Money and credit in capitalist economies: the endogenous money approach 1852783567

Citation preview

Sts

SRN

RALSSSPaCAT aye

esi a0 :

oh

s,

sayy

pirat X33

z

3

ARS.

ta)

‘

APA OLoEaTOrAr

rereSy

;

BOAtAN

id wih Je ESHA eS , S HS

; Sa

H‘

SNES

} SAMSON

SD

4

"

eS

4

‘

5 :

BERR N

ceS Soe

, ASSAM SAE "

Se:

banat

PAAR

(

:

3 Re

yiRASS ‘ 4

5 } . wie s

) ‘

ANh

oer

’

RAIN

;

:

sh

‘

y

i

siatentnnitse

2

d

Wwe

A

\ y

Ma

eyed \.

:“d y

i

Rye \

«

i

:

i

:

x

Sy yi

i

¥

5 h

é

‘

? ‘ way) ¢

aaet f

Y

y

; Z Pity y

ae

‘

i seh

,

,

t

" >

By ,

hatid

seer ielesc bee?

:

y

, :

d

ons 1

2

4

fhe

i,

Weg

sen

¢ fe

i

indy

i

aifote 19

leneETRY eit PREM WEE

th

ds

5ie

saybaese ¥Wes

Coes

‘

i

i, ie rn aypareyagey

;

a sr!

athoe

mY Lonny

oi: @ By eyegege

Ape agate, 65 tO05)

sree

ARON:

Ne RTI

“hy

kts

AT

Ab Peet

sh

A Sat aud) ata? fig

as Ata

sha

AAR

CL Rete derek hcke

es

ct es ro

AROS

POP

Se, a=

Deis

>— CLIFTON

M. MILLER

LIBRARY

Washington College Chestertown, MD 21620-1197

-—

a

—

—

=

MONEY

AND CREDIT IN CAPITALIST ECONOMIES

Digitized by the Internet Archive in 2022 with funding from Kahle/Austin Foundation

https://archive.org/details/moneycreditincap0000wray

MONEY AND CREDIT IN CAPITALIST ECONOMIES THE ENDOGENOUS

L. Randall Wray

University of Denver

EDWARD ELGAR

MONEY APPROACH

©L. Randall Wray 1990 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited Gower House Croft Road Aldershot Hants GU11

England

3HR

Edward Elgar Publishing Company Old Post Road Brookfield Vermont 05036 USA

, isNd G -

Nn} AAS

A7 * ie

W073

wer

{

7G

fa

.

Ms

me

/

“a

Pi

British Library Cataloguing in Publication Data

Wray, L. Randall 1953Money and credit in capitalist economies: the endogenous money approach. 1. Capitalist countries. Economic policies I. Title 330.9 Library of Congress Cataloging-in-Publication Data Wray, L. Randall, 1953-

Money and credit in capitalist economies: the endogenous money approach / by L. Randall Wray. po cm: Includes bibliographical references (p. Includes index. 1. Money supply. 2. Monetary policy. 3. Credit. I. Title HG226.3.W73 1990 332.4’14--dce20 90-13856 CIP ISBN 1 85278 356 7 Printed in Great Britain by Billing & Sons Ltd, Worcester

CONTENTS

Preface and Acknowledgements

Introduction 1

The Endogenous Approach to Money

Introduction Origins of money The Post Keynesian conception of money Which assets constitute money? Liquidity, money, and spending 2

Money and Institutional Evolution

Introduction Definitions Endogenous money without banks: the case of Islamic society Premodern financial institutions and the rise of banking in Western Europe The role of the state in the development of a monetary system The development of modern banking: the case of England Money and the logic of capitalism The third world and endogenous money Conclusions 3

Endogenous Money versus Exogenous Money

Introduction The endogenous money approach Exogenous control of the money supply Statistical exogeneity Endogenous money and central bank behavior Conclusion 4

History of the Endogenous Money Approach Brief overview of the history of the approach The Currency School and the Banking School Vv

100

Marx and endogenous money Keynes and endogenous money Schumpeter on credit and innovation Revival of the Endogenous Money Approach Introduction

N. Kaldor and the Radcliffe Committee Hyman Minsky and the financial instability hypothesis James Tobin and the new view John G. Gurley and E.S. Shaw: intermediation and diversification Basil Moore: the horzontalist approach Marc Lavoie: the flow of credit

110 116 123 131 ihe

135 138 142 147 150

Endogenous Money and Interest Rates

Introduction Liquidity preference Keynesian thought Flows, stocks, money preference Liquidity preference Endogenous money, markup approach Conclusions

jee in Keynesian and Post 155 demand, and liquidity and IS-LM models expectations theory, and the

162 170

diel 187

Institutional Analysis Introduction Reserve requirements Fed funds market Repurchase agreements Certificates of deposit Borrowing from foreign sources Discount window and open market operations Summary of sources of funds Analysis of uses of funds Reserves Government bonds Loans Off-balance sheet commitments Conclusions Appendix v1

193 195 198 200 201 202 203 207 208 209 210 210 213 218 Sap

8

Review of Evidence Supporting the Endogenous Money Approach

Introduction Institutional innovation Federal reserve bank behavior Formal empirical tests Money and the business cycle Conclusion 9

242 242 247 aye) Zoo 263

The Balance Sheet Approach to Money

Introduction A simple model Reserve requirements Simple numerical example A model with production and consumption Extensions of this model Expectations, uncertainty, and investment finance Conclusions 10 Conclusions and Policy Implications Summary of the endogenous money approach

Policy implications

267 267 Ziel ZA 1G) 278 280 283 287 291

Bibliography

303

Index

319

PREFACE AND ACKNOWLEDGEMENTS

This book grew out of work I did for my dissertation at Washington University, St. Louis, under the direction of Hyman » Minsky, Steve Fazzari, and David Felix. However, they should not

be held responsible for errors in the present work, as I have extensively altered each chapter of the dissertation and added several new ones. There is an extensive list of individuals I must thank for helping to make this book possible. First, I must thank John Henry, who convinced me to study economics and whose guidance over the past decade has been essential in shaping my ideas. Hyman Minsky has taught me just about everything I know about banking and financial instability. Steve Fazzari has read almost everything I’ve written in the past five years and has continually forced me to throw out sloppy ideas and work. I must thank the Fulbright Commission for a fellowship which allowed me to spend a year in Bologna, Italy doing research under Jan Kregel. Thanks are due to Jan for introducing me to much of the material which was used in my analysis, for graciously sharing his office with me, and for initially bringing my manuscript to the attention of Edward Elgar. A number of individuals have read various parts of the manuscript and provided valuable comments. Among these, I must thank Shiela Dow, Paul Davidson,

Perry Mehrling, Tracy Mott,

Basil Moore, and Rod St. Hill. The analysis in various sections of the manuscript benefited from conversations with Tom Asimakopulos,

Mike Carhill, Gladys Foster, David Levine, Colin

Rogers, and Otto Steiger. I must also thank my colleagues at the University of Denver, Peter Ho, Robert Urquhart, and Spencer Wellhofer, who have helped to make the academic environment

pleasant and stimulating. Chris Sundberg provided essential assistance in obtaining permission to quote from the various

published works I have used, and helped in preparation of the bibliography. Aneita Gullet and Frank Suyat also provided valuable help in preparation of the manuscript. Carol Taylor and Carolyn Bolden provided essential help with the word processing. Finally, I must thank Edward Elgar and Julie Leppard for all their editorial guidance.

In all cases, emphasized material in quotes was emphasized in the original. Material in quotes which is enclosed by square brackets [] has been added. The following publishers have granted special permission to quote from the indicated material. The University of Chicago Press, Chicago (60637), and The University of Chicago Press, Ltd., London, have given permission to quote from: Friedman, Milton, Capitalism and Freedom, copyright 1982, The

University of Chicago; and Fischer, Stanley, ‘Long Term Contracts, Rational Expectations, and the Optimal Money Supply Rule’, Journal of Political Economy, 85(1), copyright 1977, The University of Chicago. All rights reserved.

Lawrence Wishart provided permission to reprint from: Marx, Karl, Capital, Volume III, Lawrence and Wishart, London,

copyright 1972; and Marx, Karl, Theories of Surplus Value, Lawrence and Wishart, London, copyright 1972. Material from The Theory of Economic Development: An Inquiry into

Profits, Capital, Credit, Interest, and the Business Cycle, by J.A. Schumpeter, Cambridge, Mass., Harvard University Press, copyright 1934 by the President and Fellows of Harvard College, and copyright 1962 by Redvers Opie, was reprinted by permission of the publishers.

Yale University Press provided permission to quote from the following: Minsky, Hyman, Stabilizing an Unstable Economy, copyright 1986

by Yale University Press. All rights reserved. This book may not be reproduced in whole or in part, in any form (beyond that copying permitted by sections 107 and 108 of the U.S. Copyright Law and except by reviewers for the public press), without written permission from the publishers; and from the following articles from The Dawn of Modern Banking, Center for Medieval and Renaissance Studies, University of California, Los Angeles, New

Haven and London, copyright 1979 by Yale University press: Blomquist, Thomas, ‘The Dawn of Banking in an _ Italian Commune’; Lopez, Robert S., ‘The Dawn of Medieval Banking’;

Munro, John H., ‘Bullionism and the Bill of Exchange in England, 1272-1663’; and Udovitch, Abraham, L., ‘Bankers without Banks: Commerce, Banking, and Society in the Islamic World of the

Middle Ages’. All rights reserved. This book may not be reproduced in whole or in part, in any form (beyond that copying permitted by sections 107 and 108 of the U.S. Copyright Law and except by reviewers for the public press), without written permission from the publishers.

Wiley and Sons provided permission to quote from: Minsky, Hyman, ‘Central Banking and Money Market Changes’, Quarterly Journal of Economics, May 1957; and Poole, William,

Optimal Choices of Monetary Policy Instruments in a Simple Stochastic Macro Model’, Quarterly Journal of Economics, May 1970.

Regnery Gatway, Inc., provided special permission to reprint from: Polanyi, Karl, ‘Aristotle Discovers the Economy’, in Polanyi, Karl,

Conrad M. Arensberg, and Harry W. Pearson (editors), Trade and Market in the Early Empires, copyright 1971 by Regnery Gateway. All rights reserved.

xi

INTRODUCTION

In orthodox works (whether of the Monetarist or Keynesian orthodoxy), the money supply is treated as an exogenous variable whose quantity is increased or decreased at the whim of the central bank. Banks passively adjust their balance sheets as the central bank increases or contracts reserves. Indeed, money is treated as if it were dropped from helicopters. In these models, money doesn’t really matter, at least in the long run, since nominal values can matter only to those who suffer from money illusion. In these pages I will introduce the reader to an alternative, the

endogenous money approach, in which money enters the economy through the normal economic processes of a capitalist economy. Because the creation of money is tied to profit seeking behavior, ' to deficit spending, and to contracted payment commitments, it can never be neutral even in the long run. This non-neutrality is not based on money illusion, but on the characteristics of a capitalist economy which is fundamentally a monetary economy. An endogenous money approach will be shown to be applicable to a private property economy in which production is for the market and in which historical time ensures that uncertainty exists. I trace the history of the idea that money is created during normal economic processes from the Banking School through the modern revival led by Post Keynesians. In retrospect, it is difficult to understand how the orthodox approach became the dominant foundation of monetary theory. An examination of the writings of monetary theorists of 1800 generally shows that the level of analysis and understanding of monetary phenomena was superior in many respects to that of the Monetarist-Keynesian debates of today.’ When the writings of Thomas Tooke, John Stuart Mill, Karl Marx, J.M. Keynes, and J.G. Gurley and E.S. Shaw are examined, one finds most of the elements of an endogenous money approach at an advanced stage of development. An examination of the reasons why these insights were dropped from mainstream analysis would make an interesting project, but is beyond the scope of this

book.

