter buildson these embryonic ideas, organising them into the coherent and qualitativelydifferent framework of economic evolution. More specifically, the influence ofearly Austrians on Schumpeter’s analysis is manifest at three levels of his anal-ysis: his focus on the economic process, on the nature of economic rationalityand on the role of institutions in economic dynamics.

The circular flow and economic evolution: a qualitativeturning point

Austrians have long been struggling to assert their originality with respect, first,to marginalism and, later, to neo-classical economics. Basically, the problemstems from the ambiguity inherent in the use of identical terms to whichdifferent meanings are attributed: Austrian conceptions of competition,economic rationality and subjectivism differ from those of neo-classicaleconomics. It was only during the inter-war period that, as a consequence of thedebate on planning, Austrians progressively grew fully aware of the conceptual

22 History of economic analysis

differences separating them from mainstream economics.3 As opposed to this, itis Schumpeter’s particular merit to have located himself clearly with regard tothe Walrasian framework as soon as 1912, thereby avoiding the ambiguity initi-ated by Menger and passed on to two generations of economists.

In The Theory of Economic Development (TED) Schumpeter distinguishesbetween the circular flow and economic evolution. While Walrasian theory bearson the reality of the circular flow, it loses all relevance as soon as we turn toeconomic evolution. To a certain extent, general equilibrium theory is realistic:it is an abstraction from reality which captures the essence of the circular flow.4

Indeed, according to Schumpeter, it is realistic to consider human needs as thefundamental basis for the explanation of economic phenomena, to take indi-vidual choices as the basic conceptual unit of analysis and to focus on thegeneral interdependence of plans which is beyond the grasp of consciousness ofthe individual agent.

It is, however, quite clear that Schumpeter does not adopt the framework ofthe circular flow uncritically as part and parcel of Walrasian theory. This isparticularly evident when examining the assumption of individual behaviour. In1908, Schumpeter identifies ‘hedonistic egoism’ with Walrasian rationality anduses this term to criticise the Austrian analysis of individual rationality, andWieser’s psychologism in particular.5 However, in the TED, the Walrasian equi-librium framework is replaced with the circular flow. Simultaneously,‘hedonistic rationality’ is no longer equated with explicit and consciousmaximisation.6 Rather, choices are the result of past experience and are basedon something close to adaptive rationality to which the concepts of rule androutine are central:

In this system of values a person’s whole economy is expressed, all the rela-tions of his life, his outlook, his wants, all his economic combinations. Theindividual is never equally conscious of all parts of this value system; ratherat any moment the greater part of it lies beneath the threshold ofconsciousness. Also, when he makes his decisions concerning his economicconduct he does not pay attention to all the facts given expression to inthis value system, but only to certain indices ready at hand. He acts in theordinary daily round according to the general custom and experience.

(Schumpeter 1912 [1934]: 39)

Individuals automatically reproduce the actions that have proved efficient inthe past. Such behaviour is compatible with the framework of the circular flowas this concerns a stationary economy, free of endogenous disturbances. Astationary economy here is not simply one that reproduces itself. It may grow,but growth will be smooth, determined by gradual exogenous changes of thefundamental variables (endowments, preferences and techniques) to whichindividuals adapt continually. Hence, the circular flow framework coincideswith the analysis of the process of convergence towards a general equilibriumposition whose movement is, in turn, governed by exogenous circumstances. It

Schumpeter and the old Austrian school 23

describes the adaptive process of an economy whose dynamics are determinedby extra-economic factors. Taking particular circumstances as a given, individ-uals know from experience which forms of action are best suited to satisfy theirneeds:

The individual household or firm acts, then, according to empirically givendata and in an equally empirically determined manner. Obviously this doesnot mean that no changes can take place in their economic activity. Thedata may change, and everyone will act accordingly as soon as it is noticed.But everyone will cling as tightly as possible to habitual economic methodsand only submit to the pressure of circumstances as it becomes necessary.Thus the economic system will not change capriciously on its own initia-tive but will be at all times connected with the preceding state of affairs.This may be called Wieser’s principle of continuity.

(ibid.: 8–9)

What is important to notice here is that, even if the fundamental variableschange spontaneously (in response to exogenous factors), no qualitative break-down of the logic of the circular flow follows: agents simply react to thismodification by a process of trial and error. Their present choices are backward-looking in the sense that they are determined by past experience. Economicgrowth is the result of successive quantitative changes smoothly crystallised intoroutinised behaviour:

[A]ll knowledge and habit once acquired becomes as firmly rooted inourselves as a railway embankment in the earth. It does not require to becontinually renewed and consciously reproduced, but sinks into the strataof subconsciousness. It is normally transmitted almost without frictions byinheritance, teaching, upbringing, pressure of the environment. Everythingwe think, feel, or do often enough becomes automatic and our consciouslife is unburdened of it.

(ibid.: 84)

The circular flow is not, however, identical with general equilibrium theory. Itconstitutes a more general framework which takes the Walrasian equilibrium asa central point of reference and elaborates on it. As will be argued, it is,however, also compatible with early Austrian analysis.

If the framework of the circular flow allows us to grasp the essential mecha-nisms of a stationary economy, it cannot provide any insights into the verydifferent phenomenon of economic evolution which concerns a qualitativelydifferent – endogenous – set of dynamics; changes of the fundamental variablesare here the result of the economic process itself:

The position of the ideal state of equilibrium in the economic system,never attained, continually ‘striven after’ (of course not consciously),


changes, because the data change … If the change occurs in the non-socialdata (natural conditions) or in non-economic social data (here belongs theeffects of war, changes in commercial, social, or economic policy), or inconsumers’ tastes, then to this extent no fundamental overhaul of the theo-retical tools seems to be required. These tools only fail … where economiclife itself changes its own data by fits and starts.

(ibid.: 62)

The dynamics of the circular flow results from changes in the exogenouscircumstances which determine the optimal equilibrium position on which theeconomic process converges (though unconsciously as far as individuals areconcerned); evolutionary dynamics results from endogenous shocks, self-gener-ated by the normal working of the economy. But what then is the ‘source ofenergy within the economic system which would of itself disrupt any equilib-rium that might be attained’?7 A new type of individual rationality enters thescene, to wit ‘energetic rationality’. ‘Energetic rationality’ corresponds to aforward-looking form of behaviour, something more akin to creation, ‘[to] doingsomething different from other conduct’ (ibid.: 81 fn. 2). More precisely, inorder to define this alternative type of rationality and, therefore, to grasp theessence of economic evolution, three new factors need to be introduced intothe analysis which are absent from the stationary economy: creative actionsmaterialise through innovations; creation is impossible without the support ofthe institution of credit; the initiator of change is the entrepreneur.

As is well known, the concept of innovation is defined in very broad terms asa new combination of factors leading to the production of a new product, theimplementation of a new method of production, the creation of a new market,the reorganisation of sectors of the economy or to the utilisation of new rawmaterials. However, not all inventions are actually implemented. Most face aselection process by banks deciding on the allocation of finance to specific newcombinations that have become available. Credit is the standard source offunding for the implementation of new combinations (ibid.: 104), because inno-vations disrupt the normal working of the circular flow in which all means ofproduction are employed, no profits are made and, thus, no savings can be accu-mulated. If an entrepreneur uses his savings to finance an innovation, thismeans that past evolution has allowed him to accumulate savings. Even so, bankcredit remains by far the most common and also the logically prior form offinance: ‘And [credit] is the source from which new combinations are oftenfinanced, and from which they would have to be financed always, if results ofprevious development did not actually exist at any moment’ (ibid.: 73).

The entrepreneur’s main role is precisely that of obtaining financial resourcesand employing them in the implementation of new combinations. In this,Schumpeter does not follow the traditional definition of the entrepreneur as arisk-bearer. Nor does he accept the idea that the entrepreneur is simply therecipient of profits, for profit is only the consequence of the entrepreneur’sinnovative role. Entrepreneurs are agents guided by energetic rationality, whose


plan of action does not originate in the (unconscious and routinised) search forways to satisfy their needs but from novel intuitions about future developments.In deciding to switch from a mechanically adaptive response to existing condi-tions to an attitude which impels them to act upon these conditions,entrepreneurs represent the disruptive force in the circular flow: ‘Carrying out anew plan [energetic rationality] and acting according to a customary one [hedo-nist rationality] are things as different as making a road and walking along it’(ibid.: 85).

Given that Schumpeter, the historian of thought, bases his appraisal ofMenger and the early Austrians on the qualitative distinction between thecircular flow and economic evolution, a closer examination of the essence ofthis distinction is in place here. As we have pointed out above, the concept ofeconomic evolution is, of course, directly linked to the introduction of theconcepts of innovation, credit and entrepreneurship. However, given that weare, in the first place, engaged in an exercise of logical reconstruction, what isrequired is a more radical criterion to locate authors either in the logic of thecircular flow or in that of economic evolution. What we need is a criterion thatmay help us understand why, despite their undisputed originality, it is possible tosubsume Austrians along with marginalists under the logic of the circular flow.The criterion we have in mind concerns the idea of consumer sovereignty. Inthe circular flow, the production structure continually adapts to consumers’needs. Individual needs are the ultimate cause of all economic phenomena, andindividual behaviour is governed by the drive for need satisfaction. Givenexternal conditions and individual needs are the two decisive factors that co-operate in determining the result of the economic process: ‘Production followsneeds; it is so to speak pulled after them’ (ibid.: 12).

The logic of economic evolution requires a fundamental alteration of therole of consumer sovereignty: an innovation is not a better way to satisfy a givenneed. Rather, it anticipates a change of tastes and preferences. An innovation isnot a hedonistic response – an adaptation to the existing structure of prefer-ences – but a real activity which prompts adaptation on the side of theconsumer:

Yet innovations in the economic system do not as a rule take place insuch a way that first new wants arise spontaneously in consumers andthen the productive apparatus swings round through their pressure. Wedo not deny the presence of this nexus. It is, however, the producer whoas a rule initiates economic change, and consumers are educated by himif necessary.

(ibid.: 65)

To sum up, in the circular flow the production structure adapts to consumers’needs, whereas in the course of economic evolution needs are endogenouslymodified by the creative actions of entrepreneurs on the supply side.


The Austrians and the logic of the circular flow

We can now return to Schumpeter’s classification of the early Austrians asmarginalists, subjecting it to closer scrutiny. In Schumpeter’s opinion, Menger,Böhm-Bawerk and Wieser are to be regarded as marginalists for one mainreason: they develop the principle of marginal utility, thus breaking with theclassical objective theory of value. The marginalist revolution was, above all, asubjectivist revolution, and the only element of originality in Austrian thoughtwhich Schumpeter recognises concerns their attitude towards the use of mathe-matical tools in economic analysis, that is the fact that their technical weaknessprevented them from attaining the level of abstraction of Walras’s model ofgeneral equilibrium. Menger, in particular, opposed Walras’s view on the rele-vance of the mathematical method for economics. Following Schumpeter, thereason is to be found in Menger’s own ‘defective technique [that] prevented[Menger] from climbing the top of it … mainly because [he] did not understandthe meaning of a set of simultaneous equations’ (Schumpeter 1954: 918);Böhm-Bawerk ‘was an architect, not an interior decorator’ (Schumpeter 1951:159) who left it to his epigones to do ‘the necessary polishing’ and to sort out allthe technical imprecisions of his theory of capital and interest; Wieser ‘was theworst technician of the three great Austrians’ (Schumpeter 1954: 913).8

However, Schumpeter also maintains that, on the whole, their technical weak-nesses led them to provide one of the most clear presentations of the marginalistprinciple.

This view of Austrian economists is based on a rather crude argument. EarlyAustrians are marginalists simply because they contribute to ‘the great reform ofthe theory of value’ (Schumpeter 1951: 85) – that is, to the replacement of theclassical labour theory of value with the idea that human needs are the drivingforce of economic mechanisms. However, this argument is by no means suffi-cient to justify Schumpeter’s unwillingness to recognise the distinctiveness ofthe Austrian tradition. A more profound examination of Menger’s principle ofmarginal utility immediately calls into question his congruence with marginal-ists in this regard. The differences are not simply limited to a differingappreciation of mathematical methods. Rather, a closer comparison of Walras’sand Jevons’s formal approach with Menger’s non-technical presentation revealssubstantial differences.9

Hence, a more subtle argument is called for to explain why Schumpeter doesnot recognise the distinctiveness of the Austrian tradition vis-à-vis margina-lism. One such argument relates to the mentioned distinction between thecircular flow and economic evolution. In this interpretative framework, bothAustrians and marginalists can be regarded as adhering to the logic of thecircular flow. Which are the main elements of Austrian analysis that allow usto uphold this proposition? In what follows, we concentrate on Menger’s anal-ysis and provide a reconstruction of Schumpeter’s interpretation of Menger inlight of the former’s distinction between the circular flow and economic evolu-tion.10


Menger’s analysis is based on a precise methodological stance. The scientificapproach, in economics in particular, is purely analytical and consists inbreaking down complex phenomena into their most simple elements, a logicaldecomposition in terms of relations of causality. On a methodological level,Menger’s objective is thus

to reduce the complex phenomena of human economic activity to thesimplest elements that can still be subjected to accurate observation, toapply to these elements the measure corresponding to their nature, andconstantly adhering to this measure, to investigate the manner in whichmore complex phenomena evolve from their elements according to definiteprinciples.

(Menger 1871 [1950]: 46–7)

The simplest element from which Menger infers universal laws, explaining themost complex economic phenomena by exploring causal relations, is the indi-vidual and his or her behaviour. In Menger’s view the most basic building blockof all analysis is human behaviour directed towards need satisfaction. Thisapproach is defined as the principle of ‘economising’, and, being at the root ofall economic explanation, is the bedrock on which the whole Mengerian edificerests.11

The theory of imputation exemplifies this quest for universalism. Thegeneral principle of imputation consists in evaluating production factors –higher order goods – by attributing to them that fraction of the value of thefinished product – a first order good – which they have contributed to produce.Schumpeter regards this as Menger’s most important contribution to economics,because imputation theory allows the extension of the marginalist principle –‘economising’ in Menger’s terms – to the sphere of production. This theory isthe result of Menger’s attempt to offer a universal theory of value, applicable toall types of economic goods, and to production goods in particular.

This approach clearly implies a hierarchy of consumption and productionstructures: production goods are evaluated on the basis of our knowledge aboutcausal links relating goods (of different orders) to individual needs. The produc-tion structure follows the consumption structure, trying to adapt to individualneeds. Moreover, according to Menger, economic progress results precisely fromimprovements in our knowledge about the causal links between higher ordergoods and individual needs:

If it is generally correct that clarity about the objective of their endeavoursis an essential factor in the success of every activity of men, it is also certainthat knowledge of requirements for goods in future time periods is the firstprerequisite for the planning of all human activity directed to the satisfac-tion of needs. … The second factor that determines the success of humanactivity is the knowledge gained by men of the means available to them forthe attainment of the desired ends. Wherever, therefore, men may be


observed in activities directed to the satisfaction of their needs, they areseen to be seriously concerned to obtain as exact a knowledge as possible ofthe quantities of goods available to them for this purpose.

(ibid.: 89–90)

Individual needs are the ultimate determinant of the whole economic structure.There is then no room for innovation in the Schumpeterian sense because theonly improvements possible result from the acquisition of knowledge about thecausal links between goods and needs. In other words, economic progress flowsfrom a better use of resources due to a better knowledge of production processes.Thus, Menger remains within the logic of the circular flow, providing an accu-rate analysis of the process by which the economy develops towards a higherlevel of consumer need satisfaction.

There is, of course, one potentially serious objection to this interpretation:the figure of the entrepreneur is not absent from Menger’s theory, pulling himtowards the logic of economic evolution. Given that our aim here is to recon-struct Schumpeter’s view of Menger, the main question is how Schumpeterinterprets the Mengerian producer. Essentially, this kind of criticism loses itsforce once it is recognised that Schumpeter considers Menger’s entrepreneur asa mere arbitrator without creative capacity. This firmly shuts the door on anynotion of innovation and, thus, confirms Schumpeter’s classification of thefounding fathers of Austrian thought as analysts of the circular flow. However,a few more details need to be added to the picture: Schumpeter’s makes hisview explicit in the course of his debate with Knight about Menger’s theory ofthe entrepreneur.12 Following Knight, Menger lacks a rigorous concept of theentrepreneur, whereas Schumpeter maintains that not only does Mengerdevelop an exact theory of the entrepreneur, but one that is very close toKnight’s own view. In other words, Schumpeter (1954) identifies in Mengerthe basic elements of the Knightian theory of the entrepreneur. In bothMenger and Knight, the figure of the entrepreneur emerges in response to theuncertainties inherent in the economic process. The entrepreneur is defined asthe residual uncertainty-bearer. In Menger, uncertainty results from the factthat production takes time. To recall, in the theory of imputation the presentvalue of higher order goods depends on the expected value of related first ordergoods. Profits originate in correct or relevant expectations and are a manifesta-tion of a well-managed risk. The entrepreneur’s remuneration depends on hisability to detect the best combinations of goods of different orders so as tosatisfy future consumers’ wants as completely as possible. To put it anotherway, the entrepreneur’s remuneration depends on his ability to process infor-mation about causal relations between higher order goods and individualneeds, and to foresee what these future needs will be. In this view of Menger’sapproach, the entrepreneurial role consists in adapting the production processto the expected future structure of consumption. Menger’s following descrip-tion of the functions of the entrepreneur cannot but confirm Schumpeter inhis interpretation:


Entrepreneurial activity includes: (a) obtaining information about theeconomic situation, (b) economic calculation – all the various computa-tions that must be made if a production process is to be efficient … (c) theact of will by which goods of higher order … are designed to a particularproduction process; and finally, (d) supervision of the execution of theproduction plan, so that it may be carried through as economically aspossible.

(Menger 1871 [1950]: 160)

What this debate with Knight demonstrates very clearly is that, to Schumpeter,the Mengerian entrepreneur is not an innovator.

For the discussion to be exhaustive, it would, of course, be necessary toreconstruct Schumpeter’s view of Böhm-Bawerk’s and Wieser’s position insimilar detail. However, we will restrict the discussion to a few insights, basedon an intuitive interpretation of Schumpeter’s perception that the analysis ofthese authors, too, is concerned with the logic of the circular flow only.

There can be little doubt that, according to Schumpeter, Böhm-Bawerk’scontribution is to be situated in the circular flow framework. It is not our inten-tion to doubt Böhm-Bawerk’s originality with respect to the marginalists. Theanalytical implications of the temporal dimension of the production process aresufficient proof of his independence of thought. However, Schumpeter does notthink this is sufficient to classify Böhm-Bawerk in the alternative framework ofeconomic evolution. His main argument is that innovation – in the sense ofnew combinations – is absent from his analysis:

And what of the third, the ‘roundabout method of production’? If Böhm-Bawerk had kept strictly to his expression ‘adoption of roundabout methodof production’ and if he had followed the indication that it contains, thiswould be an entrepreneurial act – one of the many subordinate cases of myconcept of carrying out new combinations. He did not do this.

(Schumpeter 1912 [1934]: 159)

The only variable upon which producers can act in order to implement themost productive techniques under given conditions is duration of the produc-tion process; the production structure remains static in a Schumpeterian sense,that is, it adapts only to the given structure of consumer needs. In his PositiveTheory of Capital, Böhm-Bawerk (1889 [1959]: 106–7) provides a graphicalrepresentation of the concept of maturity classes of goods, by depicting a cross-section of processes as a series of concentric circles. Schumpeter points to theambiguity inherent in this representation, which conflates all stages of produc-tion into a single framework. This allows a synchronic view of production interms of the simultaneity of the different stages of the process, thus emphasisingthe static nature of Böhm-Bawerk’s approach.13

Schumpeter’s interpretation of Wieser is based exclusively on his theory ofvalue. Schumpeter restricts his analysis to the development of Wieser’s thought


on pure economics, ignoring his writings on societal conflicts of power whichWieser conceptualises in terms of the interactions between leaders and themasses. Yet, these interactions are at the core of Wieser’s analysis of socialdynamics within which the forces guiding the evolution of the realm ofeconomics form, in Wieser’s view, the most powerful engine. Notwithstanding,the theory of value is an indispensable theoretical element of his analysis. Morespecifically it is the essential point of departure for the normative theory of thesocial economy Wieser wishes to develop, as the title of his 1914 book suggests.According to Wieser, economic relationships are the most conspicuous socialrelationships, and economics, in turn, cannot be understood without an under-standing of value.14 The place occupied in Wieser’s work by the theory of valueand the principle of marginal utility is then quite clear: neither is at the core ofhis analysis. Rather, they serve as a point of departure for an inquiry which isreally directed towards a better understanding of the dynamics of socialphenomena.

Schumpeter, however, strictly separates Wieser’s contributions to pureeconomics from his sociological writings, although he recognises that ‘everyelement of [Wieser’s work] formed part of an harmonious whole, which unfoldeditself slowly and grew organically to an imposing height and breadth’(Schumpeter 1951: 301). Even so, Schumpeter does, in effect, focus his atten-tion on Wieser’s contribution to pure economics which basically consists in anelaboration of the Austrian theory of costs and distribution which had alreadybeen outlined by Menger. Schumpeter emphasises the fact that Wieser coinedthe terms ‘marginal utility’ and ‘imputation’ and pioneered the law of opportu-nity costs, or ‘Wieser’s Law’.15 However, by ignoring Wieser’s theory of socialpower, Schumpeter gives weight to the strictly analytical dimension of his work– the theory of value and the question of imputation – that propels Wieser,along with Menger, towards the logic of the circular flow.16

The Austrian influence on Schumpeter

It would be difficult to deny the presence, in the work of early Austrians, oftheoretical elements which, even though they remained at an embryonic stage,are hardly compatible with the marginalist logic. Conventional interpretationsusually pay little attention to these elements, regarding them as ‘introductoryremarks’ (Stigler 1941) or as the manifestation of mathematical weaknesses(Schumpeter 1954). However, it is precisely these elements which have beentaken up and developed by later generations of Austrian authors and whichhave, ultimately, come to constitute the core of Austrian thought. We are herereferring mainly to three elements of Austrian thought: the causal-geneticmethod, the radical conception of subjectivism and the role of institutions.Schumpeter’s writings are clearly influenced by these emergent concepts of the‘old’ Austrian school, which he subsequently organised into the coherent frame-work of economic evolution.

A key aspect of Austrian originality which Schumpeter inherited is a


dynamic vision of economic phenomena. The concern with the economicprocess takes over from that with equilibrium conditions. However, while thefocus on economic processes is often presented as an important distinctivefeature of Austrian thought vis-à-vis mainstream economics, this is not suffi-cient to position Austrians within the framework of economic evolution asSchumpeter understands it. The point is that the notion of process is central tothe whole of Schumpeter’s analysis and, thus, is not specific to the framework ofeconomic evolution. Even when analysing the circular flow, Schumpeter is notso much interested in the characteristics of the predetermined state of generalequilibrium as in the process of convergence towards this point of reference.17

In many ways, the analysis of the circular flow allows Schumpeter to reconcilehis primary concern for the economic process with his admiration for Walrasiangeneral equilibrium analysis.18 His model of the circular flow analyses indi-vidual adjustments to a predetermined state of equilibrium; it concerns theprocess of convergence towards equilibrium. While the Walrasian general equi-librium model remains a key point of reference, Schumpeter’s main concernshifts to the question of how such a state can be attained. When adopting theframework of economic evolution, Schumpeter questions Walras’ static vision.The process he now examines requires a break with the idea of passive adapta-tion to the equilibrium position: ‘I felt very strongly that this was wrong, andthat there was a source of energy within the economic system which would ofitself disrupt any equilibrium that might be attained’ (Schumpeter 1912 [1937]:preface).

To summarise, the economic process is always at the centre of Schumpeter’sanalysis, be this the process of convergence towards equilibrium in the circularflow model or the process of creative destruction in the context of economicevolution. As will be shown below, from the start the analysis of the economicprocess is also at the centre of research in the Austrian tradition.

As regards Austrian thought, the focus on the economic process originates inthe adoption of a more general approach to economic phenomena, to wit acausal-genetic method. The first Austrian author to have made this approachexplicit was Mayer (1932 [1994]: 57) who distinguished between two theoret-ical approaches to the question of price determination: causal-genetic theorieswhich ‘by explaining the formation of prices, aim to provide an understandingof price correlations via knowledge of the laws of their genesis’, and functionaltheories which ‘by precisely determining the conditions of equilibrium, aim todescribe the relation of correspondence between already existing prices in theequilibrium situation’. According to Mayer, functional theories fail to improveour understanding of the economic system since formal relationships depict aparticular situation – a state of equilibrium – which implicitly assumes that theprocess of price formation has already taken place. Causal-genetic approaches,on the contrary, emphasise the search for the primary causes of an economicphenomenon. Understanding an economic event means identifying the forceswhich set in motion the process that, through time, results in an outcomewhich is the phenomenon under scrutiny.19 In the same vein, Schumpeter


(1912 [1934]: 3) defines the activity of the theorist as consisting in identifyingcausal links between economic variables until a primary extra-economic cause isreached: ‘Always we are concerned with describing the general form of thecausal links that connect economic with non-economic data’. Economic expla-nation should focus on the study of causal links rather than be based on theanalysis of interdependence.

A causal-genetic approach to methodology is, without doubt, part of theearly Austrians’ originality and is evidence of their influence on Schumpeter’sanalysis. That the founders of the Austrian tradition advocated such anapproach can be seen from the fact that Mayer’s article was a tribute to Wieserand that it presents Menger’s theory of price as a typical causal-geneticanalysis.20 Whether Böhm-Bawerk, too, developed a causal-genetic approach ofcapital and interest is a more controversial question. However, for our purpose,it is sufficient to note that the aspect of Böhm-Bawerk’s work whichSchumpeter celebrates most concerns precisely his emphasis on the economicprocess. According to Schumpeter, Böhm-Bawerk’s great achievement has beento combine:

his own ideas with Menger’s teachings into a coherent structure, into atheory of the economic process. … [H]e became one of the five or six greateconomists of all time. He gave us an all-embracing theory of the economicprocess – one of the great analyses of economic life on the scale of theClassics and of Marx – conceived on a Mengerian foundation.

Schumpeter (1951: 147)

Subjectivism is normally regarded a central feature of the Austrian school. Atfirst glance, however, this also seems to be the only feature undoubtedly sharedwith Walras and Jevons, and which is at the core of the marginalist, or subjec-tivist, revolution. Our main argument here is that Menger, Wieser and, to alesser extent also Böhm-Bawerk, paved the way for a particular, more generaland more dynamic conception of subjectivism. From the start, subjectivism didnot remain restricted to the demand side, but Menger’s theory of imputationand Wieser’s concept of opportunity costs extended it to the production side.21

In the early Austrians we can, moreover, find the beginnings of a dynamic viewof subjectivism, as defined by O’Driscoll and Rizzo (1985: 22): ‘[dynamic subjec-tivism] views the mind as an active, creative entity in which decision-makingbears no determinate relationship to what went before’, whereas static subjec-tivism is characterised by the fact that ‘the mind is viewed as a passive filterthrough which data of decision-making are perceived. To the extent that thefilter can be understood, the whole process of decision-making is perfectlydeterminate’.

This specific conception of subjectivism has direct implications for thenotion of individual rationality. Austrian subjectivism challenges themarginalist homo economicus whose objective is limited to utility maximisationand whose choice results from comparisons which are limited to a closed set of


alternatives, themselves a sub-set of all available possibilities. The Austrianagents take decisions in a constantly changing environment; the passage of timebeing the cause of the evolution of their perceptions and knowledge at theorigin of the continuous modification of their plans of action.

Note, however, that the writings of Menger, Wieser and Böhm-Bawerkcontain only the basic premises from which dynamic subjectivism can be devel-oped. Menger insists on the importance of acquiring knowledge of causal links;Wieser pioneers opportunity cost theory, and Böhm-Bawerk develops the impli-cations of the temporal dimension of the production process for individualchoices. As was argued above, these authors did not, however, go as far asSchumpeter in that they did not consider the creative role of the entrepreneurproper. Even so, the fact remains that the notion of energetic rationality –entrepreneurial rationality – presupposes a dynamic conception of subjectivism,thus revealing the Austrian influence on Schumpeter’s theory of economicevolution. In addition, Menger and Wieser had already developed the idea thateconomic progress – evolution – results from the interaction of two forms ofrationality which correspond closely to what Schumpeter was later to call hedo-nistic and energetic rationality. What we have in mind is Menger’s and Wieser’stheory of money or, more generally, of institutions: money emerges here as theresult of the interaction between two groups of agents displaying different formsof behaviour. On one side, we find Wieser’s ‘leaders’ or Menger’s ‘innovators’,endowed with superior natural abilities that allow them to detect new processesor new tools, such as, for example, using a unique commodity as an intermediaryof exchange. On the other side, we find Wieser’s ‘masses’ or Menger’s ‘imitators’who simply adapt to the decisions taken by members of the other group as soonas they grow aware of the positive effects, if any, of these new processes ortools.22

Schumpeter’s theory of economic evolution rests, among other things, on ananalysis of the role of institutions, or of particular institutional settings, in theeconomy. More specifically, Schumpeter offers an analysis of credit as an institu-tion that is indispensable to the implementation of innovations byentrepreneurs.

As opposed to marginalism, the topic of institutions is not absent from theold Austrian tradition either, and progressively becomes an essential part of theAustrian research programme. Menger only clarifies his analytical objective inhis 1883 book which focuses precisely on the emergence of organic institutions:‘How can it be that institutions which serve the common welfare and areextremely significant for its development come into being without a commonwill directed toward establishing them?’ (Menger 1883 [1963]: 146). A theoret-ical understanding of the origin and the changes of organically created socialstructures is, according to Menger, the fundamental question any theorist in thesocial sciences should aim to resolve. However, this focus on institutions doesnot change the basic logic of the Mengerian approach, which remains withinthe confines of the circular flow in that he advocates a particular hierarchybetween consumption, production and institutional structures of society: the


institutional structure provides the environment within which economicactivity, organised around the two remaining structures can unfold, and indi-vidual well-being depends on the ability of the production structure to respondto consumer needs.

Wieser’s theory of the emergence of institutions is slightly more complexthan Menger’s in that he adds a theory of power conflicts to the explanation ofinstitutions in terms of the interaction between two types of rationality.According to Wieser, institutions emerge from a process of interaction betweenmasses and leaders. Masses, however, do not react passively to decisions takenby leaders. Rather, the attitude of the masses is essential in that they can acceptor reject the impulses provided by the leaders. Leaders initiate a social move-ment whose actual outcome, owing to the selective influence of the masses, maywell go beyond original expectations: ‘Only a part of the force that builds socialinstitutions is directed by purposes; the final decisive mass-influence operatesbeyond the purpose’ (Wieser 1914 [1967]: 165).

Even if the ‘old’ Austrians – in this case, Menger and Wieser – can be cred-ited with having raised the issue of institutions, a fundamental difference withregard to Schumpeter’s analysis remains: while the basic concept is similar, it isembedded in a totally different perspective. Schumpeter is not at all interestedin the emergence of institutions in abstracto. Rather, his interest is focused onidentifying the nature of capitalist institutions and their role in economicevolution.23 Schumpeter does not, therefore, analyse the factors explaining thedevelopment of credit, taking them as historically given. Instead, he concen-trates on the effect of credit on the process of creative destruction: the role ofcredit is to select between innovative projects of entrepreneurs by decidingwhich of these projects to finance:

The essential function of credit in our sense consists in enabling theentrepreneur to withdraw the producer’s goods which he needs from theirprevious employments, by exercising a demand for them, and thereby toforce the economic system into new channels.

(Schumpeter 1912 [1934]: 106)

Conclusion

The similarities are too striking to deny any influence of the early Austrians onSchumpeter’s theory of economic evolution. The main difference is, however,that Schumpeter understood the path-breaking nature of these ideas, whereasthe Austrians retained an ambiguous stance with regard to marginalism. InSchumpeter’s view, Menger, Wieser and Böhm-Bawerk ultimately have to beclassified within the framework of the circular flow despite certain atypicalelements of thought in their analyses which we have discussed in the lastsection of this chapter. Schumpeter seems to think that these elements are notreally central to their analysis. It is necessary to keep in mind that Menger’stheory of money and institutions, although already outlined in the Grundsätze,


is the main theme of the Untersuchungen über die Methode which Schumpeterdoes not regard as a major contribution to economics in the same sense as theGrundsätze.24 As for Wieser’s concept of the leader and of power conflicts, thisis part of the sociological aspect of his work which Schumpeter, concentratingon Wieser’s theories of value, imputation and subjective cost, rather neglects.Finally, Böhm-Bawerk fails to grasp fully the implications of the temporaldimension of the production process and to extend his concept of roundaboutmethods of production to new combinations in Schumpeter’s sense.

The distinction Schumpeter draws between the circular flow and economicevolution may still be of use today for the classification of the different versionsof the Austrian revival. In particular, Kirzner’s and Lachmann’s theories couldbe classified in the circular flow framework and in that of economic evolutionrespectively: Kirzner’s theory of entrepreneurship focuses on the equilibriumfunction of the entrepreneur, depicted as an arbitrator, whereas Lachmann’sanalysis of the market process emphasises the disrupting force of creation.Lachmann’s analysis may thus be regarded as a fruitful attempt to extendMengerian logic to the framework of economic evolution.

Notes1 Schumpeter (1954: 825) agrees with this ‘familiar tradition from which it is conve-

nient to start, [according to which] this revolution centered in the rise of themarginal utility theory of value is associated with the names of three leaders: Jevons,Menger and Walras’.

2 ‘Menger belongs to those who have demolished the existing structure of a scienceand put it on entirely new foundations’ (Schumpeter 1951: 83).

3 Vaughn 1994, Lavoie 1985, Gloria-Palermo 1999.4 Schumpeter 1912 [1934]: 8.5 ‘We have good reasons to be suspicious of remarks we find everywhere in the litera-

ture under the disguise of psychological assertions … Our examples show clearly thatwhen recourse is taken to the psychology of crises nothing but banalities are theresult’ (our translation from Schumpeter 1908: 545).

6 Note, however, that in the first edition of the TED individual behaviour in thecircular flow still reflects the Walrasian meaning. The change in concept was rathergradual and is explicitly stated in the second (1926) edition only.

7 Preface of the Japanese (1937) edition of Schumpeter 1912.8 It is interesting to note that Schumpeter addresses the same criticism to both

Menger and Jevons.9 Karl Menger (1973), the mathematician son of the economist, provides a mathemat-

ical translation of the Austrian definition of the marginalist principle which revealssubstantial differences with the marginalist definition of Walras and Jevons. Inparticular, the Austrian version does not use any implicit assumptions regarding thecontinuity or differentiability of functions. See Gloria-Palermo 1999 for the implica-tions of these formal differences.

10 This choice can be justified by a series of reasons: first of all, from the point of viewof the history of thought, Menger is unanimously accepted as the founder of theAustrian movement. His work is the sole point of reference common to all Austrianauthors over several generations. Moreover, from an analytical viewpoint, Mengerdeveloped the principles and concepts which were to be adopted and refined by hissuccessors.


11 The term used by Menger 1871 [1950]: 116 is ‘Bedürfnisbefriedigung’, literally thesatisfaction of needs and desires.

12 See Martin 1979 for a more detailed analysis of the debate between Knight andSchumpeter on the theory of the entrepreneur and their interpretation of Menger’streatment.

13 See Egidi 1983 for a more detailed analysis of Schumpeter’s criticism of the staticcharacter of Böhm-Bawerk’s approach.

14 Wieser 1926, translated from German by Hayek (1926 [1952]: 556).15 Schumpeter 1954: 848.16 Schumpeter 1951: 300.17 Recall the celebrated opening sentence of the TED: ‘The social process is really one

indivisible whole. Out of its great stream the classifying hand of the investigator arti-ficially extracts economic facts.’

18 Schumpeter’s expresses his approval in categorical terms: ‘as far as pure theory isconcerned, Walras is in my opinion the greatest of all economists’ (Schumpeter1954: 827).

19 For an analysis of genetic-causal thinking in economics see Cowan and Rizzo 1996.20 It is interesting to notice that Schumpeter develops a similar theory of price determi-

nation. Like Menger, he does not deal with the determination of a single equilibriumprice but looks at an interval of potential values delimited by the respective subjec-tivity of traders (Schumpeter 1912 [1934]: 52).

21 The subjective dimension of marginalist analysis is restricted to the introduction ofsubjective factors on the demand side, the aim being to counterbalance the impor-tance of objective factors passed down from classical theory and still present in themarginalist analysis of the supply side: objective production costs determine supplyin the analysis of Jevons, while Marshall rejects the theory of opportunity costsprovided by Wicksteed and Davenport.

22 See Arena and Gloria-Palermo (2001) for a more detailed comparison betweenMenger, Wieser and Schumpeter on institutions and rationality.

23 De Vecchi 1995.24 ‘It would be unfair to his chief contribution to present this later work as equally

important’ (Schumpeter 1951: 88).

Bibliography

Arena R. and Gloria-Palermo, S. (2001) ‘Evolutionary themes in the Austrian tradition:Menger, Wieser and Schumpeter on institutions and rationality’, in P. Garrouste andS. Ioannides (eds) Evolution and Path Dependence in Economic Ideas, Aldershot:Edward Elgar.

Böhm-Bawerk, E. von (1889) Positive Theorie des Kapitales. Translated into English asCapital and Interest, 2: The Positive Theory of Capital, South Holland, Ill.: LibertarianPress, 1959.

Cowan, R. and Rizzo, M. (1996) ‘The genetic-causal tradition and modern economictheory’, Kyklos: 273–317.

De Vecchi, N. (1995) Entrepreneurs, Institutions and Economic Change: The EconomicThought of J.A. Schumpeter, Aldershot: Edward Elgar.

Egidi, M. (1983) ‘Schumpeter: la genesi della teoria e le transizioni’, Ricerche EconomicheXXXVII: 713–26.

Gloria-Palermo, S. (1999) The Evolution of Austrian Economics: from Menger toLachmann, London: Routledge.


Hayek, F.A. von (1926) ‘Friedrich von Wieser’, Jahrbücher für Nationalökonomie undStatistik. Translated into English in H. Spiegel (ed.) The Development of EconomicThought: Great Economists in Perspective, Durham, NC: Duke University Press, 1952.

Hicks, J. and Weber, W. (eds) (1973) C. Menger and the Austrian School of Economics,Oxford: Clarendon Press.

Higgs, H. (1927) ‘Friedrich von Wieser’, Economica June: 150–4.Jaffé, W. (1976) ‘Menger, Jevons and Walras de-homogenized’, Economic Inquiry 14:

511–24.Kirzner, I.M. (ed.) (1994) Classics in Austrian Economics, Vols 1–3, London: Pickering.Lavoie, D. (1985) Rivalry and Central Planning: The Socialist Calculation Debate Reconsid-

ered, Cambridge: Cambridge University Press.Leser N. (ed.) (1986) Die Wiener Schule der Nationalökonomie, Vienna: Böhlan.Martin, D.T. (1979) ‘Alternative views of Mengerian entrepreneurship’. History of Polit-

ical Economy, vol. 1: 271–85Mayer, H. (1932) ‘Der Erkenntniswert der Funktionellen Priestheorien’. Translated into

English as ‘The cognitive value of functional theories of price’, in I.M. Kirzner (ed.)(1994).

Menger, C. (1871) Grundsätze der Volkswirtschaftslehre. Translated into English as Princi-ples of Economics, Glencoe: Free Press, 1950.

—— (1883) Untersuchungen über die Methode der Socialwissenschaften und der PolitischenOekonomie insbesondere. Translated into English as Problems of Economics andSociology, Urbana: University of Illinois Press, 1963.

Menger, K. (1973) ‘Austrian Marginalism and Mathematical Economics’, in J. Hicks andW. Weber (eds) (1973).

Mitchell, W. (1917) ‘Wieser’s theory of Social Economics’, Political Science Quarterly 32:125–40.

O’Driscoll, G. and Rizzo, M. (1985) The Economics of Time and Ignorance, Oxford: BasilBlackwell.


—— (1912) Theorie der wirtschaftlichen Entwicklung, Leipzig: Duncker und Humblot.Preface dated Vienna, July 1911. English translation of 2nd edn as The Theory ofEconomic Development: An Inquiry into Profits, Capital, Credit, Interest, and the Busi-ness Cycle, Harvard University Press: Cambridge, Mass., 1934. Preface to the Japaneseedition (1937) translated in J.A. Schumpeter (1989).

—— (1951) Ten Great Economists: From Marx to Keynes, New York: Oxford UniversityPress.

—— (1954) History of Economic Analysis, London: Allen & Unwin. Reprinted, London:Routledge, 1994.

—— (1989) Essays on Entrepreneurs, Innovations, Business Cycles and the Evolution ofCapitalism (ed. R.V. Clemence), New Brunswick and Oxford: Transaction Publishers.Originally published in 1951 by Addison-Wesley.

Stigler, G. (1941) Production and Distribution Theories, New York: Macmillan.Streissler, E. (1972) ‘To what extent was the Austrian school marginalist?’, History of

Political Economy 4: 2.—— (1986) ‘Arma virumque cano: Friedrich von Wieser, the Bard as Economist’, in N.

Leser (ed.) (1986).Vaughn, K. (1994) Austrian Economics in America: The Migration of a Tradition,

Cambridge: Cambridge University Press.


Wieser, F. von (1914) Theorie der gesellschaftlichen Wirtschaft. Translated into English asSocial Economics, 1927. Reprinted, New York: Augustus M. Kelley, 1967.

—— (1926) Das Gesetz der Macht, Vienna: Springer.


Introduction

Ten years ago, I looked at the influence Leon Walras’s approach to economicshad on the formation of the Schumpeterian system of thought (Arena 1992).What I tried to argue then was that, in spite of his many assurances to thecontrary, Schumpeter’s own approach to economics clearly differed from that ofWalras. Essentially – and notwithstanding his insistence, in all of his majorwritings, on his high regard for Walras’s economics – Schumpeter built a theoryin which institutions and ‘stylized facts’ occupy an important space, and whichis thus clearly different from Walrasian ‘pure economics’.

This contribution, too, is concerned with the relationship between the worksof Walras and Schumpeter, but focuses on rather a different aspect, namely,Schumpeter’s contribution to Walrasian historiography. My main argumentscentre around two rather unexpected upshots of Schumpeter’s rehabilitation ofWalras’s contribution to economic theory: first, what was intended as a rehabili-tation turned into a reinterpretation of the Walrasian system of thought thathardly qualifies as a faithful representation of Walras’s intellectual project. Yet itis Schumpeter’s interpretation that came to dominate the mainstream under-standing of Walras, thus paving the way to the standard view of Walras’s legacythat has been endorsed and promoted by modern general equilibrium theorists.Second, even though Schumpeter never tired of his enthusiasm for Walras’smerits, much of his discussion of Walras actually stresses important limitationsand shortcomings of the latter’s contribution to economics. This is often over-looked. Yet it was this dissatisfaction with Walras that led Schumpeter totranscend the Walrasian framework and, in so doing, to construct his ownapproach to economic theory.

Why did Schumpeter rehabilitate Walras? The ‘magna charta’argument

In all his major contributions, Schumpeter insisted on his ‘opinion [that] Walrasis … the greatest of all economists’ (Schumpeter 1954: 827). Published in 1908,Das Wesen und der Hauptinhalt der theoretischen Nationalökonomie (DW: The

2 Schumpeter on Walras

Richard Arena

Essence and Main Content of Theoretical Political Economy) is a statement on thenature and limitations of Walras’s pure economics. In it, Walras is described as‘the great master of exact theory’ (Schumpeter 1908: 261). A couple of yearslater, Schumpeter notes in his obituary to ‘Marie Esprit Léon Walras’ that

the theory of economic equilibrium confers to Walras a legitimate claim toimmortality; and this is the great theory, the concept of which, as limpid ascrystal, illuminated the structure of purely economic relations with thelight of a unique fundamental principle.

(Schumpeter 1910 [1951]: 252)

The preface to the Japanese edition of The Theory of Economic Development(TED 1937) similarly emphasises what Schumpeter regards as his lasting debt toWalrasian economic theory (Schumpeter 1951: 159–60). Business Cycles (BC)ascribes to Walras the role of founder of the general equilibrium theory and, inparticular, credits him with having formulated its essential ‘principle’, namely,the transformation of economic relations into a system of mathematical equa-tions for which a unique set of positive solutions exists (Schumpeter 1939: 46).In Capitalism, Socialism, and Democracy (CSD), Schumpeter insists that,amongst the makers of the marginal revolution, it is Walras to whom the claimto scientific precedence must be conferred. And, last not least, the History ofEconomic Analysis (HEA) contains the following well-known praise:

So far as pure theory is concerned, Walras is in my opinion the greatest ofall economists. This system of economic equilibrium, uniting, as it does,the quality of ‘revolutionary’ creativeness with the quality of classicsynthesis, is the only work by an economist that will stand comparisonwith the achievements of theoretical physics. Compared with it, most ofthe theoretical writings of that period – and beyond – however valuable inthemselves and however original subjectively, look like boats beside a liner,like inadequate attempts to catch some particular aspect of Walrasiantruth. It is the outstanding landmark on the road that economics travelstowards the status of a rigorous or exact science and, though outmoded bynow, still stands at the back of much of the best theoretical work of ourtime.


For Schumpeter, Walras’s indisputable scientific pre-eminence contrasts sharplywith the unfair treatment accorded to him in the history of economic analysis.Not only had Walras effectively anticipated all the major contributions by disci-ples of the Lausanne School – Pareto and Barone, for example, built theirarguments directly along Walrasian lines, and ‘there are no unimportant pointsin which Walras’s system remained superior’ ibid.: 861, 987) – but, more oftenthan not, the founding fathers of the marginal revolution unfairly neglectedWalras’s contribution. Hence, even though Walras’s theoretical results had

Schumpeter on Walras 41

predated their own, Marshall, Edgeworth and Cassel were all rather ungenerousin their treatment of Walras’s work (ibid.: 853–4, 862, 893, 953). Finally, Walraswas long ignored by the teaching profession, and it was not before the 1920sthat his contributions received any attention at all in this area of professionaleconomic activity. Bowley’s Mathematical Groundwork of Economics seems to bethe first English textbook to actually make use of Walras’s results (ibid.: 1954:829). It is interesting to note, however, that throughout the book Walras isnever once mentioned explicitly, Pareto being the only theorist of the LausanneSchool to whom the author pays tribute (Bowley 1924: V).

It does not come as a surprise, therefore, that Schumpeter should have beenso insistent on what he regarded as a rehabilitation of Walras’s contribution toeconomics. However, before entering into a more detailed discussion of thisattempt at rehabilitation, it is important to look briefly at Schumpeter’s reasonsfor crediting Walras’s work with such scientific importance and pre-eminence inthe first place. While there are, of course, many good reasons for doing so,Schumpeter attributes particular importance to what we will call the ‘magnacharta’ argument. The expression ‘magna charta’ appears in BC where, afterhaving examined the concept of interdependence central to Walras’s approachto general equilibrium, Schumpeter argues that

[t]he first and foremost task of economic analysis is to explore the proper-ties of that system … What we want to learn before anything else iswhether or not the relations known to subsist between the elements of thesystem are, together with the data, sufficient to determine these elements,prices and quantities, uniquely. For our system is logically self containedonly if this is the case: we can be sure that we understand the nature ofeconomic phenomena only if it is possible to deduce prices and quantitiesfrom the data by means of those relations and to prove that no other set ofprices and physical quantities is compatible with both the data and therelations. The proof that this is so is the magna charta of economic theoryas an autonomous science, assuring us that its subject matter is a cosmosand not a chaos.


For Schumpeter, there is no doubt that the credit for having provided economictheory with its magna charta, with ‘that monument of constitutional law’, mustgo to Léon Walras (ibid.: 45–7; 1954: 242, 268). Essentially, the notion ofmagna charta here refers to the logical foundation or, at least, the consistentframework that Walras provided, thus enabling the economist to describe ‘thegeneral forms of the causal links that connect economic with non economicdata’ (Schumpeter 1934: 5).

To develop this argument had already been a central concern of DW, sinceone of the main objectives of the book is to analyse the nature and limitationsof the general equilibrium approach in economics. It also explains why


Schumpeter was intent on showing how Walras constructed a ‘pure’ model –that is, one built exclusively on an ‘economic logic’ which is autonomous fromany considerations emanating from other social sciences.1 These latter couldobviously be of help in providing and describing the required data, but this wasas far as any association with them could go. Put differently, for Schumpeter,Walras’s pure model is proof of the possibility of constructing a generaleconomic theory that is both self-contained and consistent. It is in this sensethat Walras can actually be seen to have founded economic theory or, to bemore precise, to have drawn up its constitution or magna charta. Schumpeter’s1908 contribution undoubtedly provides the first complete outline of this argu-ment. While, in his later writings, Schumpeter never went into quite as muchdetail as in this early book, he nevertheless maintained the magna charta argu-ment throughout his work: in the TED (e.g. 1934: X), in BC (e.g. 1939: 41) orin the HEA (1954: 16). What one can observe, however, is a gradual shift inemphasis away from praise for the analytical content of Walras’s contribution toadmiration for its usefulness as a methodological device. While in 1908, thetribute paid to Walras extends to both the analytical content and the method-ological device provided by his theory of general equilibrium, in the measure inwhich he developed his own approach to economic analysis Schumpeter grewmore critical of the analytical structure of the Walrasian system. In the TED, hedescribes with great clarity and precision the task of the theoretician asperceived by Walras:

When we inquire about the general forms of economic phenomena, abouttheir uniformities, or about a key to understanding them, we ipso facto indi-cate that we wish at that moment to consider them as something to beinvestigated, to be sought for, as the ‘unknown’; and that we wish to tracethem to the relatively ‘known’, just as any science deals with its object ofinquiry. When we succeed in finding a definite causal relation between twophenomena, our problem is solved if the one which plays the ‘causal’ rôle isnon-economic. We have then accomplished what we, as economists, arecapable of in the case in question, and we must give place to other disci-plines. If, on the other hand, the causal factor is itself economic in nature,we must continue our explanatory efforts until we ground upon a non-economic bottom.


In spite of the limitations of Walras’s theory that he came to emphasise in hislater writings, Schumpeter never changed this view on the magna charta statusof Walras’s methodological ‘discovery’, maintaining throughout that Walras’sgreatest merit had been to show that it was possible to define what, in 1908,Schumpeter called ‘the core of economics’ (1908: 523). Moreover, according toSchumpeter, the existence of this core establishes economics as a science in itsown right. After Walras, economists can legitimately claim that economic


analysis is a perfectly self-contained and consistent science that does not requireany analytical or methodological support from other social sciences.

How did Schumpeter rehabilitate Walras? The emergence ofthe modern Walras interpretation

If Schumpeter thus rehabilitated Walras, the interpretation he provided of hissystem of general equilibrium was not exactly faithful to Walras’s intellectualproject. It did, however, pave the way to what was to become the predominantreading of Walras’s legacy by modern economists.

The analytical content of Walras’s self-contained pure economics

In DW, Schumpeter dedicates an entire chapter to the description of the ‘Natureand essence of exact economics’ (Schumpeter 1908, Pt V, Ch. 1: 523–53) inwhich he is concerned mainly with the relation between economics and othersciences such as, in particular, biology, sociology, psychology, ethics andethnology. His treatment of this question goes beyond a merely epistemologicaldiscussion of the boundaries of specific disciplines in that it focuses on the natureof the non-economic data on which the Walrasian system is built in order todetermine economic unknowns, i.e. prices and quantities. These data correspondto the givens of the Walrasian models of exchange and production, or to what wenow call the ‘fundamentals’ of a general equilibrium model, namely, consumerpreferences, factor endowments and production techniques. Of these, consumerpreferences are clearly exogenous to economic theory since the ‘theory of needs’is situated outside its scope: the explanation of human action is the subjectmatter of other social sciences, and it is thus not up to the economist to uncoverand analyse the motives underlying individual behaviour (ibid.: 146). This viewamounts to an implicit criticism by Schumpeter of the Austrian as well as theGerman Historical School, both of which favour ‘psychologism’ (Calzoni 1982:VII). As regards the two remaining fundamentals – production techniques andfactor endowments – their case is slightly different since, according toSchumpeter, there is no reason, in principle, why the economist could not enterinto a study of their determinants other than that the static general equilibriumframework does not allow this possibility. It is in this sense that Schumpeterinterprets the model of the ‘circular flow’, described in detail in his TED asequivalent to a Walrasian equilibrium:

[The] picture [of the circular flow] may be refined, and made to yield moreinsight into the functioning of the economic system, by means of a wellknown device. We assume all this experience to be nonexistent, and recon-struct it ab ovo [here, Schumpeter writes in a footnote: ‘this method is dueto Léon Walras’], as if the same people, still having the same culture, tastes,technical knowledge, and the same initial stocks of consumers’ andproducers’ goods, but unaided by experience, had to find their way towards


the goal of the greatest possible economic welfare by conscious andnational effort.

We do not thereby imply that people would in practical life be capableof such an effort. We merely want to bring out the rationale of economicbehaviour irrespective of the actual psychology of the households and thefirms under observation. Neither do we aim at giving a sketch of economichistory. Not how the economic process developed historically to the statein which we actually find it, but the working of its mechanism or organismat any given stage of development, is what we want to analyse.


What this passage shows is that, for Schumpeter, there is no place for history inthe Walrasian realm of pure economics. Instead, the picture captured by thegeneral equilibrium model depicts the economy at ‘a given stage of develop-ment’, thus separating pure economics from historical evolution. It should benoted, that this view describes not only Schumpeter’s interpretation of Walras’sposition, but is also compatible with his own more general conception of therelation between history and economic theory discussed at greater length in theIntroduction to this volume. While this does not contradict the important roleSchumpeter attributes to history in economic analysis, it highlights the factthat he clearly regards pure economic theory as distinct from economic analysis.

More originally, Schumpeter also includes ‘economic organisation’ amongthe parameters relevant to pure economics. This implies that, for him, the anal-ysis of economic organisation is mainly situated outside the realm of pureeconomics. At the same time, it provides information required by theeconomist to organise a basic set of data on which to build economic theory,just as production techniques and preferences are treated as givens explained byother social sciences. It is in this sense that Schumpeter argues that:

[a]ll we state is that both, pure economics and the science of organisation,can be separated; that they are independent from one another and thatneither contributes to the concrete results of the other.

(Schumpeter 1908: 157, emphasis in the original)

This ‘science of organisation’ includes ‘the science of state forms’, that of ‘socialrelations and structures’ and that ‘of economic organisation’ (ibid.: 133). Inaccordance with the magna charta argument, the autonomy of pure economicsfrom the theory of organisation derives from the absence of any direct referenceto law, politics and sociology in the Walrasian system. According toSchumpeter, this also explains the assumption of free competition in labour andproducts markets in this system: since the analysis of different forms of competi-tion is part of the theory of organisation, the assumption of free competitionsimply serves the purpose of allowing general equilibrium theory to remainneutral – that is, strictly insulated from issues relating to other sciences, in thiscase organisational theory. Clearly, this interpretation of Walras is strongly


reminiscent of Schumpeter’s own view of the relation between economictheory, economic sociology and the study of institutions in economics.2

This view of the analytical content of Walras’s pure economics is alsoreflected in Schumpeter’s understanding of the process of abstraction adoptedby Walras. For him, Walras’s methodological position is rationalist, holding thatat the start of any scientific investigation is a process of abstraction that allowsthe scientist to transform concrete and complex forms into abstract and simplecategories or concepts. It should be noted in passing that, rather than the use ofmathematics, it is the fact that they share this same process of abstraction thatunderlies the analogy between pure economics and pure mechanics. In Walras’sown words: ‘I continue to believe that my conception of the equilibrium ofproduction is not a fiction but an abstraction completely analogous to theconception of mechanics’ (Walras 1893). In sum, then, pure economics buildson given techniques, preferences and endowments and assumes free competi-tion understood in terms of pure mechanics, thus abstracting from the realphysical world and from any form of potential friction in our daily life.

Interestingly, Schumpeter discusses fundamentals not only in the context ofthe process of scientific abstraction, but also in relation to the magna chartaargument.3 As we have already noted, referring to the self-contained system ofpure economics, Schumpeter interpreted fundamentals as the elements formingthe boundary between non-economic and ‘unknown’ economic data(Schumpeter 1934: 4–5):

Furthermore, pure economics is ‘pure’ not only because it is self-contained,but also because the tools used to analyse the processes of a pure economyare ‘immutable’. That is, they invariably follow ‘the same formula’, even ifthe concrete results of these processes ‘are determined by circumstances,just as the value judgements about them’.


Hence, were we to follow Schumpeter’s interpretation of Walrasian fundamen-tals, it would appear that the main purpose of pure economics is methodologicalin nature: it reveals an accurate method or toolbox for the study of a staticmarket economy. This, however, is far from reflecting Walras’s own perceptionof what pure economics is about. Far from simply providing a technical devicefor the determination of prices and quantities, pure economics is concernedwith the conceptualisation of an ideal society. Put differently, it allows therational investigation of the features of an ideal society, thereby providing abenchmark for the study of the real world.

Is pure economics a rational device for analysing the features of anideal society?

Which parts of Walras’s writings should be regarded as normative analysis, andwhich as positive, has long been a bone of contention among Walrasian scholars.


Jaffé’s well-known 1977 contribution to this discussion has established somecommon ground. Hence, there is now almost complete agreement that a properanalysis of Walras’s general message must take account not only of the Elémentsd’Economie Pure but also of the Etudes d’Economie Politique Appliquée and theEtudes d’Economie Sociale. Moreover, there is little debate as to the contents ofthe two last-mentioned contributions. While the Etudes d’Economie Socialeprovide a normative theory of distribution of societal wealth organised mainlyaround the discussion of a norm of social justice, the Etudes d’Economie PolitiqueAppliquée specify under which conditions and for which goods free competitionis superior, in terms of the normative principles of utility and efficiency, andunder which conditions monopoly is a necessity. More contentious is the inter-pretation of the main message of the Eléments d’Economie Pure. Thus, whereasJaffé (1977), for example, regards the theory of general equilibrium in Walras’sEléments as a description of an ‘imaginary system’ – whose realisation, if it wereto take place, would have to be based on the principles of social justice and thenormative arguments developed in both the Economie Appliquée and theEconomie Sociale – Morishima (1980) interprets the Eléments as a ‘scientificdescription of the real world’, namely ‘the capitalist economy’ (ibid.: 552).Whatever the truth of this matter, the Eléments undoubtedly contain somenormative aspects that must be taken into consideration for any comprehensiveunderstanding of Walras’s main message.

Among others, Lallement (1997) and Herland (2000) have provided arecent discussion of certain aspects of this normative dimension of the Eléments.Specifically, they highlight two concepts or forms of justice. The first of these iscalled ‘commutative’ justice and requires that agents be equal whenever theyare party to an exchange. For Walras, the main safeguards of this form of justiceare, first, unique prices and, second, free competition. Unique prices ensure thatno agent pays more than any other for identical goods. In this case, exchangeexcludes asymmetric pecuniary gain and the conditions of exchange are thesame everywhere for all agents. Free competition, and thus zero entrepreneurialprofits, means that entrepreneurs cannot exploit an artificial advantage (suchas, for example, a position of monopoly) to obtain some unjustifiable share ofsocial wealth. The second form of justice is ‘distributive’ justice. This excludesthe possibility that ‘those who produce little, consume a lot’ and that those‘who produce a lot, consume little’ (Walras 1990: 60). In other words, itrequires that the quantity of goods and utility agents receive in exchange beproportional to their productive contribution. Free competition is a major safe-guard of this form of justice, too, since without commutative justice there canbe no distributive justice. Moreover, free competition implies respect for indi-vidual freedom. This, in turn, ensures that, once exchange has taken place, theinitial inequality of ‘positions’ is reinstated. Put differently, free competitionmeets the requirement of distributive justice that exchange must not interferewith the distribution of initial positions. This notion of distributive justice alsoreceives strong support from the microeconomic marginal productivity theory ofincome distribution that holds that this respect for initial individual positions is


compatible with the highest possible degree of social welfare or need satisfac-tion. Dockès (1996: 58) neatly sums up the importance of these normativeaspects and, in so doing, he also reconciles Morishima’s and Jaffé’s interpreta-tions to a certain extent:

For L. Walras, the Eléments represent, at the same time, a blueprint of the realeconomy (an a priori construction arrived at by way of deduction based onideal types that are simultaneously drawn from reality and being defined), thenecessarily flawless ideal economy, true (in the first place) to the best of self-interests and (less fundamentally, though no less inevitably so) to justice inexchange and, finally, the rational economy that the future will bring bynecessity.

(ibid.: 58)

The contrast between this view of Walras’s message and Schumpeter is obvious.For Schumpeter, not only do both the Economie Appliquée and the EconomieSociale represent the weakest part of Walras’s legacy, but pure economics doesnot need them:

It is the outstanding landmark on the road that economics travels towardthe status of a rigorous or exact science and though outmoded by now, stillstands at the back of much of the best theoretical work of our time.Unfortunately, Walras himself attached as much importance to his ques-tionable philosophies about social justice, his land-nationalization scheme,his projects of monetary management, and other things that have nothingto do with his superb achievement in pure theory. They have cost him thegoodwill of many a competent critic, and must, I imagine, try the patienceof many of his readers. In any case, the tribute above must be understood torefer to his pure theory alone.

(Schumpeter 1954: 827–8)

This scepticism regarding Economie Appliquée and Economie Sociale reflectsSchumpeter’s general view of the relation between ethics and pure economicsaccording to which it is the purpose of economics to describe reality and not toform moral judgements on the nature of this reality. This, in turn, is based onthe argument that

[Pure economics] is not a theory of economic motives. We are notconcerned with the question of whether these motives play a larger or aminor role for human desires and actions. We do not ask what the motivesare that govern men. And this is the only reason why we cannot takeethical motives into account.



Consequently, Schumpeter not only rejects all reference to social optimality(e.g. Schumpeter 1908: 35), but he also refuses to see in Walras’s theory ofincome distribution a manifestation of the features of an ideal society. To him,the microeconomic principle of distribution according to marginal productivityis simply a feature of a static market economy (or the circular flow, discussedin more detail below). It is thus a (positive) fact, rather than a (normative)principle.

This refusal, based on his own view of the purpose of economics, to interpretWalras’s message as an attempt to specify the conditions under which a marketeconomy will function in accordance with principles of social justice ledSchumpeter to insulate pure economics from the Walrasian project as a whole.In so doing, he paved the way to the modern mainstream interpretation ofWalrasian general equilibrium theory as a positive theory of the workings ofmarket economies.

The core of Walras’s theory of general equilibrium

Schumpeter identifies the core of Walras’s general equilibrium theory, and itsthree main characteristics, with great precision. The first of these is clearly thatit places the concept of exchange at the very heart of the Eléments. Walras’spoint of departure is, in fact, a pure exchange between two consumers eachendowed with an initial bundle of goods with all subsequent stages of the anal-ysis representing mere extensions of this basic model of pure exchange. There isno doubt that in 1908 Schumpeter approved of and shared this approach,describing its main assumptions rather eloquently thus:

The exchange relation does not … always exist, not in the isolatedeconomy and not in those elements of isolated economies that can actuallybe found in exchange economies. Even so, in order not to have to dispensewith an already available tool or with general validity of our results, we willsupplement it even where it is absent by interpreting all economic activity asexchange and by assuming that even where no exchange relation exists theeconomy functions as if it existed. Hence, exchange provides, so to speak,the bracket that holds together the economic system or, to use a differentimage, its main wiring.

(Schumpeter 1908: 49–50)4

Of course, Schumpeter’s own view evolved and changed with time. Thus, in theHEA, he remarks that, ‘in the spirit of J.B. Say’, production was reduced to anexchange between goods and services (1954: 1011), noting though ‘the costs ofthis attempt’ – namely, the neglect of the role played by the entrepreneur andby time in the production process. More fundamentally, in CSD, commentingon Marx, he discusses the effects of the productive structure on socialbehaviour, seemingly implying that production cannot be equated with or


reduced to some general notion of exchange and that production more thanmarkets matters for the understanding of capitalist history and reality.

Secondly, based on his distinction of faits humanitaires (that result from theinteraction between individual wills and forms of behaviour) from faits naturels(that are independent of such interaction and result from ‘blind and fatalforces’), Walras clearly considers exchange a ‘natural fact’, a view of exchangethat is incompatible with a contractualist approach. Interestingly, Schumpeter’snotion of exchange is, to some extent, quite close to that of Walras. He argues,for instance, that economic theory must eradicate any trace of psychologismfrom the theory of exchange, rejecting Austrian subjectivism as well as theMarshallian notion of the ‘ordinary business man’ and the conventional notionof homo economicus (Schumpeter 1908: 87). Consequently, a formalistic treat-ment of individual economic behaviour is the only remaining option, implyingthat pure economics, like pure physics, must treat forms of behaviour as if theycould be described objectively. However, if Schumpeter does not reject theWalrasian idea of exchange as a natural fact, he makes it clear that pureeconomics is only a part of economic analysis or ‘economic theory’. Morespecifically, pure economics is a theory of ‘limited validity’ in the sense that it isconcerned only with static analysis (ibid.: XIX, 32–3; 1912 [1934]: 81). Thischaracterisation of static analysis as a theory of limited validity is importantsince it will help to explain Schumpeter’s views on the nature and methods ofdynamic analysis.

The third central feature of the theoretical core of Walras’s analysis is obvi-ously the notion of general interdependence. While Schumpeter highlights theoriginality of Walras’s technical achievement of handling n goods consistently(1954: 911, 991), he argues that, for Walras, the significance of the concept ofinterdependence goes beyond its technical dimension in that it allows Walras togive more precise meaning to the notion of a self-contained system:

Now, if we find that the relation between them [the economic quantities] issuch that to a given value of one or more of these there corresponds oneand only one value of another, we will then call the system uniquely deter-mined. We use the term ‘corresponds’ here in order to indicate that thisvalue of unknown quantities will tend to emerge on its own and that, onceit has emerged, any tendency towards further change will be absent fromthe system. We call this state the state of equilibrium. The individual quan-tities associated with it we call normal or natural.


In sum then, while Schumpeter manages to pin down the theoretical core ofWalras’s theory of general equilibrium very precisely, he does so at a doubleexpense: he removes all elements with a potentially normative content, and heaccepts the limited validity of pure economic theory in this sense.


Metholodogy in Walras and Schumpeter

Following Schumpeter, Walras adopted a particular form of hypothetical deduc-tivism based on the idea that the scientist chooses ‘either hypotheses or axiomsor postulates of assumptions or even principles … [that] … are also suggested byfacts – they are framed with an eye to observations made – but in strict logicthey are arbitrary creations of the analyst’ (1954: 15). Important examples ofthis approach in Walras include, according to Schumpeter, ‘the principle ofvalue’ (i.e. the principle that, in exchange, prices are determined by the interac-tion of supply and demand) and his theory of ‘tâtonnement’. In his view, theseprinciples and hypotheses are adopted not because they were the only validview, but because they provide ‘the most practical way of obtaining results’(1908: 57). Schumpeter thus clearly discards a priorism (e.g. Schumpeter 1908:XIII–XIV, 7, 25) since, in his view, assumptions are neither intrinsically rightnor intrinsically false. Rather, they ‘explain’ reality more or less accurately, anddepending on their degree of pertinence and generality, they allow theeconomist to prefer this or that hypothesis (ibid.: 192–6). His main objection toWalras’s theory of tâtonnement is, therefore, not strictly speaking a logical one.Rather, what he objects to is that it is based on ‘heroic’ assumptions (1954:1002, 1004) and characterised by a ‘hopeless discrepancy from any process ofreal life’ (ibid.: 1015; see also 1939: 46–7). In BC, Schumpeter provides a longlist of the drastic limitations Walras’s theory of tâtonnement faces regarding itsability to account for even the most general features of the workings of a marketeconomy. This comprises the absence of any relation with real world processes,the timelessness of the tâtonnement mechanisms, the absence of technological orsupply and demand lags, the impossibility of trading at false prices, and theabsence of any frictions, any inflexibility of prices and of uncertainty(Schumpeter 1939: 46–7). Schumpeter’s main qualification of hypotheticaldeductivism, as he discerns it in Walras, is then that the choice of theoreticaltools cannot be totally arbitrary since such tools must be sufficiently general inscope to enable the scholar to deal with some set of phenomena characteristicof the real world. This is why, referring to Walras’s pure economics, he arguesthat:

Why then do we replace the observation of reality with such an unsatisfac-tory theory? All we can answer to this is that, in spite of all, we believe thatits results are of sufficient significance, that they cover a considerable, evena very considerable, part of the field to be analysed and that they performwell, even though within limits that must never be lost sight of.


In some sense, Schumpeter’s view of the relation between assumptions andobserved reality could be seen to anticipate Popperian ideas. Referring oncemore to the ‘principle of value’, he argues, for example, that


[S]ince naturally we cannot observe all exchange relations that exist inreality, this principle always has the characteristics of an arbitrary hypoth-esis of which we can make use as long as it does not lead us to aninexplicable contradiction with the facts.

(ibid.: 55)

He does not, therefore, imply that observation is sufficient to either falsify orconfirm a theory. Rather, it allows the scientist to specify the explanatory scopeand, thus, the degree of generality of a theory. This is why Shionaya regardedSchumpeter’s methodological position as an example of ‘instrumentalism’(Shionoya 1997: 56).

This interpretation of Walrasian methodology is clearly highly questionable.Walras never actually adopted assumptions on the basis that they would fit withor explain a specific part of reality. As Dockès has argued recently,

Walras does not construct a hypothetical model. Rather, what he is lookingfor are ‘ideal types’ that are true in the sense geometrical figures are. Thescholar develops his ‘ideas’ and if his reasoning unfolds in a logical manner,not retaining any contradictions, nothing remains to be verified!Therefore, the abstract sciences ‘leave experiments behind them themoment they borrow their “types”. From such “real types” they abstract anddefine “ideal types”. All their theorems are deducted in this way. Thus,geometry arrives at ‘abstract and ideal’ forms (note the requisite coupling ofthese two adjectives) and defines them, starting from real objects disclosedby observation. The scholar contemplates these a priori in order to establishrelations between them as well as laws. In the same manner, starting fromreal phenomena that experience draws from nature or from society, one hasto abstract ideal types, to proceed via ‘definitions’ and then to reflect onthese in order to construct, by way of deduction, natural as well as ‘humani-tarian’ sciences. This is what constitutes the rational method.5

Léon Walras did not, in fact, think it necessary to take recourse to theempirical sciences, situated below their abstract counterparts in the hier-archy of sciences, to ensure verification: Deductions are true so long as thethought process is logical.

