[rogerio studart] investment finance in economic d(bookos.org)

Investment finance in economicdevelopment

As a result of the liberalisation of the 1980s, the financial system hasacquired a prominent role in developing economies. It is nowconventional wisdom that financial liberalisation is the means tostimulate economic development.

Investment Finance in Economic Development challenges thisassumption and offers an alternative view. The book presents a post-Keynesian approach to the role of banks, financial markets and savingsin economic development. It departs from the conventional belief thatfinancial institutions are mere intermediaries between savers andinvestors, to show that banks have a key, active role in the process ofinvestment finance and growth. Further, financial markets, as the loci ofallocation of financial savings, are shown to have an important role insupporting financial stability during the process of growth.

Rogrio Studart is Associate Professor of Economics and Tutor ofUndergraduate Studies at the Federal University of Rio de Janeiro.

Investment finance ineconomic development

Rogrio Studart

London and New York

First published 1995by Routledge

11 New Fetter Lane, London EC4P 4EE

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

To purchase your own copy of this or any of Taylor & Francis or Routledgescollection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.

Simultaneously published in the USA and Canadaby Routledge

29 West 35th Street, New York, NY 10001

1995 Rogrio Studart

All rights reserved. No part of this book may be reprinted orreproduced or utilized 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 in writing from thepublishers.

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

Library of Congress Cataloguing in Publication DataA catalogue record for this book has been requested

ISBN 0-203-20223-6 Master e-book ISBN

ISBN 0-203-26614-5 (Adobe eReader Format)ISBN 0-415-10866-7 (Print Edition)

To Christine, Rafaella, Camila and Marlene

Contents

ForewordVictoria Chick

vi

Preface ix

Acknowledgements xi

List of abbreviations xiii

1 Introduction 1

2 Finance and economic development: The dominance ofthe prior-saving argument

5

3 Departing from the prior-saving argument: Finance in amonetary production economy

26

4 Saving and financial markets in economic growth 47

5 Financial structures, financial development andeconomic development

63

6 From theory to evidence 78

7 Financial system and industrialisation in Brazil, 194766 94

8 The financial reforms and the economic miracle 117

9 Towards the lost decade: The financial system in theimbalanced growth (197383)

141

10 Conclusion 171

Notes 179

Statistical appendix 199

Bibliography 219

Index 230

ForewordVictoria Chick

Rogrio Studart worked out the ideas in this book as a PhD student atUniversity College London from 1988 to 1992. It was great funwatching the thesis develop through many intense discussions, seminarsand summer barbecue parties. While these ideas were developingRogrios wife Christine was pursuing a Masters degree and they hadtwo lovely daughters. Their time in London was fruitful indeed.

The feature which has long served as the core for developmenteconomics as a separate subject is surely the recognition of the specificcircumstances of individual countries, yet the mainstream analysis ofthe financial problems of developing countries is characterised by aproposition long left behind by the institutional structure of most ofthose countries: that the main inhibitor of investment is insufficientsaving. This proposition was valid in eighteenth and even earlynineteenth century Britain, where the development of financialinstitutions lagged behind the needs of the transforming economy; but itis no longer valid in countries with developed banking and financialsystems of their own or access to the international financial markets.

Rogrio Studart takes this proposition as his starting point and showsthat it is the foundation of two of Keyness propositions: thatinvestment precedes and causes changes in saving, and that finance isindependent of saving. Studart uses these propositions to provide athorough critique of two-gap analysis and the potent modern panacea offinancial liberalisation. He substitutes for these conceptions an analysiswhich develops the institutional foundation of Keyness distinctionbetween finance and saving, finding a role for saving in the funding,rather than the financing, of investment.

Funding is shown to be important as a means by which the liquidity ofthe banking system is perpetually restored, thus mitigating the danger ofyet further investment: it is essential to find finance, but it is

equally essential to fund, that is, to find a permanent home forinvestment in stable portfolios.

Studart draws a lesson from the now advanced countries during theirstage of early development. The Industrial Revolution was, by contrastto todays development ambitions, a very gradual affair, in whichfinancial institutions were growing up along with forms of corporateownership and capitalist organisation. In this climate of gradual growth,it was possible to imagine steady, stable growth, though in fact theprocess was punctuated by crises. Why then should we be surprised tofind modern development, where both the rate of growth of theeconomy and its structural transformation is expected to proceed somuch more rapidly, devoid of difficulty?

Studarts framework unifies the experience not only of modern andearly development but of developed and underdeveloped countries, forit fully acknowledges the influence of institutions, real and financial, onthe process of growth and development. The unifying principle whichhe proposes, as a replacement for the institution-free notion ofefficiency, is functionality: a financial system is deemed functionalwhen it is able to provide finance and funding in a way which supportsdevelopment without substantially increasing the fragility or illiquidityof the financial system.

The theoretical framework is then applied to the Brazilian experiencefrom 1947 to 1983. Detailed examination of the institutional structurethe interrelations between the banks, foreign lending, the central bank,and governmentexposes the shifting sources of stress in the system.He particularly demonstrates that the reforms of 19645 were a mixedblessing. After 1983 currency reform plans were attempted at frequentintervals. Under these circumstances there was not sufficient stability inthe institutional structure to permit the application of a theoreticalframework without getting into great detail with little prospect ofderiving general results. In any case the basic point has been made: thedevelopment of the Brazilian banking system and access to theinternational system is such that finance is not the operative constrainton growth and development in Brazil. The problem lies in the realinvestment decisions perhaps, but also, unambiguously, in anunderdeveloped and flawed mechanism for funding investment andcontrolling the liquidity pressure of rapid development. The liquiditypressure has, in various ways which are documented here, been taken onby Government rather than the private sector. Thus the experience doesnot mirror the crises and bankruptcies of the early British experience; ithas, rather, resulted in inflation of staggering proportions.

vii

Rogrio Studart presents in this book an integrated approach to thefinancial preconditions for growth and development, with applicabilityto any country, advanced or developing. Simultaneously he shows thatthe general principles here developed manifest themselves differentlyaccording to the different institutional structures prevailing. The theoryis general; the application is necessarily specific. There is great scope forfurther applications.

The thesis on which this book is based won the Sayers prize for thebest thesis on a topic of monetary theory or monetary history in theUniversity of London. Richard Sayers would have been pleased, and Iam very proud.

Victoria ChickUniversity College London

viii

Preface

One common assumption in models of finance and economicdevelopment is that saving is a precondition to investment andeconomic growth an assumption which we call hereafter the prior-saving argument. The two-gap models, for instance, claim that externalsaving is required for development if both the investment-saving andthe import-export gaps are to be overcome. The prior-saving argumentis also present in the financial liberalisation models, which maintainthat internal saving/ investment can be increased by stimulating savingswith positive interest rates and by enhancing the competition betweenfinancial institutions through financial deregulation.

The mention of these two models in this preface is not random, asthey have been influential for policy purposes in multilateral agencies asmuch as in less developed countries (LDCs) governments. Forinstance, the two-gap models were fashionable in the 1970s, and were toa great extent used to support the mounting of foreign debt by LDCs.When the debt crisis began in 1982 and voluntary capital inflowsrapidly declined, the financial liberalisation models became the maintheoretical foundation behind the financial policies advocated bymultilateral development agencies such as the World Bank and theInternational Monetary Fund to LDCs.

This book fully accepts the importance for economic development ofefficiency of the allocation of real and financial resources, andsubscribes the view that financial development is an important facet ofthe process of development. However, it also claims that the prior-saving argument is a fallacious foundation to understand the problemsconcerning the financing of growth and is misleading as the basis for apolicy towards financially sustainable development.

That the prior-saving argument is a pre-Keynesian concept isrecognised by many post-Keynesians. Notwithstanding, few haveexplicitly acknowledged the full consequence for the analysis of finance

and growth, let alone of finance and economic development, of thereversal of causality between investment and saving proposed byKeynes. This reversal in turn relies on a sharp distinction betweenfinance and saving. In this book, this distinction is used in search for analternative approach to the role of banks, saving and financial marketsin the process of development, along postKeynesian lines.

Our adherence to post-Keynesian theory does not mean, though, thatthis theory can be straightforwardly used to analyse problemsconcerning the financing of economic development. Post-Keynesiantheory is based on a specific type of financial system, with well-developed banking and non-banking financial institutions and marketsfor a diversified range of financial assets. Most developing countries, incontrast, do not have developed financial markets, and growth has todepend heavily on bank credit. Such credit-based financial structuresneed to develop alternative institutions to financeand, especially, fundlong-term investment, to avoid the risk of financial instability andother possible adverse side-effects of growth. Therefore, the theoreticalmodifications to post-Keynesian theory, necessary in applications todeveloping countries, are discussed in this book in the context of theBrazilian case.

x

Acknowledgements

This book was developed out of my Ph.D. dissertation at the Universityof London and therefore represents the final stage of a long trajectorywhich began in January 1989. I cannot thank by name all those who, atdifferent points of this trajectory, inspired my ideas and gave me supportto continue. Among the people I want to single out the first is ProfessorVictoria Chick, my former supervisor, to whom I am intellectuallyindebted and who discussed substantial matters relating to this book,always providing brilliant comments and suggestions on how to proceedwith my research.