Perhaps the endogenous money approach has been neglected because it has never been fully formulated. This approach is xiii

certainly much more complicated than the orthodox approach. One finds strands of the endogenous approach running throughout various theoretical analyses over the past two centuries, but no one has yet completely laid out the approach.’ That is what I attempt to do here. In the first chapter, a definition of money is developed, which is consistent with the endogenous approach. I start with an historical analysis of the development of money, using anthropological evidence presented by Polanyi, Malinowski, Keynes, and Heinsohn

and Steiger. Their analyses are expanded to show how money naturally develops in a capitalist economy in which property is privately owned and in which production for the market is made possible by the existence of propertyless workers. This anthropological evidence is then contrasted with the orthodox story that money was created to facilitate exchange, and show how these

two contrasting versions of the origin of money lead to different definitions of money. The different definitions of money, in turn, lead to different foci for analysis: orthodoxy focuses on exchange or portfolios, while the endogenous approach focuses on asset production and ownership. These different foci, in turn, lead to different conclusions regarding the role of money in the economy and regarding appropriate policies. Chapter 2 is an analysis of the development of financial institutions from pre-modern times to the development of a financial system based on a central bank. The money supply (properly defined) has always been endogenously determined - that is, has never been constrained by the quantity of commodity money. Modern financial relations and institutions evolved with continual

interference

from the state, and will show

how

state

actions helped to shape the financial system. Modern financial arrangements are a joint product of profit seeking behavior of private institutions and of attempts by the state to obtain sufficient purchasing power to meet its needs. The logic of the capitalist system, which requires accumulation, relies on an endogenous money supply and that reversion to commodity money occurs only during crisis. Indeed, the money supply could be made exogenous only by suppressing capitalist relations. Finally, the financial system of the less developed country today is examined to compare and contrast this situation with that of premodern financial arrangements and with modern financial systems in developed

countries.

In Chapter 3, I distinguish between exogeneity in the control XIV

sense and exogeneity in the statistical sense, and argue that even if money were exogenous in the control sense, it is unlikely to be exogenous in the statistical sense. This implies that virtually all empirical work testing money demand equations or relating money to spending suffers from simultaneity bias. I go on to argue that the money supply is not exogenous in the control sense, either. Historical experience broadly confirms the hypothesis that central banks are accommodative. Finally, I conclude with succint definitions of exogenous and endogenous money and with a comparison of alternative endogenous money approaches. A pre-publication reader has expressed concern that a rigorous distinction between exogenous money and endogenous money is not provided until Chapter 3. After giving this some thought, I decided to postpone the discussion until after I had given sufficient treatment to the origins of money and to the evolution of the financial system. The obvious disadvantage of this approach is that readers may get lost in the historical detail presented in Chapters 1 and 2. The advantage, however, is that the historical background

illuminates the primary argument presented in Chapter 3 that the money supply is not and cannot be made to be exogenous in either the control sense or the statistical sense. Impatient readers may wish merely to skim the historical detail and then skip to the chase. Minsky always warned me to avoid writing the story of Genesis, but the temptation proved too strong. I hope that the historical detail will help to convert some readers to the endogenous approach without boring too many others. My own, specific, contributions in chapters 1 through 3 include combining an examination of the origins of money and financial institutions with the endogenous approach. This has not, as far as I know, been done explicitly previously. I believe that I have defined money in a way which is consistent with Post Keynesian thought, but have more clearly distinguished between money and liquidity and between money demand and liquidity preference. These distinctions prove useful in showing how money, but not necessarily liquidity, can be created on demand. I have also shown why a fairly stable relation between final expenditures and narrowly defined money exists, and have argued that this does not validate the Monetarist conclusion that money causes income. Indeed, I system requires an that the capitalist economic argue accommodative credit money supply to allow economic growth and

capital accumulation. monetary

policy which

Finally, I have presented an analysis of shows why the central bank ultimately

cannot constrain reserves. While this idea can also be found in Moore (1984, 1985, 1986), I believe I have provided a more rigorous analysis by combining Moore’s insights with those of

Minsky (1957, 1975, 1986b), with those of Fazzari and Minsky (1984), and with the business cycle theories of Niggle (1986) and Wolfson (1986). Chapter 4 traces the history of the endogenous money approach, beginning with the Classical economists, and particularly with the Banking School. Tooke and Mill examined the relation between money and banking and were able to formulate a rudimentary endogenous money approach. This led to the Currency-Banking School controversy, which has been examined by Panico (1988) and Cramp (1962, 1970). I examine the works of Tooke in more detail to show how his ideas relate to the endogenous money theme. I next examine Marx’s analysis of money. My analysis synthesizes insights contained in Lianos (no date) and Panico (1980, 1982) with the endogenous money approach. While their analyses were not directly related to a study of the endogenous approach, I show that Marx did recognize the endogeneity of money and incorporated this within some of his analysis. Finally, I examine the endogenous money thread which ran through the works of Keynes and Schumpeter. I argue that Keynes’s definition of money is consistent with my examination of the origin of money and with my definition of money. His analysis of liquidity preference is also consistent with my distinction between money demand and liquidity preference. Finally, his definition of a ‘monetary economy’ is consistent with my analysis of an economy with an endogenously determined money supply.’ I also show that Schumpeter’s analysis of the relations among credit, innovation, and development is consistent with the endogenous money approach. Credit allows spending to increase, while innovation funded through creation of credit allows economic development to proceed. Innovation arises as financial and nonfinancial capitalists pursue profits by finding less costly ways of doing things. Thus, credit creation and innovation by financial institutions are explained as profit seeking behavior. Schumpeter’s

insights are consistent with my argument that the logic of the capitalist system rests upon credit money and provide a powerful argument against the orthodox approach.

The revival of the endogenous money approach is traced in

Chapter 5, beginning with the Radcliffe Committee and continuing through the early insights of Kaldor, Minsky, Tobin, and Gurley

and Shaw, and finishing with the arguments of Post Keynesians such as Moore and Lavoie. My contribution in this chapter is primarily to show how the work of Minsky, Tobin, and Gurley and Shaw strengthen the Post Keynesian theory of an endogenous money supply. Recent Post Keynesian analyses (such as Jarsulic 1985, Moore 1986, and Lavoie 1984) have traced the origins of the

endogenous approach from the Radcliffe Committee, Kaldor, and later Minsky. However, the contributions of early Minsky, Gurley and Shaw, and Tobin are generally neglected.

In Chapter 6, a theory of the determination of interest rates is

presented, which is consistent with both Keynesian liquidity preference theory and with the endogenous money approach. I criticize the textbook version of liquidity preference and show that the IS-LM analysis is inconsistent with both Keynes’s liquidity preference theory and with the endogenous money approach. Finally, I show that the Post Keynesian interest rate theory advocated by Kregel, Mott, and Davidson is consistent with an endogenous money approach. I provide an alternative to the markup approach to interest rates (which has been adopted by Moore 1988a, Rousseas 1985b, and Kaldor 1980) which I believe is consistent with the liquidity preference theory of interest rates and with the endogenous money approach. This approach is also more consistent with the observed behavior of banks and other financial institutions than are the orthodox and markup approaches. A detailed analysis of financial institutions is presented in Chapter 7. I argue that one of the reasons orthodoxy reaches incorrect conclusions regarding money is due to the lack of institutional detail in its analysis. Ishow how banks actively engage in asset and liability management, how they economize on reserves, and how they innovate to circumvent reserve constraints. The recent trend in bank balance sheets is consistent with the view of banks as innovators and profit seekers which comes out of an endogenous money approach. I show that the relation between reserves and bank liabilities is not a fixed one, and that the portion of reserves newly supplied by the Fed varies widely depending on monetary policy and bank behavior. Finally, I show that off-balance sheet commitments which don’t show up on bank balance sheets have increasingly subverted central bank control over banks and over money. While no regressions are presented here, I do present a great deal of empirical evidence which is consistent with the endogenous money approach, while it is inconsistent with some of the propositions of the exogenous money approach. XVli

Chapter 8 briefly reviews the evidence in support of the endogenous money approach. The evidence reviewed includes an analysis of institutional innovation and an analysis of the role of the Fed, including a presentation of what central bankers say they do. Next I examine formal empirical tests, including causality tests. The results of these are not inconsistent with the endogenous money propositions that money causes reserves and that spending causes money. However, the tests probably do little but to confirm strong priors. Finally, I show that the endogenous money approach is consistent with several analyses of the business cycle. My contribution is to link business cycle analysis explicitly with an endogenous money approach. Although this link is implicit in Wolfson (1986) and Wojnilower (1980), I believe only Minsky (1986b) has tried to make it explicit (in his financial instability hypothesis). Chapter 9 uses some simple models of banking to show how it is possible for bank credit to finance spending. Balance sheets are used to show how expansion of bank balance sheets can fund an increase in investment and how a rise in investment generates sufficient income to allow debts to be retired. In short, I show how

it is possible for simple expansion of balance sheets to allow income and expenditure to rise even in the absence of prior saving or central bank provision of reserves. This confirms the Keynesian proposition that investment determines saving, so that prior accumulation of saving is not necessary. It also confirms the Post Keynesian proposition that banks can behave as if they are free from quantity constraints. In this way, the link between investment (freed from prior saving) and credit advanced by banks, and the role credit plays in allowing income and expenditure to rise are shown in a simple model. Finally, these models are used to help clarify the relation between spending and money (that is, velocity) in an endogenous money approach. The models, of course, are too simple and mechanistic to stand by themselves as representations of a real economy. However, taken in conjunction with the previous chapters, these models may clarify how banks can finance spending, how short term credits generate ‘savings’, and how ‘savings’ finance long term positions in investment goods. The final chapter very briefly summarizes the conclusions of the manuscript, and briefly discusses the policy implications which come out of an endogenous money approach. I argue that central bank attempts to constrain the rate of growth of monetary aggregates through quantity constraints on bank reserves only XVlii

increase financial instability. As an alternative, I propose that the central bank provide liquidity needed by the system, but that it Supervise the types of assets held by banks. Rather than quantity constraints, I advocate quality constraints on expansion of bank balance sheets. The primary responsibility of the central bank should be to ensure parity between bank liabilities and government issued currency, to maintain solvency of banks, and to help maintain low and stable interest rates. While this book will primarily examine the US case, much of the analysis is applicable to other developed capitalist economies. Indeed, much of the historical analysis is based on the evolution of financial institutions in Europe and elsewhere. However, as I

discuss in Chapter 2, the situtation of the less developed country today is quite different and requires further research. Thus, readers are cautioned that this analysis is institution-specific: I analyze the evolution of financial institutions from the development of capitalism in Western Europe to the development of a modern financial system such as that of the US today. One further caveat may be necessary: I consider Chapters 2 and 6 to be preliminary reports on work in progress. In Chapter 2, I provide many examples of early societies in which credit money is endogenously created to meet the demands of credit worthy customers. However, before it can be argued that credit money has always been created on demand in all societies in which private property existed, further research is required. Regarding Chapter 6, more

research must be done on the determination

of interest

rates in a model with liquidity preference and endogenous money - particularly in the area of the relation between interest rates and rates of profits. Keynes, of course, argued that the rate of interest ‘sets the standard’ for the marginal efficiency of capital. Further analysis is required to determine how the rate of profit goes into the determination of interest rates in a model in which banks actively pursue profits as they finance accumulation of capital.