(Dockès, 1996: 21–2)

The contrast between Walras’s own methodology and Schumpeter’s interpreta-tion of it is quite clearly striking. Walras welcomes deductivism but does notbelieve that the conclusions thus obtained should be contrasted with someobserved reality in order to confirm or falsify them. He essentially adopts a neo-Kantian approach according to which the only a priori method to establishwhat are true and necessary assumptions is to engage in a process of progressiveabstraction from reality on the basis of which preliminary concepts are beingchosen and adopted. Deduction in this sense relates only to these true andnecessary assumptions. Since this is a strictly logical process, its results are also


true a priori. There is therefore no trace of instrumentalism, logical empiricismor any anticipation of Popper in Walras. What this highlights is the extent towhich Schumpeter contributed to the reinterpretation of Walras’s purely ratio-nalist methodology as one with a much more empiricist emphasis, and thus alsoone with which most modern economists feel more comfortable.

Methodological individualism in Walras?

According to Donzelli (1983: 639), Schumpeter was the first to coin the expres-sion ‘methodological individualism’. A complete chapter of DW is dedicated toa careful definition and detailed discussion of both methodological individu-alism and its main opponent, methodological organicism. There can be littledoubt that Schumpeter’s main conclusion is instrumentalist. For him, the pointis not so much to establish whether one of the two approaches is superior butrather under which circumstances they are relevant: ‘As we have stressedconstantly, analogies and generalities lead to nothing: Only detailed analysescan produce worthwhile results’ (Schumpeter 1908: 93).

The basic question is how society and its economic systems can be analysedwithout reference to the existence of social classes or groups. Schumpeter clari-fies his own position in his article on social classes written in 1927. Referring tothe Marxian analysis of investment, he argues that

[t]his view is a typical example of how bias in favor of a theory blinds thetheorist to the simplest facts, grotesquely distorting their proportions.Manifestly, the captured surplus value does not invest itself but must beinvested. This means on the one hand that it must not be consumed by thecapitalist, and on the other hand that the important point is how it isinvested. Both factors lead away from the idea of objective automatism tothe field of behavior and motive – in other words, from the social ‘force’ tothe individual – physical or family; from the objective to the subjective. It maybe objected that the logic of the social situation forces the individual toinvest his profits, that the individual motivation is only a fleeting interme-diate phase. This is true, as far as it goes, and must be acknowledged by anyreasonable person. Naturally the individual psyche is no more than aproduct, an offshoot, a reflex, and a conductor of the inner necessities ofany given situation. But the crucial factor is that the social logic or objec-tive situation does not unequivocally determine how much profit shall beinvested, and how it shall be invested, unless individual disposition is takeninto account. Yet, when that is done, the logic is no longer inherent solely inthe system as distinct from the individuality of the industrialist himself.

(Schumpeter 1927 [1951]: 155, emphasis in the original)

For Schumpeter, therefore, the task of the social scientist is to study both, indi-vidual particularities and their context as defined by social stratification, inconjunction. The importance of social classes for societal analysis derives from a


number of reasons. First, ‘[t]he class membership of an individual is a primaryfact, originally quite independent of his will’ (Schumpeter 1919 [1951]: 143).Secondly, path dependency is an undeniable fact of social reality and musttherefore be taken into account by the social scientist:

Every social situation is the heritage of preceding situations and takes overfrom them not only their cultures, their dispositions, and their ‘spirit’, butalso elements of their social structure and concentration of power. …When applied to our problem, this means, first, that any theory of classstructure, in dealing with a given historical period, must include prior classstructures among its data; and then, that any general theory of classes andclass formation must explain the fact that classes coexisting at any giventime bear the marks of different centuries on their brow, so to speak – thatthey stem from varying conditions. … [T]here are no amorphous societiesin this sense – societies, that is, in which the absence of our phenomenon[social differentiation and classes] can be demonstrated beyond doubt.

(ibid.: 144–6)

Finally, in any given society, social classes correspond to specific social func-tions:

Every class, in other words, has a definite function, which it must fulfillaccording to its whole concept and orientation, and which it actually doesdischarge as a class and through the class conduct of its members.Moreover, the position of each class in the total national structure,depends, on the one hand, on the significance that is attributed to thatfunction, and, on the other hand, on the degree to which the class success-fully performs the function.

(Schumpeter 1927 [1951]: 179–80)

What follows from this is not however an outright rejection of methodologicalindividualism since

[t]he ultimate foundation on which the class phenomenon rests consists ofindividual differences in aptitude. What is meant is not differences in anabsolute sense, but differences in aptitude with respect to those functionswhich the environment makes ‘socially necessary’ – in our sense – at anygiven time; and with respect to leadership, along lines that are in keepingwith those functions.

(ibid.: 210)

Rather, both self-interest and class interest coexist in every given society(Schumpeter 1919 [1951]: 34). The relative importance of individual motiva-tion, on the one hand, and class interests, on the other, differs according to thetype of society under investigation. While in ‘traditional economies … the


economy is the concern of the whole group or at least is subject to a super-indi-vidual system’ (1918 [1954]: 18), in the case of pure economics, the individualistapproach is essential, in particular since social organisation is included in thedata. This, however, also contrasts with ‘the science of organisation, for instance,and … sociology generally, [where] individualism would not lead us very far’(Schumpeter 1908: 95; also 1912 [1934]: 60 fn. 1; 1939: 75).

Clearly, his reduction of Walrasian economics to pure economics, as heunderstands it, also leads Schumpeter to see in Walras the true father ofmethodological individualism. This is particularly evident in Schumpeter’s latercriticism of the Walrasian entrepreneur (1954: 1000–1), conceptualised interms of an atomistic individual, in contrast to Schumpeter’s own conception ofentrepreneurs in the context of the ‘epoch of social history’ with which they areassociated, ‘a style of life’ and a ‘system of moral and aesthetic values’(Schumpeter 1934: 78).6

Once more, the association of methodological individualism with Walras’séconomie pure amounts to a rather peculiar interpretation of Walrasian views.Contrary to this interpretation, Walras never actually regarded society as theoutcome of the interaction of individual behaviour. Rather, for Walras ‘societyis a natural fact’ and as such cannot be derived from a social contract betweenindividuals:

It was the idea of eighteenth century philosophers, and it is still the idea ofmost of the writers of our epoch, to pretend that society is a conventionalor a free fact and not a natural and necessary fact. From this point of view,man has left the state of nature on a given day to enter into the social stateand, therefore, the latter state rests on a social pact or contract.

(Walras 1867–8 [1990], Vol. IX: 129)

For Walras, this conception ‘within social science, leads to that absolute indi-vidualism which considers the individual as the basis and the unique purpose ofsociety’. This is, however, ‘irrational and contradictory’ (ibid.),

saying that society has been constituted on a given day or is conservedevery day thanks to the agreement of some moral persons, is as if one wouldsay that art or science had been founded on a given day or continue everyday thanks to the agreement of some artists or some scientists … I askwhere these artists or these scientists have been formed, if, before themthere was no art and no science … It is certain that if one suppresses all theartists and all the scientists, art and science would disappear; but the recip-rocal is also true: let us suppress art and science, and there are no moreartists and no more scientists.

(ibid.: 132)

Walras thus distinguishes two different entities that cannot be reduced to oneanother. More specifically, he distinguishes ‘individuals’ from what he calls


‘general social conditions’ (ibid.: 134). Individuals are defined as ‘men consid-ered in abstraction from the society which they belong to’. They are ‘moralpersons seen as accomplishing destinies which differ from all other ones’ (ibid.:134). General social conditions correspond to society as a whole. They are‘society considered in abstraction from the men who belong to it’, which meansthat they correspond to ‘the set of all the moral persons seen as accomplishingdestinies which are in mutual solidarity’ (ibid.). Walras’s conception thereforeexcludes methodological individualism stricto sensu. For him, men and societyboth exist; they are interdependent but neither does society precede individualsnor do individuals precede society. This point of view is clearly very differentfrom what has become the mainstream interpretation of Walras to whichSchumpeter made such an important contribution.

It should be clear by now that, far from being a pure rehabilitation,Schumpeter’s interpretation of Walras is better understood as a reductionist exer-cise that paved the way to the modern perception of Walras. His interpretationcontributed to the ascendance and predominance of an understanding ofWalrasian economics as combining a self-contained analysis based on givenfundamentals with a positive (as opposed to normative) perspective on theconcrete working of a market economy, a hypothetical-deductive method and anindividualist methodology. This interpretation of Walras is rather close to themistaken perception prevailing among those microeconomists for whom theArrow–Debreu theory of general equilibrium represents a reasonably faithfulmodernisation of Walras’s approach. Schumpeter did not, however, limit himselfto an attempt at rehabilitating and reconstructing Walras. As already pointedout, the development of his own theory led to an increasing distance from pureeconomics. However, rather than rejecting Walrasian economics, Schumpeternever wavered in his admiration for Walras and consequently tried to make spacefor his approach within his own different conceptions of statics and dynamics.The remainder of this chapter provides a brief summary of Schumpeter’s attemptat integrating Walras with his own approach to economic theory.7

What is the place of Walras’s pure economics in Schumpeter’sown contribution to economic analysis? Statics and dynamicsrevisited

The first step Schumpeter took in order to afford Walras’s pure economics aplace within his own theoretical framework or approach was to interpret thelatter in terms of static economic analysis.

The circular flow: general equilibrium or stationary state?

In several of his major contributions, Schumpeter (1908, 1912 [1934], 1939)depicted Walras’s general equilibrium as a specific version of his circular flow,describing it as ‘a well known device’, ‘made to yield more insight into the func-tioning of the economic system’ (Schumpeter 1934: 10). The most detailed


version of this interpretation can be found in Chapter 2 of BC. Schumpeter’s inter-pretation is not, however, as straightforward as might appear at first sight. Thus, inhis comment on the first edition of the TED, Böhm-Bawerk (1913a: 13–14)immediately criticised Schumpeter for making use of two incompatible notions ofstatic analysis: one, corresponding to Walras’s general equilibrium – that is,abstract and virtual – and another, corresponding to Schumpeter’s twin notions ofreproduction and routine – that is, concrete, adaptive and meant to depict anactual process. Similarly, Perroux (1965: 66) noted the substitution of a ‘psycho-logical and sociological’ circular flow for a ‘mechanical’ one. These commentsillustrate the difficulty of equating the general concept of economic statics withwhat Schumpeter (1912 [1986]: 206) called a ‘static vision of economic life’.

Essentially, the problem is that Schumpeter’s concept of the circular flowdescribes the process of the simple reproduction of an economy. There is, there-fore, no reason to equate it with Walras’s general equilibrium in particular,rather than with any type of stationary state. This is particularly evident in thedescription of the concept of the circular flow Schumpeter provides in ‘Moneyand the social product’ (1917/1918 [1956]: 151–2). The succession of circularflows over time, which Schumpeter describes here in some detail, fits well withwhat elsewhere he calls ‘the classic scheme of the economic process’ in astationary state, that is ‘advance economics’ (Schumpeter 1954: 554, 564–5). If,instead, we use the notion of ‘synchronisation economics’ he attributes to themarginalist approach, the idea of successive circular flows over time is, however,also compatible with Wicksell’s circular flow, as well as with Walras’s staticequilibrium.

The view that Schumpeter’s circular flow is best understood as describing theprocess of simple reproduction is confirmed by his reformulation of a bartercircular flow as a ‘monetary circular flow’ (1917/1918 [1956]: 152–5). In thisversion, the circular flow is now depicted in two separate monetary processes(Arena 1992). This facilitates the understanding of the transition from thecircular flow to economic development by introducing liquidity constraints foreconomic agents. Note that what this shows is that while money is alreadypresent in Schumpeter’s circular flow, it is absent from Walras’s pure theory ofexchange.

The circular flow: routine or tâtonnement?

The distance between Schumpeter and Walras’s general equilibrium grows oncethe question of the stability (of general equilibrium) is taken into account. In1908, Schumpeter still accepted the theory of tâtonnement as part of Walras’sresearch programme. As he saw it then, the theory of tâtonnement made itpossible to distinguish questions of adjustment, or ‘small variations’, from thoserelating to dynamic analysis, where the former are an integral part of the realsphere dealt with within the framework of static analysis.

The TED already takes a more critical stance (e.g. in particular, 1912 [1934]:83 fn. 1) and, as has been pointed out above, BC goes much further by


attempting to prove that the theory of tâtonnement is inappropriate to analysethe workings of real markets. This sceptical attitude is confirmed in the HEA(Schumpeter 1954: 1009).

Schumpeter attributes this failure to the fact that Walras confused twodifferent problems: ‘I wish to re-emphasize that in general it seems to me anerror to identify the problem of the “tendency” with the problem of “stability” ’(ibid.: 1007 fn. 25). Walras reduced the first problem to the second; he actually

treated the problem of stability in a peculiar way, because it posed itself tohim in connection with what in strict logic is an entirely different problem,namely, the problem of the relation between the mathematical solution ofhis equations and the processes of any actual market.

(ibid.: 1008)

Schumpeter himself abandoned this approach and instead tried to explain anempirical tendency to equilibrium by employing the concept of ‘routine’:

The individual household or firm acts, then, according to empirically givendata and in an equally empirically determined manner … Everyone willcling as tightly as possible to habitual economic methods and only submitto the pressure of circumstances as it becomes necessary. Thus theeconomic system will not change capriciously on its own initiative but willbe at all times connected with the preceding state of affairs.

(Schumpeter 1912 [1934]: 8–9)

The solution to the existence of a tendency towards equilibrium is therefore tobe found in the nature of economic behaviour:

All knowledge and habit once acquired become as firmly rooted inourselves as a railway embankment in the earth. It does not require to becontinually reserved and consciously reproduced, but sinks into the strataof subconsciousness. It is normally transmitted almost without friction byinheritance, teaching, upbringing, pressure of the environment. Everythingwe think, feel or do often enough becomes automatic and our conscious lifeis unburdened of it. The enormous economy of force, in the race and in theindividual, here involved is not great enough, however, to make daily life alight burden and to prevent its demands from exhausting the averageenergy all the same. But it is great enough to make it possible to meet theordinary claims. This holds good likewise for economic daily life.

(ibid.: 84, emphasis added)

The tendency towards equilibrium is therefore ensured, according toSchumpeter, by ‘the rules by which the businessmen form their judgementabout existing business situations’ (Schumpeter 1939: 6). It corresponds towhat, in BC, Schumpeter calls the ‘general business situation’ and to the


‘ordinary routine’ this implies (ibid.: 3 and e.g. 40). Hence, Schumpeter’s viewdoes not coincide with that of Walras on this point since, contrary to what he argued in the TED (1912 [1934]: 10), it is not obvious that optimalbehaviour is identical with routines. Not only is optimisation based on asubstantive view of economic rationality that contrasts with the proceduralview adopted by Schumpeter, but routines can imply satisfactory results that,given the knowledge limitation characteristic of the social environment inwhich agents operate, are not necessarily optimal. Schumpeter appears to beaware of these differences when he distinguishes ‘normal’ from ‘optimal’ actions, noting that

it is society that shapes the particular desires we observe; … wants must betaken with reference to the group which the individual thinks of whendeciding his course of action – the family or any other group, smaller orlarger than the family; … action does not promptly follow upon desire butonly more or less imperfectly corresponds to it; … the field of individualchoice is always, though in very different ways and to very different degrees,fenced in by social habits or conventions and the like.

(ibid.: 91)

In sum then, it appears to be quite clear that it would be mistaken to regardSchumpeter’s notion of the circular flow in real time as equivalent to theWalrasian concept of a given general equilibrium arrived at through individualoptimisation and the tâtonnement.

Beyond the circular flow: from routine to innovation

In spite of these differences, Schumpeter maintained that the Walrasian equi-librium could be understood as a specific design suited to illuminate the notionof circular flow. However, his keenness to assign Walras’s equilibrium concept aprominent role in the construction of modern economic analysis producesfurther difficulties in the context of the analysis of innovation. WhileSchumpeter noted the difficulty Walrasian economic rationality poses for thetheory of production and capitalisation – referring in particular to the problemof making sense of stocks or inventories in a static framework (Schumpeter1954: 1002) – the main point of contention here concerns the conception ofthe entrepreneur.8

Schumpeter argued that, even in the circular flow, explicit and consciousmaximisation was inadequate to describe the behaviour of managers andeconomic agents in general:

[T]he individual is never equally conscious of all parts of this value system;rather at any moment the greater part of it lies beneath the threshold ofconsciousness. Also, when he makes decisions concerning his economicconduct he does not pay attention to all the facts given expression to in


this value system, but only to certain indices ready at hand. He acts in theordinary daily round according to general custom and experience, and inevery use of a given good he starts from its value, which is given to him byexperience.

(Schumpeter 1912 [1934]: 39)9

This experience is the result of the past activity that has taught individuals‘sternly what [they had] to do’ (ibid.: 6). Agents may form false expectations but,through trial and error, they succeed in rectifying wrong decisions. Hence, inSchumpeter’s concept of the circular flow, it is adaptability rather than opti-mality that rules individual behaviour.

Finally, the circular flow is not perfectly stationary, or at least to assume sowould amount to taking an artificially abstract stance. Things become morecomplex once we leave a ‘downright changeless system’ – that is ‘an abstraction’(ibid.: 9). In real life, there is no doubt indeed that ‘the data may change’ (ibid.:8). Faced with these changes – not to be confused with innovations occurringin the process of economic development – ‘everyone will cling as tightly aspossible to habitual economic methods and only submit to the pressure ofcircumstances as it becomes necessary’ (ibid.: 8–9). However, these changes indata

create new situations, adaptations to which require time. And before thatcan happen a great many positive or negative discrepancies between costand receipt occur in the economic system. Adaptation always offers diffi-culties. The mere knowledge of the changed state of affairs is not attainedin most cases with the desirable promptness. … Often perfect adaptationrelative to the formerly existing products is impossible, and of course espe-cially in the case of durable producers’ goods.

(ibid.: 33)

Obviously, things become even more complex once innovations and, thus,economic development are introduced. At this point, the Walrasian approachbecomes entirely useless.10

How did Schumpeter reconcile this new analysis of the entrepreneur withthe Walrasian approach? The answer is already contained in DW and in thefirst German edition of the TED. Schumpeter here distinguishes between twotypes of ‘egoism’ (1908: 86–7): ‘hedonistic egoism’ describes Walrasian rationalbehaviour whereas the notion of ‘energetic egoism’ is reserved to describe anactive and ‘voluntaristic’ behaviour based on a different kind of rationality.11

While these terms disappeared from Schumpeter’s subsequent writings, mainlyto allow him to avoid the charges of ‘sociologism’ or ‘psychologism’ levelled athim, he implicitly held on to the distinction, transforming it into a method-ological device that allowed him to retain Walrasian rationality as a specificform of behaviour valid only in the limited context of static or stationaryeconomic states. In BC, Schumpeter provides a precise explanation of his view:


What we are doing amounts to this: we do not attack traditional theory,Walrasian or Marshallian, on its own ground. In particular, we do not takeoffense at its fundamental assumptions about business behavior – at thepicture of prompt recognition of the data of a situation and of rationalaction in response to them. We know, of course, that their assumptions arevery far from reality; but we hold that the logical schema of that theory isyet right ‘in principle’ and that deviations from it can be adequately takencare of by introducing frictions, lags and so on, and that they are, in fact,being taken care of with increasing success, by recent work developing fromtraditional bases. We also hold, however, that this model covers less groundthan is commonly supposed and that the whole economic process cannotbe adequately described by it or in terms of (secondary) deviations from it… the reasonable thing for us to do, therefore, seems to confine the tradi-tional analysis to the ground on which we find it useful, and to adopt otherassumptions … for the purpose of describing a class of facts which liesbeyond that ground.


Beyond the circular flow: from general equilibrium to economicdevelopment

This characterisation of the Walrasian construction as a theory of limitedvalidity is not confined to the concept of economic rationality, but is further-more required for the explanation of economic change. In the preface to theEnglish edition of the TED, Schumpeter stresses that his book was dedicated tothe elaboration of a ‘body of theory [that] might be contrasted with the theoryof equilibrium, which explicitly or implicitly always has been and still is at thecentre of traditional theory’ (Schumpeter 1934: xi). In other words, the ‘tradi-tional theory of equilibrium’ is of limited validity, requiring another ‘body oftheory’ to complement it. What we are interested in here is not so much thisbody of theory in itself, but what differentiates it from traditional theory and inwhat sense the former can be regarded as more ‘general’.

The first point to note is that, contrary to the theory of equilibrium, thetheory of development is not self-contained in the Schumpeterian sense. As hasbeen pointed out in the Introduction to this volume, this theory requires theuse of economic sociology as well as history.

Secondly, traditional theory is of limited validity since economic progress isnot mainly the adaptive process of adjustment to changes in data assumed by‘the tendency towards an equilibrium position’, but implies endogenous changeand an irreversible process of transition from an existing to a new equilibriumposition.

So far, it might be possible to view Schumpeter’s notion of economic devel-opment as describing not more than changes in Walrasian fundamentals. This,however, disregards the fact that Schumpeter regards capital accumulation anddemographic changes as mere ‘causes of disturbance’ but not as factors of


development per se.12 Changes in consumer preferences are different in thatthey are essentially endogenous; they are more the consequences than thecauses of economic progress. What really matters are ‘changes in technique andin productive organization’.13 This is also why credit institutions, interest andprofit play a role in the Schumpeterian framework of economic developmentbut are denied in the Walrasian theory of capitalisation and money.

Beyond the circular flow: from commodity money and exchange tocredit and development

The substantial tribute Schumpeter paid to Walras also extends to his contribu-tion to monetary theory (1954: 1082). This praise did not prevent him fromrejecting the idea that the concept of marginal utility could be used to deter-mine the value of money in terms of another goods. In his view, using marginalutilities for this purpose implies that

goods are obtained for money because it has exchange value, whereas infact it has exchange value because goods can be obtained for it. The pointis that money not only has no use-value, but also, as a consequence, cannothave exchange value in the same sense as commodities.

(Schumpeter, 1917/1918 [1956]: 162)

From then onwards, Schumpeter always maintained that ‘money not being acommodity, the traditional apparatus of supply and demand cannot be appliedto the solution of the problem of money prices of commodities and of the pricelevels’ (Schumpeter 1939: 547).

Schumpeter’s rejection of the concept of a marginal utility of money was nothis sole disagreement with Walras’s theory of money. Another point ofcontention concerned ‘this Walrasian idea of an encaisse désirée’ which ‘is one ofthe least valuable elements in the great Frenchman’s mighty structure’ (ibid.).Schumpeter, in fact, excluded the function of store of value from the definitionof money proper, precisely because, in his view, money was not a commodity.While we cannot discuss Schumpeter’s own well-known theory of money, creditand finance here in any length, it is important to note that it contrastedstrongly with Walras’s approach.14 According to Schumpeter, Walras totallyunderestimated the role of credit in the working of an economic system:

Walras saw the phenomenon of credit creation quite clearly (though heconfined himself to banknotes). But he considered it as an abuse that oughtto be suppressed and refused for this reason, to make it a normal element ofhis general scheme.

(Schumpeter 1954: 1116 fn. 2)

The origins of this disagreement with Walras go back a long way. As early as1908 Schumpeter criticised Walras for his tendency to treat monetary theory as


an ‘annex to the rest of our scientific edifice’ (Schumpeter 1908: 282) and forconsequently ignoring the fact that money was not only a useful but also an indis-pensable element of economic life.

What this brief comparison of Walras’s and Schumpeter’s approach to thetheory of money can show is that, here too, Schumpeter failed to disentanglehimself convincingly from an essential analytical difficulty. While he appears toconsider Walras’s theory of money as valid within the static context of thecircular flow – once again restricting its validity to a specific case – his ownconception of monetary exchange clearly differed from Walras’s, even in astationary economic world. As a matter of fact, monetary theory is in manyways the area of analysis in which the analytical tension between the Walrasianand the Schumpeterian approaches is particularly evident.

Concluding remarks

Schumpeter’s interpretation of Walras is of great interest as well as impressivelythorough. There are three main points that emerge from our analysis and that,by way of conclusion, we would like to highlight.

First, there can be little doubt that the consistent tribute Schumpeter paid toWalras in all of his major writings strongly contributed to the latter’s rehabilita-tion, in particular in the English-speaking world. What we have labelled the‘magna charta’ argument, in fact, establishes Walras as the true founder ofeconomic theory as a self-contained scientific discipline.

Second, it should have become clear that Schumpeter’s interpretation wasinstrumental in promoting an understanding of Walras that fails to do justice tohis broader social, as well as scientific, project. The picture Schumpeter paintsof Walras, even though definitely not faithful to Walras’s actual message,manages to convey a powerful and consistent view of his work. As such, itpaved the way to the standard interpretation of Walras endorsed by moderngeneral equilibrium theorists. From this point of view, Schumpeter’s contribu-tion is not so much that of a faithful historian of Walras’s economic analysis butof the founder of contemporary neo-Walrasian thought.

Finally, and paradoxically, Schumpeter developed a framework of his ownthat contradicted the Walrasian theoretical system more often than not. Hisway of solving this paradox was to consider the Walrasian system as a constructof limited validity that could be subsumed under his own more general concep-tion of capitalist market economies. However, as we have argued, this device isnot always convincing. Analytical tensions remained, and still remain today,between the modern inheritors of the Walrasian and Schumpeterian traditions.

Notes1 The term ‘economic logic’ is used by Schumpeter himself (1908: 613).2 See Legris and the Introduction to this volume.3 Following Walras, Schumpeter (1908: 429) also compares economic theory to ‘pure

mechanics’ because it is an accurate science.


4 Note that the term ‘exchange relation’ (Tauschrelation) here has a purely quantita-tive meaning, i.e. relative price. Schumpeter provides the following definition: ‘aready relation between the economic quantities … the price, or better the exchangerelation’ (ibid.).

5 Note that this distinction is specific to Walras. While sciences naturelles refers tonatural phenomena, the notion of sciences humanitaires designates, in Walras’s termi-nology, phenomena resulting from the exercise of human will that is free but notarbitrary.

6 Schumpeter (1908: 438) refers to Walras’s ‘entrepreneur, ne faisant ni benefice, ne perte’.7 The interested reader can find a more detailed account of some of the points made

in the following section in Arena (1992).8 See Arena and Romani in this volume.9 This argument echoes another distinction Schumpeter made in a paper on ‘The

meaning of rationality in the social sciences’ (1940), reprinted in R. Swedberg, TheEconomics and Sociology of Capitalism, Princeton: Princeton University Press, 1991,namely, that between the rationality of the observer and the rationality in theobserved (see also Arena and Romani in this volume).

10 See Arena and Romani in this volume, and also Arena 1992: 133–5.11 See Santanelli and Pesciarelli 1990: 684–7 and Arena 1992: 132–5.12 See Schumpeter 1912 [1934]: 60, fn. 1 or 1939: 75.13 Schumpeter 1912 [1934]: 60, fn. 1 and 1939: 84.14 On this point, see the articles by Dangel-Hagnauer, Festré and Nasica in this

volume. See also Arena 1992: 134–40.

Bibliography

Arena, R. (1992) ‘Schumpeter after Walras: “economie pure” or “stylised facts”?’, in T.Lowry (ed.) Perspectives on the History of Economic Thought, Vol. VIII, Aldershot:Edward Elgar.

Böhm-Bawerk, E. von (1913a) ‘Eine “dynamische” Theorie des Kapitalzinses’, Zeitschriftfür Volkswirtschaft, Sozialpolitik und Verwaltung, 22: 1–62.

—— (1913b) ‘Eine “dynamische” Theorie des Kapitalzinses: Schlussbemerkungen’,Zeitschrift für Volkswirtschaft, Sozialpolitik und Verwaltung, 22: 640–56.

Bowley, A. (1924) The Mathematical Groundwork of Economics – An Introductory Treatise,Oxford: Claredon Press.

Calzoni, G. (1982) Introduction to J.A. Schumpeter (1908 [1982]), Rome: Laterza.Dockès, P. (1996) La Société N’est Pas un Pique-nique – Léon Walras et l’Economie Sociale,

Paris: Economica.Donzelli, F. (1983) ‘Schumpeter e la teoria economica neoclassica’, Ricerche Economiche,

XXXVII, 4: 634–90.Herland, M. (2000) ‘Léon Walras ou l’Apothéose de l’Economie Politique Sociale’, in

P. Dockès, L. Frobert, G. Klotz, J.-P. Potier, A. Tiran (eds), Les Traditions EconomiquesFrançaises – 1848–1939, Paris: CNRS Editions.

Jaffé, W. (1977) ‘The normative bias of the Walrasian model: Walras vs. Gossen’, Quar-terly Journal of Economics, 91, August: 371–87.

Lallement, J. (1997) ‘L’économie pure de Walras est-elle normative?’, in H. Brochier,R. Frydman, B. Gazier and J. Lallement, L’Economie Normative, Collection GrandsDébats, Paris: Economica.

Morishima, M. (1980) ‘William Jaffé on Léon Walras’, Journal of Economic Literature,18(2) June: 550–8.

Perroux, F. (1965) La pensée économique de Joseph Schumpeter, Geneva: Editions Droz.


Santarelli, E. and Pesciarelli, E. (1990) ‘The emergence of a vision: The development ofSchumpeter’s theory of entrepreneurship’, History of Political Economy, 22(4) Winter:677–96.


——(1910) ‘Marie Esprit Léon Walras (1834–1910)’, Zeitschrift für Volkswirtschaft,Sozialpolitik und Verwaltung, Vol. XIX. Translated into English by W.F. Stolper andreprinted in Ten Great Economists: From Marx to Keynes, New York: Oxford Univer-sity Press, 1951.

—— (1912) Theorie der wirtschaftlichen Entwicklung, Leipzig: Duncker und Humblot.Preface dated Vienna, July 1911. English translation of 2nd edn as The Theory ofEconomic Development: An Inquiry into Profits, Capital, Credit, Interest, and the BusinessCycle, Harvard University Press: Cambridge, Mass., 1934. Preface to the Japaneseedition (1937) translated in J.A. Schumpeter (1989). Italian translation of ch. 2 of1st edn in J.A. Schumpter (1986).

—— (1917/1918) ‘Das Sozialproduct und die Rechenpfennige. Glossen und Beiträge zurGeldtheorie von heute’, Archiv für Sozialwissenschaft und Sozialpolitik, Vol. 44. Trans-lated into English by A.W. Marget as ‘Money and the social product’, InternationalEconomic Papers, vol. 6, 1956.

—— (1918) ‘Die Krise des Steuerstaates’, Zeitfragen aus dem Gebiet der Soziologie, 4.Translated into English as ‘The crisis of the tax state’, International Economic Papers,vol. 4, 1954.

—— (1919) ‘Zur Soziologie der Imperialismen, Archiv für Sozialwissenschaft undSozialpolitik, 46: 1–39; 275–310. Translated into English as ‘The sociology of imperi-alisms’ in J.A. Schumpeter (1951).

—— (1927) ‘Die sozialen Klassen im ethnisch homogenen Milieu’, Archiv für Sozialwis-senschaft und Sozialpolitik, vol. 57: 1–67. Translated into English as ‘Social classes inan ethnically homogenous environment’ in J.A. Schumpeter (1951).

—— (1939) Business Cycles. A Theoretical, Historical, and Statistical Analysis of the Capi-talist Process, 2 vols, New York: McGraw-Hill.

—— (1951) Imperialism and Social Classes (edited by P. Sweezy), New York: Augustus M.Kelley.


—— (1986) ‘Il fenomeno fondamentale dello sviluppo economico’. Italian translation ofch. 2 of the 1st edn of J.A. Schumpeter (1912). With an introduction by E. Pescia-relli and E. Santarelli, Quaderni di Storia dell’ Economia Politica, IV/1–2: 171–271.


Shionoya, Y. (1997) Schumpeter and the Idea of Social Science: A Metatheoretical Study,Cambridge: Cambridge University Press.

Walras, L. (1867–8) ‘Théorie Générale de la Société’, in P. Dockès, P.-H. Goutte,C. Hubert, C. Mouchot, J.-P. Potier, J.-M. Servet (eds), Etudes d’Economie Sociale,Les Oeuvres Economiques Complètes d’Auguste et Léon Walras, Vol. IX, Paris:Economica, 1990.

—— (1893) ‘To Johan Gustave Knut Wicksell’ (letter No. 1170), in W. Jaffé (ed.)Correspondence of Léon Walras and Related Papers, Vol. II (1884–97), Amsterdam:Royal Netherlands Academy of Sciences and Letters, 1965.


Introduction

For anyone wishing to understand Schumpeter as an historian of economicthought, the first port of call is his History of Economic Analysis (HEA). Writingsuch a history involves three successive steps: the definition of a researchsubject, the adoption of a research method and the retrospective evaluation – orthe reconstruction – of past thought. It is along these lines that we will discussthe main features of Schumpeter’s research programme in the first section ofthis chapter. This will allow us to gain a better understanding of Schumpeter’smethodological approach to historical analysis, which has been well captured inBlaug’s tongue-in-cheek expression of ‘anthropomorphic sin’, referring to thetendency among historians of analytical thought to judge ‘older writers [exclu-sively] by the canons of modern theory’ (Blaug 1962 [1997]: 1).

As we will argue in the second section, one of the most revealing examplesof Schumpeter’s approach to the history of economic thought is his reading ofMarshall. His interpretation, which pays homage to the innovative nature ofMarshall’s contribution to economic theory, seems logically well founded andhas been well received by scholars of economic thought. However, on closerexamination, it is evident that Schumpeter often twists his material – that is,Marshallian thought – and in places even rids it of contentious implications tosuit his own hypothesis.

Schumpeter’s reconstructive approach

Questions put to history

In the introduction to his HEA, Schumpeter clarifies the nature of his researchproject, arguing that it differs from that of an historian of economic thought:‘By History of Economic Analysis I mean the history of the intellectual effortsthat men have made in order to understand economic phenomena or, whichcomes to the same thing, the history of the analytic or scientific aspects ofeconomic thought’ (1954: 3, emphasis in the original). The approach to thehistory of thought implicit in this remark has been discussed, and to someextent also embraced, by Blaug in his Economic Theory in Retrospect.

3 Schumpeter on Marshall

Nathalie Duval

Blaug discusses two polar approaches to the history of economic thought:‘Relativism’ regards the understanding and justification of any theory as possibleonly in terms of the meta-economic contingencies of its time, while ‘absolutism’‘has an eye only for the strictly intellectual development of the subject, regardedas a steady progression from error to truth’ (1962 [1997]: 2). The historian ofthought thus appears to be confronted with a methodological choice, although,as Blaug points out, one that ‘depends entirely on the questions that [he]wish[es] to raise’ (ibid.: 7). Blaug then outlines his own ‘question’, which is toput the internal coherence and explanatory scope of modern economic theoryto the test by investigating a past perceived as a succession of analytical effortsdriven by the desire to achieve scientific perfection. He essentially advocateswhat could be called a ‘reconstructive approach’ to the history of economicthought, one based on a retrospective evaluation of past theories through thefilter of modern economic analysis:

Rational reconstructions … treat the great thinkers of the past as if they arecontemporaries with whom we are exchanging views; we analyse their ideasin our terms in order to locate their mistakes and to verify our fond beliefthat there has been progress in the course of intellectual history.


To a large extent, Schumpeter’s view of the history of economic thought issimilar to what Blaug calls ‘absolutism’. Even though he persistently emphasisesthe importance of the political and socio-historical context of economic analy-sis, this Zeitgeist remains autonomous from analysis per se:

We shall … never neglect the general environment of economic thought… But these environments and their historical changes are never our mainobject of interest. They come in as favorable or inhibiting influences uponanalytic work, which shall remain the hero throughout our play.


Schumpeter justifies his methodological choice by arguing that, as the disci-pline becomes more autonomous and professional, economists begin to applyprocedural rules that guarantee a rigorous, objective and universal discourse.This also enables the historian of thought to separate out theoretical thoughtfrom contextual influences, and to reconstitute the course of intellectual historyculminating in modern economics. It is along these lines that Schumpeterformulates his own ‘question’: in the HEA, he stresses explicitly that his prin-cipal objective is to describe the ‘process of the Filiation of Scientific Ideas …by which men’s efforts to understand economic phenomena produce, improve,and pull down analytic structures in an unending sequence’ (ibid.: 6).

Given his definition of science as ‘tooled knowledge’ (ibid.: 7), Schumpeterperceives the progress of economic science mainly in terms of the developmentof its techniques of analysis:

Schumpeter on Marshall 67

[A] new apparatus poses and solves problems for which the older authorscould hardly have found answers even if they had been aware of them. Thisdefines … in what sense there has been ‘scientific progress’ between Milland Samuelson. It is the same sense in which we may say that there hasbeen technological progress in the extraction of teeth between the times ofJohn Stuart Mill and our own.

(ibid.: 39)

In the HEA, he defines ‘Economic Analysis’ as the ensemble of theoretical,historical, statistical and socio-economic techniques. However, it soon becomesevident that he effectively restricts his inquiry to ‘analysis’ in the more limitedmodern sense of ‘theory’ only, that is, to a ‘box of tools’ (‘hypotheses’ andconcepts ‘by which results may be extracted from the hypotheses’ (ibid.: 15)).

To Schumpeter, the only meaningful way of talking about progress ineconomics is in terms of this instrumentalist or analytical angle (ibid.: 40). Asthe very idea of progress involves the use of comparisons, it presupposes anirrefutable point of reference that does not exist in economic thought. Thus,because of the nature of his inquiry into the history of thought, he effectivelyadopts the methodological principles of rational reconstruction, which leadshim to apply to the theories he examines an implicit ‘standard’ of evaluation:

Now, our ability to speak of progress … is obviously due to the fact thatthere is a widely accepted standard … that enables us to array differenttheories … in a series, each member of which can be unambiguouslylabelled superior to the preceding one. … [T]his array is associated with thelapse of time in the sense that the later theory … almost always holdshigher rank in the array of analytic perfection.

(ibid.: 39–40)

It thus seems that Schumpeter accepts the risk of bias since his perception ofscientific progress as being of an essentially cumulative nature induces him toseek the precursors of modern economic theory.

The standard of reconstructive evaluation

Schumpeter’s standard consists of a contemporary economic model elevated tothe status of arbiter of the scientific contents of past contributions. In applyingthis standard, the historian effectively adopts a subjective perspective on thepast. In Schumpeter’s case this perspective is shaped by his definition of moderneconomics:

We have … no choice but to interpret and to appraise every piece of tooledknowledge … in the light of our standards, since we have no others. Theyare the results of a development of more than six centuries, during whichthe realm of scientifically admissible procedures or techniques has been


more and more restricted … We mean this critically restricted realm onlywhen we speak of ‘modern’ … science.

(ibid.: 8)

To Schumpeter, economic discourse attains the status of a modern science byfocusing on the understanding and formalisation of what he regards as the over-arching problem of economic analysis – namely, the general interdependence ofeconomic phenomena (ibid.: 242) In Business Cycles (BC), he had already main-tained that the first and foremost task of economic analysis consists inestablishing whether the economic system is uniquely determined by the knownrelations between its elements:

For our system is logically self-contained only if this is the case … Theproof that this is so is the magna charta of economic theory as anautonomous science, assuring us that its subject matter is a cosmos and nota chaos.


According to Schumpeter, this ‘magna charta’ has, for the first time, beenprovided by Walras’s concept of general equilibrium, a static and theoreticallyrigorous framework built on a deductive approach that guarantees its logicalcoherence.1 As such, it provides a scientific canon in the form of ‘PureEconomics’ (1908: 38). Schumpeter’s reconstructive perspective on the scien-tific progress of economics is, thus, based on elevating the principles of ‘PureEconomics’ to the status of an epistemological ‘standard’ of evaluation, whichprovides him with a basis on which to judge the internal coherence of thetheory under scrutiny. Clearly, for Schumpeter, the criterion of internal consis-tency is a necessary but not a sufficient criterion for the assessment of economictheory, especially in the area of economic dynamics.

Obviously, the adoption of such a standard was not neutral for Schumpeter’scommentators. In their view, reduced to its logical foundations, static equilib-rium theory is limited to a theory of catallactics, built on a specific understandingof economic phenomena (i.e. in terms of the logic of the maximisation of indi-vidual preferences and the optimisation of resource allocation). Schumpeterseems to be well aware of these implications when he writes that:

[T]he first question we need to ask in order to gain a clear understanding ofa theory and to be in a position to subject it to analysis, is whether or not itcan be defined as ‘static’, and our appreciation of it will depend on theanswer to this question. A theory is ‘static’ whenever the apparatus ofhypotheses characteristic of static analysis can be applied to it; … such atheory must be interpreted exclusively in terms of the foundations of oursystem, and it is in this sense only that it gains coherence and relevance.

(Schumpeter 1908 [1982]: 154)


This highlights the implications of adopting a reconstructive approach to thestudy of economic thought. By imposing a specific filter on his reading, thehistorian runs the risk of twisting his material to suit his preconceptions and ofpresenting a biased interpretation. However, the real danger is not so much thathe should search past works for theoretical elements that fit his standard model,but that he will misinterpret these works as a representation of this model andjudge their internal relevance on this basis.

The impact of an analytical standard of evaluation

The choice of a reconstructive approach is, thus, neither innocuous nor is itimpartial since it is implicitly driven by Schumpeter’s own expectations aboutthe history of economic thought and, more specifically, by his own desire tocontribute to the progress of economics and to refine analytical concepts:

Now, our minds are apt to derive new inspiration from the study of thehistory of science. … A man’s mind must be indeed sluggish if, standingback from the work of his time and … beholding the wide mountain rangesof past thought, he does not experience a widening of his own horizon …We learn about both the futility and the fertility of controversies; aboutdetours, wasted efforts, and blind alleys; … about spells of arrested growth,about our dependence on chance, how not to do things, about leeways tomake up for.


In other words, Schumpeter’s approach testifies to the close interdependencebetween his historical investigation and his own theoretical ambition, betweenthe economist–historian and the economist–theoretician. Evidence of this canbe found in the HEA where he points out that static equilibrium theory is nomore than the first stage of a much larger research project:

Now, an observer fresh from Mars might excusably think that the humanmind … would start analysis with the relatively concrete and then, as moresubtle relations reveal themselves, proceed to the relatively abstract, that isto say, to start from dynamic relations and then … to … the static ones.But this has not been so in any field of scientific endeavor whatsoever: alwaysstatic theory has historically preceded dynamic theory … static theory ismuch simpler to work out; its propositions are easier to prove, and it seemscloser to (logical) essentials. The history of economic analysis is no excep-tion.


For Schumpeter, this first step is by no means unimportant: by setting out firstto explain basic interdependencies in the abstract and rigorous terms of staticanalysis, economists not only define the boundaries of static analysis, but they


also specify those ‘shadow areas’ for which ‘it is of vital importance that they beliberated of the strait jacket that is, in this case, the static apparatus’(Schumpeter 1908 [1982]: 155). In this way, Schumpeter underlines the needfor a more general theory, capable of dealing with the totality of economicphenomena (Schumpeter 1954: 963): the theory of economic dynamics.

Schumpeter is, of course, very familiar with dynamic analysis, which heregards as a ‘land of the future’ in which purely economic phenomena continueto play their role (Schumpeter 1908 [1982]: 154, 490). Nonetheless, he has avery clear view of its nature. As early as 1908, he distinguishes between‘Theoretical Economics’ and ‘Economic Theory’, stating that ‘[t]he latter coversa much larger field than the former’ (ibid.: 41) and adding that there is littledifference between ‘static’ and ‘pure theory’ so long as ‘economics does notventure beyond static analysis’ (ibid.: 490). There is no shortage of proof inSchumpeter’s writings that he perceives of dynamic analysis, such as it willresult from scientific progress, as including the logic of pure economics. In theHEA, borrowing from Frisch, he specifies that dynamic theory differs stronglyfrom ‘historical analysis’ in that the former refers to ‘theoretical and not histor-ical datings’ (Schumpeter 1954: 965, fn. 5). Dynamic theory, in this sense,combines economics and history, which means both that history cannot bereduced to the study of mere details and that there is room for economic analy-sis per se.

As Elizabeth Boody Schumpeter notes in her introduction to the HEA, sinceearly in his career, as a theoretician, Schumpeter ‘envisaged a theory whichmight some day synthesize dynamic economics in the same way that theWalrasian system summed up static economics’ (1954 [1994]: xli). In BC, awork entirely dedicated to the study of dynamic phenomena, Schumpetercombines statistics with history and economic analysis. Adopting this methoddoes not however mean that there is no longer the need to refer to the standardof evaluation specific to economic analysis. This is how we might interpretSchumpeter’s observation that ‘equilibrium is indispensable as a standard ofreference’ because it presents ‘the bare bones of the economic logic’(Schumpeter 1939: 68–70).

Thus, much of Schumpeter’s approach is guided by his own research agendaand the link this constitutes between the perspective of the economist–histo-rian and that of the economist–theoretician. At one level, his investigation isorganised around one particular point of reference, namely Walras’s system. Hefocuses on those areas of economic research that provide insights into thenotion of general equilibrium – which he regards as constitutive of economics– and into its static analysis. However, at another level, Schumpeter alsosearches for historical confirmation of the necessity to provide an altogetherdifferent explanation for dynamic phenomena. However, as already stressed,this explanation must be based upon economic analysis and the reference to itsstandard of scientific evaluation.

This twofold enquiry corresponds to Schumpeter’s twofold preoccupationwith static and dynamic analysis and is further proof of the overlap between his


historical investigation and the analytical claims of the theoretician. Bothresearch interests are, thus, subjected to the same standard of evaluation. Thehistorian of thought reads the works of the past through a filter based on thisstandard of evaluation. In imposing it on the course of intellectual history, heachieves a rational reconstruction of this history – a ‘filiation of ideas’ – alongthe lines he has defined at the outset. The theoretician, preoccupied with thesearch for the building blocks of his theory, is constrained by the historian’sperspective and discerns only what is left visible after a rational reconstructionhas already occurred.

Schumpeter’s interpretation of Marshall

Schumpeter is full of praise for Marshall’s pioneering achievements – compa-rable to those of Walras – in the field of pure theory: Marshall deserves creditfor his comprehensive understanding of the problem of interdependencies inthe economy, as well as for introducing the concept of partial equilibrium intostatic analysis (Schumpeter 1951: 99). Schumpeter also highlights the impor-tance of Marshall’s historical analyses, which he finds particularly helpful forgaining a better understanding of the dynamic workings of a capitalist system.

However, this judgement is not so much a final appraisal but a preliminaryhypothesis that Schumpeter sets out to defend. In order to rehabilitate what heconsiders to be the essence of its originality – Marshall’s static equilibriumtheory – he often distorts Marshall’s original thought.

As with any such reconstructive effort, the first step is to separate out basicarguments – that is, those that are constitutive to the overall line of reasoningof the author under scrutiny – from the more trivial and less significant ones.Doing this leads Schumpeter to ‘sanitise’ Marshallian thought and to readMarshall’s work through the filter implied by his standard of evaluation. Hence,Schumpeter’s reconstructive reading begins by ‘moulding’ Marshall’s work ‘toshape’. This involves a certain amount of distortion (i.e. interpreting certainpassages in a misleading way), omission (i.e. ignoring awkward passages that donot fit the hypothesis) and addition (i.e. filling in elements which are ‘missing’to complete the argument).

Moulding Marshall to shape

Schumpeter clearly embarks on a process of signposting specific aspects ofMarshall’s work when he effectively purges it of its ‘unscientific’ elementswhich reflect those of Marshall’s views that are more evolutionary. Accordingto Schumpeter, what renders the task of the historian of thought difficult in thefirst place are certain traits of Marshall’s personality since these find their wayinto his work in the form of ambiguities that tend to blur what Schumpeterconsiders as Marshall’s most important contributions to marginalism.

For instance, referring to Marshall’s leading role in the English-speakingscientific community, Schumpeter invokes the image of an over-powering


‘father-figure’ of the profession whose works overshadowed many other equallyrigorous contributions to economic theory, most notably among these the worksof Jevons, Wickstead and Edgeworth (Schumpeter 1954: 826, 830). However,according to Schumpeter, the most damaging consequence of Marshall’s pre-eminence was that it led many of his students to adopt an often unjustifiablycritical stance towards economists of whom Marshall had not thought veryhighly.2 To Schumpeter, this is the main reason why Walras never succeeded inmaking himself known in England during Marshall’s lifetime (ibid.: 829). Heultimately attributes this development to Marshall’s desire to promote a certain‘scientific nationalism’, i.e. to his tendency to overrate the British tradition ofclassical political economy.

Moreover, Schumpeter disapproves of what he regards as the populistelement in Marshall’s work. He argues that Marshall’s desire to make his writ-ings accessible to the non-economist finds its way into his purely theoreticaldiscourse and accounts for its conciliatory tone (e.g. Schumpeter 1951: 94,102–3). In a similar vein, Schumpeter is critical of the ‘scientific utilitarianism’to which Marshall subscribes. Schumpeter argues that science cannot possiblyprogress if it is ‘incessantly harassed by imperious demands for the immediatelyuseful result’. But Marshall

felt no repugnance to the credo that prompts those demands. … L’art pourl’art had no place in his eminently Anglo-Saxon soul. To serve his nationand his time, and to teach what would be immediately helpful, that waswhat he himself wished to do more than anything else.

(ibid.: 102–3)

For Schumpeter, this Marshallian philosophy encourages the trivialisation of hiswork, that is, it leads the reader to underestimate the significance of the scien-tific elements of his discourse. Therefore, the historian of thought is all themore justified in focusing on Marshall, the theoretician of equilibrium.

The second element of Schumpeter’s attempt to mould Marshallian thoughtto shape concerns his selective use of Marshall’s writings when arguing his case.Schumpeter knew Marshall’s work inside out, and frequently emphasised that‘nobody knows Marshall who knows only the Principles’ (Schumpeter 1954:834). However, it is precisely the Principles that are his main, and almost exclu-sive source for assessing Marshall’s theoretical achievements. He thereforeseems to imply that the Principles are of paramount importance, even though heacknowledges that for Marshall they represent no more than a preliminarystage.

On the one hand, the Principles – as Schumpeter interprets them – epitomisethe third of Schumpeter’s ‘Classical Situations’ in the HEA, i.e. a theoreticalsituation characterised primarily by the consolidation of scientific progress,based on the mobilisation, co-ordination and synthesis of previous advances inpure theory. According to Schumpeter, this explains why the Principles haveattained canonical status, and are, consequently, an indispensable reference for


the historian of thought. In this sense, the work is of capital importance and, assuch, necessarily in a class of its own that sets it apart from Marshall’s otherwritings: because their contents are of historical or statistical nature, the latterare relegated to the status of ‘illustrative appendices’.

On the other hand, in view of both the normative and the ‘Classical’ natureof the Principles, it is almost self-evident that, to Schumpeter, they cannot butstand for Marshall’s theoretical as well as his innovative achievements, as theycontain the substance of Marshall’s contribution to modern pure theory. This isprecisely what interests Schumpeter as a historian of analysis. In his view, thePrinciples are the cornerstone of Marshall’s theoretical work, and its noveltyboils down to its analytical apparatus. This then is also the main reason why, toSchumpeter, understanding the Principles is crucial to the assessment ofMarshall’s work as a whole.

However, the reconstructive approach that leads Schumpeter to focus on thePrinciples in the first place also influences his reading of the book. He justifieshis procedure of signposting suitable elements of the work by pointing toMarshall’s tendency to intersperse his pure theory with digressions of a lessstringent nature: ‘The analytic skeleton does not grin at you. It is clothed inflesh and skin’ (Schumpeter 1951: 102). To Schumpeter, most of these digres-sions are illustrative and reflect Marshall’s pluralist concerns, but are of littleconsequence for his theory. Others, however, result in a methodological biaswith the potential of undermining his analysis. They arise from Marshall’sexcessive realism: ‘[H]is mastery of historical fact and … analytic habit of minddid not dwell in separate compartments but formed so close a union that thelive fact intrudes into the theorem and the theorem into purely historical obser-vations’ (ibid.: 94).

Hence, Schumpeter highlights the analytical scheme of the Principles. Tohim, Book V, generally recognised as a ‘classic masterpiece’, constitutes thetheoretical core of the work. Together with notes xiv and xxi of the Appendix,it represents the ‘analytical core or kernel of the Principles’ which ‘consists of acourse in a theory of economic statics’, not unlike that presented by Walras(ibid. 1951: 95, also 1954: 835–6).

Having thus placed emphasis on Marshall’s pure theory, Schumpeter hascleared the way to make his point. There remains, however, one furtherobstacle in the form of Marshall’s own choice of words which tends toencourage a simplistic view of his theoretical originality and of the man himselfas ‘an eclectic, who tried to reconcile and to combine . . . the analytic principlesof the English “classical school” (meaning Ricardianism) and the analyticalprinciples of the “marginal utility school” ’ (Schumpeter 1954: 837).Schumpeter resolves this issue by arguing that ‘in order to appraise a man prop-erly and to put him into the right place, it is sometimes necessary to defend himnot only against his enemies but also against his friends and even againsthimself’ (ibid.: 491).

Schumpeter’s defence of Marshall is achieved by disowning certain aspects ofMarshall’s views insofar as they contradict what Schumpeter perceives as the


norms of scientific historiography. In his view, Marshall’s understanding of hisown intellectual roots is confused and contradictory. Schumpeter’s reconstruc-tive effort is thus organised around two main goals. First, he is intent on cuttingthe influence of classical political economy in the formation of Marshallianthought down to size. Second, he sets out to rehabilitate the role of what heregards as the ‘true’ precursors of the innovative aspects of Marshall’s contribu-tion.

The classical roots of Marshallian thought

Schumpeter has no difficulty in accepting that Marshall was obviously influ-enced by the English classics, since he himself regards their works as milestonesin his history and since, furthermore, they clearly informed Marshall’s earlyviews on economics. However, Schumpeter also holds that Marshall tends tooverestimate his intellectual debt to this school of thought because he commitsthe mistake of interpreting scientific progress in terms of ‘kinship’, i.e. toattribute too much importance ‘to the fact that today’s work necessarily growsout of yesterday’s’ (ibid.: 835). In reality, Schumpeter finds in Marshall ‘merelythe form, not the essence, of the classical approach. … Moreover, a closecontact with the classical economists was maintained by him merely because hecompletely re-interpreted their doctrines’ (1914 [1954]: 185–6).

Schumpeter concedes Mill’s indirect influence on Marshall, given that Mill’swork epitomises the second ‘Classical Situation’ of the HEA. However,Marshall focuses on Mill’s intuitive understanding of equilibrium theory and,even more revealingly (and in line with the overall argument of the HEA), onthe gaps and shortcomings in Mill’s argument (Schumpeter 1954: 838).

Schumpeter’s view of Marshall’s debt to Smith – whose work he classifies asbelonging to the first ‘Classical Situation’ of the HEA – is very similar. Eventhough, according to Schumpeter, ‘ “It is all in A. Smith” was a favourite sayingof Marshall’s’ (ibid.: 309), he does not regard this as proof of the relevance ofSmith’s work to Marshall’s static theory. This is not surprising once we take intoaccount Schumpeter’s conceptualisation of scientific progress in economictheory as a cumulative process. From this point of view, a more up-dated version(that is, more methodologically and scientifically advanced) of Smith’s analysisis contained in Mill. However, Schumpeter is quite aware that Marshall takesfrom Smith other elements of his approach, such as his taste for empirical anal-ysis, his tendency to intermingle theory with facts and even a somewhat‘pre-scientific vision’ of economic dynamics. Thus, he tends to belittleMarshall’s appreciation of Smith, attributing it to his insularity (ibid.: 307,fn. 15).

All in all then, Schumpeter grants Mill and Smith a limited role asprecursors of Marshallian thought. His main point is that Marshall fails to graspthe true nature of the scientific or analytical progress of economic theory. Morespecifically, he argues that Marshall fails to understand that what he regards as agradual refinement of techniques first developed by Mill in fact represents an


innovative and revolutionary breakthrough (e.g. ibid.: 838) for which theground has been prepared less by Mill than by the technical achievements ofthe marginal revolution.

As concerns Ricardo, Schumpeter is quick to dismiss his legacy to Marshall– in the same terms as his overall contribution to pure theory – as a detour(ibid.: 473, 568). While a case can be made for a link between Smith, Milland Marshall, ‘there is no really practicable bridge between Ricardo andMarshall’: ‘in spite of his statement to the contrary, Marshall is connectedwith him only by a very loose tie (almost that of filial piety)’ (ibid.: 838, fn. 11;1914 [1954]: 71).

Throughout the HEA, but in particular in the subsection dealing with‘Marshall’s attitude and real cost’, Schumpeter leaves little doubt that, in hisview, the abyss that separates Marshall from the English classics and, above all,from Ricardo cannot be concealed. This is most evident when he explains hisdisagreement with Marshall’s pro-Ricardian views on value theory, arguing that‘though Marshall did not admit this … he never espoused the specificallyRicardian elements in the “classic” structure, such as, e.g. the labor-quantitytheory of value’ (Schumpeter 1954: 921, fn. 4). Hence, according toSchumpeter, Marshall’s theory of real cost – despite having been ‘the olivebranch presented to his classical predecessors’ (ibid.: 1057) – cannot possibly beinterpreted as Ricardian by the historian of thought.

Schumpeter also comments on Marshall’s criticism of both Jevons’s and theAustrians’ failure to understand Ricardo. It is in this context that Marshallemploys the famous analogy of the ‘two blades of a pair of scissors’. Schumpeter,however, questions the true nature of this disagreement, arguing that Marshallreally only rephrased Jevons’s and the Austrians’ clumsily put views in moreeloquent terms:

[I]t is meaningless to accuse either Jevons or the Austrians of wishing tominimize the importance of the very theorem which they were the first todeduce rationally and which Wieser called the ‘law of cost’. They stood inno need of being told about the two blades of Marshall’s pair of scissors. …Half the generosity lavished upon Ricardo might have revealed the greatachievement behind the poor technique and reduced criticism to the onepoint that could have been justifiably made though Marshall never madeit: that Jevons knew not enough mathematics and the Austrians none atall.

(ibid.: 922–3)Before turning to Schumpeter’s views on the role played by the forerunners

of marginalist analysis in the formation of Marshallian thought, one last pointremains to be made. Ironically, while Schumpeter disagrees with Marshall’sappraisal of both Ricardo’s and Mill’s contributions and disapproves ofMarshall’s assessment of the way Jevons and the Austrians interpret Englishclassical theory, his own criticism opens with the following words: ‘I submit thatthe impression the reader of Marshall’s Principles is bound to receive, in spite of


qualifications that Marshall inserted, is …’ (ibid.: 920, emphasis added). Thisdemonstrates, once again, that underlying Schumpeter’s reconstructiveapproach is a point of view that is directly related to his conception of thedirection of progress in economic thought.. However, as far as Schumpeter isconcerned, far from invalidating his point, this admission only reinforces itsince it is understood as a justification for reading Marshall’s work through thefilter of a scientific standard, thus rehabilitating the non-eclectic aspects of histheory.

Marshall’s marginalist precursors

Schumpeter is less concerned with Marshall’s ‘subjective originality’, i.e. hisdisregard for economists to whom, in retrospect, he seems to be indebted, thanwith Marshall’s unwillingness to recognise the objective anteriority ofmarginalist thought to his own approach (Schumpeter 1951: 95). This ignoranceof the marginalists by Marshall is of uttermost importance to him since he fearsthat it is in part to blame for an erroneous view of the history of economicthought: ‘Time and again, I have been impressed by the fact that competent andeven eminent economists have an uncritical habit of attributing to Marshallwhat should, in the “objective” sense, be attributed to others’ (Schumpeter 1954:839 fn. 13). This calls for the corrective intervention of rational reconstruction:

According to what I believe to be the ordinary standards of scientific histo-riography, … the rediscovery of marginal utility is Jevons’; the system ofgeneral equilibrium … is Walras’; the principle of substitution and themarginal productivity theory are Thünen’s; the demand and supply curvesand the static theory of monopoly are Cournot’s … the consumers’ rent isDupuit’s; the ‘diagrammatic method’ of presentation is also Dupuit’s or elseJenkin’s. If this had been always clearly understood, there would be nomore to be said. But it has not been generally understood …. [T]he reasonfor this state of opinion is largely Marshall’s own fault.

(ibid.: 838–9)

It is with this goal of ex-post rectification in mind, that Schumpeter invokes aline of argument that, once again, brings him into conflict with Marshall’s posi-tion. He basically reinterprets the analytical scheme underlying Marshallianthought such that it allows him to emphasise a close connection between thePrinciples, on the one hand, and the works of leading marginalists, on the other.

For reasons of expositional expedience, we have so far let Schumpeter’sassumption go unquestioned, according to which the ‘scientific age’ beginningwith Marshall’s ‘classical’ contribution is represented by an essentially unifiedbody of work, despite some apparent degree of diversity:

For within serious economic theory there are no such things as ‘schools’ ordifferences of principle, and the only fundamental cleavage in modern


economics is between good work and bad. The basic lines are the same inall lands and in all hands: there are differences in exposition, in the manner– and mannerism – of putting things, for example according to rigour andgenerality or vicinity to ‘real life’. Then there are differences in technique,… differences in individual pieces of the analytic machine … differences asto particular problems …. But this is all.

(Schumpeter 1928 [1989]: 49)

Schumpeter argues, more specifically, that in its modern form economicdiscourse converges on the theme of equilibrium and, thus, on an analyticalscheme that is invariant to different areas of application since it is of an essen-tially mathematical nature: this is what gives the theory its quintessentiallyscientific quality. According to Schumpeter, the ‘mathematisation’ of equilib-rium theory merely provides a more rigorous treatment of the relation betweenvariables that have already been defined on theoretical grounds. As a matter offact, pure theory is, by its very nature, based on a deductive approach thatensures its logical coherence. In this sense, modern economic theory is mathe-matical, first in essence and second in method:

Since this kind of service [the one ‘rendered to economic theory’ by ‘math-ematics’] consists simply in sharpening the edges of our analytic tools … aman’s mathematics does not necessarily show on the surface of an argu-ment: mathematical theory is more than a translation of non-mathematicaltheory into the language of symbols, but its results can, in general, be trans-lated into non-mathematical language.


Two things follow for Schumpeter: first, the fact that few equilibrium theorists ofthe time actually used mathematical language by no means proves that theirreasoning was therefore of a non-mathematical nature. Second, in the realm ofstatics, any fundamental opposition to marginalist principles is profoundly absurd(ibid.: 956, fn. 4). This also applies to Marshall, whose reluctance to produce apurely abstract piece of work leads him to erase all traces of mathematics from hiswritings (ibid.: 956) and to complain about the limitations of static analysis:

The point is – not merely that his mathematical turn of mind was favorableto his achievement in the fields of economic theory, but – that the actualuse of the methods of mathematical analysis produced that achievement …That performance of the Marshallian kind practically presupposes a mathe-matical schema. But this Marshall always refused to admit. … He nevergave full credit to the faithful ally. He hid the tool that had done the work.


Schumpeter is all the more critical of Marshall’s position as what makes histheory modern and innovative is precisely its quasi-mathematical nature:


‘Marshall had built an “engine of analysis … machinery of universal applicationin the discovery of a certain class of truths” ’ (Keynes 1933: 101).3 In this sense,his theory is essentially identical to that developed by other leading equilibriumtheorists (1928 [1989]: 49–54).

In the HEA, Schumpeter further substantiates his effort at restoring fore-runners of marginalist analysis to their rightful place in the formation ofMarshallian thought. He now not only regards the analytical structure of theirtheories as essentially identical, but treats Marshall’s and Walras’s respectiveworks, in particular, as a standard of evaluation:

Jevons, Menger, and Walras taught essentially the same doctrine. … Jevons’and Marshall’s analytic structures do not, in essence, differ more than thescaffolding differs from the completed and furnished house, and note XXI inthe Appendix to Marshall’s Principles is conclusive proof of the fundamentalsameness of his and Walras’ models. Wicksell’s engaging frankness revealsthe two pillars of this arch … to the most perfunctory glance: the one isWalrasian, the other Böhm-Bawerkian. J.B. Clark’s blueprint … embodiedsubstantially the same principles as Marshall’s Book vi; Pareto and Fisherdeveloped Walras. … [T]hese names cover what we may call the period’sprimary work in ‘general theory’; the teaching associated with them …shaped practically all of the secondary or derivative work of the period.


This direct association of Marshall with Walras is present in all of Schumpeter’swritings, as the following quotations indicate: ‘[A]ll that follows applied toMarshallian as well as to Walrasian economics … on Marshallian lines. But Icould equally well call them Walrasian lines … Marshall … will come in …both in this section and in the next … where we … move on to the higherlevel of Walras.’4

Schumpeter is, of course, aware that the connection established betweenMarshall and Walras requires careful handling since Marshall is, above all, themaster of partial equilibrium analysis, ‘condemned … by theorists of the sternertype, especially by Walras …’ (ibid.: 991). His line of defence consists in provingthat ‘it is obvious from his [Marshall’s] appendix (note XXI) that, had hewished to go further, he would have sought the necessary complements … inthe general microanalysis of the Walrasian type’ (ibid.: 996–7). Other than this,Schumpeter maintains that the widespread perception of Marshall’s and Walras’swritings as representing different, even antagonistic approaches – a perceptionshared not only by their readers but also by the authors themselves – is to beblamed on Marshall’s insularity.

Our analysis of Schumpeter’s reconstructive approach to Marshall would beincomplete without a brief discussion of the main arguments he employs toshow that Marshall’s pure theory can, indeed, be understood as a replication ofWalras’ general equilibrium theory. A few examples will suffice to illustrateSchumpeter’s reconstructive approach.


To begin with, Schumpeter argues that Marshall’s substitution principle ‘risesto the proud position of Thünen’s great law of substitution’ which pervades andcontrols the whole economic process and opens one of several possible roadstoward the recognition of the universal interdependence of economic quanti-ties’ (ibid.: 995). To Schumpeter, this principle is an important indication of thesimilarity between Marshall’s and Walras’s analytical approaches since thenotion of substitution not only presupposes the logic of allocation but is one ofthe analytical concepts on which the determinacy of general equilibrium isbased. However, this conclusion is clearly based on Schumpeter’s reconstructiveperspective: in order to render Marshall more transparent to the reader, hecomplements and amends his analysis:

This becomes obvious if we add to the technical or factor substitution intro-duced on p. 420 of the Principles, the still more fundamental productsubstitution … Though Marshall also recognized the latter he never fullyco-ordinated the two … In consequence, Marshall’s principle of substitu-tion never appeared in his work … in its true light.

(ibid.: 995–6 fn. 11)

Another example concerns Schumpeter’s praise for Marshall’s elaboration ofmarginal productivity and value theory and his specification of two funda-mental requirements for the existence of static equilibrium. Referring toEdgeworth and Berry, Schumpeter stresses that ‘their contributions help us torealize the breadth of the wave, at the crest of which stands Marshall’s work’(ibid.: 1032 fn. 16). We would gladly subscribe to this statement, hadSchumpeter not added that: ‘If … he [Marshall] would never admit the fullextent to which he actually did so [accept the whole of the marginal utilityapparatus], this is, I think, adequately explained by … [his] reluctance to throwin his lot with the non-English economists who did the same thing’ (ibid.: 1032fn. 17).

Finally, regarding Marshall’s doctrine of normal profit, Schumpeter disagreeswith its interpretation by Marshall’s followers. He argues that they arrive at theconclusion that Marshall’s view is incompatible with Walras’s notion of zeroprofit in equilibrium only by basing their argument on a simplified notion of theconcept of normal profit in Marshall. In Schumpeter’s view, the two conceptsare, on the contrary, effectively identical, once Marshall’s concept of normalprofit is stripped of its reference to ‘the balance sheets of business practice’ andreduced to a notion of ‘pure profit’ (ibid.: 1048): ‘If we are resolved to displaythe logical properties of perfect equilibrium in pure competition, Marshall’sprofits will in fact vanish as completely as will Walras’s’ (ibid.: 1050). However,the essentially subjective nature of Schumpeter’s interpretation of Marshallianthought is particularly apparent at this point, if we recall that, elsewhere in theHEA, he himself repeatedly emphasises that, true to his English predecessors,Marshall is unwilling to abstract from reality to the point of accepting Walras’sdefinition of perfect competition.


In sum then, Schumpeter sees himself constantly obliged to rectify anddisown aspects of Marshall’s analysis in order to justify his own rational recon-struction of his work. An obvious question is whether Schumpeter does not, infact, provide what could be called a ‘Schumpeterian–Walrasian’ interpretationof Marshall which, in the end, has little to do with Marshall’s actual work.

Marshall, a theoretician of dynamics?

As we have stressed throughout, Schumpeter is often confronted with what heregards as the ‘extra-scientific’ element in Marshall’s reasoning which rendersthe task of the historian of thought difficult and complex and which heattributes mainly to Marshall’s unreasonable attachment to the English tradi-tion and its ‘realist’ tendencies (ibid.: 974).

Schumpeter, thus, highlights one of the major weaknesses of Marshallianthought which, in his view, explains why Marshall never reached the levels ofperfection characteristic of Walras’s pure system, even though he was in posses-sion of the means to do so. This weakness also explains a lack of coherence inthose parts of his work where Marshall attempts to substantiate his theoreticalinsights with empirical findings and, thus, to breathe some empirical life intothe pure postulates of static analysis. In so doing, Marshall, according toSchumpeter, potentially undermines the internal logic of his equilibriumsystem, most evidently so in the context of his theory of production (ibid.:1043). Schumpeter argues that the theoretical framework Marshall uses isstrongly constrained by the logical requirements of equilibrium analysis, whichis why Marshall’s attempt to introduce an ‘industrial fact’ (increasing returns toscale) into a theoretical concept (a decreasing long-run industrial supply curve)is doomed to fail:

He insisted on including internal and external economies in his … ‘supply’schedules … in spite of the fact that he thereby destroyed their reversibilityand rendered them useless for the purposes of static theory: they reallyrepresent pieces of economic history in the form of diagrams.

(ibid.: 1046)

Schumpeter regards this as a pertinent example of a recurrent methodologicalbias in Marshall’s work, i.e. his tendency to treat what Schumpeter regardsas the qualitatively different phenomenon of economic development as anextension of static analysis by introducing historical time into the theorems ofstatic analysis.

This, then, sheds some more light on Schumpeter’s perception of Marshall’sdigressions from static analysis: to the historian of thought, they are no morethan the ‘casualties’ of rational reconstruction based on the standard of evalua-tion that is implied. To the theoretician, however, they constitute dead ends ofthe analysis which must not lead him astray. But despite this stern methodolog-ical judgement, Schumpeter is highly appreciative of what he regards as


Marshall’s exceptional grasp of the driving forces of capitalism and, in partic-ular, the workings of the world of business. To Schumpeter, Marshall’sreflections on this subject are those of a visionary (Schumpeter 1951: 104).