I also wish to thank Professor Fernando Cardim de Carvalho, now afriend and a colleague at the Universidade Federal do Rio de Janeiro,who patiently read several draft versions of this study, contributingsubstantially to it. It was under Professor Carvalhos influence that Ibegan studying post-Keynesian economics and to a great extent it wasthanks to his unwavering support and encouragement ever since we firstmet in 1988 that this book became a reality.

Thanks are also due to Dr Sheila Dow and Rogrio Sobreira Bezerra,who read parts of my dissertation and provided me with insightfulcomments; and to Professors Philip Arestis and Laurence Harris, whogave detailed comments on earlier versions of this study.

I also need to mention the late Professor George Shackle and hismost kind wife, Mrs Catherine Shackle. I met Professor Shackle in aconference in Great Malvern, England, in 1988, and was later his and MrsShackles guest for a memorable lunch at his home in Aldeburgh. Eventhough I was only beginning my postgraduate research, ProfessorShackle carefully discussed my ideas, providing me with lessons whichI shall never forget. Even now, it is difficult to express in a few wordshow honoured I feel to have met, to have discussed my ideas withand to have been encouraged by a brilliant academic of the stature ofProfessor Shackle.

Christine, my wife, has been more than just an understandingpartner of this enterprise. She has read, re-read and edited drafts ofseveral versions of this study. In addition, her support and tendercomradeship made it possible for me to bear the most difficult timesduring which this book was written. My family, and especially mymother and my sister Carla, always stood by me, creating an enablingenvironment for writing this book. Adhemar Mineiro offered hiswholehearted friendship when I needed it most. I sincerely thank allthese people.

Finally, the financial support received from the National ResearchCouncil of Brazil (CNPq) is thankfully acknowledged.

xii

List of abbreviations

BCB Banco Central do Brasil (Central Bank of Brazil)BB Banco do Brasil S.A. (Bank of Brazil)BNCC Banco Nacional de Crdito Cooperativo (National

Bank of Cooperative Credit)BNDE(S) Banco Nacional de Desenvolvimento Econmico e

Social (National Economic and Social DevelopmentBank)

BNH Banco Nacional da Habitao (National HousingBank)

CDB Certificados de Depsitos Bancrios (certificates ofbank deposit)

CDC Crdito Direto ao Consumidor (direct consumercredit)

CEF Caixa Econmica Federal (Federal Savings Bank)CMN Conselho Monetrio Nacional (National Monetary

Council)FGTS Fundo de Garantia de Tempo de Servio (Job Tenure

Guarantee Fund)FGV Fundao Getlio Vargas (Getlio Vargas

Foundation)FINAME Agncia Especial de Financiamento Industrial

(Special Agency for Industrial Financing)Financeiras Sociedades de Crdito e Financiamento (finance

companies)FINSOCIAL Fundo de Investimento Social (Fund for Social

Investment)

GDP Gross domestic productIBGE Instituto Brasileiro de Geografia e Estatstica

(Brazilian Institute of Geography and Statistics)IPEA Instituto de Pesquisas Econmicas Aplicadas

(Institute of Applied Economic Research)LDC(s) Less developed country(ies) LIBOR London interbank offer rateLTN Letras do Tesouro National (National Treasury Bill)ORTN Obrigaao Reajustvel do Tesouro Nacional

(Readjustable Treasury Bond)OTN Obrigaao do Tesouro Nacional (National Treasury

Bond)PAEG Programa de Ao do Governo (Governments Action

Plan)PASEP Program do Patrimnio do Trabalhador do Setor

Pblico (Asset Accumulation Program for PublicServants)

PIS Programa de Integraao Social (Program for SocialIntegration)

SFH Sistema Financeiro de Habitaao (Housing FinanceSystem)

SNCR Sistema Nacional de Crdito Rural (NationalAgriculture Credit System)

SUMOC Superintendncia da Moeda e do Crdito(Superintendence of Money and Credit)

xiv

Chapter 1Introduction

In the preface to this book, we expressed our dissatisfaction with theconventional, prior-saving view on finance and economic development.This book, in effect, should be seen as a long essay in persuasion of theneed to pursue alternatives to such an established view. ThepostKeynesian theory is used as large shoulders in this search for onepossible theoretical alternative.

In contemporary post-Keynesian theory, finance in a monetaryproduction economy is sharply distinguished from savingwhich issaid to derive from, rather than be a pre-condition for, growth.Investment is the motor of accumulation and finance is what permitsinvestment decisions to materialise. The supply of finance is causallydetermined by banks: it is banks, and not savers, who hold a keyposition in the process of growth. Only if they share the optimism ofentrepreneurs in periods of growth or are led, for any other reason, toaccommodate the demand for investment finance, can the monetaryproduction economy grow. This conclusion would appear to leave norole for savings and, hence, for capital markets, but such is far frombeing the case. Saving, which funds (but does not finance) capitalaccumulation, has an important role, as we shall see, in maintaining thefinancial stability of the growing economy.

Keyness monetary production economy may represent a paradigmon which an alternative view on finance and growth can be built.Nevertheless such a view is based on a specific type of financialsystem, with well-developed banking and non-banking financialinstitutions and markets for a diversified range of financial assets. Mostdeveloping countries, in contrast, do not have developed financialmarkets, and growth has to depend heavily on bank credit. Such credit-based financial structures need to develop alternative institutions tofinanceand, especially, fundlong-term investment, to avoid the riskof financial instability and other possible adverse side-effects of

growth. The theoretical modifications to post-Keynesian theory,necessary in applications to developing countries, are also discussed inthis book, especially in the context of the Brazilian case.

As regards its organisation, this book is set out as follows. Chapter 2surveys the economic literature on finance and economic development.This aims at establishing the theoretical foundations of the prior-savingargument and at determining how such a unifying principle has beenused for policy purposes. As illustrations of the use of the prior-savingargument in models of finance and economic development, the two-gapand the financial liberalisation models are appraised. The choice ofthese two models was not random, as they have been influential forpolicy purposes in multilateral agencies as much as in less developedcountry (LDC) governments, as noted earlier in the preface.

Chapter 3 presents the post-Keynesian foundations for the criticalappraisal of the prior-saving argument and for the alternative view onfinance and economic development proposed in this book. This is doneby discussing three common assumptions in most post-Keynesiananalyses of finance and growth. These are: (1) finance, and not saving,is the pre-condition to investment; (2) banks, and not savers, play themost fundamental role in the process of finance; and (3) saving fundsbut does not finance accumulation.

In Chapter 4, using the Minskian financial fragility hypothesis, webuilt on the above-mentioned assumptions in order to complete ourpostKeynesian approach to finance and growth with the followingcomplementary assumptions: (4) growth increases the economysfinancial fragility; (5) this financial fragility can be mitigated by activefunding, i.e. the issue of long-term securities by the investing firms toconsolidate their short-term liabilities; (6) well-structured and well-functioning financial markets may play a fundamental role in afinancially stable process of growth/development, a role which isnevertheless ambiguous because of the inherent volatility of thosemarkets.

The book is mainly concerned with the internal mechanisms tofinance accumulation. Nevertheless, given the importance of theexternal debt problem in many LDCs, the role of foreign debt indevelopment is also addressed. The main conclusion that we want toestablish here, which is soundly based on a post-Keynesian perspective,is that foreign capital inflows as sources of finance are only functionalto the process of development if they finance transfers of real resourcesfrom abroad, which can complement internal accumulation.

2 INVESTMENT FINANCE IN ECONOMIC DEVELOPMENT

Post-Keynesian theory emphasises the role of institutions andconven tions in mitigating the uncertainties of decision making inmarket economies. But most post-Keynesian models implicitly assumea specific institutional backgroundi.e. one with a developed bankingsystem and organised financial markets. In the context of LDCs,therefore, a post-Keynesian alternative to the prior-saving argumentmust attempt to generalise its conclusions to other types of financialstructures.