Notes 1. Readers are urged to read (or reread) the excellent history of banking theories by Mints (1945). Criticism of the ‘real bills doctrine’ has diverted attention away from other important discussions of monetary theory which took place during the nineteenth century. I have not attempted to analyze the debate over such issues as velocity and the determination of prices, but I believe that modern theory xix

could benefit from an examination of this early work. As I was preparing this manuscript, Basil Moore’s book on endogenous money (Moore 1988b) was published. Much of his analysis complements this manuscript. The primary differences between our analyses are over the shape of the money supply curve and the determination of interest rates. Moore argues that the money supply curve is, quite simply, horizontal at any given interest rate, and that the short term interest rate is set by the central bank. In contrast, I have argued that the money supply curve is upward sloping. While I accept some influence of the central bank over interest rates, I also accept Keynes’s notion that the primary determinant of the interest rate is liquidity preference. Our focus is somewhat divergent: I concentrate more upon the origins of money, on the intellectual history of the endogenous money approach, on an evaluation of institutions, and on the relation of the endogenous money approach to business cycle theory. While Moore does include a discussion of each of these topics, he concentrates on the reasons why money is supplied on demand and why central banks provide reserves on demand in all countries with developed financial systems. Because Moore accepts a money supply curve which is horizontal at the interest rate determined by the central bank, he is critical of Keynesian liquidity preference theory. Indeed, Moore argues that Keynes’s adoption of liquidity preference theory was a mistake and acted as a barrier to the development of an endogenous approach to money. In contrast, I will incorporate liquidity preference theory into my analysis as an essential component of an endogenous money approach. Furthermore, while Moore’s analysis would seem to imply that the orthodox approach would be consistent with an economy without well-developed overdraft facilities, | have argued that the endogenous approach to money can be applied at least to some degree to all monetary economies, and have traced the development of the financial system from a system with endogenous money but without a central bank, to the modern system with endogenous money and a central bank. Foster (1986a, 1986b) has argued that Keynes realized the money supply is endogenous, but adopted an exogenously determined money supply in the General Theory, and that his adoption of an exogenous money supply weakened the general argument. Robinson (1971) argued that Keynes adopted an exogenous money supply to forestall criticism. I do not pursue either of these arguments.

1

THE ENDOGENOUS APPROACH TO MONEY

Introduction

This chapter discusses the origins of money and the Post Keynesian conception of money. I will show how this naturally leads to an ‘endogenous money’ approach, in which the creation of money is tied to the normal operations of a monetary economy. I will then discuss the various assets which constitute the ‘money supply’, and will compare and contrast money demand with liquidity preference. It will be argued that money, but not necessarily liquidity, can be created essentially on demand to finance spending in excess of income. Thus, the money supply normally increases as spending rises. However, a rise in liquidity preference can generate a run to liquidity - manifested as an attempt to liquidate positions in assets - which cannot be met within private institutional relationships. In Chapter 3, I will compare and contrast the endogenous and exogenous money approaches. However, I will now briefly define what I mean by each. The exogenous money approach is the approach adopted by virtually all mainstream macroeconomic theory, including both the Monetarist school and the orthodox Keynesian school (or IS-LM Neoclassical Synthesis). In this approach, the real and monetary sectors can be analyzed separately - at least in the long run. For example, an increase in investment spending can be analyzed separately from an increase in the money supply. Thus, holding the money supply constant, rising investment spending can lead to excess money demand and force up interest rates. Alternatively, the money supply can be increased, causing excess money supply which is eliminated by falling interest rates and/or increased spending. Furthermore, it is normally argued by

practitioners of this approach that the money supply is ultimately under the control of the central bank. In contrast, in the endogenous money approach, the real and monetary sectors cannot be independently analyzed. Thus, it makes no sense to attempt to analyze the effects of rising investment with a constant money supply. Rising expenditures necessarily require deficits, and deficits require expansion of balance sheets. Exactly 1

which assets and liabilities will expand to finance such spending depends on institutional practices and relations and on central bank behavior. Conversely, since money is supplied to finance spending, it makes no sense to assume that an increase in the money supply might cause excessive money balances. Money is issued because someone wants it. There is no strict dividing line between ‘narrow money’ and ‘broad money’. Therefore, I will include as money any balance sheet item which transfers purchasing power across time. The issuer of an IOU transfers purchasing power from the future to the present, while the holder obtains a claim on future purchasing power. This emphasizes that the economy cannot be dichotomized into ‘real’ and ‘monetary’ sectors. Finally, for reasons to be discussed throughout later chapters, the central bank cannot control the money supply, no matter how it is defined.

Origins of money Money is usually defined as a medium of exchange, created to eliminate the necessity of double coincidence of wants which allows barter to take place. For example, Ritter and Silber define money as follows: Money is just what you think it is - what you spend when you want to buy something. The Indians used beads, Eskimos used fishhooks, and we use currency (coins and bills) and, most of all,

checking accounts... [MJost prominently money is what you can spend, a generally acceptable means of payment or medium of exchange that you can use to buy things or settle debts. (Ritter and Silber 1986, p. 3)

The typical definition thus concentrates on the use of money. The endogenous money approach, however, concentrates on the source of money: how is money created, and how does it enter a capitalist economy? Thus, the approach taken here will concern endogenous money, created in a capitalist economy, with its creation tied to the essential features of capitalism: private ownership, production for the market, and capital accumulation. In this approach, money is not a veil which can be ignored when theory discusses the ‘real’ sector, and later added when theory determines nominal prices. Milton Friedman argues that money, in fact, is a veil: 2

Despite the important role of enterprises and of money in our actual economy, and despite the numerous and complex problems they raise, the central characteristic of the market technique of achieving co-ordination is fully displayed in the simple exchange economy that contains neither enterprises nor money. (Friedman 1982a, p. 14)’

This view follows that of A.C. Pigou: Monetary facts...differ from ‘real’ facts and happenings in that unlike these they have no direct significance for economic welfare...take money away and, whatever else might follow, economic life would not become meaningless: there is nothing absurd about the conception of a self-sufficing family, or village group, without any money at all. In this sense money clearly is a veil. It does not comprise any of the essentials of economic life.

(Pigou 1949, p. 24)

Keynesian economics as it is presente .1 in textbooks (henceforth, ‘textbook Keynesianism’) does not differ greatly from this view. In the textbook view of Keynesianism, money is neutral in the long run, but has real effects in the short run due to real balance

effects, rigidities, or illusions. For example, Samuelson argues that ‘even in the most advanced industrial economies, if we strip exchange down to its barest essentials and peel off the obscuring layer of money, we find that trade between individuals or nations largely boils down to barter...’ (Samuelson 1973, p. 55).’ Typically, a discussion of money will begin with an hypothesized barter economy which discovers that money can be used to lubricate the market mechanism. For example, Samuelson presents the ‘historical stages of money’ as follows: Inconvenient as barter obviously is, it represents a great step forward from a state of self-sufficiency in which every man had to be a jack-of-all-trades and master of none.... Nevertheless, simple barter operates under grave disadvantages....In all but the most primitive cultures, men do not directly exchange one good for another. Instead they sell one good for money, and then use money to buy the goods they wish...Money does simplify economic life. If we were to reconstruct history along hypothetical, logical of lines, we should naturally follow the age of barter by the age commodities of variety great a commodity money. Historically, has served at one time or another as a medium of exchange:

cattle...tobacco, leather and hides, furs, olive oil, beer or spirits,

slaves or wives, copper, iron, gold, silver, rings, diamonds, wampum beads or shells, huge rocks and landmarks, and cigarette butts. The age of commodity money gives way to the age of paper money....Finally, along with the age of paper money, there is the age of bank money, or bank checking deposits. (Samuelson 1973, pp. 274-6)

Heinsohn and Steiger, however, argue: the concept of a pure barter economy from which the monetary economy devolves is no more than a historical speculation and does not correspond to the true course of history.... (Heinsohn and Steiger 1989, p. 180)

In fact, an examination of anthropological evidence reveals that early societies were

not barter economies,

that markets

did not

spring forth from barter, and that money was not invented to facilitate exchange. In the orthodox conception, individuals produce for a market which is created before money exists in order to exchange the goods they produce for needed goods which others produce, and only later is money invented to facilitate these exchanges (Levine 1983). However, the evidence indicates that money was developed first, and that markets were later developed (often with great difficulty). Barter has never been an important exchange

mechanism,

but was

ceremonial,

infrequent, petty, or

emergency behavior (Heinsohn and Steiger 1989). Indeed, the objects which were traded in pre-private property economies were not traded on the basis of needs since exchange in such societies had as its main aim ‘to exchange articles which are of no practical use’ (Malinowski 1932, p. 86). Polanyi (1971) examines ‘exchange’ in pre-private property societies. He notes that gifts were often given which were ‘construed as counter-gifts, given in return for some fictitious service rendered to the giver’ (Polanyi 1971, p. 74). Furthermore, ‘occasionally the identically same object is exchanged back and forth between the partners, thus depriving the transaction of any conceivable economic purpose or meaning...the sole purpose of the exchange is to draw relationships closer by strengthening the ties of reciprocity’ (ibid., p. 74). In such economies, parties to the exchange enter into the transactions on fundamentally unequal grounds so that ‘the exchange is usually unequal, as befits the relationship’ (ibid., p. 74). In these cases, such transactions are

irreversible, and the goods are ‘noninterchangeable’ (ibid., p. 74). Thus, while transactions occurred in early societies, they clearly were of a different character from those in modern society and do not conform to the hypothesized barter economy of orthodox theory. “Transactions in early times are public acts performed in regard to the status of persons and other self-propelling things...’ (ibid., p. 75). Polanyi goes on to argue that production for sale in markets for money did not ‘naturally’ arise out of ‘barter’. Indeed, he argues: Throughout the course of its early history the use of small coin and the retailing of food went together....Broadly, coins spread much faster than markets. While trade was abounding and money as a Standard was common, markets were few and far between....

Coinage had spread like wildfire, but outside of Athens the market habit was not particularly popular...As late as the beginning of the fourth century the Ionian countryside possessed no regular food markets. The chief promoters of markets were at that time the Greek armies, notably the mercenary troops now more and more frequently employed as a business venture. (ibid.,

pp. 83-5)

Thus, he argues that markets were developed by direct government intervention: [W]e have evidence of organizational and financial activities initiated by kings, generals or governments responsible for the military undertaking. The campaign itself was quite often no more than a rationalized booty raid.... An expedition’s sale of booty, if only for reasons of military tactics, formed as much part of efficiency as did the regular provisioning of the troops, while it avoided, as far as could be, the antagonizing of friendly neutrals. Go-ahead generals devised up-to-date methods of stimulating local market activities, financing sutlers to wait upon the troops, and engaging local craftsmen in improvised markets for the supply of armaments. They boosted market supply and market services by all means at their disposal, however tentative and hesitant local initiative sometimes may have been. There was, in effect, but little reliance on the spontaneous

business

spirit of the residents. (ibid., p. 85)

Finally, the view that early markets flourished on the basis of irec trade, with ‘higgling and haggling’ determining prices is also disputed by the evidence.