Thus, he stresses, for instance, the relevance of Marshall’s treatment ofdecreasing costs for the theory of imperfect competition, as it emerged in the1920s and 1930s. The same holds true for Marshall’s awareness of the hybridcase of what later came to be labelled ‘monopolistic competition’ in marketsdominated by product differentiation. Schumpeter, furthermore, draws atten-tion to Marshall’s suggestion that perfect competition fails to maximise outputunder certain circumstances, thus anticipating the core concern of welfareeconomics, and he regards his awareness of the role of time for economic anal-ysis as decisive for the emergence of short-term analysis. Finally, he argues thatMarshall’s preoccupation with statistics and business practices ultimatelyencouraged, if not prompted, the emergence of econometrics.

Not surprisingly, then, Schumpeter concludes his HEA by arguing thatfurther progress in economic theory will consist of improvements based onMarshall’s basic insights, summarised in his ‘Classical’ contribution and basedon the increased availability and use of mathematical tools. As Schumpeter seesit, such progress will lead the next generation of economists to modernise, criti-cally assess and specify the postulates of equilibrium theory and thus to extendthem to incorporate some of the more intuitive aspects of Marshall’s analysisinto a new analytical framework – that is, in particular, those concerningdynamic processes.

Schumpeter has a very clear understanding of the methodological nature ofMarshall’s excursions into dynamic analysis:

He [Marshall] … added plenty of extra-static considerations, chiefly aboutgrowth but also about sequences, so much in fact that he may be said tohave posited the task of future dynamic theory … but though he presentedmaterial, viewpoints, and desiderata, he did not cross the Rubicon … [bywhich] I mean this: however important those occasional excursions intosequence analysis may have been, they left the main body of economictheory on the ‘static’ bank of the river; the thing to do is not to supplementstatic theory by the booty brought back from these excursions but toreplace it by a system of general economic dynamics into which staticswould enter as a special case. The realization of the fact that even a statictheory cannot be fully developed without an explicit dynamical schema …is a first step in this direction.


In this view, Marshall takes the first step towards dynamic analysis: he under-stands the workings of capitalism (ibid.: 836), the malleable and changingnature of human behaviour (Schumpeter 1951: 93), and the realities of theindustrial production and organisation of his time. Furthermore, his awarenessof historical process introduces ‘an evolution-mindedness’ into his theoretical


work (ibid.; also 1954: 1165). Most importantly, Schumpeter adds that Marshall,by turning the stationary state into a methodological device associated withsmooth growth, sheds some light on the question of economic development ordynamics.

Even so, according to Schumpeter, Marshall’s understanding of dynamicanalysis remains limited since he misconceives it to be a theory of historicalchange (Schumpeter 1954: 966). Marshall’s ultimate inability to perceive thetrue nature of dynamic theory is directly related to his aversion towards pureanalytical schema since, had he admitted the importance ‘of logically rigorousdefinitions, the problems of dynamics would have emerged almost of them-selves’ (ibid.: 967). In other words, Marshall fails to understand that in order todeal with the dynamic aspects of economic development a totally differenttheory is required.

His theoretical apparatus is strictly static. This does not prevent him fromdealing with evolutionary phenomena of economic life that are refractoryto the application of the methods of statics … [I]n order to do so he had toget off the driver’s seat of his analytic engine, the arms of which do notreach these problems.

(ibid.: 836–7)

Once again, it is tempting to question Schumpeter’s reconstruction of Marshallon the basis of his own comments. The thrust of his argument is this: if onlyMarshall had understood the close connection between static equilibrium anal-ysis, as embraced by most marginalists, and his own work – if, so to speak, hehad been aware of Schumpeter’s main thesis – his extraordinary talents as aneconomist (his technical skills in combination with his historical knowledge)would have been enough to ensure a much faster progress of economic theorythan was actually achieved. However, as things stand, the historian of thoughthas to be content with recognising the visionary – rather than the revolutionary– in Marshall: ‘Much good could have been accomplished if Marshall had reso-lutely stood for the line of advance which he had done more than anyone elseto open’ (Schumpeter 1951: 98).

Conclusion

The constraints implicit in the reconstructive approach Schumpeter adopts inhis role of historian lead him to scrutinise the history of economic ideas forconcepts and formalisations of ‘pure’ dynamic analysis and, thus, to draw adistinction between what he regards as ‘economic theory’ and other social disci-plines. This perspective is what underlies his assessment of Marshall’s excursionsinto dynamic analysis as representing no more than historical or statisticaldigressions from pure static analysis. As we know, Schumpeter’s view ofdynamics is different. It consists in the combination (and not the confusion)between history, economic sociology and economic analysis.5


In many ways, one could also think of Marshall’s approach as ‘multidisci-plinary’, close to the approach advocated by Schumpeter in the HEA when hedefines the scope of ‘Economic Analysis’. However, when dealing withMarshall, Schumpeter can see no more in it than Marshall’s desire to upholdthe classical English tradition. Had he adopted a more ‘relativist’ approach tothe history of economic thought – to take up Blaug’s expression – he mighthave come to the same conclusion as Robbins that ‘the characteristic features ofthe Marshallian system … are all in a sphere which begins … where theWalrasian system tends to leave off (Robbins 1955 [1991]: 40). To put itanother way, Schumpeter, the economist–historian, closes the mind ofSchumpeter, the economist–theoretician, to ways of perceiving progress ineconomic theory other than those highlighted by his rational reconstruction ofeconomic thought.

In sum then, our discussion of Schumpeter’s approach to the history ofeconomic thought has led us to touch upon a debate that goes far beyond theconcerns we have raised in this chapter in that it opens the Pandora’s box of atheory of dynamic processes. More importantly, though, it also seems to us thatthis debate will go a long way towards rehabilitating the role of the historian ofthought in economic analysis.

Notes1 See, for example, Schumpeter 1908 [1982]: 424–7. Also see Arena’s contribution to

this volume.2 ‘Edgeworth, too, was sadly ungenerous to Walras as well as to the Austrians. But his

lack of generosity was somewhat like the loveable ungenerosity of devoted mothersand wives who cannot see any merit in competitors of their wholly admirable sons orhusbands’ (Schumpeter 1954: 839 fn. 14).

3 The quotation is from J.M. Keynes, Essays in Biography, 1933.4 Schumpeter 1954: 1020 fn. 60; 1928 [1989]: 49–54; 1954: 908.5 See the Introduction to this volume.

Bibliography

Blaug, M. (1962) Economic Theory in Retrospect, 7th edn, Cambridge: Cambridge Univer-sity Press, 1997.

Keynes, J.M. (1933) Essays in Biography, reprinted in The Collected Writings of JohnMaynard Keynes, Vol. X, London: Macmillan (1972).

Robbins, L. (1955) ‘Schumpeter’s history of economic analysis’, Quarterly Journal ofEconomics, 69:1–22. Reprinted in J. Cunningham Wood (ed.), J.A. Schumpeter: Crit-ical Assessments, vol. II, London: Routledge, 1991.

Schumpeter, J.A. (1908) Das Wesen und der Hauptinhalt der theoretischen Nation-alökonomie, Munich and Leipzig: Duncker & Humblot. Italian translation Rome:Laterza, 1982.

—— (1914) Epochen der Dogmen- und Methodengeschichte, Tubingen: J.E.B. Mohr. Trans-lated into English as Economic Doctrine and Method: An Historical Sketch, London:Allen & Unwin, 1954.


—— (1928) ‘The instability of capitalism’, Economic Journal 38, September II: 361–86.Reprinted in J.A. Schumpeter (1989).



—— (1954) The History of Economic Analysis, Allen & Unwin. Reprinted, London:Routledge (1994).



Part II

Methodology

Introduction

At the beginning of the twentieth century, while Schumpeter was nearing theend of his university studies in Vienna under such teachers as F. von Wieser andE. von Böhm-Bawerk, political economy was still reeling from recent powerfulblows to its reputation. Throughout the 1870s, members of the academiccommunity had raised doubts as to its relevance vis-à-vis other disciplines, inparticular history and the emergent sociology. Francis Galton, for instance,suggested at a meeting of the British Association for Science in 1877, that polit-ical economy be struck off the list of scientific subjects.1 About the same time,the Methodenstreit set the followers of the Historical School against thesupporters of a more rigorous approach to methodology, such as that embracedin the theoretical contributions to the emerging marginalist school. In thiscontext, the survival of political economy as an autonomous discipline wouldhinge on two developments of an essentially epistemological nature.

On the one hand, what was to be at stake was the development of a method-ology specific to the discipline that would be based on the rigour of thehypothetical deductive approach, largely inspired by its counterpart in the so-called natural sciences. Thus, Léon Walras, for instance, suggested that ‘it ispossible to apply algebra or geometry to political economy’, because ‘this mighttransform political economy in the same way as it has transformed mechanics orastronomy’ (Walras 1874 [1987]: 310–11).

On the other hand, the debate would evolve around redrawing the bound-aries of economic research so as to extend it to ‘new objects’, to borrow a fittingexpression by Jevons who maintained that economists must learn to takeaccount of the past out of which we are always in the process of emerging. Hewent on to argue that whether or not this would be called ‘sociology’, whatmattered was that economists needed to develop a scientific analysis of theprinciples of evolution such as they manifest themselves in every aspect ofsocial existence (Jevons 1876 [1965]: 195). This statement by one of thefounding fathers of marginalism is of interest to us in more than one respect.

First of all, it conveys a sense of what Jevons regarded as an inevitableencroachment of the history of institutions and of economic sociology upon thehabitual themes of classical economic theory. This sense of inevitability is all

4 On the boundaries betweeneconomic analysis and economicsociology

André Legris

the more significant as the exponents of the ‘old’ narrative history and theemerging sociology had by then adopted a competitive stance vis-à-vis politicaleconomy. To extend the boundaries of the latter might well prove a forcefulresponse to Galton’s disastrous suggestion.

Furthermore, the spirit of Jevons’s remarks suggests that, until then, ‘theprinciples of evolution’ had been totally ignored by economists, and that itseemed to him to be essential not to continue in this vein. On the one hand, itappeared prudent to put a stop to the then dominant and dangerous develop-ment of Spencerian sociology, which claimed to explain the whole of humanrelations, economic relations included, by applying the law of natural selectionto social relations. On the other hand, it seemed hardly promising to scrutinisethe heuristics of the emerging marginalist revolution – of which Jevons hadbecome one of the most active promoters – for a justification to extend theboundaries of the discipline’s subject matter. As a matter of fact, in replacingthe theories of wealth and/or of value – of such great importance to the classicaleconomists – with the theory of equilibrium, be it in the shape of partial or ofgeneral equilibria, marginalism explored static phenomena with great rigour butencountered great difficulties in dealing with the dynamics of the historicalevolution of real societies.

Apart from Jevons, a great many other marginalists were aware of this limita-tion. Among others, we may recall Enrico Barone’s view who states in theintroduction to his Principi di Economia Politica (1913) that, ‘if the concept ofequilibrium was not subsequently complemented by an analysis of all dynamicphenomena, it would give rise to conclusions which have little to do with realphenomena’ (Barone 1913 [1929]: 5).2

This historical context, some of whose basic aspects we have just explored,will serve us as a general backdrop against which to organise our discussion ofthe boundary between ‘pure’ economic theory and economic sociology – aboundary that Schumpeter strove to specify from the outset of his career – andof the place accordingly occupied in his work by the history of institutions. Aswill become clear, Schumpeter’s main concerns are best understood as aresponse to both the desire and search for methodological rigour and the needto amplify and refine the subject area of political economy.

We begin by examining Schumpeter’s support for the hypothetical deductivemethod. This will be followed by a discussion of the question of where to drawthe boundaries of economic inquiry.

From the hypothetical deductive approach to the boundariesof ‘pure’ economic theory

During the 1960s in particular, the scientific community insisted on the impor-tance of Schumpeter’s later contributions, published a few years before hisdeath, which were considered the culmination of his thinking. It has, forexample, been emphasised that the topic of Capitalism, Socialism and Democracyis not limited to a discussion of the theoretical links between firm concentra-

90 Methodology

tion and the cyclical nature of the innovative process. Rather, these links areseen to have been put into perspective by relating them to certain endogenoustendencies towards a breakdown of coordination mechanisms in decentralisedeconomies.

More recent contributions (Costabile 1986; Pesciarelli and Santarelli 1986;Arena 1992) have, on the contrary, shifted the focus of attention back to thepublications of the ‘younger’ Schumpeter. This is where we find most of hiscontributions to the definition of political economy as a scientific discipline.

Static analysis and ‘ pure’ economic theory

Starting in 1908 with the publication of Das Wesen und der Hauptinhalt der theo-retischen Nationalökonomie (DW), followed in 1912 by Die Theorie derwirtschaftlichen Entwicklung (The Theory of Economic Development), Schumpeterdevelops with great precision the principal concepts which, with only a fewalterations, were to remain at the heart of his analysis (Pesciarelli and Santarelli1986: 171).

Thus, DW constitutes a first attempt at establishing not only the basicfeatures but also the limitations of ‘pure’ economic theory. In Schumpeter’sopinion, the stylised facts describing specific features of a national economy areof little use to the theoretician who is mainly concerned with defining thesubject matter of his science and who, therefore, needs to adopt a purelyabstract approach. The procedures that are constitutive to static economictheory require the use of a hypothetical deductive methodology, tailored specifi-cally to the particularities of economics. Hypotheses put forth by an economistcan, thus, be detached from all historical and sociological context and can beregarded as mere postulates, upheld exclusively on grounds of their consistency.

According to Schumpeter, the best illustration of such a coherent system ofpropositions can be found in the principles of the Walrasian competitiveuniverse. This universe is, therefore, relevant not because it offers a more or lessplausible description of capitalism at the end of the nineteenth century, butbecause it provides a point of departure – valid only in the sphere of formallogic – for deductive propositions. Certain features of the deductive element ofthis approach are also of great value in clarifying the definition of pureeconomic theory.

The first argument is of an essentially instrumental nature in that it concernsthe use of mathematical tools which had recently become available and whoserelevance Schumpeter emphasises as early as 1906 in his first paper (Arena1992: 125). The use of mathematical tools allows the rigorous specification ofthe nature of as well as the links between stock quantities, on the one hand,and of the relation between values and quantities, on the other. Wheneverquantities are coordinated in such a way that the value of one corresponds toone and only one value of all the others, the system adjusts to a deterministicstate of general equilibrium. In this way, the natural level of quantities isdisclosed (Schumpeter 1908 [1982]: 37–41).

Economic analysis and economic sociology 91

The second argument is of a more analytical nature. Taking as a given anyone equilibrium configuration, the investigative purpose consists in describingthe most important quantitative changes in the absence of exogenous riskfactors. In this way it becomes possible to establish fundamental propositions, inthe form of economic laws, without reference to any methods or tools borrowedfrom other disciplines. Under these conditions, pure economic theory can claimthe status of a self-contained science (Schumpeter 1954: 15).

The last point of interest to us here concerns the methodological approachto social analysis, a methodology aimed at understanding how certain proto-typical forms of behaviour can be distilled from social reality. Schumpeter’sapproach to the reconstruction of the social has a distinct ‘individualist’ flavour(Gislain and Steiner 1995: 25–30). According to Donzelli (1983: 639),Schumpeter is the first author to have introduced the term methodological indi-vidualism into the language of social sciences. As opposed to political andsociological individualism (Schumpeter 1954: 888), methodological individu-alism, as Schumpeter understands it, is concerned with the analysis of theeffects of different individual actions, based on different forms of behaviourallogic, on the social mechanisms of a particular economic system. From here on,the reconstruction of the social can take recourse to a unique methodology,integrating several logical levels of the analysis.

At the first level, an explanation of different prototypes of individualbehaviour is provided. These have been shaped by specific historical and insti-tutional environments, and it is economic sociology that provides a description ofthese environments. At a second level, it is possible to narrow the analysisdown to the economic effects of one specific type of behaviour only, and toexamine the ways in which the workings of the economy adapt to this type ofactor. This level of the analysis concerns the prototype of the homo economicusoperating in a pure economy. It shows how the hedonistic actions and routinesof this prototype, whatever the concrete economic activity under scrutiny, set inmotion certain mechanisms which lead to a static equilibrium, without there-fore implying that such agents will always behave as explicit and consciousmaximisers:

In this system of values a person’s whole economy is expressed, all the rela-tions of his life, his outlook, his wants, all his economic combinations. Theindividual is never equally conscious of all parts of this value system; ratherat any moment the greater part of it lies beneath the threshold ofconsciousness. Also, when he makes his decisions concerning his economicconduct he does not pay attention to all the facts given expression to inthis value system, but only to certain indices ready at hand. He acts in theordinary daily round according to the general custom and experience.

(Schumpeter 1934 [1912]: 39)

Finally, at a third potential level of the analysis, methodological individualismpermits the incorporation of other actor types into the analysis whose behaviour

92 Methodology

now is specific to certain areas of the economy, such as work, new productionprocesses or the financing of economic activity.

At this point, one may wish to follow Arena (1992: 125) in raising the ques-tion whether a society reconstructed on the basis of the aggregation of specificindividual prototypes is really that different from a social whole structured instable social groups, such as entrepreneurs, bankers or wage-earners. It seemsthat Schumpeter does not believe that to belong to a social category determinesindividual behaviour, but that, on the contrary, history and custom haveproduced a certain range of regularities that individuals are free to embrace orto ignore.

From 1908 onwards, Schumpeter thus seems to hold the view that pureeconomic theory should form part of a much larger enterprise called politicaleconomy in which, however, only theoretical or ‘pure’ economics can claim tobe self-contained (1908 [1982]: 40–1). In rendering the outline of pureeconomics more precise, Schumpeter describes that part of political economywhich he considers to be at the heart of the discipline and whose essence isstatic analysis. Static analysis, therefore, provides the scientific ground ontowhich a more complete theoretical edifice can be built.

‘Pure’ economic theory and the circular flow

According to Schumpeter, pure economic theory at its best is illustrated by theWalrasian equilibrium model, such as it emerges from a reading of the succes-sive editions of the Eléments d’Economie Pure.

However, so far as pure theory is concerned, Walras is in my opinion thegreatest of all economists. His system of economic equilibrium, uniting, asit does, the quality of ‘revolutionary’ creativeness with the quality of classicsynthesis, is the only work by an economist that will stand comparison withthe achievement of theoretical physics.


Thus, Walrasian equilibrium theory constitutes the key point of reference (the‘magna charta’) of economic analysis, in both historical and conceptual terms(Costabile 1986: 156; Arena 1992: 124–5).

Even so it is important to realise that Schumpeter’s reading of the Walrasianmodel of the pure economy is rather specific inasmuch as Schumpeter equatesthis model with the circular flow. Ever since the first edition of The Theory ofEconomic Development (TED) his main concern is to develop an analysis of theevolution of capitalism. Within this overall perspective, he begins by describinga static economy which corresponds to a model of general equilibrium. Thenotion of a static economy, as employed by Schumpeter, differs from that of astationary state in that it is compatible with smooth growth. In this sense, staticanalysis is, then, synonymous with the economy of the circular flow. The latteroperates to produce consumption goods in the absence of all change and


technical progress. Agents take their decisions in an environment of totalcertainty where accidental events cannot occur. Given these assumptions, theeconomic system is limited to the subject area of pure economics, knowingneither crises nor the underemployment of factors of production, nor a disrup-tion of established patterns of interaction between the structural variables.Thus, the circular flow can, for example, follow the trajectory of a stationaryeconomy with zero growth of its basic aggregates just as well as it can follow thatof a smoothly growing economy with a constant rate of growth.

The implications of Schumpeter’s interpretation of the Walrasian generalequilibrium should not be underestimated.3 One may, in fact, be tempted toraise doubts about the explanatory power the author of the TED attributes topure economic theory, as well as about the nature of the links between thecircular flow and the workings of real capitalism. Thus, one reading of the TEDviews the model of dynamic evolution as a logical sequence to the analysis ofthe static framework. The static situation is here seen as a conceptual form ofthe real workings of capitalism. The static stages are a theoretical representationof concrete situations occurring in decentralised economies, stages that undergochanges as dynamic evolution unfolds. That Schumpeter presented these twoconcepts consecutively is attributed to the requirements of academic exposition,its principal goal being clarity. A critical review by C. Grilli, for example,published shortly after the first edition of the TED, subscribes to this view:

He [Schumpeter] never misses an opportunity to remind us of the extraor-dinary scientific value of the concept of economic equilibrium which is notjust a simple abstraction employed to facilitate theoretical reasoning butwhich corresponds perfectly to reality … But, so the author goes on toemphasize, it is necessary to resolve the problem of the dynamic move-ments of the economy which is no less important and real than that ofstatics to which it is closely linked. The equilibrium is, in effect, the pointof origin of the movement towards a new equilibrium.

(Grilli 1912: 556)

Another reading points to the fact that the second part of the TED is dedicatedto dynamic analysis as an alternative concept to its static counterpart.Adherents of this position consider static analysis to be nothing but themethodological underpinning of a purely abstract model, detached from allreality. They regard Schumpeter’s treatment of the subject as a prelude to a crit-icism of the whole of neo-classical analysis (see, for example, Graziani 1989),where the first part of the book is seen to have served the purpose of allowing itsauthor to outline the basic features of the evolution of capitalism.

These conflicting interpretations can be put into perspective if we take intoaccount the suspicion with which some of these authors view Schumpeter’stheory of interest. If the static and the dynamic dimension of the analysis arebut two aspects of the evolution of capitalism, and if the Walrasian equilibriumremains the main overall framework, the nature of the rate of interest is that of

94 Methodology

a real phenomenon in that it originates in the scarcity of accumulated savingsvis-à-vis firms’ investment demand in purely financial markets. Banks will exist,of course, but their exclusive role is that of passive intermediaries betweensavers and investors, and they are of no real relevance to the analytical frame-work. The proposition of a monetary market superimposing itself on the realeconomy will, thus, not change anything. When, in the TED, Schumpeterassumes the absence of the interest rate from pure economic theory, somecommentators have tried therefore to rectify what they take to be, at best, asimplification and, at worst, a mistake. Ricci (1927), for instance, concludesthat the rate of interest should logically be nullified, which appears to him tomake no sense at all. If, on the other hand, the interest rate, as perceived bySchumpeter, has a place only in an economy with credit, interest becomes apurely monetary phenomenon and corresponds to the price paid by firms tobanks which take the risk of financing an innovation. Such an interest rate hasan existence only outside the boundaries of a pure economy, once the systemhas entered the dynamic phase.

Be this as it may, Schumpeter’s view continues to inspire conflicting inter-pretations inasmuch as he appears to lend his support to one version or theother in different parts of his work. In DW Schumpeter writes, for instance, that

static and dynamic analysis belong to completely different domains, andnot only do they deal with completely different problems, but they alsoapply different methods to different materials. They are not so much twochapters of the same theoretical edifice as two completely different edifices.

(Schumpeter 1908 [1982]: 509)

Similarly, in Capitalism, Socialism and Democracy one can find unequivocalsupport for the idea of a solely methodological role for pure economics, forexample, when Schumpeter asserts that ‘[c]apitalism, then, is by nature a formor method of economic change and not only never is but never can bestationary’ (1942: 82).4

On the other hand, Schumpeter emphasises the limited nature of develop-mental phenomena with regard to space and time, in particular, in the secondchapter of the first edition of the TED: ‘Economic development is so far simplythe object of economic history, which in turn is merely a part of universalhistory, only separated from the rest for purposes of exposition’ (1934 [1912]:58).

If we go along with this statement, the circular flow would not just play apurely methodological and normative role in Schumpeter’s writings, but wouldconstitute an aspect of real social history. While it is true that the situation of astatic equilibrium appears less frequently in capitalist societies after the indus-trial revolution, this does not affect the reality of the static state. Pure economictheory provides a theoretical framework for the exploration of the mechanismsleading to equilibrium. However, it also mirrors concrete situations experiencedby economic systems when the effects of real shocks, responsible for dynamic


imbalances, have subsided. An important illustration of this view is provided bythe process of creative destruction through ‘new combinations’ (Schumpeter1939: 613). In the real world, the reality of the static state largely assumes theexistence of a ‘subterraneous force’. In this context then, Schumpeter’s refer-ence to the neo-classical paradigm cannot be seen as merely formal, reflectingonly his admiration for its coherence.

Whichever interpretation one prefers – both of which ultimately raise theproblem of the boundary between theory and reality – pure economic theory,understood as a conceptualisation of static analysis, enables us to explore theadjustment mechanisms determining the return to a position of equilibriumwhenever disturbances have occurred. This aspect of the analysis is concernedwith the formation of market prices, the components of the internal structure ofthe system within the framework of the circular flow, the main economic factsof the non-dynamic world, the stages of reorganisation the system undergoesfollowing disorder caused by the development process and the conduct of actors,motivated by hedonistic egoism and whose daily routines enable the system toreturn to an equilibrium path once the real shocks provoked by theentrepreneur-innovator at the dynamic level have been brought under control.At this point, equilibrium values and price structures will take over again. Inthis way then, the purpose and subject area of pure economic theory seem to bewell defined. The question that remains, however, is that of the workability ofsuch a construction.

What seems to be clear though is that for Schumpeter static analysis consti-tutes but a part of economic theory, and not necessarily the most important partin qualitative terms, even though a historical analysis shows that static ‘reality’occupies an important place in social organisation. It is at the boundary of pureeconomics that we enter the terrain of dynamic analysis.

From economic sociology to dynamic analysis

The problem of methodological rigour thus encounters a solution inSchumpeter’s interpretation of the circular flow – that is, the Walrasian generalequilibrium. At the same time, this also renders the boundaries of pureeconomic theory more precise.

This leaves us with the task of shedding some more light on the problemsSchumpeter encounters once he turns his attention to the notion of develop-ment, thus paying heed to Jevons’s suggestion some decades earlier. Schumpeternow intends to narrate the history of human societies from the point of view ofthe economist, and in order to explain the emergence of the organisation ofsocial life, he broadens his area of investigation to include what he callseconomic sociology.

While one can agree with Swedberg (1991: 94) that Schumpeter neverattached much importance to an academic definition of sociology, it appears, onthe contrary, that he took economic sociology to describe a well-establishedsubject area which he identified as the analysis of institutions.5 To the extent

96 Methodology

that institutions evolve throughout history to sustain mechanisms specific to aneconomic system, they encourage particular forms of behaviour and guide ittowards specific actions that are no longer simply a manifestation of timelessrationality. Schumpeter thus incorporates economic sociology into his analysisin order to extend the subject area of political economy and to enrich the anal-ysis of systems by asking those questions which hitherto had been largelyneglected by economists. In other words, by embedding economic phenomenain a sociological perspective, Schumpeter re-establishes a connection betweentheory and history.

In his introduction to the History of Economic Analysis (HEA), he defines‘Economic Analysis’ as ‘dealing with questions of how people behave at anytime and the economic effects they produce by so behaving’, while ‘EconomicSociology [deals] with the question of how they come to behave as they do’(1954: 21). We therefore agree with Arena (1992: 135) who regards ‘economicsociology as a valuable tool in lending assistance to economists when they try toselect what N. Kaldor called stylised facts’.

Schumpeter now confronts two new fields of investigation: the first concernsthe specific ways in which institutions – governments, contracts, legacies orbanks – affect economic life, the role they play and the conditions under whichthey emerge. The second relates to the theory of agency insofar as historicallyspecific institutions encourage, or not, particular forms of behaviour which, inturn, permit certain economic structures to operate. In this overall context,Schumpeter suggests the need for a specific research agenda to develop amethodology suited to the study of the first of these two areas of investigation.

The evolution of institutions

‘The crisis of the tax state’ (1918 [1991]), published shortly before the end ofWorld War I, if not the best known of Schumpeter’s contributions, is neverthe-less of great interest to anyone wishing to understand Schumpeter’s method ofanalysis. Two themes run through this contribution: Schumpeter’s interpreta-tion of the evolution of an economic system based on private property, and theway in which historically evolved institutions produce and shape mechanismsadapted to this evolution.

The starting point of ‘The crisis of the tax state’ (CTS) is the state of publicfinances in Austria in the aftermath of the World War I, plagued by high publicdebt which, in turn, is at the root of monetary and financial instability. Thissituation, which concerns not only Austria but all other parties to the war withthe exception, undoubtedly, of the United States, worried many observers, someof whom even felt it necessary to discuss the danger of a total collapse of thesystem of decentralised economies.

From this discussion of Austria’s domestic economy at a specific point intime, Schumpeter proceeds to develop a more general theory of the origin of themodern state, covering the period from the Middle Ages to contemporary fiscalstructures. His main contention is that these origins are to be located in the


fiscal needs of the state. A few more detailed remarks may be in place at thispoint.

First, the way in which modern institutions emerge has an evolutionaryflavour, because:

Most important of all is the insight which the events of fiscal historyprovide into the laws of social being and becoming and into the driving ofthe fate of nations, as well as into the manner in which concrete situations,and in particular organizational forms, grow and pass away.

(Schumpeter 1918 [1991]: 101)

However, the term ‘evolutionary’ here does not correspond to the Lamarckianor Darwinian definition (Arena and Gloria-Palermo, 2001). Rather, as Hodgson(1993: ch. 10) has pointed out, it refers more to ‘a disturbance of existing struc-tures and more [to] a series of explosions than gentle, though incessanttransformations’ (Schumpeter 1939: 102). Thus, Schumpeterian evolution is ahistorical rather than a biological concept.

Second, it would be impossible to understand the nature of a modern indus-trial system without taking into account the history of its fiscal needs. As amatter of fact,

fiscal measures have created and destroyed industries, industrial forms, andindustrial regions even where this was not their intent, and have in thismanner contributed directly to the construction (and distortion) of theedifice of the modern economy and through it of modern spirit.

(Schumpeter 1918 [1991]: 101)

It is clear, then, that it becomes inevitable to extend the field of economicinvestigation to include issues which, owing to the existence of sociology, havethus far remained outside the conventional subject area of economics.Sociology here becomes ‘economic sociology’ insofar as it refers to the scientificinvestigation of institutions which are of major importance to economicactivity and which have gradually evolved through a historical process.Following Schumpeter, to the extent that the working of these institutionsaffects the trajectories of real economies, recourse to economic sociologybecomes indispensable if economists do not wish to ignore evolution.

The CTS should not, therefore, be read merely as a contribution dealingwith the analysis of a historically and geographically specific problem. Rather, itshould be seen to develop a more general methodological approach aimed atbroadening the subject area of pure economic science to include economic soci-ology, so as to account for the historical evolution of social organisation – thatis, of exchange and coordination within an institutional setting shaping indi-vidual adaptive behaviour. In the case of the tax state, its emergence signals thegradual ascendance of the capitalist state and the equally gradual demise of thefeudal state. From this point of view, specific social activities typical of the

98 Methodology

feudal order (such as those contributing to the maintenance of feudal lordship)will disappear because the institutions (knighthood) on which they have beenbased change. The prototypical behaviour of the feudal landlord will come backto life in the attitude of the heads of enterprises as the new institutions redirecttheir activity towards a purely economic purpose.

Entrepreneurs and institutions

Ever since his earliest contributions, Schumpeter has maintained that a moderneconomy is regularly faced with dynamic imbalances. The TED opens with theassertion that the aspect of the capitalist system which has received the leastattention is not equilibrium but structural change caused by innovation.6 Inconventional approaches, innovation is conceptualised as an exogenous shock,largely located outside the scientific subject area of political economy.7

Schumpeter, on the contrary, asks his readers from 1912 onwards to accept thatinnovations constitute the central concept of dynamic analysis.

Schumpeter’s research interest at the time focuses on the role of innovationswhich he regards as having two main effects: they lead to a reorganisation of theproduction structure in the economies concerned, and they provoke a series ofdisequilibria whose nature Schumpeter studies more closely in Business Cycles(BC) (1939). The next step consists of explaining the conditions under whichinnovations take place by reference to the role of the ‘captains of industry’ –that is, of entrepreneurs, whose activities require, in turn, the existence ofcertain institutions that replace the conventional coordination mechanisms ofthe competitive Walrasian market.

In the early stages, innovations remain firmly embedded within the micro-economic organisation of leader firms, but, to the extent that they result fromthe application of major inventions whose selection depends on financial andprofitability criteria, they tend to diffuse rapidly through the market in the formof ‘new combinations’. Economic history appears to be governed by such wavesof innovation triggering – this time at the macroeconomic level – periodicalfluctuations. In BC, Schumpeter provides an explanation of Kondratieff cyclesalong these lines.

Within the boundaries of pure economic theory, Walrasian perfect competi-tion alone determines which mechanisms are best suited to maintain anequilibrium. In the event of external shocks, imbalances may, of course, occur.But there will always be forces ensuring the return to equilibrium, which may ormay not coincide with the pre-disturbance position. Furthermore, there is noreason at all why the occurrence of disequilibria should follow any recurrentpattern.

In Schumpeter’s approach, innovations are the only factor disrupting thecircular flow, and one comes to appreciate the considerable impact innovationshave on the industrial structure. Even so, the necessity remains ‘to encounter,within the economic system, a source of energy’ (ibid.: 447) capable of coun-tering the resistance of the forces, mentalities and activities at work in the


circular flow. In accordance with the principles of methodological individu-alism, Schumpeter takes a view of the agent as an endogenous factor in theemergence of innovations (1912 [1934]: 106). What is needed, therefore, is ananalysis of the entrepreneurs or, more specifically, of the nature of theentrepreneurs’ economic behaviour and of the social circumstances whichallows them, with the emergence of capitalism, to carry out their activities. Inother words, what is required is an investigation of the economic sociology ofinnovations. Economic sociology, thus, leads to an examination of the institu-tional environment that encourages the emergence of an actor-type differentfrom the Walrasian agent.8 In this sense, Schumpeter analyses theentrepreneur’s activity by contrasting it with that of the homo economicus.

As noted by Donzelli (1986), the fundamental merit of Schumpeter’sapproach lies in the definition of different and contrasting forms of behaviour.What Schumpeter labels ‘hedonistic egoism’ prevails in the circular flow,whereas ‘energetic egoism’ plays a central part in the realm of economic devel-opment (1908 [1982] and 1912 [1934]). While hedonistic egoism can bethought of as an equivalent to routine-based behaviour and adaptive rationality(Arena 1992: 133), energetic egoism applies to entrepreneurs only and is essen-tially an active form of behaviour. Entrepreneurs do not adapt to theirenvironment but, instead, shape it according to their own needs and purposes.Therefore, their rationality is not compatible with maximisation or optimisa-tion, and their desire, or their will, cannot be measured or calculated because‘the will to found personal kingdom’ or ‘the joy of creating’ (Schumpeter 1912[1934]: 93) cannot be maximised on the basis of a single Walrasian utility func-tion.

The entrepreneur exists then by virtue of a specific task, to wit that offorcing through new productive combinations. This sets him apart from routinebehaviour based on an individualistic, rational and hedonistic logic. The logicof economic sociology thus intervenes to allow us to gain a better understandingof the social mechanisms which, in the course of economic evolution, result indifferent forms of action.

Generally speaking, the prototype of energetic behaviour is not limitedeither to the economic sphere at the capitalist stage of evolution nor to the caseof the entrepreneur alone:

By far the greatest part of all types of activity, the battle of ideas and thehabitual modes of thought amongst artists, intellectuals or politiciansgenerally, can be traced back to some dominant personality where theepigones continue what these personalities have initiated.

(Schumpeter 1912)9

Therefore, the Schumpeterian entrepreneur is by no means a figure whoseappearance is limited to the modern economy. Rather, its basic features can betraced back to long bygone eras. However, in the more remote past suchentrepreneurial figures had to rely on support from institutionalised sources of

100 Methodology

power, such as the church, knighthood or the state apparatus. In addition,according to Schumpeter, the mechanisms underlying capitalism are no longercompatible with the traditional imperial relations between nations.10 Suchimperialism as can still be observed would be regarded as no more thanwitchcraft, inherited from ancient times or from the feudal age, an age inwhich, in stark contrast to capitalism, an attitude prevailed that knew nopurpose and that was characteristic of states which strove for territorial expan-sion by force, independently of any more specific objective (Schumpeter 1919[1991]). With the emergence of capitalism, entrepreneurial activity can nolonger rely on such institutional structures. Therefore, in order to reaffirm theirnatural tendency to bring new combinations into place, entrepreneurs will haveto find new institutions which are adapted to the changed circumstances.

To cut a long story short, economic sociology explores the ways in whichentrepreneurs adopt new forms of behaviour and new institutions. Bankers, andthrough them the logic of credit allocation, take on the role of ‘relays’ or insti-tutional intermediation between entrepreneurial aspirations and the financialmeans necessary to accomplish them. Credit only helps latent entrepreneurs tobecome active because it provides them with access to the means of production.It therefore becomes necessary to complete the overall picture by adding theparticular activity of bankers, as well as the institutions they control, in order toprovide an opening for the entrepreneurs’ latent managerial abilities byproviding funds for innovative projects. Once the boundary of static analysishas been crossed, banks are no longer limited to the passive role of intermedi-aries in the sphere of exchange, and the bankers’ task is no longer restricted tothat of a broker (Arena and Festré 1996: 167).

Schumpeter’s analysis of development – that is, of dynamics – is based ontwo propositions: first, the capitalist economy is characterised by the divisionbetween those who offer their labour force and those who, because of theiraccess to bank credit, take decisions about the process and the organisation ofproduction. Second, the capitalist economy evolves continuously. The forcesthat drive the system towards an equilibrium position and towards the annihila-tion of profits are perpetually subverted by the introduction of innovations. Inthe context of dynamic analysis located outside the limits of static analysis,Schumpeter regards banks as an essential regulatory element of the economy.One can reconstruct Schumpeter’s reasoning in the second part of the TED,present also in his 1917/1918 and 1970 publications.11 Thus, at the beginningof each production cycle the demand for finance will equal the firms’ wage bill.Banks provide the requested credit on the promise of repayment by creatingmoney. The rate of interest is then a monetary phenomenon because it is aproduct of the workings of the monetary market, not the real markets wheresavings and investments are exchanged:

In other words, firms sell to banks promises of payment against liquidmeans of payment in the form of bank deposits. As soon as wages are paid,


wage-earners become the creditors and firms the debtors of banks. Loansand deposits thus come into existence as the result of bank credit.

(Graziani 1989: 44)

In this view, the money initially created by the banks is destroyed. Money isthus no longer a mere good, as in the static framework, but it assumes the role ofunit of account and is transformed into a social institution. In this model,following the Wicksellian tradition (Arena and Festré 1996: 168), the moneysupply is endogenously determined and depends on the level of economicactivity which entrepreneurs, driven by the profit motive and by their desire forpower, anticipate. The role of credit is no longer that of simple aide toexchange. Rather, it becomes the only viable way of initiating the innovativeprocess. In order to save capitalism from facing definite financial constraints,the credit investment contract emerges as the main institution, sustainingevolution or progress. From what has been said so far, it follows that banks,when exchanging liquid means of payment in the form of loans against promisesof payment, create new credit money. According to Schumpeter in the TED,given full employment of all factors, bankers will limit loans to the amount theyjudge necessary for the realisation of a particular investment project. Two pointsfollow: first, some fraction of the factors employed in ‘old’ activities will bediverted to ‘new’ ones. Second, the bankers will submit entrepreneurial projectsto a process of selection, the criterion of selection being their future prof-itability. Ultimately, then, structural change and the dynamic path of economicdevelopment are heavily influenced by the banks’ choice of entrepreneurialprojects.

Concluding remarks

For Schumpeter, the creation of endogenous money becomes an indispensableelement of the innovative process. But the role of banks – and of the bankers’rationality – does not simply consist in providing the highest possible creditsupply. Banks replace the coordination mechanism of static competition as theinstitution which organises the selection of financially viable entrepreneurialprojects. If the entrepreneur essentially appears as a debtor in the process ofmoney creation, the banker assumes centre stage in the evolutionary process asthe agent who redirects the factors of production towards ‘new combinations’.In agreeing to provide funds for particular projects, it is the banker who decideson the range of innovation actually introduced into the sphere of production.Thus, the banks, whose role has so often been underestimated by scholars in theSchumpeterian tradition, exert a very decisive influence on economic life.

Schumpeter’s efforts to transcend the static framework and to explain evolu-tion necessarily leads to the analysis of institutions such as contracts, moneyand banks, as well as different forms of rationality. The contribution ofeconomic sociology consists in the analysis of precisely those institutions whichneo-classical analysis regards as data. In taking recourse to institutions and to

102 Methodology

the theory of agency, Schumpeter’s writings clearly bear witness to the difficultyof drawing a boundary between ‘pure’ economic theory/equilibrium, on the onehand, and dynamics/creative destruction, on the other.

Notes1 See Gislain and Steiner 1995: 5.2 For further details, see Legris 1995: 103–4.3 It should be pointed out that Schumpeter regarded only Walras’s Eléments

d’Economie Pure (1874) as being of scientific value. The Eléments d’EconomieAppliquée and the Etudes d’Economie Sociale were, in his opinion, only of minorinterest, a point on which he agreed with Pareto (Schumpeter 1954: 828).

4 This seems to suggest that capitalism represents the exception in the history of soci-eties, all other forms of social organisation, covering most of human history, beingsubsumed under the model of the circular flow. Capitalism seems to require an alto-gether different type of approach, whence the need to theorise on the developmentprocess. Yet, this reading contrasts with other passages in Schumpeter’s early writingswhere the equilibrium norm seems to explain considerable parts of the history ofcapitalism (Pesciarelli and Santarelli 1986: 172). Arguments along this line can alsobe found in Schumpeter’s more mature writings, even though these are much moreelaborate in comparison to those of the ‘young’ Schumpeter. (Guellec and Ralle1995: 93).

5 According to Swedberg, Schumpeter does not provide a consistent definition of soci-ology in his various publications. In Economic Doctrine and Method (1914), he takessociology to be ‘the theory of social institutions and of the principles of social organi-sation’, while in an article published the same year, he describes it as ‘the doctrineconcerning the interactions between individuals and groups of individuals withinthe social whole’ (1991: 92).

6 Three years earlier, this lack of attention to dynamic phenomena – in the sense ofstructural change – had been pointed out by Pantaleoni in the Giornale degliEconomisti (1909: 211–54) in an article entitled ‘Di alcuni fenomeni di dinamicaeconomica’, quoted by Schumpeter in the HEA (1954: 967 fn. 11).

7 Whence, for instance, the well-known treatment of technical progress by Solow(1956).

8 Gislain and Steiner 1995: 152.9 Quoted in Pesciarelli and Santarelli 1986: 233.

10 Schumpeter is here at loggerheads with Marxist analyses according to which imperi-alism is the most advanced stage of capitalism!

11 Das Wesen des Geldes (The Essence of Money), an incomplete work first published in1970. For further details, see the introduction to the Italian translation by G.Nardozzi, entitled ‘Dalla contabilità sociale alla essenza della moneta’ (Schumpeter1970 [1990]: IX–XXIII). Also see Dangel-Hagnauer’s contribution to this book.

Bibliography

Arena, R. (1992) ‘Schumpeter after Walras: “economie pure” or “stylised facts”?’ in T.Lowry (ed.) Perspectives on the History of Economic Thought, Vol. VIII, Aldershot:Edward Elgar.

Arena, R. and Festré, A. (1996) ‘Banks, credit and the financial system in Schumpeter:an interpretation’, in L.S. Moss (ed.) Joseph Schumpeter, Historian of Economics,Perspectives on the History of Economic Thought, London: Routledge.


Arena R. and Gloria-Palermo S. (2001) ‘Evolutionary themes in the Austrian tradition:Menger, Wieser and Schumpeter on institutions and rationality’, in P. Garrouste andS. Ioannides (eds) Evolution and Path Dependence in Economic Ideas, Aldershot:Edward Elgar.

Barone, E. (1908 [1929]) ‘Il ministro della produzione nello stato collectivista’, Giornaledegli Economisti, September–October: 267–414.

—— (1913) Principi di Economia Politica, Rome: Athenaeum. Reprinted in 1929.Costabile, L. (1986) ‘Metodo delle scienze e teoria economica in Schumpeter. Note su

L’Essenza e i Principi dell’Economia Teorica’, Studi Economici, 29: 147–68.Donizelli, F. (1983) ‘Schumpeter e la teoria economica neoclassica’, Ricerche Economiche,

vol. XXXVII, 4: 634–90.—— (1986) Il Concetto di Equilibrio nella Teoria Economica Neoclassica, Rome: La Nuova

Italia Scientifica.Gislain, J.-J. and Steiner, P. (1995) La Sociologie Economique (1890–1920), Paris: PUF

Sociologies.Graziani, A. (1989) ‘Schumpeter and Italian economic thought in the inter-war period’,

Studi Economici, 37/1: 41–83.Grilli, C. (1910) ‘Sunto di J.A. Schumpeter’, Rivista Internazionale de Scienze Sociale, vol.

52: 113–15.—— (1912) ‘Review of J.A. Schumpeter, Theorie der Wirtschaftlichen Entwicklung’,

Rivista Internazionale di Scienze Sociale, vol. 58: 554–6.Guellec, D. and Ralle, P. (1995) Les Nouvelles Théories de la Croissance, Paris: Repères, La

Découverte.Heertje, A. (1987) ‘Schumpeter, Joseph Alois (1883–1950)’, in J. Eatwell, M. Milgate

and P. Newman (eds), The New Palgrave: A Dictionary of Economics, Vol. 4, London:Macmillan.

Hodgson, G. (1993) Economics and Evolution: Bringing Life Back into Economics,Cambridge: Polity Press.

Jevons, W.S. (1876) ‘The future of political economy’, in H. Higgs (ed.) The Principle ofEconomics: A Fragment of a Treatise on the Industrial Mechanism of Society, New York:Kelley, 1965.

Legris, A. (1995) ‘La théorie économique italienne de l’entre-deux-guerres’, RevueEconomique, 46/1: 103–4.

MacDonald, R. (1963) ‘Schumpeter and Max Weber’, Quarterly Journal of Economics, 80:373–6.

Pareto,V. (1906) Manuale d’Economia Politica, Milan: Società Editrice Libraria. Trans-lated into French as Manuel d’Economie Politique. Reprinted in Geneva and Paris:Droz, 1981.

Pesciarelli, E. and Santarelli, E. (1986) ‘Teoria dello sviluppo economico (1912). Cap. II:il fenomeno fondamentale dello sviluppo economico’, Quaderni di Storia dell’EconomiaPolitica, IV, 1/2: 171–200.

Realfonzo, R. (1996) Moneta e Banca – La Teoria e il Dibattito (1900–1940), Naples:Edizioni Scientifiche Italiane.

Ricci, U. (1927) ‘Review of J.A. Schumpeter: Theorie der Wirtschaftlichen Entwick-lung’, Giornale degli Economisti, vol. 10: 651–658.

Romani, P.-M. (1988) ‘Le concept d’entrepreneur’, Traité d’Economie Industrielle, 81–94,2nd edn, Paris: Economica, 1991.

Schumpeter, J.A. (1906) ‘Über die mathematische Methode der theoretischenÖkonomie’, Zeitschrift für Volkswirtschaft, Sozialpolitik und Verwaltung, 15: 30–49.

104 Methodology

—— (1908) Das Wesen und der Hauptinhalt der theoretischen Nationalökonomie, Munichand Leipzig: Duncker & Humblot. Italian translation, Rome: Laterza, 1982.

—— (1912) Theorie der wirtschaftlichen Entwicklung, Leipzig: Duncker und Humblot.Preface dated Vienna, July 1911. English translation of 2nd edn as The Theory ofEconomic Development: An Inquiry into Profits, Capital, Credit, Interest, and the Busi-ness Cycle, Harvard University Press: Cambridge, Mass., 1934. Preface to the Japaneseedition (1937) translated in J.A. Schumpeter (1989).

—— (1914) Epochen der Dogmen- und Methodengeschichte, Tubingen: J.E.B. Mohr. Trans-lated into English as Economic Doctrine and Method: An Historical Sketch, London:Allen & Unwin, 1954.

—— (1917/1918) ‘Das Sozialproduct und die Rechenpfennige. Glossen und Beiträge zurGeldtheorie von heute’, Archiv für Sozialwissenschaft und Sozialpolitik, Vol. 44. Trans-lated into English by A.W. Marget as ‘Money and the Social Product’, InternationalEconomic Papers, 1956.

—— (1918) ‘Die Krise des Steuerstaates’, Zeitfragen aus dem Gebiet der Soziologie, 4.Translated into English as ‘The crisis of the tax state’, International Economic Papers,vol. 4, 1954. Reprinted in J.A. Schumpeter (1991).

—— (1919) ‘Zur Soziologie der Imperialismen, Archiv für Sozialwissenschaft undSozialpolitik, 46: 1–39; 275–310. Translated into English as ‘The sociology of imperi-alisms’ in J.A. Schumpeter (1951b). Reprinted in J.A. Schumpeter (1991).

—— (1927) ‘Die sozialen Klassen im ethnisch homogenen Milieu’, Archiv für Sozialwis-senschaft und Sozialpolitik, vol. 57: 1–67. Translated into English as ‘Social classes inan ethnically homogenous environment’ in J.A. Schumpeter (1951b). Reprinted inJ.A. Schumpeter (1991).


—— (1942) Capitalism, Socialism and Democracy, New York: Harper; 3rd edn, London:Allen & Unwin, 1950. Reprinted (with a new introduction by R. Swedberg),London: Routledge, 1992.

—— (1951a) Ten Great Economists: From Marx to Keynes, New York: Oxford UniversityPress.

—— (1951b) Imperialism and Social Classes (ed. P. Sweezy), New York: Augustus M.Kelley.


—— (1970) Das Wesen des Geldes (ed. F.K. Mann), Göttengen: Vandenhoeck &Ruprecht. Translated into Italian as L’Essenza della Moneta, by E. Dal Bosco, Turin:Cassa di Risparmio di Torino, 1990.


—— (1991) The Economics and Sociology of Capitalism (ed. R. Swedberg), Princeton:Princeton University Press.

Seidl, C. (1984) Lectures on Schumpeterian Economics, Berlin: Springer-Verlag.Solow, R. (1956) ‘A contribution to the theory of economic growth’, Quarterly Journal of

Economics, 70 (February): 65–94.Swedberg, R. (1991) Schumpeter. A Biography, Princeton: Princeton University Press.Walras, L. (1874) ‘Mélanges d’économie politique et sociale. Une branche nouvelle de la

mathématique’. Reprinted in Economica, VII, 310–11, 1987.


Introduction

With the exception of a few notes, Schumpeter never directly addressed thequestion of time in economic theory, but his principal concern – the study ofmovement and economic change – undoubtedly meant that the concept of timewas of pivotal importance to his thought. Moreover, students of the problem oftime in economics frequently refer to his work.1 Thus, according to Teboul(1992), the red thread running through Schumpeter’s work is his emphasis onthe general importance of time in economic analysis.

While Walrasian thought dominated economic theory, Schumpeter’s effortat a conceptual reconstruction of economics was aimed at developing a dynamicanalysis which went beyond a mere ‘dynamisation’ of the static framework. Theimportance he attributed to innovative activity as a driver of economic changeand, hence, to the role of disruptions and the non-determinism of movement inthe economy, is usually regarded as his main contribution. As a consequence,most economists have focused exclusively on the creative properties of time ineconomic evolution as the key to Schumpeter’s dynamic approach, thusneglecting the time dimension present also in his analysis of the circular flow.2

Time is, indeed, present at different levels of analysis in Schumpeter’s theory.However, as Teboul (1992) has noticed, the characteristics of time are notalways the same throughout his writings. Rather, for each conceptual stage ofhis analysis, Schumpeter emphasises the temporality of specific events thataccount for the economy’s movement. The view defended here is that, althoughthe circular flow cannot generate evolution – i.e. structural transformations –strictly on its own, it nevertheless provides the key to a better understanding ofthe nature of economic change that Schumpeter wished to explain, in partic-ular with regard to the role of routines and behavioural regularities, over andabove innovative activity. Thus, the purpose of this chapter is to clarify thedifferent functions of time in Schumpeter’s theory as well as the nature ofSchumpeterian ‘dynamics’.3

In the first part of the chapter, we examine the implicit treatment of time inSchumpeter’s basic theoretical frameworks: the circular flow and developmentThe difference between these two models is not based on the absence or

5 Time and rationality inSchumpeter’s construct

Anne Châteauneuf-Malclès

presence of time, as Perroux (1935) has observed. Rather, it reflects the opposi-tion between two antagonistic models of time, namely ‘duration-time’ and‘rupture-time’ which we will refer to as ‘retrospective time’ and ‘prospective time’,respectively, to highlight the fact that the first model is past-oriented while thesecond is future-oriented. Both models define a specific function of time ineconomic activity and a particular way of taking account of the element of timein economic decision-making. Furthermore, they imply a substantially differentunderstanding of the institutions of capitalism within which individual actionstake place. Thus, this interpretation lends support to the idea of the presence ofa dual view of capitalism in Schumpeter’s work (Heilbroner 1988).

In the second part of the chapter, we suggest that these two functions oftime, which, to begin with and following Schumpeter’s own line of reasoning,we have considered separately and from a purely abstract perspective in order tohighlight the driving forces of economic movement as well as the processes ofdecision-making associated with it, come to full fruition in Schumpeter’s theoryof the business cycle. In our view, it is here that Schumpeter combines the twoapproaches, thus developing a more complex concept of time which accountsfor the fundamentally unstable and cyclical character of economic movement.

Time in the theory of the circular flow versus time in thetheory of development

In The Theory of Economic Development (TED) (1912 [1934]) and in the firstvolume of Business Cycles (BC) (1939) Schumpeter developed a detailedconceptual structure of his approach to economic change. What emerges fromthese writings is a fundamental duality in the treatment of time, i.e. the conceptof time in the circular flow and the concept of time within economic develop-ment or evolution. This opposition is somewhat similar to the distinctionbetween reversible and irreversible time, suggested notably by Georgescu-Roegen (1966). However, on closer examination this comparison isquestionable since, in his description of the circular flow, Schumpeter adopts animplicit conception of time that does not exactly coincide with the reversibleand immutable time of classical mechanics.

Before going into more detail regarding the treatment of time inSchumpeter, a terminological clarification is needed. To begin with,Schumpeter identified the circular flow model – as well as the concept of‘growth’ which he regarded as a simple extension of the circular flow – with the‘static’ method, while the theory of development corresponded to ‘dynamics’.4

However, for Schumpeter, the circular flow is not equivalent to a ‘stationary’economy as it does not exclude changes in the data.5 Thus, the circular flowand development are both affected by economic changes, though the nature ofthese changes is different.6 Later on, Schumpeter abandoned the ambiguousterms ‘static’ and ‘dynamic’. He did so partly in the third edition of the TED,and then fully in BC, replacing them with the terms ‘theory (or method) ofequilibrium’ and ‘theory of evolution’, respectively. Moreover, the use of terms

Time and rationality in Schumpeter’s construct 107

borrowed from mechanics seems inappropriate in any case given thatSchumpeterian static and dynamic analysis have more in common with biolog-ical, rather than mechanical, reasoning.7

Time in the circular flow

As with Marshall, Schumpeter’s theory proceeds in stages: he begins with thestudy of the state of static equilibrium, referring explicitly to the Walrasianmodel, from whence he proceeds to dynamic analysis. However, Schumpeterturned the timeless Walrasian model into the model of the circular flow – thestarting point of the TED – whose object is ‘the normal circular flow withinaccustomed channels’ (Schumpeter 1912 [1934]: 70). The circular flow is not aneconomy ‘at rest’ as the mechanical analogy would require: it ‘evolves unceas-ingly’, it is the ‘most striking of all economic rhythms’ (ibid.: 6; 1935a: 76).8 AsArena (1992) has stressed, the nature of the circular flow is inseparably linkedwith its time dimension.

More specifically, Schumpeter compares the circular flow to ‘the circulationof the blood in an animal organism’ (Schumpeter 1912 [1934]: 61), thus high-lighting the continuity of the circular flow. The passage of time is here, therefore,seen to generate continuity of the economic process and of its reproduction,both in time and in space. This continuity of time reflects the interlocking ofsuccessive periods of economic activity on the basis of experience accumulatedin the past since, in the circular flow, all economic activity is based on theoutcomes (quantities and values) of the previous period and on the reproduc-tion of successful experience.9

This continuity in time is coupled with a spatial dimension of continuity inthe form of a ‘network of connections’ characterising the circular flow in whichagents, all of whom are at once buyers and sellers, are interdependent.10 Spatialcontinuity appears to be a consequence of continuity in time since, forSchumpeter, the permanence of economic relations in the course of timerenders possible the emergence of close economic and social links betweenactivities and agents.

Thus, the model of the circular flow emphasises the role of duration: thepassage of time creates permanent and stable conditions. However, thisapproach to time is not exactly equivalent to the conventional treatment oftime, inherited from classical mechanics and characterising Walrasianeconomics.11 In the conventional approach, time is analogous to space: there isno difference between past and future, time only produces determinist andreversible paths. This static time is Georgescu-Roegen’s (1966) ‘mechanicalclock-time’, measured by a cardinal variable, Robinson’s (1980) ‘logical time’ orO’Driscoll and Rizzo’s (1985) ‘Newtonian time’.12 In Schumpeter’s approach,the passage of time in the circular flow differs from such mechanical time in twoaspects: it is a vehicle for experience and, therefore, marked with ‘time’s arrow’;as a consequence, time establishes irreversibility of behaviour.13

In the first place, experience accumulated through repetitive activities in

108 Methodology

successive short periods and under identical conditions means that agents learnabout and adopt the best methods of production as well as acquire a betterknowledge of market conditions and consumer habits, thus allowing supply toadjust to demand. Cumulative experience is also what underlies the emergenceand establishment of well-functioning networks of relations between activitiesand agents (Schumpeter 1912 [1934]: 5–7). Thus, time, understood as contin-uous duration, improves the knowledge and co-ordination of economic actions.By reducing uncertainty, time confers an objective meaning on economic activ-ities. In other words, it generates ‘normal’ values and behaviour.14 AsHeilbroner (1988) and Oakley (1990) have pointed out, Schumpeter’s circularflow model displays similarities with the classical reproduction schemata.However, Schumpeter insists on the role of experience in the formation ofpermanent values:

[T]he old values are customary values. Long experience has determinedthem, and they are established in the consciousness of individuals. They areonly altered in the course of time and under the pressure of further longexperience.

(Schumpeter 1912 [1934]: 142)15

‘Normal’ quantities, prices and productive arrangements are the outcome of alearning process that takes time. This is why Schumpeter, both in the TED andin BC, defines the circular flow not as a state of equilibrium but as a ‘movement’or a ‘tendency towards equilibrium’.

In the second place, according to Schumpeter, the repetition of the sameeconomic process over many periods of production leads to the formation ofpersistent habits in the spheres of production and economic exchange; in short,it leads to the emergence of ‘routine’ behaviour. Economic decisions are deter-mined by past data. Using the example of a farmer, Schumpeter provides thefollowing description of such behaviour: ‘All the preceding periods have …entangled him in a net of social and economic connections which he cannoteasily shake off. They have bequeathed him definite means and methods ofproduction. All these hold him in iron fetters fast in his tracks’ (ibid.: 6).Schumpeter here stresses that individuals become locked into ‘a fixed habit ofthinking’ (ibid.: 86) and are reluctant to face change and uncertainty. Thestrengthening of habits from one period to another is, therefore, a source of theirreversibility of behaviour.

Time in the circular flow is also different from strictly reversible ‘mechanical’time insofar as what matters is not that there is no difference between today andtomorrow, but that the present is determined by cumulative past experiences.Furthermore, this cumulative and irreversible time is also of a deterministnature. It should be emphasised that Schumpeter’s theory of growth employs thesame conception of time. As he sees it, economic growth phenomena – relatedto the slow and continuous changes of population, means of production andeven tastes – are also essentially ‘static’, since they are an expansion ‘by small


steps’ of the circular flow.16 The theory of growth deals with gradual quantita-tive variations in the data which are progressively and smoothly absorbedthrough continuous infinitesimally small adjustments which are based on habitsand experience (i.e. routines) and produce cumulative effects. Because of theseconstant small displacements, the economy is never in equilibrium but alwaysin the ‘neighborhood of equilibrium’ (Schumpeter 1935b, 1939). Qualitativechanges, on the other hand, which produce major disequilibria, are of signifi-cance for the theory of development.17

Finally, time in the circular flow can be considered as the time necessary forthe establishment of an economic organisation. However, this concept of timewhich we call ‘retrospective time’, to highlight that its only function is tostrengthen past habits, does not give any space to the future, or more specifi-cally to the ‘time of becoming’. In Schumpeter’s theory of development time is,on the contrary, basically the time of becoming.

Time in the theory of development

The difference between Schumpeter’s theory of the circular flow and his theoryof development or evolution is not of a quantitative but of a qualitative nature(Perroux 1935). This is what accounts for the originality of his approach.Schumpeterian dynamics is not a mere extension of the apparatus of thecircular flow, but is based on a new conceptual framework that is fundamentallyopposed to the model of the circular flow. As Schumpeter explains, ‘[d]evelop-ment in our sense … is change in the channels of the flow as opposed to thatmovement; it is displacement of the equilibrium state as opposed to the move-ment towards an equilibrium state’ (1935a: 92). While the circular flow stressescustom and habits, evolution emphasises the rupture with habits since it dealswith radical qualitative changes, that is ‘productive revolutions’ in the indus-trial and commercial sphere, such as the introduction of railways, whoseimplementation depends on innovative activities.

It follows that the implicit conception of time in the theory of developmentis antagonistic to the conception of time in the circular flow. In the former,time is not only irreversible, but also discontinuous, future-oriented, and non-deterministic. It is associated with the breakdown of the continuity of theeconomic process, a breakdown based on events that change the direction ofhistory. This discontinuity arises from the actions of the entrepreneurs whointroduce ‘new combinations’ or innovation:

This historic and irreversible change in the way of doing things we call‘innovation’ and we define: innovations are changes in production func-tions which cannot be decomposed into infinitesimal steps.

(Schumpeter 1935b [1989]: 138)

Thus innovations are qualitative jumps that cannot be absorbed gradually likepopulation changes. Continuity in time is broken since the implementation of

110 Methodology

the new combinations is based neither on the outcome of the precedingeconomic periods nor on cumulative experience.

In the first place, new combinations do not arise from previously existingproduction as a transformation of old combinations, so to speak, nor do newcombinations automatically replace old ones between two periods. Rather, they‘start producing beside them’ (Schumpeter 1912 [1934]: 66). From the point ofview of development, there is no link of cause and effect between new and oldproductive combinations. The rhythm of the circular flow is interrupted by anexogenous shock (Teboul 1992). Even though Schumpeter is aware that thiscontradicts the historical reality of evolution, this methodological choice isjustified by his wish to get as close as possible to a ‘pure’ notion of evolution.18

Furthermore, by insisting on the total novelty of productive combinations,Schumpeter highlights the absence of determinism from change.

A corollary of this is the assumption that new possibilities require a differentuse of existing productive means. Therefore, novelty cannot be integrated intoexisting networks connected to the old combinations. It follows that entry intoa new phase, i.e. a new context in terms of time, also disrupts the spatial conti-nuity of the circular flow (Guilbaud 1951), giving rise to new forms of economicand social relations.

In the second place, contrary to ‘duration-time’ in the circular flow whichgenerates passive adaptation, time in development generates discontinuitiesbecause it redirects economic actions towards totally new goals for which expe-rience and practice have not as yet been gained. Time here is basically relatedto the conscious action of agents who develop reflective plans based on theirviews of the future. According to Schumpeter, ‘[c]arrying out a new plan andacting according to a customary one are things as different as making a road andwalking along it’ (Schumpeter 1912 [1934]: 85). This notion of time related tothe formulation of projects produces uncertainty because it calls into questionthe stable features of the economy while the impact of the new plans is still notwell understood. Moreover, rather than representing a continuous effort tomove towards equilibrium, it creates long-term disequilibria since the decisionto innovate requires ‘leaving the accustomed channels’, and the replacement ofexisting productive processes by the new ones is not instantaneous.

Hence, whereas ‘duration-time’ in the circular flow generated behaviouralregularities and coherent organisation, ‘rupture-time’ in development causesdisorder, ‘disharmonies’ and ‘disturbance of existing structures’ (Schumpeter1939: 102).

Time and economic decision-making

The two conceptions of time identified above help to clarify the role of deci-sion-making in the economy, which is crucial to Schumpeter’s theory. As amatter of fact, the role of decision-making takes on its full significance in rela-tion to the element of time, both in the circular flow and with evolution.According to Schumpeter (ibid.: 99), in order for firms to work out their choices


they have to define, consciously or unconsciously, the ‘horizon’ of economicactions, that is the range of options at their disposal which, in turn, depend onthe degree to which they are open to new productive possibilities. The differ-ences in horizon between firms determine two kinds of response to a shock.More specifically, each theoretical framework defines different ways of inte-grating time in the process of economic decision-making: in the circular flow,decisions are taken on the basis of memory – the synthesis of experience – whilein a situation of development what matters is expectation, i.e. the views on thefuture.

In the circular flow, experience and habits inherited from the past dominateindividual decision-making: decisions are based on a retrospective approach totime. In Schumpeter’s terms, appropriate behaviour takes the form of ‘passiveadaptation’ (Schumpeter 1927 [1989]: 27; 1939: 72), that is adaptation tochange ‘without materially deviating from familiar lines’ (Schumpeter 1912[1934]: 81). When adopting this ‘normal’ behaviour, responses to a shock arenot creative but automatic. Economic agents do not exploit new possibilitieseven if, as Schumpeter specifies, they ‘could act in another way’ (1935a: 76,original emphasis). The reason is that their time horizon is limited to thepresent period. The ‘manager’ of the circular flow does not need to anticipatefuture wants: he is preoccupied solely with the immediate satisfaction of presentwants since he expect wants and returns on production to remain unchangedover time (Schumpeter 1912 [1934]: 35–6; 1935a: 46). Short-term (adaptive)expectations are feasible in the context of quasi-stationary dynamic analysis(‘growth’ in Schumpeter’s terminology), but all they can achieve is to facilitateconvergence towards equilibrium since required adjustments are small andpurely quantitative.

Furthermore, with regard to individual motives, the business manager’sbehaviour is not driven by the conventional assumption of individual utility-maximising rationality, but is socially or objectively determined. The ‘normal’motivation prevailing in the circular flow is the satisfaction of wants: ‘Theremay be rational conduct even in the absence of rational motive. … There is verylittle of conscious rationality, still less of hedonism and of individual egoismabout [motivation]’ (Schumpeter 1912 [1934]: 91). This is why ‘individualbehaviour in the circular flow appears … to be adaptive rather than optimal’(Arena 1992: 133). This kind of objective, ‘automatic’ and ‘subconscious’rational conduct is a construct of duration-time. It is the outcome of long-termlearning based on trial and error, and it is ‘transmitted … by inheritance,teaching, upbringing, pressure of environment’ (Schumpeter 1912 [1934]: 84).

Once development takes place, individual decisions are taken in ‘prospec-tive’ time. The entrepreneur, the central actor of evolution, has a wider timehorizon since his defining characteristic is that he carries out new, as yetuntested, plans. His rationality is, therefore, of a different nature: theentrepreneur acts consciously on the basis of egoistic, but not hedonisticmotives.19 Contrary to the mere producer, the entrepreneur has the ability toanticipate and to ‘project himself’ into the future. ‘Initiative’, ‘authority’, ‘fore-

112 Methodology

sight’ are the terms Schumpeter (ibid.: 75) uses to describe the entrepreneur’sattitude. What we are confronted with here is a more subjective and intentionalrationality. Entrepreneurs design their plans on the basis of long-term expecta-tions: they have to anticipate future gains from new combinations as well as thewants of tomorrow. Furthermore, to ‘produce differently’ rather than to ‘producemore’ takes time. To run an ‘enterprise’ is a long-term activity the results ofwhich are not easily foreseeable. Therefore, the entrepreneur is faced withuncertainty and the shortage of information: he ‘is without those data for hisdecisions and those rules of conduct’ to guide his activities (ibid.: 84–5).Uncertainty is, thus, inherent to time in the framework of development. ForSchumpeter, the entrepreneur’s success ‘depends upon intuition, the capacity ofseeing things in a way which afterwards proves to be true, even though itcannot be established at the moment’ (ibid.: 85). The reduction of uncertainty,therefore, indicates that duration-time has produced some effects, that alearning process which corroborates the new ‘vision of the world’ has been setinto motion.

Finally, Schumpeter’s theoretical models describe two different ‘apparatusesof response’.20 The agents’ response is either built on past values and shaped byexperience or based on expected future values which have not as yet been deter-mined. Thus, what is involved is an ex-post rationality in the first case, and anex-ante rationality in the second.

In sum, two antagonistic functions of time emerge from Schumpeter’s anal-ysis which we can interpret as the formation of habits and regularities, on theone hand, and the emergence of a new path of economic development, on theother. These two functions are closely related to the ‘two categories of dynamicanalysis’ distinguished by Guilbaud (1951) in his analysis of Schumpeterianthought. As mentioned above, given the abstract perspective adopted bySchumpeter, what we encounter are theoretical concepts of time without refer-ence to any historical data. The following section is concerned with theanalytical role of both concepts of time in Schumpeter’s theory of businesscycles. In actual fact, the process of economic change can be described as theresult of the combination of these two antagonistic functions of time.

Schumpeter’s theory of economic change: relatingretrospective and prospective time

The TED provides a first outline of Schumpeter’s theory of fluctuations whichhe was to develop more fully in BC. Schumpeter’s main goal is to explain thedynamics of a capitalist economy whose very essence is the constant alternationbetween periods of prosperity and depression. In his view, this cyclical processcan be explained from the point of view of ‘pure theory’: that is, independentlyof the particular context in which this process takes place and discarding allexternal and internal disturbances (Schumpeter 1912 [1934]: 220–3). Thismodel of the ‘purely economic’ or ‘normal’ cycle, associated with the impact ofindustrial innovations, is part of what Schumpeter later called ‘dynamic


analysis’.21. Teboul (1992) labels this ‘functional’ dynamics, as opposed tohistorical dynamic analysis.