Hence Chapter 5 concludes the theoretical part of the book byapplying the model of finance and growth (established in the previoustwo chapters) to the context of development. It begins by distinguishingbetween credit-based and capital market-based financial structures inthe context of the post-Keynesian theoretical framework developed inChapters 3 and 4. It also shows that the pace and the path of financialdevelopment is not independent of the historic circumstances in whichdevelopment takes place, and it points out the reasons why mostlatecoming industrialising economies tended to start with bank-dominated financial structures and to develop credit-based systems.Further, this chapter formalises the concept of functionality, which isthe counterpart from our post-Keynesian perspective of the concept ofefficient capital markets. A functional financial system is defined as onewhich, irrespective of its stage of development or institutional fabric,finances accumulation with the least increase of financial fragilitythrough the process of growth.

Once the post-Keynesian view on finance and economic developmentis discussed, Chapter 6 establishes a transition between theory andapplication. This justifies concentrating the case study on the period194783; and presents the method to be used in the analysis and themain hypothesis that we want to analyse in the case study.

In the empirical part of the book (Chapters 79), the Brazilianexperience of financial and economic development between 1947 and1983 is analysed. The period is divided into two distinct phases ofBrazils development. The first one (194766) was characterised by theimport-substitution of durable consumption goods and of a substantialpart of light capital goods (194761), followed by a recession between1962 and 1966. The second phase (196783) marks the resumption ofgrowth (196773) and industrialisation (197480) with the SecondNational Development Plan which promoted the import-substitution ofheavier capital goods and chemicals.

As regards the financial structure, the two phases are also divided bythe 19646 financial reform, which transformed the bank-dominated

INTRODUCTION 3

system into a more complex and segmented structure. The relevanceof the Brazilian experience lies in the fact that the reform, at least asregards the provisions for the development of mechanisms to financelong-term investment, was guided by what was previously called theprior-saving argument. This reform attempted to enhance the countrysinternal saving, and at the same time creating a mechanism to increasethe absorption of external saving. This was done by a mixture ofinstitutional reforms (creation of investment banks and incentives toacquire stocks), indexation of financial assets and other measures whichwere viewed as stimuli to saving. We claim that the misleadingtheoretical foundations of the 19646 reform created a financial systemwhich was even less functional to Brazils economic development thanthe one which existed before the reform. In addition, it is claimed thatmuch of the financial chaos which the country increasingly had to facein the 1980si.e. internal and external debt, severe financial instabilityand the highly speculative character of the financial systemcan bepartly blamed on the 19645 financial reform.

Chapter 10 summarises the findings of this book and then presentsthe conclusions.


Chapter 2Finance and economic development

The dominance of the prior-saving argument

INTRODUCTION

One common assumption of the models concerned with finance anddevelopment is that saving is a pre-condition to investment andeconomic growtha view which was called the prior-saving (PS)argument in the introduction to this book. The PS argument has hadboth theoretical and policy consequences for development economics.From the theoretical perspective, it implies a hierarchy in the dynamicsof a capitalist economy: savers, as suppliers of saving/capital, ultimatelydetermine the pace of accumulation.1

In what concerns policy-making, the need to increase the availabilityof either internal or external saving is normally implicit in modelsattached to the PS argument. Thus policy recommendations havestressed the need to create internal institutional mechanisms to stimulatesaving; to attract foreign saving by opening the internal financial systemto foreign capital inflows; and to eliminate financial repression and tocorrect other constraints to the functioning of the market-clearingmechanisms.

This chapter aims at discussing the origins of the PS argument and itsconsequences to economic development models. The relevance of thisdiscussion will become evident in the next chapters, when Keynessdisputing assumption (that finance and investment precede saving) isused to build an alternative view of finance in economic development.

The chapter is set out as follows. First, we discuss the theoreticalfoundations of the view that identifies finance with saving. Second, wedescribe how macroeconomic/financial theories have convergedtowards the implicit acceptance of this view. Third, the relevance of theabove-cited identification to economic development models (the twogapmodel and the financial liberalisation model are specifically discussed)

is assessedin view of their relevance in policy recommendation fordeveloping countries in the last thirty years.

THE ORIGINS OF THE PRIOR-SAVINGARGUMENT

Even though it is difficult to trace the origins of any established view insocial science, one may at least conclude from reading SchumpetersHistory of Economic Analysis that the identification of finance andsaving has been present in the economic theory every since its foundingpillars were laid.2 For instance, Adam Smith stressed that parsimony,and not industry, is the immediate cause of the increase of capital(op.cit.: 301; as quoted in Schumpeter 1954:193). Indeed, Smithsinfluence on this topic of economic theory, as in many others, wascrucial: as Schumpeter points out, Smiths position on themacroeconomic virtues of thrift marks the victory for more than 150years to come of a pro-saving theory (ibid.).

In modern economic theory, the PS argument is in fact a legacy of theclassical theory, which, as Chick (1983:184) correctly reminds us, hadits beginnings in the setting of an agricultural economy, where thearchetypal form of saving was the seed-corn: production not consumed,a real resource. In such an agricultural economy, incomethe harvestis predetermined; hence the only source of investment is the corn-seed saved, which must logically exist previously to the act ofinvesting.

If the PS argument is a reasonable working assumption in theanalysis of an agricultural, barter economy, it is much less acceptable inan industrial economy, where significant part of the means ofproduction are produced for order, thus after, and pour cause, the act ofinvesting. The question of investment finance is not so much how toallocate a predetermined output between consumption and investment,but to define how much of the current output will be made available toinvestment.

Further, in a monetary economy, the only possible associationbetween the finance and saving is the use of accumulated stocks ofmoney to finance investment, rather than consumption. However, in aneconomy which has moved beyond metallic money to the use of creditand loansand this is the case in England since, at least, the seventhcentury3investment can be financed by new money as much as bythe transfer of existing money savings.


Thus, the view that saving is the motor of accumulationor thecreator of capitalcontrasts with both real and financial evolutionsof capitalism. Paradoxically, this view represents one of theimportant pillars of mainstream economics. The persistence of such aparadigm can only be explained by other non-economic reasons. Forinstance, Chick sees the persistence of the principle in the Anglo-Saxoneconomic in association with the principle of the virtue of the thrift:

[T]hat Victorian principle was urged in that subliminal way thatsocieties find their way to, because its basic aim wasindustrialisation, and the banks, in the 19th century stage ofdevelopment, were not adequate to cope with the needs of therapidly-expanding industry. Most borrowing had to be direct, andthat meant prior saving.

(Chick 1983:191)4

Chicks hypothesis is corroborated by the fact that the fast developmentof the banking system at the end of the nineteenth century did inspire arevision of this view in economic theory. Indeed, the denial of theidentification between finance and saving was a central feature of bothKeyness macroeconomics (e.g. Keynes 1936:178; 1937a:249) andSchumpeters (1934) theory of economic development.

Nevertheless, and despite these authors emphasis that thisidentification was fallacious as the theoretical foundation ofmacroeconomics, the pre-Keynesian view became one of thefundamental aspects of modern macroeconomic theory. How thistheoretical tendency evolved and how it affected development economicmodels is assessed below.

Neutral money, neutral credit and the loanablefunds theory

As argued above, the prior-argument is only conceivable for theanalysis of a barter economy, or if the analyst assumes that the capitalisteconomy behaves fundamentally as a barter economy. A bartereconomy is one where output and income are predetermined and hencethe process of exchangerather than productionis the main concernof economic analysis.

To picture an industrial economy as a barter economy is thus toassume that, given the state of technique, income is predetermined and,for the purpose of analysis, may be assumed to be unchanging. Indeed,

FINANCE AND ECONOMIC DEVELOPMENT 7

it is one of the principal results of neoclassical theory thatfullemployment equilibrium is a natural outcome of a competitiveeconomy. Two assumptions are crucial to such a result: first, thattechniques and prices are flexible and the latter contain all informationrequired for agents to trade; second, that Says lawnamely that outputalways creates equal demandprevails. Price flexibility entails thatmarketclearing equilibrium is incompatible with the existence ofinvoluntary idle resources. Says law guarantees that aggregate demandwill always correspond to aggregate supply: any discrepancy betweenindividual supply and demand will generate the migration of resourcestowards the more profitable sectors.

In such a market-clearing competitive economy, money plays twodistinct roles: a medium of exchange and the unit of account. Oncemoney is a non-interest-bearing asset, holding money as a store of valuecan only be justified in the interval between the receipt of income andthe acquisition of a good. In the long term, money is necessarily neutral.In this case, the quantity theory of money cannot be disputed: changes ofmoney supply cannot affect real variables, but only money variables.