Traditionally, trade carried no taint of commerce. In its origins a semi-warlike occupation, it never cut loose from governmental associations, apart from which but little trading could take place under archaic conditions.....Treaty prices were matters of negotiation, with much diplomatic higgling-haggling to precede them. Once a treaty was established, bargaining was at an end. For treaty meant a set price at which trading took its course. As there was no trade without treaty, so the existence of a treaty precluded the practices of the market. (ibid., pp. 86-7)

In summary, barter has never been an important economic activity, barter exchange did not lead to the development of markets, and money could not have developed out of barter. In fact, the earliest money was not created to facilitate exchange, but was created as a unit of account. Keynes argued in his studies on ancient currencies: ‘Now for most important social and economic purposes what matters is the money of account; for it is the money of account which is the subject of contract’ (Keynes 1982, p. 252). In the Treatise on Money, Keynes argued: A money of account comes into existence along with debts, which are contracts for deferred

payment,

and price lists, which

are

offers of contracts for sale or purchase. Such debts and price lists, whether they are recorded by word of mouth or by book entry on baked bricks or paper documents, can only be expressed in terms of a money of account. Money itself, namely that by delivery of which debt contracts and price contracts are discharged, and in the shape of which a store of general purchasing power is held, derives its character from its relationship to the money of account, since the debts and prices must first have been expressed in terms of the latter... Money proper in the full sense of the term can only exist in relation to a money of account. (Keynes 1971b, p. 3)

Heinsohn and Steiger (1989) argue that money and interest were developed at the same

time - not out of a barter economy,

but

when private property and futures contracts were developed. The development of private property destroys the collective security of tribal or even command society and makes each member of society responsible for his own security. Private property makes loans from individuals possible, and individual responsibility induces creditors to require interest. Furthermore, each proprietor is responsible to ‘set up his own reserves as a source of security which will protect him, in an unknowable future, from the need to ask credit of other

- similarly isolated - proprietors’, which generates the drive to accumulate (Heinsohn and Steiger 1989, p. 192). Money is created as private ownership of the means of production is developed: With the establishment of private property, we at once have the elements of a money economy, that is an economy which is based on interest-bearing contracts between creditor and debtor, and brings about monetary production. This production arises from the structural insecurity of the proprietors.... Private ownership production is a monetary production because the private reserves embodied in the creditor-debtor agreement are money of account.... (ibid., 1989, p. 193)

When the private proprietor extends credit to others, he must be concerned with the risk to which he is now subject since he is giving up some of his security. Thus, a ‘premium’ must be charged, whose size ‘is not determined so much by the volume of his reserves, but rather by his estimate of his existential uncertainty: i.e., by his estimate of the existential risk he runs during the period for which he will part with liquidity’ (Heinsohn and Steiger 1989, pp. 192-3). Money thus appears in private property economies. These have occurred at various times and places in human history since the Greek city states. Keynes shows that early monetary units were based on a specific number of grains of wheat or barley. (Keynes 1982, pp. 233-36)’ Later, metals (such as iron, copper, silver, or gold) would be used as money, but the value of the metal would be determined by weight in terms of the number of grains of wheat or barley it represented. Coins were a later development, which came some 2,000 years after the development of money. (ibid., p. 255) Initially, the metal used for coinage would merely be stamped to indicate fineness, while the value of the coins would still be

determined by weight. The first purely nominal coins were produced much later, under Pheidon of Argos in the seventh century B.C. (Heinsohn and Steiger 1984). but was

was

not created

to facilitate exchange,

to reduce

transactions

costs. Thus, the loan of a cow

Thus, money

created as part of a forward contract after the development of private property.” The first money existed when private property was loaned with the expectation of payment of a greater sumin the future. Eventually, grains of barley would be used as a universal equivalent to act as a unit of account in all creditor-debtor relations

could be repaid in barley equivalents. Gradually, representative

money, in the form of metal weights, could be loaned, used as the terms in which debts were written, and used to settle debts. However,

This optimising of the storage potential of private property issued as credit, which eventually produced coinage does not represent an ‘essential’, but merely technical element of the monetary economy.... (Heinsohn and Steiger 1989, p. 193) The

market,

‘then, is not a place of barter...but

a place for

earning the means of settling debts, i.e. money’ (Heinsohn and Steiger 1989, p. 195). Production occurs to retire debt, or to accumulate reserves of wealth to reduce the probability of becoming indebted. Money ‘is private property of lowest carrying costs which must be acquired by all members of the society in question to maintain or increase their (individual) security or to avoid falling too deeply into debt...’ (ibid. p. 196). A hoard of money increases individual security, whether that hoard is merely in terms of corn or in terms of the right to another’s reserves which is given by holding an IOU. An increase in concern regarding individual security raises the premium which is required to induce one to give up reserves. With the fall of economies based on private property and the rise of feudalism, the need for a money of account was diminished. Banks disappeared around the third century A.D., ‘and the medieval feudal command economy with its internally money-less estate began its more than thousand year domination in Europe’ (Heinsohn and Steiger 1984, p. 57). Even though exchange occurred in feudal society, a monetary economy did not re-emerge until feudal society began to disintegrate in the late fourteenth century, when private property (and free, but landless, laborers) reappeared. That is, money did not completely disappear as it was still used in some relations (particularly in foreign trade) but most production was for own use and responsibilities were shared by the

communal organizations.°®

To summarize this position, money is a unit of account, created when private property arises and an individual becomes a creditor or potential creditor. In the absence of private property, money as a unit of account is unnecessary because the community (or at least a portion of the community) is responsible for the well-being of individuals. Exchange may occur (e.g. the ceremonial exchanges of tribes), and payments may occur (e.g. the payment of a portion of 8

the peasant’s produce to the lord), but these societies have no need for money. Although commodities with some of the characteristics of money

did exist in tribal society, these differed

in essential

aspects from money. Most importantly, there was no fixed exchange rate between this ‘commodity money’ (e.g. beads, shells) and other goods. Second, in societies without private property, when goods were loaned, often the society as a whole suffered the loss if they could not be returned. Individuals were normally not individually responsible for such debts. If return of a loaned item would result in individual hardship, its return was not expected. Third, in many instances, exchange occurred for the purpose of equalizing wealth, not to achieve mutually beneficial allocations of resources. The type of exchange observed in modern Western society which comes closest to the types of exchange which occurred in pre-private property societies is that of the exchange of Christmas gifts.’ This is in stark contrast to the orthodox story in which markets based on barter arise naturally out of the utility maximizing behavior of individuals, and in which money is merely created to facilitate exchange. In reality, money was created ‘naturally’ in creditor-debtor relations (which require the existence of private property), but markets developed only with great difficulty and usually only with government support. With the collapse of the institution of private property after the fall of Rome, money and . markets nearly disappeared. Money re-emerged with the development of private property and a class of propertyless workers during the evolution to capitalism. This not only leads to the development of a pool of employable workers, but also leads to the possibility of the development of a market on a large scale. In order to produce for the market, the capitalist hires from the pool of propertyless workers. As production and sale take time, the capitalist must advance the means

of subsistence,

plus raw

materials,

to his workers.

This

requires either that he advance his own accumulated reserves (of food and raw materials), or borrow from those who are willing to relinquish reserves. Because the future is uncertain and failure is possible, borrowing reserves will necessitate repayment of a greater amount to compensate the lender for the security he has relinquished. Thus, production for the market involves a ‘wheat now for more wheat later’ proposition. Money as a unit of account naturally evolves out of this process, whether money is in the form

of wheat or coins.”

Of course, as reliance on the market increases, and as production

and ‘round about’, financial complicated more becomes and more complicated. important more arrangements become important when workers becomes exchange of medium Money as a are paid in medium of exchange rather than in wage goods directly? Since wage goods constitute the majority of goods produced (particularly in the early stages of development), it is tempting (but misleading) to focus on exchanges and on money as a medium of exchange. Once capitalist production dominates the economy, money becomes universally important: money operates as a medium of exchange and money hoards provide a measure of security. However, production involves goods and services now in exchange for a promise to pay in the future. That is, money is involved in the production process because production is time-based

and

involves

debt

commitments.

Thus,

it is also

necessary to focus on production and on the role money plays in the production process. If one only focuses on the use of money in exchange or as a store of value, one ignores how money creation is inextricably related to time-based production in private property economies. Capital accumulation involves the use of money to accumulate the means of further production, which can be used to increase reserves and security. Money is created when an individual loans

reserves to an entrepreneur, who uses these reserves to purchase or create means of production which will be used to produce reserves to retire debt and to accumulate means of further production. (Money is also created to finance the production of consumption goods, but it is the accumulation process which is central to capitalist production.) As I will discuss below in Chapter 4, Schumpeter clearly recognized this when he argued: ‘Credit is essentially the creation of purchasing power for the purpose of transferring it to the entrepreneur’ (Schumpeter 1934, p. 107), and this credit then circulates as money in the form of means of payment (ibid., p. 99).

The Post Keynesian conception of money This long discussion of the origins of money serves as an introduction to the Post Keynesian conception of money and its role in a capitalist economy. There are four main views concerning money in Post Keynesian thought: Keynes emphasized money as a unit of account; Minsky (1986b) emphasizes money as a debt which 10

is issued to finance positions in assets; Lavoie (1985) emphasizes money as credit which finances a flow of spending; and Keynes (and many others) emphasized the holding of money as a Safe asset

in a world of uncertainty. These can be integrated to provide a coherent definition of money consistent with the endogenous approach. In the previous section, it was argued that money existed first as a unit of account. Money is an outstanding debt obligation of one economic agent against another, and money is created in the process of financing ownership or production of goods, services, and assets. Thus, money bridges a time gap, allowing one to purchase today and to pay later - money transfers purchasing power through time, from the future to the present. It is created as a contract is created - it is created as part of an agreement for money today in return for money tomorrow. In a capitalist economy, money is most important because it finances production of means of production and funds positions in capital assets. These assets are then used to generate money income to meet contracted debt payments. Money is created as positions are financed, money is the unit of account in which debts are calculated, and money is universally accepted as the form in which debts are repaid. Minsky argues: In a capitalist economy money is tied up with the process of creating and controlling capital assets...it is a type of bond that arises as banks finance activity and positions in capital and financial assets.... Money is created as bankers go about their business of arranging for the financing of trade, investment, and positions in capital assets. An increase in the quantity of money through bank lending to business transforms a desire for investment or capital assets into an effective demand.... (Minsky 1986b, pp. 223-4)

Money is thus created in the process of financing positions in

financial and capital assets. Such positions entail payment commitments, but the income expected from the holding of assets is uncertain. Therefore, ‘there is a demand for money to make payments on financial contracts’, since ‘money and easily negotiable financial instruments are desired to provide a margin of safety to assure the fulfillment of contracts’ (Minsky 1986b, p. 218). Since money is accepted for debt retirement, money offers security. Reserves held in money form increase security since these can be used to satisfy the demands of creditors (and potential creditors) 11

money should the future prove disappointing. In this sense, money hold to desire ‘..our provides insurance. As Keynes argued of our distrust of degree the of ter barome as a store of wealth is a The .. future.. the ning concer tions conven and own calculations p. 1973, s (Keyne ’ tude... disquie our lulls money of ion possess 115-16). ae (1988b) has distinguished commodity or fiat money from credit money. When money takes the form of commodities (such as gold) or of government fiat notes, it represents an asset with no corresponding private liability. However, credit money clearly represents both a private liability and a private asset. Credit money is the proximate liability of commercial banks, which are simply one type of financial intermediary... It follows directly that the supply of credit money is credit-driven. Whenever economic units choose to borrow from their bank, deposits and so bank money are created in the process. Whenever economic units choose to repay their bank loans,

deposits are destroyed. (Moore 1988b, p. 20)"

In a credit money economy (as opposed to a commodity money economy), much spending involves both sides of two balance sheets: that of the spender, whose liabilities increase by the amount of borrowing and whose assets increase by the value of the goods, services, or assets purchased; and that of the banker, whose assets

increase by the amount of the IOU of the borrower, and whose liabilities increase by the amount of demand deposits issued to finance the spending of the borrower. The value of the liabilities in each of these balance sheets is fixed by contract. However, the value of the assets purchased by the borrower is not fixed. For example, if the borrower has purchased capital equipment to be used in production to generate a stream of returns, then the value

of this asset will be determined by success in generating returns from production. If the borrower has purchased a financial asset, then its value will be determined by yields and capital gains. In either case, the value of the purchased asset may fluctuate below the value of the payment commitments entailed by the liability. If the borrower is forced to suspend payments or to default altogether, this leads to devaluation of the bank’s asset. Again, the

value of the bank’s liabilities are fixed so that it also experiences

problems if its assets are devalued below the value of its liabilities. This explains why nominal values in a monetary economy based on credit must matter. The concern with nominal values is not due 12

to money illusion, but is due to the fact that asset values can fall but liabilities are normally contractually fixed. In any economy in which creditor-debtor relations exist, money cannot be neutral. Money finances positions in assets, and debts are written in terms of money. Only if money were dropped from helicopters (or otherwise exogenously injected into the system) would it represent an asset without a corresponding liability. Finally, one’s spending power is determined by one’s income plus the amount of money one can obtain - that is, by income plus the amount

of debt one can issue. In other words, credit allows an

individual to spend in excess of income, and thereby finances growth of aggregate spending and income. In a capitalist economy, it is important that purchasing power can be placed into the hands of those who will increase productive capacity. As I will discuss in later chapters, deficit spending, in the aggregate, must be financed through credit creation. Deficit spending always generates economic growth in nominal terms, but it will generate real economic growth only if it finances certain types of purchases. The social purpose of credit in a capitalist economy is to finance capital accumulation. Of course, this does not mean that individual banks

and individual borrowers pursue this social purpose, since each merely seeks to maximize profits (or other personal objectives). In any case, money clearly cannot be neutral where its creation is tied to the creation of purchasing power. In summary, money is a balance sheet item, or a unit of account,

which finances a flow of spending. It is created as one transfers purchasing power from the future to the present, and can be held in an uncertain world as insurance to meet expected and unexpected payment commitments because it is the generally recognized form of purchasing power and the universally recognized means of retiring debts.