According to Schumpeter, in the ‘pure model’ of the cycle time is ‘theoretictime’: that is, ‘a time which serves as an axis for a logical (and not merelyhistorical) sequence of events’ (Schumpeter 1939: 138, fn. 1). This theoretical– or dynamic – time is irreversible but it refers to theoretical data only and is,therefore, to be distinguished from ‘historic time’ which Schumpeter associateswith the analysis of historical events and phases.22 However, for Schumpeter,both concepts and types of analysis are complementary and indispensable forthe study of change and economic cycles. In effect, the empirical observation ofcycles – the sequence of values and quantities in historical time – requires atheory to explain them: that is, a ‘cyclical’ model that provides the tool to inter-pret historical cycles. However, according to Schumpeter, a logical constructalone is insufficient to account for cyclical phenomena in their totality, and thetheoretician must also take recourse to industrial history in order to specify andto countercheck his analytical work, as well as to discover what lies behind the‘surface’ manifestations of cycles.23

On the basis of our discussion in the first part of this chapter, we are in aposition to show that what underlies the cyclical movement of a capitalisteconomy is the interlinking of the ‘two categories of dynamic analysis’.24 Thus,purely economic change is not characterised by a unique time dimension.Schumpeter’s ‘dynamic time’ is complex in that it combines both of the theoret-ical times highlighted above; the operation of prospective time is necessary toinitiate a process of transformation, but only if it is linked to retrospective timecan it account for the cyclical nature of this process. The cycle is, therefore,characterised by an interaction of ex-post and ex-ante rationality.

Evolution requires the operation of prospective time

In BC, the term ‘Economic Evolution’ refers to ‘[t]he changes in the economicprocess brought about by innovation, together with all their effects, and theresponse to them by the economic system’ (ibid.: 86). Thus, evolutionary theoryin this sense is a theory of the process of change in an economic system under-going structural transformations in the sphere of production. However, forSchumpeter, evolution does not take the form of a gradual or ‘organic’ develop-ment comparable to the growth of a tree (Schumpeter 1912 [1934]: 216). Thetheory of evolution cannot, therefore, ignore prospective time which generatesqualitative change and irreversibilities in production.

In fact, outside of ‘rupture-time’, the economy’s movement is continuous andregular, as if following a linear trend, because agents can only replicate themodes of production and consumption which experience has impressed onthem. However, according to Schumpeter, the real economic process cannot bedescribed solely in terms of a tendency towards equilibrium, as neither themodel of a stationary economy, nor that of regular ‘growth’, are capable of

114 Methodology

explaining the important fluctuations and the lasting disequilibria inherent inthe movement of a capitalist economy:

The kind of wave-like movement, which we call the business cycle, is inci-dent to industrial change and would be impossible in an economic worlddisplaying nothing except unchanging repetition of the productive andconsumptive process.

(Schumpeter 1935b [1989]: 139)25

‘Rupture-time’, or the time of evolution, on the other hand, can generatecyclical fluctuations by creating a qualitative difference between the past andthe future and by transforming historical states. Schumpeter maintains that theentrepreneurs’ innovative activities – which drive the cyclical movement – willonly emerge in an environment that is very close to a state of general equilib-rium, and where, therefore, the risk of failure is minimal.26 The reason is that,as opposed to a situation of recession or depression, in a context resembling astationary state or regular growth, a climate of confidence prevails. New possi-bilities exist without as yet being exploited – that is, the economicenvironment is potentially creative – and business practices are generally stan-dardised. Thus, we face a ‘normal business situation’, in Schumpeter’s terms(1939: 4). In these circumstances, some firms will change their conduct andadopt new forms of behaviour to gain future profits. Schumpeter argues that, ina capitalist society characterised by the competitive struggle of a typical marketeconomy, there will always be firms which will abandon familiar practices andstrive to gain abnormal profit by exploiting new opportunities. Thus, firmcompetition is dynamic since it encourages economic agents to act and tosituate themselves in prospective time. As Schumpeter explains, from themoment new enterprises appear on the scene, ‘the existing or “old” firms for atime work on as before, and then react – with various characteristic lags and invarious characteristic ways – adaptively to the new state of things under thepressure of competition’ (ibid.: 95). This is the starting point of the upwardphase of the cycle.

But the emergence of new enterprises is a temporary process characterised by‘clustering’, as Schumpeter puts it, and the countervailing movements that tendto set in after a while put a halt to the operation of prospective time whichbegins to give way to the operation of retrospective time. Prospective timealone does not allow one to grasp the entirety of time-related mechanismsunderlying the cycle and, in particular, the phase of recession (or depression)that ‘logically’ follows prosperity (or the upswing), and at whose end ‘funda-mental data’ are modified. In the process of evolution, time does not onlygenerate breakdowns but also adjustments and regularities. The repetitive andcontinuous time of the circular flow, therefore, has a role to play inSchumpeter’s theory of economic evolution and, what is more, it is now alsoendogenous.


The role of retrospective time in the movement of the cycle

Schumpeter never saw the need to reject the analytical method of ‘thetendency towards equilibrium’ to explain the cyclical process. To the contrary,he considered this method useful for dealing with some of the phenomena ofevolution.27 In BC, Schumpeter suggests that the ‘traditional theory of equilib-rium’ has its place in the context of a changing, disturbed environment as ‘adescription of an apparatus of response’ (ibid.: 68, 99). In the process of develop-ment, the agents’ response to innovation can be based either on routine or itcan be creative, with both types of behaviour playing their part in economiclife.28 In a world that is very heterogeneous, not all economic agents will besituated in the same time-frame. Therefore, their decision-making can be basedon past values as well as on anticipated future values. This is the main reasonwhy the decline or disappearance of some firms or sectors and the emergence ofothers can occur simultaneously.29 For some firms, however, innovation, whilecreating a ‘new economic space’, also broadens their time horizon and opens upnew opportunities for expansion (ibid.: 134). Consequently, neither firms’behaviour nor the agents’ rationality is unchangeable.

At a more basic level, what we need to show is that retrospective time, orthe time of the circular flow, has a ‘logical’ place in the unfolding of the cyclicalprocess. Schumpeter’s analysis of the cycle demonstrates that the circular flowreacts to the development process: retrospective time operates at the level ofthe adaptive and stabilising forces of the economy: that is, it operates ex postonce the qualitative changes have been introduced. It is in this particular sensethat retrospective time is endogenous. More specifically, in Schumpeter’s theoryof the cycle, two of the above-mentioned features of the circular flow – theeffect of the passage of time on improvements in the state of knowledge, on theone hand, and its effect on the creation of routines and stable economic rela-tions, on the other – appear once more, whereas the more Walrasian features ofthe circular flow – such as individual cost–benefit calculus, value, etc. – areemphasised to a much lesser extent. Hence, what is interesting in Schumpeter’stheory of the circular flow is highlighted here – namely, the role of time and oflearning. As a matter of fact, retrospective time exerts its influence at twolevels: the sector level and the level of the economy as a whole.

Learning and knowledge improvements in the innovative sector

To begin with, the time of the circular flow, associated with the reproduction ofsuccessive economic periods over time, comes into play in that it encouragesthe adoption of an innovation by those firms in the innovating sectors.30 In thiscase, both kinds of rationality coexist since learning proceeds in the context ofinnovation and relative uncertainty. Every ‘entrepreneur’ reorients his strategyand expectations on the basis of other producers’ reactions to innovation (DeVecchi 1995). Repetitive applications of the new practice over some time andthe resulting accumulation of experience lead to the new practice spreadingprogressively throughout the sector via imitation of the ‘pioneering’

116 Methodology

entrepreneurs. After a period of gestation, ‘other entrepreneurs follow, afterthem still others in increasing number, in the path of innovation, whichbecomes progressively smoothed for successors by accumulating experience andvanishing obstacles’ (ibid.: 131). Thus, as the innovation is diffused, its specifictrajectory is being shaped by the pattern of learning and improvements thataccompany its diffusion.

The operation of retrospective time reduces uncertainty in the innovativesector, allowing the boom to expand. In actual fact, the new combination canonly gain a permanent foothold in the economy if uncertainty decreases owingto the success of the first entrepreneurs and the removal of the main obstaclesto innovation. Thus, the time horizon of agents narrows once they begin toimitate the new industrial practice, since its effects and results have becomemore predictable. Once an innovation has been successfully introduced into theeconomy and its superiority established, ‘[i]t is only necessary to repeat what hasbeen done before to acquire the equivalent advantages’, and behaviour nowaims ‘only to assure the continuity of the existing stream of goods’, that is‘maintaining something already in existence’ (Schumpeter 1912 [1934]: 147–8).Hence, in the long run, the implementation over time of new combinationseliminates all durable profits.

The formation of persistent habits and stable economic relations inthe economy as a whole

Moreover, the retrospective time of the circular flow brings about the reorgani-sation and the structural changes required to restabilise the economic system asa whole, once the operation of the prospective time of evolution has disturbedthe ‘normal’ data of economic activity. The operation of prospective time,however much a force of ‘progress’, ultimately destabilises the whole economywhose modes of organisation and co-ordination are not well-adapted to the‘industrial revolution’ underway. This revolution disrupts the spatial continuity,i.e. the network of connections, and industrial relations in particular. Accordingto Schumpeter, the generalised adoption of the new practice throughout theinnovative sector unsettles overall industrial organisation, ultimately forcing allsectors to adapt to the new situation.31 But it also increases general uncertainty,since, in a rapidly changing environment, firms will lack experience and newdata are uncertain. This hinders the process of adjustment and causes numerouserrors of anticipation which are, in turn, at the root of the ‘secondaryphenomena’ of evolution described in the TED and BC. Hence, the overalleconomic environment becomes unfavourable to innovation and the originalimpetus provided by the implementation of new combinations ceases to func-tion.

A phase of successive readjustments and the search for a new global equilib-rium follows. Schumpeter first called this phase ‘normal depression’(Schumpeter 1927, 1912 [1934]), and later on ‘recession’ (1935b, 1939). ForSchumpeter,


[i]t constitutes the response by the system to the results of entrepreneurialactivity – adaptation to the new things created, including the eliminationof what is incapable of adaptation, resorption of the results of innovationinto the system, reorganization of economic life so as to make it conform tothe data as altered by enterprise, remodelling of the system of values, liqui-dation of indebtedness.


The normal depression is hence ‘the reaction of business life to the situationcreated by the boom or, more precisely … the movement of business lifetowards a new state of equilibrium conforming to the data created by the boom’(1927 [1989]: 29). Schumpeter goes on to state that ‘[t]he new data, created bythe boom and upsetting all the bases of industrial and commercial calculation,are an “objective” fact. As such they enforce “objective” adjustments’ (ibid.).

More specifically, this process of adjustment and reorganisation which trans-forms novelty into routine produces a new industrial network and gives rise to anew stable state with little uncertainty – in short the process that ‘leads up to anew neighborhood of equilibrium’, ‘takes time’ and is sequential (Schumpeter1939: 137). This process is the product of ‘duration-time’, of a learning processand of the adoption of behaviour aimed at stabilising economic relations on thebasis of the new, but already more familiar, combinations. Schumpeter argues, ineffect, that an innovative attitude will prevail and ‘prove to be sufficiently nearto reality, if things have time to hammer logic into men’ (Schumpeter 1912[1934]: 80, our emphasis). And further: ‘The customary data are altered forevery business. The extent and nature of the change, however, can only belearned from experience’ (ibid.: 238, emphasis added).

Finally, ‘duration-time’ intervenes to give some coherence to creative deci-sions and to institute the new norm – the ‘normal business situation’. It bringsabout the adaptation of the economic system to new conditions and defines anew historically specific growth regime. In other words, retrospective timeprovides Schumpeterian evolution with an element of self-organisation: ‘order’can emerge out of ‘chaos’. Moreover, while irreversibility cannot exist in aworld without creative decision-making and, thus, outside ‘rupture-time’, ‘dura-tion-time’ reinforces the irreversibility generated in prospective time. In effect,it is impossible to return to the old productive combinations once innovationhas been ‘incorporated in the circular flow’, i.e. once economic agents havedecided to follow the new trajectory and the economic system has been entirelyreorganised in order to absorb the innovation and its impact.

Concluding remarks: theoretical times, historical time andinstitutions

A reading of Schumpeterian dynamics from the point of view of the implicitconcepts of time of the circular flow and of development permits a better under-standing of the mechanisms of change and diffusion that characterise the

118 Methodology

capitalist economy. It also gives a certain coherence to his work which, asTeboul (1992) has observed, takes its inspiration from a priori contradictorymethodological frameworks (i.e. the Austrian school, Walras, the GermanHistorical School and sociological thought). Schumpeter’s analysis highlightsperiods where the operation of prospective time is predominant (the accumula-tion of innovative behaviour), as well as others where the operation ofretrospective time is prevalent (the establishment of new routines).

However, these models of time are of an abstract nature in that their signifi-cance is independent of a specific socio-institutional and historical framework.They constitute the basis of logical, and not historical dynamic analysis. But theeconomic process takes place in historical time. As Schumpeter himself under-lined, ‘[c]ycles run their courses in the historical evolution of the capitalisteconomy’ (1954: 1167, author’s emphasis). And he also claimed that ‘we mustinvestigate historically the actual industrial processes that … revolutionizeexisting economic structures’ (1949 [1989]: 326, author’s italics). As a matter offact, both levels of analysis are not clearly separated and Schumpeter’s model ofthe cycle is not a-institutional:

[O]ur model and its working is … strongly institutional in character. Itpresupposes the presence, not only of the general features of capitalistsociety, but also of several others which … are not logically implied in theconcepts either of economic action or of capitalism. … In this sense theanalysis presented has, in fact, itself been called historical.


Thus, the analysis of the economic process in ‘theoretical time’ provides aframework to interpret the sequence of events in ‘historical time’, as they wouldtake place under ‘Competitive Capitalism’. It specifies the logical connectionbetween two models of time in an economy with a particular institutional andsociological background, i.e. a capitalist institutional setting in which privateproperty and competition are predominant, but also a particular ‘spirit’ – that ofthe ‘industrial bourgeoisie’ – as well as particular ‘schema of motivation’ (ibid.:144–5).

More fundamentally, two forms of institutional support for individual actionsemerge from these two temporalities. Schumpeter’s theory of economic changeis based on a dual mode of decision-making, which is characteristic of capi-talism where individuals and institutions interact (De Vecchi 1995). Theappearance of credit institutions – and of interest – is directly connected to theoperation of prospective time: credit creation comes to support entrepreneurialactivities and to co-ordinate the investment required to implement the newcombinations. The general purpose of this institution is to maintain and toexploit the potential of change of an economy. Retrospective time, on the otherhand, is related to the emergence of another type of institution whose purpose itis to preserve the viability and reproductive capacity of the economic system inthe presence of change by safeguarding the coherence and mutual consistency


of individual actions and by giving a meaning to collective action. Institutionalresponses of this kind encompass institutions in the sense suggested by Veblen –new habits and rules of behaviour for consumers and producers – as well as morespecific organisations for the co-ordination of decision-making in markets andfirms.

Finally, once the analysis has been ‘historicised’, the operation of bothmodels of time depends on actual economic organisation, behaviour and insti-tutions. On the basis of Schumpeter’s comprehensive historical analysis (1939,1942), it could be shown how institutional changes in capitalism have trans-formed the time-path of the economy. Under ‘Trustified’ or ‘OrganizedCapitalism’, firms’ innovative strategies, based on a high degree of institutional-isation of research activities, are connected to the reorganisation of theeconomy in a specific way, closely associated with imperfect competition, thedominance of large firms and the development of regulatory economic policies.

In a similar vein, albeit at a lower level of abstraction, one could shed somelight on another important issue: namely, the question of the causes behind thequalitative shock that triggers off disturbance. According to Teboul (1992),time is endogenous in Schumpeter’s cycle theory. However, while this appearsto be true for retrospective time, prospective time appears to be exogenous sincethe qualitative change is – in general – not generated by the learning processesoccurring in the circular flow.32 Experience occurs ex post: it corroborates acertain ‘vision of the world’ and institutionalises it throughout the economicsystem. But the ‘dynamic continuity’ of time stressed by O’Driscoll and Rizzo(1985) – the interrelation between successive instants through memory andexpectation – seems to be absent from Schumpeter’s approach. In 1966,Georgescu-Roegen put the question thus:

[H]istory, of an individual or of a society, seems to be the result of twofactors: a hysteresis process and the emergence of novelty. Whether noveltyis an entirely independent element or only a consequence of the hysteresisprocess is perhaps the greatest of all moot questions.

(Georgescu-Roegen 1966: 66)

Notes1 See e.g. Georgescu-Roegen 1966, 1994; O’Driscoll and Rizzo 1985; Prou and

Walliser 1988.2 See, however, Arena (1992); Oakley (1990).3 The term is here broadly understood to mean the theory of economic change.4 See the TED (first and second edition) and Schumpeter’s 1927 and 1928 articles.5 ‘ “Static” theory does not assume a stationary economy; it also deals with the effects

of changes in data’ (Schumpeter 1912 [1934]: 82 fn. 1).6 See Schumpeter 1934: 61–4 and 82.7 In the TED, Schumpeter rejects the mechanical analogy to describe static reasoning:

‘[T]he phrase “statics” is not very felicitous: It arouses the idea, which is extraneousto us, that we are referring to mechanics’ (1935a: 76). Schumpeter (1912 [1934]: 61;1939: 36) prefers to compare the economy to a living organism that can be observedfrom two perspectives: the functioning of the organism – the circular flow – and the

120 Methodology

process of selection and mutation that produced this organism – evolution.Nevertheless, as Hodgson (1993) noticed, Schumpeter does not conceive evolutionin terms of a Lamarckian or Darwinian process of selection, despite these raremetaphors. On this point, see the Introduction to this volume.

8 We refer in this chapter to both the French and the English translations of Theorieder wirtschaftlichen Entwicklung. Both are based on the second 1926 German edition,but as Schumpeter indicates in the Preface to the 1934 English translation, ‘[I]nsome places, the exposition has been modified and a number of pages have beenrewritten’, thus, there are differences some of which are of importance for us here.

9 ‘[E]veryone will cling as tightly as possible to habitual economic methods … Thusthe economic system … will be at all times connected with the preceding state ofaffairs. This may be called Wieser’s principle of continuity’ (Schumpeter, 1912[1934]: 8–9).

10 The expression ‘network of connections’ is borrowed from Guilbaud (1951).11 For the concept of ‘fundamental time’, inherited from classical dynamics in physics,

see Prigogine and Stengers (1988).12 In contrast, the ‘real time’ (O’Driscoll and Rizzo 1985), the ‘historical time’

(Robinson 1980), the ‘time of devenir’ (Prigogine and Stengers 1988), or timemeasured by an ordinal variable (Georgescu-Roegen 1966) is the irreversible timethat is marked with ‘time’s arrow’ – in Eddington’s terms – and from which noveltyemerges.

13 According to O’Driscoll and Rizzo, one of the characteristic features of ‘Newtoniantime’ is the absence of learning. Furthermore, Prigogine and Stengers specify that theancient ‘circular time’ vision, corresponding to the rhythm of seasons or generations(growing, maturing, dying), is marked with ‘time’s arrow’.

14 The reference to ‘normal values’ is more explicit in BC.15 In BC, we also find the idea that every position of equilibrium can be reached only

after experience has been gained, i.e. time has passed: ‘[I]t will eventually bring thewhole system to equilibrium, provided that all actions and reactions are performed withinthe bounds of familiar practice that has evolved from long experience and frequentrepetition’ (1939: 47, author’s emphasis).

16 For the signification of ‘economic growth’ in Schumpeter, see, for example,Schumpeter 1927 [1989]: 25; 1912 [1934]: 67–8; 1935b [1989]: 138, 1939: 83–4.

17 According to Schumpeter, growth ‘calls forth no qualitatively new phenomena, butonly processes of adaptation’ (1912 [1934]: 63).

18 ‘Every concrete process of development finally rests upon preceding development.But in order to see the essence of the thing clearly, we shall abstract from this andallow the development to arise out of a position without evolution’ (Schumpeter1912 [1934]: 64). This assumption is however relaxed in the ‘second approximation’of the cycle (Schumpeter 1939: 157).

19 According to Arena (1992), ‘energetic egoism’ defines the entrepreneur’s rationality.20 See Schumpeter 1939: 68, 72.21 In his later writings, Schumpeter called static ‘a relation’ which ‘connects economic

quantities that refer to the same point of time’, and dynamic ‘a relation’ which‘connects economic quantities that refer to different points of time’ (1954: 1142). Inthis respect, all present quantities ‘are in reality related to other economic quantitiesthat belong not to the same moment but to the past or to the expected future’ (ibid.:1143). If one accepts this definition of dynamics, the circular flow – that ties presentvariables to past variables – and the model of evolution – that ties present variablesto future expected variables – are both part of the dynamic method.

22 On the distinction between ‘theoretic’ time and ‘historic’ time, respectively, thereference to times of dynamic (or sequence) analysis and historical analysis, seeSchumpeter 1935b [1989]: 137; 1939: 138 fn. 1; 1949 [1989]: 327; 1954: 965 fn. 5,1160, 1167 fn. 19. As a result, Georgescu-Roegen’s (1966) comparison between this


distinction and his own distinction between mechanical time (time of the ‘timelesslaws’) and irreversible time (time of the ‘temporal laws’) is questionable.Nevertheless, Schumpeter does, indeed, refer to mechanical time when he definesthe ‘theoretical variables’ in terms of the functional relations of Walrasianeconomics, where time ‘serves only as one of the coordinates’ (1939: 194). Thus,relations between variables are ‘timeless laws’. However, in our opinion, the modelof the circular flow, although inspired by Walras’s theory, suggests a different concep-tion of time.

23 See Schumpeter 1935b, 1949, 1954: 1166–7. According to Schumpeter, all empiri-cally observed economic fluctuations have historically specific features, dependingon the particular circumstances in which they take place, contingent events, and, inparticular, the effect of institutional, political or legal changes on economic activi-ties.

24 The analysis presented here is limited to the ‘first approximation’ of the cycle, i.e. tothe two-phase cycle (prosperity and recession) that retains only the logically neces-sary and sufficient elements required to create a cyclical movement (Schumpeter1935b, 1939). However, our argument also applies to the other approximations ofthe cycle, in particular the three-phase cycle whose duration or length depends onthe period of gestation of an innovation and on the absorption of the effects of inno-vation by the economic system.

25 See also Schumpeter 1949 [1989]: 328.26 Schumpeter 1939: 136, fn. 1.27 In the second German edition of the TED, where Schumpeter still uses the terms

‘static’ and ‘dynamic’, he states: ‘This explains why the ideas used by statics can solvemany problems of evolution in its usual sense … We shall also use “static” reasoningto deal with the secondary phenomena of evolution in our sense’ (1935a: 93 fn. 1).

28 Moreover, in 1927 Schumpeter held the view that: ‘[b]usiness life, like any other,consists mainly of routine work based on well-tried experience, partly ancestral’(1927 [1989]: 32).

29 We can add that, in Schumpeter’s opinion, bankruptcies and failures are not fullyexplained by ‘errors’ (like too strong pessimism) which cause some of the ‘secondaryphenomena’ of evolution and have depressive effects. As a matter of fact, errors arenot inherent in the cyclical mechanism: ‘There is a routine procedure for dealingwith new business propositions which does not make it very easy to “get away” witheither a foolish or a fraudulent scheme. … A definite reason other than error andmisconduct is, as a rule, necessary to account for failure’ (1939: 140 fn. 1).

30 In the TED and BC, Schumpeter insists that innovation generally arises in a partic-ular industrial sector where it diffuses progressively before affecting activities outsidethe sector but related to it.

31 This explains why ‘partial disequilibria’, created by innovation and the response toinnovation in a particular industry, can cause ‘a general disequilibrium in the systemas a whole’ (Schumpeter 1939: 144).

32 Nevertheless, among those innovations that cause the cycle, Schumpeter (1939), forexample, counts changes in production methods that provide an answer to problemsin raising output or innovation that can only be implemented at a certain level ofoutput. These types of innovation originate in the previous growth phase.

Bibliography

Arena, R. (1992) ‘Schumpeter after Walras: “économie pure” or “stylized facts”?’, in T.Lowry (ed.) Perspectives on the History of Economic Thought, Vol. VIII, Aldershot:Edward Elgar.

122 Methodology

De Vecchi, N. (1995) Entrepreneurs, Institutions and Economic Change: The EconomicThought of J.A. Schumpeter (1905–1925), Aldershot: Edward Elgar.

Georgescu-Roegen, N. (1966) Analytical Economics. Issues and Problems, Cambridge,Mass.: Harvard University Press.

—— (1974) ‘Dynamic models and economic growth’, reprinted in N. Georgescu-Roegen, Energy and Economic Myths, Oxford: Pergamon, 1976.

—— (1994) ‘Time in economics’, in Hagemann and Hamouda (eds) The Legacy ofHicks: His Contributions to Economic Analysis, Routledge: London.

Guilbaud, G.T. (1951) ‘En marge de Schumpeter: quelques espérances mathématiques’,Economie Appliquée, Vol. IV (2): 243–70.

Heilbroner, R.L. (1988) Behind the Veil of Economics. Essays in the Worldly Philosophy,New York: Norton.

Hodgson, G.M. (1993) Economics and Evolution. Bringing Life Back into Economics,Cambridge: Polity Press.

Oakley, A. (1990) Schumpeter’s Theory of Capitalist Motion. A Critical Exposition andReassessment, Aldershot: Edward Elgar.

O’Driscoll, G.P. Jr and Rizzo, R.W. (1985) The Economics of Time and Ignorance, Oxford:Basil Blackwell.

Perroux, F. (1935) ‘Preface’ to Schumpeter (1935a) Théorie de l’Evolution Economique.Recherches sur le Profit, le Crédit, l’Intérêt et le Cycle de la Conjoncture, Paris: Dalloz.

Prigogine, I. and Stengers, I. (1988) Entre le Temps et l’Eternité, Paris: Fayard.Prou, C. and Walliser, B. (1988) La Science Economique, Paris: Seuil.Robinson, J. (1980) ‘Time in economic theory’, Kyklos, Vol. 33 (2): 219–9.Schumpeter, J.A. (1912) Theorie der wirtschaftlichen Entwicklung, Leipzig: Duncker und

Humblot. Preface dated Vienna, July 1911. English translation of 2nd edn as TheTheory of Economic Development: An Inquiry into Profits, Capital, Credit, Interest, andthe Business Cycle, Harvard University Press: Cambridge, Mass., 1934.

—— (1927) ‘The explanation of the business cycle’, Economica, December: 286–311.Reprinted in J.A. Schumpeter (1989).

—— (1928) ‘The instability of capitalism’, Economic Journal, 38, September II: 361–86.Reprinted in J.A. Schumpeter (1989).

—— (1935a) Théorie de l’Evolution Economique. Recherches sur le Profit, le Crédit, l’Intérêtet le Cycle de la Conjoncture, Paris: Dalloz (French trans. by J.-J. Anstett of 2nd edn(1926) of Schumpeter (1912) with introduction by François Perroux).

—— (1935b) ‘The analysis of economic change’, Review of Economic Statistics, 2–10 May.Reprinted in J.A. Schumpeter (1989).


—— (1942) Capitalism, Socialism and Democracy, New York: Harpers.—— (1949) ‘The historical approach to the analysis of business cycles’, Universities-

National Bureau Conference on Business Cycle Research, 25–7 November. Reprinted inJ.A. Schumpeter (1989).



Teboul, R. (1992) ‘Temps et dynamique dans l’oeuvre de J.A. Schumpeter’, RevueFrançaise d’Economie 7, Summer: 75–93.


Part III

Economic development andsocial change

Introduction

There is something surprising in the fact that Schumpeter’s theory of businesscycles has received considerably less attention from members of the professionthan some of his other contributions such as, in particular, his theory of innova-tion and entrepreneurship or his ideas on firm concentration and technologicalcompetition. A number of commentators have drawn attention to this differen-tial treatment of Schumpeter’s theory of the business cycle vis-à-vis other partsof his work, most notably among these some of the contributors to CunninghamWood’s collection of Critical Assessments of Schumpeter who have provided anumber of explanations for its relative lack of success.

Madaràsz, author of the introduction to the Hungarian edition of The Theoryof Economic Development, quotes Perroux, a French expert on Schumpeter, tohighlight the fact that

the inner tension of Schumpeter’s work derives from the centuries olddilemma of political economics and economic theory that is due to therelationship between abstract logical analysis and the historical and socio-logical approach … Schumpeter’s theory of development is an attempt toreconcile these two approaches, translating into the language of marginalutility theory and abstract deductive general equilibrium theory the wealthof the historical and sociological material accumulated by the Germanhistorical school concerning succeeding economic systems, primarily capi-talism.

(Madaràsz 1980 [1991]: 235)

Madaràsz concludes that Schumpeter did not succeed in this attempt. Thisplausible explanation contrasts with the view provided by Elliot who arguesthat the failure of Schumpeter’s theory of the business cycle to achieve the sameseminal reputation as other aspects of his work was mainly due to unfortunatetiming. Not only did the outbreak of World War I shift economists’ attention tomore immediate practical questions which clearly did not enter intoSchumpeter’s attempt to reconstruct economic analysis and to redirect ittowards a radically new path. More importantly, the publication in 1936 of

6 Innovation and business cycles

Agnès Festré

Keynes’s General Theory further reinforced this shift in emphasis since, by thetime Schumpeter published Business Cycles (BC) in 1939, depression had begunto abate and the exigencies of World War II focused most economists’ minds onthe more immediately relevant issues raised by Keynes. While this was the greatbenefit of Keynesian analysis, it was also to the detriment of Schumpeter’s moreabstract concerns.1

Focusing on the relation between Schumpeter’s two main contributions onbusiness cycles, The Theory of Economic Development (TED) (1912) and BC,Hansen provides a still different account of the profession’s lack of receptivitySchumpeter’s theory of the business cycle. While acknowledging that the TEDestablished to Schumpeter’s international reputation, Hansen maintains that itwould be difficult to make a similar claim about his later work. In his view,

the first book comparatively small, presents a central idea in a bold, imagi-native, dashing, colorful, and eloquent style; the second, a massivetwo-volume work, rich in historical learning, takes cognizance of a vastanalytical literature but only as a side issue in the process of unfolding theauthor’s own argument.

(Hansen 1951 [1991]: 212–13)

This view is shared by Freeman who, however, rightly points out thatSchumpeter’s later work on business cycles was the subject of renewed interestin the 1980s judging, in particular, by the explosion of international confer-ences dedicated to the theory of long cycles to which Schumpeter’s work is ofcentral importance.2 This recent reappraisal of BC prompted an extensivedebate on controversial aspects of Schumpeter’s approach, such as his views onequilibrium and the role he attributes to innovation in the theory of businesscycles.

Even though these debates have certainly been illuminating by focusingattention on the inconsistencies in Schumpeter’s work, their major drawback isthat they have sidelined the truly original element of his theory of businesscycles. As Schumpeter himself has always claimed, the specificity of hisapproach consists precisely in providing a first approximation towards a betterunderstanding of the relationship between innovation and business cycles. As amatter of fact, Schumpeter’s theory of business cycles differs considerably frommost of his contemporaries’ contributions in the field.

On the one hand, Schumpeter’s views contrast with the tradition of endoge-nous business cycle theory represented in the works of Aftalion or Keynes.These authors hold that the occurrence of cycles reveals the existence of short-term disequilibria due either to the time-lag between changes in the data andthe adaptation of productive structures, or to the destabilising influence ofmoney and the portfolio choices of economic agents. However, these theoriesleave no room for the role of technological factors in the explanation of busi-ness cycles.

On the other hand, Schumpeter’s approach differs from the Walrasian tradi-

128 Economic development and social change

tion to the extent that the latter conceptualises observed paths of cycles andgrowth as deviations from the hypothetical long-term equilibrium path of theeconomy, ensuring steady and sustained growth. In this framework, technicalchange and innovation are treated as exogenous shocks disturbing this equilib-rium. Much the same could be said about the more recent real business cycletheory and the new endogenous growth theory, with the exception of thecontribution of Aghion and Howitt who can be regarded as direct successors toSchumpeter insofar as they attempt to provide an endogenous explanation ofthe occurrence of technological innovations.

The main purpose of this chapter is to assess the originality of Schumpeter’stheory of business cycles. The first section outlines the distinctive features ofSchumpeter’s approach to business cycles and economic dynamics. Section twolooks at the mechanisms constituting the cycle in Schumpeter’s two majorcontributions on this subject: the TED and BC.

The distinctive features of Schumpeter’s business cycle theory

Schumpeter always regarded business cycles not as a minor or highly specialisedsubject of economic theorising but as the most important manifestation of thedynamics of a capitalist economy. Hence, in the preface to BC, he argued that

[a]nalyzing business cycles means neither more nor less than analyzing theeconomic process of the capitalist era. … Cycles are not, like, tonsils, sepa-rable things that might be treated by themselves, but are, like the beat ofthe heart, of the essence of the organism that display them.

(Schumpeter 1939: v)

According to Hansen, Schumpeter was one of five continental economistswhose work on business cycles laid the foundations for modern macro-economics.3 More specifically, he maintained that Schumpeter’s central messagewas ‘that a dynamic society is constantly being drawn away from neighborhoodsof equilibrium by reason of the pioneering activities of daring innovators whoselightning successes entice a swarm of imitators into a wild outpouring of newinvestment activity’ (Hansen 1951 [1991]: 209). It is worthwhile mentioninghere that Schumpeter’s views on the stability of capitalism seem to have beenmuch more radical than those endorsed by Keynes. As emphasised by Wolfson(1958 [1991]: 198), Keynes raised questions about the stability of capitalism byanalysing the values of crucial economic variables in situations of involuntaryunemployment. However, the suggestion implicit in this analysis is that, giventhe ‘right’ values, stability at full employment is attainable. In contrast,Schumpeter believes that instability is of a structural nature and an inherentcharacteristic of the process of capitalist accumulation.

In his article on Mitchell, Schumpeter distinguishes between two opposingviews of economic processes:

Innovation and business cycles 129

There is the ‘theory’ that the economic process is essentially non-oscilla-tory and that the explanation of cyclical as well as other fluctuations musttherefore be sought in the particular circumstances … which disturb thateven flow. … And there is the ‘theory’ that the economic process itself isessentially wave-like – that cycles are the form of capitalist evolution.


Schumpeter clearly adheres to the latter view, whereas his Austrian contempo-rary Hayek, whose work on business cycles grew out of the same continentalbusiness cycle school, advocates the former. The comparison with Hayek is notaccidental since Schumpeter himself dedicated several passages of BC to Hayek’stheory of the trade cycle, contrasting it with his own views on economicdynamics including his conception of the rate of interest and capital accumula-tion.4 The main differences in their respective approaches may be summarisedas follows.

To begin with, even though both Schumpeter and Hayek take as the startingpoint of their analysis a stationary equilibrium state, they do not attribute thesame meaning nor the same analytical scope to this notion. Schumpeter effec-tively interprets the notion of equilibrium in terms of his well-known conceptof the circular flow economy without investment or savings. The same is obvi-ously not true of Hayek’s approach according to which the amount of realsavings constitutes the upper limit of sustainable investment. However, thisinitial difference is not as essential as might appear at first sight. In particular,the route taken by Schumpeter does not imply that he ignores growth-relatedfactors such as saving. Rather, in his view, observed fluctuations are the result ofa combination of this and other factors driving change. Thus, it would be morecorrect to say that Schumpeter eliminates savings not from his model as suchbut rather from its ‘basic skeleton’, as he puts it. Before going into more detail,it should be remembered that Schumpeter’s treatment of saving and investmentis, generally speaking, rather idiosyncratic, as his basic definition of savingdemonstrates:5

By Saving we mean the earmarking, by an household, of an element of itscurrent receipts – as distinguished from ‘capital gains’ – for the acquisitionof titles to income or for the payment of debt. If a firm does the same thingwith an element of its net receipts from the sale of products and services,we shall speak of Accumulation. The distinction between Saving andAccumulation also applies, although it may be difficult to carry out, incases in which, as in the case of many farmers, ‘firm’ and ‘household’ areone. We confine both concepts to decisions about monetary funds and weneglect, for convenience’s sake, any similar decision that may be taken withrespect to commodities. Saving and Accumulation will thus be treated aselements of a monetary process: the complementary process in the world ofgoods constitute a distinct problem.



In other words, his notion of saving stresses the fact that not only the bulk ofindividual saving but also all business saving is done with a specific investmentpurpose in mind. This implies that, as a rule, the decision to invest precedes thedecision to save. It follows, first, that unless agents perceive investment oppor-tunities, they will not normally save and that a situation characterised by adecline in investment opportunities is likely also to be characterised by adecline in saving. Consequently, most sources of savings as well as most motivesfor the act of saving would be absent from a stationary state. To put it anotherway, the bulk of savings flows from income or elements of income, namely,profits, which do not exist in a stationary state. Second, since cash holdings arenot regarded as claims to income, Schumpeter’s definition of saving excludes allconsiderations related to the Keynesian notion of liquidity.6 As a matter of fact,Schumpeter regarded the concepts of demand for money stocks and of liquiditypreference as two examples of Keynes’ ‘deus ex machina’ approach which, hefelt, should be replaced by one that derived its concepts from a thorough anal-ysis of the economic processes underlying surface phenomena.7

A second point of difference between Schumpeter’s business cycle theoryand that of Hayek – as well as the Keynesian–Robertsonian–Swedish approach– concerns Schumpeter’s claim that the malfunctioning of the equilibriummechanism between saving and investment cannot, as such, provide an expla-nation of crises or depression:

Actually, of course, we find that that equilibrium mechanism very oftendoes not work. But sound diagnostic cannot be expected from denying itsexistence or from setting up such entities as ‘optimism’, ‘pessimism’, ‘savinginstinct’, or from simply asserting that people elect to act in such a way thatmaladjustment will ensue and that saving and investment can each go itsown way indefinitely. In order to make headway, we must locate the sourcesof the trouble. They will be found in the business situations incident to theprocess of economic change we are about to describe, and link up with notspending and with variations in real investment rather than with savings.At the moment, however, it is desirable, since the ground is so fertile inmisconceptions, to make quite sure that the saving-investment mechanism,as such, does not produce anything that could qualify for the role of anexplanation of crises or depressions.

(ibid.: 78)

More generally, Schumpeter excludes growth factors, such as saving or accumu-lation by firms and households or even changes in population, from hisdefinition of dynamics. This contrasts with Hayek who regards the voluntarydecision to save as a driving factor of capital accumulation and therefore expan-sion, whereas Schumpeter maintains that economic development and businesscycles involve qualitative change. Given that the economic system is capable ofabsorbing the effect of the above-mentioned growth factors without undergoingany substantial transformation, these factors cannot, by themselves, create the


alternation of booms and depressions that can be observed. Rather, such growthfactors, and saving in particular, owe their mainly quantitative role for capitalistdynamics to another driver of change – innovation – without which the modusoperandi of saving in the capitalist world cannot be understood. Schumpeterputs it thus:

That rudiment of a pure economic theory of development which is impliedin the traditional doctrine of the formation of capital always refers merelyto saving and to the investment of the small yearly increase attributable toit. In this it asserts nothing false, but it entirely overlooks much moreessential things. The slow and continuous increase in time of the nationalsupply of productive means and of savings is obviously an important factorin explaining the course of economic history though the centuries, but it iscompletely overshadowed by the fact that development consists primarilyin employing existing resources in a different way, in doing new thingswith them, irrespective of whether those resources increase or not. In thetreatment of short epochs, moreover, this is even true in a more tangiblesense. Different methods of employment, and not saving and increases inthe available quantity of labor, have changed the face of the economicworld in the last fifty years. The increase in population especially, but alsoof the sources from which savings can be made, was first made possible inlarge measure through the different employment of the then existingmeans.

( Schumpeter 1912 [1934]: 68)

This immediately clarifies that, starting from Schumpeter’s basic framework ofthe circular flow, the only logically conceivable means of financing innovations,if they are to occur at all, consists in credit creation, not saving. Since theprocess of innovation involves the displacement of already existing factors ofproduction, and since there are no previously accumulated idle financialresources available, new sources of purchasing power have to be created. Thisprecisely is Schumpeter’s rationale for the introduction of bank credit into hisTED.

To be sure, saving and innovation are interdependent but innovation is theprime driver of cyclical dynamics. This does not imply that once the cyclicalprocess has started, a model based on savings as the only means of financingeconomic activity could not be perceived of. However, since the modusoperandi of innovation can be understood without taking account of accumu-lated savings once one allows for credit-creation, the point is that there is noneed to take recourse to saving as an explanatory factor.

Of course, Schumpeter is well aware of other ways to finance firm expendi-tures (out of previous receipts, the sale of assets, issuing bills of exchange, etc.).8

He chooses to abstract from these alternatives in order to avoid complicationsarising from the mechanical effects of these additional flows of funds on therunning of businesses.9 As has been mentioned above, what this means is that


saving can be ignored only in Schumpeter’s basic analytical skeleton, i.e. thecircular flow, not however at those stages of his theorising geared towards amore encompassing approximation of reality. This is a typical example ofSchumpeter’s methodology. As he himself reminds the reader in BC: ‘Thestationary assumptions … have importance only for the purpose of preliminaryclarification and are admitted from the outset to yield an inadequate picturewhich, taken by itself, would only mislead’ (Schumpeter 1939: 78, fn. 1).

As is well known, Schumpeter’s theory of the business cycle comprises threesuccessive approximations to reality. The first of these is contained in the lastchapter of the TED and is referred to as his ‘pure model’ of business cycles; itconsists of only two phases – prosperity and depression. Even though this basicframework abstracts from many specific features of economic movement,Schumpeter regards it as useful precisely because it isolates the most essentialelements of the cycle from the ‘innumerable layers of secondary, incidental,accidental and “external” facts and reactions’ that ‘cover the skeleton ofeconomic life, sometimes so as to hide it entirely’ (ibid.: 137).The secondapproximation incorporates more complex features of business cycles alreadymentioned in the TED but analysed in more detail in the BC. These additionalaspects of cycles, such as, for example, the intrusion of speculation, are seen tocause ‘secondary waves’ over and above the underlying ‘primary waves’. Finally,the third approximation, developed in BC and known as the ‘Kondratieff–Juglar–Kitchin three-cycle scheme’, distinguishes cycles of different lengths and ampli-tudes. Hence, once Schumpeter’s analysis progresses from the first to subsequentapproximations, the afore-mentioned growth factors can no longer bediscarded.10

A typical example of analytical extensions of this kind is provided bySchumpeter’s description of the case of ‘steady growth’:

[W]e will envisage a society, stationary in every respect, except in that itdisplays a positive rate of saving. Production functions are invariant andexternal disturbances are absent. There is a positive rate of interest. Weexclude – but this is only for the sake of convenience and brevity – allinvestment opportunities except lending to firms (this merely excludesconsumers’ credit) and assume that saving is the only source of supply ofsuch monetary means as these firms may wish to have in addition to theircurrent receipts (this assumption excludes credit-creation …). … We startfrom competitive equilibrium … Now, that equilibrium is incessantlydisturbed by the flow of new savings which are being offered to firms. If,however, the system is adapted to the actual rate of savings … this distur-bance will be currently absorbed; for, as long as saving goes on at all, eachinstallment will depress the rate of interest to the extent required to createits own investment opportunity. … And … the combination which isoptimal now requires an increase in the more durable elements, let us callthem machinery, such as will exactly equal the additional saving offeredboth in value and cost, which is what we mean by saving creating its own


demand. … The result would, in fact, be a steady growth of the system’sindustrial outfit by the steady addition to it of new units of plants andmachinery, which, however, must be of the same types as those which arealready in use or would be in use but for lumpiness, in order to exclude anew and different element which would otherwise intrude.

(ibid. 79–80)

Even though unusual, it is not surprising that Schumpeter includes the case ofsteady growth under ‘statics’. He simply regards it as an extension of the puremodel of the circular flow to allow for a positive rate of saving. Even so his ulti-mate purpose in introducing this case to the analysis is to create an organic linkbetween the circular flow and the case of economic development.

To sum up, the main reason for excluding some basic factors governing acapitalist economy from the framework of the circular flow, or the first approxi-mation of business cycles, is that Schumpeter ‘interprets them not as part of thecapitalist process as such, but merely as the consequence of capitalist development’(Madaràsz 1980 [1991]: 227). This does not, however, imply that they shouldalso be ignored at later stages of approximation, in particular not when it comesto gaining a better understanding of how new actors, new institutions or neweconomic functions emerge once the domain of static analysis is left behind.

The ‘steady growth’ case described above exemplifies this. What it tells us isthat positive rates of interest and saving or even the existence of banks may bebrought into the picture in the process of broadening the concept of the circularflow. However, the raison d’être of these additional factors is essentially linked tothe logic of dynamic economic analysis. From this perspective, the pure case ofeconomic development cannot properly be conceived of as ‘the counter-pole ofthe circular flow’ in that ‘the negative definitions of the circular flow wouldconstitute the essence of development’ (ibid.: 227). Rather, it describes theinterplay of economic factors resulting exclusively from the emergence of inno-vations. In other words, for Schumpeter economic development cannot be setinto motion by any of the conventional growth factors. Even though these willexert some influence on the path of economic development, such developmentessentially takes place due to innovation only.

Obviously, this notion of dynamics is rather unconventional. All develop-ment begins with the ‘entrepreneur-innovator’ who ‘steps on the scene as thedemiurgos of the capitalist process of development [and] every category is tailor-made to suit him’ (ibid.: 229). With this in mind, we now turn to a moredetailed discussion of the mechanisms constituting the business cycle inSchumpeter’s theory.

The mechanisms in Schumpeter’s theory of the businesscycles

In the above section, we have focused on the originality of Schumpeter’sconception of dynamic as opposed to static analysis. We have, in particular,


pointed out that what Schumpeter means by ‘statics’ contrasts with themeaning more commonly attributed to the term and clarified first by Frisch andlater by Samuelson:

Statical refers to the form and structure of the postulated laws determiningthe behavior of the system. An equilibrium defined as the intersection of apair of curves would be statical. Ordinarily, it is ‘timeless’ in that nothing isspecified concerning the duration of the process, but it may very well bedefined as holding over time.

(Samuelson 1943: 59)

By contrast, and as seen above, Schumpeter considers a state of ‘steady-growth’of the form defined in the previous section as belonging to the domain ofstatics, even though positive values for both the rates of interest and saving aswell as a constant rate of growth of consumption and production are possible.Thus, it seems that Schumpeter’s use of the term ‘statics’ also refers to what isusually associated with a ‘stationary’ situation, defined by Samuelson as follows:

Stationary is a descriptive term characterizing the behavior of an economicvariable over time; it usually implies constancy, but is occasionally general-ized to include behavior periodically repetitive over time. Used in thissense, the motion of a dynamic system may be stationary: e.g. the behaviorof a pendulum satisfying Newton’s laws of motion, but subject to no distur-bance and hence remaining at rest; or the behavior of national incomeafter a change in investment has given rise to dwindling transientgeometric progressions of the usual ‘block-diagram’ character.

(ibid.)

Likewise, Schumpeter’s concept of ‘dynamics’ does not coincide with a moreconventional notion of dynamic analysis since the latter does not exclude thestudy of stationary systems. In Schumpeter’s view, the conventional growthfactors, and saving in particular, do not qualify as explanatory factors ofeconomic development. The kind of change produced by, for instance, anincrease in the rate of saving is likely to be absorbed by a circular flow economyand cannot, hence, account for the occurrence of recurrent business cycles.Consequently, it is not the time dimension that is the distinctive feature ofdynamic analysis. However, as we shall see, Schumpeter essentially regardsdynamics as referring to non-stationary phenomena, i.e. systems on an irre-versible path.

Thus, Schumpeter draws a clear-cut distinction between statics anddynamics on the grounds that they not only deal with distinct subject mattersbut also require different methods and data. As he puts it, ‘they are not twodistinct chapters of a same edifice but rather two entirely separate edifices’(Schumpeter 1908: 182). For Schumpeter, the essential difference between thecircular flow or steady growth, on the one hand, and dynamic analysis, on the


other, is that the latter is concerned with a type of change that cannot be dealtwith on the basis of established routines, but rather upsets these. In other words,the existence of dynamics or development means doing things in a new way ordoing new things. Although, at this stage, this appears to be rather a broadconcept, Schumpeter, in fact, attributes a very specific meaning to the notion ofdevelopment: it proceeds by way of innovations defined as ‘the setting up of anew production function’ (Schumpeter 1939: 87), including the emergence ofnew commodities, new forms of organisation, the opening up of new markets ornew sources of supply, changes in the way existing goods are made or any way ofdoing things differently. Moreover, development is a dynamic process that isendogenously driven, i.e. from within the system, and has its own endogenousmechanism of propagation or diffusion. This latter is characterised by non-linearities which account for the recurrent emergence of business cycles. AsSchumpeter puts it:

By ‘development’, therefore, we shall understand only such changes ineconomic life as are not forced upon it from without but arise by its owninitiative, from within. Should it turn out that there are no such changesarising in the economic sphere itself, and that the phenomenon that wecall economic development is in practice simply founded upon the fact thatthe data change and that the economy continuously adapts itself to them,then we should say that there is no economic development.

(Schumpeter 1912 [1934]: 63)

In terms of this definition, economic development cannot be generated by anyexternal factors. This, however, raises the question as to what differentiatesinnovation from any other factor driving change. Paradoxical though this mayappear, Schumpeter regards innovation as the single cause of developmentwithout providing a clear explanation of its endogenous emergence. WhatSchumpeter is saying is that innovation implies historical and irreversiblechanges in the way of doing things and is generally associated with the activityof particular individuals (the ‘entrepreneur-hero’) establishing new businesses,thus triggering off an endogenous process of adaptation that unfolds its owndynamics within the economic system.

As Witt (1995) has stressed, Schumpeter contrasts the notion of innovationwith the notion of invention. This distinction may help to clarify the notion ofan endogenous process of innovation in Schumpeter’s analysis. It draws atten-tion to the fact that there is, indeed, an exogenous factor driving the emergenceof novelty which is, however, confined to the domain of inventions. ForSchumpeter, inventions are abundantly available and known to many people.Moreover, he argues that how such inventions came about is irrelevant foreconomic analysis. As Witt has argued, ‘all that matters in his theory is thepioneering initiative, the “doing it”, the carrying out of what is already availablebut which no one has yet ventured to realize. Consequently, Schumpeter elabo-rates upon the psychology of his “entrepreneur-hero” ’ (ibid.: 85). Under the


impact of novelty, the economic system moves away from the initial equilibriumposition and experiences comprehensive changes which govern the process ofadaptation and adjustment, and which ultimately account for the return of thesystem towards a neighbourhood of equilibrium. Thus, Schumpeter’s theory isall about the dissemination of novelty in the market and in the economy as awhole, starting from a position of equilibrium. Since, in his view, a state ofstability is also the condition for a new surge of innovation, the business cyclereally is nothing else than the ebb and flow of innovation, including the reper-cussions, or ‘secondary waves’, generated by it.

This leads us to a closer examination of how the process of developmentstarts. In Schumpeter’s pure model – that is, in his first approximation of busi-ness cycles – innovation comes to bear on a system in full employment generalequilibrium with zero profits, a zero rate of interest, where total receipts equaltotal costs which, in turn, equal total wages plus firm rents, and where, conse-quently, there is no incentive for change. Innovation is the work of a new firmcreating new production facilities. This new firm is led by a new economicagent called the ‘entrepreneur’ to be distinguished from the ‘mere manager’prevailing in the circular flow whose activities are the product of ‘routinised’behaviour. This entrepreneur has recourse to credit newly created by banks inorder to finance the undertaking. As a matter of fact, since there are no previ-ously accumulated financial resources in the circular flow, a new source ofpurchasing power has to be introduced into the economic system: ‘ [Theentrepreneur] … must resort to credit if he wishes to carry out a new combina-tion, which cannot like an established business be financed by returns fromprevious production’ (Schumpeter 1912 [1934]: 69).

Schumpeter’s preoccupation with the financial aspect of innovation andbusiness cycles, which is the subject of another contribution to this volume,may explain his relative neglect of the role technology.11 As has been pointedout with frequency by critics, Schumpeter’s conceptualisation of innovation isfairly limited. Freeman, for instance, stresses that Schumpeter fails to pay suffi-cient attention to some aspects of the diffusion of innovations, such as theinteraction of science and technology, the cumulative nature of technologicaldevelopment, the technological as well as economic interdependence of manyinnovations and the existence of technological trajectories. Instead, he substi-tuted a theory of entrepreneurship for a theory of the firm and innovation.12

This line of criticism is supported by Witt who attributes this flaw inSchumpeter’s analysis to his distinction between invention and innovation.The chief feature of this distinction is that it permits him to circumvent boththe problem of the emergence of novelty as well as the related problem of‘subjective action knowledge’ (Witt 1995: 85).

How then does Schumpeter describe the cycle? The money borrowed fromthe banks provides the innovator with the funds to employ the factors ofproduction (acquire the factor inputs) required to set up his business. Given theassumption of initial full employment, these factors will have to be divertedaway from old businesses.


However, this employment in new processes does not produce a rise in globaloutput until after the period of gestation of the innovation. This defines theprosperity phase of the cycle, characterised by rising prices in the sector ormarket affected by the innovation as well as in the economy as a whole due tothe fact that, while the output of old industries that lose factors of productionfalls, aggregate monetary demand increases in line with the newly createdcredit. Once the innovators’ output finally reaches the market, two successivephases take place:

First, there is the process of the dissemination of the innovation via a‘swarm’ of imitators based in related industrial sectors, beginning with theproduction lines that are closest to the innovation and ending with the mostdistantly related areas of production:

We know the reasons why this is likely to happen in the same field or in –technologically, as well as economically – related fields: although in somerespects a successful innovation will make other innovations easier to carryout in any field, it primarily facilitates them in the lines in which it may bedirectly copied as a whole or in part or for which it opens up new opportu-nities.


It should be stressed that at this stage, there are no losses in the economy as awhole, since, although some old firms may experience diminishing receipts oreven losses, all old firms taken together benefit from the increasing aggregatedemand and, thus, show a net surplus. More precisely, while meeting part of therising cost of labour, this additional cost is covered by the increase in earningsfrom workers’ spending on consumption:

In spite of the losses in some industries which must, under such circum-stances, be expected to be a feature of the situation, all old firms takentogether will, of course, show a net surplus. Of this we can satisfy ourselvesif, disregarding everything except the first two steps – i.e. disbursements byentrepreneurs and again the next disbursement by income receivers – weassume that labor is the only factor, wages are the only cost. Then old firmswill, obviously, have to pay but a part of the increase in the income of thoseworkmen whom they still retain while they will, at the second turn of thewheel and before the new products reach their markets, receive the wholeof it.

(ibid.: 132)

Hence, the diffusion or imitation phase of the innovation is characterised byhigher values of the new products, reduced relative costs in general and pronoun-ced industrial expansion. By way of illustration, Schumpeter discusses the caseof existing products being produced on the basis of a new production function,even though, in the TED, he argues that ‘the vast majority of new combina-


tions will not grow out of the old firms or immediately take their place, butappear side by side, and compete, with them’ (Schumpeter 1912 [1934]: 226).However, in his view, it is immaterial

whether the innovator withdraws the necessary means of production fromthe branch of industry which hitherto manufactured the same commodity,or whether he allows existing firms to continue working in the habitualmanner and begins to produce alongside of them with the new method andwithdraws the necessary means of production from quite different branchesof industry.

(ibid.: 141)

Ex hypothesis, the result will always be a higher value of the new products rela-tive to those produced prior to the innovation with the same quantities ofmeans of production.13

This raises the question as to where this higher value originates. Schumpetermaintains that it is not to be imputed to the factors of production because ‘therewould be no sense in imputing beforehand the surplus value of the new combi-nation to the means of production, since the carrying out of it would then nolonger appear as an advantage’ (ibid.). There are two main reasons why theservices of labour and land have to be estimated at their old values, that is, thevalues they were assigned in the circular flow. On one hand, ‘the old values arecustomary values. Long experience has determined them, and they are estab-lished in the consciousness of individuals. They are only altered in the course oftime and under the pressure of further long experience’ (ibid.: 142). On theother hand, ‘the entrepreneurial activity of the leader, which is indeed a neces-sary condition of the realization of the combination, may be conceived as ameans of production’ (ibid.: 143). By contrast, the values of the new products,just as their prices, do not form part of the pre-existing system of values. Theyare disequilibrium values that foreshadow the new equilibrium price structurewhich will establish itself once the process of adaptation is complete. They are,thus, essentially of a temporary nature. To begin with, absolute as well as rela-tive prices do not change. Relative prices may vary only to the extent that ‘themarginal value [of the means of production] in the previous uses indeed rises inconsequence of the withdrawal of means of production from them’ (ibid.: 144).As expansion proceeds, price variations affect all means of production and prod-ucts as the massive expansion of entrepreneurial demand implies a generalincrease in purchasing power across all areas of production affected by the orig-inal innovation:

Only because new purchasing power goes in bulk from the hands ofentrepreneurs to the owners of material means of production, to allproducers of goods for ‘reproductive consumption’ … and to the workers,and then oozes into every economic channel, are all existing consumptiongoods finally sold at ever-rising prices. Retailers thereupon place bigger


orders, manufacturers extend operations, and for this purpose increasinglymore unfavorable and often already abandoned means of production comeinto use again. And only on this account do production and trade every-where temporarily yield a profit, just as in a period of inflation … Manythings float on this ‘secondary wave’, without any new or direct impulsefrom the real driving force, and speculative anticipation in the end acquiresa causal significance. The symptoms of prosperity themselves finallybecome, in the well known manner, a factor of prosperity.

(ibid.: 226)

In this perspective, the surplus generated by successfully implementing newcombinations, that is, the capitalist entrepreneur’s profit, ‘is not only a privatebut also a social phenomenon’ (ibid.: 143). In fact, even though, to begin with,this profit is captured by the innovator – Schumpeter labels it the ‘promoter’sprofit’ (ibid.: 137) – the forces of competition will gradually work towards theemergence and establishment of a new structure of values. But the intrusion ofnew products will not, at first, affect the supply side of the economy so as tocause any tangible change in the business situation as a whole.

However, this should not distract from the more ‘disaggregated’ features ofthe process of diffusion. Thus, it is important to recall the fact that innovationshave an asymmetric effect on economic agents and their productive activities.As Schumpeter points out, under the impact of novelty,

for some of the ‘old’ firms new opportunities for expansion open up: thenew methods or commodities create New Economic Space. But for othersthe emergence of the new methods means economic death; for still others,contraction and drifting into the background. Finally, there are firms andindustries which are forced to undergo a difficult and painful process ofmodernization, rationalization and reconstruction.


These sectoral investment shifts constitute a distinctive feature of Schumpeter’sbusiness cycles analysis. In contrast to most of his contemporaries, Schumpeterregards production structures or the various levels of activity that constitute thewhole production system – individual firms, inter-firm or inter-sector links aswell as the resulting aggregate level of production – as essential components ofeconomic activity. Since innovation and evolution alter the way in whichcommodities are produced, the resulting changes in the composition of theproductive system are an important part of dynamic analysis. For instance, apreoccupation with the vertical composition of capital, as defined by Austrianslike Hayek or Mises, implies already a rejection of aggregate analysis. However,according to Schumpeter, this particular notion fails to account for the transferof resources that characterises ‘the process by which the effects of theentrepreneurial activity spread … over the whole system, dislocating values,disrupting the equilibrium that existed before’ (ibid.: 132). Therefore,


Schumpter’s dislike of the aggregated approach differs from Hayek’s. That is, hefocuses on that level of analysis which permits a better, albeit always incom-plete, understanding of the essential characteristics of the mechanisms oftechnological diffusion:

These vital parts of the mechanism of economic evolution, which arereadily seen to dominate many business situations and to produce results offundamental importance, can never be revealed statistically by measuringvariation in an index of production, or analyzed theoretically in terms oftotal output. Such an index would display nothing except increase. Butmere increase in total output would not produce those effects. It is dishar-monious or one-sided increase and shifts within the aggregative quantitywhich matter. Aggregative analysis, here, as elsewhere, not only does nottell the whole tale but necessarily obliterates the main (and the only inter-esting) point of the tale.

(ibid.: 134)

Similarly, the end of the expansion phase and the ensuing process of adaptationcannot be accurately understood if one sticks to aggregate analysis. Indeed,Schumpeter focuses attention on the fact that as the process of expansiongathers momentum and the mechanisms of imitation take over, propagationand self-reinforcement effects steadily gain in importance and disequilibriumbegins to show. What this implies is that the sectoral shifts and the redistribu-tion of income and wealth that follow the introduction of novelty are the chieffeature of economic movement. Moreover, taking account of these factors alsoprovides the key to understanding the cyclical nature of economic activity overtime.

As for the second phase of the cycle, namely, depression, it is characterisedby a fall in prices and profits. Now, if, as Schumpeter quoting Juglar maintains‘the only cause of the depression is prosperity’ (Schumpeter 1912 [1934]: 223),the causes of a slack in entrepreneurial activity must be determined on theoret-ical grounds. Schumpeter argument runs as follows:

[S]ince entrepreneurial activity characteristically starts off in a definitedirection and does not distribute itself equally over the industrial field –since it aims typically at production of a given commodity or group ofcommodities – its possibilities are, in every instance and in any given stateof the economic body, definitely limited. The results of innovation actdirectly on certain individual prices, and therefore set definite limits onfurther advance in that direction or related directions.


More specifically, even if all existing and future producers correctly foresee thefall in prices due to the increased production and availability of the new


product, a point should be reached at which this product will be produced andsold at its minimum unit cost.

Moreover, the disturbance of the equilibrium system of values caused by aninnovation and reinforced by the release of the new products triggers a processof adaptation to the changing situation. This ongoing revision of the values ofall elements of the system impairs the calculation and planning abilities ofeconomic agents. Under these circumstances, potential new entrepreneursprefer to wait until things settle down rather than to run the risk of businessfailure. As a consequence, the decline of entrepreneurial demand for financialmeans together with the repayment of loans granted to entrepreneurs push theeconomy towards deflation and depression. Even though Schumpeter does notprovide a general explanation, he suggests that, just as in the expansionaryphase, the cumulative downward process is also limited, that is it graduallypeters out. By way of illustration, he considers the case of ‘the stoppage of a firmwhich induces unemployment that in turn causes the failure of a grocer whosecustomers the unemployed workmen were’. He then adds that ‘this grocer’smarket is not completely annihilated, however, and if he disappears there willbe some space for other grocers to expand into’ (ibid.: 153).

Therefore, once depression has run its course, the system settles in a newneighbourhood of equilibrium: that is, the third stage – namely, recovery – setsin. At this point, a new innovation is likely to occur since the conditions for itsemergence – both the stability and the reliability of the system of economicvalues – have been restored.

This simplified account is potentially complicated by differences in themagnitude of innovations as well as by the interference of external factors, suchas, for instance, the intervention of the central bank. Under certain conditions,the economy may overshoot in the process of downward readjustment and thusremain in recession for longer than would otherwise have been the case.

However, for Schumpeter recession is of a transitory nature: once bankrupt-cies have ceased, a phase of recovery is inevitable. These ‘pathological’ featuresof the cycle, as Schumpeter labels them, constitute the secondary waves of thebusiness cycle that occur in addition to the primary process and result from thediverse reactions to this process, such as speculation or self-reinforcing errors ofexpectation:

The cyclical clusters of errors, excesses of optimism and pessimism and thelike are … not necessarily inherent in the primary process – which processwould produce ups and downs and, be it particularly remembered, alsolosses without any error – although they can be adequately motivated by it.But now they acquire additional importance. Part of the phenomena of thesecondary wave consists, in part, of nothing else.

(ibid.: 146)

However, even though these secondary waves may gather momentum and mayconsequently appear to be essential features of the cycle, they are merely effects


derived from the working of the primary wave, itself a direct result of innova-tion. In Schumpeter’s words:

The phenomena of this secondary wave may be and generally are quantita-tively more important than those of the primary wave. Covering as they doa much wider surface, they are also much easier to observe; in fact they arewhat strikes the eye first, while it may be difficult, especially if the innova-tions are individually small, to find the torch responsible for theconflagration.

(ibid.)

Conclusion

We began by highlighting the original features of Schumpeter’s theory of busi-ness cycles arguing, in particular, that, in contrast to the majority of existingtheories of growth and cycles, Schumpeter’s approach provides a tentativeanswer to the question of how to integrate innovation with the study of busi-ness cycles.

We then proceeded to a discussion of Schumpeter’s conception of the differ-ence between dynamic and static analysis by way of examining his treatment ofconventional growth factors, and saving in particular. Obviously, Schumpeter’sapproach to the analysis of business cycles is far from conventional. The mainreason, however, is that his conception of dynamics is directly linked to hisview that innovations are at the core of the movements of a capitalist economy.It is only by taking account of this that we can understand why Schumpeterconcentrated on innovation and abstracted from all other growth factors whoseeffects, he maintained, cannot generate recurrent business cycles.

Finally, we discussed the mechanisms underlying the business cycle inSchumpeter’s theory. This re-examination left several questions unanswered. Inparticular, the relative neglect of significant aspects of the diffusion of innova-tion highlighted in this chapter constitutes a major shortcoming ofSchumpeter’s approach. What seems to be the case is that Schumpeter assertsrather than explains the cumulative nature of innovation. More generally,Schumpeter fails to describe the succession of events insofar as he does notprovide a detailed analysis of the mechanisms that link the scarcity ofentrepreneurial abilities to the ‘swarm-like’ appearance of imitators and to thediffusion of innovations at any given point in time. As Kuznets has emphasised,one is permitted to wonder why, given the existence of an infinite supply ofpossible innovations (including inventions and other combinations), anentrepreneurial genius should systematically manage to postpone the appear-ance of the next pioneer on the scene until a particular innovation has beenimitated and diffused to such an extent that the disturbances of the equilibriumposition should stop even this genius in his tracks.14

Nevertheless, and in line with Witt, it should be stressed that this flaw inSchumpeter’s theory later proved to be the vantage point for modern adapta-


tions of his theory of innovation.15 From this point of view, Schumpeter’s desireto see others verify and criticise his ideas in the light of further evidence hasbeen fulfilled. Today, there is indeed a vast literature on industrial innovationand R&D, even though its relation to Schumpeter’s work is often regrettablytenuous.

Notes1 See Elliot 1985.2 See Freeman 1993: 18.3 The four other economists mentioned by Hansen are Wicksell (Geldzins und

Güterpreise, 1898), Tugan-Baranowsky (Studien für Geschichte der Handelskrisen inEngland, 1901), Spiethoff (‘Vorbemerkungen zu einer Theorie der Überproduction’,Jahrbuch für Gestzgebung, Verwaltung und Volkswirtschaft, 1902) and Aftalion (‘Essaid’une théorie des crises générales et périodiques’, Revue d’Economie Politique, 1909).

4 See Schumpeter 1939: 78 (fn.), 296 (fn.), 333, 345 (fn.), 603 (fn.), and 634.5 See Staley 1986: 9.6 Cash holdings do not, in fact, belong to the ‘business sphere’ (Schumpeter, 1939:

124) also referred to as the ‘commodity sphere’ or ‘sphere of circulation’(Schumpeter 1917/1918 [1956]: 176), in which the national product circulatesamong social groups with the help of bank credit. They are, instead, included in the‘sphere of hoards and reserves’ which, together with the ‘sphere of capital’, consti-tutes the ‘money market’ (ibid.). This market permits, in turn, the working of stockmarkets. See Arena and Festré 1996: 168–9.

7 See Schumpeter 1936: 795.8 See Schumpeter 1939: 578.9 See Staley 1986: 306.

10 On this point, see Date 1961.11 See Festré in this volume on ‘Money, banking and dynamics: Hayek vs Schumpeter’.

Also see Nasica’s contribution to this volume.12 See Freeman 1993: 22–3.13 See Schumpeter 1912 [1934]: 141.14 See Kuznets 1940: 262.15 See Witt 1995: 85.

Bibliography

Aftalion A. (1913) Les Crises Périodiques de Surproduction, Paris: Rivière.Aghion, P. and Howitt, P. (1998) Endogenous Growth Theory, Cambridge, Mass.: MIT

Press.Arena, R. and Festré, A. (1996) ‘Banks, credit and the financial system in Schumpeter:

An interpretation’ in L. Moss (ed.), Joseph Schumpeter, Historian of Economics,London: Routledge.

Date, K. (1961) ‘The relation of cycles and trends in Schumpeter’s model’, WasedaEconomic Papers 5: 22–34. Reprinted in J. Cunningham Wood (ed.), J. A. Schumpeter:Critical Assessments, Vol. II, London: Routledge, 1991.

Elliot, J.E. (1985) ‘Schumpeter’s theory of economic development and social change:Exposition and assessment’, International Journal of Social Economics 12: 6–33.Reprinted in J. Cunningham Wood (ed.), J. A. Schumpeter: Critical Assessments, Vol.IV, London: Routledge, 1991.


Freeman, C. (1993) ‘Schumpeter’s business cycles revisited’ in U. Witt (ed.), Evolu-tionary Economics: 17–38, Aldershot: Edward Elgar.

Hansen, A.H. (1951) ‘Schumpeter’s contribution to business cycle theory’, Review ofEconomic Statistics 33: 129–32. Reprinted in J. Cunningham Wood (ed.), J.A. Schum-peter: Critical Assessments, Vol. I, London: Routledge, 1991.

Kuznets, S. (1940) ‘Schumpeter’s business cycles’, American Economic Review 30, 2:257–71.

Madaràsz, A. (1980) ‘Schumpeter’s theory of economic development’, Acta Oeconomica25: 337–56. Reprinted in J. Cunningham Wood (ed.), J. A. Schumpeter: CriticalAssessments, Vol. III, London: Routledge, 1991.

Perroux, F. (1965) La pensée économique de Joseph Schumpeter, Geneva: Editions Droz.Samuelson, P. A. (1943) ‘Dynamics, statics, and the stationary state’, Review of Economic

Statistics 25: 58–68.Schumpeter, J.A. (1908) Das Wesen und der Hauptinhalt der theorischen Nationalökonomie,

Munich and Leipzig: Duncker und Humblot.—— (1912) Theorie der wirtschaftlichen Entwicklung, Leipzig: Duncker und Humblot.

Preface dated Vienna, July 1911. English translation of the second edition as TheTheory of Economic Development: An Inquiry into Profits, Capital, Credit, Interest, andthe Business Cycle, Harvard University Press: Cambridge, Mass., 1934.

—— (1917/1918) ‘Das Sozialprodukt und die Rechenpfennige: Glossen und Beiträge zurGeldtheorie von heute’, Archiv für Sozialwissenschaft und Sozialpolitik Vol. 44:627–715. Translated from German by A.W. Marget as ‘Money and the socialproduct’, International Economic Papers , vol. 6, 1956.

—— (1936), ‘Review of Keynes’ General Theory of Employment, Interest and Money’,Journal of the American Statistical Association: 791–5.




Staley, C.E. (1986), ‘Schumpeter’s business cycles’, New York Economic Review XVI:300–13.

Witt, U. (1995) ‘Schumpeter vs. Hayek: Two approaches to evolutionary economics’ inG. Meijer (ed.), New Perspectives on Austrian Economics, London: Routledge.

Wolfson, R.J. (1958) ‘The economic dynamics of Joseph Schumpeter’, Economic Develop-ment and Cultural Change, 7: 31–54. Reprinted in J. Cunningham Wood (ed.), J. A.Schumpeter: Critical Assessments, Vol. II, London: Routledge, 1991.