The postulate of long-term neutrality of money is one of thefundamental principles of classical and neoclassical economics. Thispostulate has permitted the mainstream economist to distinguishshortterm monetary phenomena from long-term equilibrium values, sothat all the fundamental theorems can be established in real termsbythe direct consideration of goods, preferences and technical constraints(Carvalho 1992:32).5 In Schumpeters words:

Real analysis proceeds from the principle that all the essentialphenomena of economic life are capable of being described interms of goods and services, of decisions about them, and ofrelations between them. Money enters the picture only in themodest role of a technical device that has been adopted in order tofacilitate transactionsit does not affect the economic process,which behaves in the same way as it would in a barter economy:this is essentially what the concept of Neutral Money implies.

(1954:277)

Even though constructed for the analysis of an economy where metallicmoney prevails, the evolution of bank money does not change thispostulate as long as banks as well as other financial institutions arepictured as mere intermediaries between saving and investment. That is,as long as credit is also neutral, in the sense that it does not interfere


with the real forces behind accumulation (thrift and productivity). Thishas been the role of the loanable funds theory (LFT) in monetaryeconomics (Rogers 1989: ch. 2).

In LFT models, output is neoclassically determined in the labourmarket. Equilibrium is only achieved when all factors are remuneratedaccording to their productivity.6 For the purpose of analysis of thecapital (saving) market, income can be assumed to be predetermined.In other words, there is no fallacy in the passage from the micro economicto the macroeconomic level. Consequently, the supply of real saving/capital (S) is seen as determined according to households inter-temporal preferences (and is inversely related to the interest rate r). Inturn, the demand for capital (investment) (I) is a direct function of thereturn on capital (or the marginal productivity of capital, mpc). Hencethe equilibrium (or natural) rate of interest (rn) is the rate whichequilibrates the quantity of capital saved (saving) and of capital invested(investment), thrift and productivity.

Now, assume that some technical innovation is introduced into ourLFT economy, so that the marginal productivity of investment riseswhich is illustrated in Figure 2.1 as a rightward shift of the demand forloanable funds schedule (I). If the savers intertemporal preferences donot change, this would normally imply an increase in the natural rate ofinterest, causing a decrease in consumption symmetrical to the increaseof investment. The banking system may decide to maintain the marketrate of interest below the natural one, but this can only be achieved inthe LFT if the banks accommodate the increase in the demand forloanable funds by creating deposits ex nihilo (Ms).

For identical reasons, an increase in the supply of bank money mayprovoke a departure from equilibrium by reducing the market rate ofinterest. However, if full employment is assumed and consumers preferences are unchanging, this disequilibrium in the money andsaving-investment markets will set off a cumulative inflationaryprocess.7 This is the logic behind the forced-saving hypothesis, whichhas been, from Wicksell to Robertson, from Pigou to Friedman, a mainpillar of monetary economics.

To sum up the argument thus far, the first pillar of the PS argument isthe view that savings are the main source of supply of real capital.The second theoretical basis of identification of finance with saving isthe application of the perfect market paradigm to monetary economics.This application has been put forth by Lewis (1992:204) as follows:


In an ideal world of complete and perfect capital markets, withfull and symmetric information amongst all market participants,economic decisions do not depend in any way upon the financialstructure. All the potential gains from adding banks are assumedaway because every transactor is completely informed and honestabout the environment, and frictions and indivisibilities do notexist. If banks do operate, they do so as traders or equity-financedmutual funds since households and firms have the informationto arrange their own risk diversification.

The perfect market paradigm establishes that in the long run, ifcompetition prevails in the financial system, the real interest rate willequate saving and investment optimally. For the purposes of thischapter the application of this paradigm to monetary economics has oneimportant outcome: it makes the financial system, including banks, aneutral intermediary between savers and investors. There are, however,two other consequences which are worth remembering.

First, the above-cited application determines that, in equilibrium,there will be no mismatch between investment and savinghencebetween aggregate supply and aggregate demand. This is an importanthypothesis to sustain Says Law.8 Second, the application of the perfect

Figure 2.1 The saving and investment schedules in a simplified loanable fundsmodel


market paradigm is important for policy reasons. If the competitivecapital market becomes the standard of efficiency of financialintermediation, inefficiency is associated with anything outside thatparadigm.9 This permits the analyst to view a disequilibrium in theoptimal allocation of capital as the result of constraints on thefunctioning of the capital markets: lack of perfect competition orasymmetric distribution of information become distortions in relationto the optimal, idealised structure. The role of analysis is thus to spotsuch imperfections and, perhaps, to point ways of re-establishing thesovereignty of the marketclearing forces (e.g. financial liberalisation).

From Hicks and Tobin to the monetarists

The convergence of monetary economics back to the PS argument hasfollowed the track of the neoclassicisation of macroeconomics afterKeynes.10 This theoretical convergence was achieved by the almostuniversal acceptance by monetary theorists of the Hicksian neoclassicalsynthesis (Hicks 1937) and the portfolio balance analysis (e.g. Tobin1958; 1965). Already in Hickss (1937) Mr. Keynes and the classicsKeyness denial of the identification of saving and finance was playeddown, as aggregate saving is determined by the rate of interest (andhence is dependent on individual saving decisions):11

As against the three equations of the classical theory,

Mr. Keynes begins with three equations,

These differ from the classical equations in two ways. On the onehand, the demand for money is conceived as depending upon therate of interest (Liquidity Preference). On the other hand, anypossible influence of the rate of interest on the amount saved outof a given income is neglected. Although it means that the thirdequation becomes the multiplier equation, which performs suchqueer tricks, nevertheless this second amendment is a meresimplification, and ultimately insignificant. It is the liquiditypreference doctrine which is vital.

(Hicks 1937:1323)

Later Hicks will point out that:


Mathematical elegance would suggest that we ought to have I andi in all three equations, if theory is to be really General. Why nothave them like this:

(ibid.: 138)

The second logical step towards the re-establishment of thepreKeynesian view on finance/saving is Tobins (1958) reinterpretationof Keyness liquidity preference theory. Keyness liquidity preferencetheory maintains that the interest rate is mainly a monetary phenomenon,determined by supply and demand for money. Especially, heemphasised the role of uncertainty in provoking changes in the demandfor money as a store of value. In his framework, money was not hoardedfor saving purposes, but as a means of postponing the decision (tospend) when the uncertainty of the future changed substantially. Thisimplies that individual saving decisions have little, if any, importanteffect over the level of interest rates.12

The portfolio-balance analysis reinterprets Keyness liquiditypreference theory within a general equilibrium framework, substitutingKeyness concept of uncertainty (as affecting the allocation of financialwealth and its relation to money interest rate) by the one of risk (seePettenati 1977; and Chick 1983). Risk here is defined as the variance ofthe return of an asset, which is assumed to be known by all agents. Oncethe risk preference is established and agents allocate their savingsaccordingly, the aggregate portfolio of financial assets is determined(Tobin 1958). Variation within the portfolio of assets can only beachieved by changes in interest rates.13

The acceptance of the portfolio-choice analysis by Keynesiansopened one front of the Keynesians/Monetarists debate, focusing on theperiod of expectation adjustment and substitutability between assets.14This debate can be summarised as follows. In equilibrium savers choosetheir portfolios according to their preferences and the relativeremuneration of different assets. However, in the short run their choicecan suffer from monetary illusion and credit can force saving and thusfinance levels of investment which are higher than the equilibrium one:monetary phenomena can affect the real forces of thrift.15

The convergence towards the view that saving is independentlydetermined from investment partly explains why Keynesian and theMonetarist visions only differ in the short-term analysis (see, forinstance, Goodhart 1975:21921). For, in the long run, when all


monetary illusions disappear, the balance between investment andsaving must be re-established. If investment exceeds ex ante saving,either the propensity to save rises or saving will have to be forcedthrough inflation. Thus, if the question of the relation between savingand investment became more subtle in the Keynesians/Monetarists thanin the Keynes/Classics debate, it is because of the common theoreticalbackground of the former.

From rational expectations to asymmetricinformation

The rational expectations revolution represented a frontal attack on boththe Monetarist and the Keynesian positions.16 If agents form rationalexpectations then the possibility of monetary illusion is limited tothe very short term. If a strong form of rational expectations is used thenmoney is neutral both in the short and in the long term: any anticipatedincrease of money supply would just affect prices leaving the level ofinvestment and output intact.

By incorporating rational expectations, neoclassical monetaryeconomics finally abandoned the Wicksellian connection (Leijonhufvud1981:131202) for neo-Walrasian monetary theory. This conversion,however, had a high cost: as Rogers (1989:3) rightly put it, one horn ofa dilemma facing neoclassical monetary theorists is the inessentialityof money in modern neoclassical theory (see also Hahn 1981).