Which assets constitute money? In this section, I will discuss those assets which should be included

in a definition of money consistent with the endogenous money approach. According to the view presented above, money Is simply a debt (transferring purchasing power through time from the future to the present), which may or may not serve as a medium of

exchange.”

Thus, commercial paper may suffice as money, but

might not serve as a medium of exchange. Often money must be 13

converted into demand deposits or cash in order for it to serve in this way. Currency is the debt of the government and is the legal tender and must therefore be accepted as a medium of exchange.” Demand deposits usually serve as a medium of exchange, and (for reasons to be discussed later) almost always trade at par with legal currency. In a company town, chits may serve as a medium of exchange, but normally only the IOUs of the government and commercial banks are widely accepted in exchange, even though a wide variety of assets function as money. There is no specific relation between the quantity of money and the quantity of currency plus demand deposits (or media of exchange), because the relation depends upon institutional practices, government guarantees,

rules

of

and_

thumb,

technological

various

considerations. Orthodoxy tends to confuse the medium of exchange with money and emphasizes the medium of exchange function of money. In this way,

it concentrates

on

the

relation

between

the

medium

of

exchange (a stock) and spending (a flow), where the exogenously determined stock affects flows. I distinguish between money and the medium of exchange, between money and spending, and between the medium of exchange and spending. Money consists of the terms in which debts are written; it is the unit of account; it is the balance sheet item created as part of a forward contract. Thus, money is a stock created to facilitate flows. Some forms of money

can be used as media of exchange to finance purchases which don’t entail a continuing debt commitment. Some forms of the created money can function as means of payment used to settle debts, to reduce the size of balance sheets, and to destroy money.

The use of a medium of exchange is the use of another’s debt to finance a purchase, while the use of a means of payment is the use of another’s debt to settle one’s own debt. Only the government can print up a means of payment to settle its debts. Although commercial banks (and, potentially, other institutions) do create the medium of exchange, they do so to go into debt, not to settle debt. Money is privately created when one party is willing to go into debt and another is willing to hold that debt. This money may then circulate among ‘third parties’ as a medium of exchange to the extent that it is universally recognized as such. Normally a wide variety of IOUs function as money, allowing one to ‘spend’ now in exchange for a promise to pay later. Some of this money also functions as

a medium

of exchange, allowing one to

spend now without incurring a debt. Many forms of money may be 14

held as a store of value - as a hedge against an uncertain future. Finally, some of this money also functions as a means of payment, enabling one to retire debts and destroy money. A given quantity of demand deposits plus currency can support a wide range of levels of production and income, and a wide range of numbers of transactions, since the length of time between an act of spending (in which money is created) and an act of payment (using the means of payment to destroy the debt and money) is extremely variable. Of course, this is the notion behind velocity. Many transactions occur without involving the universally recognized medium of exchange (M1, or whatever). Others require medium of exchange, and still others fall in between - sometimes

requiring use of the medium of exchange and other times not. For example: a wide range of transactions can be made using a credit card. The use of a credit card involves a contracted debt commitment, that is, money. However, only the final transaction at the end of the month involves use of a means of payment, that is,

when the debt is retired using a demand deposit. Clearly, the use of credit cards does not in any way reduce the quantity of money, but does reduce the need for a means of payment. The consumer need only hold a demand deposit during the brief settlement period (and only in the amount of net debits) at the end of the month. During the rest of the month, income may be used to purchase interest earning assets and purchases can be made using credit cards. Finally, at the end of the month, the earning assets

are sold (or otherwise converted to demand deposits) and demand deposits are temporarily held until the credit card debt is destroyed. I have, therefore, defined

money

as the unit of account,

as a

balance sheet entry, as a promise to pay. Money is often created when one takes a position in an asset. If a transaction takes place, money is normally used, but only some transactions require use of

a medium of exchange.”

Only certain liabilities function as media of exchange. These are typically the debts of the government, or private debts which trade at par with government debt. The willingness of the Fed to enter

as a lender of last resort, the existence of the Federal Deposit Insurance Corporation (FDIC) to guarantee safety of deposits, the existence of bank equity to stand behind bank liabilities, and the ability of banks to diversify asset holdings all contribute to the ability of demand deposits to function as a medium of exchange under

most

circumstances.

Furthermore,

15

demand

deposits

are

valuable because most debt commitments can be retired using demand deposits, functioning as means of payment. Each economic unit accepts payment in demand deposits because its own debt commitments are written in terms of demand deposits (Minsky 1986b, p. 231). The importance of demand deposits lies not only in their usefulness as a means of payment or as a medium of exchange, but also in their creation when loans are made. The close relation between narrow money (defined as M1 or M2) and income is primarily due to the close relation between demand deposits and loans, between loans and investment (or the wage bill, as in Moore 1985), and between investment and income, as I will argue below

in Chapter 3. Thus, by focusing on money, I focus on the process of debt creation, which is an important part of the capitalist production process. In contrast, orthodox quantity theory begins with a fixed quantity of money and focuses on the use of money as a medium of exchange, while the portfolio balance approach takes an exogenously given money supply and analyzes the effects of changes in asset prices on portfolios. Neither of these orthodox approaches concentrates on the process by which money is endogenously created as assets and liabilities are produced and

purchased in a capitalist economy.” Liquidity, money, and spending Finally, I will distinguish between liquidity and money, and between liquidity preference and money demand. A liquid asset is one which can be quickly converted into the medium of exchange with little loss of value.” As Cramp (1989) argues, liquid assets are _ normally those with a short period to maturity - so that currency (which requires no wait) is the most liquid asset. The most liquid assets are now the debt of the government and those private debts which the government is willing to guarantee. Liquidity is one of the characteristics of money, and the amount of liquidity embodied in alternative forms of money varies. Liquidity preference is a preference for assets of greater liquidity. It can be measured by the premium required to induce someone to exchange a liquid asset for an asset with less liquidity. A rise in liquidity preference means that this premium rises. The liquidity of an asset is partially determined by characteristics 16

of the asset, but also by general economic conditions and by Subjective determinants. Those characteristics of an asset which give it liquidity include: good secondary markets; guarantees of its value given by the government or by other trustworthy institutions; the wealth and other characteristics of its issuer; and low wastage

or carrying costs. One of the determinants of the amount of liquidity embodied in a monetary asset is the degree of parity between that asset and the assets which function as media of exchange or means of payment. Currency, as the debt of the government, is legal tender and today always has a value of one. Because demand deposits play a special role in our economy, they trade at par with currency. The central bank guarantees their parity, so they are nearly as liquid as currency. The liquidity of other private debts is normally less than this because there is at least the possibility that they will not trade at par with currency. In fact, the liquidity of most privately issued money varies over the cycle. When expectations are high, commercial paper may be nearly as liquid as demand deposits. When expectations fall, commercial paper may suddenly become very illiquid: it will be traded for demand deposits only at a great discount. Because an expansion is normally based on credit expansion relative to the growth in demand deposits and currency, it is not possible to

liquidate credit when expectations fall (that is, to trade other private debts for cash or demand deposits) unless the credit is traded below par against cash and demand deposits." As Cramp (1989) notes (and as will be discussed in later chapters), the liquidity of an economy tends to fall over the course of a business cycle expansion as the maturity structure shifts toward assets which have a longer term to maturity. When preferences shift toward short term assets, large capital losses are incurred by the holders of long term assets. Finally, subjective evaluations and ‘crowd psychology’ can also go into the determination of the amount of liquidity embodied in an asset. An asset can become liquid merely if it is believed to be so by a sufficient number of people. Liquidity preference is not the same as money demand. A firm may desire the liabilities of a commercial bank in order to finance a position in capital assets. That is, the firm exchanges its own IOU

for the liabilities of a bank by taking out a loan and receiving a

checkable deposit. Both ‘money demand’ and ‘money supply’ rise because the firm has willingly increased its debt to finance a purchase and because the bank has willingly substituted its liability for that of the firm.’” Clearly, however, liquidity preference has not 17

risen. In fact, the firm’s willingess to go into debt to purchase a capital asset indicates a fall of liquidity preference: it has taken a less-liquid position. A liquid position might be defined as an ability to meet contractual payment obligations as they come due (Davidson 1986b). Thus, a liquid balance sheet contains means of payment, lines of credit, or assets which can be quickly exchanged for means of payment. These balance sheet items can be used to meet payment commitments should income flows prove insufficient. An increase in liquidity preference means a desire to shift portfolios into more liquid positions. This may involve a desire to liquidate assets (exchange less liquid assets for the means of payment), or to retire debts (which may involve the use of means of payment to settle debts). As a last resort, one might borrow to meet payment commitments. For example, a firm might try to obtain an advance on inventories, to discount commercial paper, or to utilize a line of credit in order.to obtain the medium of exchange to meet payment commitments or to obtain the means of payment to retire debt. However, if banks share the pessimistic outlook, it is likely that they will cut off credit to all but established customers. Thus,

increased liquidity preference means that the rate of growth of the money supply is likely to fall since willingness to expand balance Sheets will fall. A rise in liquidity preference may even be associated with a fall in money demand and money supply. In orthodoxy, spending is related (either directly or through the interest rate) to the quantity of exogenously supplied money. Money demand is narrowly defined as a demand for those assets included in M1 or M2. A rise in ‘money demand’ then raises interest rates because the supply of money (narrowly defined) is fixed. The rising interest rates then might decrease spending. Alternatively, if the money supply (again, narrowly defined) is increased (dropped from helicopters or injected by the central bank), spending rises directly due to excess real balances (Monetarism), or the excess money supply lowers interest rates which causes spending to rise indirectly (textbook Keynesianism). Thus, in the orthodox approach, the money supply does not increase as positions in assets are financed, since orthodoxy begins with a given quantity of money and simply analyzes portfolio adjustments which might occur if the money supply is increased or decreased exogenously. Alternatively, orthodoxy holds the money

supply constant and then analyzes the effects of increased spending (or attempts at increasing hoards) on the interest rate. An increase 18

in spending is supposed to raise income and money demand, which pushes up the interest rate because the money supply is fixed. The demand for money in the orthodox approach is not a demand for finance,

but

a

demand

for

hoards,

whose

quantity

is fixed

exogenously." However, in the endogenous money approach, money demand and supply increase when desired spending increases.” A rise in liquidity preference represents a fall in the _ desire to spend, so is likely correlated with a fall in money supply and money demand (or at least a fall in their rates of growth). Liquidity preference will not normally be met by an increase in privately supplied liquidity, so interest rates can be expected to rise. Even if the central bank intervenes to supply the liquidity desired, it is unlikely that spending will rise (even with a fall in interest rates) until expectations recover. Keynes argues: [T]he premium which we require to make us part with money is the measure of the degree of our disquietude...fluctuations in the degree of confidence are capable of modifying not the amount that is actually hoarded, but the amount of the premium which has to be offered to induce people not to hoard. (Keynes 1973, p. 116) Moreover, no amount of anxiety by the public to increase their hoards can affect the amount of hoarding, which depends on the willingness of the banks to acquire (or dispose of) additional assets beyond what is required to offset changes in the active balances. (Keynes 1973, p. 213)