Introduction

Reducing Schumpeter’s contribution to economic thought to the study offactors accounting for the economic dynamism of capitalism – that is, theentrepreneur, innovation and credit – fails to do justice to the much broaderscope of his research agenda which is encapsulated in his view of capitalism as aprocess of historical transformation. Schumpeter always acknowledged his intel-lectual debt to two great exponents of economic thought – Walras and Marx –and it is to the latter that he owes his conception of capitalism as an historicallyspecific set of institutions. Like Marx, Schumpeter believes that the capitalistsystem carries within it the seeds of its own destruction and that it will eventu-ally be replaced by socialism. Schumpeter is, however, primarily interested inthe role played by institutional and cultural factors in this incessant movementtowards destruction, and much of his work is dedicated to highlighting theirsignificance. Thus, his analysis of the long-term perspectives of capitalism,contained mainly in Capitalism, Socialism and Democracy (CSD) provides aperfect opportunity to study the links between the dynamic theory of theprocess of economic evolution, on the one hand, and the analysis of social andcultural disruptions, on the other – both intrinsically linked to one another inthe reality of capitalist development. At the time of the publication of CSD(1942), the possibility of a decline of capitalism and the spectre of socialismwere particularly popular subjects of debate. In developing an analytical frame-work of his own, Schumpeter takes a fairly original stance, distinct from theposition of both the liberals of the day and the proponents of regulation andinterventionism.

The evolution of capitalism: from economic stability toinstitutional instability

Schumpeter believes that the threat to the capitalist system does not stem fromeconomic malfunctions or deficiencies. Quite the opposite, over time capitalismexperiences improved economic performance and increased economic stability.The transition from a situation characterised by a high degree of competition to

7 The long-term perspectiveSchumpeter’s prediction of the end ofcapitalism

Odile Lakomski

one in which monopolies are dominant has only added to this success. At thesame time, however, profound changes of a different nature threaten to under-mine the very foundations of the capitalist economy. These concern the linkbetween the entrepreneur and the banker.

The stability of the economic system versus the instability ofcapitalist order

The theme of the long-term transformation of capitalism appeared inSchumpeter’s writings as early as 1928 in an article entitled ‘The instability ofcapitalism’, which raises the problem as to ‘whether or not [the capitalisticsystem] would … show any tendency towards self-destruction from inherenteconomic causes, or towards out-growing its own frame’ (Schumpeter 1928[1989]: 47).

Before entering into the detail of Schumpeter’s argument, it is essential toclarify the basic concepts on which his analysis is based. First of all, a distinctionmust be drawn between stability and instability within the context of the capi-talist system, and between stability and instability relative to capitalist order. Thefirst notion refers to the set of economic factors that are distinctive of a capitalistsociety, whereas the second concerns the institutions and the set of values onwhich capitalist society is built, as well as its political, social or even natural envi-ronment. It may also be useful to remind ourselves of what exactly Schumpeterincludes in his definition of a capitalist system. Above all, capitalism is a systemundergoing a continuous process of historical transformation. It is characterisedby cyclical dynamics generated by successive waves of innovations penetratingthe sphere of production. Such innovation is the work of talented economicagents, entrepreneurs, who are capable of imagining a very different future and ofovercoming the obstacles arising from society’s natural resistance to technolog-ical upheavals and novelty. Hence, capitalism is driven by an endogenous forcewhich is likely periodically to disrupt economic equilibrium.

To succeed in their innovative efforts, entrepreneurs require the means toact: bank credit is a sine qua non for the evolution of capitalism. Why shouldthis be the case? The fact is that banks are the only agents willing to endowentrepreneurs with the means they require and to manage the risk inherent ininnovation. Banks, in fact, have the power to create money and they canexploit this power within the framework of rules and constraints governing thebanking system, a system overseen by a higher authority, the central bank.Individual bankers, moreover, perform a social role that is crucial to thedynamics of innovation: they evaluate and select the projects suggested to themby entrepreneurs and they supervise the production process.1 Such is the crucialpart played by credit that Schumpeter regards it as one of the defining charac-teristics of a capitalist economic system:

[C]apitalism will be defined by three features of industrial society: privateownership of the physical means of production; private profits and private

The long-term perspective 147

responsibility for losses; and the creation of means of payments – banknotesor deposits – by private banks. The first two features suffice to defineprivate enterprise. But no concept of capitalism can be satisfactory withoutincluding the set of typically capitalist phenomena covered by the third.


It is no accident that credit takes pride of place in Schumpeter’s definition ofthe capitalist system: monetary phenomena play a major role in his theory. Tobe more precise, it is the bonds created between bankers and entrepreneursthrough the credit contract that is the driving force behind change. One couldeven argue that Schumpeter attempted to replace the Marxist view of capitalistdynamics with an approach that puts not exploitation but the lender–debtorrelationship at the heart of the analysis.

To return to our initial question of the long-term perspectives of the institu-tions of capitalism, as far as Schumpeter is concerned, the most importantquestion is that of identifying the endogenous forces at the root of the stabilityor instability of capitalism as a socio-historical formation. Whilst recognisingthat the emergence of capitalism goes back as far as the mid-seventeenthcentury in some countries, and in England in particular, Schumpeter argues thatit was not until the nineteenth century that the ‘competitive’ nature of thingstended to take over. From then onwards, history tells a story of ‘violent fluctua-tions’, even ‘disasters’. If the latter were mainly caused by the logic of capitalismitself, they did not, to begin with, challenge the stability of the overall order ofsociety. Society and the economy functioned such that the disruptions broughtabout by innovation created the conditions for a renewed search for equilib-rium. Therefore, according to Schumpeter, the resulting cyclical fluctuationsdid not accumulate, but rather displayed a tendency to self-correct: ‘We mayphrase the result we reach in our terminology by saying that there is, thoughinstability of the System, no economic instability of the Order’ (Schumpeter1928 [1989]: 70). However, the nature of the problem changed with the emer-gence of ‘trustified’, ‘regulated’ or even ‘organised’ capitalism. As we will arguebelow, monopolies tend to bring about profound changes in the economicsphere that are likely to challenge the foundations of a capitalist system in thelong term. As mentioned above, Schumpeter stresses repeatedly that the insti-tutional instability of capitalism does not emanate from a malfunctioning of itseconomic logic. Contrary to Marx, the idea that capitalism is untenable becauseof the internal contradictions of its economy is absent from Schumpeter’stheory. The capitalist order of things is not doomed by repetitive crises; there isno self-destructive tendency of a purely economic origin. To put it another way,instability of order is not the result of a growing instability in the system:

The Marxist argument which holds that, due to the nature of things, thereis a necessary movement toward the socialist regime, is wrong. The capi-talist system will never collapse by itself. Quite the opposite, it increases ineconomic stability. But capitalism brings about psychological, moral and


political changes. Changes in habits and attitudes which do produce a move-ment toward socialism.

(Schumpeter 1931 [1990]: 403–4)

If Schumpeter’s thesis is already partially explicit in his 1928 article, it is inCSD that he really dedicates all his energy to developing it fully. In a sense,towards the end of his career, Schumpeterian thought reaches its full spectrum;while, at the beginning of the century, it was focused exclusively on the‘essence’ of economic theory (Schumpeter 1908), it experiences a profound re-examination and reversal in 1942 when, rather than abstracting frominstitutional and cultural influences on theoretical economics, Schumpetermakes it a priority to investigate and explore precisely these factors and theirrelevance for the essence of economic theory.

However, before going into more detail, we first need to have a closer look atSchumpeter’s analysis of the growing stability of the capitalist system.

Monopoly prices: a two-sided weapon

One of the main arguments Schumpeter puts forward to explain the success ofcapitalism refers to the increasing market power of large companies and theparallel development towards high levels of concentration, prompting the emer-gence of monopolistic practises in industrial sectors.

The analysis underlying this argument has, in part, a critical dimension inthat it is based on a critique of the conclusions drawn from the theoreticalmodel of perfect competition. Schumpeter disagrees with the widely held beliefat the time that any restrictions on a perfectly competitive organisation ofmarkets is likely to undermine the efficiency of the system.2 In this view,monopolies are synonymous with waste, price increases, a reduction in thevolume of output and the existence of lasting super-profits, all resulting inwelfare losses for society. Schumpeter’s critique of this line of argument is basedon his rejection of the validity of the hypotheses underlying the model ofperfect competition. Essentially, he argues that the results obtained within thisframework are based on static analysis. However, as soon as one adopts thedynamic perspective of Schumpeterian development, the properties of themodel of perfect competition become meaningless, and the anti-trust measuresderived from them are called into doubt. The main idea defended bySchumpeter is that, in the context of an economy experiencing waves of inno-vation, the nature of the problem is reversed in the sense that the competitivemodel, up to this point referred to as the ‘normal’ case, should henceforth beconsigned to the class of exceptions (Schumpeter 1950a: 78–9).

However, not only will monopolistic practises become the rule but theywill be at the root of capitalist expansion: ‘A shocking suspicion dawns uponus that big business may have had more to do with creating that standard oflife than with keeping it down’ (ibid.: 82). Consequently, the kind ofbehaviour that would have been criticised by conventional economic theory


on the grounds that it restrains competition, is given a certain legitimacywithin the Schumpeterian framework: ‘There is no more of paradox in thisthan in saying that motorcars are travelling faster than they otherwise wouldbecause they are provided with brakes’ (ibid.: 88). In fact, monopolisticbehaviour can be explained by the very nature of innovation, the major forcebehind capitalism. Schumpeter argues that any innovative project representsa long-term investment characterised by a considerable degree of risk. Theuncertainty regarding the outcome of such projects must be offset by theentrepreneur’s profit expectations. Furthermore, taking into account thecyclical nature of evolution, the period over which profits can be realisedmust be long enough to allow entrepreneurs to cope with difficult times, suchas recessions or depression (ibid.: 88–92). Innovation, thus, requires measuresdesigned to protect innovators from the threat of competition as well as fromthe general uncertainties of their trade (insurance, arbitration) and to enablethem to maintain their technological advantage (patents). Schumpeter evenhints at procedures aimed at ‘discouraging’ or ‘checkmating’ competitors(ibid.: 89).

Such practices are not only the logical price to be paid for risk-taking in anuncertain environment, but they also guarantee stability to entrepreneurs whomeet with temporary difficulties. Trade restrictions, cartels, tacit agreements onpricing, all these are efficient ways to cope with transitory difficulties and evento ‘steady the ship’ (ibid.: 87). In the same vein, Schumpeter defends the logic ofprice rigidities in general as based on the same need to preserve a position in themarket. Underlying his argument are doubts as to the regulatory capacity ofperfectly flexible prices. Schumpeter believes that, in a dynamic context, suchflexibility could accentuate the instability inherent in periods of depression: ‘Inother words, under the conditions created by capitalist evolution, perfect anduniversal flexibility of prices might in depression further unstabilize the system,instead of stabilizing it as it no doubt would under the conditions envisaged bygeneral theory’ (ibid.: 95). As a matter of fact, Schumpeter is intent on arguingthat any systematic condemnation, such as that of monopolistic practices bystandard competition theory, is unfounded and that the time has come to ‘refutea prevalent theory and the inferences drawn therefrom about the relationbetween modern capitalism and the development of total output’ (ibid.: 91–2).His final conclusion speaks for itself:

Thus it is not sufficient to argue that because perfect competition is impos-sible under modern industrial conditions – or because it always has beenimpossible – the large-scale establishment or unit of control must beaccepted as a necessary evil inseparable from the economic progress …What we have got to accept is that it has come to be the most powerfulengine of that progress and in particular of the long-run expansion of totaloutput.

(ibid.: 106)


Increased industrial concentration and market power of large enterprises is thusseen as resulting from the way in which economic agents adapt to innovationand, thus, to the outcomes of capitalist evolution. This development is rein-forced by an additional factor, namely, the alliance between banks and industry.Schumpeter here follows the train of thought developed by Hilferding in hisFinancial Capital (1910), according to which banks tend to assert their controlover a large part of industry and to organise it into monopolies, thus stabilisingcapitalism. This link between banks and industry can also be explained in termsof the logic of the evolution of capitalism. With free entry, a company will findit much easier to impose itself on the market if it can be sure of firm backing bythe bank providing it with the funds required to achieve its objectives. Thisargument is corroborated by some contributions to the contemporary theory offinancial intermediation which argue that the long-term relationship betweenbanks and companies is central not only to the financing of innovations butalso to the success of an innovative project. A bank’s overall evaluation of acompany acts as a guarantee of its productive efficiency, thus determining thelatter’s status as a debtor. At the same time, a bank’s willingness to grant creditsto a company that has fallen on hard times will increase with the existence of along-term relationship based on trust. Hence, close links between the bank andthe industrial sector are likely to result in an improved adjustment of supply anddemand for credit, increased control over (financial) capital as well as a greaterroom for manoeuvre on the part of companies. Moreover, this alliance also leadsto greater overall stability as, on the one hand, entrepreneurs can be reasonablysure of obtaining the necessary funds to continue their projects, while, on theother, banks have greater control over the risk of defaulting since their long-standing relationship with individual firms provides them with valuableinformation about their clientele. Hence, as Schumpeter points out, ‘the poli-cies of high finance are based on control of a large proportion of the nationalcapital’ (Schumpeter 1951: 107).

However, this alliance also functions as a two-sided weapon in a capitalisteconomy. To the extent that bank credit becomes a sine qua non for the emer-gence of innovation, the driving force of capitalist evolution becomes directlydependent on the way in which banks manage their social role of evaluatingand selecting innovators. In Business Cycles (BC), Schumpeter insists that amajor premise for this sort of alliance to work is that the political and industrialsphere be totally independent from the interests of the banking sector(Schumpeter 1939: 92). This will not be the case where a monopolisticallyorganised economy relies on a hierarchical banking system dominated by a fewbanks. In this case, if the central bank is directly controlled by the state, thelatter effectively gains control over a powerful tool to control the entire produc-tive machinery and the market economy is heading towards socialism. AsSchumpeter puts it:

Everywhere except, significantly, in England, there has come into being aclose alliance between high finance and the cartel magnates, often going as


far as personal identity. Although the relation between capitalists andentrepreneurs is one of the typical and fundamental conflicts of the capi-talist economy, monopoly capitalism has virtually fused the big banks andcartels into one. Leading bankers are often leaders of the national economy.Here capitalism has found a central organ that supplants its automatism byconscious decisions.


More fundamentally, monopolistic practises and the confusion of interestsinherent in the alliances sustaining these practices tend to undermine the pillarof capitalist development, namely, the provision of credit-capital based on adirect relation of trust between individual banks and entrepreneurs. This rela-tionship is increasingly being replaced by a much more centralised organisationdependent on the sphere of politics. To put it another way, the alliance betweenbanks and industry is the first step towards socialism.

The institutional decline of capitalism: an anti-heroiccivilisation

Schumpeter argues that, while monopolies contribute, to a certain extent, tothe success of the capitalist economy, they are also somewhat of a Damocles’sword to capitalism in that they are vulnerable to take over by the state, thuseffectively creating socialism. In order to grasp fully Schumpeter’s view of thethreat to the future of capitalism, we need to take a closer look at the complexrelationship between the economic system, on the one hand, and the institu-tions maintaining the order of a capitalist society, on the other.

The rationalisation of the human mind

Schumpeter’s notion of capitalist order is best understood through the conceptof civilisation. Capitalism is not defined merely by a specific mode of economicproduction, but also by a mentality, a set of values and attitudes that structurethe life of individual members of society.

The capitalist era has been marked by a process of ‘rationalisation’ of bothindividual behaviour and of societal forms of reasoning, or what Schumpeterrefers to as the human mind in general. According to Schumpeter, this is theresult of a gradual expansion of the economic sphere, in the course of whichrational thought has developed and turned into a generalised paradigm. It isunder the pressures emanating from economic life that individuals have foundthemselves faced with the necessity to employ logical reasoning: ‘They wererationalized, because the instability of their economic position made theirsurvival hinge on continual, deliberately rationalistic decisions’ (ibid.: 89; also1950a: 122). This form of reasoning is, above all, driven by the monetarydimension of economic calculation: capitalism ‘exalts the monetary unit … intoa unit of account’; that is to say, ‘capitalist practice turns the unit of money into


a tool of rational cost-profit calculations’. This ‘cost-profit calculus in turnreacts upon the rationality; by crystallising and defining numerically, it power-fully propels the logic of enterprise’ (Schumpeter 1950a: 123). Thisrationalisation of the mind subsequently spreads to the social sphere as a whole,a characteristic of modern society described in great detail by Max Weber.There are two implications of this pre-eminence of logical reasoning. First, itestablishes a reign of ‘free thinking’, meaning ‘materialistic monism, laicism andpragmatic acceptance of the world this side of the grave’ (ibid.). To the extentthat individuals no longer regard societal organisation as derived from divineinfluences, rational thought tends to ‘exorcise the spirits’, that is ‘metaphysicalbeliefs’, ‘magic’ and ‘romantic and mystical notions of any nature’. Thistendency gives rise to what Schumpeter calls an ‘anti-heroic’ civilisation char-acterised, in particular, by the brandishing of a model of social success thatreplaces the image of the feudal knight with that of a company director rakingin profits (ibid.). Second, by gradually permeating all spheres of social life, ratio-nalism also engenders a critical and inquisitive attitude towards ‘the mass ofcollective ideas’ or, more crucially, towards the established order (ibid.: 122; seealso Schumpeter 1954: 114). Schumpeter’s explanation of this type of behaviourrefers to two main factors: on the one hand, utilitarian conceptions regardingthe betterment of mankind’s lot tend to establish themselves firmly in the mindof individuals; on the other hand, the ‘rationalization of the soul rubs off all theglamour of super-empirical sanction from every species of classwise rights’(Schumpeter 1950a: 127). These two factors, together with, and reinforced by, aparticular enthusiasm for ‘Efficiency’ and ‘Service’, typically expressed by indi-viduals in a capitalist society, breed what Schumpeter calls ‘that “will” withinthe bourgeoisie itself ’ (ibid.). These elements, and the particular ‘will’ theycreate from within the bourgeoisie, constitute an institutional dynamics whichgives rise to a powerful force of disintegration of bourgeois mentality, precipi-tating the downfall of the capitalist order. Hence, ‘[c]apitalism, whilsteconomically stable, and even gaining in stability, creates, by rationalising thehuman mind, a mentality and a style of life incompatible with its own funda-mental conditions, motives and social institutions’ (Schumpeter 1928 [1989]:71–2).

The disappearance of the entrepreneur as a driving force ofcapitalism

Schumpeter defines capitalism as an economic system that is constantly revolu-tionised through the innovative efforts of particularly gifted economic agents,the entrepreneurs. However, the expansion of monopolistic organisationthreatens the very nature of this source of permanent renewal. Progress becomesincreasingly automatised, undermining the role of leadership. According toSchumpeter, this phenomenon may ‘affect entrepreneurship and capitalistsociety nearly as much as the cessation of economic progress would’(Schumpeter 1950a: 131). The reason is twofold: first, by qualifying the role


played by private ownership and by imposing bureaucratic procedures, largecompanies tend to interpose a distance between innovative activity and thepersonality of an individual. Secondly, this effective erosion of the purpose ofthe individual entrepreneur has serious repercussions for the social or classstructure of a capitalist society in that it undermines the role of the bourgeoisie.

Let us consider the first point in more detail. If, as we have previously shown,the increase in company size provides the economic system with stability andefficiency, it nevertheless promotes a bureaucratic form of organisation whichradically modifies the nature of innovation. An increasingly bureaucratic formof organisation implies that innovative projects are more and more entrusted toteams of specialists (e.g. R&D departments) working to order and whose testedmethods make the outcome of research efforts foreseeable. Consequently,

[r]ationalized and specialized office work will eventually blot out person-ality, the calculable result, the ‘vision’. The leading man no longer has theopportunity to fling himself into the fray. He is becoming just anotheroffice worker – and one who is not always difficult to replace.

(ibid.: 133)

Hence, the rationalisation of the production process effectively does away withthe need for imagination that was the main asset of the entrepreneur, whoseeffort also went hand in hand with the radical uncertainty characterising theprocess of innovation. In an economy of trusts, it is as if large enterprises, basedon the increasingly large powers of calculation at their disposal, were suddenlyable to control the hazards of the future to the point of controlling the marketsin which they operate. However, even if we admit that bureaucratisation elimi-nates the need for Schumpeterian entrepreneurs, it is difficult to believe thatthe act of innovation would lose its destabilising nature. Be it the individualentrepreneur who introduces a new technology, or an R&D department, onewonders whether there would not remain the same level of disturbance associ-ated with novelty for society at large. The mere fact that a technology isdeveloped within large structures does not remove uncertainty and the possi-bility of failure which always surrounds innovative projects. Even if thetendency towards bureaucratisation were to turn innovation into a routineactivity, why should this necessarily lead to the decline of capitalism? After all, itwould be equally plausible to assume that such bureaucratisation would be char-acterised by a stationary process instead.

It seems that what we are faced with is a certain element of romanticism inSchumpeter’s critique that leads him to perceive of capitalism as dependent on‘heroic’ behaviour and of its end as brought about by the impersonal nature oflarge enterprises. There can be little doubt as to Schumpeter’s thoughts on thissubject: ‘The romance of earlier commercial adventure is rapidly wearing away,because so many more things can be strictly calculated that had of old to bevisualized in a flash of genius’ (ibid.: 132). Shareholding and the legal structureof publicly limited companies complete the picture of the symbolic separation


of man from their enterprises, undermining the concept of private propertyrights and with it one of the most important pillars of capitalism:

The capitalist process, by substituting a mere parcel of shares for the wallsof and the machines in a factory, takes the life out of the idea of property. Itloosens the grip that once was so strong … the grip also in the sense thatthe holder of the title loses the will to fight, economically, physically, polit-ically, for ‘his’ factory and his control over it, to die if necessary on its steps.

(ibid.: 142)

As a result of this process, the entrepreneur ceases to identify with his firm andthe success of innovation ceases to be linked to the individual’s desire to winand to climb the social ladder (Schumpeter 1928 [1989]: 71). Theentrepreneurial function is, of course, not the only area of social and economiclife to suffer the consequences of rationalisation. Schumpeter also refers tobanking, central to the development process, as succumbing to the phenomenonof bureaucratisation:

In the past, banking was the sector where personal factors, the trust oneinspired or the initiative one showed, were more important than anywhereelse. This type of banker … helped create modern industry especially viacredit overdrafts: this consisted in selecting strong personalities and placingthe means of production within the grasp of any talented but pennilessman. Nowadays this function has not ceased to exist but, more often thannot, has passed from the hands of private bankers to the directors ofmodern banks. Moreover, banking operations as such … have developed awell-known and well-established technique which may be learned, andwhich has often led to a bureaucratic numbness.

(Schumpeter 1931 [1990]: 426)

In fact, it is here that Schumpeter’s argument takes on its full significance. Inaffecting both the nature of the entrepreneurial function as well as the role ofthe banker, the expansion of large enterprises radically alters what was thedriving force of capitalism: that is, the granting of credit as a means of evalu-ating and selecting economic agents and innovative projects.

As for the second aspect of Schumpeter’s analysis of the disappearance of theentrepreneur, this emphasises the more strictly sociological implications of theprocess, and, in particular, its effect on the class structure of capitalist society.According to Schumpeter, although entrepreneurs do not form a social class oftheir own, they are a vital component of the bourgeoisie which depends ontheir dynamic influence on the system. Hence, the position of the whole of thebourgeoisie is greatly affected by their disappearance: ‘Economically and socio-logically, directly and indirectly, the bourgeoisie therefore depends on theentrepreneur and, as a class, lives and will die with him’ (Schumpeter 1950a:134). In fact, by helping to eliminate the individual entrepreneur, ‘[t]he


perfectly bureaucratized giant industrial unit not only ousts the small ormedium-sized firm and “expropriates” its owners, but in the end it also ousts theentrepreneur and expropriates the bourgeoisie as a class’ (ibid.).

Challenging the legitimacy of capitalist institutions

To recall, Schumpeter’s concept of capitalism is that of a dynamic processprompted by waves of innovation which periodically disrupt productive activi-ties and flood the markets with new products. This being the case, capitalismpossesses a rather special property: that of ‘creative destruction’ which elimi-nates obsolete elements so as to keep only those seen to be the most efficient.The results of entrepreneurial actions are beneficial to society as a whole asliving standards rise. However, considering the way in which capitalist develop-ment takes place, these beneficial effects can only be appreciated in the longterm: there is no point in attempting to evaluate output in the short term.

It is here that the real problem emerges because support for capitalismpresupposes an impossible squaring of the circle:

The success of the indictment becomes quite understandable as soon as werealize what acceptance of the case for capitalism would imply. That case… could never be made simple. People at large would have to be possessedof an insight and a power of analysis which are altogether beyond them.

(ibid.: 144)

Only the detachment of the theorist and his or her power of abstraction allowsa full appreciation of the phenomenon. Schumpeter seems to phrase theproblem in terms of the legitimacy of capitalist rule: how can one expect thevictims of depressions and all those whose positions are constantly threatenedby the devastation caused by competition to be persuaded that capitalism will,in the end, improve societal standards at large? Thus,

rational recognition of the economic performance of capitalism and of thehopes it holds for the future would require an almost impossible moral featby the have not. That performance stands out only if we take a long-runview; any pro-capitalist argument must rest on long-run considerations. Inthe short run, it is profit and inefficiencies that dominate the picture.

(ibid.: 145)

Consequently, the success of anti-capitalist activities is easily understandable asthe fruit of progress can only be harvested at the cost of fear, suffering anddiscontent brought about by the dislocation of industrial structures and theimmediate consequence of introducing innovations in general (Schumpeter1954: 760). These costs, even though they were only the superficial repercus-sions of the transitional process of adaptation between two phases of prosperity,cannot be accepted by those who find themselves in the direct line of fire, be


this through bankruptcy, unemployment or the permanent threat of unemploy-ment (ibid.). All of these factors are likely to offset the positive effects thatchange may bring and they explain that ‘[f]or the masses, it is the short-runview that counts. Like Louis XV, they feel après nous le déluge, and from thestandpoint of individualist utilitarianism they are of course being perfectlyrational if they feel like that’ (Schumpeter 1950a: 145).

Hence, what needs to be understood is that the capitalist system will find itdifficult to resist the pressures of protest movements, especially those emanatingfrom those parts of society who feel threatened by the process of creativedestruction. Their power is further strengthened by a new interest group – thatconstituted of intellectuals. These pure products of capitalism add to socialprotests by defending positions hostile to the economic logic of capitalism.

Why, however, should it be that economic change in a capitalist society isundermined by relatively marginalised groups? Or, to put it another way, whyshould such protests threaten capitalist order? The nub of the question is this:the sacrifices imposed on parts of society in the process of the evolution of thecapitalist system are incompatible with the basic principles of an individualisticsociety proclaiming as it does the twin ideals of democracy and equality. Hence,to understand what is at stake, one needs to understand the inner contradic-tions of the political institutions on which capitalist order is built.

Capitalism and democracy: an inner contradiction

The examination of the political dimension of capitalism is a crucial element inSchumpeter’s analysis of its historical evolution. In a way, political factors arethe cornerstone of his argument, while at the same time forming a centralaspect in his overall consideration of the influence of institutions on the work-ings of the capitalist system. Schumpeter argues that, true to its own logic,modern society, has created a political regime – democracy – which rapidlyproves to be incompatible with its economic foundations.

Democracy or the perverse effect of a belief in ‘the common good’

The final element of Schumpeter’s argument is based on a strong criticism ofthe ‘democratic ideal’. Essentially, he regards this ideal as no more than a ratherpernicious political ‘trick’ that ends up threatening capitalist order.

Schumpeter basically makes two related points. First, the establishment ofdemocracy implies that the bourgeoisie will take control of public functionsformerly in the hands of an aristocratic elite. As a matter of fact, this step goes along way in explaining the self-destructive tendencies inherent in the set ofvalues and institutions established by the bourgeoisie. In Schumpeter’s opinion,the early capitalist organisation of society was synonymous with a political andsocial equilibrium based on the symbiosis of two classes: while the bourgeoisieenlisted the support of the feudal class (ibid.: 135–6), the aristocracy was able toprolong its hold on political power until the end of competitive capitalism.


But sooner or later, the effect of the rationalisation of the human mind wasto turn this aristocratic regime into an obstacle to capitalist expansion.However, in Schumpeter’s view, the bourgeoisie’s objective of political emanci-pation was not accompanied by a parallel development of its skills in the artof governing. Hence, Schumpeter argues that ‘[i]n breaking down the pre-capitalist framework of society, capitalism thus broke not only the barriers thatimpeded its progress but also flying buttresses that prevented its collapse’ (ibid.:139). Once again, an element of romanticism enters into Schumpeter’s argu-ment: the bourgeois industrialist and trader – fundamentally ‘anti-heroic’ and‘rationalistic’ – cannot lay claim to the same power as the feudal lord who isbacked by a ‘quasi-mystical prestige’ and ‘is used to giving orders and beingobeyed’. Equipped with only his economic performance, the bourgeois is notsufficiently qualified to take on public responsibilities. Hence Schumpeter’sconclusion: ‘[W]ithout some protection by some non-bourgeois group, the bour-geoisie is politically helpless and unable not only to lead its nation but even totake care of its particular class interest. Which amounts to saying that it needs amaster’ (ibid.: 138).

Second, Schumpeter criticises the elevation of the democratic regime to therank of a supreme ideal, and almost to the status of a religion (ibid.: 265). Aboveall, he reminds us of the fact that democracy has two meanings: on the onehand, no form of societal organisation can go beyond democracy. On the other,democracy is simply a political method.3 However, what Schumpeter has in mindis not so much the method itself, but the ‘classical doctrine’ on which it isbased. Having its roots in a naive utilitarian philosophy (inherited fromBentham and Rousseau, in particular), the idea of democracy is based on twoelements:

1 the possibility of realising the individual quest for happiness at the level ofthe collective;

2 the endowment of each individual with the means to achieve this happi-ness or, more generally, its goals.

Schumpeter sets out to question the foundations of this doctrine by challengingthe following three principles:

1 the possibility of defining a common good accessible to one and all andwhich could form the basis of a collective agreement;

2 the possibility of public authorities giving a valid expression to thecommon will, which would also be a true formal representation of actualindividual desires;

3 the possibility of the common will corresponding to the common good(ibid.: 250–1).

To be precise, the distrust Schumpeter displays for democracy as a politicalmethod is fuelled by his deep-rooted conviction that individuals are unable to


behave rationally as soon as they act collectively. The picture Schumpeterpaints of the average citizen is indeed explicit:

Thus the typical citizen drops down to a lower level of mental performanceas soon as he enters the political field. He argues and analyzes in a waywhich he would readily recognize as infantile within the sphere of its realinterests. He becomes a primitive again.

(ibid.: 262)

Schumpeter’s critique is thus mainly aimed at the denunciation of a naive viewof human nature, which he regards as a consequence of utilitarianism. In hisview, a more appropriate perspective on the political method of democracy isthat of a competitive struggle between politicians for access to political power(ibid.: 269). Their success will essentially depend on their ability to win over asufficiently large part of the electorate. From this point of view, the efficiency ofbourgeois democracy as well as its long-term viability, essential to preserving thelegitimacy of the capitalist order, becomes doubtful:

The incessant competitive struggle to get into office or to stay in it impartsto every consideration of policies and measures the bias so admirablyexpressed by the phrase about ‘dealing in votes’. The fact that in a democ-racy government must attend primarily to the political values of a policy ora bill or an administrative act … is likely to distort all the pro’s and con’s.In particular, it forces upon the men at or near the helm a short-run viewand makes it extremely difficult for them to serve such long-run interests ofthe nation.

(ibid.: 287)

Therefore, a further complication, highlighting the perverse relation developingbetween the economic logic and the political regime of capitalism, has to betaken into account. As a political career depends upon the politician’s ability toattract votes, and given the growing dissatisfaction of the general public withthe drawbacks of economic change, there is scope for the introduction ofeconomic policies aimed at minimising the negative effects of economic cycles.This logic, inherent to capitalist society, is likely to increase the involvement ofthe state in matters of the economy which, in turn, may well prove to under-mine the capitalist order in the long term.

A symptomatic development: increasing state involvement ineconomic life

Following the logic of Schumpeter’s approach, the advocacy of interventionisteconomic policies based on Keynes’s arguments is symptomatic of an imminent‘march into socialism’. This tendency is evident in the emerging gulf betweenpublic opinion and bourgeois values which underpin capitalist order and


economics. In many ways, the pre-eminence of Keynesian ideas which markedthe post-war era is an essential part of the argument contained in CSD.

It is the combination of the three elements of Schumpeter’s overall argu-ment, discussed above, on which his prediction is based: the nature of thepolitical game underlying the democratic regime (element 1) will be reinforced,on the one hand, by the egalitarian ideal imprinted on the minds of individualsin the process of the rationalisation of society (element 2) and by growingprotest movements, fuelled by the insecurities arising from permanent change,on the other (element 3). More specifically, it is through the meeting of theshort-term interests of both the politician and the average citizen that the legit-imacy of the capitalist regime is threatened. As Schumpeter puts it, ‘[i]n orderto identify himself with the capitalist system, the unemployed of today wouldhave completely to forget his personal fate and the politician of today hispersonal ambition’ (ibid.: 145). These social factors go a long way in justifyingincreased political and state intervention into the economy aimed at fullemployment. Thus, a rising standard of living and a new perception of the roleof government are just as important in explaining the state of capitalist societyand order as the threat of unemployment. For Schumpeter, these factors alsolargely explain the popularity of social reform projects, the tendency of industryto reorganise itself (especially into cartels) and the expansion of governmentactivity (1954: 760). From a theoretical point of view, he also notes that ‘thealliance between economics and liberalism … was broken’ (ibid.: 766).

This would explain the popularity of Keynesian views and policy recommen-dations on whose harmful implications Schumpeter comments in particular inhis remarks on the Tract on Monetary Reform (Schumpeter 1925: 312–20). It ishere that the disagreement between Keynes and Schumpeter is the mostapparent. Whereas the Cambridge economist regards both state interventionand the monitoring of entrepreneurial behaviour as the only solution to thesurvival of capitalism, the Austrian economist emphasises the possibilitiesinherent in this development of the emergence of anti-capitalist elements andmovements. For Schumpeter, short-term considerations are inherently incom-patible with the economic logic of capitalism, whose net impact on society canonly be observed over the long term. Hence, in Keynes’s Tract on MonetaryReform Schumpeter detects much more than a simple programme of monetaryreform. He is acutely aware of the likely long-term implications of the centralbank and private banks being assigned tasks akin to those of the centralmanagers of a planned economy. In his view, Keynes may be seen to havecontributed to the exploration of an appropriate method to transform capitalisminto a new social organisation, thereby opening the door to ‘serious socialism’(ibid.: 328). But what Schumpeter condemns, above all, is the fact thatKeynesian arguments, derived from The General Theory, have made the goal ofserving the common good respectable even outside socialist tendencies, therebyencouraging modern governments to adopt increasingly anti-capitalist measureswhich Keynes himself would most probably not have been prepared to accept(Smithies 1951: 164). It is mainly in the context of this particular development


that Schumpeter develops his already mentioned critique of the democraticideal.

‘The march into socialism’ has begun: the values of capitalist society losetheir grip both on public opinion and on the capitalist class itself. It is at thispoint that Schumpeter stresses the ease with which businessmen and post-wareconomists – who declare their fervent opposition to socialism, denying anytendency whatsoever towards such a regime – nevertheless approve of policyrecommendations based on Keynesian theory (Schumpeter 1950b: 424).4 Thisattitude exemplifies the disintegration of bourgeois mentality to the extent thatit becomes possible ‘to develop and regulate capitalist institutions as to condi-tion the working of private enterprise in a manner that differs but little fromgenuinely socialist planning’ (ibid.: 424–5). According to Schumpeter, thisprocess is irreversible. On this point, he tends to be in agreement with the posi-tion of the Austrian school, and with Hayek and Mises in particular, who arguethat economic policy agendas create the conditions for increased state involve-ment in the economy. In the long term, this cumulative and self-propellingprocess can only culminate in the replacement of the market economy by asocialist form of societal organisation (Mises 1949: 753). However, their adher-ence to the principles of liberalism leads authors such as Hayek or Mises tosuggest ways of slowing down this development and of restoring a pure marketeconomy. To put it another way, they believe that there is still time to savecapitalism from socialism.

Schumpeter’s position is different. First of all, he is careful to steer clear of allnormative considerations:

I do wish … to preclude any interpretation that I regard capitalism as thefinal phase of social evolution, as something that exists of natural necessity,that cannot be adequately explained. Still less do I regard it as an ideal inany sense.

(Schumpeter 1951: 108 fn.)

Secondly, he shares with Marx the strong belief that competitive capitalismmust necessarily be supplanted by another form of societal organisation. In hisview, it is vain to try and oppose the relentless force of capitalist developmentthrough a defence of liberalism, for growing interventionism is fully part of thelogic of capitalism:

I have pointed out before that social legislation or, more generally, institu-tional change for the benefit of the masses is not simply something whichhas been forced upon capitalist society by an ineluctable necessity to alle-viate the ever-deepening misery of the poor but that, besides raising thestandard of living of the masses by virtue of its automatic effects, the capi-talist process also provided for that legislation the means and the will.

(Schumpeter 1950a: 127)


To sum up, we can illustrate Schumpeter’s argument using the followingdiagram:

Conclusion

As a student of Böhm-Bawerk and Wieser and a member of the second-generation Austrian school, Schumpeter was one of the leading economists ofhis time. As a distinguished representative of the Austrian tradition, he must becredited not only with having provided a view of the economy as a dynamicprocess, but furthermore with having integrated a general theory of economicchange and a penetrating analysis of the role of institutional factors. Thus,Schumpeter’s contribution and, in particular, the way in which he highlightedthe crucial role of institutions, must not be forgotten, especially at a time wheneconomic theory is paying renewed attention to institutionalism.

In this chapter, we have stressed that beyond his Austrian heritage,Schumpeter developed an original view of the capitalist process that promptedhim to emphasise the role of institutions. We can summarise his view at twolevels of the discussion. First, we referred to his definition of capitalism. ForSchumpeter, capitalism is a process of discontinuous accumulation and change,both inextricably linked to one another. However, in modern capitalism thedriving force of change cannot be reduced to the creative impulse of theentrepreneur. Rather, the latter must be understood in relation to the roleplayed by credit and by the rules and institutions created by the banking system.The key role of this system is to select innovative projects. Hence, the monetarysphere is responsible for two features of the capitalist system, disequilibrium and


Figure 7.1 Institutions and instability of the capitalist order

instability, which cannot be eliminated without, at the same time, eliminatingcapitalism itself.

Second, we emphasised that for Schumpeter, economic analysis consists ofthe combination of analytical, historical and statistical tools. The main thesis ofCapitalism, Socialism and Democracy illustrates perfectly how Schumpeter’sevolutionary approach is inseparable from historical analysis, something helearned from Marx. And from Marx he also gained a dialectical perspective thatcaused him to search for the seeds of destruction in capitalism as an order. ButSchumpeter’s argument also contains a refutation of Marxian views. He, in fact,reverses the subordination of the ‘superstructure’ to the ‘basis’: it is not theeconomic crises and contradictions that threaten to bring capitalist civilisationto its knees, but the crises and contradictions in its socio-psychological super-structure. That is why, in Schumpeter’s view, institutions cannot be regarded asneutral factors.

Notes1 For further details of Schumpeter’s theory of credit and banking, see Schumpeter

1970 and 1996. Elsewhere, we suggest an interpretation of Schumpeter’s view of thecapitalist process as based on monetary factors. In particular, we suggest that it canbe viewed as an attempt to define the capitalist economy through the conflictualrelation between banks and entrepreneurs, representing the central element ofevolution (O. Lakomski, Monnaie, Banques et Crédit dans l’Oeuvre de J.A.Schumpeter, PhD dissertation, University of Picardie – Jules Verne, 1999).

2 Schumpeter often emphasises the hostility of economists and business towards largeenterprises, particularly in the United States where anti-trust laws have played animportant role (Schumpeter 1950a: 98–100).

3 That is, a type of institutional organisation whose purpose is to arrive at politicaldecisions (Schumpeter 1950a: 242).

4 Especially: ‘(1) the various stabilization policies which are to prevent recessions or atleast depressions, that is, a large amount of public management of business situationseven if not the principle of full employment; (2) the “desirability of greater equalityof incomes” … (3) a rich assortment of regulative measures … (4) public control …over the labor and the money market; (5) indefinite extension of the sphere of wantsthat are, now or eventually, to be satisfied by public enterprise … and (6) of courseall types of security legislation’ (Schumpeter 1950b: 424).

Bibliography

De Vecchi, N. (1995) Entrepreneurs, Institutions and Economic Change: The EconomicThought of J.A. Schumpeter (1905–1925), Aldershot: Edward Elgar.

Hilferding, R. (1910) Finance Capital: A Study of the Latest Phase of Capitalist Development,introduction by T. Bottomore (ed.), London: Routledge and Kegan, 1981.

Mises, L. von (1949) Human Action. A Treatise in Economics, London: Hodge.Schumpeter, J.A. (1908) Das Wesen und der Hauptinhalt der theoretischen Nation-

alökonomie, Munich and Leipzig: Duncker and Humblot.—— (1925) ‘Kreditkontrolle’, Archiv für Sozialwissenschaft und Sozialpolitik, 54: 289–328.—— (1928) ‘The instability of capitalism’, Economic Journal 38, September, II: 361–86.

Reprinted in J.A. Schumpeter (1989).


—— (1931) ‘Les possibilités actuelles du socialisme’, L’Année Politique Française etEtrangère 24: 385–418. Reprinted in Schumpeter (1990).

—— (1939) Business Cycles. A Theoretical, Historical, and Statistical Analysis of the Capi-talist Process, 2 vols, New York and London: McGraw-Hill.

—— (1943) ‘Capitalism in the postwar world’, in S.E. Harris (ed.) Postwar EconomicProblems, New York and London: McGraw-Hill.

—— (1950a) Capitalism, Socialism and Democracy, 3rd edn, London: Allen & Unwin; 1stedn, New York: Harper, 1942; 4th edn (with a new chapter, ‘The march intosocialism’), London: Allen & Unwin, 1952. Reprinted (with a new introduction byR. Swedberg), London: Routledge, 1992.

—— (1950b) ‘The march into socialism’, American Economic Review 40, 2, May: 446–56.—— (1951) Imperialism and Social Classes, (ed. P. Sweezy), New York: Augustus M.

Kelley.—— (1954) History of Economic Analysis, London: Allen & Unwin. Reprinted, London:

Routledge, 1994.—— (1970) Das Wesen des Geldes, introduction by F.K. Mann (ed.), Göttingen: Vanden-

hoeck & Ruprecht.—— (1989) Essays on Entrepreneurs, Innovations, Business Cycles and the Evolution of

Capitalism (ed. R.V. Clemence), New Brunswick and Oxford: Transaction Publishers.Originally published in 1951 by Addison-Wesley.

—— (1990) Capitalisme, Socialisme et Démocratie, Paris: Payot (2nd edn of French trans-lation of J.A. Schumpeter, Capitalism, Socialism and Democracy, New York: Harper,1942).

—— (1996) Trattato della Moneta – Capitoli inediti (eds L. Berti and M. Messori), Naples:Edizioni Scientifiche Italiane.

Simmel, G. (1907) Philosophie des Geldes, 2nd edn, Munich: Duncker & Humblot.Smithies, A. (1951) ‘Schumpeter and Keynes’, Review of Economic Statistics, vol. 33,

May: 163–9.Weber, M. (1930) The Protestant Ethic and the Spirit of Capitalism, English translation by

Talcott Parsons, Foreword by R.H. Tawney, New York: Charles Scribner.


Part IV

Entrepreneurship andcompetition

Schumpeter’s approach to entrepreneurship has been the topic of numerouscontributions. Rather than providing one more exhaustive account ofSchumpeter’s view of entrepreneurship, this chapter concentrates on ways tointerpret his approach from the institutionalist perspective described in theintroduction. This implies that we will not trace the evolution of Schumpeter’stheory of the entrepreneur from The Theory of Economic Development (TED) toCapitalism, Socialism and Democracy (CSD), as is usually done. Instead, we willfocus on the question of whether this theory can be understood as an importantapplication of his methodological approach of combining history with economictheory, using economic sociology as the intermediary. We will argue that thismethodological approach was instrumental in constructing a version ofeconomic dynamics that links economic development and institutional change.This differs from standard interpretations of Schumpeter’s theory ofentrepreneurship that emphasise either an evolutionary selection view ofentrepreneurial competition or the so-called ‘Schumpeterian hypothesis’.1

Entrepreneurs and social leadership

Following the perspective developed in the introduction, our point of departurehere is simple: entrepreneurship is no more than a specific form of social leader-ship. Schumpeter defines social leadership as a ‘special function’ that entails ‘todecide, to command, to prevail, to advance’ (Schumpeter 1927 [1951]: 217).2

This function is passing in nature since ‘it emerges only with respect to ever newindividual and social situations and would never exist if individual and nationallife always ran its course in the same way and by the same routine’ (ibid.). Putdifferently, Schumpeter defines social leadership in relation to social and institu-tional change. This is not surprising since this view of social leadership waswidespread in the Austrian literature of the end of the nineteenth and earlytwentieth century. Max Weber, for instance, introduced the concept of the‘charismatic leader’, even if different from Schumpeter’s entrepreneur, such thatit helped to explain how social systems change. Weber’s charismatic leadersappear when organisational routines become insufficient to cope with emerging

8 Schumpeter onentrepreneurship

Richard Arena and Paul-Marie Romani

new situations. Contrary to other individuals, leaders are not submitted to theprevailing rules of social order. Rather, their main function is that ofcontributing to the formulation and introduction of new social norms and rules.Once these new rules have been routinised, that is, accepted by the mass offollowers, these leaders disappear again. Interestingly, Carlin (1956) comparedWeber’s and Schumpeter’s approaches to social leadership when he tried todefine what he called ‘Schumpeter’s constructed type’ of ‘the entrepreneur’. Inhis Principles, Carl Menger also employed the notion of social leadership toanalyse the emergence of money as a fundamental institution of marketeconomies. As we know, he explained the emergence of this institution in termsof an unexpected consequence of decisions taken by social leaders. Wiesergeneralised the notion of social leadership and Menger’s analysis in his Law ofPower (1983 [1926]) and Social Economics (1927 [1914]), explaining the emer-gence not only of money but also of the market economy from a system of‘production on order’. He provided a detailed description of the roles played byleaders, on the one hand, and followers (or what he called ‘the masses’), on theother, arguing that institutional failure became a possibility when followersdisagreed with or rejected the innovations introduced by social leaders.3

In his sociological writings, Schumpeter provides a number of examples ofsocial leadership taken from different historical contexts. One such example isthe introduction of agricultural activities to German tribes. This new activityprompted the emergence of a social distinction between workers–soldiers, onthe one hand, and a ‘military leadership’, on the other, whose function was the‘command of forces’ and the ‘actual execution of combat actions’ (Schumpeter1927 [1951]: 182–3). This was ‘reflected and objectified in the rise of a definiteinstitution among the Germans in their new territories – the creation of mano-rial estates’ (ibid.: 183). Another example is the rise of ‘aristocratic businessmen’(ibid.: 199) in England who were able to benefit from the ‘patrimonialization ofoffice’, of ‘landed property’ and of ‘the individual’ (ibid.: 190–1) and often becamemodern entrepreneurs (ibid.: 202–3). Schumpeter uses these examples to illus-trate how new types of social leadership and, therefore, new forms of productionand of (net) output circulation, emerge in the course of historical time throughagriculture, war, industry and any combination of these activities. Put differ-ently, these examples demonstrate how history and economic sociology mightbe combined with economic theory to study institutional and economicchanges.

Entrepreneurs being no more than social leaders, it becomes possible todistinguish them from capitalists:

In the institutional pattern of capitalism there is machinery, the presence ofwhich forms an essential characteristic of it, which makes it possible forpeople to function as entrepreneurs without having previously acquired thenecessary means. It is leadership rather than ownership that matters. Thefailure to see this and, as a consequence, to visualize clearly entrepreneurial

168 Entrepreneurship and competition

activity as a distinct function sui generis, is the common fault of both theeconomic and the sociological analysis of the classics and of Karl Marx.


This definition of entrepreneurship as a form of social leadership appears to be adirect application of Schumpeter’s methodology of drawing together history,economic sociology and economic theory. This suggests that, following thismethodological logic, a useful way of looking at Schumpeter’s approach toentrepreneurship is by analysing its historical, sociological and economiccomponents. The next section deals with the first two of these components,while sections three and four discuss the economic aspects of entrepreneurialinnovation and rationality, respectively.

The historical and sociological aspects of Schumpeter’sconcept of entrepreneurship

When Schumpeter defined ‘the term Enterprise’ in a famous passage of BusinessCycles (BC), he immediately specified that this definition ‘is based on a histor-ical fact and a theoretical proposition, namely that carrying out innovations isthe only function which is fundamental in history and essential in theory to thetype usually designated by that term’ (Schumpeter 1939: 102).

Schumpeter’s work provides clear explanations of, as well as ample supportfor, a definition of enterprise ‘based on a historical fact’. Again, this emphasis onhistory is not surprising and is in line with Schumpeter’s general methodology.Thus, in the introduction of his ‘Historical outlines’ in BC, he notes that:

[t]he importance of [the historical] approach has been emphasized from theoutset. Since what we are trying to understand is economic change inhistoric time, there is little exaggeration in saying that the ultimate goal issimply a reasoned (= conceptually clarified) history, not of crises only, norof cycles or waves, but of the economic progress in all its aspects and bear-ings to which theory merely supplies some tools and schemata, and staticsmerely part of the material. It is obvious that only detailed historic know-ledge can definitely answer most of the questions of individual causationand mechanism and that without it the study of time series must remaininconclusive and theoretical analysis empty.

(ibid.: 220)

Substantial parts of BC as well as CSD are devoted to the analysis and discussionof ‘Enterprise’ as a ‘historical fact’. Unfortunately, these passages do not alwaysreceive the attention they deserve although as Swedberg has noted, there can belittle doubt that ‘Schumpeter produced the first competent history ofentrepreneurship in economic theory’ (Swedberg 2000: 12). This historydescribes the rise of entrepreneurship in the tenth century and its evolutionthrough historical time, from the artisans of the middle ages through to the

Schumpeter on entrepreneurship 169

entrepreneurs of the putting-out system, the English agricultural entrepreneurswho introduced the system of enclosures, the industrial entrepreneurs of the firstand second industrial revolutions and the modern North American ‘captains ofindustry’ overseeing the organisation of large firms in a ‘trustified’ capitalism. Itwould, however, be misleading to reduce Schumpeter’s contribution to a simplehistory of entrepreneurship. Rather, reflecting his methodological stance, hecombines this history with a systematic analysis of the rise and decline of varioussociological types of leadership.

In order to acquaint his readers with his approach, Schumpeter begins bydiscussing the conventional approach to history that he sees as resting on ‘theprinciple of historic continuity’ (Schumpeter 1939: 226) and then proceeds tocompare this to his own view of institutional, social and economic history.Schumpeter does not deny the usefulness of standard history when it assumes that‘every change seems to consist in the accumulation of many small influences andevents and comes about precisely by steps so small as to make any exact datingand any sharp distinction of epochs almost meaningless’ (ibid.: 227). He argues, inparticular, that this principle of continuity allows one to highlight the shortcom-ings and limitations of theories of history that, in his view, exaggerate theimportance of historical breaks, such as, for instance, those by Marx or Sombart(ibid.: 226–331). Schumpeter does, however, stress that this conventionalapproach to the study of economic history is both insufficient and incompatiblewith his own emphasis on institutional and social changes. This emphasis followsdirectly from Schumpeter’s methodological choice since this leads him to takeaccount, in his own analysis, not only of standard economic history but also of therise and decline of ‘legal institutions but also [of] attitudes of the public mind andpolicies’ (1950: 135). This conception clearly appears when Schumpeter definescapitalism and the entrepreneurial function: ‘Capitalism is that form of privateproperty economy in which innovations are carried out by means of borrowedmoney, which in general, though not by logical necessity, implies credit creation’(1939: 223). Commenting on this definition, he adds that

it should also be observed that, like most other definitions of capitalism,ours is institutional. But of course the institutions which, with very rareexceptions, we treat as data throughout, are themselves the results of andelements in the process we wish to study.

(ibid.: 224)

This clearly means that the analysis of the economic dynamics of capitalistdevelopment cannot be purely historical, but must take account also ofeconomic sociology and, therefore, study the transformations of social leader-ship – that is, of entrepreneurship in the present context. One importantconcomitant of this approach is that some forms of entrepreneurship can existoutside the institutions characteristic of capitalist societies proper in that theycan predate these and/or survive them: ‘The entrepreneurial function itself isnot confined to capitalist society, since such economic leadership as it implies


would be present, though in other forms, even in a primitive tribe or in asocialist community’ (ibid.: 223). To understand why the entrepreneurial func-tion cannot be limited to capitalist societies, it is necessary to have a closer lookat what, in BC, Schumpeter calls his ‘historical outline’ of entrepreneurship.This contains a number of illustrative examples of what Schumpeter means by‘pure entrepreneurship’.

A first example is provided by the transition from militarily organised tocapitalist societies. As Schumpeter notes in CSD, referring to the case ofGerman tribes, in the former societies, the individual art of warfare formed thebasis of social leadership and even of a predominant social class. However,

social and technological change undermined and eventually destroyed boththe function and the position of that class. Warfare itself did not cease onthat account. It simply became more and more mechanized – eventually somuch so that success in what now is a mere profession no longer carries thatconnotation of individual achievement which would raise not only the manbut also his group into a durable position of social leadership.


Schumpeter’s approach also allows us to understand why entrepreneurshippredates the period to which the emergence of capitalism is normally dated. Inhis view, some ‘rudimentary forms of capitalist existence’ (Schumpeter 1939:228) can be traced back to a time as early as the tenth century, in particular inVenice. However, at this time the strength of entrepreneurial innovation and itspower of social attraction are insufficient to impose entrepreneurship as a newform of leadership to the economic system. Similarly, the innovations linkedto the emergence of ‘the type of medieval artisans, their organization andbehavior’ (ibid.) were sufficiently strong to make them increasingly relevant totheir environment. These artisans could have become new economic and socialleaders replacing the prevailing military aristocracy had the new forms of indus-trial and social organisation they were setting up lasted sufficiently long tobecome predominant. According to Schumpeter, instead they ‘succumbed towhat then was a commercially superior method, the putting-out system’ (ibid.).This example is particularly interesting since it shows how a new form of organi-sation was superseded by a more efficient one even before it could be generalised.

Schumpeter’s discussion of the historical phenomenon of enclosures isanother example. Referring to the English landlords and their agriculturalentrepreneurial activity, Schumpeter describes a typical case of the emergence ofa major innovation driven by economic leaders:

Individuals, proceeding of their own accord or accepting the teaching ofsome advocate of improvement, went ahead and set up new productionfunctions, the success of which induced others to follow – first few, thenmany. This was first the case in the matter of enclosures, which in them-selves implied no other innovation than one in organisation – as did the


simple exchange of inconveniently dispersed strips of land – but quickly ledon to others.


Next, the two industrial revolutions provided Schumpeter with ample materialto illustrate his methodological approach. We will not enter here into the detailof Schumpeter’s analysis that is both rich in material and dense. Essentially, thediscussion here centres around the paradigmatic type of innovations promotedand nurtured by typically capitalist entrepreneurship. Among these Schumpeterincludes the emblematic cases of ‘[r]ailroad construction in its earlier stages,electric power production before the First World War, steam and steel, themotorcar, colonial ventures’ (Schumpeter 1950: 132). These cases are also inter-esting because they gave Schumpeter the opportunity to develop the concept ofresistance to innovation.4 Industrialisation, for instance, was faced with thedistrust of consumers in machine-made products, a workers’ revolt againstmachinery and unemployment, and the resistance of old-style firms (such as, forexample, craft guilds) prompting the introduction of prohibitions and regula-tions that ran counter to the interests of factory owners. These multiple sourcesof resistance explain why ‘new men’ do not always become social leaders or whytheir proposals and innovations do not always find sufficient supporters, at leastnot immediately.

Finally, for Schumpeter, the emergence and consequences of trustified capi-talism can also be explained in relation to entrepreneurship.5 Again, this is notthe place to go into more detail regarding the concept and history of trustifiedcapitalism. Rather, what is of interest to us here is to explore the ways in whichthis discussion builds on and involves analytical concepts belonging toeconomic sociology.

To begin with the concept of institutional inertia, it is clear that, forSchumpeter, the rise of large firms reduces resistance to change. Innovationbecomes routinised and is transformed into a normal and permanent form ofeconomic activity.6 Technological progress now results from a process ofnumerous minor changes enacted by R&D specialists. On the consumers’ side,resistance also vanishes since agents have grown accustomed to the ‘incessantstream’ of new consumer goods.

Regarding the impact of large firms on the existing institutional setting ofcapitalism, Schumpeter’s emphasis is their undermining effect on the institu-tional underpinnings of entrepreneurship. On the one hand, the rise of largefirms goes hand in hand with a progressive elimination of small producers. Thisincrease of industrial concentration contributes to the decline of the classicaltype of entrepreneurship associated with the two industrial revolutions. Thedecline of entrepreneurship, in turn, has drastic repercussions on other relatedinstitutions, triggering further changes: ‘The very foundation of private prop-erty and free contracting wears away in a nation in which its most vital, mostconcrete, most meaningful types disappear from the moral horizon of thepeople’ (ibid.: 140–1). On the other hand, the internal organisation of big


business implies a new division of economic agents into three groups:managers, ‘big stockholders’ and ‘small stockholders’ (ibid.: 141). For a numberof different reasons, none of these groups correspond to capitalistentrepreneurs. This tendency towards a decline of entrepreneurship is rein-forced by a progressive marginalisation of the ‘freedom of contracting’ in favourof ‘stereotyped, unindividual, impersonal and bureaucratized’ contracts (ibid.).The institutional foundations of entrepreneurship are therefore destroyed andthis situation is compounded by the disappearance of the classical ‘type’ ofentrepreneurship.

Moreover, once we take account also of the organisational rather than onlythe purely institutional aspects of this development, another point frequentlystressed by Schumpeterian scholars must not be overlooked, namely, the factthat the entrepreneur is no longer a single person. In his later articles, writtenwhile he was involved with the Research Center in Entrepreneurial History atHarvard University, Schumpeter rewrote some of his earlier passages dealingwith this organisational tendency, in particular with regard to the first andsecond editions of CSD (Schumpeter 1950 [1942, 1947]). He now argued thatthe concept of entrepreneur could equally well refer to an organisation, be thispolitical or economic, stressing entrepreneurial behaviour more than theentrepreneur as an individual actor (Schumpeter 1947 [1989] and 1949 [1989].As Blaug has recently reminded us,

The entrepreneur in Schumpeter is a functional role which is not neces-sarily embodied in a single physical person and certainly not in awell-defined group of people. The entrepreneur may be a capitalist or evena corporate manager but whether all these different functions are combinedin one or more persons depends on the nature of capital markets and on theforms of industrial organisation.

(Blaug 2000: 83)

Schumpeter’s position on this point, however, seems to be somewhat ambiguous.In BC, he insisted rather on the technical ‘difficulties’ which arise for theeconomic historian ‘in settling who the entrepreneur [is]’ (Schumpeter 1939:404), quoting the example of the construction of railroads in the United States(ibid.: 404–5). In the last edition of CSD overseen by himself, Schumpeter inter-preted the rise of big business organisation as a major factor in the decline ofentrepreneurship. Referring to the disappearance of the social leadership of‘armored knights’ in the middle ages in the wake of the mechanisation ofwarfare activity, he argues that

a similar process – in the last analysis the same social process – underminesthe role and, along with the role, the social position of the capitalistentrepreneur. His role, though less glamorous than that of medievalwarlords, great or small, also is or was just another form of individual leader-ship acting by virtue of personal force and personal responsibility for


success. His position, like that of warrior classes, is threatened as soon asthis function in the social process loses its importance, and no less if this isdue to the cessation of the social needs it served than if those needs arebeing served by other, more impersonal methods.


Most likely, the tension between these two aspects can be resolved once we takeaccount of Schumpeter’s motivation to develop a ‘comprehensive historyof entrepreneurship’ (Schumpeter 1949 [1989]: 264). More specifically,according to Schumpeter, the drastic transformations modern social scientistsfaced with the rise of big business meant that their interest had to shiftaway from preconceived views of the entrepreneur and entrepreneurship(Schumpeter’s own included) towards a focus on ‘the actual activity of theentrepreneur’, that is, on their actual role in the real world as it had become.(ibid.) This implies that what matters in Schumpeter’s historical outline of thedevelopment of entrepreneurship is not the accurate depiction of historicaldetail, but its logical element – the combination of history, economic sociologyand economic theory.

In Schumpeter’s view, the emergence of trustified capitalism not only under-mines entrepreneurship through its institutional and organisational effects butalso through its systemic effects. Schumpeter here refers to the way in which therise of big business transformed the social structure and social ‘habits of mind’(Schumpeter 1943 [1989]: 181) in capitalist society and in which these transfor-mations helped to destroy the social system that supports entrepreneurship.Without entering into any detail here, two main developments highlighted bySchumpeter deserve mention. First, Schumpeter discussed what he called thedestruction of the ‘protective strata’, by which he meant the different groups(artisans, small traders and producers, farmers, peasants, etc.) that constitute thesocial support of capitalist society and of entrepreneurship. Second, he arguedthat the rise of rationalism, itself aided by the evolution of capitalism, favouredthe systematic use of economic calculus, thereby diminishing the importance ofthe qualities of imagination and audacity characteristic of entrepreneurialbehaviour. Both these developments are good examples of the ideas that rein-forced Schumpeter in his view that economic sociology represented an essentialintegral element of the analysis of economic development.

The economic aspects of entrepreneurial innovation inSchumpeter

The economic aspects of Schumpeter’s concept of the entrepreneur havereceived ample attention There is therefore no need here to dwell on them inany great detail. Rather, what we will focus on here is the link between thehistorical and sociological aspects of Schumpeter’s analysis of entrepreneurshipand its economic elements. More precisely, what we are interested in is how hecombines the two and what, if any, are the analytical implications of this


combination for the economic concept of entrepreneurship. We begin with adiscussion of the relation between Schumpeter’s economic analysis ofentrepreneurship and the innovative function per se. As we know, Schumpeterheld that innovation is not limited to invention or to what is normally called‘technical progress’ in economic literature. Instead, he analysed entrepreneurialinnovation in terms of the social function that permits capitalist economicdevelopment.

This definition, which derives directly from the characterisation of entre-preneurship as a specific type of leadership, excludes a view of entrepreneursas permanent economic agents. This feature of his analysis already made anappearance in our discussion of Schumpeter’s historical outline of entrepreneur-ship. It is therefore easy to see why, from an economic standpoint, ‘being anentrepreneur is not a profession and as a rule not a lasting condition’(Schumpeter 1912 [1934]: 78). This lack of permanence also explains why,according to Schumpeter, entrepreneurs cannot form a social class, contrary tolandowners or capitalists (ibid.). Their social destiny is, in fact, to disappear assoon as the changes they have planned are realised. Moreover, membership of adominant social class and being a social leader are different in that these rolespertain to different realities, even if, historically, there might be some overlap-ping between them.7 That the existence of entrepreneurs is thus strictly linkedto the planning and implementation of innovations is also the reason why ‘thefunction of the entrepreneur itself cannot be inherited’ (ibid.: 79). Finally, itprovides an explanation for a feature of Schumpeter’s analysis that is oftenstressed in the literature, namely, that Schumpeterian entrepreneurs cannotsimply be compared to, or equated with, lasting economic actors, such amanagers–employees (ibid.: 75), buying and selling agents or technical experts(ibid.: 77), even if, in real life, entrepreneurs can also engage in these comple-mentary activities.

The fate of entrepreneurs to rise and vanish also explains whyentrepreneurial profit is ‘temporary by nature’ (Schumpeter 1939: 105), some-times including ‘windfall gains’ (ibid.: 106). This characteristic ofentrepreneurial profit must, in turn, be seen in the context of Schumpeter’scritique of the concept of pure competition and his interest in the phenomenonof imperfect competition: it is essential to his theory of economic developmentthat there be enough imperfection in the competitive market system for theinnovating entrepreneur to see his innovative behaviour rewarded by super-normal monopoly profits.8 If competition were perfect, imitative behaviourwould be instantaneous and profits would immediately fall back to their normallevels, leaving no incentive for technological change.9

Finally, the link between the entrepreneurial function and major economicchanges also helps to illuminate the central importance Schumpeter attributedto credit in his analysis of capitalism. Access to credit allows the Schumpeterianentrepreneur to face the ‘time lag between the deadlines that [he] has to meetin his tasks and his ability to fulfil them’ (Schumpeter 1912 [1934]: 195). This isthe reason for another feature, arising from the discontinuous function of


entrepreneurs in society, namely that, for Schumpeter, they are not the naturalowners of the means of production. In passing, it may be noted that the roleplayed by credit in Schumpeter’s theory of entrepreneurship makes it possible to understand why medieval artisans are not, after all, typical entrepreneurs:they owned the means of production and did not operate on the basis of adeveloped system of credit. Obviously, firm managers dealing with minoreconomic changes are excluded from this analysis since their behaviour is typi-cally characterised by ‘adaptive responses’ whereas ‘creative’ responses are acentral feature of entrepreneurship (Schumpeter 1947 [1989]: 222). As notedabove, entrepreneurs inevitably meet with resistance to change or with institu-tional inertia emanating from the environment in which they attempt topromote a particular process of innovation. Such resistance is intrinsic to thevery notion of social leadership: ‘[In] economic life … every step outside theboundary of routine has difficulties and involves a new element. It is thiselement that constitutes the phenomenon of leadership’ (Schumpeter 1912[1934]: 84). Resistance to change therefore appears to be an inevitable conse-quence of entrepreneurial activity:

The whole economic history of capitalism would be different from what itis, if new ideas had been currently and smoothly adopted, as a matter ofcourse, by all firms to whose business they were relevant. But they were not.It is in most cases only one man or a few men who see the new possibilityand are able to cope with the resistances and difficulties with which actionalways meets outside the ruts of established practice. This accounts, on theone hand, for the size of the gains that success often entails, and, on theother hand, for the losses and vicissitudes that it produces for other people.

(Schumpeter 1946 [1991]: 413)

Finally, the transitory nature of entrepreneurial activity also explains thespecific features that account for the difference between this form of social lead-ership and those that prevail in non-capitalist societies. First, contrary to othertypes of social leaders, entrepreneurs have ‘no cultural tradition or attitude’;their only role is that of an ‘upstart’ in capitalist society (Schumpeter 1934: 90).Secondly, they lack the ‘glamour’ of other social leaders such as, for example,medieval warlords, which is why ‘only in rare cases [do they] appeal to the imag-ination of the public’ (ibid.: 89).

The economic characteristics of entrepreneurship just described are relatedto the innovative function per se. They show that entrepreneurs are fundamen-tally the ‘prime movers’ of capitalist economic change. This emphasis on thetransitory function of entrepreneurs explains why Schumpeter was so critical ofother theories of entrepreneurship. We have already noted that, by definition,entrepreneurs cannot be the owners of the means of production. This logicallyled Schumpeter to reject Say’s theory of entrepreneurship that definesentrepreneurs as co-ordinators of productive services (Schumpeter 1908: 434).He also excluded the proposition that entrepreneurs might be risk-bearers, since


the risk associated with innovation is carried by capitalists (ibid.). Furthermore,he of course refused the Walrasian concept of the entrepreneur as ‘ne faisant nibénéfice ni perte’10 (ibid.: 1908: 438). And finally, he rejected the classical theorysince, in his view, this confused capitalist with entrepreneurial functions(Schumpeter 1912 [1934]: 77 and 1939: 105).

The economic aspects of entrepreneurial rationality inSchumpeter

We now turn to another aspect of Schumpeter’s economic analysis ofentrepreneurship, namely, his conception of entrepreneurial rationality.11 Thispart of our discussion draws, in particular, on Schumpeter’s reflections on ‘Themeaning of rationality in the social sciences’ (Schumpeter 1940 [1991]). Thiswork is primarily of methodological interest; it takes up and discusses a distinc-tion he had already introduced in 1908 and 1912 between energetic andhedonistic rationality. However, Schumpeter begins his methodological reflec-tions with another distinction, that between the ‘rationality of the observer’and the ‘rationality in the observed’ (ibid.: 319, 323). This distinction contrastswith the views expressed by conventional social analysis which assumes

that the individuals under research – sometimes even ‘the people’ as such –are themselves actuated by clearly perceived motives, and regulate theirbehavior with conscious rationality working in the full daylight of the egos(subjective rationality). In doing so, those analysts, especially those ofbygone generations, have almost invariably overrated the actual range ofconsciously rational action.

(ibid.: 326)

Put differently, according to Schumpeter, social theorists, and in particulareconomists, writing in the conventional tradition attribute to the consciousnessof observed agents the knowledge of and the ability to employ the ‘rationality ofthe observer’. At its most extreme, this view also assumes that these ‘consciouslyrational actions’ are mutually compatible and therefore produce socially optimalresults. Schumpeter raises several objections to this received view. First, even ifit is clearly permissible for the social scientist to employ the ‘rationality of theobserver’, this does not release him or her from the need to examine whetherthe appropriate conditions for the application (that is the conformity with facts)of this type of rationality are given: ‘The rationality of a model must be supple-mented by the rationality of its application which is part of the analyst’s task’(ibid.: 324). This is not an easy task since it consists in determining the condi-tions under which the rationality of the observer and the rationality in theobserved are compatible. The list of difficulties the social scientist must over-come is long and includes: a precise definition of the conditions of compatibility(ibid.: 322–3); the possibility that an infinite variety of cognate ends can beassociated with the same behaviour (ibid.: 324); the possible existence of


conflicting ends associated with the same behaviour (ibid.); the necessity ofdistinguishing rational behaviour from rational results (ibid.: 325);12 and thenecessity of a clear understanding of the social and historical context ofrational behaviour (ibid.). This is not all, however. Compatibility of the ratio-nality of the observer with the rationality in the observed does not onlyrequire what Schumpeter called ‘objective rationality’, namely, the removal ofthe difficulties that we referred to. It can also require taking into account the‘subjective rationality’ of the observed individuals.