Furthermore, another interrelated horn exists: in neo-Walrasianmonetary economics the distinctions between credit, saving, capital andinvestment, which so much concerned monetary economists sinceWicksell, became irrelevant. This has allowed mainstream economics toaddress saving without even mentioning the mechanisms by which it istransformed into investmenti.e. the financial system. For instance,this is the case of the long survey on saving and development byGersovitz, which is introduced by the following statement:

Saving, a sacrifice of current consumption, provides for theaccumulation of capital which, in turn, produces additionaloutput that can be used for consumption in the future. The processis thus inherently intertemporal. Its presumed operation makes thesaving behavior of citizens and their governments central to thedevelopment of poor countries. Moreover, threats ofexpropriation, repudiation and other hostile acts against foreignsuppliers of capital, and donor resistance to significant increases


in aid, mean that domestic savings is likely to remain thepredominant source of capital accumulation in developingcountries.

(Gersovitz 19889:382; my emphasis)

If the neo-Walrasian economics renders no role for the financial system,the question of finance also becomes useless. The only way thatfinancial analysis becomes relevant is when the strong assumptionsbehind the perfect-market paradigm are criticised. That is, the only wayout from the straight-jacket of the sound microfoundations ofnewWalrasian economics is to point to market failures which inhibit theachievement of the equilibrium results of such economics. This hasbeen the core of the so-called New Keynesian economics.17

As regards the role of financial markets as efficient (neutral)intermediaries between savers and investors, New Keynesianmodels have focused on heroic assumptions about the availability anddistribution of information between borrowers, lenders and financialinstitutions.18 The relaxation of the perfect information hypothesispermitted these models to explain the role for financial intermediaries ina competitive economy: the lack of costless information gives financialintermediaries the role of assessing the credit-worthiness ofborrowers.19

A common theme in this new line of thought in financial economicsis, in a nutshell, that informational asymmetries may introduceinefficiencies in financial markets which may have quantitativelysignificant real effects.20 Furthermore, the growing literature oninformation and incentives information also points to two problems infinancial intermediation which can jeopardise the allocative role playedby intermediation: adverse selection, where trading parties haveasymmetric information prior to contracting, and moral hazard, wherethe asymmetries arise after contracting.21 Asymmetric informationoccurs when lenders have trouble determining whether a borrower is agood risk (i.e. good investment projects with low default risk) or a badrisk (bad investment projects with high default risk). Because of thislack of information, lenders will desire to pay for a security that reflectsthe average quality of firms issuing the securitiesa price which ishigher than the market value for high quality firms and too high for thelow quality onesa classic case of the lemons problem proposed byAkerlof (1970). Hence, only low quality firms will be willing to selltheir securities.


Stiglitz and Weiss (1981) also demonstrated, using a loanable fundsframework, that adverse selection would generate credit rationingbecause low quality firms with riskier projects will be the ones willingto pay the highest interest rates. If lenders cannot identify the riskiestprojects, then the supply of loanable funds will shrink when interestrates increase, exactly the opposite result from that theoreticallyexpected. The danger of moral hazard may prevent lenders fromextending credit, if the interest rate makes it very attractive to do so. Inother words, lending would be at sub-optimal levels.

The two cases just point to the possibility that, due to asymmetricinformation, the financial system may not play its role as broker in thesaving-investment process efficiently. In other words, in these casesloanable funds to investment will be lower than potential and allocationof resources will be distorted. This literature presents an embarrassingchallenge to the view that financial markets are efficient allocators ofcapital. But, in reality, it does not seem to advance much fromthe neoclassical perspective: the role of the financial system is still to bean intermediary between saving and investment;22 what is at stake ishow well this role is performed. This type of argument leads to the viewthat, were it not for the problems generated by imperfect informationdistribution or other market failures, then that role would be fullyrestored and allocative efficiency of capital would prevail.

To sum up, these recent developments explore deviations from theperfect markets paradigm as regards the optimal allocation of the assetsof financial intermediaries. As far as the theory is concerned, theliabilities side is an exogenous variable, determined by the preferencesof consumers/savers. The identification of finance and saving is stillsovereign.

THE IDENTIFICATION OF FINANCE ANDSAVING IN GROWTH MODELS

In growth theory the question of finance has traditionally been seen asone of availability and allocation of saving, normally denoted by themarginal propensity to save s. This is surprising given the Keynesiancredentials of one of the founders of modern growth theory: Harrod.Again, this paradox has to do with the convergence towards the PSargument of the models which followed Harrods.

Harrods (1939) model is concerned with establishing the level ofinvestment throughout time that will maintain a dynamic equilibrium,i.e. that will continuously fulfil expectations of previously made


investments. The fulfilment of such expectations depends on the currentlevel of expected demand in relation to the level of output capacitycreated by previous investment projects. Thus the model can be built onkey variables determining the two outcomes of investment undertakings:namely, income creation (and thus the level of saving) and additionalproductive capacity (and thus the resulting aggregate productionfunction). The result of Harrods model is the well-known warrantedgrowth rate which is defined as follows:

(2.1)

where x is output, I is investment, s is the marginal propensity to saveand C is the capitaloutput ratio.

It is important to stress that in Harrods equation the marginalpropensity to save (s) only expresses the gap between income createdby a certain level of investment and consumption. Hence, Harrodsperspective is very Keynesian: effective demand must be maintained ata level that satisfies producers. The lower the level of saving, the higherthe multiplier and the lower the level of investment required to fill thegap. Or, what is the same, if growth is to be maintained this wouldrequire increasing levels of investment in order to compensate for amarginal propensity to consume which is normally lower than 1.Nothing, in Harrods model, refers to the question of the financing ofgrowth in market economies, and it may be said that financing isimplicitly assumed to be always forthcoming.

Already in Domars version of Harrods dynamic model, the causalitybetween the marginal propensity to save and the rate of growth isblurred. Harrods equation was reinterpreted in order to show that thebalanced growth would require an improbable co-ordination of thesaving ratio, of the capital-output ratio and of the natural rate on a sharpknife edge. If capital is assumed to be scarce, however, saving canconstrain growth.23

As a matter of fact, after Solows (1956) interpretation of theHarrodDomar model, very little remained even of the saving side of thegrowth model.24 Solow reinterpreted Harrods Y as a result of aproduction function Y=f (K, L) homogeneous of the first degree andwith substitution elasticities lower than 1. With that reinterpretation hewas able to prove that, in the long run, the real product growth ratedepended only on the rate of growth of labour supply, and not on therate of savings. Suddenly, even the saving constraint, together with the


question of finance, disappeared from the theoretical concern of growthmodels.25

THE PRIOR-SAVING ARGUMENT INDEVELOPMENT ECONOMICS

The identification of saving with finance in mainstream economics hashad a direct consequence on models concerned with finance anddevelopment. Capital is commonly assumed to be the scarce factor inless developed countries (LDCs), so saving becomes a constraint togrowth (see, for example, Todaro 1981:5863; Simonsen 1991): indevelopment economics, the identification between saving andinvestment becomes the prior-saving argument. This is confirmed by aextensive survey on saving and economic development found in theHandbook of Development Economics, which begins by the statementby Gersovitz already quoted above (p. 13). That statement is anextreme representation of the view that has permeated the discussion ofthe question about finance and development ever since Gurley and Shawattempted (and did not manage) to emphasize the financial aspects ofgrowth. Below we briefly describe this theoretical trajectory.

The Gurley and Shaw heritage

The mainstream literature on finance and economic development owesmuch to the seminal works of Gurley and Shaw (1955; 1960). For theseauthors:

Development is associated with debt issue at some points in theeconomic system and corresponding accretion of financial assetselsewhere. It is accompanied, too, by the institutionalization ofsaving and investment that diversifies the channels for the flow ofloanable funds and multiplies varieties of financial claims.Development also implies, as cause and effect, change in marketprices of financial claims and in other terms of trading in loanablefunds.

(Gurley and Shaw 1955:515; their emphasis)

Gurleys and Shaws debt-intermediation view established a clearinterrelation between financial and economic development. First,economic development would be associated with financial developmentbecause financial intermediation (external indirect finance) provides


potential surplus units with the capacity to spend beyond their earnings.Second, growth would stimulate and be stimulated by theinstitutionalization of saving and investment: income grows, richerwealth-holders will increase their desire to diversify their asset portfolio;if financial innovation is such as to accommodate this diversificationdemand, financial institutions can enhance their lending capacity andthus boost growth; the process becomes then a benign circle.

To sum up, for Gurley and Shaw financial development enhances theintermediation of loanable funds and therefore growth could bestimulated. Nevertheless they also warned of the dangers of financialdevelopment. In their view, the separation between the acts ofinvestment and saving meant that financial development would permithigher levels of indebtedness and growth; but this would be closelyfollowed by the deterioration of debt profiles, which would result inhigher interest rates.26 Hence financial development widens the scopefor regulatory management.