Keynes clearly does not equate liquidity preference with money demand. Liquidity preference is a demand for hoards. The quantity of money available for hoards is determined primarily by the willingness of banks to issue liabilities as they finance positions in assets. If one concentrates solely on liquidity preference, one focuses on one of the functions of money: the security it offers in a world of uncertainty. Of course, this is a very important function of money. However, money is also important because it finances production and ownership of assets. I have argued that money demand is related to the demand for finance of spending in excess of income. In the following chapters, I will argue that money supply is a function of the demand for money. However, as Keynes has argued, the supply of hoards is not a function of the demand for

them. A rise in the propensity to hoard (a rise in liquidity preference) raises the price of hoards, but not normally the 19

quantity. ’ As I will argue below in Chapter 8, a rise in liquidity preference can degenerate into a panic as those who hold various assets attempt to liquidate them. Keynes argues that when liquidity preference increases, the interest rate rises sufficiently that the available quantity of hoards satisfies the demand. In Keynes's statement, it appears that this could be a smoothly functioning process: the interest rate rises so that people give up the idea of liquidating assets to increase hoards. But in a crisis, the adjustment is not smooth. As expectations decline, economic units liquidate assets to retire debt or to increase hoards. However, as Keynes argues, the quantity of liquidity does not rise to meet the increasing liquidity preference, therefore, assets can’t be liquidated in the aggregate without devaluation. Indeed, the problem is made worse to the extent that assets which were formerly regarded as liquid lose this characteristic - so the supply of liquidity may actually fall. As assets are devalued, units may actually increase attempts to liquidate them to meet payment commitments or to accumulate hoards as a way of dealing with rising uncertainty. However, each wave of attempts at liquidation only serves to further devalue assets. A panic can result due to the feedback from falling asset values to liquidity preference. Thus, Keynes is quite correct to assert that the quantity of hoards does not adjust. However, price adjustments may not smoothly and quickly equate the supply and demand for hoards: the adjustment process may take on the character of a financial panic. In summary, I will use the term money demand to indicate a willingness to issue debt, or a willingness to expand one’s balance sheet in order to spend on goods, services, or assets. This definition clearly distinguishes money demand from liquidity preference. Liquidity preference is a preference to exchange illiquid items on a balance sheet for more liquid items, or even to decrease the size of a balance sheet by retiring debt.

Notes 1. In fact, Friedman has been known to argue that one may assume that money is dropped from helicopters. Clearly, money is not an essential element of his theory of a capitalist economy. (See Friedman 1969, p. 4.) 2. Pigou is correct when he argues that money doesn’t constitute the essentials of life under certain forms of economies. However, money 20

is essential to a capitalist economy, as will be argued below. Pigou does go on to argue that ‘money is not merely a veil or a garment or a wrapper. Like the laws of property and contract, it constitutes at least a very useful lubricant, enabling the economic machine to function continuously and smoothly’ (Pigou 1949, pes): Weintraub characterizes textbook Keynesianism as follows:

It can be charged that the dominant brand of Keynesianism...has accomplished a literal return to a barter economy when the veil of money and the complications of a price system have been removed. Even a superficial examination of its form reveals that all its analyses devolve about a real system that is relieved of the basic characteristics of a money economy. (Weintraub 1961, p. 3) In the General Theory, Keynes gives an example of a monetary economy in which the unit of account is wheat (and other commodities). Each commodity has its own rate of interest. His monetary economy can function even without ‘money’ (narrowly defined). However, an asset with the lowest carrying costs emerges as the unit of account. (Keynes 1964, p. 222) One would expect that external trade would offer the most favorable conditions for the creation of money in order to reduce transactions costs. In fact, however, ‘it is specifically external trade that offers the most resistance to coinage. We find the Phoenicians of the 4th Century BC, for instance, conducting trade with foreign tribes and feudal lords in the form of real exchange of goods...’ (Heinsohn and Steiger 1989, p. 186). The main obstacle to coinage in foreign trade was the absence of a believable guarantee behind the value of the coin. Dobb argues that money never completely disappeared during the feudal ages, and that feudal lords increasingly accepted payment in terms of money during the later stages of the feudal period. However, he argues that the existence of money did not provide an incentive to abandon the feudal relations. It was only after these were abandoned that a monetary economy developed. (See Dobb 1963, Chapter II.) While Dobb’s argument contradicts the extreme position taken by Heinsohn and Steiger, it is consistent with the argument that money did not reappear as an essential component of the economy until private property reappeared at the end of the feudal period. See Chapter 2, in which I discuss the rise of banks in the Middle Ages. For a discussion of these points, see Heinsohn and Steiger (1983b). The term ‘reserves’ is used here in the broad sense to indicate reserves of wealth, and not to indicate ‘high powered money’. As Moore argues, ‘historically the need for borrowing to finance working capital has been the major force behind the development of money and banking institutions’ (Moore 1988b, p. 222). 21

10. Moore goes on to argue that credit money, which is demand-driven, is primarily a function of the wage bill. Thus, when firms decide to increase production, banks advance credit to meet working capital

expenses.

,

tL According to Moore, ‘Bank money is evidence of a private debt contract... What prevents other types of private debt (e.g. trade

credit, commercial paper) from fully becoming money is the absence of a specific clearing institution plus the absence of confidence that they will continually be convertible into legal tender on demand at a fixed parity...” (Moore 1988b, p. 21). Here, Moore identifies money as the medium of exchange. This is a narrower definition than I am developing. {2: In Chapter 2, I will discuss how government debt became fiat money and legal tender. 13. Transactions can occur without money. For example, at a ski swap, two individuals may exchange items in a barter which doesn’t involve a medium of exchange or contracted commitments. A spot barter exchange doesn’t involve money either as medium of exchange or as unit of account. 14. The differences between the exogenous and endogenous approaches will be contrasted in Chapter 3. I will also discuss in more detail why a close relation between money (narrowly defined) and spending should exist. 13: See Cramp (1989) and Davidson (1986b) for definitions of liquidity. 16. Wolfson (1986) shows that liquidity falls in an expansion as firms increase leverage ratios. For example, in Figure 11.1 (p. 134), he shows that the debt-equity ratio for nonfinancial corporations increases in expansion. Figure 11.3 (p. 136) shows that liquid assets held by nonfinancial corporations fall as a percentage of total short term liabilities issued by them. Figure 11.8 (p. 143) shows that the ‘financing gap’ (which is the ratio of capital expenditures less internal funds to capital expenditures) increases in an expansion, requiring nonfinancial corporations to rely on debt finance (which increases in an expansion, as is shown in Figure 11.16, p. 152). Finally, he shows in Figure 11.22 (p. 161) that banks increasingly use purchased funds (including large time deposits) during expansions. 1s Thus, the money supply and money demand functions are not independent. (See Moore 1988b, p. 109.) 18. As it turns out, the supply of finance in the orthodox approach comes from the supply of loanable funds, which represents income not consumed, or from money dropped exogenously into the system. I will analyze loanable funds theory below in Chapter 6. 19. This does not mean that an increase of money demand will always induce an increase of money supply. As I will discuss in later chapters, credit-rationing is a normal feature of all monetary economies. An endogenous money approach merely indicates that an 22

increase in money demand can be met by an increase in money supply, and, conversely, that money is supplied because someone wishes to go into debt. Acceptance of the endogenous money approach certainly does not mean that one is arguing that money demand is always and everywhere met (although Basil Moore often appears close to accepting this extreme position).

23

2

MONEY AND INSTITUTIONAL EVOLUTION

Introduction

In the previous chapter, I defined money, examined the origins of money, and discussed the range of assets which should be included in a definition of money which is consistent with the endogenous approach to money. I argued that money is a debt which arises with the creation of private property and which transfers purchasing power from the future to the present. In Chapter " below, I will analyze how money is created by modern financial institutions as part of the normal functioning of a capitalist economy. In this chapter, I will argue that money has always been endogenously created. Hence, the endogeneity of money does NOT rely on modern institutional relations. I will first provide some definitions which will be useful in the subsequent analysis. A discussion of early credit arrangements of Islamic society and early Western Europe will demonstrate that credit has never been constrained by the quantity of commodity money. Rather, credit relations develop with the development of a monetary economy.’ It is not the quantity of commodity money that constrains output, but general economic conditions, liquidity preference, institutional relations, technology, and the degree of development of the credit system. I will argue that credit rationing is a normal constraint faced by monetary economies, and that this

constraint results because profit seeking behavior of banks does not always coincide with profit seeking behavior of nonbanks. I will also discuss the role of the state in the development of monetary relations and institutions, and then will examine in detail how modern banking and central banking evolved in England. The current situation of money and banking in less developed countries (LDCs) will be shown to be similar in certain respects to that of earlier societies examined in previous sections.”

24

Definitions

I will define several terms which will be used throughout this chapter. Most of these signify various financial instruments or practices which may not be familiar to modern readers. A bank note is paper currency issued for circulation by banks, and may be convertible on demand into other forms of currency. Bank notes might be convertible into state issued money (gold coins or fiat money), or into privately issued liabilities (for example, Bank of England notes). However, convertibility is not a necessary characteristic of bank notes - they have often circulated even when not convertible. Reflux will be defined as the return of issued liabilities to the issuer.’ For example, when bank notes return to the issuing bank,

reflux has occurred. Returned notes may be deposited, may be used to repay debt, or (in the case of convertible notes) may be redeemed. Note redemption causes a loss of reserves so that this form of reflux is undesirable. It is also unnecessary - bank notes may circulate for long periods before reflux occurs. On the other hand, reflux of checks occurs as soon as depositors draw upon their

accounts: the checks will be presented to the bank and reserves will be lost through redemption during the check clearing process. The bank may attempt to ‘recapture’ these reserves through another reflux by attracting deposits (or through the use of some other form of liability issues). In the case of deposit banking, reflux is not undesirable as long as the public can be convinced to hold some sort of bank liability rather than the reserve asset. For example, assume a bank issues a loan by creating a demand deposit for agent A, who writes a check to agent B. When B then presents the check to the bank for payment, reflux has occurred. This is not undesirable, and will not lead to a loss of reserves, if the bank can

convince B to hold a deposit rather than the reserve asset. Even if B demands the reserve currency, the bank may still be able to recapture these reserves by issuing another liability. As will be discussed later, the bill of exchange was used in Islamic society by the eighth century, and was rediscovered and introduced into Europe as early as the thirteenth century as a means to purchase goods on credit which would be retired later Typically, the issuer of the bill when the goods were sold. and would promise to redeem goods the receive would (borrower) the bill at a specified location, in a specified currency, and on a specified date. Interest would be included in the bill of exchange, 25

as the borrower would promise a greater quantity of currency on the deferred date. The acceptor of the bill (lender) would have to find a way to return the currency received on the deferred date to his domicile. Rather than transporting currency (which was often foreign), he would prefer to have his agent purchase goods or another bill of exchange - this time drawn on the domicile of the original lender. Bills of exchange permitted the circulation of goods on the basis of credit, eliminating

the need

for circulation

of

foreign currency or bullion. Eventually, bills of exchange became negotiable: the original lender might prefer to sell the bill of exchange rather than wait to receive payment on the specified date at the specified location. As a secondary market for bills developed, they gradually were used as a medium of exchange and as a means of payment of debts. A giro might be defined as a ‘payment society’, in which the members agree to accept credit instruments of one or more of the members as media of exchange and as means of payment. The wool trade of England in the fourteenth century serves as an example of a giro. English wool exporters (staplers) required immediate payment in cash of one-third, and accepted bills of exchange issued by Flemish importers (drapers) for the remaining two-thirds. The drapers would sell the wool and buy a bill of exchange issued by a Flemish merchant payable on London. The merchant would export goods to London to be sold so that his bill of exchange could be retired. The stapler would finally be paid when these imported goods were sold in the London fairs. Since the stapler had usually bought the wool from producers by issuing bills of exchange, he would now be able to retire his own bills. Thus, the bill of exchange allowed the wool to move from the producer, to the stapler, and to the draper. This circuit gave rise to the bill issued by the Flemish merchant and to another circuit of goods going in the reverse direction: from Flanders to London.* After the development of the negotiable bill of exchange, the size of the giro and the complexity of the circuits would be increased.