These general remarks by Schumpeter on rationality in social sciences aredirectly relevant to the discussion of his view of entrepreneurial rationality.Hence, the distinction between the ‘rationality of the observer’ and the ‘ratio-nality in the observed’ immediately highlights the shortcoming of theWalrasian conception of entrepreneurial rationality: this conception grosslyconfuses both rationalities (independently of whether the rationality in theobserved is regarded as objective or as subjective). Walrasian economicsregards optimal behaviour as universal. It therefore provides a typical examplefor a situation in which the observer simply assumes, or takes it for granted,that the logical form of rationality he has built has a general power of applica-tion, thus allowing him to discard the Schumpeterian problem of thecompatibility between the rationality of the observer and the rationality in theobserved.

We know that Schumpeter was opposed to this Walrasian conception andcountered it by distinguishing two forms of rationality which Santarelli andPesciarelli have described by contrasting ‘energetic-dynamic’ and ‘hedonistic-static’ modes of behaviour (1990: 682). Hedonistic-static rationality is therationality of followers who prefer to minimise their efforts to attain their endsand, therefore, to rely on routinised modes of behaviour. This form of rationalitycorresponds to the behaviour of firm owners within the circular flow. Accordingto Schumpeter’s view of economic development, it is, on the contrary, ‘only inthe contemporary economy that the energetic type has developed to such asignificant extent in the economic field as to constitute a special class ofeconomic subject and be given his own name: “entrepreneur” ’ (Schumpeter1912 [1934]: 171).

Within this dualistic scheme, Schumpeter first reinterpreted the Walrasianconception of rationality as a hedonistic-static mode of behaviour, arguing thatin this sense it was of limited validity.13 This reinterpretation is only a first steptowards examining the conditions of the validity of the rationality of theobserver – that is, of its compatibility with the rationality in the observed. It isin this sense that Schumpeter considered that Walrasian rationality could beapplied to stationary phases or to some phases of the business cycle. The secondstep entails Schumpeter’s observation, based on his equating entrepreneurialrationality with energetic-dynamic types of rationality, that the latter prevailsonly under specific conditions corresponding to economic development or toother phases of the business cycle. In Schumpeterian terms, these two steps


taken together can be interpreted as an attempt to demonstrate the existence ofan objective rationality in the observed.

However, Schumpeter went further, distinguishing between ‘objective’ and‘subjective’ rationalities in the observed. This distinction comes into play in hisdiscussion of egoism and the concept of homo economicus. In this context,Schumpeter draws the attention of his readers to what he describes as a logical‘danger’. Referring to entrepreneurial behaviour, he observes that

If our firm failed to adapt to the new environmental conditions from habit,laziness, lethargy, its behavior may, from the standpoint of its manager, yetconform to a rational model – he may be maximizing his welfare by notbothering. Now, just as in the case of altruistic behavior practiced from a‘egotistical’ wish for the satisfaction such a behavior yields, or, in the case ofenergetic and aggressive activity pursued ‘hedonistically’ – because it givespleasure – so we are in cases such as the one of the lazy manager in dangerof losing our criterion of rational behavior: all behavior, so its seems, wouldhave to be looked upon as tautologically rational ex visu of suitably chosenends and horizons.

(Schumpeter 1940 [1991]: 330)

The removal of this danger requires the introduction of a twofold distinction.First, it becomes necessary to distinguish objective from subjective forms ofrationality in the observed. For instance, the case of ‘energetic and aggressiveactivity pursued “hedonistically”’ can be understood if it is considered that theactor’s rationality is of both a subjective hedonistic kind (i.e. that of homoeconomicus) as well as an objective energetic rationality (of the entrepreneurialtype). Secondly, another distinction, referring to two different psychologicaltypes, may also be of use here: rationality in behaviour and rationality inmotives. From this point of view, energetic-dynamic and hedonistic-staticmodes of behaviour could be seen to belong to the same type of rationality sinceboth follow the same logic of intentionality of connecting means to ends.However, rationality in behaviour also entails a substantial difference betweenentrepreneurs who use their excess of energy for innovative efforts and followerswho minimise their efforts through passive imitation. In the same vein, thenotion of rationality in motives also implies a substantial difference betweenentrepreneurial motives (desire for supremacy, will to conquer, joy of creation)and routinised motives (maximisation of profit). Finally, the list of difficulties inreconciling the rationality in the observed with the rationality of the observer,referred to earlier, also comes into the equation. However, if one attempted toresolve these difficulties, this would actually reinforce the distinction betweenthe two types of rationality or, in Schumpeter’s terminology of 1947, thedistinction between ‘creative and ‘adaptive responses’ (Schumpeter 1547[1989]: 222).


Concluding remarks

Among the criticisms of Schumpeter’s theory of entrepreneurship, made afterhis death, three stand out in particular.

First, it is true that Schumpeter paid particular attention to major innova-tions. In the view of many commentators, this emphasis led his successors toneglect other aspects of technical change, such as learning by doing or learningby using that are rarely considered by Schumpeter. Rosenberg’s judgement isparticularly severe in this respect. Because of Schumpeter’s emphasis on majortechnical changes,

we have become saddled with an analytic framework which simply does notexplicitly recognize or lead us into a consideration of sources of productivityimprovements other than those emanating from major innovations. Theconsequences of this have been particularly serious in impeding our under-standing of the origin and nature of technical change.

(Rosenberg 1976: 66)

Blaug confirms this judgement in his assessment of Schumpeter’s contribution:

Schumpeter never managed to get away from the concept of theentrepreneur as a heroic adventurer and even his discussion of innovationsis too much focused on the introduction of dramatic novelties with far-reaching consequences … losing sight of the fact that so much technicalprogress consists of small, cumulative improvements.

(Blaug 1996: 446)

Second, Schumpeter has been accused of having paid too little attention to theprecise mechanisms or processes that lead to the emergence of innovations. It is,once again, Rosenberg who complains most forcefully about the strict distinc-tion Schumpeter made between invention, innovation and imitation, arguingthat these ‘characteristics’ are often combined in differing ways that Schumpeternever considered explicitly (Rosenberg 1976: 66–7). More recently, Heertjetakes up this critique and stresses consequent analytical limitations inSchumpeter:

Schumpeter’s insight is indispensable for a better understanding of thedynamics of capitalism but many questions are left open, such as the timepattern and nature of applications of new technology. When will the newtechnology be implemented? What is the role of demand factors, in partic-ular, in the case of new products? What kind of equipment will beintroduced?

(Heertje 1988: 87)

Finally, commentators have also expressed their scepticism with regard toSchumpeter’s claim that he provided an endogenous explanation of the emer-


gence of innovations. Demsetz notes that, in Schumpeter, ‘the entrepreneur isneither an object of analysis nor of research, but is rather a deus ex machina ofeconomic change’ (1982: 275). The same criticism is developed by Heertje wholinks it to the two above-mentioned points of criticism (Heertje 1988: 87).

These criticisms certainly contain a substantial element of truth. However, inour view they should not be overestimated. On the one hand, if it is true thatSchumpeter neglected cumulative minor changes as well as the mechanisms oftechnical progress, it does not therefore follow that the analysis of such changesare logically incompatible with his framework. We would, for example, arguethat many more recent contributions that emphasise the self-organisationalnature of technical change – and that, by the way, often refer to theSchumpeterian legacy – can be combined with Schumpeter’s analysis of the roleof imitative behaviour as well as with the insights he developed at Harvardduring the last years of his life.

On the other hand, there can be little doubt that Schumpeter’s analysis ofentrepreneurship remains vulnerable to the criticism of having treatedeconomic change as exogenous so long as, in line with much recent work in thefield of evolutionary economics, his contribution is understood in terms of apurely economic theory of entrepreneurial competition seen as a selectiveDarwinian process. We have tried to argue that such an interpretation is inade-quate. If, instead, we consider entrepreneurship as the form of social leadershipprevailing under capitalism, this critique loses its force. Schumpeter explicitlyrejected the idea that capitalist dynamics can be analysed within a self-contained conception of economic theory, as opposed to the theory of thecircular flow. Once these dynamics are analysed within a methodological frame-work that combines economic theory with sociology and history – asSchumpeter does in the historical outline we have discussed – they clearlyreflect an endogenous process of change. While presently this may not be thepredominant view among scholars of technical change, this does not mean thatthe predominant view is the right one and that Schumpeter’s perspective can orshould no longer be revived.

Notes1 See, for instance, Kamien and Schwartz (eds) 1982, chs 2, 3.2 See the introduction to this volume for a discussion of social leadership in

Schumpeter’s writings.3 For a more detailed discussion of the role of the concept of social leadership in the

Austrian tradition, see the introduction and Arena and Gloria-Palermo (2001).4 Defined as a form of ‘institutional inertia’ in the introductory chapter.5 See the introduction.6 For details, see the introduction.7 See the introduction to this volume for details.8 See the chapter by Raybaut and Sosthé in this volume.9 This remark shows, by the way, why so-called ‘Schumpeterian hypotheses’ are reduc-

tionist when their builders pretend to grasp with them the main reason forSchumpeter’s interest in imperfect competition.


10 ‘[M]aking neither a profit nor a loss’.11 Aspects of this theme are dealt with in the introduction and in Châteauneuf-

Malclès’s contribution.12 This difficulty includes the possible existence of socially unexpected consequences of

voluntary individual behaviours.13 See the chapter by Arena in this volume.

Bibliography

Arena R. and Gloria-Palermo, S. (2001) ‘Evolutionary themes in the Austrian tradition:Menger, Wieser and Schumpeter on institutions and rationality’, in P. Garrouste andS. Ioannides (eds), Evolution and Path Dependence in Economic Ideas, Aldershot:Edward Elgar.

Blaug, M. (1996) Economic Theory in Retrospect, 5th edn, Cambridge: Cambridge Univer-sity Press.

—— (2000) ‘Entrepreneurship before and after Schumpeter’, in R. Swedberg (ed.)Entrepreneurship: The Social Science View, Oxford: Oxford University Press.

Carlin, E.A. (1956) ‘Schumpeter’s constructed type: the entrepreneur’, Kyklos, vol. 9, no.1: 27–43.

Demsetz, H. (1982) ‘Concluding comments’, in J. Ronen (ed.) Entrepreneurship,Lexington, Mass.: Lexington Books.

Heertje, A. (1988) ‘Schumpeter and technical change’, in H. Hanusch (ed.) EvolutionaryEconomics: Applications of Schumpeter’s ideas, Cambridge: Cambridge University Press.

Kamien, M.I. and Schwartz, N.L. (1982) Market Structure and Innovation, Cambridge:Cambridge University Press.

Rosenberg (1976) Perspectives on Technology, Cambridge: Cambridge University Press.Santarelli, E. and Pesciarelli, E. (1990) ‘The emergence of a vision: the development of

Schumpeter’s theory of entrepreneurship’, History of Political Economy, vol. 22, no. 4:677–96.

Schumpeter, J.A. (1908) Das Wesen und der Hauptinhalt der theoretischenNationalökonomie, Munich and Leipzig: Dunker und Humblot.

—— (1912) Theorie der wirtschaftlichen Entwicklung, Leipzig: Duncker und Humblot.Preface dated Vienna, July 1911. English translation of 2nd edn as The Theory ofEconomic Development: An Inquiry into Profits, Capital, Credit, Interest, and the BusinessCycle, Cambridge, Mass.: Harvard University Press, 1934.

—— (1927) ‘Die sozialen Klassen im ethnisch homogenen Milieu’, Archiv für Sozialwis-senschaft und Sozialpolitik, vol. 57: 1–67. Translated into English as ‘Social classes in anethnically homogenous environment’ in J.A. Schumpeter (1951). Reprinted in J.A.Schumpeter (1991).


—— (1940) ‘The meaning of rationality in the social sciences’, introduction by W.F.Stolper and R. Richter (eds) (from typescript of a Faculty Seminar given at Harvardin 1940), Zeitschrift fuer die Gesamte Staatswissenschaft, 140, 1984. Reprinted in J.A.Schumpeter (1991).

—— (1943) ‘Capitalism in the postwar world’, in S.E. Harris (ed.) Postwar EconomicProblems, New York and London: McGraw-Hill. Reprinted in J.A. Schumpeter(1989).


—— (1946) ‘Comments on a plan for the study of entrepreneurship’ (to be found inHarvard University Archives). Reprinted in J.A. Schumpeter (1991).

—— (1947) ‘The creative response in economic history’, Journal of Economic History,November: 149–59. Reprinted in J.A. Schumpeter (1989).

—— (1949) ‘Economic theory and entrepreneurial history’, in Research Center inEntrepreneurial History (ed.), Change and the Entrepreneur, Cambridge, Mass.:Cambridge University Press. Reprinted in J.A. Schumpeter (1989).

—— (1950) Capitalism, Socialism and Democracy, 3rd edn, London: Allen & Unwin; 1stedn, New York: Harper, 1942; 2nd edn, London: Allen & Unwin, 1947. Reprinted(with a new introduction by R. Swedberg), London: Routledge, 1992.

—— (1951) Imperialism and Social Classes (ed. P. Sweezy), New York: Augustus M.Kelley.


—— (1991) The Economics and Sociology of Capitalism (ed. R. Swedberg), Princeton:Princeton University Press.

Swedberg, R. (2000) ‘The social science view of entrepreneurship: introduction andpractical applications’, in R. Swedberg (ed.) Entrepreneurship: The Social Science View,Oxford: Oxford University Press.

Wieser, F. von (1914) Theorie der gesellschaftlichen Wirtschaft. Translated into English asSocial Economics, 1927. Reprinted, New York: Augustus M. Kelley, 1967.

—— (1926) Das Gesetz der Macht, Vienna: Springer. Translated into English as The Lawof Power, Lincoln: Bureau of Business Research, University of Nebraska, 1983.


Introduction

The concept of competition has always been central to economic thinking. It isnot, therefore, surprising that it should have given rise to numerous debates andinterpretations and continues to be a subject of controversy. Schumpeter is,without doubt, one of the economists most often called upon in this contro-versy. References to his work are perhaps most widespread in the context of thedebate on market structure and efficiency, and here, in particular, regarding histreatment of this question in Capitalism, Socialism and Democracy (CSD).1

However, Schumpeter’s work has also been extensively influential in other areasof competition theory concerned with innovation, prices and the formation ofmarket equilibria. This aspect of the debate has led to the formulation ofdifferent neo-Schumpeterian frameworks, such as recent models of technolog-ical competition and endogenous growth, whose main preoccupation is with therelation between static and dynamic analysis. Finally, Schumpeter’s writings arealso frequently invoked to describe alternative approaches to price competitionof which there are two main strands: the first of these, the evolutionaryapproach, develops a competition analysis based on technological diffusion andselection processes, while the second emphasises the role of institutions, organi-sations and structural change.

On the whole, these diverse references to Schumpeter’s work indicate thecomplexity of his conception of competition. There can, thus, be little doubtthat Schumpeter’s views on competition provide ample space for further investi-gation, and this is then the main focus of this chapter. The chapter is organisedas follows: section 2 deals with Schumpeter’s assessment of the main assump-tions made by the theory of competition until 1914; section 3 looks atSchumpeter’s appraisal of perfect or free competition within the circular floweconomy. Section 4 discusses the meaning(s) of competition in the process ofeconomic development, while section 5 is concerned with Schumpeter’s atti-tude towards the emerging theories of imperfect competition. The sixth andfinal section contains concluding remarks.

9 Schumpeter on competition

Alain Raybaut and Franck Sosthé

Schumpeter on the competitive hypothesis in the History ofEconomic Analysis

To begin with, it is worth noting that, according to Schumpeter, what the clas-sics called free competition is ‘an institutional assumption rather than the resultof certain market conditions. And so firmly were they convinced that thecompetitive case was the obvious thing, familiar to all, that they did not botherto analyse its logical content’ (Schumpeter 1954: 545). He asserts, more specifi-cally, that since Adam Smith, two institutions – namely, private ownership andfree competition – characterise the economic organisation of the society studiedby the classics. In this context, Schumpeter mentions that J.S. Mill deals withcompetition in a chapter of the Principles entitled ‘Competition and customs’(Book II, ch. 4). In his view, the meaning of Mill’s famous assertion, accordingto which it is only through the principle of competition that political economyhas any pretension to being a science, is simply that prices and quantities aredeterminate whenever free competition prevails, that is, in the absence ofmonopoly which was considered as abnormal and vigorously condemned.

From this point of view, J.S. Mill takes credit for having made two importantpoints. First, he emphasised the importance of customary prices, ‘custom beingthe only reason he gave for the existence of competition’ (ibid.: 546). Second,he pointed out that competition ‘often “falls short of the maximum” and that inthis case, a general correction must be applied … Into such a picture co-opera-tive price setting could enter only … as another deviation from normalpractices like straight monopoly’ (ibid.). On the whole, however, Schumpeterstresses that the classics did not ‘think necessary to state what competition is’(ibid.: 545 fn.). The first author in the early nineteenth century to handleperfect or pure competition correctly was Cournot, whose contribution toimperfect competition will be dealt with below.

Schumpeter goes on to argue that, generally speaking, economists of theperiod 1870–1914 ‘substantially retained the habits of their “classic” predeces-sors, which was to consider “competition” as the normal case from which tobuild up their general analysis’ (ibid.: 972).2 This does not, however, precludethe existence of substantial differences in the status accorded to competition inthese theories. According to Schumpeter, for Walras and the Austrians purecompetition covers most of actual business practices, whereas Pareto denies thatcompetition actually governs society. Next, Marshall and Wicksell emphasisefrequent deviations from the competitive scheme, while Clark underlines thatcompetition ought to be the normal case and should be enforced by appropriatepolicies. Finally, Schumpeter discusses Cassel’s position that the actual system,however non-competitive in parts, nevertheless works out on the whole as if itwere competitive. Even so, Schumpeter reiterates his main point, to wit thatnone of these authors adds anything to the view of competition already devel-oped. As a matter of fact:

Schumpeter on competition 185

To a surprising extent, they continued to look upon the competitive casenot only as the standard case that, for certain purposes, the theorist mightfind it useful to construct, but also as the normal case of reality. Even theowner-managed firm survived much better in economic theory than it didin actual life. The great merit that must nevertheless be put to their creditis that they complemented this vision by an analysis that was far superior tothat of the ‘classics’. … [T]hey defined competition and analyzed its modusoperandi with ever-increasing success; they worked out the theory of othercases such as straight monopoly, oligopoly and so on.

(ibid.: 892)

Given this overall assessment, he then sets himself the task of investigatingthese contributions to the pure theory of competition more closely.3 To beginwith he emphasises that, ‘while not all of them were uncritical eulogists ofcompetition … nearly all of them were apt to yield to the specific bias of theeconomic theorist … the bias for easily manageable patterns’ (ibid.: 972), whichis why they assumed perfect competition in the first place. Consequently,product and factor prices cannot be influenced by individual households orfirms, and can thus be treated as parameters given a theory of their respectivebehaviour. Schumpeter comments that ‘[w]e may call this the principle ofexcluded strategy’ (ibid.). In his view, the exclusion of pricing strategiesdescribes best what Walras meant by libre concurrence, with Pareto’s definitioncoming to much the same thing.4 Schumpeter then develops a lengthy argu-ment on the difficulties raised by this assumption of excluded strategy, whichreveals his own perception of the issue:

Exclude strategy as much as you please, there will still remain the fact thisadaptation … will produce results that differ according to the range ofknowledge, promptness of decisions and rationality of actors, and alsoaccording to expectations they entertain about the future course of prices,not to mention the further fact that their action is subject to additionalrestrictions that proceed from the situation they have created for them-selves by their past actions.

(ibid.: 973)

Schumpeter notes that Walras was obviously aware of these difficulties, eventhough he does not seem to have realised to what extent they qualified thepractical importance of his competitive assumption.5 However, according toSchumpeter, absorbed as Walras was in the pioneer task of working out theessentials of the mathematical theory of the economic process, ‘he had nochoice but to simplify heroically’ (ibid.: 974).

To sum up, for Schumpeter the ‘bulk of the [1870–1914] period’s pure theorywas a pure theory of static equilibrium that excluded strategy. The all-round riseof the scientific level produced, if not the term yet the substance of, what we


now call pure or perfect competition’ (ibid.: 972–3). With this criticism in mind,we now turn to Schumpeter’s own contribution to the theory of competition.

Competition within the circular flow

It is generally agreed that the circular flow refers to a state of equilibrium whichis not far removed from a Walrasian general equilibrium. First, according toSchumpeter, the prices of individual goods ‘form a value system, the separateelements of which are mutually dependent’ (1912 [1934]: 39). Second, in anexchange economy without development, the prices of all products must, underfree competition, be equal to the prices of services, of labour and natureembodied in them. In The Theory of Economic Development (TED), Schumpeteris, in fact, quite explicit about the link he makes between the Walrasian theoryof general equilibrium and economic facts. In his view, the Walrasian schemerefers to a special field of economic research, namely the study of regularphenomena in an exchange economy with private property, division of labourand free competition.

Schumpeter clearly thinks that this aspect of economic investigation refersto the ‘normal cycle’ of an exchange economy. He argues that the working ofthe exchange economy is related to the experience of producers who know theintensity of demand. In other words, the description of ‘normal business situa-tions’ of the economic system centres around routine choices. And this does notmean that no change in economic magnitudes may appear: all this means is thatthese changes are taken into account by agents who develop adaptivebehaviour. According to him,

this rationale is afforded by the concept of equilibrium. Hence, in thesecases, stationary flow and equilibrium are analytically equivalent and,describing the same mass of facts, have the same empirical basis.

(Schumpeter 1939: 42, fn.)

Consequently, Schumpeter does not make a distinction between static equilib-rium and regular growth. In both cases, the agents’ experiences explain the realtendency towards an equilibrium position. For Schumpeter,

it is however possible to prove beyond reasonable doubt and with butunimportant qualifications that there exists a uniquely determined equilib-rium state of the economic system in the special case which, followingusage, we shall call the case of perfect competition.

(ibid.: 47)

That is, Schumpeter accepts the idea that a unique or growth equilibrium existsand regards as the important question the stability of such an equilibrium.Accordingly, to Schumpeter, this concept is to be useful as a tool of business


cycle analysis, only if in it the economic system strives ‘to re-establish equilib-rium whenever it has been disturbed, in such a way as to absorb the change.This problem has first been seen by Walras, although some critics do not seemto be aware of the fact’ (ibid.: 9).

Thus, in the TED, Schumpeter draws together his thoughts on the subjectarguing that, while the existence of a general equilibrium is still to be proved, itis a convenient description of reality and of facts, and that the main open ques-tions confronting the economist relate to the stability of such an equilibrium.He essentially defends the idea that in a normal cycle routine behaviour willalways drive the system towards an equilibrium position without producing anydrastic changes in the main variables. In this context, Schumpeter also criticisesthe Walrasian tâtonnement as a purely theoretical construct used to describereality. Referring to the important role of experience and routine for thestability of equilibrium, he makes the point that, if this experience did not exist,if agents lost this experience, they would have to restore it by tâtonnement.Schumpeter, here, seems to distance himself from Walras by arguing that theconvergence mechanism is really based on institutional characteristics of thesystem such as habits, routines and learning processes, and that consequentlythe Walrasian tatônnement is at best an approximation of reality.

To fully grasp his point, it is important to understand that Schumpeter’s criti-cism of the tatônnement is directed at its usefulness in explaining the relationbetween microeconomic behaviour and macroeconomic equilibrium. Thus, heargues that if an agent were to find himself in a disequilibrium situation, freecompetition will indeed play its role and will

drive him toward equilibrium, and if all firms and households simultane-ously react in the same manner, it will eventually bring the whole system toequilibrium, provided that all actions and reactions are performed within thebounds of familiar practice that has evolved from long experience and repetition.

(Schumpeter 1939: 47, italics in the original)

In the same vein, he adds that: ‘Common sense tells us that this mechanism forestablishing or re-establishing equilibrium is not a figment devised as an exer-cise in the pure logic of economics but actually operative in the reality aroundus’ (ibid.). To Schumpeter, the tatônnement is a fiction not per se, but insofar asit neglects the behavioural factors that drive economic development. Hence,the tatônnement ‘leaves out of account many facts that may practically, if notlogically, be just as important as those it includes and even go far towardproducing exactly opposite results’ (ibid.).

This argument is of great importance in relation to the TED. ForSchumpeter, the study of a stationary economy is a first step towards the anal-ysis of other aspects of the economy, such as business cycles and disequilibriumtrajectories. He clearly recognises that in describing both a stationary and agrowing economy ‘we admit only small deviations at the margin, such as everyindividual can accomplish by adapting to economic environment, without


materially deviating from familiar lines’ (1912 [1934]: 81). However, once otherforms of behaviour, such as entrepreneurial behaviour, are included in the anal-ysis, new combinations have to be taken into account and a dynamic analysis ofthe system’s evolution becomes necessary. Hence, Schumpeter clearly embarkson an analytical transition from the theory of general equilibrium with purecompetition to a new, as yet unexplored, approach which is required to describethe working of competition in a dynamic context.

Competition and economic development

When economic development enters into the picture, spontaneous and discon-tinuous changes in the basic data disturb the state of equilibrium. This hasa number of important implications, especially in a competitive economy‘in which new combinations mean the competitive elimination of the old’(ibid.: 67).

The first of these implications concerns the relation of profits to monopolyrevenues. On the one hand, since entrepreneurs do not face competition whentheir new products first appear on the market, the determination of prices forthese new products temporarily proceeds wholly according to the principle ofmonopoly price. On the other hand, new combinations may also lead to apermanent monopoly being established, ‘perhaps in forming a trust which needfear absolutely no competing outsiders. Then, profit is obviously to be consid-ered simply as permanent monopoly revenue and monopoly revenue simply asprofit’ (ibid.: 152). Thus, according to Schumpeter, a monopoly elementcontinues to play a certain role even in an evolving – innovating – capitalisteconomy.

Moreover, Schumpeter notes that the majority of new combinations are notdeveloped in established or old firms, nor do they immediately replace oldercombinations. Rather, new combinations appear side by side with older ones,sooner or later entering into competition with them. This competitive processleads to a fall in prices, putting an end to the boom phase associated with theemergence of new combinations. In other words, it ‘may lead to a crisis, mustlead to a depression, and starts all the rest’ (ibid.: 233). While, in depression,‘the stream of goods is enriched, production is partly reorganised, costs ofproduction are diminished and what appears as entrepreneurial profit finallyincreases the permanent real income of other classes’ (ibid.), these positiveeffects of the competitive process also need to be qualified since ‘[t]the exis-tences, connected with the old business, which are now being competitivelyvanquished, of course suffer’ (ibid.: 247).

Finally, the uneven movement of the economic system renders accuratecalculations based on experience impossible. Not only is every business unitfaced with changes in its customary data, but new competitors now appear onthe scene. Consequently, ‘old customers and dealers fail to appear; the right atti-tude towards new economic facts has to be found … The “mere businessman”faces problems which lie outside his routine’ (ibid.: 238–9). Producers will, thus,


decide in favour of expansion or contraction as the best response to circum-stances ‘without it being possible at the moment to advance trustworthy reasonsfor the one or the other’ (ibid.). This is why producers’ perception of the turningpoint of the cycle is often highly subjective. Schumpeter argues that, ultimately,this struggle ‘must actually lead to a closed approach to an equilibrium position’(ibid.: 243). Of course, the final position never completely corresponds to thetheoretical concept of a system without development, but it constitutes thestarting point for a new phase of expansion.

It is worth noting that the form of competition described by Schumpeter inthe TED takes on an evolutionary flavour when he deals with the processunderlying the economic as well as the social rise and fall of enterprises andfamilies. In the process of competition, the successful entrepreneur destroys oldbusinesses and, hence, also the existences of those dependent on them.Consequently,

there always corresponds to it [the rise in the social scale in the capitalistworld] a process of decline, of lost costs, of elimination. This fate alsothreatens the entrepreneur whose powers are declining, or his heirs whohave inherited his wealth without his ability. This is not only because allindividual profits dry up, the competitive mechanism tolerating no perma-nent surplus values, but rather annihilating them by means of just thisstimulus of the striving for profit which is the mechanism’s driving force;but also because in the normal case things so happen that entrepreneurialsuccess embodies itself in the ownership of a business; and this business isusually carried on further by the heirs on what soon become traditionallines until new entrepreneurs supplant it.

(ibid.: 156)

However, Schumpeter also contends that older firms control buffer quasi rents,because they are embedded in a wider network of firms, have accumulatedresources at their disposal and are often supported by long-term links to thebanking sector. Old firms – ‘that is theoretically all existing ones with theexception of the ones formed during the boom and … those removed fromdanger by a monopoly position, the possession of a peculiar advantage or lastingsuperior technique’ (ibid.: 242) – face three options: first, if they are unable toadapt, they will decay; second, ‘to take in sail and try to survive in a modestposition’ (ibid.); and third, either to move production to another industry or toadopt technical or commercial methods which lower their unit costs. Hence,the fact that new enterprises never completely eliminate older firms distorts thepure evolutionary picture. This is

the reason why a selective process in crises may be spoken of only with animportant qualification, for that firm which is well supported, and not theone that is most perfect in itself, has the best chance of surviving a crisis.

(ibid.: 241)


This feature becomes particularly important when a competitive economy isbeing undermined by the growth of large conglomerates, a development whichSchumpeter judged to be on the rise in most capitalist economies. Under thesecircumstances,

the carrying on of new combinations must become in ever greater measurethe internal concern of one and the same economic body. The difference somade is great enough to serve as the water-shed between two epochs in thesocial history of capitalism.

(ibid.: 67)

On the one hand,

the progressive trustification of economic life facilitates the permanentcontinuance of maladjustment in the great combines themselves and henceoutside of them, for practically there can only be complete equilibrium ifthere is free competition in all branches of production.

(ibid.: 244)

This is also the case when firms, or whole industries, in difficulty receive finan-cial support – for example, government subsidies ‘upon the bona or mala fideassumption that difficulties are only temporary’ (ibid.). Obviously, where largefirms and active industrial policies are predominant, the normal adjustmentprocess will be slowed down. On the other hand, Schumpeter points out thatthe role played by government enterprises or private conglomerates duringdepression

appears from our standpoint as a moderation of the consequences of theswarm-like appearance of new combinations, and as an attenuation of theinflation of the boom and of the deflation of the depression, hence as aneffective means of alleviating the cyclical movement and the danger ofcrises.

(ibid.: 253)

This last remark leads on to Schumpeter’s position on the status of competitionin BC, a contribution that is generally seen to deal with the second ‘epoch inthe social history of capitalism’ – that is, big business capitalism.

Schumpeter begins by noting that ‘the importance of the case of purecompetition does not, of course, rest with the frequency of its occurrence inactual life. A system satisfying its conditions in all its parts has probably neverexisted’. More significantly, he goes on to argue that ‘even if there were no prac-tical instances of the case, it would still retain scientific importance as aninstrument for proving that purely economic logic is capable of determininguniquely purely economic variables’ (Schumpeter 1939: 46, fn.). And he insists


that in a book devoted to the historical analysis of the real dynamics of capi-talism, the concept of perfect competition

would not have been mentioned if it were to serve a purpose no moreconcrete. The case does serve primarily as a steppingstone from which toproceed to more lifelike patterns. But it also held that within the periodcovered by our material it affords a sufficiently close approximation toreality in many cases and that in others the actual patterns, although notfulfilling requirements, yet work in a way not fundamentally different fromthe working of perfect competition.

(ibid.: 46)

He then focuses on the equilibrium mechanism. What really is of interest toSchumpeter is a precise analysis of the conditions under which a real economytends towards equilibrium. Only if this problem can be solved will the conceptbe of any use to the analysis of business cycles. To put it another way,Schumpeter is interested in an economic system that experiences disturbancesbut will strive to return to an equilibrium position.

From this point of view, Schumpeter develops a more accurate critique of theWalrasian device of tatônnement than in the TED. Unlike Walras, he maintainsthat, even in the case of perfect competition, a competitive equilibratingprocess is at work, based on institutional factors and learning mechanisms. Thisprocess works effectively both in the static case as well as in an economy withsmooth growth. Recall, that, in Schumpeter’s analysis, the upswing of a typicalcycle disrupts the prevalent equilibrium, and the adjustment process describedabove continues to function in the downswing.

However, as pointed out early on by Marshall, problems do arise when timeis introduced into the analysis. Schumpeter emphasises, in particular, the roleplayed by technological lags in ‘the setup of a firm, as well as in the economicsystem’ (ibid.: 48). The importance of such lags ‘does not lie in the obvious factthat, since it takes so much time to come about, [full or perfect equilibrium]may fail to come about at all and that, therefore, new disturbances alwaysimpinge on an imperfectly equilibrated system’ (ibid.). Rather, in order to pose aserious problem for the concept of equilibrium, lags must be associated withtechnological change. More specifically, the agents’ reaction to a situation inwhich new combinations are being introduced – that is, in which innovationtakes place – ‘would have to counteract or to reverse that tendency and to leadaway from instead of toward full equilibrium’ (ibid.). Schumpeter makes it clearthat this is not generally the case, except in a boom situation when the ‘neces-sity for intermediate adaptation and for reaction to measures of intermediateadaptation, of course, alters the path the system takes … but does not in itselfbar the way to some equilibrium’ (ibid.). Finally, he points out that lags, andtime in general, will give rise to some form of path dependency in any normalmarket that is not organised in a very specific manner or whose activities arenot concentrated in a single point in time.


Overall, however, Schumpeter believes that the economy will eventuallyreach some competitive equilibrium, or settle in the near neighbourhood ofsuch an equilibrium, since the agents’ reactions to the various intermediate situ-ations that arise are, in general, of a corrective rather than a disruptive nature.Once again, Schumpeter’s argument relies on the stabilising role played bydurable relationships, that is by learning and routinised behaviour:

[E]xperience acquired in dealing with other people and the possibility ofprofiting in each market period from the lessons taught by the precedingones, tend to reduce the practical importance of the pattern under consid-eration and to make results approach those of the Walras–Edgeworthschema.

(ibid.: 50)

Although it is precisely the incessant changes of economic data which createwhat one could describe as indeterminate pricing, this indeterminacy must, ofcourse, be dealt with, but it does not impend the tendency towards equilibriumduring depression.

Schumpeter next discusses the problem of stickiness and rigidities in theeconomy. Obviously, stickiness and rigidities can result from a number offactors. Referring to both real and nominal rigidities, Schumpeter lists theprevailing economic organisation, behavioural inertia, the existence of menucosts as well as of contracts, among such factors:

The reader may think of costs incident to change of occupation or to anyshift from the production of one kind of commodity to the production ofanother kind or quality, or to the exchange … of one asset for another, orof the resistance to change of some prices or the difficulty of adapting long-time contracts or of persuading oneself or other people to act and so on.

(ibid.)

The combined effect of these rigidities or frictions is that of serious distortionsimpeding the formation of a perfect competitive equilibrium. As Schumpeter putsit, ‘if different sectors of the system work with different amounts of friction, as ingeneral they do, lack of harmony will ensue, the more slowly and the more quicklyadaptable elements [get] out of step with each other’ (ibid.: 51). Finally, Schumpeteralso points out that price rigidities, often the result of state intervention, have asimilarly retarding effect on the adjustment process as frictions in general. Hence,Schumpeter here emphasises the role of rigidities and frictions in drawing out theperiod of adjustment and, thus, in lengthening the phases of the cycle. However,he also maintains that frictions do not necessarily play a purely negative role.Quite to the contrary, they can ‘steady adaptation by making it impossible to reactto any disturbance instantaneously and to the full extent’ (ibid. 51). Moreover,those rigidities and frictions, which slow down the adjustment process on thesupply side, may even be


necessary for the economic system to function at all …. Just as the physicalworld would be an uninhabitable chaos if the slightest temperature sufficedto transfer all heat instantaneously to the region of the minimum, so theeconomic world could not function if, for example, the slightest variationin the rate of exchange sufficed to set all gold flowing at once.

(ibid.)

In this context, Schumpeter stresses that modern production techniques oftencontribute to economic instability – a theme he takes up in some detail in thesecond volume of BC. In this book, as well as in later works, he analyses theworkings of the actual competitive process characterising industrial capitalismfrom a historical and empirical perspective. His main contention is that whatprevails is a kind of hybrid competition that cannot be regarded as a deviationfrom, or an alteration of, the basic cases of perfect competition and puremonopoly. Rather, the latter is best described ‘as limiting cases’ (Schumpeter1954: 975). This, then, brings us to Schumpeter’s appraisal of theories of imper-fect competition.

Schumpeter and the theory of imperfect competition

In the History of Economic Analysis (HEA), Schumpeter makes an importantdistinction. If we consider, on the one hand,

that from all the infinite variety of market patterns pure or perfectmonopoly and pure or perfect competition stand out by virtue of certainproperties – of which the most important is that both cases lend themselvesto treatment by means of relatively simple and in general unique deter-mined rational schemata – and on the other hand, that the large majorityof cases that occur in practice are nothing but mixture hybrids of these two,then it seems natural to accept pure monopoly and pure competition as thetwo genuine or fundamental patterns and to proceed by investigating howtheir hybrids work out.

(1954: 975)

This, so the argument goes, is the attitude taken by most theorists of imperfectcompetition. However, Schumpeter then distinguishes this from what heconsiders a more fruitful approach, namely, that

instead of considering the hybrid cases as deviations from or adulterationsof the fundamental ones, we may also look upon the hybrids as funda-mental and on pure monopoly and pure competition as limiting cases inwhich the content of actual business behaviour has been redefined away.

(ibid.)


As we shall see, this distinction sheds considerable light on Schumpeter’s viewof the development of theories of imperfect competition.

Schumpeter quickly deals with the limit case of pure monopoly. By analogywith the case of perfect competition, the case of pure monopoly is characterisedby a determinate outcome and the existence of an equilibrium mechanism, ‘irre-spective even of whether the monopolist sets the price or offers, as it were forauction, the quantity most advantageous to him’ (Schumpeter 1939: 56). Solong as each monopoly is surrounded by a sufficiently large ‘zone of perfectcompetition’, no difficulties arise with regard to determinacy. However, oncemonopolies operate in close proximity and can exert some influence on oneanother, strategic behaviour becomes important. Hence, the situation changesso ‘as to make it necessary for each monopolist to shape his policy with regard tothe policy of one or more of the others’ (ibid.: 57). Schumpeter agrees with JoanRobinson’s view, expressed in the chapter on “A world of monopolies” in her1933 Economics of Imperfect Competition, that the difficulty in this case is not toprove the existence of an equilibrium. Rather, what matters is ‘our inability toprove that there is any tendency for reality to conform to it’ (ibid.). However, hedoes not discuss the question any further, but instead goes on to consider threestandard cases of imperfect competition: bilateral monopoly, oligopoly andmonopolistic competition.

The case of bilateral monopoly, though he judges it to be of little interest initself, has yet for Schumpeter

some bearing on situations which actually arise in the course of the businesscycles: momentary situations emerge that are very imperfectly understoodby the actors on the business stage and often lead to erratic actions more orless conforming to that type. Selling and buying a going concern amidst theexcesses of a violent boom may serve as an example. The only thing we cando, even in less extreme instances, is to replace an equilibrium point by anequilibrium zone.

(ibid.: 58)

In a developing economy, situations and information quickly change.Schumpeter argues, on the one hand, that these changes may, in themselves,provide precisely the information needed to reduce the range of indeterminacy,since ‘temporary necessity, consciously planned strategy and fluctuating antici-pation of the general course of events acquire a very much wider scope than wasassumed in the foregoing analysis’ (ibid.: 59). On the other hand, the range ofindeterminacy is also widened since we are left not only with ‘zones’ rather thanpoints of equilibrium, but with shifting zones. Moreover, in many cases thedemand and supply curves are interdependent. Thus, Schumpeter concludesthat indeterminacy becomes the rule as soon as perfect competition ceases toprevail. However, ‘the equilibrating mechanism does not work thus in vacuo’,but within a specific institutional context.


The extent of indeterminacy increases with oligopolistic competition.Again, Schumpeter regards the pure oligopolistic case as being of very littleinterest. As a matter of fact, ‘[T]his pattern, implying as it does that allcustomers will instantly transfer their allegiance from one firm to another onthe slightest provocation, is of very little interest to us because it is anotherlimiting case which in practice must be rare, if not altogether absent’ (ibid.: 60).This is not to say that oligopoly is uninteresting from the standpoint of puretheory. According to Schumpeter:

The chief performance was Cournot’s and the period’s work may bedescribed as a series of successful attempts to develop his statics of straightmonopoly and as another series of much less successful attempts to developand to correct his theories of oligopoly and bilateral monopoly.


Given that Cournot’s work set off a debate that lasted well into the 1930s,Schumpeter stresses that ‘it should be clear that Cournot’s solution is notabsurd’ (ibid.: 980). However, his main focus is clearly on the real competitivepatterns observed during the business cycle which tend to alter the pure logic ofoligopoly. Schumpeter distinguishes three types of business strategies.

First, firms may ‘attack to kill or cow’. This can result in either precariousmonopoly requiring ‘defensive moves’ or in situations which ‘while stoppingshort of technical monopoly, yet [give] the aggressor more or less completecontrol, the unconquered positions being insignificant or submitted to his lead-ership’ (Schumpeter 1939: 61). The effect of these struggles on the economywill differ according to whether they occur in a phase of expansion whendemand curves shift upwards, or during depression when they shift downwards.Essentially, whatever the nature of the struggle, in a downswing no equilibriumcan exist, ‘a fact which is of considerable importance to the picture of themechanism of business cycles in a society in which big units prevail’ (ibid.).

Second, there may be collusion, whether tacit or explicit. The creation ofexcess capacity is particularly characteristic of this case in which no firm isstrong enough ‘to venture on a fight to a finish’ (ibid.: 62). However, accordingto Schumpeter, a tendency towards equilibrium may still exist in this case, eventhough, due to indeterminacy, ‘the resulting set of values will be different fromany of those that would follow from any other course’ (ibid.).

Finally, firms may try to get rid of product homogeneity or rather to ‘increasethat lack of perfect homogeneity which already exists in most cases’ (ibid.). Thisis obviously a strategy typically associated with monopolistic competition. InSchumpeter’s view, Chamberlinian monopolistic competition is worthy of moreattention than is the theory of oligopoly, since he regards product differentia-tion as being at the heart of modern industrial production:

[I]t comprises not only ‘real’ but also ‘putative’ differences, not only differ-ences in the product itself, but also differences in the services incident to


supplying it … and every device that enables the buyer to associate thething he buys with the name of a particular firm. Differences in locationand other factors which will induce customers to prefer, rationally or a-rationally, one firm to another are of course unavoidable, irrespective ofany intention to create them. And there is simply no such thing as a homo-geneous commodity, motorcar or liver pill.

(ibid.: 63)

Schumpeter concedes that with monopolistic competition ‘we can gain in thedirection of competition some of the ground we thus lose in the direction ofmonopoly’ (ibid.: 64). However, product differentiation has two main implica-tions for the economy. First, it increases the amount of friction and rigidities inthe system and reinforces traditionalistic and cooperative forms of behaviour.Second, firms’ immediate reactions to technological change are very different inthe case of monopolistic competition: ‘This is due to the fact that the possessionof a special market … gives scope for short-time strategy, for moves and countermoves which would not otherwise exist’ (ibid.: 66). Schumpeter mentions, inparticular, the tendency to create excess capacity by reducing output ratherthan prices. Thus, monopolistic competition may sometimes lead to a differentadjustment process ‘characterized by many movements that seem, and sometimes are, erratic, but possibly also by a different equilibrium, if indeed any equi-librium be eventually reached’ (ibid.: 67). Consequently, what is to be expectedis a lessening of the tendency to equilibrium, and equilibrium points will bereplaced by zones or neighbourhoods of equilibria.

Schumpeter’s positive view of monopolistic competition can be easily under-stood in the broader context of his business cycle theory. On the one hand,endogenous and exogenous changes are here seen to induce short-term adapta-tions and produce short-time equilibria, a feature which is also stressed by thetheory of monopolistic competition. Recall further that, during the upswing,innovative behaviour results in an increased variety of goods and of methods ofproduction. On the other hand, new firms producing new products oremploying new methods to produce existing products will, ‘as a rule, try tobehave according to it [monopolistic competition], for that is the obviousmethod of exploiting to the full, and keeping alive, the temporary advantagesthey enjoy’ (ibid.). Consequently, Schumpeter reckons that the theory ofmonopolistic competition has clearly improved our understanding of the mech-anisms underlying business cycles. Moreover, this approach avoids one of themain difficulties of the theory of imperfect market patterns which

neglected or overlooked the fact that, as we leave the case of puremonopoly, factors assert themselves that are absent in this case and vanishagain as we approach pure competition, in other words, that the unbrokenline from monopoly to competition is a treacherous guide.



The recognition of this fact ‘would have wedded pure analysis to institutionalfacts and would have produced a richer and more useful theory of prices’ (ibid.).Schumpeter concludes that during the 1930s and the early 1940s, ‘with a lag ofmore than half a century, we have more or less reached this position, althoughmuch remains to be done. Professor Chamberlin’s work may be mentioned atonce as an outstanding landmark on this road’ (ibid.: 982).

Concluding remarks

It is a widely held view that Schumpeter is outspoken in his preference ofmonopoly and oligopoly over perfect competition as regards the static alloca-tion of resources.6 In our opinion, some qualifications of this view are called for.Evidently, Schumpeter has the highest possible opinion of the dynamic effi-ciency of monopolistic structures and rejects the traditional conception ofatomised firms as the modus vivendi of competition, in particular so inCapitalism, Socialism and Democracy. However, in Schumpeter’s writings, there ismore to competition than its effect on efficiency. In this chapter, we havefocused on the relation between competition and prices as well as market equi-librium formation. It hardly deserves mention that, for Schumpeter, we neveractually come across perfect competition or a perfect equilibrium in real lifewhere, in particular in the period covered by him, stickiness, rigidities and bigbusiness prevail. Most notably in Business Cycles, his historical and statisticalanalysis centres around the actual workings of competition in industrial capi-talism.

However, Schumpeter’s appraisal of the emerging theory of imperfect compe-tition is rather reserved. Whereas he expresses his reservations about theusefulness of the concept of oligopolistic competition, monopolistic competi-tion appears to him to carry more promise, coming closer as it does to his ownconception of economic fluctuations and growth. In a nutshell, for Schumpeter,the true meaning of competitive business

is the scheme of motives, decisions and actions imposed upon a businessfirm by the necessity of doing things better or at any rate more successfullythan the fellow next door; … it is this situation to which we trace the tech-nological and commercial efficiency of ‘competitive’ business, and thispattern of behavior would be entirely absent both in the cases of puremonopoly and pure competition, which therefore seem to have more claimto being called degenerate than being called fundamental cases.

(ibid.: 975)

Thus, we cannot but agree with McNulty’s view that ‘Schumpeter’s defence ofmonopoly and big business has tended to overshadow his insights in thecompetitive process, insights extended beyond those of Chamberlin, becausethey include an appreciation of the importance of changing methods of produc-tion and focus on industrial organisation’ (McNulty 1968: 654). The core of


Schumpeter’s analysis of competition in industrial capitalism is that it is a formof hybrid competition which does not constitute a deviation from the funda-mental cases of perfect competition and pure monopoly, but captures the actualinstitutional and organisational framework of business behaviour.

Notes1 This aspect of the debate on Schumpeter is dealt with in Odile Lakomski’s contribu-

tion to this volume.2 Referring to Value and Capital, Schumpeter notes that as late as 1939 ‘Professor

Hicks was just as convinced that successful theoretical analysis is substantiallyconfined to the competitive case as J.S. Mill had been in 1848’ (1954: 972, fn. 3).

3 Schumpeter remarks that the term ‘pure competition’ was introduced by Chamberlinin his Theory of Monopolistic Competition, but that ‘the substance of the argument, inall essentials, is contained in an unpublished Ph.D. thesis presented in 1927’ (1954:973 fn. 4).

4 He judiciously adds that this approach does not however dispose of all the logicaldifficulties associated with the concept of a competitive market. According toSchumpeter, the first author to express some discomfort with the concept of perfectcompetitive market was H.L Moore in his 1906 article ‘Paradoxes of competition’.

5 However, Schumpeter considers that Marshall was, on the contrary, aware of this.Not only did he point out the ‘trivial truth’ of perfect competition. But he went on‘to show that even if we disregard this truth, we cannot assert that the prices andquantities of competitive equilibrium are necessarily the ones that maximize aggre-gate satisfaction’ (1954: 985) This can be illustrated in cases in which welfare maybe increased by subsidising certain industries.

6 On this point see, for instance, Heertje (1987).

Bibliography

Chamberlin, E.H. (1951) ‘The impact of recent monopoly theory on the Schumpeteriansystem’, Review of Economic Statistics, 33, May: 133–8, Reprinted in J. CunninghamWood (ed.), J. A. Schumpeter: Critical Assessments, Vol. I, London: Routledge, 1991.

Donzelli, F. (1983) ‘Schumpeter e la teoria economica neoclassica’, Ricerche Economiche,XXXVII, 4: 634–90.

Heertje, A. (1987) ‘Schumpeter, Joseph Alois (1883–1950)’, in J. Eatwell, M. Milgateand P. Newman (eds), The New Palgrave: A Dictionary of Economics, Vol. 4, London:Macmillan.

Longhi, C. and Raybaut, A. (1998) ‘Free competition’, in R. Arena and C. Longhi (eds)Markets and Organisation, Berlin: Springer Verlag.

McNulty, P.J. (1968) ‘Economic theory and the meaning of competition’, QuarterlyJournal of Economics 82: 639–56.

Robinson, J. (1933) The Economics of Imperfect Competition, London: Macmillan.Santarelli, E. and Pescarelli, E. (1990) ‘The emergence of a vision: the development of

Schumpeter’s theory of entrepreneurship’, History of Political Economy, 22: 677–96.Schumpeter, J.A. (1912) Theorie der wirtschaftlichen Entwicklung, Leipzig: Duncker und

Humblot. Preface dated Vienna, July 1911. English translation of 2nd edn as TheTheory of Economic Development: An Inquiry into Profits, Capital, Credit, Interest, andthe Business Cycle, Cambridge, Mass.: Harvard University Press, 1934.


—— (1934a) ‘The nature and necessity of a price system’, Economic Reconstruction(Report of the Columbia University Commission), New York: Columbia UniversityPress. Reprinted in J.A. Schumpeter (1989).

—— (1934b) ‘Joan Robinson: The Economics of Imperfect Competition’, Journal of PoliticalEconomy, XLII. Reprinted in J.A. Schumpeter (1989).


—— (1942a) ‘George J. Stigler: The Theory of Competitive Price’, American EconomicReview, XXXII. Reprinted in J.A. Schumpeter (1989).

—— (1942b) Capitalism, Socialism and Democracy, New York: Harper; 11th edn, London:Allen & Unwin, 1966.

—— (1943) ‘Capitalism in the postwar world’, in S.E. Harris (ed.) Postwar EconomicProblems, New York and London: McGraw-Hill. Reprinted in J.A. Schumpeter(1989).

—— (1946) ‘Capitalism’, in Encyclopaedia Britannica IV, London. Reprinted in J.A.Schumpeter (1989).

—— (1949) ‘English economists and the state-managed economy’, Journal of PoliticalEconomy, LVII. Reprinted in J.A. Schumpeter (1989).



Stigler, G.J. (1987) ‘Competition’, in J. Eatwell, M. Milgate and P. Newman (eds), TheNew Palgrave: A Dictionary of Economics, Vol. 4, London: Macmillan.


Part V

Money, banking andfinance

Introduction

The importance Schumpeter ascribes to credit and banking is well known, ifonly because he devotes an entire chapter of The Theory of EconomicDevelopment (TED) to credit and capital. Development, he explains, would nottake place should banks fail to finance innovators’ projects. Numerous passagesof Business Cycles (BC) echo this analysis, in particular those where Schumpeterinsists on how important it is for banks to have full knowledge of the useborrowers make of the loans they are granted. Thus, banks need to appraise,select and follow up investment projects. Moreover, they must respect certainprofessional norms. These ideas are familiar to specialists of modern bankingtheory, as they are cited by Diamond (1984), who traces the insight that bankshave a role to play in the monitoring of borrowers’ behaviour back toSchumpeter.

Leaving to one side the passages on money in the History of EconomicAnalysis (HEA), which do not always convey a clear picture of Schumpeter’sown position on the subject, his writings on money proper are not well knownbeyond the circle of specialists. An exception is ‘Das Sozialprodukt und dieRechenpfennige’, an article published in 1917/1918 and translated into Englishin the 1950s by Marget under the title ‘Money and the social product’.

From these writings, it would appear that Schumpeter treats the analysis ofmoney proper separately from questions relating to the financing of economicactivity by means of bank loans. Only in The Theory of Money and Banking(TMB), a treatise he never managed to publish, does he seek to establish aconnection between these two themes. The story behind the long-delayedpublication of the book has all the attributes of a detective novel and has beenthoroughly investigated and enjoyably recounted by Messori (1996). Thoughnot the most fortunate of choices, Das Wesen des Geldes (The Essence of Money),the title chosen by the editor when the book was finally and posthumouslypublished in 1970, has nevertheless the advantage of drawing the reader’s atten-tion to the book’s central chapter with the same title. It is this chapter withwhich our contribution will be mainly concerned.

Schumpeter’s goal in the TMB is to provide his own approach to money, onethat is distinct from both the ‘metallic’ as well as the ‘cartalist’ theories, which

10 Schumpeter on the institutionof money

Cécile Dangel-Hagnauer

he identifies as the leading paradigms in the history of monetary thought.1 Aswe shall see, Schumpeter emphasises and elaborates on the institutional dimen-sion of money.

Schumpeter tells his story in his usual manner: he sets the stage by consid-ering a simple institutional framework valid for the case of an economy in staticequilibrium. Once the plot starts to unravel and economic development entersthe scene, Schumpeter finds that the monetary concept developed at the firststage of the analysis leaves important questions unanswered. This leads him totake recourse to the institutional features of money which he had not consid-ered in his original framework.

This chapter is organised as follows: the next section briefly recounts thecircumstances surrounding the writing and non-publication of the TMB.Section 3 describes the preliminary theoretical framework underlyingSchumpeter’s conception of money. This static equilibrium framework serves asa backdrop for section 4 which discusses the concept at the heart of the TMB,the ‘critical figure’. Section 5 considers the role played by the institution ofmoney once the economy in no longer in static equilibrium, and the finalsection concludes and suggests a tentative assessment of Schumpeter’s contribu-tion to the theory of money.

The non-publication of The Theory of Money and Banking:missed opportunities

The fact that The Theory of Money and Banking was never published duringSchumpeter’s lifetime can be explained by a series of unfortunate circumstances.As already mentioned, the book was published posthumously under the titleDas Wesen des Geldes. However, it turned out that the editor, F.K. Mann, usedan unrevised and incomplete carbon copy of the initial manuscript containingonly the first twelve chapters. Also, the title chosen for this publication – theheading of its Chapter 9 – was not among the three versions Schumpeterhimself had contemplated.2 That only the first twelve chapters of themanuscript were published (before 1996) goes back to an initial and difficult tojustify decision by Elizabeth Schumpeter and Marget shortly after Schumpeter’sdeath in 1950.3 Fortunately, the three missing chapters were ‘rediscovered’ inthe late 1980s and published in Italian in 1996, thanks to the editorial efforts ofBerti and Messori. Although an edition of the first twelve chapters that wouldbe faithful to Schumpeter’s original manuscript is still not forthcoming, consid-erable progress has been made since the various unfruitful attempts by Margetand Stolper to publish them.

In the opinion of several commentators, in particular Tichy (1984), parts ofSchumpeter’s book on money were written between 1925 and 1930, a period inhis life that was not the most conducive to scholarly occupation, as much of histime was taken up by teaching in order to repay the debts of a bank that hadgone bankrupt while he was its chairman. According to Mann’s reconstructionof events, when Schumpeter left for Harvard in 1931, and later for two short

204 Money, banking and finance

stays in Japan, he entrusted the manuscript to Spiethoff with the explicitrequest to publish it in the event of some misfortune befalling him during histravels. It seems that, upon his return, Schumpeter expressed his intention toabandon all attempts at completing the book. However, a number of footnotesin BC indicate that he went back to work on the manuscript between 1933 and1935. This is confirmed by the fact that the published version contains severalquotations from works published in the 1930s.4

Why did Schumpeter never publish the book? Some commentators haveargued that his alleged intention to give up on the project, and even to set fireto the first version in 1930, was a reaction to the publication of Keynes’s Treatiseon Money. This view is supported somewhat by Schumpeter’s refusal to reviewKeynes’s book for a German journal.5 However, a more straightforward explana-tion is that Schumpeter’s own work was not advanced enough to provide a basisfor an adequate appraisal of Keynes’s arguments.6

The unfinished state of the manuscript is, of course, sufficient explanationfor the book remaining unpublished for so long. Why Schumpeter never actu-ally completed the book is, however, another story that has intrigued manyscholars of Schumpeterian thought. Schumpeter’s letters show that he nevercompletely abandoned his project, but kept postponing the revision of themanuscript while he was writing BC, Capitalism, Socialism and Democracy (CSD)and the HEA. The last letter in which he makes mention of the book indicates,though, that he planned to have it ready for publication by the summer of1951.7

In the form in which it has finally reached us, the manuscript is available intwo fragments, Mann’s 1970 German edition of the first twelve chapters as DasWesen des Geldes (translated into Italian in 1990) and the 1996 Berti andMessori edition in Italian of the last three chapters.8 The introductory chapteremphasises the importance of monetary theory. The second and third chaptersare, in many ways, a preview of Schumpeter’s treatment of money in the HEA,while the fourth chapter examines the workings of a socialist economy.Chapters 5 to 8 introduce the theoretical framework which we will discuss insome detail below and which underlies the core of Schumpeter’s monetarytheory, developed in Chapter 9. The three remaining chapters of Das Wesen desGeldes discuss the theoretical implications of Schumpeter’s approach, while theseparately published Chapters 13 to 15 deal with questions related to monetarypolicy.9

Berti and Messori have made a strong case for retaining the original title ofSchumpeter’s book, i.e. The Theory of Money and Banking (TMB), and in thischapter we follow their example.10

The general setting: banks as social accountants

In the second chapter of the HEA, Schumpeter distinguishes three ‘techniques’– history, statistics and ‘theory’ – which together constitute ‘EconomicAnalysis’. The aim of economic theory is then to construct, by way of

Schumpeter on the institution of money 205

‘generalizing abstraction’, a ‘composite instrument of economic analysis …which functions formally in the same way, whatever the economic problem towhich we turn it’. A few pages later, Schumpeter notes however that ‘theschemata of economic theory derive the institutional frameworks within whichthey are supposed to function from economic history’. This leads him to add afourth technique, ‘Economic Sociology’, to the three previously listed. Theneed for this additional technique arises because, ‘when we introduce … institu-tions, … we are introducing a sort of generalized or typified or stylized economichistory’. Thus, to paraphrase Schumpeter, economic sociology is the disciplinethat deals with the question of how social institutions come to operate as theydo (Schumpeter 1954: 12, 16 and 20–1).11

The proposition we shall try to defend in this chapter is that Schumpeter’stheme in the TMB is located at the frontier between economic theory andeconomic sociology. As already indicated, Schumpeter regards money as aninstitution. Thus, he begins his analysis by constructing a ‘stylized’ institutionalframework within which he embeds the concept of money. Next, he turns toeconomic theory with a view to developing a ‘science of money’(Geldwissenschaft) (Schumpeter 1970: 15). This he accomplishes by establishinga connection between the concept of money as an institution, as outlined in thefirst part of the book, and the ‘composite instrument of economic analysis’ heregards as the most advanced, namely, Walrasian general equilibrium theory.This leads to a theory of money that applies to the circular flow. However, assoon as Schumpeter departs from the framework of the circular flow, not onlydoes the Walrasian approach no longer apply, but what is required is also a re-examination of his initial conceptualisation of money as an institution. This heaccomplishes by referring to stylised historical forms of monetary institutions.

It is important to emphasise that Schumpeter does not open his analysis witha discussion of the historical forms in which money has existed as, in his view,this would be a misleading way of organising the debate. He argues, in partic-ular, that the fact that gold and silver have been a prevalent form of moneythroughout history may lead scholars to identify the nature of money with itsmetallic form. Moreover, as gold and silver are commodities, it might beinferred that money is a commodity of some particular kind. This is a view that,according to Schumpeter, clearly needs to be refuted.

This is also why, in his 1917/1918 article, quoting Bendixen, he favours theview that ‘money is a claim ticket and a receipt voucher’ (Schumpeter 1917/18[1956]: 154–5).12 The problem with this definition, though, is that it is some-what deceptive, which explains why Schumpeter will not retain it in the TMB,since the conceptualisation of money as a claim might imply that the nature ofmoney is legal in origin, as suggested by the cartalist school, and by Knapp(1924) in particular. According to Schumpeter, the view supported by thisschool is equally mistaken: money is not ‘a creature of the law’ (ibid.: 160).

The opposition between metallism and cartalism and the rejection of boththese monetary theories is present in all of Schumpeter’s writings, as witnessedby many passages of the HEA.13 Consequently, Schumpeter sets out to develop


an alternative approach. The notion of money as an institution he advances inthe TMB is complex and has many aspects to it, which we shall examine indetail as we move along. The most important of these is that the institution ofmoney provides the economy with a social accounting and clearing system. ForSchumpeter, the market economy is not anarchy. On the contrary, it gives riseto a stream of payments and settlements that can be represented by a vastaccounting and clearing system:

Fundamental to all understanding of economics is the insight that even anexchange economy is a planned economy, and that the capriciousness ofindustrialists and merchants who, as legal subjects, are the sole masters oftheir destiny, is thus controlled effectively by the social conditions whichprevail in the economy, and which are experienced by these industrialistsin the form of expected profits and impending losses such that an outsideobserver would believe their actions to be guided by some determinateinstructions. Similarly – and more particularly as a result of this – all indi-vidual settlements and accounts, all payments and receipts, merge into theorganic whole of a social account which is no less real because it does notactually exist as a central bureau of accounts.

(Schumpeter 1970: 125, italics in the original)14

This notion of a social accounting and clearing system is borrowed from E.Solvay and is already present in Schumpeter’s 1917/1918 article. However, inthe TMB it becomes an integral part of the basic framework for ‘this notionserves us, above all, to explain the essence of money as a social institution’(Schumpter 1970: 206). It is worth noting here that this system requires a unit ofaccount and a means of payment for accounting and clearing purposes respec-tively. This corresponds to Schumpeter’s view according to which money hastwo functions – that of unit of account and that of a means of payment.

However, given Schumpeter’s intention to provide an ‘economic theory’ ofmoney – a ‘science of money’ – merely to assert that money is an institutionwould miss the point. Hence, in order to build his theory, Schumpeter goes backto two basic and related concepts underpinning his general approach, thestationary circular flow and the social product. In Chapter 5 of the TMB, hediscusses the case – or ‘fiction’, as he calls it – of simple reproduction, referringexplicitly to Marx: the economy is in equilibrium, there is no growth, nor doesan development process take place. The time interval considered is equal to theperiod of production, during which the social product, the sum of goods andservices for consumption, is produced and consumed. All means of productionlast one period. Both types of goods (for consumption and production) resultfrom the ‘productive services’ of labour and ‘nature’ (ibid.: 113).

In the 1917/1918 article, consumers receive a certain amount of monetaryincome in the form of ‘receipt vouchers’ in exchange for their participation inthe production process. As, at this stage, there are no savings, all receiptvouchers are exchanged for consumption goods.15 Thus, these vouchers are, at


the same time, a ‘claim ticket’ in that they entitle their holders to a certainshare of the social product. This implies that in a market economy productionand distribution take place simultaneously. Money is thus a technical ‘device’that allows this twofold process to unfold.

The same idea appears in the TMB, but is now expressed in terms of thesocial accounting and clearing system managed by banks, or rather by theeconomy’s ‘social accountants’.16 Each economic agent holds a current accountthat is credited with their contribution to the creation of the social product anddebited with the share of it they consume. Banks are, thus, important players inthe economy. In fact, Schumpeter includes them in the group of what he calls‘primary agents’. As might be expected, this group comprises ‘households’ and‘firms’, but also ‘commercial banks’ as well as the ‘central bank’ (Schumpeter1970: 128). At this stage, the latter’s role is merely to keep the accounts of thecommercial banks in the social accounting system.

Obviously, as Schumpeter himself points out in the passage quoted above, nosuch pure accounting system actually exists in practice. He believes, however,that all monetary systems operate in a way that is consistent with the principlesjust outlined. More importantly, though, the notion of such a system is mean-ingful and relevant for understanding the nature of money.

Hence, the social accounting and clearing system replaces the system ofreceipts and claims of the 1917/1918 article. At the end of the period of produc-tion, all accounts clear. Moreover, as the stationary flow describes an economyin equilibrium, not only is the overall balance nil, but individual accounts, too,will balance over the period of production. However, an important feature ofthis framework is that, at different points in time during the period of produc-tion, temporary imbalances will occur. In order to settle these, means ofpayment are required. In addition, the economy’s social accountants need toknow what numerical value to assign to the quantities circulating in the system.Thus, what is also required is a unit of account.

The reference to equilibrium: the critical figure

In the economy depicted by the stationary flow, ‘money [is] only of technicalsignificance for the market’: it is a ‘device’ to facilitate the payment of thefactors of production and the distribution of the social product (Schumpeter1917/18 [1956]: 154–5). However, this observation made by Schumpeter in his1917/1918 article provides only a partial account of the role played by money ina market economy. This, at least, is how he sees it in the TMB. One question,in particular, which deserves our further attention is the actual function of thesocial accounting and clearing system in the organisation of a market economy.Contrary to the impression they convey, the ‘industrialists and merchants’ oper-ating on the market do not really act in accordance with a set of ‘determinateinstructions’, nor does the existence of a social accounting and clearing systemimply anything of the sort since it is not the equivalent of a central planningagency in a socialist economy. In fact, economic agents are driven by their own


interests and, more specifically in the case of firms, by the prospect of making aprofit. The question then arises as to what extent the management of individualagents’ accounts actually contributes to the working of a market economy.

To answer this question, we must first return to Schumpeter’s definition ofthe stationary flow referring to an economy that has reached a stable state ofequilibrium. As Schumpeter writes, this economy is described by a uniquelydetermined system of simultaneous equations – that is, a system comprising thesame number of equations describing agents’ behaviour as variables to be deter-mined. This system solves for quantities of goods and exchange ratios which aremutually compatible (Schumpeter 1970: 114). Thus, the cohesion of themarket economy results from the behaviour of individual agents acting withinthe framework of Walrasian general equilibrium. At this stage, this organisationowes nothing to money. Even so, Schumpeter does, in effect, establish a linkbetween his conceptualisation of money and an economy in general equilib-rium, as a careful reading of Chapter 9 of his TMB, entitled ‘The essence ofmoney’, reveals.

This chapter contains an analysis of the functions of money. It opens withthe assertion that the essence of the institution of money is best understood byreference to the social accounting system. Next, the author sets out to explainthe purpose of a means of payment, ‘setting aside, for the present moment, thequestion of assigning a numerical value to the size of payments and receipts, asthey materialise and vanish again in the economic process’ (ibid.: 206–7).

For Schumpeter, a means of payment is defined neither by the particularform it takes nor by its legal meaning. From an economic point of view, moneythat is legal tender is no more a means of payment than, say, a bill of exchange.What matters for the understanding of monetary relations is that they originatein a stream of payments and receipts which, over time, cancel out but which, inthe meanwhile, give rise to temporary imbalances that need to be settled. Forinstance, in the course of the productive process, some agents, and firms inparticular, will be faced with temporary debts. Once again, this is best under-stood within the framework of the social accounting and clearing system thatcan be used to show how a large proportion of debts and claims cancel out.Thus, in a monetary system where gold coins are legal tender, agents running adeficit can issue bills of exchange. As the productive process unfolds and thebills circulate among agents, a large proportion of the debt represented by billsof exchange will cancel out, requiring that only the remainder be settled by adeterminate amount of gold coins in circulation.

As we shall see, Schumpeter repeatedly acknowledges the practical advan-tages of relying on commodity money, and gold in particular. One suchadvantage is that this form of money might conveniently provide a concreteanchor for the unit of account, ‘insofar as association with the market value of acertain quantity of metal – or the quantity of any other good’ could enable‘households and firms … to envisage [the unit of account] as somethingconcrete’ (ibid.: 213). However, Schumpeter discards this idea and argues, quiteto the contrary, in favour of ‘the absolute logical autonomy of the unit of


account from any notion of “something having a value” ’. The line of reasoningthat leads to this assertion will, so Schumpeter maintains, ‘at the same time,lead us to the essence and peculiarity of the institution of money’ (ibid.: 214).This is also our next subject.

Interestingly, Schumpeter relates the need for a unit of account to a centralproblem in economic theory. Thus, at one point in his argument, he ratherabruptly raises the problem of the determination of prices and quantities withina general equilibrium framework. The fundamental problem of economictheory, he argues, is to establish whether such an equilibrium exists and, if so,whether it is uniquely determined (ibid.). For the sake of simplicity, Schumpeterconsiders a basic exchange economy in which n individuals trade in m goods.There are thus m�mn magnitudes to be computed. However, the system onlycontains m�mn�1 independent equations. Therefore, although mn quantitiescan be computed, it is only possible to determine m�1 prices. Thus, whiledetermining quantities and relative prices, the economic process left to itselfleaves absolute prices indeterminate. What is, therefore, lacking is a unit ofaccount, which must be provided from outside the system of equations. That is,it must be chosen arbitrarily by some external actor.

It is worth noting that Schumpeter raises two problems here which call for asingle solution: a unit of account is required both for the general accountingsystem to function and to determine absolute prices. In this way, Schumpeterrelates the problem of the workings of the institution of money to a basicproblem of economic theory.