Notwithstanding the innovative aspect of the debt-intermediationview the prior-saving argument was still the basic tenet: by puttingthe banking system with the rest of financial institutions in the passiverole of mere brokers, Gurley and Shaw denied their capacity toinfluence the growth path:

Neither banks nor other intermediaries create loanable funds. Thatis the prerogative of spending units with surpluses on income andproduct account. Both banks and other intermediaries have thecapacity to create special forms of financial assets that surplusunits may accumulate as the reward for restraint on current orcapital account. Banks alone have the capacity to create demanddeposits and currency, to be sure, but only savings and loansassociations can create savings and loans shares: both createcredit, both transmit loanable funds, both enable spending unitsto diversify their portfolios.

(1955:5212)

At the end of the day, it is the forces of productivity and thrift whichmatter in the financing of development. The financial dimension onlymatters if it is underdeveloped.

Gurley and Shaws failure to present a fundamentally differentapproach seemingly had a high cost: some of their ideas seemed to havebeen forgotten by the mainstream economics.27 Indeed developmentfinance has for the last thirty years been mainly treated by development


economics as a problem of availability and allocation of internal andexternal saving. Two models have dominated the literature: the two-gapmodel and the Shaw-McKinnon financial liberalisation model.

The two models have different emphases: the two-gap model isconcerned with the external finance for resources to supportdevelopment and with development planning; whereas the ShawMcKinnon model deals with the increase and mobilisation of internalresources, mainly through the liberalisation and deregulation of internalsaving. Both approaches are based on the PS argument. In thesubsequent subsections we describe these models, establishcomparisons between the two and seek for the common aspects whichultimately link them to the same theoretical roots.

The role of external saving: the two-gap model

The identification of finance with saving is as much rooted inclosedeconomy macroeconomics as it is in open-economymacroeconomics. For instance, the saving-investment approach to thebalance of payments defines external saving as the difference betweeninternal investment and internal saving, which is identical to the resultin the current account of the balance of payments:

(2.2)(2.3)

where Y is income, C is consumption, I is investment, S is internalsaving, Sc is external saving, G is government expenditure and (XM)is the deficit/surplus of the current transactions of the balance ofpayments (see, for example, Dismoor 1990:34n).

The identification of external finance with external saving may seeminconsequential. However, when it comes to policy it shows itsimportance. For instance, if a continuous balance-of-payments deficitproblem is interpreted as scarcity of internal saving, then the solution isto increase internal saving. Finally, if the PS argument is carried outconsistently, interest rates must be raised and internal absorption mustbe reduced in order to re-establish the external sector equilibrium. Thiswas, for instance, the line followed by the International Monetary Fund(IMF) and the World Bank to the debt crisis in the 1980s (seeFitzGerald and Vos 1989; Felix and Caskey 1990).28

An application of the PS argument to the analysis of the role ofexternal saving in the context of an open developing economy is thetwo-gap model. Two-gap models (TGMs) are based on the idea that the


lack of internal saving can posit a constraint on economicdevelopment.29 External saving is thus required in two initial stages ofdevelopment: first, to overcome the difference between plannedinvestment and saving; second, to finance the increasing gap betweenplanned imports and exports. The two gaps are expressed ex post by thenational account identity

(2.4)where the left-hand side of the equation represents the investmentsavinggap and the right-hand side the import-export gap. By definition,equation (2.1) is concerned with ex post concepts, whereas the TGMsare concerned with the relation of the two gaps with economic growthand with the means to overcome them. The model can be summarisedas follows:

The investment-saving gap(2.5)

(2.6)

(2.7)(2.8)

The import-export gap(2.9)(2.10)(2.11)

where r represents the target rate of growth, k is the constant capitaloutput ratio (Y=kK), is the marginal propensity to save, m is themarginal propensity to import and x is the expected rate of growth ofexports (determined by the conditions of world trade).

The condition for balanced growth is (i) that the economy starts offfrom balanced trade, (ii) that r=x and (iii) that kr=m.30 Since there is noautomatic mechanism that will guarantee any of these three conditions,the TGMs conclude that development will be a disequilibrium process.Such a state of disequilibrium cannot be held forever: the continuanceof growth depends on the LDCs capacity to raise internal saving or onthe supply of external saving. Furthermore, since the saving capacity ofLDCs is assumed to be limited because of the low per capita income,foreign aid is desirable in the starting-up of the growth process(Chenery and Strout 1966; 1968).


The short-term disequilibrium between internal saving and investmentwill, according to the above-cited authors, be overcome in the long run:it is assumed that once the developing economy has sufficientlyincreased its per capita income (which allows for a higher propensity tosave) and its export capacity (by improving productivity in exportsectors), it can finally achieve self-sustaining growth and even repay thedebt acquired in complement of financial aid (Chenery and Strout 1968:913).

Many Latin American countries launched growth-cum-debtdevelopment strategies in the 1970s using as theoretical guidance viewswhich were very similar to the two-gap model. This took the form ofproposals to open the countries financially in order to profit from theexcess of liquidity of the international financial market of the 1970s(Felix and Caskey 1990). The main idea behind this view hasneoclassical roots: since the return on capital had to be higher in theunderdeveloped countries than in the developed economies (due to theformers lower capital intensity), international interest rates would belower then the prospective returns of investment made in LDCs.

If the observation made in the paragraph above is accepted, it is easyto understand the development within the mainstream view after the so-called debt crisis of the 1980s. For, after the interest rate shock of 1979and the Mexican de facto moratorium in 1982 led to a fast contractionof credit to the highly indebted LDCs, the growth-cum-debt strategywas completely discredited.

Indeed the model has been abandoned in the 1980s, and substitutedby the financial liberalisation models, which emphasise the need forLDCs to increase internal saving rather than counting on externalsaving.31 The PS argument remained; it was only the historicalcircumstances that changed.

The financial liberalisation models

The financial liberalisation model (FLM), developed after Shaws(1973) and McKinnons (1973) classics, has exerted considerableinfluence on macroeconomic policy in developing countries in the1970s and 1980s, particularly through the recommendations of the IMFand the World Bank.32 The analysis is based on the idea that manydeveloping economies suffer from financial repression, a misguideddevelopment strategy of low interest rate ceilings and selective creditpolicies. It is argued that financial repression inhibits saving bydeliberately maintaining interest rates below their natural level. With


financial repression, the argument goes, even though investmentopportunities abound, growth is kept below its potential (e.g. Fry 1989:19; McKinnon 1973:5961; Shaw 1973:8).33

Perhaps the most interesting aspect about most FLMs is that they arecompletely devoid of institutional content; they completely disregardthe institutional aspects of LDCs financial systems.34 Thus, within itsrationale, solutions to problems related to the financing of developmentneed only consider re-adjusting relative prices. For instance, the liftingof deposit ceilings is pictured as a panacea that would lead to theestablishment of a superior equilibrium position with higher levels ofsavings, investment and growth. It is also assumed that a less regulatedand less repressed financial market would equilibrate saving andinvestment optimally.

The model presented below (based on Kumar 1983) is fairlyrepresentative of FLMs developed after the original works of Shaw(1973) and McKinnon (1973).35 First of all an unlimited supply oflabour is assumed, so equilibrium output is determined not necessarily atfull employment but at the point where the marginal productivity oflabour equals an institutionally determined wage deflated by a priceindex. As usual in this type of model, output is assumed to be at itsequilibrium level, so the goods equation is simply dropped, and thediscussion centres on the money and savinginvestment markets.

Investment is still a function of the marginal productivity of capital,but the saving function is restated. It is claimed that the lack ofdiversified financial markets implies no alternative for savings otherthan in physical capital and bank deposits. Furthermore, followingMcKinnons (1973:5761) hypothesis of money-capitalcomplementarity, investment is assumed to be indivisible and hence alump-sum expenditure.36 So individual deposits are also used for theaccumulation of savings in order to finance a minimum investmentlevel (Imin).

In his model, for instance, Kumar assumes that the net acquisition ofreal cash balances [d(M/P)/dt] by savers is channelled via bank credit toeither private borrowers who wish to accumulate capital [Ib], or togovernment for current consumption [G]. Investment can be eitherselffinanced (Is) or financed by the loanable funds intermediatedthrough the banking system (Ib). Finally, the total supply of loanablefunds available for private investment equals the newly-created moneybalances (d(MIP)/dt) minus the part appropriated by government tofinance consumption (G). In algebraic terms:


and

where Y is current income, s' is private propensity to save out of currentincome,37 m* is the additional money balances created by the bankingsystem and held by savers, M is the nominal money balances held bysavers, P is the price index, G is the increase in real cash balancesappropriated by the government for current consumption purposes, S isaggregate saving38 and IF is the aggregate supply of funds available forinvestment. Thus the following flow of funds equation can bewritten:39

(2.12)Sources of finance are thus basically savings from current income andreal balances accumulated in the form of bank deposits. On the demandside, government and private sectors compete for the existing loanablefunds. Finally, investment is a function of the average return to physicalcapital (rk) and the real deposit rate (r*):

(2.13)

where r is the deposit interest rate and e is the expected rate ofinflation.