The functions of a bank might be characterized as the Lombard business and the giro business. (Knapp 1924, p. 129) In the Lombard business, banks provide money now for the promise of money later. This may take the form either of a discount of eligible paper (for example, a bill of exchange), or of an advance on certain kinds of goods which will be sold. In the first case, the bank earns interest by buying the paper at a price below face value, and in the second it explicitly charges a rate of interest on the loan.

26

Banks also facilitate circulation by becoming part of the giro. The accounts of the bank are used to transfer money (as a unit of account) between debtors and creditors. The bank may also act as a guarantor - endorsing bills of exchange so they may more freely circulate within the giro. The business of a bank can be greatly expanded if it can get its customers to accept bank debt in payment. The bank can increase its profits by issuing notes in its

Lombard business, so that the volume of its discount business is

not tied to the quantity of minted coin (or other form of money) held in its vaults (Knapp 1924, p. 131). From the earliest period,

banks issued liabilities, such as notes, in order to make loans. As

the notes become generally accepted, the size of the giro is increased and the ability of the bank to expand is enhanced. Eventually, the bank debt can displace the debt of the other members of the giro, so that the bank becomes the apex through which most other debts must pass. That is, the bank becomes the intermediary, issuing its notes in loans and accepting its notes in deposits. Bank customers {members of the giro) will use bank notes between themselves because these notes can be used to settle debts with the bank. Neither convertibility into other forms of money (for example, state currency), nor circulation outside the giro is a requirement for circulation within the giro (Knapp 1924, p. 135). However, the possibility of use outside the giro will clearly become important as the scale of potential payments outside the giro becomes important, and convertibility may be required before the giro can expand. As I will discuss below, the state eventually becomes a member of the giro and in fact encourages the process through which the bank’s liabilities supplant all other private liabilities in the giro. I will define commodity money as circulating currency whose value is determined by an objective measure - usually weight. As discussed above, the weight of grains of barley or wheat formed the first unit of measurement, and equivalent weights of gold, silver, bronze,

and

iron were

later

used

as more

convenient

substitutes (Keynes 1982, pp. 233-6). Thus, minted coins containing a specific quantity of a precious metal are commodity money. Fiat money will be defined as currency issued by the state whose value is purely nominal. While precious metals may be used in the production of the fiat money, the value of this currency is unrelated to the quantity of metal contained in the coin. Legal tender will be defined as any form of money which the state has officially recognized for payment of debts. In most cases, any form of money eA

accepted in payment of taxes may be called legal tender. By accepting private liabilities (such as bank notes or checks) in payment of debts to the state, the state becomes a member of the private giro.

Endogenous society

money without banks: the case of Islamic

The medieval Islamic world serves as a useful example for an exposition of the endogeneity of money in the absence of banks, deposits, and explicit interest. Islamic law provided severe sanctions against

interest

bearing

loans,

speculative

transactions,

certain

forms of delayed payment, and other commercial and exchange

operations.’ (Udovitch 1979, pp. 256-7)

The prohibition against usury and interest ‘precluded the coalescence of two financial practices...the generation of capital by means of numerous deposits on the one hand, and moneylending and the provision of credit on the other’ (Udovitch 1979, p. 259). Under Islamic law, a depositor can neither earn interest on deposits nor can he be charged a fee. Furthermore, anyone who accepts a deposit must hold it for safekeeping and return it on demand. That is, it was impossible for intermediaries

to develop which would accept deposits and loan them at interest. However, such proscriptions did not hinder the development of sophisticated market activities normally associated with societies with evolved banking institutions, for ‘as early as the late eighth century, credit arrangements of various types were an important feature of both trade and industry in the Near East’ (ibid., p. 260). Merchants bought and sold goods on the basis of credit - whether for local or for international trade. One of the credit techniques utilized was the ‘credit partnership’ in which ‘people form a partnership without any capital in order to buy on credit and then sell’ (ibid., p. 263). It was called a ‘partnership of good reputations’ because the credit was advanced on the basis of the reputations of the partners. The partnership could thus obtain commodities on credit, hoping to later sell them at a higher price so that they could retire their debt and retain profit. The sanctions against interest charges were avoided by quoting two prices: one for immediate payment and one for future payment. Islamic law permitted a premium for deferred payment (which is implicitly an interest rate) because the merchant faced

28

some risk that the partnership would default. Bills of exchange (‘suftajas’) were widely used as a means of deferred payment. However, unlike the Western bill of exchange, the suftaja did not involve the exchange of currencies and did not develop into a negotiable instrument. Paper notes (‘ruqcas’) were issued by well known merchants and ‘could be used not only in commerce but even to pay government toll collectors’ (Udovitch 1979, p. 270). These were also used much like modern checks to

settle petty debts. Most merchants engaged in banking activities such as: ‘money changing, issuing letters of credit, accepting deposits, [and] acting as a clearinghouse for payments...’ (ibid., p. 271). Merchants allowed ‘overdrafts’ by good customers, and accepted promissory notes in lieu of cash payment. Thus, they provided ‘checking accounts’ and made ‘loans’ by issuing their own liabilities even though they were prohibited from loaning the deposits of depositors. It is important to emphasize that the medieval Islamic society was not a ‘monetary economy’. Medieval Arabic has no words for banks, banking or money. Banking activities were closely tied to commercial

activities, and credit was

advanced

on the basis of

social-personal relationships among merchants and their customers. Most production was undertaken for nonmarket use, so ‘the scale of economic activities was necessarily restricted to numerous small, even intimate, circles. The possibility of expansion into a larger, more cohesive structure was precluded by the comparatively narrow social basis on which economic life was conducted’ (ibid., p. 273). Thus, the development and organization of banking was also limited. However, it is clear that the quantity of commodity money (gold and silver coins) did not constrain commercial activity. While the demand for coins ‘was constantly greater than the available supply’ (ibid., p. 270), the excess demand was met through the creation of credit and banking techniques. The prohibition of interest payments on deposits (as well as the restricted development of markets) may have retarded the development of banks, but this did not prevent

the development of alternative credit arrangements. The credit advanced by merchants represented liabilities of the merchants and was constrained funds deposited intermediaries. Thus, the case view that credit

by their ‘good reputation’. As they could not loan with them, they clearly were not functioning as

of Islamic society supports the endogenous money money can be created on demand. Credit money

29

is a private liability, created to give purchasing power to the borrower. The creation of money is necessarily a ‘money now for more money later’ contract, so it involves interest, even if usury is explicitly prohibited. In contrast, the orthodox story in which money creation occurs where a goldsmith realizes he need only hold a small reserve against notes issued cannot apply to Islamic society, where this intermediary function was prohibited by law.

Premodern financial institutions and the rise of banking in Western Europe For the purposes of this chapter, I will define ancient society to cover the period up to the fall of Rome, while the Middle Ages will cover the period from ancient society through the twelfth or thirteenth centuries, and the premodern period will cover the period from the Middle Ages to the modern period, which begins in the late seventeenth century. As discussed above in Chapter 1, money was first developed as a unit of account with the rise of private property. The ancient Near East developed a variety of monetary institutions and financial instruments between 3000 and 500 BC. In the ancient,

middle, and premodern periods, there was typically a small monetized sector within a large nonmonetary economy. Barley (or other grain) was used as the first money of account, with silver (or other precious metals) later substituted on a weight-equivalent basis. The financial instruments in use were primarily trade and consumer credit which were privately created. Credit was usually obtained to finance inventories (especially in foreign trade), consumption by the aristocracy, or provision of seed corn to peasants. The major suppliers of credit included temples, private merchants and landowners, or royal treasuries. They loaned their own reserves of grain, money, or other goods - financial intermediation was rare. Indeed, claims against financial institutions were not very important until near the end of the premodern period (after the thirteenth century). Interest rates were typically between 20 and 33 per cent, although interest rates ran as high as 50 per cent on consumer loans. Except in the last case, the interest rate seems to have been fixed according to the typical rent received from land. In general, the extent of foreign trade serves as a good indication of the degree of development of financial instruments. That is, those societies which were more involved in foreign trade

30

had a higher level of development of financial institutions, due to

the need to finance goods in the process of transport.‘ I will first examine ancient society, taking Greece and Rome as examples, before turning to financial practices during the Middle Ages. Since the extent of private ownership of the means of production in ancient Rome and Greece was limited, money and financial institutions remained primitive. While money originated as a unit of account with the development of private property, it did not disappear with the rise of the Greek and Roman empires. However, the scope for private creation of money was limited by the narrow field of private contractual relations, which were primarily limited to consumer loans and mercantile shipping. Most production remained outside the market. While government spending was significant, it relied on the use of commodity money rather than on credit. Strong imperial power enabled the state to extract taxes, seigniorage, or tribute in the form of commodity money in order to raise sufficient revenues. Thus, credit money was nearly insignificant in ancient society, and became important only _ near the end of the premodern period. Credit was not constrained by the quantity of commodity money - it simply wasn’t very important in these societies.’ In Periclean Athens, the citizens owned the land, but resident foreigners were allowed to engage in manufacturing (and, indeed,

dominated trade and finance), while slaves did much of the work. In early Athens, the state minted silver coins, as commodity money, whose invariable fineness led to their use throughout the Greek world. However, because other Greek cities minted coins, money changers were necessary. Banks existed as early as the late fifth

century BC, but only on a small scale, accepting deposits and making loans to finance consumption and liturgies. Banks did not finance commercial trade. The only common commercial loans were bottomry loans, which were apparently made out of private reserves at high rates of interest (15 to 30 per cent per six months).* Otherwise, transactions were financed by the proprietor’s own funds. Government spending on the navy, on public works, and on ‘welfare’ (such as the mass juries) was equal to about one-fifth of total national product. Perhaps half of the government revenue came from tribute paid by allies of Athens, while the rest came from state silver mines, from other state properties, and from customs duties and taxes on noncitizens. The state did not borrow from the public, nor did it issue fiat money to finance expenditures. The early Roman Empire was the largest economic unit In the 31