That the mathematical system describing the Walrasian framework does notpermit the determination of absolute prices implies that one equation is lacking.Accordingly, Schumpeter’s solution consists simply in introducing an additionalequation which defines the unit of account. What he suggests is that ‘we set anyone nominal economic magnitude equal to a figure to be chosen arbitrarily’ (ibid.: 217,emphasis in the original). More specifically, it is the social product that ‘shall beequated to this arbitrary figure’ (ibid.). This boils down to the following equa-tion: �piqi = M, where qi are the quantities of consumer goods and pi their prices.The ‘arbitrary’ figure plays a crucial role insofar as, in addition to permitting thedetermination of absolute prices, it provides the social accounting and clearingsystem with a standard by which to express the quantities that are recorded bythe social accountants. As this figure is not generated by the Walrasian systemitself, it ‘has to be introduced into it from the outside through an act of choice’(ibid.). To emphasise its importance, Schumpeter calls this figure the ‘criticalfigure’ (die kritische Ziffer).17 He argues that it is by way of introducing thisconcept into the analysis that a basis can be provided for understanding theinstitution of money. This is all the more so as the task of setting the criticalfigure is entrusted to a specific institution, namely the central bank whose role,at this stage of the analysis, is thus limited to the act of assigning an arbitraryvalue to some economic magnitude such as, for instance, the social product.18

Thus, the institutional nature of money appears with more emphasis here.The arbitrary nature of the unit of account means that there is a wide range


of economic magnitudes in which absolute prices can be expressed (ibid.:216–17). Although Schumpeter’s preference is for defining the unit of accountby setting the critical figure equal in value to the social product, he insists thatother magnitudes could do the trick as well.19

As always with Schumpeter, the important question is what happens oncewe abandon the setting depicted by the stationary circular flow. Before tacklingthis question, though, it is worth recalling that Schumpeter’s reference togeneral equilibrium is somewhat anecdotal.20 What, basically, he has in mind isthe interdependence of markets, as emphasised by Marget in the followingpassage which also points to the particular connection between money andmarkets found both in Schumpeter and Walras:

Viewed as a system of flows of money expenditure and of objects soldagainst such expenditure, the Walrasian system is to be thought of as asystem of inter-dependent markets. It is, moreover, the flow of money whichestablishes the inter-dependence between the markets. It is money whichentrepreneurs disburse to create income to the factors of production; it istherefore money which the factors of production, in their turn, disburse outof income in purchase of the products of industry. Money, in short, is themeans whereby a link is established in time between the successive discreterealized events of the economic process.

(Marget 1951: 181, emphasis in the original)

The emphasis of this passage is, however, on the means of payment function ofmoney and not on that of the unit of account. This is understandable as Margetrefers only to Schumpeter’s 1917/1918 article and some of his other works, suchas BC, and given that he was presumably not as yet familiar with the contentsof the TMB at the time he made the above comment. As we shall see, the roleof means of payment is, in fact, more specifically dealt with as part of the‘monetary methods’ Schumpeter considers, once he abandons the fiction of thestationary flow.

Money in a changing economy

While the reference to the social accounting and clearing system and to thestationary circular flow provides the grounds for conceiving of money as aninstitution, it is by gradually considering the real conditions affecting the way inwhich the economic process unfolds and is transformed under the impact ofdevelopment that Schumpeter actually fills this concept with life. However, asSchumpeter quickly realises, this departure from the initial framework alsoraises problems when it comes to defining the critical figure. This does notmean, however, that the monetary concept which applies to the stationary flowcannot also be used to understand money in a developing economy. Had thisbeen the case, Schumpeter’s approach would have ended in failure. Instead, thedifficulties that arise are seen to be of an objective nature insofar as they have to


be faced not only by economists but by society itself which will, therefore, equipitself with ‘monetary methods’. However, because these methods have theirimperfections, money always imposes some form of constraint on the economy –what Schumpeter calls the ‘money tie’. It is thus not neutral. Yet, there are waysof relaxing this constraint, namely, by relying on credit. In this section, weanalyse these points in more detail.

Modifying the initial analytical framework

As Schumpeter repeatedly reminds us, a capitalist economy never actually takesthe form of the stationary flow. This has two consequences. First, what weobserve when looking at the real world are not equilibrium values. This raises apreliminary problem in that, when Schumpeter defines the critical figure andsuggests that the central bank set it equal in value to the social product – sothat Σpiqi = M – pi and qi are equilibrium values of the prices and quantities ofconsumer goods. Now, what the central bank observes are, in fact, the actualquantities and prices of goods that determine consumer spending. It cannot,therefore, provide the economy with a critical figure that would have anyrigorous meaning. Moreover, since the central bank is not the equivalent of acentral planning bureau in a socialist economy, it will not actually possess allthe information required to enable it to set the critical figure at its equilibriumvalue. But even assuming the central bank were in a position to accomplishsuch a feat, doing so would make little sense inasmuch as the usual state of theeconomy is not one of equilibrium. To complicate things further, in aneconomy out of equilibrium, to make the necessary corrections would alsorequire that goods that have been produced but not sold be accounted for, aswell as those that would have been offered for sale had the economy been inequilibrium.

Second, outside the stationary circular flow, the economy is no longerdescribed by a simple reproduction scheme. To show what this implies, whilekeeping the argument simple, Schumpeter considers an economy that hasmoved from equilibrium A to equilibrium B. The problem, he argues, would notbe serious, if it were possible to regard the two states as independent systemssince, in this case, all that was required would be to define two critical figures,one for each equilibrium state. However, such independence presupposes thatthere be no interrelations between the components of the two economies. As amatter of fact, such relations do exist and they are important. For instance, theentries recorded on a particular date by the social accounting and clearingsystem refer, more often than not, both to economic activities originating in thepast as well as to others that will unfold in the future.

Hence, Schumpeter concludes that there is a need for a unit of account thatcan handle these difficulties and remain functional in dynamic conditions –that is, once the passage of time is explicitly taken into account. This does not,however, require that the critical figure will always ‘mean the same thing’(Schumpeter 1970: 220). A simple example can help us understand what


Schumpeter means by this. Assuming that the critical figure has been definedpreviously by setting the value of the social product equal to 1,000 and thatrelative prices remain unchanged while the quantity of goods doubles, to keepthe measure of the social product constant (i.e. at 1,000) would imply that thelevel of absolute prices be halved. Even in this very simple case, this seems quiteimpractical if not absurd. Thus, Schumpeter argues, the central bank must, onthe contrary, ‘adapt the critical figure continuously to the changing economicbody’ (ibid.).

It appears, then, that the central bank faces an impossible task: out of equi-librium, it is already quite impossible to define the critical figure, and it is evenless possible to adapt the figure to the changing circumstances of an evolvingeconomy: ‘[O]nly in a totally rationalised civilisation could such a “pure”accounting system operate’ (ibid.: 221).

At this point, the reader begins to wonder how Schumpeter can possiblyextricate himself from what appears to be a dead end in his argument. In fact, itis through a closer examination of the workings of the institution of moneythat he shows how society has contrived of at least a partial solution to thedilemma:

Even though [life] has not resolved the problem of a unit whose substanceremains stable over time, it has, on the other hand, in its own way resolvedthe problem of how to set the critical figure and of interlinking the criticalfigures associated with different states of the economy, and the monetarymethods that have been devised for this purpose constitute the essence of the socialinstitution that we call money.

(ibid.)

The institutional forms of money: monetary methods

It is thus by returning to the institutional nature of money that Schumpetersolves the problem he has raised. As a matter of fact, the monetary methods herefers to are the stylised historical forms of the institution of money. Out of thewide variety of such forms of money, Schumpeter focuses on two, in particular.The first method he investigates is not the first to have emerged historically.Rather, it displays a close affinity with the ‘pure’ or ‘ideal’ accounting andclearing system that we have discussed above. To explain how this methodworks, the author begins by considering an economy in static equilibrium. Theagents hold current accounts. The critical figure is defined by assigning a valueto the number of accounting and clearing units that are necessary to settle thetemporary imbalances generated by agents’ transactions. These units are then‘embodied in physical tokens’ – banknotes – that circulate among agents,enabling them to settle their accounts (ibid.). Interestingly thus, by choosing aparticular monetary method, the central bank simultaneously defines the unit ofaccount and the amount of means of payment that serve to settle the temporaryimbalances: ‘Every such method creates reckoning tokens [Rechenpfennige]


which exist as such – either physically or in accounts. We call these reckoningtokens money’ (ibid.: 224).21

In an economy in static equilibrium, the reproduction of temporary imbal-ances follows an invariant pattern and the numerical relation between thequantities these imbalances represent and all the other economic magnitudesremains constant. The critical figure and the value of these economic magni-tudes can, thus, be set once and for all. In sharp contrast, when quantities andrelative prices vary over time, maintaining an identical numerical expression forthe imbalances is no longer practical nor feasible, and would, in fact, produceodd, if not absurd, results. One could compare this to a situation in which, inorder to find out how tall people are, a measurement system was adopted inwhich height was always set equal to, say, two metres. Hence, the task of thesocial institution in charge of managing the critical figure consists in modifyingthe number of units in accordance with the changes occurring on the marketsfor goods, a task it can only fulfil imperfectly.22

This task is facilitated in the case of the second monetary methodcontemplated by Schumpeter. Here, the institution responsible for managingthe critical figure sets the price of one commodity, say gold, at an arbitrarylevel (ibid.: 222). Both the numerical determination of the critical figure andits modification are achieved by altering the quantity of gold ‘contained’ inthe unit of account. However, keeping constant the price of this commodityalso provides an adequate solution to the problem of adapting the criticalfigure to a changing economy (ibid.: 222–3). This is precisely what makescommodity money special, or, as Schumpeter puts it, ‘a stroke of genius ofcivilisation’ (ibid.: 224). However, there is a major drawback to this methodinasmuch as it subjects the economic magnitudes to the variations of supplyand demand for the commodity that is the unit of account. To highlight theimportance of this drawback, Schumpeter compares the method ofcommodity money to a technique which would calculate people’s heightusing an elastic yardstick. Such a method provides no more than an ‘indirect’way of setting the critical figure because, like all monetary methods encoun-tered in history, including Schumpeter’s first method mentioned above, itcontains an element that is exogenous to the social accounting system (ibid.:225). In the case of commodity money, this exogenous element is preciselythe fact that economic magnitudes are subjected to the uncertainties affectingthe supply and demand for the commodity that is money (ibid.: 223). It ishowever this indirect and ‘absurd’ method that defines the essence of money(ibid.: 224).

Consequently – because it is impossible for the central bank to adapt thecritical figure perfectly to the dynamics of the economy – money is notneutral. Schumpeter states it thus: ‘Any such method subjects economicmagnitudes to a new condition to which they have no choice but to adapt.This condition we call the money tie [das Geldligamen]’ (ibid.: 224).23


The money tie and the role of credit

Thus the way in which the institution of money operates inevitably creates aconstraint in the form of the ‘money tie’, a ‘checkrein’ that is ‘buckled on theprocess of clearing and accounting’ (ibid.: 227).

This is self-evident in the case of both paper and metallic currencies when,as a consequence of changes in the volume of production as well as in the struc-ture of the production process, the quantity of notes and coins in circulationbecomes inadequate, as Schumpeter points out using the following example:

Even if everybody had the same inspiration and would henceforth supplyand demand at ‘double prices’, so long as gold coins or bits of paper were tobe used as a means of payment in some segment of the economic sphere, inthat segment a certain quantity of goods could not be sold at the newprices. This is precisely what constitutes the jerk of the checkrein.

(ibid.: 227–8)

There are, however, ways to escape the grip of the money tie: ‘As a matter offact, economic life rebels against the money tie … and, to some extent, itmanages to extricate itself from it’ (ibid.: 227). One route of escape is providedby the different forms of credit (Guthaben) and, in particular, bank credit.

According to Schumpeter, if bank credit is generalised throughout theeconomy, it creates a system akin to the pure system of accounting. Commercialbanks thus have the power to relax the constraint money imposes on theeconomy, and to provide economic agents with the flexibility they require asthe economy develops and undergoes innovation. As Schumpeter notes, byissuing additional means of payment, banks manage the critical figure not ‘indi-rectly’, but ‘directly’ (ibid.: 229–30). Although somewhat vague, this remarkreminds us that banks are regarded as a constitutive component of the economyinsofar as they are the economy’s social accountants: as the managers of thesocial accounting and clearing system, they have a ringside seat that permitsthem to provide agents directly with the means of payment they require.

However, as the history of money and banking proves, their capacity toinfluence the critical figure invests banks with such power as is open to thepossibility of abuse. Therefore, some sort of regulation of the banking sector isrequired, and monetary policy will tighten or loosen the money tie according tocircumstances (ibid.: 230). This, then, is an additional reason why the agency incharge of managing the critical figure must be an institution external to thesocial accounting and clearing system.24

The economic role Schumpeter assigns to credit is well known. In the TED,Schumpeter shows how the ad hoc creation of means of payment by banksallows the transfer of means of production from established to innovative activi-ties. The TMB does not have much more to say on this subject. Schumpeter iscontent with recalling his theory of interest in broad terms, and in particular


the determination of the rate of interest on the money market, of which banksare a part. 25

Concluding remarks

If the TMB is to be considered as an attempt, on the part of Schumpeter, toprovide an original contribution to the theory of money and credit, what wehave to assess is the relevance of the concepts he introduces. As we have seen,it is within the framework of the stationary flow and Walrasian general equilib-rium that Schumpeter introduces the first of these concepts, the critical figure,in a way he regards as rigorous. It is worth noting that, even then, some factorexternal to the static equilibrium setting is required to determine absoluteprices. Thus, we are already at the frontier of economic theory, since theWalrasian system provides no internal explanation of money. In fact, what weencounter here is a manifestation of the as yet unresolved problem of inte-grating money into the general equilibrium framework. Interestingly,Schumpeter suggests that the solution to this problem can be found by payingexplicit attention to the institutional nature of money.

However, what has led many commentators to argue that Schumpeter fails inhis attempt to provide a new approach to money is that, when he shifts fromstatic to dynamic analysis by studying an economy that moves from one state ofequilibrium to another, he finds – and seemingly deplores the fact – that it is nolonger possible to define the critical figure accurately.26 In fact, the problemSchumpeter raises here is that, once the economy is set into motion, the criticalfigure can no longer be determined in terms of the ‘pure’ logic of economictheory. This assigns considerable importance and responsibility to the institu-tion in charge of defining the critical figure, whose interventionist role containsnot only an element of arbitrariness, but is also very demanding in that thisinstitution is expected to adapt the critical figure to the economy as it movesalong and changes in size and structure. The ‘monetary method’ – the secondconcept Schumpeter introduces – that such an institution chooses to imple-ment in order to set the critical figure ‘indirectly’ will affect the way in which itwill manage the provision of ‘reckoning tokens’.

All of this involves issues pertaining to the field of economic sociology.Thus, the institution of money can be interpreted as an analytical deviceSchumpeter introduces in order to solve a problem that pure economic theorycannot deal with. It must then be recalled that Schumpeter’s definition ofeconomic analysis implies that, where economic theory proper is at the centreof the analysis, institutions have to be ‘taken for granted’, just like the‘hypotheses or axioms or postulates or assumptions or even principles’ under-lying it (Schumpeter 1954: 15). However, economic reasoning enables us tograsp one essential aspect of money, namely, that in the form of the unit ofaccount it cannot be derived from within the general equilibrium frameworkand that it must be included alongside the ‘institutional frameworks withinwhich [the schemata of economic theory] are supposed to function’ – like


‘government, property inheritance, contract, and so on’ (ibid: 20–21). The taskof economic sociology then becomes that of analysing money as a stylised insti-tutional fact and of describing the working of the social institution we callmoney. Only on this basis can we then go back to history and make some senseof prior forms of money, and only then can we understand the organisationaladvantages of commodity money, despite the fact that there is no theoreticalreason why money should take the form of a commodity. Money thus offers anemblematic illustration of an economic concept that can only take on its fullmeaning once it has been carefully scrutinised with the help of the tools andmethods provided by economic theory, economic sociology and history.

Hence, the main conclusion to be drawn from a re-examination ofSchumpeter’s approach is that there can be no self-contained discourse onmoney based exclusively on economic theory. Rather, Schumpeter providesevidence that money is part of the institutional framework within which theeconomic process unfolds. As such, money constrains the behaviour of theeconomic agents. This is why they are subjected to the ‘money tie’, the thirdconcept introduced by Schumpeter. However, as he himself tells us, agents havealso managed to devise ways of relaxing this constraint, in particular byresorting to credit. The institution of banking and credit is thus intimatelyconnected with the institution of money.

It is worth noting that the mere existence of the money tie means thatmoney is not neutral. This brings us to Schumpeter’s view of the role of mone-tary policy. One of the tasks of the central bank is to monitor commercial banksand where necessary to limit their capacity to grant loans. This is because,fundamentally, Schumpeter is concerned with the value of money – that is,with its purchasing power – as is evident from both his 1917/1918 article andthe TMB. Thus, Schumpeter is not indifferent to the question raised by thequantity theorists. However, he is not a supporter of their approach. The reasonis not so much that, in his view, the supply of money is endogenous, eventhough one could be forgiven for reaching this conclusion on the basis ofSchumpeter’s analysis in the TED. However, this reading of Schumpeter’s posi-tion towards the quantity theory of money ceases to be convincing in view ofhis 1917/1918 article which examines at great length – in actually over morethan half of the text – the effect of the quantity of money on prices. Moreover,in the TMB credit is not the starting point of the analysis. Rather, the authorbegins his analysis by developing the notion of a general accounting andclearing system in which each agent holds a current account. Credit, or ratherthe role it plays in the TED, is introduced only towards the end of the analysisas a means of loosening the money tie.27

All this explains why Schumpeter believes that the quantity theory of moneyis irrelevant. Time and again he stresses that it is no more than a ‘theorem’which fails to explain anything. It describes an economy in static or stationaryequilibrium, whereas the important and interesting problems are those whicharise when the institution of money faces an economy undergoing continuouschange. To try and keep absolute prices constant is then obviously impossible,


although the focus on the level of prices remains important, at least from atheoretical point of view.28 Thus, Schumpeter’s main concern is with the needfor a nominal anchor in order to determine absolute prices, an important issuefor modern economists confronted with a world that has abandoned the goldstandard for good.

Notes1 The terms ‘metallism’ and ‘cartalism’ are borrowed from Knapp (see Schumpeter

1954: 288 fn. 2).2 See Messori 1996: XXVIII and XXXVIII. In fact, Mann could have chosen either

the initially projected title, Geld und Währung (Money and Currency), or the titleSchumpeter alludes to in BC, i.e. his ‘treatise on money’ (e.g. p. 75) or the final titlehe used, The Theory of Money and Banking, when referring to the book in the lastletter in which it is mentioned, dated 19 November 1949 and addressed to S.E.Harris, the editor of the Harvard Economic Handbooks where the book would havebeen published had it been completed (see Messori 1996: XVII).

3 No such justification is found by Messori 1996 (see pp. XLIII–XLV).4 Two quotations are from works published in 1936 and 1943 respectively (Messori

1996: XV, fn. 13).5 Tichy 1984: 136.6 Messori 1996: XIX.7 As evidenced by the letter to Harris.8 Translations of some chapters are available in English and French: see Schumpeter

1991 for a translation of the first two chapters into English, as well as Schumpeter1998 and 1999 for translations into French.

9 The titles are ‘International’, ‘Problems of the monetary standard’, ‘Inflation, defla-tion, reflation’, in accordance with an index presumably provided by Marget, whothus appears to have been perfectly aware of their existence (see Messori 1996:XXII–XXIII and 3).

10 The main reason being that this is the title Schumpeter had suggested to Harris in1949 (see note 2 of this chapter).

11 Capital letters are the author’s, ‘theory’ and ‘generalizing abstraction’ are in quota-tion marks in the text. For an in-depth analysis of the relationship betweeneconomic analysis and economic sociology, see the introduction to this volume aswell as Legris’s contribution.

12 Throughout this chapter, page numbers for this reference relate to the English trans-lation (1956).

13 See, for instance, Schumpeter 1954: 288–99.14 The author of this chapter would like to address special thanks to S. Bankenburg for

helping her with the translation into English of the quotes from the German editionof Schumpeter 1970, and of this passage in particular. The usual disclaimer applies.

15 See Festré’s chapter in this book on ‘Innovation and business cycles’ for an accountof what Schumpeter means when he considers there are no savings.

16 As Stiglitz and Weiss (1988) have fittingly called them.17 We have chosen to follow Shah and Yeager (1994) and translate kritische Ziffer as

‘critical figure’, although ‘critical number’ or ‘critical value’ might have been moreappropriate. In the Italian translation, Dal Bosco uses indicatore critico.

18 In a footnote, Schumpeter indicates that the institution setting the critical figurecan also be the result of a ‘social habit that has emerged gradually and “sponta-neously” ’ (1970: 219 fn.). However, he does not explain how the numerical value orthe critical figure would then be determined, which, as we shall see, is a question ofcrucial importance.


19 This preference goes back to the 1917/1918 article and is related to the fact thatonly on the market for consumption goods is money a ‘claim ticket’ and a ‘receiptvoucher’. As Schumpeter notes, this is a point stressed by Wieser (1917/1918 [1956]:155). See also Schumpeter (1939: 457–8).

20 For an in-depth analysis on this point, see Arena’s chapter ‘Schumpeter on Walras’in this volume.

21 Schumpeter 1970: 224. See Shah and Yeager 1994: 452, for an English translation ofthe entire passage referred to here. See also Schumpeter 1939: 452–61, where theauthor alludes to the problems he raises in chapter 9 of the TMB. Interestingly,Schumpeter uses the terms ‘unit of calculation and clearing’, ‘unit of accounting andclearing’ and ‘unit of account’ interchangeably.

22 See Schumpeter 1970: 222. Schumpeter suggests in passing that the paper moneymethod could be modified so as to resemble even more the pure accounting system.In this case, the central bank would determine a total amount that would be settledby cheque. The same problems would arise, however, as soon as allowances neededto be made to adapt to economic change.

23 Schumpeter writes Geldligamen (which the editor changes into Geldligament in theindex). Again, we have chosen to follow Shah and Yeager and translate this notionas ‘money tie’. Dal Bosco has chosen legame monetario (monetary link).

24 There is an obvious contradiction in Schumpeter’s argument: on the one hand, thecentral bank is regarded as a constitutive component of the economy; on the otherhand, it is external to the social accounting and clearing system.

25 For a detailed analysis of these aspects, see Arena and Festré 1996.26 See, for instance, Hutter 1998. See Messori (1996: XIX and CI) for an account of

appraisals by other authors. Also see Tichy’s (1984) conclusion.27 For a slightly different point of view, see Lakomski 1999.28 In the first chapter of TMB, Schumpeter raises the issue from the point of view of

monetary policy without really taking sides. Thus, throughout his work his attitude isconsistent: see, for instance, BC: vi.

Bibliography

Allen, R.L. (1991) Opening Doors - The Life and Work of Joseph Schumpeter, 2 vols, NewBrunswick and London: Transaction Publishers.

Arena, R. and Festré, A. (1996) ‘Banks, credit, and the financial system in Schumpeter:an interpretation’, in L.S. Moss (ed.) Joseph A. Schumpeter, Historian of Economics,London: Routledge.

Diamond, D.W. (1984) ‘Financial intermediation and delegated monitoring’, Review ofEconomic Studies, LI: 393–414.

Hutter, M. (1998) ‘Theories and their metrified environment. The case of money inGerman texts, 1916–1946’, Diskussionspapiere, Heft 56, Fakultät für Wirtschaftswis-senschaft, Universität Witten/Herdecke.

Knapp, G.F. (1924) The State Theory of Money, London: Macmillan. First published inGerman in 1905. Reprinted by Augustus M. Kelley, Clifton, 1973.

Lakomski, O. (1999) Monnaie, Banques et Crédit dans l’Oeuvre de J.A. Schumpeter, PhDdissertation, University of Picardie – Jules Verne.

Marget, A. (1951) ‘The monetary aspects of the Schumpeterian system’, Review ofEconomic Statistics, vol. 33: 112–21.

Messori, M. (1996), ‘Nota ai testi’ in J.A. Schumpeter (1996).Reclam, M. (1984) J.A. Schumpeter’s ‘Credit’ Theory of Money, Riverside: University of

California.


Schumpeter, J.A. (1912) Theorie der wirtschaftlichen Entwicklung, Leipzig: Duncker undHumblot. Preface dated Vienna, July 1911. English translation of 2nd edn as TheTheory of Economic Development: An Inquiry into Profits, Capital, Credit, Interest, andthe Business Cycle, Cambridge, Mass.: Harvard University Press, 1934.

—— (1917/1918) ‘Das Sozialproduct und die Rechenpfennige. Glossen und Beiträge zurGeldtheorie von heute’, Archiv für Sozialwissenschaft und Sozialpolitik, Vol. 44. Trans-lated into English by A.W. Marget as ‘Money and the social product’, InternationalEconomic Papers, 6, 1956.



—— (1970) Das Wesen des Geldes, F.K. Mann (ed.), Göttengen: Vandenhoeck &Ruprecht.

—— (1990) L’Essenza della Moneta, Italian translation by E. Dal Bosco of J.A. Schum-peter (1970), Turin: Cassa di Risparmio di Torino.

—— (1991) ‘Money and Currency’, Social Research (with an introduction by R. Swed-berg), Vol. 58, pp. 499–543.

—— (1996) Trattato Della Moneta: Capitoli Inedeti, L. Berti and M. Messori (eds),Naples: Edizioni Scientifiche Italiane.

—— (1998) ‘L’essence de la monnaie’ (ch. 9 of Schumpeter 1970), with introduction byC. Jaeger, Journal des Economistes et des Etudes Humaines, 8: 283–91.

—— (1999) ‘Le calcul économique dans une communauté socialiste’ (ch. 4 of Schum-peter 1970), with introduction by O. Lakomski, Cahiers d’Economie Politique, 35:89–123.

Shah, P.J. and Yeager, L.B. (1994) ‘Schumpeter on monetary determinacy’, History ofPolitical Economy 26: 443–64.

Stiglitz, J.E. and Weiss, A. (1988) ‘Banks as social accountants and screening devices forthe allocation of credit’, NBER Working Paper, No. 2710.

Swedberg R. (1991) Schumpeter: A Biography, Princeton: Princeton University Press.Tichy, G. (1984) ‘Schumpeter’s monetary theory: an unjustly neglected part of his work’,

in C. Steindl (ed.), Lectures on Schumpeterian Economics, Berlin: Springer-Verlag.


Introduction

It is a fairly commonly held view these days that Schumpeter’s work on moneyand banking is by no means easy to access, making it particularly difficult toarrive at clear conclusions regarding the author’s own point of view. This diffi-culty is compounded by Schumpeter’s extensive work on the history of thoughtin this field, encompassing, as it does, thorough investigations of numerouscontributions to the debate on money and credit by authors writing in differenttraditions of economic thought. Among these, the Wicksellian tradition is ofparticular interest. First, it is from this perspective that Schumpeter’s view ofthe role of bank credit for dynamic economic processes can be best understood.1

Secondly, provided a fairly broad definition of the Austrian school is acceptable,the discussion can be extended to an assessment of the Wicksellian influence,in the field of monetary and business cycle theory, on the Austrian school ingeneral. Although a generation apart and representing different strands, bothSchumpeter and Hayek were part of the Austrian tradition. The strand ofAustrianism associated with Mises and Hayek led to the modern Austrianschool, whereas Schumpeter developed his own version of Austrian economicswhich has been considered as unique ever since. However, as we shall develop,these two strands of the Austrian tradition share a common Wicksellianheritage, even though Hayek’s and Schumpeter’s respective accounts of thedynamic interactions between credit and productive activity appear to be ratherdistinct analytical extensions of the original Wicksellian cumulative process. Tobe more specific, comparing the two authors’ views on the role played by bankcredit through its effects on income distribution – via forced saving – and oninflation during the different phases of cyclical dynamics – impulsion, propaga-tion, reversal – is not only a worthwhile exercise in its own right, but is ofparticular interest with a view to clarifying Schumpeter’s conception of capi-talist dynamics.

In the second section we discuss the Wicksellian origins of Schumpeter’s andHayek’s approaches to money and banking in the context of dynamic economicanalysis. The third section compares the role played by banks and credit inSchumpeter’s and Hayek’s explanation of economic fluctuations. We conclude

11 Money, banking and dynamicsSchumpeter vs Hayek

Agnès Festré

in the fourth section by contrasting both authors’ perception of economicdynamics.

Schumpeter and Hayek on money, banking and dynamics:Wicksell’s legacy

In the aftermath of the so-called ‘marginal revolution’, many economistswriting at the end of the nineteenth and beginning of the twentieth centuriesconsidered the static or stationary state of the economy as the methodologicalstarting-point of their analyses. This could then be extended to account formonetary and financial factors as well as for dynamic movements of theeconomy. In this framework, the introduction of money, bank credit or financialfactors does not affect the essential features of the basic case of a static bartereconomy or a two-commodity exchange. Similarly, dynamics are conceptualisedas representing a higher degree of complexity of the basic setting, triggered offby changes in the ‘fundamentals’.

This view was not, however, shared by all economists. Wicksell, Schumpeterand Hayek adopted a different approach that can be characterised as follows.

First, they do not believe that a monetary economy can be described as amere extension of a barter economy, a line of thought also explicitly taken upby Keynes. Wicksell, Hayek and Schumpeter argue that, once bank credit isintroduced into economic analysis, the working of the economic system experi-ences fundamental changes. Indeed, as soon as one accounts for the existence ofa banking system, money ceases to be exogenous and to represent no more thanthe mere counterpart of real exchanges. The ex novo and endogenous nature ofcredit money modifies the conditions governing the co-ordination of saving andinvestment. While, in a barter economy, both saving and investment aredefined in real terms, in a monetary economy credit money can, to a certainextent, come to act as a substitute for real savings. By prompting changes in themechanisms adjusting saving and investment, it can give rise to global disequi-libria.

Wicksell provided the first analytical attempt to address these issues, hismain contributions in this context being his analysis of the dynamic processesunderlying saving and investment movements as well as of the conditionsrequired to establish a saving–investment equilibrium. The ‘working hypothesis’at the core of his argument is contained in his well-known distinction betweenthe monetary rate – which he takes as a given since it is set by the bankingsystem irrespective of real productivity considerations – and the natural rate ofinterest defined as the rate at which all aggregate savings are invested. Incontrast to the monetary rate, the natural or real rate of interest is likely to varywith investment opportunities based on a rise in productivity. This samedistinction is also present in Hayek’s analysis and, though less clearly so, inSchumpeter’s theory of business cycles and economic development.2 Underlyingit is another more important distinction, namely, that between a cash and acredit economy, which can be found in both Hayek and Schumpeter’s business


cycle theories and which provides the theoretical link to Wicksell’s analysis.Moreover, in all three cases, this distinction is more than a mere pedagogicdevice in that it highlights the important theoretical implications of a shift offocus to dynamic theory.

A second characteristic of Wicksell’s, Hayek’s and Schumpeter’s approachesis that, in contrast to Walras, they interpret the real exchange economy,characterised by a stationary state and full employment, as a preliminarymethodological step, allowing the theoretician to describe and characterise themain tools of economic inquiry before moving on to the core subject ofeconomic analysis – that is, dynamic analysis. The importance given by allthree authors to the stationary state can thus be explained by the specific rolethey attribute to money. The passage from a barter to a monetary economy doesnot in itself imply the emergence of disequilibria. It is possible to conceive ofthe special case of a monetary economy where money is ‘neutral’ insofar as itnever hinders the inner tendency of the economic system to return to equilib-rium. However, not all monetary economies function in this way: money canalso be ‘non neutral’. One sufficient condition of such non-neutrality is theexistence of an organised system of bank credit. Credit creation by banks allowsthe quantity of money in circulation to vary and, consequently, a disequilibriumbetween the money demand of entrepreneurs for investment, on the one hand,and the supply of savings, on the other, becomes a possibility. Wicksell addressesthis problem by introducing a different concept of neutral money. He assumeseither a monetary system in ‘tranquil’ conditions or a ‘properly functioning one’– that is, a state of equilibrium in which savings are invested instantly.3 Thissystem corresponds to what Wicksell describes as a ‘pure cash economy’ inwhich money is neutral, credit is absent and the velocity of circulation is quasistable.4 The extreme other case is that of ‘pure credit economies’ with almostno practical limits to the quantity of money in circulation.5 Wicksell introducesthen the distinction between the monetary rate of interest and the ‘natural’ (or‘normal’) rate of interest. In his own words:

The rate of interest at which the demand for loan capital and the supply ofsavings exactly agree, and which more or less corresponds to the expectedyields on the newly created real capital will then be the normal or naturalrate. At the same time equilibrium must ipso facto obtain … in the marketfor goods and services, so that wages and prices remain unchanged. Thesum of money income will then usually exceed the money value ofconsumption goods annually produced, but the excess of income – i.e.,what is annually saved and invested in production – will not produce anydemand for present goods but only for labour and future production.

(Wicksell [1906] 1967: 192–3)

What Wicksell is arguing here is that the natural rate of interest corresponds tothat level of the rate of interest that would be determined by demand andsupply of capital, if the latter were lent without the mediation of the banking

Schumpeter vs Hayek 223

system. This mechanism is disrupted as soon as bank credit is introduced intothe system as part of the money market. Credit accelerates the velocity of circu-lation and may therefore cause a disequilibrium between the entrepreneurialdemand for cash balances and the supply of ‘accumulating savings awaiting suit-able investment opportunities’.6 In the second volume of his Lectures, Wicksellargues that

[t]he influence of credit on currency may, under all circumstances be regardedas accelerating the circulation of money … The occasions on which creditactually replaces money and thereby renders it superfluous may, quitesimply be regarded as special cases of the general acceleration of circula-tion; for instead of a purely physical transfer of money, we have a virtual,i.e., merely imaginary or possible transfer, but of the same effectiveness.

(Wicksell [1906] 1967: 67)

Thus, if the system is not in proper order – that is, if the monetary rate ofinterest diverges from the natural rate – deposits can be created to meetentrepreneurial demand independently of the amount of previously accumu-lated savings. This disequibrium situation constitutes the ‘working hypothesis’of Wicksell’s analysis of the ‘cumulative process’.

Schumpeter’s approach is not fundamentally different from Wicksell’s. TheSchumpeterian ‘circular flow’ simply replaces the Wicksellian ‘cash economy’.In the basic framework of the circular flow, money is primarily perceived as aspecial good serving the purpose of a unit of account and facilitating the circu-lation of commodities within the economy.7 As Schumpeter puts it: ‘Let us nowintroduce this denominator of price and medium of exchange and let us choosegold for the role of “money commodity” ’ (Schumpeter 1912 [1934]: 47).However, metal money is not the only conceivable means of payment. Credit –in this case ‘normal credit’ – also plays a part.8 Together with commoditymoney, collateral, or asset-backing requirements, it serves as a counterpart toreal exchanges. Contrary to Wicksell, when dealing with the case of a ‘simplecredit economy’, Schumpeter assumes that ‘normal credit’ does not increase thevelocity of money circulation.9

However, Schumpeter regards credit creation by banks as the ‘differentiaspecifica’ of capitalism. Indeed, in the same vein as Wicksell, Schumpeterclaims that the emergence of a banking system signifies a departure from thestatic case in that it gives rise to a new category of credit which he refers to as‘abnormal credit’.10 This form of credit is associated with the case of economicdevelopment, that is, with dynamic analysis, since without it neither innova-tion nor cycles were possible. Thus, the process of economic developmentcreates a situation where the nature and role of money is dominated by its bankcredit form. The key role of credit is the creation of purchasing power for thepurpose of transferring it to innovators in order for them to finance their newproductive activities. If the stationary state is confined to the mere transfer ofalready existing purchasing power, the economic development of a system char-


acterised by private property and division of labour requires the creation of newmeans of payment. As Schumpeter puts it, by means of credit entrepreneursgain access to the social stream of goods before they have acquired a normalclaim to it.11 Bank credit thus consists in ‘new means of payment created ad hoc,since the entrepreneurs have no means of their own and since there are – so far– no savings’ (Schumpeter 1912 [1934]: 107). These means of payment do notonly include money as such. Schumpeter lists them as ‘commodities which infact circulate as money’, ‘money made of a material the market price of which isless than the purchasing power of the monetary unit made of it’, ‘bank notes’,but also ‘current accounts and clearing accounts’, ‘the amount of all paymentswhich are disbursements out of income and are handled exclusively by compen-sation’ and, finally, ‘credit instruments and claim titles of all kinds, to theextent that they in fact perform the role of money’.12 Therefore, money strictlydefined no longer plays the same role as in the circular flow. Schumpeter arguesthat now ‘not only a part but the whole of the exchange process can be settledby … credit media’ (Schumpeter 1912 [1934]: 53).

This generalised role of credit cannot but affect the market for loanablefunds. Like Wicksell, Schumpeter shows that credit creation by banks togetherwith the institutional setting that renders it possible – i.e. an organised bankingsystem producing new sources of purchasing power within the economy –disrupts the Walrasian adjustment mechanism of the supply and demand forcash balances. In such an environment, it is logically impossible to interpret themarket for money and credit in the same way as any other market in whichsupply and demand would be represented by independent functions. Thus,Schumpeter maintains that the

demand for credit is self-propagating, in that the consequences of its expan-sion and increasing satisfaction go on creating the economic conditions foreven more credit demand. The more bank money is issued, the more creditis necessary for the purchase of one and the same quantity of means ofproduction, and the more, also, can be economically invested in theiracquisition …. The demand for credit makes possible not only itself, butalso a corresponding supply; and every supply makes possible a corre-sponding demand, so that supply and demand in this case do not confronteach other as independent forces. To this extent, therefore, the banksdetermine not only to whom they will grant credit but also how muchcredit as a whole they wish to grant and what demand to call forth.

(Schumpeter 1917/1918 [1956]: 207)

Schumpeter’s adherence to Wicksell is less clear with regard to the definition ofinterest rates. In the first place, Schumpeter insists that the short-term rate ofinterest is essentially a ‘monetary phenomenon’. This is a consequence of thedefinition of saving and investment in monetary terms. Indeed, according toSchumpeter, investment gives rise to an equivalent amount of saving since thelatter is defined independently of its real source. In his own terms:


[B]y Saving we mean the earmarking, by an household, of an element of itscurrent receipts – as distinguished from ‘capital gains’ – for the acquisitionof titles to income or for the payment of debt. If a firm does the same thingwith an element of its net receipts from the sale of products and services,we shall speak of Accumulation. The distinction between Saving andAccumulation also applies, although it may be difficult to carry out, incases in which, as in the case of many farmers, ‘firm’ and ‘household’ areone. We confine both concepts to decisions about monetary funds and weneglect, for convenience’s sake, any similar decision that may be taken withrespect to commodities. Saving and Accumulation will thus be treated aselements of a monetary process: the complementary processes in the worldof goods constitute a distinct problem.


Given this definition of saving, the concept of the real rate of interest loses itsrelevance since the interest factor is a purely monetary phenomenon.Schumpeter writes:

[I]nterest – more correctly, the capital sum plus interest – is, to use our turnof phrase, the price paid by borrowers for a social permit to acquirecommodities and services without having previously fulfilled the conditionwhich in the institutional pattern of capitalism is normally set on the issueof such a social permit, i.e., without having previously contributed othercommodities and services to the social stream.


Moreover, as the rate of interest is derived from the positive rate of profit associ-ated with the operation of innovative productive activities, it is also ashort-term phenomenon. Contrary to Keynes, interest is therefore related to‘income-money’ or ‘transaction money’ rather than to ‘stock’ or ‘speculativemoney’. From this point of view, the Wicksellian dichotomy between a real anda monetary rate of interest becomes meaningless. As a matter of fact,Schumpeter never refers to a real rate of interest since he argues that

nominal and real rates … are only different measurements of the samething or, if we prefer to speak of different things even in this case, it is themonetary rate which represents the fundamental phenomenon and the realrate which represents the derived phenomenon.


In other words, the real and the monetary rates are no longer determined inde-pendently from one another since the level of the real rate derives from thedifference between the monetary rate of interest and the rate of inflation.

Turning now to Hayek’s views on interest, saving and investment, thedistinction between neutral money and credit can be found in his two major


contributions on the subject, Monetary Theory and the Trade Cycle (MTTC) andPrices and Production (PP). As is well known, Hayek assumes that a bartereconomy is characterised by a tendency towards equilibrium and that thistendency continues to exist when money is introduced. He notes that ‘in abarter economy, interest forms a sufficient regulator for the proportional devel-opment of the production of capital goods and consumption goods, respectively’(Hayek 1929 [1966]: 91–2). More specifically, the introduction of a supply ofmoney – that is, the transition to a monetary economy – has no impact on thetendency towards stability so long as it is backed by an equivalent amount ofaccumulated savings.13 Disequilibria only become a possibility when the organi-sation of credit disturbs the adjustment process towards equilibrium betweensupply and demand. Credit expansion affects the velocity of circulation.14 Now,a change in the velocity of circulation ‘represents as it were a one-sided changein demand which is not counterbalanced by an equivalent change in supply’(Hayek 1929 [1966]: 93). Thus, this violation of the logic of supply and demandexplains why banks are likely to charge a rate of interest that deviates from theequilibrium level and will not instantly adjust to it. As Hayek puts it:

Either because the supply of bank credits is, within certain limits, funda-mentally independent of changes in the supply of savings or because thebanks have no particular interest in keeping the supply of bank credit inequilibrium with the supply of savings and because it is, in any case, impos-sible for them to do so – then we shall have proved that, under the existingcredit organisation, monetary fluctuations must inevitably occur and mustrepresent an immanent feature of our economic system.

(Hayek 1929 [1966]: 152)

This passage clearly indicates Hayek’s adherence to Wicksell as regards thetreatment of bank credit and its impact on the determination of the rate ofinterest. As is well known, the discrepancy between the monetary and the realrates of interest plays a central role in Hayek’s theory of business cycles.However, unlike Wicksell’s cumulative process, Hayek’s theory of the tradecycle is not confined to the explanation of fluctuations in the general pricelevel. In Hayek’s words:

The monetary starting point makes it possible, in fact, to show deductivelythe inevitability of fluctuations under the existing monetary system and,indeed, under almost any other which can be imagined. It will be shown, inparticular, that the Wicksell–Mises theory of the effects of a divergencebetween the ‘natural’ and the money rate of interest already contains themost important elements of an explanation, and has only to be freed fromany direct reference to a purely imaginary ‘general money value’ … in orderto form the basis of a Trade Cycle theory sufficing for a deductive explana-tion of all elements in the Trade Cycle.

(Hayek 1929 [1966]: 147)


To sum up, our discussion of Schumpeter’s and Hayek’s respective positions oninterest, savings and investment has emphasised their similarities. Theseinclude their choice of a stationary equilibrium as the analytical point of depar-ture, the distinction between neutral and active money, and the importance ofthe role attributed to bank credit in the process of the emergence of non-equi-librium dynamics. Important though these similarities are, they should notdistract from the differences between the two approaches. These are mainlyrelated to the authors’ understanding of economic dynamics as such, and in thefollowing section we therefore turn to Schumpeter’s and Hayek’s explanationsof business cycles.

Bank credit and productive dynamics in Schumpeter andHayek: two distinct views of how credit shapes the economy

So far, we have highlighted the Wicksellian origin of both Schumpeter’s andHayek’s conceptions of money and banking, focusing on the theoretical impli-cations of the introduction of bank credit for dynamic analysis. Although bothauthors can be said to adhere to the same Wicksellian tradition, they differmarkedly with regard to the mechanisms they regard as constituting economicdynamics.

To begin with, let us recall the main features of Hayek’s and Schumpeter’sapproach to this question. As we have already pointed out, in both explanationsof business cycles a state of stationary equilibrium serves as the point of depar-ture. The upswing is then triggered by an increased activity in the investmentgoods sector that does not meet with sufficient resources, that is, voluntarysavings by the public. Schumpeter argues that this situation is caused by a spurtof innovations, financed through new credit created by banks and organised byentrepreneurs, as opposed to the ‘mere managers’ who prevail in the circularflow and whose activities are driven by the logic of ‘routine’.15 Once the gesta-tion period for the new goods has come to an end, the economy adjusts towardsa new equilibrium position, in the process eliminating some old firms. This isthe core of Schumpeter’s description of the primary wave – or ‘first approxima-tion’ – consisting of only two phases: prosperity and depression. Schumpeterthen extends his analysis to take account of secondary effects – optimistic orpessimistic expectations, miscalculations and income effects – which reinforcethe primary process. Owing to these secondary effects, the economic processwill overshoot the new equilibrium position at the end of a period of prosperity.Recession will deteriorate into depression from where a process of recovery willresult in the system settling in the near neighbourhood of a new equilibriumposition.

For Hayek, deviation from equilibrium is typically caused by a monetaryphenomenon, namely, a money rate of interest lower than the natural or equi-librium rate. As we shall develop, the thesis of a monetary origin of economicfluctuations needs to be understood in broad terms, since it refers to theinability of the banking system to ensure that the level of the monetary and the


natural rate of interest coincide at every point in time, independently ofwhether or not the impulse for the initial deviation from equilibrium is of amonetary kind. Essentially, the lower money rate of interest induces firms toundertake excessive investment expenditures, financed by money creationrather than by savings. However, this disproportionate increase in investmentexpenditure cannot be maintained forever, in particular, when inflation controlis important. The necessary adjustment of the economy’s productive structuresleads to depression.

As noted by Streissler (1983) and Bellofiore (1991), both Hayek andSchumpeter make use of the mechanism of forced saving in their analyses of thecyclical upswing in order to describe the real effects of credit creation. InSchumpeter’s framework, the relevant redistribution of purchasing power isfrom traditional producers to innovators with banks playing a crucial comple-mentary role in meeting demand for finance by innovating firms. The dynamicprocess thus set into motion then leads to a new quasi-equilibrium positioncharacterised by higher productivity and an improved utilisation of resources.For Hayek, however, forced saving is equivalent to a redistribution fromconsumers to investing producers as credit not backed by voluntary savings ischannelled towards investment activities, in the course of which more round-about methods of production are being implemented. In this setting, expansiondoes not lead to a new equilibrium position but is equivalent to a deviationfrom the equilibrium path, that is to an economically harmful distortion of therelative (intertemporal) price system. The eventual return to equilibrium thentakes place via an inevitable economic crisis.

These basic differences between Schumpeter’s and Hayek’s approaches areimportant for clarifying their respective understanding of dynamic processes inthe economy. By focusing on the considerable emphasis both authors put onmoney and banking in the context of explaining real dynamic processes, it ispossible to distil elements of their views on the workings of the business cyclewhich, in turn, open the way to a better understanding of the meaning theyattribute to the notion of economic fluctuations. To simplify the exposition,we will make use of Frisch’s well-known distinction between impulse andpropagation mechanisms to discuss the role played by banks and credit in theemergence, diffusion and reversal of disequilibria in both Schumpeter’s andHayek’s theories of business cycles. As we shall argue, such a comparison isnot only interesting in itself but also helps to shed some light on twoconflicting theoretical views of the problem of stability (or instability) in capi-talist economies which continue to dominate contemporary macrodynamicanalysis.

We already know that Schumpeter attributes the origin of business cycles todiscontinuous changes arising from innovations disrupting the circular flowwhich stands at the beginning of the analysis. Naturally, this position encour-ages a reading of The Theory of Economic Development (TED) as concerned withthe real sphere of economics.16 Recall, however, Hayek’s reading of it:


This group [of theories] pays close attention to the monetary inter-connec-tions and expressly emphasizes them as a necessary condition for theoccurrence of the processes described. But they fail to pass from this realiza-tion to the necessary conclusion; to make it a starting-point for theirtheoretical elaboration, from which all other particular phenomena have tobe deduced. To this group belongs the theory of Professor J. Schumpeter.

(Hayek 1929 [1966]: 97)

According to Hayek, Schumpeter discards ‘the monetary causes which start thecyclical fluctuations’ (Hayek 1929 [1966]: 17). This does not, however, meanthat monetary or financial factors play no role in Schumpeter’s explanation ofthe process of economic development. To the contrary, Schumpeter makes itclear that the new combinations brought about by innovation cannot be under-taken without the co-operation of bankers providing entrepreneurs with thenecessary financial means. More precisely, unlike existing production, innova-tion cannot be financed by a revolving fund or on the basis of returns fromproduction in the previous period. The financing and obtaining of credit is anintegral part of the process of innovation. Since entrepreneurs lack thepurchasing power required to carry out their investment plans, they mustborrow it. As Schumpeter (1912 [1934]: 102) puts it: ‘[H]e [the entrepreneur]can only become an entrepreneur by previously becoming a debtor’. Thus, inno-vations give rise to the demand for bank finance, that is, for ‘abnormal credit’,defined by Schumpeter as the creation of purchasing power to which noexisting new goods correspond.17 Clearly, the relationship between bankers andentrepreneurs is essential for getting a process of economic evolution started.Moreover, the influence of banks goes far beyond the mere provision of credit.According to Schumpeter:

Since all reserve funds and savings today usually flow to him [the banker]and the total demand for free purchasing power, whether existing or to becreated, concentrates on him, he has either replaced private capitalists orbecome their agent; he has himself become the capitalist par excellence.He stands between those who wish to form new combinations and thepossessors of productive means.

(Schumpeter 1912 [1934]: 74)

More specifically, in Schumpeter’s analysis, banks are seen to have both apermanent and an asymmetric impact on the money market, which includestwo spheres.18 The first sphere of the money market is the ‘sphere of hoards andreserves’. The second one, the ‘capital sphere’, is where the ‘income-yieldingassets’ are traded. It includes the real estate and mortgage markets as well as thestock market.19 The common feature of these two spheres, and therefore thedistinctive feature of the money market, is that they permit stock markets towork. The money market is the place where ‘cash reserves’, that is, idle non-circulating money, and ‘income yielding assets’ are mutually exchanged.20 In


this framework, then, the role of banks is clearly not limited to the control ofcredit. Schumpeter, in fact, asserts that: ‘The most cursory glance at moneymarket processes shows that the banks regulate both stock market speculationand the pulse-beat of industrial and commercial life, now restraining, now stim-ulating them’ (Schumpeter 1917/1918 [1956]: 206)

This implies that banks exert a very strong influence on economic life. Thispower derives from two factors. First, Schumpeter assumes that both spheres ofthe money market are interrelated. Therefore, the markets for short-term loansand long-term assets do not work separately but interact within a single moneymarket. Secondly, the ‘sphere of hoards and reserves’ depends heavily on bankssince the latter can manipulate the volume of available liquidity through thelending of credit. By creating means of payment through organising credit,banks effectively regulate the activity of this sphere. Moreover, the interdepen-dence of both spheres within the money market allows banks to extend theirinfluence to the sphere of income-yielding assets. On the one hand, bankscreate ex novo credit means of payment, thereby strongly contributing to theemergence of interest. This, in turn, affects the whole economy in that the exis-tence of interest now constitutes an additional motive to save on the part ofconsumers. Banks are, thus, not purely neutral intermediaries, nor are the effectsof credit creation transitory, since they give rise to a secondary wave of thecreation of new sources of purchasing power which can be mobilised to financefurther productive activity. On the other hand, during an upswing, banks inter-fere with real propagation mechanisms by allowing the transfer of productiveresources to new entrepreneurs. These reallocation effects can interfere withprice competition and alter the outcome of the process of adaptation, in thecourse of which some existent firms turn out to have become unprofitable andare out-selected, while others, seizing new profit opportunities and being backedup by banks, manage to escape bankruptcy.

Furthermore, banks interact with entrepreneurs in determining the volumeof credit. While it is the entrepreneurs who initiate the process, banks decidewhich of these initiatives to finance based on their expectations regarding theprofitability of innovative projects and the entrepreneurs’ ability to repay theirloans: ‘We know already by what forces this supply is regulated: first with regardto possible failures by entrepreneurs, and secondly with regard to the possibledepreciation of the credit means of payment’ (Schumpeter 1912 [1934]: 195).In another passage, Schumpeter explicitly argues that

[t]he banker must not only know what the transaction is which he is askedto finance and how it is likely to turn out, but he must also know thecustomer, his business, and even his private habits, and get, by frequently‘talking things over with him’, a clear picture of the situation.


On closer examination, it is possible to define the equilibrium level of theinterest rate at a given point in time by deriving a supply and a demand curve


for credit.21 However, this description of the workings of the money market isnot very satisfactory. In the first place, banks select entrepreneurs not only bysetting the rate of interest but also by evaluating innovations as well as theentrepreneurs themselves and the subsequent use they make of a loan. In thesecond place, the changes in the demand for finance occurring throughout thecycle affect not only actual but also potential credit (i.e. the maximum creditbanks can create in a given institutional context). Moreover, the question oftechnical limits to credit supply, such as may arise in a monetary system whenbanking operations are constrained by reserve requirements and when there is apreference for cash on the part of the public, is of little relevance toSchumpeter, given that banks can ration credit and manage cash–deposit ratiosin a procyclical manner, reducing them in prosperity and raising them in adepression.22 In short, the actual supply of credit shifts with the demand anddoes not face a definite ceiling of potential credit supply since the latter movesprocyclically. This also explains the manner in which the creation of purchasingpower works, leaving us with the question of how this affects real productiveactivity. This is a point particularly worth developing since, as we shall see,Schumpeter’s position on the role of inflation and forced saving sharply differsfrom Hayek’s.

According to Schumpeter, the new sources of purchasing power created bybanks are targeted at individual entrepreneurs and their specific productiveprojects. To put it differently, credit precedes the realisation of entrepreneurialprofits. While credit inflation may occur in this context, it will only be oftemporary duration. In Schumpeter’s words:

After completing his business – in our conception, therefore, after a periodat the end of which his products are on the market and his productivegoods used up – he [the entrepreneur] has, if everything has gone accordingto expectations, enriched the social stream with goods whose total price isgreater than the credit received and than the total price of the goodsdirectly and indirectly used up by him. Hence the equivalence between themoney and commodity streams is more than restored, the credit inflationmore than eliminated, the effect upon prices more than compensated for, sothat it may be said that there is no credit inflation at all in this case –rather deflation – but only a non-synchronous appearance of purchasingpower and of the commodities corresponding to it, which temporarilyproduces the semblance of inflation.

(Schumpeter 1912 [1934]: 110)

Turning now to forced saving, this appears to be of secondary importance inSchumpeter’s analysis of economic development. While it is true that, when-ever innovation entails a lengthening of the period of production, the output ofconsumer goods and, thus, of real consumption, is likely to decrease during thegestation period, voluntary saving out of income arising from the expenditure ofnew money cannot be ruled out.23 In this case, even though productive


resources are being redistributed between new entrepreneurs and ‘mere-managers’, there is no forced saving. Moreover, and more importantly, it has tobe stressed that the main phenomenon is the absolute squeeze of the purchasingpower of old firms, so that forced saving, if it occurs at all, takes the form of anindirect process based on a reduction in the purchasing power of existentproductive units.

Considering now the mechanisms underlying the reversal of the cycle, bankscan continue to exert some influence during the upward phase. Although thesupply of credit cannot be invoked directly to explain the upturn of the cycle –recall that potential credit supply increases during the upswing – banks may,however, delay the end of the expansionary phase or, more likely, anticipate it.It is, in fact, probable that they will impose a risk-premium accounting for thedevaluation of capital due to inflation or tighten the rationing of credit, sincethe risk of innovation is carried not by the entrepreneur but by the banker.24

Hence, in Schumpeter’s explanation of business cycles, banks clearly play amajor role in the dynamics of accumulation. However, their influence is notunambiguous since the rate of interest on loans is a ‘tax’ on profits and thusconstitutes a brake economic development.25

Returning to Hayek’s description of economic fluctuations, his objections toSchumpeter’s approach can now be understood more easily. According toHayek, the introduction of a banking system into economic analysis disturbs theadjustment process between capital supply and money demand. In other words,as soon as credit is allowed for, supply and demand will no longer adjust auto-matically and prices no longer determine a path towards economic equilibrium.Consequently, prices will no longer provide signals for short-term market adjust-ments. To the contrary, ‘these prices may elicit movements which not only donot lead to a new equilibrium position but which actually create new distur-bances of equilibrium’.26

Banks, interested in keeping the credit supply elastic, will, in particular, set amoney rate of interest which does not, in general, correspond to the equilib-rium or the natural rate. The former is determined by the liquidity of banks,while the latter is always determined by its role in adjusting capital supply todemand.

The above comparison of Schumpeter’s and Hayek’s views on bank creditand business cycles is of particular interest for two main reasons. First, it revealsthe importance of banks and, more generally, of the banking system as a drivingforce behind economic fluctuations. Second, it also suggests that, in a monetaryeconomy, disequilibria can be described in terms of a co-ordination failure ofthe interest rate mechanism. In such an economy, any given mechanism ororganisational device, for instance, the organisation of credit, which inherentlyrelies on this co-ordination failure, result in the formation of prices or rates ofinterest that deviate from those associated with a barter economy equilibrium.Yet, if the above analysis is valid, any other situation leading to disequilibriumprices could also be invoked to explain the same phenomenon. Hayek did, infact, consider some such situations. Thus, he refers, for instance, to ‘changes in


the relations of costs and selling prices’, to ‘shifts in the distribution of incomes’(Hayek 1929 [1966]: 129) and to the possibility of ‘a rate of interest [on loans]lowered by monetary influences’ (ibid.: 128). However, Hayek’s position on thesubject of ‘monetary influences’ is unclear. In MTTC, he insists that the elas-ticity of the volume of money is an ‘immanent necessity of the monetary andcredit mechanism’ (ibid.: 127) and conveys the impression that the endogenousnature of the mechanisms of credit creation by the banking system represents anecessary and sufficient condition for business cycles, whether or not it resultsfrom arbitrary interference by the banking authorities. By contrast, in PP, hechooses the ‘case of an increase of money in the form of credits granted toproducers’ (Hayek 1931 [1935]: 54) as the starting point of a cycle. Here, a fallin the rate of interest is clearly regarded as resulting from ‘deliberate’ decision-making (ibid.: 85). This ambiguity arises, in part, because Hayek fails sufficientlyto clarify which institutional monetary framework he has in mind. In any case,his discussion of the ways in which banks interfere with real propagation mech-anisms is a good example of the lack of coherence in his treatment of moneyand banking in the wider context of his theory of business cycles.

As is well known, in his theory the upswing of a cycle is characterised by anincrease in the demand for capital emanating from producers’ awareness of newinvestment opportunities and their access to bank credit. Capital newly raisedin this way is then employed in the implementation of more roundaboutprocesses of production. However, since full-employment prevails, the increaseof capital goods can only be achieved by withdrawing productive resources fromalready existing shorter lines of production. Therefore, the growth in theproduction of capital goods is accompanied by a decline in the output ofconsumption goods. Assuming that wages only rise with some delay, prices ofcapital goods rise faster than those of consumption goods, thus reinforcing theexpansionary movement. There are, moreover, additional reinforcing factorslinked to the ‘organization of credit’. Consider Hayek’s conceptualisation of thebanking system in some more detail. Hayek assumes a ‘mixed’ monetary systemwith both an exogenous and an endogenous form of money. Commercial banksmake their decisions on the basis of their profit expectations which depend onthe risk characteristics of borrowers as well as on the actions of their respectivecompetitors. The bank’s risk aversion grows as expansion proceeds and is notindependent of their pricing policy. At a given risk level, the decision not tosatisfy demand (by imposing too high a loan rate) implies a greater opportunitycost for the banker. This raises the winner’s curse problem, that is, it leads to asituation where banks will expand credit in the upswing, even at the cost ofdepleting their resources, so as not to lose clients and encounter additionalrisks. In this way, there is an ‘elastic’ deposit multiplier which sustains a growingproductive activity.

However, a point will be reached at which consumers will face an insuffi-cient supply of consumption goods, thus creating tension in the economy. Thesituation is aggravated by the fact that additional income has been generatedduring the upswing. This induces a counter-movement of relative prices:


consumption goods prices rise while capital goods prices fall, and the originalprice ratio is re-established. Contrary to Schumpeter, Hayek argues that theseare technical limits to the creation of credit, so that it is the specific behaviourof banks that determines the upper turning point of the cycle.27 In otherwords, the flexible deposit multiplier described above appears to be bounded.Hayek indicates, in fact, that, when the price of consumer goods begins to risefaster than the price of capital goods, the ratio between cash payments andpayments by cheque is altered in favour of the former. Consequently, in thecourse of a boom, the need for cash will increase along with prices and inducea cash drain that will force banks to restrict credit supply. Hayek’s reasoning isas follows:

Concerted action in this direction, which for competitive reasons is theonly action possible, will ensue only when the increased cash requirementsof business compel the banks to protect their cash balances by checkingfurther credit expansion, or when the Central Bank has preceded them.This, again, will only happen, as a rule, when the banks have been inducedby the growing drain on their cash to increase their re-discount. Experienceshows, moreover, that the relation between cheque-payments and cashpayments alters in favour of the latter as the boom proceeds, so that anincreased proportion of the cash is finally withdrawn from the banks.

(Hayek 1929 [1966]: 174–5)

Therefore, even without reserve restrictions, credit expansion must come to ahalt before an accelerating rate of inflation undermines the function of moneyas the unit of account. In PP, Hayek writes:

So long as the banks go on progressively increasing their loans it will,therefore, be possible to continue the prolonged methods of production orperhaps even to extend them still further. But for obvious reasons the bankscannot continue indefinitely to extend credits; and even if they could, theother effects of a rapid and continuous rise of prices would, after a while,make it necessary to stop this process of inflation.

(Hayek 1931 [1935]: 89–90)

This brings us to Hayek’s view on forced saving. When dealing with the caseof forced – as opposed to voluntary – saving, Hayek assumes given and stableconsumer preferences. It is supposed that banks start the cycle, whereas forSchumpeter the prime mover is entrepreneurial action. Credit supply isregarded as normatively limited by savings or, to be precise, even by a definiteupper limit since a lengthening of the time structure of production, madepossible by bank credit, proves not to be sustainable once credit supply meetsthis limit. By contrast, Schumpeter argues that ex novo credit creation may notonly make saving dependent on investment but also may have no intrinsiclimit. It should, however, be pointed out that, in his earlier writings, Hayek’s


views on forced saving are not all that different from Schumpeter’s.28 Thus,Hayek writes, for example, in 1925:

The losses which arise from the revelation that the capital outlay made isnot yet economically justified are the price of an undesirably rapid progress,a rate of progress which exceeds that which people are ready to purchase forthemselves by a corresponding voluntary sacrifice of current enjoyments. –There can be no doubt at all that the development of the capitalisteconomy over the last 100 years would not have been possible without the‘forced saving’ effected by the extension of additional bank credit. Henceeconomic fluctuations must probably be regarded as a necessary accompani-ment of the accelerated development experienced by countries of theWestern world in the last 150 years. Such fluctuations, in turn, could beentirely eliminated only if the tempo of this development was substantiallylessened.

(Hayek 1925 [1984]: 21)

This passage points to a puzzle in Hayek’s conception of the role played by thebanking system in the unfolding of economic dynamics. In MTTC, Hayekconveys the impression of taking for granted a commercial banking system themonetary liabilities of which enter circulation by way of loans to manufacturers.He also emphasises ‘the potential implicit in this institutional fact for thecreation of money to interfere with the capital market’s co-ordination of savingand investment’ (Laidler 1994: 9). However, in PP, when discussing the case of‘voluntary savings’, he refers instead to a monetary system consisting of stablebase money, thus eliminating the confusion between ‘those deposits which findtheir origin in credit and those which arose through cash payments’ (Hayek1929 [1966]: 163) that was at the root of the unsustainable cash drain inMTTC. As stressed by Trautwein, this ‘dual’ treatment of the monetary systemrenders Hayek’s distinction between the cases of ‘voluntary savings’ and ‘forcedsavings’ inconsistent. There is indeed no reason to assume that in the case of‘forced savings’ banks act as passive brokers, if we have in mind the same under-lying institutional framework as in MTTC. To put it another way, if weconceptualise banks as creators of money that cannot, therefore, distinguishprecisely between deposits originating in credit and those originating in cashpayments, an increase in voluntary savings would also imply an expansion ofbank deposits, triggering exactly the kind of destabilising mechanisms as doesthe direct creation of credit.29

Thus, a closer look at Hayek’s conception of money and banking revealsinconsistencies with regard to the institutional framework underlying his theoryof business cycles. This ‘schizophrenic treatment of the bank’s behaviour’(Trautwein 1996: 45) explains why Hayek ultimately chooses to focus on thereal side of the economy and, in particular, on the sphere of production and thequestion of factor substitution. This shift of focus is present in PP and is rein-forced in Profits, Interest and Investment.


Concluding remarks

Our re-examination of Schumpeter’s and Hayek’s theoretical frameworks hasrevealed significant similarities in their business cycle theories, which we havetraced back to their common Wicksellian heritage. This is not, however, asurprising insight for historians of economic thought. Schumpeter and Hayekwere both prominent economists, but they were also great historians ofeconomic thought.

In his early writings, Schumpeter referred to Wicksell’s fundamental contri-bution to monetary theory.30 He also paid a specific tribute to Wicksell in aGerman article entirely dedicated to the works of the Swedish author and hepersistently emphasised the importance of Wicksell’s work in his History ofEconomic Analysis.31 Hayek, too, acknowledged Wicksell’s strong influence onthe evolution of his own conception of economic theory. In one of his first arti-cles in 1925, he comments extensively on several of Wicksell’s contributions toeconomic analysis. Moreover, as we know, he frequently referred to Wicksell inPP, thereby contributing to the diffusion of Wicksell’s ideas among his English-speaking readers.

However, our comparison has also shown that Schumpeter and Hayekprovide two distinct explanations of how credit shapes the economy. Accordingto Schumpeter, business cycles are the very vehicle of progress and growth.32

Therefore, any attempt to get rid of fluctuations would amount to eliminatingthe dynamics of capitalism. Given that the upswing requires co-operationbetween banks and innovating entrepreneurs, both money creation and forcedsaving are, as a means for redistributing purchasing power to new productiveunits, necessary and beneficial components of the workings of a moderneconomic system. In this he differs sharply from Hayek who takes as his point ofreference a state of intertemporal price equilibrium. Under ideal conditions, thissituation also represents an optimal state. Deviations from this point of refer-ence are the result of a divergence between the equilibrium and the monetaryrate of interest, where the latter can be maintained for some time on the basis ofelastic credit supply by banks. Forced saving then signals a global disequilib-rium, a ‘dis-coordination’ of saving and investment plans which can no longerbe carried out as originally envisaged. However, given Hayek’s belief in the exis-tence of a tendency towards equilibrium, he attributes the causes of fluctuationsto external factors. Thus, for Hayek, the ideal state is one of an evenly evolvingeconomy without business cycles and characterised by neutral money. WhereasSchumpeter’s concept of economic dynamics is that of business cycles driven byinnovation together with the complementary and necessary role played by bankcredit, for Hayek it consists in the absence of such cycles.

Notes1 For a similar account, see Arena (1985) and de Boyer (1985), although the latter

concentrates on the distinction between money and credit which provides theconnection between Wicksell and Schumpeter.


2 There is no explicit reference to a ‘natural’ rate of interest in Schumpeter’s writings.First, Schumpeter is rather sceptical about the meaning the concept. Second, as weshall develop later, Schumpeter conceives of the rate of interest as a purely monetaryphenomenon which owes its existence to the emergence of real profits through inno-vation. It is, however, possible to consider the dynamics of the gap between the rateof interest and the rate of profit in Schumpeter’s analysis to a certain extent as asubstitute for the interest rate divergence mechanism in Wicksell and Hayek.

3 Wicksell 1906 [1967]: 11–12.4 Ibid.: 51–8.5 Wicksell 1898 [1965]: 71–80. The intermediate cases correspond to what Wicksell

calls a ‘simple credit economy’ (ibid.: 59–62; 1906 [1967]: 70) and an ‘organisedcredit economy’ (ibid.: 62–70; 1967: 72).

6 Ibid.: 53–8.7 Schumpeter 1912 [1934]: 53.8 Ibid.: 100.9 Ibid.: 55, fn.1.

10 Ibid.: 102.11 Ibid.: 107.12 Schumpeter 1917/1918 [1956]: 207.13 Hayek 1929 [1966]: 92.14 Ibid.15 Schumpeter 1912 [1934]: 83.16 In this context, Schumpeter’s position regarding the origin of disequilibrium is

similar to Wicksell’s. As is well known, the cumulative process, although describinga far less sophisticated type of dynamics, starts with a real productivity shock whichcreates a divergence between the monetary and the natural rates of interest.

17 Schumpeter 1912 [1934]: 101.18 Schumpeter 1917/1918 [1956]: 176.19 Ibid.20 Ibid.21 The rationale for this analytical development can be found in Schumpeter’s TED.

See Schumpeter 1912 [1934]: 191–8. Also see Bellofiore 1991: 378 and Messori1984.

22 Schumpeter 1912 [1934]: 112–15; 1939: 121–3; 1917/1918 [1956]: 206–8.23 This scenario is also envisaged by Robertson in Banking Policy and the Price Level

(1926).24 Schumpeter 1912 [1934]: 75–6; 1939: 104.25 Bellofiore 1991: 379.26 Hayek 1929 [1966]: 94.27 See Hansen and Tout 1933: 133–5; Colonna 1994: 41–4.28 See Klaussinger 1995: 99.29 See Trautwein 1994: 77; 1996: 45–6. These inconsistencies in Hayek’s analysis are

also taken up by Sraffa in his critique of Hayek. Sraffa argues that Hayek’s case of‘voluntary savings’ effectively describes a situation where there is no money at all(Sraffa 1932: 47). See also Hansen and Tout 1933: 139–40; Neisser 1912 [1934]:436–9.

30 Schumpeter 1917/1918 [1956].31 Schumpeter 1927.32 Note that Robertson’s view on productive credit creation and forced savings is not

very different from Schumpeter’s. Robertson did, in fact, point out ‘that a littleforced saving now and again … [might] be the price for … progress’ (1928: 57).


Bibliography

Arena, R. (1985) ‘Circulation, revenu et capital: théorie monétaire et tradition quantita-tive’, in R. Arena, A. Graziani and J. Kregel (eds), Production, Circulation et Monnaie,pp. 47–76, Paris: Presses Universitaires de France.

Bellofiore, R. (1991) ‘Money and development in Schumpeter’, in J. Cunningham (ed.),Joseph A. Schumpeter: Critical Assessments, Vol. IV, pp. 371–94, London: Routledge.

Boyer, J. de (1985) ‘Taux d’intérêt et quantité de monnaie: note sur la distinction entredeux sphères de circulation chez T. Tooke, K. Wicksell, J.M. Keynes et Schumpeter’,in R. Arena, A. Graziani and J. Kregel (eds), Production, Circulation et Monnaie, pp.371–86, Paris: Presses Universitaires de France.

Colonna, M. (1994) ‘Hayek’s trade cycle theory and its contemporary critics’, in M.Colonna and H. Hagemann (eds) Money and Business Cycles: The Economics of F.A.Hayek, Vol. I, pp. 27–53, Aldershot: Edward Elgar.

Haberler, G. (1941) Prosperity and Depression, 3rd edn, Geneva: League of Nations.Hansen, A.H. and Tout, H. (1933) ‘Annual survey of business cycle theory: investment

and saving in business cycle theory’, Econometrica, 1: 119–47.Hayek, F.A. (1925) ‘Die Währungspolitik der vereinigten Staaten seit der Überwindung

der Krise von 1920’, Zeitschrift für Volkswirtschaft und Sozialpolitik, 5. English transla-tion of an extract as ‘The monetary policy of the United States after the recovery ofthe 1920 Crisis’, in R. McCloughry (ed.) F.A. Hayek: Money Capital and Fluctuation:Early Essays, London: Routledge & Kegan Paul, 1984.