In this simplified model, the banking system is simultaneously a brokerof capital and a supplier of the means of payments. Deposits aredemanded for saving (and investment) and transactions, and hence thereal demand for money function L(')assuming it is homogeneous ofdegree ozne in Yis as follows (after Kumar 1983:21):

(2.14)

where y is output and the deposit interest rate appears as a central variable.Once the money supply (Ms) is exogenously given by the monetaryauthorities, equilibrium in the money market is given by

(2.15)Assuming adaptive expectations, a once-and-for-all expansion of moneysupply can only affect money demand during the period of adjustment ofexpectations. As in the loanable funds approach, any increase of moneysupply beyond the desired level of savings (which here can only


assume the form of bank deposits) would only result in inflation in thelong run.

Given the above framework, the policy of financial repression isdescribed as one which seeks to promote investment by maintaining thereal deposit interest rate below its (positive) equilibrium levels. Such apolicy would result in the rationing (non-price allocation) of scarcesavingwhich is seen as a fertile ground for inefficiency and an easysource of windfall profits to the banking system (Fry 1989:18). Whenassociated with government deficits, it would reduce the availability ofscarce resources to the private sectorand hence leave savings to beunproductively allocated by the short-sighted bureaucracy.

Hence, the argument continues, the appropriate financial policy inthe context of LDCs has to focus on incentives to increase the realdemand for money in the form of depositsi.e. to raise real depositrates. This would expand the resources (loanable funds) necessary tofinance current investment, and provide small investors with a way ofaccumulating saving for investment in a less inflationary manner.Furthermore, such a policy of financial liberalisation would unify thecapital market, increase the return to domestic savers and allocate thisenhanced saving to higher-return (more productive?) projects, besideseliminating other forms of fragmentation.40

SUMMING UP

Mainstream development economics has, in the last thirty years or so,moved from the view that external saving was required in the firststages of development (the two-gap model) to one which privilegespolicies towards increasing internal saving in order to achieve sustainedgrowth (the Shaw-McKinnon liberalisation model). Even though thesemodels present different prescriptions to the problem of financingdevelopment, both share the same unifying principle: the PS argument.

The identification of finance with saving is neither new norunquestionable. For instance, already in 1937, Ohlin, in a well-knowndebate with Keynes and others on finance, investment and saving, hadexpressed a position identical to that of supporters of the financialliberalisation literature:

If an authoritarian government fixes a rate of interest which ismuch lower than the rate which would prevail in a free market,then during any period saving and new investment ex-post arenevertheless equal, but the quantity of credit offered is found to


have been smaller than the quantity demanded; some form ofrationing takes place. The credit market reacts in the same wayas the commodity market, when maximum prices are fixed.

(Ohlin 1937b: 424; my emphasis)

This view was insistently criticised by Keynes, among others, as afallacious foundation for analysing the financial dimension of a marketeconomy. For Keynes, individual saving was not causally an importantvariable in the determination of the supply of investment finance.Banks, and not savers, hold the key position in the transition fromlower to a higher scale of activity[and] the investment market canbecome congested through shortage of cash[but] neverthroughshortage of saving (1979:222). Banks are more than mere socialaccountants or screening devices, their expectations and their animalspirits are at par with other entrepreneurial expectations in thedetermination of the level of investment, output and employment.

In other words, Keyness claim that finance and investment precedesaving, reversing the neoclassical causality, implies a significant changein the hierarchy in the determinants of the dynamics of capitalisteconomies. Rather than the individual saving decisions, it is the animalspirits of entrepreneurs and the readiness of bankers to financeinvestment which make the (capitalist) world go round. Keynesscritique is thus based on an alternative view of the laws of motion of amarket economy (the monetary-production economy paradigm) and thefinance-investment-saving-funding circuit. In the next three chapters,we build on this critique to present a post-Keynesian view of financeand development.


Chapter 3Departing from the prior-saving argument

Finance in a monetary production economy

INTRODUCTION

In the previous chapter it was established that the unifying principle ofmost models of finance and development is that saving is a pre-condition for accumulation. It was also indicated that Keynessdisputing view (i.e. that finance and investment precede saving) wouldbe used in this book as the starting point of an alternative approach.

In this chapter the foundations of Keyness and post-Keynesiansviews on finance are discussed. Initially it sets out the principles behindKeyness paradigm of the monetary-production economy and showswhy investment is the causa causans of output and employment in thisparadigm. Further it argues the case that Keyness view on finance isdeeply rooted in his perception of the stage of development of thebanking system. We then build on the Keyness paradigm of amonetaryproduction economy to stress the fact that banks (and notsavers) are the suppliers of finance and a simplified model of thebanking firm under uncertainty is presented. Finally we address theconsequences of the post-Keynesian view on finance for open-economyanalysis.

THE MONETARY PRODUCTION ECONOMY

Neoclassical economics presumes that, in equilibrium, a market economybehaves as if it were what Keynes called a real-wage or co-operativeeconomy. As the name suggests, in such an economy production isorganised co-operatively and the output distributed in kind: the factorsof production are rewarded by dividing up in agreed proportions theactual output of their co-operative efforts (Keynes 1979: 78). In other

words, in equilibrium the real income of each factor corresponds to itsproductivity.

If money exists in such co-operative economies, it acts as a meremedium of exchange: money can only be demanded because of thetimelags which exist in the process of trading goods. A co-operative orrealwage economy behaves like a barter economy. As stated earlier(Chapter 1), the main economic problem of such an economy is theallocation of the predetermined output. An efficient allocation ofresources will be that one which, given the consumers preference,maximises the present and future welfare of society.

In contrast, Keyness theory is based on what he called themonetaryproduction economy or entrepreneur economy. In a monetary-production economy, the means of production are privately owned andproduction takes place through the entrepreneurs hiring of labour.Because in such an economy, for the reasons that we shall discussbelow, full employment is not guaranteed, and because involuntaryunemployment means wastage and human suffering, the main problemof Keyness economics is to promote full employment. The threekeywords to Keyness economics are time, uncertainty, money andirreversible time. How these words are articulated into a coherentalternative to mainstream economics is discussed below.

Time, uncertainty and non-neutral money

One of the prominent aspects of Keyness and the post-Keynesianmodelling of the economy is the use of historical rather than logicaltime, or, what in practice means the same, of historical analysis ratherthan equilibrium analysis.1 In a nutshell the difference of approacheshave been summarised as follows:

Historical analysis is defined as the study of a series of calendarperiods in which foreseen events play an important part inchanging the agents plans, and thus the path of the system overtime. Equilibrium analysis is characterized by the comparisonbetween equilibrium positions.

(Amadeo 1991:6n)

For Keynes and for the post-Keynesians, historical analysis is theappropriate way to analyse decision making in what Davidson hascalled a non-ergodic world. The principle of ergodicity, which is thebasis of the concept of probabilistic risk, assumes that economic

INVESTMENT FINANCE IN ECONOMIC DEVELOPMENT 27

processes are basically stationary, so in the long run agents can learnhow they operate and, by adapting their behaviour to the environment,long-period equilibrium positions can be attained (Carvalho 1992:42).In contrast, in a non-ergodic environment there is no basis to formcalculable probabilities so that future probabilities are numericallyindeterminate and non-comparable.2

Unlike an ergodic world, in the uncertain, non-ergodic world moneyis more than a convenient means to trade goods: it is also a convenientmeans to delay decisions.3 This is what makes money an asset, despitethe fact that it bears no interest. In Keyness words:

partly on reasonable and partly on instinctive grounds, our desireto hold money as a store of wealth is a barometer of the degree ofour distrust of our own calculations and conventions concerningthe future. Even though this feeling about money is itselfconventional or instinctive, it operates, so to speak, at a deeperlevel of our motivation. It takes charge at the moments when thehigher, more precautious conventions have weakened. Thepossession of actual money lulls our disquietude; and the premiumwhich we require to make us part with money is the measure ofthe degree of our disquietude.

(Keynes 1937a:116, my emphasis)

The fact that money can be held as a store of wealth is the basis ofKeyness principle of the non-neutrality of money. In other words,because money is the liquid asset par excellence, it is a safe haven intimes of uncertainty: a way of postponing expenditure when the futurebecomes cloudier. Its demand, that in more certain times ispredominantly determined by the volume of transactions, can changeabruptly in such times. This is one side of Keyness postulate of the non-neutrality of money. The other is the supply side.