—— (1928) ‘Das intertemporale Gleichgewichtssystem der Preise und die Bewegungendes “Geldwertes”’, Weltwirtschaftliches Archiv, 28: 33–76. Translated from German as‘Intertemporal price equilibrium and movements in the value of money’, in R.McCloughry (ed.), F.A. Hayek: Money Capital and Fluctuations: Early Essays, London:Routledge & Kegan Paul, 1984.

—— (1929) Geldtheorie und Konjunkturtheorie, Beiträge zur Konjunkturforschung,herausgegeben vom Österreischisches Insitut für Konjunkturforschung, 1, Vienna.Translated into English by N. Kaldor and H. Croome as Monetary Theory and theTrade Cycle, London: Routledge, 1933. Reprinted New York: Augustus M. Kelly,1966.

—— (1931) Prices and Production, London: Routledge (2nd rev. edn, 1935).—— (1939) Profits, Interest and Investment and Other Essays on the Theory of Industrial

Fluctuations, London: Routledge.Klaussinger, H. (1995) ‘Schumpeter and Hayek: two views of the great depression re-

examined’, History of Economic Ideas, 3, 3: 93–127.Laidler, D. (1994) ‘Hayek on neutral money and the cycle’, in M. Colonna and H. Hage-

mann (eds) Money and Business Cycles: The Economics of F.A. Hayek, Aldershot:Edward Elgar.

—— (1995) ‘Robertson in the 1920s’, European Journal of Economic Thought, 2: 151–74.Messori, M. (1984) ‘Il credito nel modello di Schumpeter’, in M. Messori (ed.) Atti del

Convegno: Società Sviluppo Impressa. Nel Centenario della Nascita di J.A. Schumpeter,Milan: Franco Angeli.

Neisser, H. (1912 [1934]) ‘Monetary expansion and the structure of production’, SocialResearch: 434–57.


O’Driscoll, G.P. (1994) ‘An evolutionary approach to banking and money’, in J. Birnerand R. van Zijp (eds), Hayek, Co-ordination and Evolution: His Legacy in Philosophy,Politics, Economics and the History of Ideas, London: Routledge.

Robertson, D.H. (1926) Banking Policy and the Price Level, London: King.—— (1928) ‘Theories of banking policy’, reprinted in D. H. Robertson, Essays in Mone-

tary Theory, pp. 39–59, London: Staples Press, 1940.Schumpeter, J.A. (1912) Theorie der wirtschaftlichen Entwicklung, Leipzig: Duncker und


—— (1917/1918) ‘Das Sozialprodukt und die Rechenpfennige. Glossen und Beiträge zurGeldtheorie von heute’, Archiv für Sozialwissenschaft und Sozialpolitik, vol. 44:627–715. Translated from German by A.W. Marget as ‘Money and the SocialProduct’, International Economic Papers, 6, 1956.

—— (1927) ‘Zur Einführung der Folgenden Arbeit Knut Wicksells [MathematischeNationalökonomie]’, Archiv für Sozialwissenschaft, 58: 238–51.



Sraffa, P. (1932) ‘Dr. Hayek on money and capital’, Economic Journal, 42: 42–53.Reprinted in B. Caldwell (ed.) The Collected Works of F.A. Hayek, vol. 9, ContraKeynes and Cambridge – Essays, Correspondence, London: Routledge, 1995.

Streissler, E. (1983) ‘Schumpeter and Hayek: On some similarities in their thought’, in F.Machlup, G. Fels and H. Müller-Groeling (eds) Reflections on a Troubled WorldEconomy: Essays in Honour of Herbert Giersch, London: Macmillan.

Trautwein, H.–M. (1994) ‘Hayek’s double failure in business cycle theory: a note’, in M.Colonna and H. Hagemann (eds) Money and Business Cycles: The Economics of F.A.Hayek, Vol. I, Aldershot: Edward Elgar.

—— (1996) ‘Money, equilibrium, and the business cycle: Hayek’s Wickselliandichotomy’, History of Political Economy 28, 1: 27–55.

Wicksell, K. (1898) Geldzins und Güterpreize, translated into English by R. Kahn asInterest and Prices, London: Macmillan, 1936. Reprinted New York: Augustus M.Kelley, 1965.

—— (1906) Lectures in Political Economy Vol. II: Money, English translation of 2nd(1915) Swedish edn, London: Routledge & Kegan Paul, 1967.


Introduction

In this chapter we compare Schumpeter’s and Keynes’s views on the financingof economic activity. As will be seen, in section one, both economists sharecommon ideas about the working and financing of market economies. In partic-ular, both Keynes and Schumpeter reject the classical notions of the neutralityof money and the dichotomy of the real and the monetary sector, emphasisinginstead the role of monetary and financial variables in their respective explana-tions of economic fluctuations. However, the two approaches also displaysignificant differences, most notably with regard to the importance attributed tospecific financial variables and institutions for the financing of economicactivity. Section two examines these differences and emphasises the originalityof Schumpeter’s analysis of banking.

The rejection of classical monetary theory

When comparing Schumpeter’s and Keynes’s monetary analyses, one can hardlymiss the considerable degree of similarity between the two approaches: not onlydoes the inclusion of money into their respective analytical frameworks producesimilar insights in both cases, it is also at the root of a radical revision of thefeatures of the economic system they examine.

In the preparatory drafts of the General Theory (GT) (Keynes 1973a, 1973b,1979), Keynes makes a distinction between a ‘co-operative’ (or real-wage)economy and an ‘entrepreneurial’ (or money-wage) economy.1 Schumpeter, too,draws a clear distinction between two situations: the circular flow and economicdevelopment.

Keynes’s co-operative economy and Schumpeter’s circular flow have two keyfeatures in common. First, even if money exists in these specific states, as withclassical theory, it is simply regarded as a technical device for facilitating realexchange and is neutral with regard to the level of production. As Schumpeterpoints out, ‘money has, in the circular flow, no other role than that of facili-tating the circulation of commodities’ (Schumpeter 1912 [1934]: 53). Second,the co-operative economy and the circular flow both describe what might be

12 Financing economic activitySchumpeter vs Keynes

Eric Nasica

called a ‘static’ economy. In the Schumpeterian circular flow, there are neitherentrepreneurs nor capitalists, credit or interest: all factors potentially influ-encing the process of economic development are excluded from the analysis. InKeynes’s co-operative economy decisions about production and distribution aremade jointly by all producers in such a way as to maximise utility and consump-tion for all individuals and, at the same time, for society at large (Keynes 1979:66, 77). In this economy, income is paid in kind or out of the output producedby the workforce, and Say’s law necessarily holds true.

In both approaches, the departure from the static economy is made possibleby introducing a more complex concept of money. In Keynes’s analysis, this newtype of money is a store of value; in Schumpeter’s theory, it takes the form ofcredit. With this change in the nature of money, the economy becomesdynamic in that disequilibria (or underemployment equilibria) and fluctuationsof economic activity are now possible.

In Keynes’s approach, macroeconomic instability and disequilibria are aconsequence of new patterns of behaviour characteristic of the entrepreneurialeconomy, such as the long-run demand for liquidity. These types of behaviourare a feature of the money-wage economy operating in an uncertain decision-making environment, in which Keynes was mainly interested. The mereexistence of money as a store of value is likely to drive the economy towards along-period of equilibrium characterised by lasting unemployment. Indeed, thedecision not to purchase goods (that is, to adopt an attitude of saving andwaiting) does not

necessitate a decision to have dinner or to buy a pair of boots a week henceor to consume at any specified date … It is not a substitution of futureconsumption demand for current consumption demand – it is a net diminu-tion of such demand.

(Keynes 1936: 218)

In Schumpeter’s approach, the dynamic aspects of economic activity, such asinnovation and development, are closely linked to the creation of money in theform of credit. Schumpeter regards credit creation by banks as the main sourceof finance, once the stationary economy of the circular flow is left behind andthe analysis focuses on the process of economic development. More specifically,the meaning of credit, in this context, is that of ‘new means of payment createdad hoc since the entrepreneurs have no means of their own and since there are– so far – no savings’ (Schumpeter 1939: 111). These means of paymentrequired by entrepreneurs in order to finance their innovations do not onlyinclude money as such. Schumpeter (1917/1918 [1956]: 168–71) provides a fulllist of such means of payments that includes ‘commodities which in fact circu-late as money’, ‘money made of a material the market price of which is less thanthe purchasing power of the monetary unit made of it’, ‘bank notes’ but also‘current accounts and clearing accounts’, ‘the amount of all payments which aredisbursements out of income and are handled exclusively by compensation’ and,


finally, ‘credit instruments and claim titles of all kinds, to the extent that theyin fact perform the role of money’.

Money is therefore no longer a mere means of facilitating the circulation ofcommodities as in the circular flow: ‘Not only a part but the whole of theexchange process can be settled by … credit media’ (Schumpeter 1912 [1934]:53). This also implies Schumpeter’s definition of the entrepreneur as the ‘typicaldebtor’ (ibid.: 101, 103). At first sight, this specific feature of the Schumpeterianentrepreneur, together with the fact that Schumpeter considers the diversion offinance from routine to innovative activities to be the main role of credit,suggests a certain affinity between his analysis and the writings of the BankingSchool.2 However, on closer examination this interpretation ceases to beconvincing. Not only does Schumpeter accept the idea of a possible autonomouseffect of the quantity of money on the general price level (Schumpeter 1939:546–7), suggesting that he would welcome a synthesis of the Banking and theCurrency Schools, he has also made it clear himself in unambiguous terms thathe does not regard his approach to be in complete accord with that embraced bythe Banking School (Schumpeter 1917/1918 [1956]: 209).

However, it would also be quite difficult to include Schumpeter among quan-tity theorists, given his explicit and persistent rejection of the quantity theory ofmoney (Schumpeter 1954: 1095–117). A detailed analysis of Schumpeter’s mainobjections to the quantity theory of money has been provided by Graziani (1989)and Messori (1985, 1986), to which we refer the interested reader. In the presentcontext it suffices to note a few main points. First, Schumpeter rejects any apriori concept of the quantity of money as a causal influence on prices.3 Second,he argues that, because of the existence of substitutes, and of credit and reservesin particular, a precise definition of the quantity of money is quite impossible.4

Third, it is equally impossible to provide a homogenous concept of the quantityof money or a unique notion of the velocity of circulation.5 Furthermore, money,not being a commodity, has no proper value (Schumpeter 1970 [1990]) and ‘thetraditional apparatus of supply and demand [can]not be applied to the solutionof the problem of money prices of commodities and of price levels’(Schumpeter 1939: 547). Schumpeter emphasizes that the ‘reflected value’ ofmoney is rarely proportionate to the quantity of money because variations inthe latter often imply variations in relative prices and redistribution effects.6

Finally, credit money supplied by banks depends crucially on entrepreneurialdemand: without the entrepreneur there is no credit creation, and the effec-tive volume of credit supplied changes according to shifts in demand.

Schumpeter also mentions the possibility of a twofold origin of finance. Ifthe demand of enterprises for means of payment is not fully met, entrepreneursmay take recourse to accumulated non-banking savings, either through self-financing (by using profits accumulated in an earlier phase of economicdevelopment) or through the money market (Schumpeter 1912 [1934]: 199).

In essence then, Schumpeter regards the money supply not as an indepen-dent variable, but as one that varies endogenously in response toentrepreneurial action. While it remains true that, just as with the quantity

Schumpeter vs Keynes 243

theory of money, the direction of causality runs from bank money to prices,changes in the former originate in entrepreneurial demand for credit required tofinance innovations. Moreover, in the upswing, inflation (through changes inrelative prices) also plays a part in changing the distribution of productiveresources and of income: the Schumpeterian inflationary process results in atransfer of productive resources and a squeeze of the purchasing power of‘mature’ producers’.

Schumpeter’s analysis thus implies a drastic break with the quantity theory ofmoney. In contrast to traditional monetary theory, he strongly emphasises thatthere is no dichotomy between real and monetary analysis and no neutrality ofthe money supply in the long run. In Schumpeter’s own words,

[i]t cannot be argued that the effect is temporary and as such negligible. …A monetary process, the creation of money which is only a ‘claim ticket’and not also a ‘receipt voucher‘, and the rise in prices to which it leads,become a powerful lever of economic development. … It is the specificallycapitalistic method of effecting economic progress.

(Schumpeter 1917/1918 [1956]: 205–6)Keynes’s analysis of banking, developed both prior to and following the publi-

cation of the GT, is similar to Schumpeter’s in several respects. In Treatise onMoney (TM), Keynes maintained that money takes the form of credit money,and he shared the idea that bank money – that is, deposits – is created in theform of loans and is not first collected from already existing deposits. In the intro-ductory pages to the TM, he stresses over and over again that banking consists inthe creation, and not the transfer, of liquidity (Keynes 1930: 25). Moreover,Keynes argues on several occasions that the money supply is not exogenouslyfixed (e.g. ibid.: 189). In Book 4 of the TM, he specifies the reasons why, in hisview, the supply of bank money should be regarded as endogenous, emphasising,in particular, that an increase in the volume of global output requires substantialchanges in the monetary sphere, that is, it ‘requires the acquiescence of thebanking authorities’ (ibid.: 256). This thought is forcefully taken up again in hisanalysis of the credit cycle, which he concluded thus:

In countries (such as most of the continent of Europe) where the volume ofmoney partly depends on the volume of suitable bills available to bediscounted at the central bank, an increase in the volume of output has adirect tendency to produce some corresponding increase in the volume ofcirculating money.

(ibid.: 275 fn.)

The same ideas are developed in a series of articles published between 1937 and1939 (Keynes 1937a, 1937b, 1938, 1939) when Keynes was defending histheory of the determination of income against criticisms by Ohlin andRobertson.7 In these articles, he insists both on the central role of banks in thefinancing of economic activity and on the endogenous nature of bank money:


‘The finance of enterprises … is mainly supplied by specialists, in particular bythe banks’ (Keynes 1937b: 219).

Immediately following this debate, Keynes slightly modified his position tothe effect that, like Schumpeter, he highlighted the possibility of a twofoldorigin of finance and argued that the demand for finance could be met in eitherof two ways: (a) through banks creating new liquidity, or (b) through financialintermediaries bringing about an increase in the rate of interest and succeedingin making already existing liquidity available to enterprises.8

To sum up, in this first section, we have pointed to a number of importantsimilarities in Keynes’s and Schumpeter’s analyses. In particular, it has beenargued that both authors reject the dichotomy between a real and a monetarysector. They also agree that credit money created by banks plays a crucial role inthe financing of economic activity. However, as asserted by Schumpeterhimself, ‘[t]he reader should be on his guard against … surface similarities’(Schumpeter 1939: 127 fn. 2) between Keynes’s theory and his own. Payingheed to this warning, the next section focuses on other aspects of the relationbetween finance and production.

Interest rates, banking and economic activity

The argument outlined in this section is concerned with Keynes’s andSchumpeter’s analyses of the role played by specific financial tools and institu-tions in financing economic activity. As we shall see, it is here where the twoauthors differ considerably.

Short-term and long-term rates of interest

Keynes and Schumpeter coincide in their treatment of the rate of interest as amonetary phenomenon. In Chapter 17 of the GT, the liquidity preferencetheory – stating that the marginal efficiency of money is the premium to be paidfor parting with money – effectively ensures that the rate of interest is deter-mined in the money market; in Keynes’s analysis the rate of interest is the priceof liquidity.

The same monetary explanation of the rate of interest is also present in theapproach of Schumpeter. He argues that ‘interest attaches to money and not togoods’ (Schumpeter 1912 [1934]: 158). Indeed, Schumpeter points out that if

money [was] only an intermediate link, merely of technical importance,and [we] set about substituting for it the goods which are obtained with itand for which therefore in the last analysis interest is paid, we at once losethe ground from under our feet. … [I]t is impossible to pierce the moneyveil in order to get to the premiums on concrete goods. If one penetratesthrough it one penetrates into a void.

(ibid.: 184)


In short, ‘the money form [of interest] is not shell but kernel’ (ibid.). Even so,the reason why, according to Schumpeter, the rate of interest is not a real but amonetary phenomenon differs from the explanation provided by Keynes. ForSchumpeter, the monetary nature of interest derives from the fact that the rateof interest is a portion of, a ‘tax’, on monetary profits. However, the main differ-ence between Keynes’s and Schumpeter’s perception of the rate of interest lieselsewhere. To see this, a broader comparison between Schumpeter’s work andKeynes’s analysis in the GT is called for.

To begin with, recall the main assumptions and conclusions of the GTregarding the status of money, financial markets and interest rates. As seen previ-ously, money is essentially thought of as a store of value. While this interpretationis necessary to refute Say’s law and to demonstrate the possibility of permanentunemployment, it also has further consequences for the analysis. In the first place,the concept of money as a pure financial asset – as ‘fund-money’ (Arena 1982:434) – tends to be overemphasised at the expense of the role of ‘circulatory’money (Arena 1985). Consequently, money as a store of value tends to overlook,even though not necessarily to exclude, what Robertson (1966: 161) called theforgotten but simple truth that people often acquire money not in order to hold itbut in order to use it. Moreover, in the GT Keynes assumes an exogenous moneysupply.9 Remarks such as ‘the quantity of money as determined by the action ofthe central bank’ (Keynes 1936: 247) or ‘the quantity of money created by themonetary authority’ (ibid.: 205, see also ibid.: 84, 167, 174, 230, 267) can befound throughout the text. This was explicitly noted by Schumpeter, who arguesthat, in the GT, Keynes assumes that ‘the quantity of means of payment [is]externally given, i.e. [is] freely malleable by governments and central banks’(Schumpeter 1954: 1176), whereas, in his earlier writings dating from the1920s, ‘he actually only accepted the equation of exchange’ (ibid.: 1102), that is,‘a formal relation [between the money supply, its velocity, the price level andthe volume of transactions] without any causal connotation’ (ibid.: 1096).

Another central feature of the GT is that financial markets receive starbilling. Keynes gives two reasons for this. First, in chapters 13 through to 15 heargues that the cost of borrowed funds is an important determinant of invest-ment. Second, in chapters 12 and 22 we find the argument that financialmarkets will inevitably be affected by waves of shareholder optimism orpessimism which will influence entrepreneurs’ calculations of the marginal effi-ciency of capital and may, ultimately, even dictate their investment strategy.

Furthermore, as will be seen below, Keynes’s analysis of the determination ofshort- and long-term rates of interest and their effect on investment decisionsdiffers drastically from Schumpeter’s. Even before 1936, Keynes was alreadyconvinced that the short-term rate of interest was of little significance as adeterminant of investment. Thus, in his correspondence with Hawtrey in 1935,Keynes pointed to the minor importance of short-term interest charges as ashare of total cost, arguing that therefore they were a matter of little concern fordecision-making in business. Even though the argument put forward is different,he confirmed this point of view in the GT:


The short-term rate of interest is easily controlled by the monetaryauthority, both because it is not difficult to produce a conviction that itspolicy will not greatly change in the very near future, and also because thepossible loss is small compared with the running yield (unless it isapproaching vanishing point).

(Keynes 1936: 202–3)

Therefore, what matters is the long-term rate of interest, which is not alwayseasily controlled by the monetary authorities. The most obvious and suddenvariations in the long-term rate are not caused by changes in the money supply,but are instead linked to changes in the liquidity preference function.

Schumpeter’s approach contrasts strongly with this view of the workings offinancial markets. Above all, Schumpeter is not at all convinced that finan-cial markets have a decisive role to play in the provision of credit foreconomic activity. His view is based on two main arguments. First,Schumpeter perceives of the ‘money market’ as the interaction of two distinctspheres: the ‘sphere of hoards and reserves’ and the ‘sphere of capital’ or‘income yielding assets’ (Schumpeter 1917/1918 [1956]: 176), where the latterincludes the stock market. However, this distinction is of secondary impor-tance to Schumpeter who holds that both markets are essentiallyinterdependent.10 He argues, in particular, that the workings of financialmarkets are subordinate to those of ‘the sphere of hoards and reserves’ and,therefore, to the choices of the banking system. In Schumpeter’s view, theusual dichotomy between short-term loans and long-term assets is, hence,inadequate. Instead, he maintains that capitalist development inherently tendsto create the conditions for a ‘perfect negotiability of all instruments of credit,whatever their legal form may be’ (Schumpeter 1939: 613). Therefore, bondsand shares differ from short-term instruments only ‘technically and by degree’(ibid.: 614) because financial speculation implies a withdrawal of existingbalances or relies on credit creation. Credit and finance, banks and financialmarkets, and hence short- and long-term rates of interest, are intrinsicallylinked to one another, with long rates really representing a ‘trend value’ ofshort rates.

In this framework, the role of long-term rates is hugely more important thanthat of short-term rates, the main reason being that entrepreneurs’ real profitsdo not so much depend on the rate of interest paid on bonds but on interestpaid to the banking system. In Schumpeter’s view, income is divided into wages,bank interest and profits. Thus, interest payments made to banks involve atransfer of real wealth from the industrial to the financial sector. Wheneverinterest is paid to banks, real output is divided into real wages, industrial profitsand financial profits. As Graziani (1989: 27) points out, ‘a financial capital isthus gradually built up by banks, along with an industrial capital built up byfirms’. This reinforces an essential point made by Schumpeter, to wit, the factthat interest paid to the banking system is a ‘tax’ on profits and consequently abrake on development. Finally, in the Schumpeterian framework, banks are the


main financial institutions, which is why they are at the centre of much ofSchumpeter’s focus and interest.

The role of banks

While Schumpeter was writing his 1910 article ‘On the nature of economiccrises’, followed in 1912 by his Theory of Economic Development, Keynes waspreoccupied with the unsatisfactory state of monetary theory in Britain,which, in his judgement, was reduced to a ‘matter of oral tradition’ (Keynes1983: 375). Anglo-Saxon thought on the theory of bank credit had indeedfallen somewhat behind developments in continental Europe. The view ofbanks as creators of money, inherited from MacLeod (1855), had lost groundin academic circles, though retaining some popularity in banking circles.11

MacLeod’s theory of bank credit had been violently attacked by Cannan,who had instead proposed his famous ‘cloak-room theory of banking activity’(Cannan 1921), which held that no single bank, nor the banking system asa whole, could ever lend more money than the deposits it had collected.Cannan’s view, according to which banks collect savings and then use themin order to finance investment, had become the official doctrine whichKeynes had to confront, and would ultimately discard, in his Treatise.

Not surprisingly, then, Keynes’s main goal in the TM is to show that creditgranted by banks is not limited by savings, or deposits, already in existence (thatis, banks can ‘create’ deposits). Nonetheless, Keynes still felt it necessary tospecify the constraints under which banks operate and which they have torespect. It is for this reason that he brings to our attention several cases.12 First,there is the case of ‘pure credit’ granted by a single bank: the creation ofdeposits by the bank determines the means of payment deposited with thebanking system, not vice versa. Moreover, the single bank experiences no limitsto this creation of deposits. Second, in the case where there are several banks,the creation of finance by banks is governed by the average behaviour of thebanking system. Finally, in a ‘mixed-money’ system, banks as a whole areconstrained by reserve requirements which depend on the monetary policy ofthe issuing bank.

One problem with this approach is that it fails to specify the analytical deter-minants of banks’ behaviour and of the setting of interest rates. In thetheoretical part of the TM, Keynes appears to reduce banks’ behaviour to themonetary policy pursued by the issuing bank.13 Variations in the bank interestrate are governed by changes either in legal reserves or in the discount rate.14 Inso doing, Keynes makes the interest rate – and, thus, the determinants of banks’behaviour – exogenous.

Schumpeter’s approach is free of such inadequacies. Like Keynes, he showsthat banks are bound by norms of, and limits to, credit supply. Twenty yearsbefore the publication of the TM, Schumpeter concludes that these limits aredetermined by the average behaviour of banks as a whole. However, contrary toKeynes, he downplays the role of the issuing bank, assuming instead that


we have a banking system grouped around a central issuing bank, but thatthere are no other legal barriers and rules for the gestation of banking busi-ness … This represents the leading case the treatment of which is easilyapplicable to other cases.

(Schumpeter 1912 [1934]: 112–13)

Other than in the TM, on this assumption it is impossible to resort to reserverequirements as the main determinant of the credit supply.

Consequently, Schumpeter needs to specify the endogenous determinants ofthe credit supply curve and of the interest rate. His starting point is a closerexamination of each bank’s supply behaviour. According to Schumpeter, eachbank is faced with the risk of financial loss since it can only succeed in meetingfinancial requests if its customers settle their debts. In order to monitor this risk,each bank has to evaluate the economic projects or activities to be financedagainst the limits to supply set by the average behaviour of banks as a whole.Risk management, then, implies that there are two determinants of bankbehaviour in Schumpeter’s framework: the quality and the total amount ofcommitments. Each bank faces two types of risk: the total risk of bankruptcyrelated to the financing of innovative as opposed to imitative activities, and therelative risk of bankruptcy for those banks adopting a more accommodatingcredit policy than the average bank.15

Schumpeter is thus in a position to define the rules required for securingbanking activity. The crucial point is ‘neither the formal character of the busi-ness to be transacted … nor the security that makes sound banking, butknowledge and understanding of, and proper attention to, the purpose whichthe balances applied for are to serve’ (Schumpeter 1939: 641). According toSchumpeter, these rules consist of ‘[j]udging the chances of success of eachpurpose and, as a means to this end, the kind of man the borrower is, watchinghim as he proceeds’ (ibid.). This, then, is the basis on which banks determinethe amount and the composition of a loan as well as the rate of interest atwhich it is granted. Hence, the supply of credit presupposes an entrepreneurialdemand for balances. The credit requested will be granted if the bank thinks itproper to meet this demand at terms compatible with the potential debtor’seconomic targets which are, in turn, a function of the latter’s profit expecta-tions.

We can thus conclude that the Schumpeterian credit market is characterisedby a negative relationship between the demand for credit and the interest rate,on the one hand, and by a positive relationship between the supply of creditand the interest rate, on the other. The shape of the credit demand curve isdetermined by the fact that the rate of interest is a ‘tax’ on profits. The relation-ship between the rate of interest and the credit supply is positive because of thegreater risk accepted by banks, who extend their credit facilities to those poten-tial entrepreneurs who have been out-selected at lower levels of the rate ofinterest, and because of the likely depreciation of capital through inflation. Theequilibrium level of the interest rate is established on the basis of these two


curves (Schumpeter 1912 [1934]: 191–8). That is, the interest rate fixed byeach bank must be equal to the rate of risk assigned to the ‘marginal’ debtor(ibid.: 195–6). Likewise, at the aggregate level, the equilibrium level of interestmust equalise the profit rate expected by the ‘marginal’ entrepreneur or imitatorwhose demand for finance has been met with the rate of interest at which the‘marginal’ bank has actually granted the last credit.16

However, as we shall see, the complexity of the Schumpeterian analysismakes it somewhat difficult to pin down the determinants of credit supply,credit demand and the interest rate with more precision. Schumpeter’s ownview on the matter is aptly summarised in the following passage:

[T]here is always, no matter how great the amount of credit in circulation,some demand for credit which remains unsatisfied even though it is able topay the current rate of interest. The productive demand for any commodity,e.g. wool, is limited, at constant quantity of money, by the falling proba-bility of processing continually increasing quantities; by contrast, demandfor credit is self-propagating, in that the consequences of its expansion andincreasing satisfaction go on creating the economic conditions for evenmore credit demand. The more bank money is issued, the more credit isnecessary for the purchase of one and the same quantity of means ofproduction, and the more, also, can economically be invested in theiracquisition.

… The demand for credit makes possible not only itself, but also a corre-sponding supply; and every supply makes possible a corresponding demand,so that supply and demand, in this case do not confront each other as inde-pendent forces. To this extent, therefore, the banks determine not only towhom they will grant credit but also how much credit as a whole they wishto grant and what demand to call forth.

(Schumpeter 1917/1918 [1956]: 207)

On the one hand, the above passage contains important hints about the rolesplayed respectively by entrepreneurs and bankers in the process of determiningthe volume of credit.17 It demonstrates that the volume of credit is a function ofthe interaction between both agents. Firms take the initiative, but banks havethe power to select from among these initiatives, based on their expectations asto the likelihood of borrowed funds being repaid: ‘We know already by whatforces this supply is regulated: firstly, with regard to possible failures byentrepreneurs and, secondly, with regard to the possible depreciation of thecredit means of payment’ (ibid.: 195).

On the other hand, however, this passage also highlights the limitations ofSchumpeter’s analysis of the workings of the money market. In the first place,even when examining the ways in which the credit market operates from withinhis own framework, Schumpeter fails to offer exact conclusions, the mainreason being that credit supply and credit demand are mutually interdependent.Schumpeter’s reasoning suggests that the higher the demand for credit, the


more important its supply. However, an increase in credit will be inflationary,consequently raising the entrepreneurs’ demand for the quantity of creditmoney required to finance the same quantity of means of production, implyingan increase in the demand for credit. Therefore, an ‘uncertain value’ adheres tothe concepts of credit demand and supply due to the permanent instabilityinvolved. In other words, equilibrium levels of credit and of the interest ratecannot be accurately determined. In the second place, Schumpeter’s descriptionof the money market is not very useful for an analysis of the process of creditcreation. Cyclical shifts of the demand for finance affect not only actual butalso potential credit levels (i.e. the maximum credit banks can create in a giveninstitutional context). Moreover, even in a monetary system where bankingoperations are apparently constrained by reserve requirements, credit demanddoes not meet with any technical supply limits. As pointed out in the abovepassage, the reason is credit rationing (that is, the fact that the credit volumeactually supplied by banks is always lower than the potential volume so that theformer is in elastic supply).18 Furthermore, economic development will pushpotential credit in the same direction (for instance, banks are likely to reducetheir reserve ratio and the cash–deposit ratio falls in periods of prosperity).19 Inbrief, in this model, actual credit supply shifts with demand and does not face adefinite ‘ceiling’ of potential credit because the latter moves procyclically.

It follows that, in the Schumpeterian model, even though banks obviouslyperform an important task in the accumulation process because ‘purchasingpower is the vehicle of an essential process’ (Schumpeter 1912 [1934]: 97), andbecause without credit there will be no innovations and no cycles, banks do notdetermine economic fluctuations. On the one hand, at the onset of an upswinginterest is zero and cannot be reduced any further. On the other hand, whenactive innovation is going on, a lowering of the interest rate is of little analyt-ical relevance because of the shifts in entrepreneurial demand (Schumpeter1939: 634–8). In Schumpeter’s words, this means that ‘the analytical schemapresented in this book evidently does not belong to the family of monetarytheories of business cycles’ (ibid.: 142). The actual upper ceiling of the upswingis not set by credit supply, which expands in prosperity. Rather, limits to creditsupply are set by the prospect of the success of innovations since interestdepends on profits, and the risk taken is not the entrepreneur’s but the capi-talist’s (the banker’s) (Schumpeter 1912 [1934]: 75–6, 1939: 104).

Schumpeterian bankers then appear to be both Stiglitz’s and Weiss’s ‘socialaccountants’, replacing the auction market ruled by the law of supply anddemand, and the ‘ephors’ of market economies who assess innovative projects,but do not ‘think them up’.20, 21

Concluding remarks

By taking account of the relationship between finance and economic activity,both Keynes and Schumpeter are able to develop a true analysis of economicinstability which assigns a central role to financial factors. However, the nature


of economic instability clearly differs between the two authors. To use a distinc-tion suggested by Vercelli (1985), one could say that the type of instabilityinvestigated by Schumpeter is ‘physiological’, in the sense that it makes thesurvival and the development of the capitalist system possible, whereas the typeof instability Keynes is concerned with is ‘pathological’, in the sense that itimpedes the performance of an individualistic economic order.

This assertion may offer the key to a better understanding of the actualdevelopment of financial structures from commodity money to sophisticatedcredit instruments in market economies. More specifically, we suggest thatfinancial innovations have been introduced whenever the need was felt toincrease the structural flexibility of the economic system.22 Unfortunately, thedevelopment of financial structures seems to have contributed not only to morephysiological instability but also to more pathological problems, resulting attimes in financial crises, inflation, unemployment and generalised economiccrises.23 Neither Keynes nor Schumpeter paid sufficient attention to this basicambiguity of financial tools and institutions and their evolution over time. Asynthesis of the two approaches would perhaps be the most promising route to amore balanced and articulate view of the role of money in economic fluctua-tions.

Moreover, such a synthesis might be of use for the clarification of an impor-tant empirical issue, namely, the nature of the evolution of the role of financialinstitutions in capitalist economies from the time of Schumpeter and Keynes tothe present. As is well known, financial capitalists no longer act as the ‘ephors’of the economy, providing the necessary finance to ensure the growth of capitalstock. Today’s narrowly focused ‘managers of money’ are preoccupied little withthe development of capital assets and much less with the going concerns ofeconomic and technological progress. Indeed, the great crash of 1929–33marked the end of this stage of ‘finance capitalism’ in which investmentbankers dominated financial markets. Today, with the advent of ‘moneymanagerial capitalism’ (Minsky 1993: 108), financial structures seem muchcloser to Keynes’s characterisation of the financial arrangements of advancedcapitalism as a ‘casino’.

However, this evolution of financial structures does not render Schumpeter’sapproach obsolete. By enriching the analysis with a broader view of theeconomy as a set of evolving institutions, Schumpeter implicitly completesKeynes’s theory of monetary production. Indeed, nowhere is market-driveninstitutional evolution (innovation) more apparent than in the financialsphere. As Schumpeter himself argued, financial institutions are alsoentrepreneurial organisations striving to innovate in order to generate capitalgains (Schumpeter 1947 [1951]: 222). Hence, the rapid changes in the use ofmonetary and financing tools that have characterised the past forty-five years ofsuccessful capitalism would have been easily understood by Schumpeter. Thus,the framework he built remains a valid theoretical tool for explaining the evolu-tion of today’s financially sophisticated economies.


Notes1 In the published text of The General Theory, Keynes uses the terms ‘barter economy’

and ‘monetary economy’.2 Arena 1985; de Boyer 1985.3 See, for instance, Schumpeter 1917/1918 [1956].4 Schumpeter 1917/1918 [1956]: 173, 174 and 177; Schumpeter 1939: 546.5 Schumpeter 1917/1918 [1956]: 179.6 On this point, see Graziani 1978: 91 and Messori 1985: 49.7 Robertson 1966 accused Keynes of having neglected the process of the creation of

liquidity in The General Theory.8 Keynes 1937a: 208. This may be attributed the fact that Keynes was intent on modi-

fying the approach adopted in The General Theory as little as possible. As will beseen in the next section, the stock market there occupies the main position, whilethe banking system seems to be entirely absent.

9 Mainly to facilitate the analysis, see Robinson (1971: 81–2).10 A footnote to The Theory of Economic Development dealing with this problem lends

support to this interpretation. Schumpeter argues there that ‘at the most one maywith Spiethoff distinguish the capital market as the market for long-term purchasingpower from the money market as the market for loans. But purchasing power is thecommodity in each’ (Schumpeter 1912 [1934]: 124 fn. 1). For a more detailedaccount, see Arena and Festré 1996.

11 See Graziani 1989.12 For a more detailed account, see Messori 1986: 131–3.13 See, for instance, Keynes 1930: ch. 17.14 Ibid.: ch. 13, section 2.15 Schumpeter 1912 [1934]: 75 fn. and 137; 1939: 104.16 See Messori 1986: 139.17 For a more detailed account, see Arena and Festré 1996.18 See also Schumpeter 1939: 126 and 640–1.19 Schumpeter 1912 [1934]: 112–15; 1939: 121–3 ; 1917/1918 [1956]: 206–8. This

point has been stressed by Bellofiore 1985, in particular.20 Stiglitz and Weiss 1988.21 Schumpeter 1912 [1934]: 74. The ‘ephor’ was a magistrate of Sparta who contained

and controlled the kings. In Schumpeter’s framework it is the banking system of acapitalist economy which controls the finance of economic activities and only thoseactivities which are financed enter the realm of the possible.

22 As suggested by Schumpeter himself, see 1939: 122.23 In this we follow the ‘Minskyan’ interpretation of The General Theory (Minsky

1975).

Bibliography

Arena, R. (1982) ‘Reflexions sur la compatibilité des approches ricardienne et keynési-enne du fonctionnement de l’activité économique’, Economie Appliquée, 35 (3):405–48.

—— (1985) ‘Circulations, revenu et capital: théorie monétaire et tradition non quanti-tative’, in R. Arena, A. Graziani and J. Kregel (eds) Production, Circulation etMonnaie, Paris: Presses Universitaires de France.

Arena, R. and Festré, A. (1996) ‘Banks, credit and the financial system in Schumpeter:an interpretation’, in L. Moss (ed.) Joseph Schumpeter, Historian of Economics, London:Routledge.


Bellofiore, R. (1985) ‘Les aspects monétaires de la théorie de Schumpeter’, in R. Arena,A. Graziani and J. Kregel (eds) Production, Circulation et Monnaie, Paris: PressesUniversitaires de France.

Boyer, J. de (1985) ‘Taux d’intérêt et quantité de monnaie: note sur la distinction entredeux sphères de circulation chez T. Tooke, K. Wicksell, J.M. Keynes et J.A. Schum-peter’, in R. Arena, A. Graziani and J. Kregel (eds) Production, Circulation et Monnaie,Paris: Presses Universitaires de France.

Cannan, E. (1921) ‘The meaning of bank deposits’, Economica 1, January: 28–36.Graziani, A. (1978) ‘Il trattato sulla moneta di J.A. Schumpeter’, Note Economiche 1:

87–94.—— (1989) ‘Schumpeter and Italian economic thought in the inter-war period’, Studi

Economici 37: 41–83.Keynes, J.M. (1930) A Treatise on Money, reprinted in The Collected Writings of John

Maynard Keynes, Vols V and VI, London: Macmillan (1971).—— (1936) The General Theory of Employment, Interest and Money, reprinted in The

Collected Writings of John Maynard Keynes, Vol. VII, London: Macmillan (1973).—— (1937a) ‘Alternative theories of the rate of interest’, Economic Journal 47: 241–52,

reprinted in The Collected Writings of John Maynard Keynes, Vol. XIV, London:Macmillan (1973).

—— (1937b) ‘The ex-ante theory of the rate of interest’, Economic Journal 47: 663–9,reprinted in The Collected Writings of John Maynard Keynes, Vol. XIV, London:Macmillan (1973).

—— (1938) ‘Mr Keynes and “Finance” [rejoinder]’, Economic Journal 48: 314–22reprinted in The Collected Writings of John Maynard Keynes, Vol. XIV, London:Macmillan (1973).

—— (1939) ‘The process of capital formation’, Economic Journal 49: 569–74, reprintedin The Collected Writings of John Maynard Keynes, Vol. XIV, London: Macmillan(1973).

—— (1970) Das Wesen des Geldes, edited by F.K. Mann, Güttengen: Vandenhoeck &Ruprecht. Translated into Italian as L’Essenza della Moneta, by E. Dal Bosco, Turin:Cassa di Risparmio di Torino, 1990.

——(1973a) The Collected Writings of John Maynard Keynes, Vol. XIII: The GeneralTheory and After: Preparation, London: Macmillan.

—— (1973b) The Collected Writings of John Maynard Keynes, Vol. XIV: The GeneralTheory and After: Defence and Development, London: Macmillan.

—— (1979) The Collected Writings of John Maynard Keynes, Vol. XXIX: The GeneralTheory and After: A Supplement, London: Macmillan.

—— (1983) The Collected Writings of John Maynard Keynes, Vol. XI: Economic Articlesand Correspondence: Academic, London: Macmillan.

MacLeod, D. (1855) Theory and Practice of Banking, London: Brown, Green & Long-mans.

Messori, M. (1985) ‘Il credito nel modello di Schumpeter’, in C. Filippini and P. Porta(eds) Società, sviluppo, impresa. Saggi su Schumpeter, Milan: Instituto Ipsoa.

—— (1986) ‘Financement bancaire et décisions de production’, Economie et Société,Série MP, 8–9: 127–58.

Minsky, H.P. (1975) John Maynard Keynes, London: Macmillan.—— (1993) ‘Schumpeter and Finance’, in S. Biasco, A. Roncaglia and M. Salvati (eds)

Markets and Institutions in Economic Development: Essays in Honour of Paolo SylosLabini, New York: St Martin’s Press.


Robertson, D.H. (1966) ‘Mr Keynes and the rate of interest’, in Essays in Money andInterest, London: Collins.

Robinson, J. (1971) Economic Heresies, New York: Basic Books.Schumpeter, J.A. (1912) Theorie der wirtschaftlichen Entwicklung, Leipzig: Duncker und


—— (1917/1918) ‘Das Sozialproduct und die Rechenpfennige. Glossen und Beiträge zurGeldtheorie von heute’, Archiv für Sozialwissenschaft und Sozialpolitik, Vol. 44. Trans-lated into English by A.W. Marget as ‘Money and the social product’, InternationalEconomic Papers, 1956.


—— (1947) ‘The creative response in economic history’, Journal of Economic History,November: 149–59. Reprinted in R.V. Clemence (ed.) Essays on Entrepreneurs, Inno-vations, Business Cycles and the Evolution of Capitalism, New Brunswick and Oxford:Transaction Publishers, 1989. Originally published in 1951 by Addison-Wesley.


—— (1970) Das Wesen des Geldes, F.K. Mann (ed.), Göttengen: Vandenhoeck &Ruprecht. Translated into Italian by E. Dal Bosco as L’Essenza della Moneta, withintroduction by G. Nardozzi, Turin: Cassa di Risparmio di Torino, 1990.

Stiglitz, J.E. and Weiss, A. (1988) ‘Banks as social accountants and screening devices forthe allocation of credit’, NBER Working Paper, No. 2710.

Vercelli, A. (1985) ‘Money and production in Schumpeter and Keynes: twodichotomies’, in R. Arena, A. Graziani and J. Kregel (eds) Production, Circulation etMonnaie, Paris: Presses Universitaires de France.


Aftalion, A. 128agents 60, 94, 100, 108–9, 111, 131;

adaptive behaviour of 187, 192–3, 209;Austrian 34; banks as 208, 230; creditand 217; money and 7; new consumergoods and 172; responses 113, 116,time horizon 117; see also economicagents

Aghion, P. 129Allen, R.L. 219American institutionalists xiianti-trust measures 149Arena, R. xi, xiii, xv, 1–17, 40–63, 93, 98,

167–81; circulatory money 246;economic sociology 97; Schumpeter’scircular flow and 57, 108, 112

Arrow-Debreu theory of generalequilibrium 56

Austrian school 44, 119, 161; ‘causal-genetic’ explanations 12, 31–3; Hayekand 221; literature on social leadership167; logic of circular flow and 27–31;marginalist tradition and 21–2; moneyand credit 221; pure competition 185;Ricardo and 76; Schumpeter and xiii22, 26, 31–5, 162, 221; sociology ofcapitalism 16; subjectivism 33, 50;theory of costs and distribution 31

bank credit and productive dynamics,Schumpeter and Hayek 228–36

Banking and Currency School 243banking xiv, xvii , 9banks 95, 97, 101–2, 137, 147; behaviour

and risk management 249; dynamics ofaccumulation and 233; industry and151; main financial institutions 247–8;role of 231, 248–51; Schumpeter and248–9; social accountants and 205–8,

215, 251; winner’s curse problem 234Barone, E. 41; Principi di Economia Politica

90BC 128–30, 133, 151; circular flow and

109, 116; competition in 191, 194,198; discussion of ‘Enterprise’ 169;dynamic phenomena and 71; economicchange 107, 113–14; entrepreneur,function of 10, 171, 173; evolution and117; ‘How the economic systemgenerates evolution’ 7–8; innovationsand 99; money and 203, 205, 211;Schumpeter’s view on rationality 60;task of economic analysis 69; Walrasand 41; Walras, theory of tâtonnementand 51, 57–8

Bellofiore, R. 229Bentham, Jeremy 158Blaug, M. 66–7, 84, 173, 180; Economic

Theory in Retrospect 66Böhm-Bawerk, E.: Austrian school and

162; capital and interest 33; circularflow and 22, 30, 35; concept ofproduction xii, 36; dynamicsubjectivism 34; marginalism 27, 30,79, 89; Positive Theory of Capital 30;static analysis 57

Bowley, A.: Mathematical Groundwork ofEconomics 42

British Association for Science 89bureaucratisation 154, 156Business Cycles (1939) see BCbusiness cycles, analysis 187–8, 192;

banking institutions and xvii;disequilibrium trajectories 188;monopolistic competition and 197;Schumpeter and Hayek 237;Schumpeter’s theory xiv, 107, 129–35,143, 233, 237; Schumpeter’s theory,

Index

mechanisms in 134–43; Schumpeter’stheory, weaknesses of xv

Calzoni, G. 44Cannan, E. 248capital accumulation 61capital formation 13capitalism xv, 16, 94–5, 100, 101–2,

103n.4, 107; big business 191;competitive 119, 146, 157, 161;definitions 147–8, 152, 153, 156, 170;democracy and 157–62; disappearanceof entrepreneur as driving force 153–6;economic history of 176; evolution146–9, 151; institutional decline of152–6; logic of 161; Marshall and 82;Schumpeter’s historical analysis and120, 154, 162, 199, 224; Schumpeteron success of 149, 156, 252

Capitalism, Socialism, and Democracy seeCSD

capitalist economies, change and diffusionin 118–19; conflict between capitalistsand entrepreneurs 152; disequilibra in115; divisions of 101; financial marketsand 7; innovations as core movementsof 143, 150; money and credit in xii,xvi; monopoly element and 189;reasons for excluding basic factors 134;Schumpeter and dynamics of 113, 129;Schumpeter and large conglomerates191; stationary flow and 212

capitalist expansion, monopolisticpractices and 149

Carlin, E.A. 168Cassel, G. 42, 185Chamberlin, E.H. 196, 198change, endogenous process xvi, 181Châteauneuf-Malclès, Anne xiv 106–20circular flow xiv, xv, 13, 22, 27, 100;

competition and 187–9; developmentand 106; dynamic analysis and 135,181; economic evolution and 22–6, 36;from commodity money and exchangeto credit and development 62–3; fromgeneral equilibrium to economicdevelopment 61–2; from routine toinnovation 59–61; general equilibriumor stationary state 56–7; hedonisticegoism in 100; individual decision-making and 112; innovation and 118;labour and land and 139; ‘normal’motivation 112; not perfectlystationary 60; pure economics and

93–6; routine or tâtonnement? 57–9;Schumpeter and analysis 32, 44, 95–6,107–8, 132, 241–2; stationaryequilibrium and 228; theory of moneyand 206; time in 107–10; time in,theory of, versus time in theory ofdevelopment 107–8; Walrasian features116

Clark, J.B. 79, 185classical economic theory 89classical monetary theory, rejection of

241–5clustering, new enterprises and 115commodity money 209, 214, 217, 224commutative justice 47competition xii, xvi, 1, 8, 15, 146; circular

flow and 187–9; concept and economicthinking 184; economic developmentand 189–94; entrepreneurial theory ofxi; firm is dynamic 115; price 184;Schumpeter on 198

competitive equilibrium xvi , 193concept of classes 5continuity in time, network of

connections 108core of economics 43Costibile, L. 91, 93Cournot, A. 185, 196creative activity 3–4creative destruction xi, 14, 96, 156–7credit xv, 3, 14, 25–6, 34–5, 138, 151;

agents and 217; bank 9, 151, 215, 222,225, 227; capitalism and 147, 170, 175;creation 243; cycle 244; economy and237; entrepreneurs and 101, 137;innovations and 132; institutions 62,119; potential 232, 251; role of 102,176, 212, 215, 225, 243; role of andmoney tie 215–16; roles ofentrepreneurs and bankers 250;Schumpeter and capitalist system 148,242; selection of economic agents 155

credit money 244–5‘crisis of the tax state’ (CTS) 97–8critical figure (die kritishe Ziffer) 208,

210–16CSD 41, 90, 146, 163, 205; capitalism in

149, 160; competition 184, 198;entrepreneurs and 167, 169, 171, 173;pure economics 95; Schumpetercommenting on Marx 49

Index 257

Dangel-Hagnauer, Cécile xvi, 1–17,203–220

Das Sozialprodukt und die Rechenpfennige(Money and the Social Product) 203

Das Wesen des Geldes (The Essence ofmoney) 203–5

Das Wesen und der Hauptinhalt dertheoretischen Nationalökonomie see DW

De Vecchi, N. 116, 119decision-making, element of time 111decreasing costs, theory of imperfect

competition and 82democracy 157–9demographic changes 61Demsetz, H. 181depression 133, 141–2, 150, 156, 189, 228Diamond, D.W. 203disequilibria 99, 110, 222, 227, 233, 242distributive justice 47Dockès, P. 48, 52Donzelli, F. 53, 92, 100Dosi, G. 15Duval, Nathalie xiii, 66–84DW 2, 6, 12, 40–42; methodological

individualism chapter 53; ‘Nature andessence of exact economics’ chapter 44;‘pure’ economic theory 91, 95;rationality in 5–6, 60

dynamic analysis 57, 71, 94; economicsociology 96–7; innovations and 99;Marshall and 83; static analysis and 71

dynamics, economic activity and 3, 12–13,71, 181; extra-economic factors 24;Schumpeter’s analysis 101, 106, 135–6;statics xiv 56

economic agents 112, 115, 118;entrepreneurs as 137; groups of 173,175; innovations and 140, 151, 153,155, 215; interests of 208–9

economic analysis xiii, 2, 45; definition inHEA 84, 97, 205, 216; mathematicaltools 27, 41, 91

economic development 13–15, 60, 95,100, 136; competition and 189–94

economic dynamics, capitalistdevelopment and 170; economicdevelopment and institutional change167; ‘ideal equilibrium’ 14;methodological approach 13–15, 17;role of institutions and 22;Schumpeter’s concept 237

‘Economic Evolution’, definition 114economic evolutionary theory 1

economic history 1–2economic instability 194, 251–2economic life, increasing state

involvement in 159–63; the money tieand 215; trustification of 191

‘economic logic’ 43, 63economic organization, pure economics

and 45economic rationality 22economic relationships, social

relationships 31economic sociology xii, xvi, 1–2, 14–16,

46, 61, 92; analysing money and216–17; bridge between economictheory and 7, 10, 174; definition 97;dynamic analysis 96–7, 170;entrepreneurs and 101; innovationsand xiv, 100; institutions and 102, 206;intermediary 167; ‘pure’ economictheory and 90; Schumpeter’s definition3; sociology and 98

economic structure, individual needs 29economic theory 2, 7, 10, 45–6, 174; aim

of 205–6; time in 106economics and history, dynamic theory

and 71economist-historian, economist-

theoretician and 70, 84economy, dynamic and evolutionary

process 162; persistent habit and stableeconomic relations 117–18; problem ofstickiness and rigidities 193

Edgeworth, F.Y. 42, 73, 80Elliot, J.E. 127energetic egoism 60, 100energetic rationality13, 25–6, 34, 177;

entrepreneurs as agents of 25;hedonistic rationality 5, 8

entrepreneurial competition 8, 14; naturalselection and xv

entrepreneurial function 170–1, 175, 189entrepreneurial rationality xv, 34;

economic aspects in Schumpeter 177–9entrepreneurial success 190entrepreneurs 5, 8, 10, 13, 16, 100;

borrowing and 230; capitalists and 168,173; creative role of 34; credit creationand 243; dynamic concept 22;evolution and 112; not permanenteconomic agents 175; remuneration of29; risk-bearers 25; Schumpeter on112–13, 136, 175; social leadership and167–9; typical debtors 243; ‘upstarts’ in

258 Index

capitalist society 176; Walrasianconception xv 59; ‘windfall gains’ 175

entrepreneurship xv, 26; criticisms ofSchumpeter’s theory 180; decline of172–3; economic concept of 175;historical and sociological aspects ofSchumpeter’s concept 169–74;leadership under capitalism 181;Schumpeter’s approach 167

equilibrium, economic logic and 71; ‘idealequilibrium’ 14, 24; neighbourhoods of14; tendency towards 58, 109, 114,192; theory of 61, 82, 90

equilibrium theory, mathematisation of 78ethics, pure economics and 48evolutionary dynamics, endogenous shocks

25evolutionary economics, ‘neo-

Schumpeterian’ xi, xii, 163evolutionary explanations, Schumpeter

and 11–14, 98, 107‘evolutionary’ reasoning, dangers of 12excluded strategy, Schumpeter and 186exogenous circumstances, dynamics of

circular flow 25, 111exogenous shocks, technical change and

innovation 129

faits humanitaires 50faits naturels 50Festré, Agnès xiv, xv, xvi 101, 127–44,

221–37filiation of ideas 67, 72fiscal sociology 2Fisher, Irving 79forced saving, Schumpeter and 229,

232–3, 237free competition 45, 47, 184–5Freeman, C. 15, 128, 137Frisch, R.A.K 135, 229

Galton, Francis 89–90general equilibrium 22–3, 44–5; circular

flow and 24; distance betweenSchumpeter and Walras 57;fundamentals of model 44; Walras’smodel 27

general equilibrium theorists xiiiGeorgescu-Roegen, N. 107–8, 120German Historical School xii, 44, 119Gislain, J.-J. 92Gloria-Palermo, Sandye xiii , 21–36, 98Graziani, A. 94, 243, 247

Grilli, C. 94Guilbaud, G.T. 111, 113

Hansen, A.H. 128–9Hayek, Friedrich August von xv, xvii,

130–1, 140, 161, 221; banking systemand 236; deviation from equilibrium228–9; forced saving 229, 235; idealstate 237; interest and 226; MonetaryTheory and the Trade Cycle (MTTC)227, 234, 236; money and credit221–2; Prices and Production (PP) 227,234–6; Profits, Interest and Investment236; role of inflation 232; onSchumpeter’s theories 230, 233

HEA 1–2, 14; dynamic economics and 71;imperfect competition 194; Marshalland 80, 82; ‘Marshall’s Attitude andReal Cost’ 76; money and 203, 206;Schumpeter on the competitivehypothesis 185–7; Schumpeter and‘Filiation of Scientific ideas’ 67, 72;Schumpeter as historian and 66;Schumpeter’s ‘Classical Situations’ in73, 75–6; Schumpeter’s definition of‘Economic Analysis’ 68, 84, 97, 205;static equilibrium theory 70; Walras 41,43, 49; Walras and theory oftâtonnement 58; Wicksell’s work and237

hedonistic egoism 23, 96, 100, 112hedonistic rationality, energetic rationality

5, 8, 17, 26, 34, 177Heertje, A. 180–1Heilbroner, R.L. 107, 109Herland, M. 47Hilferding, R.: Financial Capital 151history xvi, 1, 15, 45, 61, 174; economic

development 95; economic thoughtand Schumpeter 66–8; institutionalchange and 8–11

History of Economic Analysis see HEAHodgson, G. 16, 98homo economicus 33, 50, 92, 179Howitt, P. 129hypothetical deductive approach, ‘pure’

economic theory and 90–6

‘ideal equilibrium’ see equilibriumimperfect competition xvi, 82, 120, 175,

184–5, 194–8imputation theory 28, 31, 33, 36individuals, definition 56innovations xii, xv, 8, 16, 25–6, 60; agents

Index 259

as endogenous factors in 100, 151;asymmetric effect on economic agents140; business cycles and 128; circularflow and 25; credit and 228, 230;diffused 6, 117, 138; exogenous shocks99; factor driving changes and 136,142; full employment generalequilibrium and 137; monopolisticbehaviour and 150; qualitative jumps110; savings and 132; secondary wavesof 137; selective competition 1;technical progress 175

innovative sector, learning and knowledgeimprovements in 116–17

institutional change 8–9institutional inertia 172, 176; ‘classes’ of

reasons 10; market economies and 11‘institutionalist’, meaning xiiinstitutions, analysis of 96–7, 102;

challenging legitimacy of capitalist156–7; entrepreneurs and 99–102;evolution of 97–9; forms oforganisation and 5–8, 17, 31, 34;instability of capitalist order and 162;Schumpeter on role of 162, 216; studyof 46; Wieser’s theory of 35

instrumentalism 52–3intellectuals, protest movements and 157interdependencies 70, 72interest rates 95, 101, 130, 135, 216, 223,

227; equilibrium level of 231;relationship with credit supply 249–50;Schumpeter and 225, 245–6; short-term and long-term 245–8; ‘tax’ onprofits 249

Jaffé, W. 21, 47–8Jevons, W.S. xiii, 73, 77, 79, 96;

marginalism 21, 89–90; Marshall and76; subjectivism 33

Kaldor, N. 97Keynes, Maynard xvii, 7, 129, 131, 159,

161; advanced capitalism and 252; co-operative and entrepreneurial economy241–2; finance and 245; GeneralTheory 128, 160, 241, 244–6; interestand 226, 246–7; monetary theory inBritain 248; money and credit 222;similarities with Schumpeter’s analyses245; Tract on Monetary Reform 160,248–9; Treatise on Money 205, 244, 248

Keynesian-Robertsonian-Swedish,business cycle theory 131

Kirzner, I.M. 36

Knapp, G.F. 7, 206Knight, Schumpeter’s debate with 29–30Kondratieff cycles 99‘Kondratieff-Juglar-Kitchin three-cycle’

scheme 133Kuznets, S.S. 143

Lachmann’s theories 36Laidler, D. 236Lakomski, Odile xv, 146–63Lallement, J. 47Lausanne School 41–2leaders xiii, xv, 3–6, 8–10, 16, 31, 34–5,

99, 139,152, 167–9, 171–2, 175–6; Weber’s

charismatic leader 167leadership xii, xv, 3–6, 8–11, 16, 54, 153,

167–71, 173,175–6, 181, 196Legris, André xiv , 89–103logic of economic theory 216

Macleod, D. 248McNulty, P.J. 198Madaràsz, A. 127, 134magna charta 42–3, 45–6, 63, 69, 93Mann, F.K. 204‘march into socialism’ xv, 159, 161 see

Schumpeter, J.A.Marget, A. 204, 211marginal productivity, Marshall and 80marginal utility principle 21, 27, 31, 62,

77marginal utility school see marginalistsmarginalism 27, 34, 72, 77, 89–90marginalist revolution 21, 27marginalists xii, 22, 26–7, 30; marginal

utility school 74; static equilibriumanalysis 76, 83

Marshall, Alfred xiii, xiv, 12, 42, 108;competition and 185; doctrine ofnormal profit 80; driving forces ofcapitalism 82; economic theory and 66;influence of English Classics 75; linkwith Smith and Mill 76; marginalistprecursors 77–81; partial equilibriumanalysis 79; Principles 73–4, 76, 79–80;problems with time 192; Schumpeter’sinterpretation of 72–5; ‘scientificutilitarianism’ 73; static equilibriumtheory 72; substitution principle 80;theoretician of dynamics 81–3; theoryof production and 81

Marshallian thought, classical roots of

260 Index

75–7, 84; impact of analytical standardof evaluation 70–2; marginalist analysisand 79; ‘ordinary business man’ 50, 61;Schumpeter’s reconstructive approach66–72; Schumpeter on weakness in 81;Schumpeter’s standard ofreconstructive evaluation 68–70

Marx, Karl: analysis of investment 53;capitalism and 148, 161, 163; CSDcomment on 49; entrepreneurialactivity 169; history and 170; referencein TMB 207; Schumpeter’s intellectualdebt to 146

mathematical methods 27, 41, 91, 210Mayer, H. 32–3Menger, C. xiii, 10, 21–3, 26–30, 33–4,

79; Grundsätze 35–6; Principles 168;Untersuchungen über die Methode 36

Menger, Karl 27Messori, M. 203–5, 243Methodenstreit 89methodological approach xv, 172;

combining history with economictheory 167; Marshall and 81–3;Schumpeter and 170; social analysisand 92; static analysis and 94

methodological individualism,Schumpeter and 53–4, 92, 100; Walrasand économie pure 55–6

methodology, Schumpeter and 67–8;Walras and Schumpeter 51–3

Mill, J.S. 68, 75; Principles 185Minsky, H.P. 252Mises, Richard von 140, 161, 221, 234money, banking and dynamics,

Schumpeter and Hayek on 222–8money, banks and 147; in changing

economy 211–16; claim ticket andreceipt voucher 7, 207–8; credit and 6;difficulty in defining quantity 243–4;institutional forms 207, 210–11,213–14, 216–17; interest attaches to245; means of payment function 211;Schumpeter and quantity theory of243; Schumpeter’s circular flow and 57;Schumpeter and value of 217, 242;store of value (Keynes) 246; technicaldevice xvii, 6–7, 208; TMB and 203;unit of account 102, 152–3, 208–10,216, 224

money market 243; definition 230–1;Schumpeter and 247, 251

money tie, role of credit and 215–17monopolies 147, 153, 189, 194–5, 198–9;

competition in 195, 196, 197–8; pricesand 149–52, 189

Morishima, M. 47–8

Nasica, Eric xvii, 241–52Nelson, R. xi, 15neo-classical economics 22, 96‘New Men’ 6, 9, 172‘normal business situation’ 115, 118, 187‘normal depression’ 117–18

Oakley, A. 109O’Driscoll, G. 33, 108, 120‘old’ Austrian school xiii, 31, 35; see also

Austrian schoololigopolistic competition 196, 198oligopoly 195, 198opportunity costs 31, 33–4, 36, 76

Pareto, V. 41–2, 79, 185–6perfect competition 82, 149–50, 184,

186–7, 189, 191–2, 198–9Perroux, F. 57, 107, 110, 127Pesciarelli, E. 91, 178Popperian ideas 51, 53prices 139, 141, 150, 193, 198, 233‘principle of economising’ 28principle of substitution 80problem of stickiness and rigidities 193product differentiation 196–7production process 36prosperity 133, 141, 156, 228psychologism 44, 50pure economics 40, 44–6, 69, 95; circular

flow and 93–6; ethics and 48; idealsociety and 46–9; individualistapproach 55; limited validity of 50, 91;self-contained science on economics92; static analysis and 96; Walras andmethodological individualism 55

‘pure’ economics/equilibrium,dynamics/creative destruction 102

R&D xv 144, 154, 172rationalisation 155; of human mind

152–3, 158; of production process 154rationalism 153, 174rationality, adaptive 23; banks and 102;

energetic 13, 25–6, 34, 177; hedonist 5,8, 17, 26, 34, 177; intentional 113;objective and subjective 179; of theobserved 177–9; of the observer 177–8

Raybaut, Alain xvi, 184–99recession 117, 142, 150, 228

Index 261

Research Center in EntrepreneurialHistory (Harvard University) 173

Ricardo, David 74, 76Ricci, U. 95Rizzo, R.W. 33, 108, 120Robbins, L. 84Robertson, D.H. 244, 246Robinson, J. 108; The Economics of

Imperfect Competition (A World ofMonopolies) 195

role of inflation, Schumpeter 232Romani, Paul-Marie xv, 167–81Rosenberg’s judgement of Schumpeter 180Rousseau, J.-J. 158

Samuelson, P.A. 68, 135Santarelli, E. 91, 178saving 131–3, 135, 207, 225–6; forced 229,

232–3, 237; voluntary 232, 236Saving and Accumulation 130, 226Say, J.B. 49, 176Say’s Law 242, 246Schumpeter, Elizabeth Boody 71, 204Schumpeter, J.A.: analysis of dynamics

101, 106, 135–6; analytical andmethodical approach xii; Austrianschool and xiii, 22, 26, 31–5, 162, 221;banks and 248–9; business cycles xiv,107, 129–35, 143, 233, 237; businesscycles, mechanisms in 134–43; businesscycles, weaknesses in xv; capital and148, 242; capitalism, historical analysis120, 154, 162, 199, 224; capitalism,success of 149, 156, 253; circular flow32, 44, 95–6, 107–8, 132, 241–2;competition xii, 198; concept ofeconomic dynamics 237; credit 148,242; criticism of theory of tâtonnement188; Das Sozialprodukt und dieRechenfennige 203; definition ofeconomic sociology 3; dynamics ofcapitalist economy 101, 106, 113, 129,135–6; economic aspects ofentrepreneurial innovation 174–7;economic aspects of entrepreneurialrationality 177–9; economicevolutionary theory 115; endogenousmoney and 102; entrepreneurs 112–13,136, 175; entrepreneurship 167,169–74, 180; evolutionaryexplanations 11–14, 98, 107; excludedstrategy 186; forced saving 229, 232–3,237; German Historical School xii 44,119; innovations and xii 106, 144, 237,

252; ‘instability of capitalism, The’,(1928) 147; institutions 9, 162, 216;interest rates 225, 245–6; largeconglomerates 191; ‘march intosocialism’ xv, 159, 161; marginal utilityand 62; Marshall, interpretation of72–83; ‘Meaning of Rationality in theSocial Sciences’ (1940) 177; ‘On thenature of economic crises’ (1910) 248;social leadership xv, 167–8;sociological thought 119; static analysis71, 135; technological diffusion 141;theory of economic change 113–18;theory of growth 109–10, 112; theoryof imperfect competition 194–8, viewson division of income 247, see alsoCSD; DW; TED;TMB

‘science, tooled knowledge’ 67science of money (Geldwissenschaft) 206–7science of organisation 2, 7, 45, 55‘scientific utilitarianism’ see Marshall,

AlfredShionoya, Y. 4–5, 15, 52Silverberg, G. 15Smith, Adam 75, 185Smithies, A. 160social accounting 207social classes 3–5, 53–4social dynamics 31social leadership, Schumpeter and xv,

167–8social motives, social classes and 3–5social product 207–8social scientists, subsidiary factors and 4social structure 9, 45sociological individualism 92Soete, L. 15Solvay, E. 207Sosthé, Franck xvi , 184–99stability of economic system, instability of

capitalist order 147–9statical, definition 135static analysis 149; banks and 101;

capitalism and 94; circular flow and 93;economists and 70; historical time andxiv; Marshall and 81; ‘pure’ economictheory and 91–3; Schumpeter and 71,135; subterraneous force and 96

static equilibrium theory xiii , 72, 95, 108,213–14; theory of catallactics 69

statics xiv 56, 78, 134–5static theory, definition of 69; Marshall

and 75; pure theory 71‘static vision of economic life’ 57, 242

262 Index

stationary: circular flow 207, 211–12;definition 135; economy 23; flow,equilibrium and 187, 209, 211, 228

stationary state 223–4steady growth 133–4Steiner, P. 92Stigler, G. 31, 251Streissler, E. 21, 229subjectivism 27, 31, 33–4, 50Swedberg, R. 96, 103n.5, 169

taxation 3Teboul, R. 106, 111, 119–20technological change, lags and 192TED 91, 167, 190, 192, 248; banks 101–2;

Böhm-Bawerk’s criticism 57; businesscycles 128–9, 133, 229; capitalism93–5, 99; circular flow 23, 44, 108–9;competition in 190; credit and capital132, 203, 215; economic development12, 95; egotism and 60; English edition61; general equilibrium and 187–8;Hungarian edition 127; innovations138–9; Japanese edition 31,46 ; magnacharta 43; money 217; rationality 6;Schumpeter’s debt to Walras 41;Schumpeter’s theory of fluctuations113; ‘secondary phenomena’ in 117;time in 107

tendency towards equilibrium 109, 114,116

‘Theoretical Economics’, ‘EconomicTheory’ 71

theory of agency 97, 103theory of competition 15–16theory of development, qualitative

changes 110theory of dynamics 13–14Theory of Economic Development (Die

Theorie der WirtschaftlichenEntwicklung) see TED

theory of economic dynamics xii, xiiitheory of growth 109–10, 112theory of innovation 8theory of money 34–5; circular flow and

206Theory of Money and Banking, The see

TMBtheory of tâtonnement 51, 57–8;

Schumpeter’s criticism of 188theory of value 27, 31, 36, 51; Marshall

and 76, 80Thünen, J.H. von, productivity theory 77;

substitution theory 80

Tichy, G. 204time: of becoming 110; continuous

duration 109; duration- 107, 111, 118;economic decision-making 111–13;evolution and operation of prospective114–15; irreversible 107; Newtonian108; prospective 107, 112, 115, 117–20and rationality xiv; reversible 107;reversible mechanical 109; rupture-107, 111, 114–15, 118; Schumpeteriandynamics and 106, 120n.3 theory ofdevelopment and 110–11; uncertaintyand 113

TMB 9, 203–4, 211, 215–17; ‘Essence ofMoney’ (chapter 9) 209; non-publication of 204–8

trade cycle 130traditional economies 54–5traditional theory, limited validity 61Trautwein, H.-M. 236Trieb (urge) 4trustified capitalism 7, 8, 11, 120, 170,

172, 174

unit of account, money as 102, 152–3,208–10, 216, 224, need for 210,212–14; see also money

Untersuchungen über die Methode seeMenger, C.

Veblen, T. 120Vercelli, A. 252

Walras, Léon xiii, 21–2, 27, 33, 40, 73–4,89; analytical content of self-containedpure economics 44–6; core of theory ofgeneral equilibrium 49–50, 77;deductivism and 52; Elémentsd’Economie Pure 47–8, 93; Etudesd’Economie Politique Appliquée 47–8;Etudes d’Economie Sociale 47–8; libreconcurrence 186; methodologicalindividualism 53–6; monetary theory62, 223; perfect competition 80, 185;Schumpeter and 40–63, 72–83, 119,146; static equilibrium 57; theory oftâtonnement 51, 57–9, 188, 192

Walrasian system: agent 100;capitalisation and money theory 62;competitive market 91, 99; concept ofentrepreneur (‘ne faisant ni bénéfice niperte’) 177; economics 108;entrepreneurial rationality 178;

Index 263

features of circular flow 116;fundamentals 44, 46, 61; generalequilibrium analysis 32, 44, 56, 59, 69,79, 93–4, 187, 206, 209, 216;historiography xiii, 23, 40;mathematical system describing 210;model and circular flow 108; non-economic data 44; ‘pure economics’ 40,45–6, 81; rationality 23, 60; staticeconomics and 71; theory of incomedistribution 49; thought and economictheory 106, 128–9; utility function 100

Weber, Max 153, 167–8Weiss, A. 251welfare economics 82Wicksell, J.G.K. 57, 73, 79, 102, 185, 237;

Hayek and bank credit and 227;interest 226; Lectures 224; money andcredit 221–2, 225; ‘pure cash economy’223–4

Wicksell-Mises theory 227Wieser, F. xiii, 3–5, 10, 22–4, 89;

institutions and 35; Law of Power 168;marginalism 27; opportunity costs 31,33–4, 36, 76; Schumpeter and 89, 162;Social Economics 168; subjectivism 34;theory of value 30–1, 36

Winter, S. xiWitt, U. 136–7, 143Wolfson, R.J. 129Wood, Cunningham, Critical Assessments

of Schumpeter 127

264 Index

ARENA the Contribution of Joseph Schumpeer to Economics - Economic Development and Institutional...

Documents

Transcript of ARENA the Contribution of Joseph Schumpeer to Economics - Economic Development and Institutional...