For Keynes, the supply side of money can be defined by two furthercharacteristics of money as an asset: money has zero elasticities both ofproduction and of substitution. This means that, on the one hand,money cannot be readily produced;labour cannot be turned on at willby entrepreneurs to produce money in increasing quantities as its pricerises in terms of the wage-unit (1936:230). On the other hand, becausethe utility of money is solely derived from its exchange value, as theexchange value of money rises there is no tendency to substitute someother factor for it.

28 DEPARTING FROM THE PRIOR-SAVING ARGUMENT

From the supply- and demand-side characteristics of money, asdescribed by Keynes, emerges his and the post-Keynesians postulate ofthe non-neutrality of money. In Davidsons (1988:330) words:

Moneys non-neutrality is associated with the fact thatindividuals recognise that, in an uncertain [non-ergodic] (asopposed to statistically predicable, but risky [ergodic]) world, thepossession of money provides both the flexibility which permitsone to take advantage of previously unforeseeable opportunitiesand the selfinsurance against unwelcome events.

That is, because the demand for money as a store of value can varyabruptly as uncertain expectations change, and the supply does notrespond endogenously, money is non-neutral. This non-neutrality iswhat makes it inconceivable to analyse the laws of motions of a marketeconomy by establishing a dichotomy between real and moneyvariables. This is the essence behind the proposition of a differentparadigm by both Keynes and the post-Keynesian: the monetaryproduction economy. This is our next, interrelated topic.

Production and non-neutral money

Because output is not directly distributed to workers and because anentrepreneur economy is decentralised, contracts (betweenentrepreneurs, workers and suppliers) are essential components of thiseconomy. These contracts must be denominated in an accepted mediumof exchange. Finally, because money is at the same time a widelyaccepted medium of exchange and a store of value, it has the power todischarge contracts. Money allows the entrepreneur to have access tothe physical resources and labour required for production4 thuswhoever has money or the capacity to create money (for example, theState and banks) can influence the allocation of resources.

In entrepreneur economies, production is a time-consuming activitywhich requires that entrepreneurs commit (their own or borrowed)resources before the return on the output is knownthe entrepreneureconomy is essentially a forward-looking system. The law of productionof an entrepreneur economy is thus defined as follows:

[A] process of production will not be started up, unless the moneyproceeds expected from the sale of the output are at least equal to


the money costs which could be avoided by not starting up theprocess.

(Keynes 1979:78)

This law is what guides Keyness principle of effective demand. Thisrepresents, according to Keynes himself (1936:25), the substance of hisgeneral theory of employment. Although detailed examination of thisprinciple is beyond the scope of the present book, some analysisis nevertheless in order.5 First, one must consider that Keyness firmsare profit-seeking entities, deciding in a non-ergodic environment. Insuch an environment, expectations are an inseparable part of thedecision making of these entities:

Expectation is necessary for two reasons: first, because time is insome sense irreversible; second, because there are costs associatedwith the use, and the transfers between uses, of resources. If itwere always possible for us to retrace our steps and to putourselves, without cost, into the most advantageous position ateach moment, the need for expectation beyond the mostimmediate period would be removed and, thereby, uncertaintywould be rendered impotent.

(Dixon 1986:586)

Unlike a barter economy, in a monetary economy a change ofexpectations can more easily affect production and employment becausemoney can be used as the instrument to postpone decisions. Indeed, ifthe firm sees prospects of decreasing demand for its products, it is onlyreasonable that it will increase its demand for money in order to face itsfuture (already contracted) financial commitments. By withdrawingmoney (or by not borrowing money from banks), firms obstruct theprocess of income creationand thus, of aggregate-demand creation.For the firms as a whole, pessimistic expectations are a self-fulfillingprophecy.

Thus, in a nutshell, the principle of effective demand is a direct resultfrom Keyness depiction of market economies as monetary-productioneconomies,6 where money is non-neutral and expected money profitsare the guide for entrepreneurial decision making. This principlemaintains that output and employment will be determined at the pointwhere the expected aggregate demand (De) curve intersects theaggregate supply curve. The De curve refers to the aggregate ofestimates of anticipated sales (Chick 1983:65). The distinction between


actual and estimated demand is an essential feature of Keynessanalysis:

The expected results are not on a par with the realised results in atheory of employment. The realised results are only relevant in sofar as they influence the ensuing expectations in the nextproduction period. This period is covered by short-periodexpectations.

(Keynes 1937a:119)

The aggregate supply curve Z represents the aggregate cost of producingan output Q, which is obtained by multiplying the marginal cost MC bythe level of output. Since the cost is assumed to be composed solely ofwages, then

so that

(3.1)

where A stand for the average physical productivity of labour and Q 'for the marginal productivity. Assuming diminishing returns, Z willincrease with N, but at a faster pace (Figure 3.1).7

Since this is a system of simultaneous equations, the figure does notimply causality. However, for the sake of analysis we can choose one ofthe three quadrants to discuss the rationale behind the diagram.

Thus, starting from the third quadrant, the marginal cost curve isdrawn by assuming that the costs of production are known and returnsare diminishing. The profit-maximising output is found at the point wherethe price entrepreneurs expect to receive from their products (pe) equalsthe marginal cost of producing them.

Once the level of output is determined, the amount of employmentwill be determined by a function (f) associating the level of output andemployment (second quadrant). The marginal cost and the level ofoutput determine the Z curve, whereas the level of aggregate demand (D)will be partly determined by the consumption out of incomecorresponding to employment offered by entrepreneurs (N) (firstquadrant). The point of effective demand will be that where D and Zintersect.


Keynes uses the p rinciple of effective demand to deny the classicalassumption that a market economy is a self-correcting systemgravitating around full employment. He explicitly criticised the viewthat wage flexibility would automatically bring the economy out of arecession. On the contrary, Keynes claimed, a decline of wages wouldaffect the aggregate supply curve as much as the aggregate demandcurve, and therefore the equilibrium result was neither necessarily fullemployment nor stable. For Keynes, the cyclical pattern of a capitalisteconomy had more to do with the volatility of investment than therigidities of the labour (or indeed any other) market. This is the topicof our next section.

THE DETERMINANTS OF AGGREGATEDEMAND

Contrary to what the quotation above (p. 30) seems to imply, Keynesclearly chose not to emphasise the distinction between expected and

Figure 3.1


actual aggregate demand in his General Theory (1937c). Instead, heproceeded to discuss, in Books III and IV of his opus magnum, thedeterminants of consumption and investment planned by consumers andentrepreneurs rather than as estimated by producers of consumer andcapital goods (Chick 1983:64). This tactic, however, seems consistentwith Keyness emphasis on investmentand, therefore, on the role of(inherently volatile) long-term expectationsas the causa causans in thedetermination of output and employment. That is, by assuming thatshort-term expectations are always fulfilled, the production of and thedemand for consumption goods can be assumed stable whereas thedemand for investment goods becomes the main determinant of thechanges in total employment and income.8 In more conventional(Keynesian) words, this involves redefining the aggregate demand curve(De), for a closed economy with a small government, as follows:

(3.2)where C is consumption, N is the amount of employment, w is theaverage wage, I is investment, r is the rate of interest and E is the stateof long-term expectations.

Having said this, it becomes clear why, after a lengthy exposition ondefinitions and ideas (1936:3585), Keynes spends forty-two pagesexplaining the determinants of consumption and dedicates another 109to the determinants of investment. The reason for this imbalance is wellknown: Keynes considered investment, of all the components ofaggregate demand, to be the most volatile; whereas consumption can beassumed a stable function of income without losing much theoretically.In turn, investment is assumed to be the independent variable in (3.2)and thus causa causans in the determination of employment and output.The reasons why investment is potentially more volatile thanconsumption are discussed below.

Long-term expectations

In the chapter of the General Theory dedicated to long-term expectations,Keynes attributes two sorts of uncertainties which make investmentrelatively more volatile than consumption. First, he emphasises thetenuity of the basis of knowledge for forecasting the yield of an assetthat makes many production periods become productive.9 The secondsource of instability is, according to Keynes, the close relation betweenstock prices and the evaluation of investment opportunities by the


entrepreneurs. While the first of Keyness assumptions is relativelystraightforward, the second needs elaboration.

The financial cost of reversing an investment decision in the earlystages of capitalism was normally very high, and sometimes such areversion might not be possible. However, the evolution of organisedinvestment markets changed the necessary long-term commitment ofinvestment (Keynes 1936:1501). Because firms could be boughtthrough mergers and take-overs, the volume of new investment becamehighly