Western world until the nineteenth century. However, it was primarily agricultural with only a small monetized sector. Thus, ‘Augustan Rome...had no fiduciary money, no credit creation, no negotiable instruments, no business loans except bottomry loans...and no long- term partnerships or corporations’ (Goldsmith 1987, p. 36). The money supply consisted of gold and silver coins which were legal tender and which would be redeemed on demand for bullion, and copper alloy coins which were not legal tender. Whenever the demand for coins exceeded the supply, debasement occurred to fill the gap, so that by the time of Diocletian, it took 130 coins to equal the same amount of gold as that contained in one coin during the time of Nero. Only the emperor could mint - coins, and at seignorage rates of 25 per cent and more, although tokens were issued on occasion to meet excess demand for coin. Consumer loans to the poor and wealthy, mortgage loans on agricultural land, and debts of tenants to landlords constituted most

credit. Lenders were mainly wealthy individuals, but occasionally loans were made by religious or benevolent institutions, and professional money lenders specialized in consumer loans. Money changers were common, and most cities had at least one banker. However, banks were not very important as a source of credit, as

they primarily specialized in loans for consumption or political expenditures. Interest rates were generally between four and six per cent, and ‘apparently did not approach the legal maximum of 12 percent’, although unsecured consumer loans to the poor had interest rates of up to 48 per cent (ibid., p. 44). There is very little evidence of commercial credit and it appears that ‘book credit would be the only possible form of financing’ (ibid. p. 46). Bottomry loans were the only major form of trade credit, and were made by groups of wealthy lenders. Foreign trade seems to have been financed through bullion flows, with the negative imperial trade balance supported by payment of gold tribute to Rome. Government expenditure probably did not exceed 5 per cent of national product, and was supported through customs duties, rents from public land, and tributes paid by colonies. The tributes were

collected by tax farmers, which were generally companies formed by groups of wealthy individuals. Tax farming contracts were very profitable and were sometimes sold in secondary markets. Taxes were paid in cash, or sometimes in kind. The existence of money and primitive financial institutions in the ancient period did not lead to a ‘monetary system’, in which production for sale on the market for money is the norm. Most

32

production occurred outside the market, and private, contractual

arrangements were relatively rare. Neither Rome nor Greece had

need of extensive credit relations. When Rome fell in AD 476, the

size of the monetized sector declined because the feudal mode of production, which was not based on private property, had no need for money for internal relations. However, this does not mean that money and financial relations disappeared, for they were still necessary for external relations. Indeed, financial institutions continued to evolve over the next thousand years. Banks were operating again in Europe as early as the eleventh and twelfth centuries. There were four main types of ‘banking institutions’ in the early Middle Ages: pawnbrokers, money changers, deposit bankers, and merchant bankers (the elite) (Lopez 1979, pp. 6-7). As commerce grew, new opportunities emerged for the merchant bankers to grant credit based on trust (ibid., p. 9). Early deposit banks in Italy created giro accounts, which could be used to settle payments where both the payor and payee had accounts with the same bank (Cameron 1967, p. 315). The government of Genoa began to regulate banks in the twelfth century (requiring them to keep cash and record books available for inspection); in return, the government guaranteed the credibility of the bankers and recognized entries in their books as legal proof of transactions (Lopez 1979, p. 11). Banks used overdrafts to expand credit or created ‘fictitious’ bills of exchange: a merchant or banker would charge a banker with supplying foreign exchange in a foreign place, but would agree to waive repayment abroad in order to receive postponed payment in the currency and place of origin (ibid., p. 15). In the Middle Ages, bullion was the primary form of means of payment and was constantly in short supply, particularly during the fourteenth century, when many mines were abandoned throughout Europe (due to floods, cave-ins, or high labor costs) (Day 1987, p. 33). Between 1370 and 1469, only about three-quarters as many coins (in current value) were struck in England as between 1270 and 1369 (and in constant value, less than 60 per cent as many). The problem was compounded because loss of coin due to wear may have been as high as 1 per cent per year (Munro 1979, pp. 181-2).° Faced with the chronic shortage of coins and insufficient state purchasing power, governments would try to debase the currency by adding copper - but ‘debasement was easily recognizable, and the purchasing power of a denier was not determined by its nominal value but by its actual silver content

33

(Lopez 1979 p. 18). This meant that privately issued liabilities were necessarily required to fill the gap and facilitate circulation. Many types of private credit instruments were increasingly used toward the end of the Middle Ages and the beginning of the period. In the thirteenth century, the West premodern ‘rediscovered bills of exchange’ (Braudel 1973, p. 359), and these ‘financed the bulk of the merchandise trade between the Mediterranean and northern Europe in the second half of the thirteenth century’ (Day 1987, p. 173). Bills of exchange circulated upon endorsement as early as 1410 (Braudel 1973, p. 359). By the late 14th century, letters obligatory were freely assignable so that privately issued debts could circulate as means of payment. In 1437, the London Mayor’s Court issued a decision that ‘recognized the transferability of a formal bill of exchange and the bearer’s legal claim to full payment’ (Munro 1979, p. 215). Full discounting wasn’t recognized until the eighteenth century, but some evidence exists of discounting of letters obligatory at the Antwerp fairs as early as 1536 (ibid., p. 215). Goldsmiths issued notes against deposited gold in the mid-seventeenth century (by 1666 goldsmiths had issued 1.2 million pounds sterling). In Venice, banks had issued redeemable notes since the fifteenth century, and bank notes were

commonly issued by the 1660s (Braudel 1973, p. 360). Proscriptions against usury played a role in

financial

developments.” In 1049, the Council of Reims proclaimed ‘no cleric or layman should be a usurer’ (Lopez 1979, p. 4). Laws against usury viewed it ‘specifically and exclusively as a fixed, predetermined,

certain return

on a loan’ (Munro

1979, p. 170).

Bills of exchange could conceal the rate of interest within the exchange rate on bills. Because the exchange rates between currencies

could fluctuate, the return

on these

instruments

was

permitted as a reward for risk. The prohibition against usury slowed the development of the negotiable bill of exchange because open discounting was hindered. The laws against usury, however, were not the principle reason for the development of the bills of exchange, for other methods existed which would permit the concealment of interest. For example, interest charges could be deducted in advance from loans; late payment charges could be added (with the late payment ‘prearranged’); or two prices could

be quoted: one for deferred payment.”

We can look at the development of banking in Italy, early England, and Aragon as specific examples of the evolution of banking during the premodern period. In Italy, banking developed 34

out of money changing. Money changers operated in Lucca by the twelfth and thirteenth centuries. They changed petty foreign coin into legal Lucchese tender, lent sums in foreign coin to ctients, dealt in gold and silver (accepted raw gold, sold it to gold-beaters, and bought thread and leaves for resale), made loans, and accepted deposits from clients (both time and demand deposits) (Blomquist 1979, p. 60). Depositors were usually from the middle and upper classes, while loans were often made to peasants - who borrowed coin to buy seeds or tools and repaid the loans in kind with grain or wine (ibid., pp. 63-4). The typical loan was smaller than the typical deposit, but deposits were generally longer term than loans. By the mid-thirteenth century, the money changers made frequent and substantial loans to merchants, while as early as 1200, a money changer would settle a debt for one client by transferring the amount owed to the account of the creditor (Blomquist 1979, pp. 65-7). Money changers began to keep accounts with one another ‘to facilitate settlement of debts through bank transfer’ (ibid., p. 67). Loans were sometimes allowed by providing overdrafts to depositors. Thus, we observe three banking functions performed by money changers as early as the thirteenth century: the giro function, the clearing mechanism, and credit creation. By the fourteenth century, banks were operating in the major port towns. These offered giro accounts which were used to settle accounts between customers, and acted as intermediaries, loaning deposits. They also created ‘banco’ money (notes) which would circulate in the community and expand the volume of the bank’s business (Day

LOS Tepe 2). Italian money changers and deposit banks operated primarily in the local market, while international commerce was in the hands

of mercantile banking partnerships operating in the major towns by the mid-thirteenth century. (Blomquist 1979, p. 68) International banking developed as early as the twelfth century in Lucca; by the 1150s Lucca was involved in the six fairs held annually in Champagne - sending silk and buying products to bring back home. This led to the development of an organized money market in Lucca: the buyer of exchange ‘delivered funds in Lucca and received from the seller, the taker, a notarial instrument promising repayment at one of the Champagne fairs in an equivalent amount of money of Provins’ (ibid., p. 71). Since repayment would occur in the future, the buyer of exchange was a lender and the seller a

borrower, and ‘with interest on the buyer’s capital built into the

fluctuating rates of exchange’

(ibid., p. 71). For example,

ee

the

borrower (seller) would receive more local (Lucchese) currency per foreign (Provinois) currency to be delivered at the fair the closer the borrowing date was to the date of the fair (ibid., p. 74).

Detailed records of the Medici bank of Florence from 1427 allow fairly accurate estimation of the importance of finance during this period in Italy. Loans made up about a third of the bank’s assets, while accounts payable equaled about 60 percent of liabilities. The major financial instruments were

coins, bills of exchange,

trade

credit, and government debt (which equaled one third of total financial instruments). The aggregate quantity of financial assets in ' Florence was equal to one half of the total value of real assets. However, financial institution liabilities were equal to only about five per cent of total assets. In other words, there was a large financial system, but most of it remained outside the banks.” While currency exchange played a major role in the development of the Italian financial system, state finance shaped the development of the English financial system. Italians dominated English finance in late thirteenth and early fourteenth centuries, and were primarily involved in the export of wool and in loans to the crown. Italians also made payments on the Continent for the English. Loans to the crown and foreign payments enabled the state to conduct foreign wars (they loaned over £125,000 sterling in 1338-39 at the outbreak of the Hundred Years War) (Prestwich 1979, p. 79). The loans to the state were made at high interest rates, and were rarely repaid in full. However, the Italians received special benefits in return for their services. Italian bankers did not play a big role as depositories - but did accept short term funds in order to make payments on behalf of the depositor (ibid., p. 96). The crown had strict prohibitions on export of coin, so Italian bankers would export letters of exchange. They would accept pounds in England, then deliver, for example, livre tournois in Paris (ibid., p. 97). The crown also maintained strict control on exchanging foreign coins for sterling: it was permitted only at the royal exchange. This was to prevent the export of English bullion

and to prohibit the import of counterfeit, clipped coin. For this reason, domestic banking couldn’t develop out of currency exchange in England, as it had in Lucca (Prestwich 1979, p. 99). The main role of Italians in England remained state finance: lending on the security of the customs so the crown could anticipate revenue (ibid., p. 103). In any case, ‘despite the fact that they were hampered by prohibitions on usury and using many techniques strange to modern bankers, the [Italian] companies

36

provided essential facilities of credit and exchange...’ (ibid., p. 104). Banking was, however, slow to develop in England. Even as late as Elizabethan England (late sixteenth century), there were no true

banks or government bonds, and short term credit was limited to

consumer debt, small advances of working capital, and loans to the crown (Goldsmith 1987, p. 186). However, this does not suggest that credit was not important. Indeed, evidence from England suggests that virtually all sales were

made on the basis of credit (McIntosh 1988, p. 560). Consumers would settle accounts with coin, but merchants would use coins

only to settle net debits. The evidence shows that debts might be carried for months (and even for years) on the accounts of merchants before finally being settled. Any two people might build up a number of outstanding debts to each other. As long as goodwill between the individuals remained firm, the balances could go uncollected for years. When the parties chose to settle on an amicable basis, they normally named auditors who totalled all current unpaid debts or deliveries and determined the sum which had to be paid to clear the slate. If trust between the parties broke down, the complainant could bring suit in the Havering court.... (McIntosh 1988, p. 561) Thus, detailed evidence shows that small credits played an essential

role in premodern England before the development of an extensive banking system. Banking in late medieval and early modern Aragon displays many modern characteristics, including substantial government intervention. Jewish banking was well established by the twelfth century, and Christian banking was developed by the thirteenth. Interest charges were authorized in 1217 (although they had existed unofficially for generations), and maximum interest rates were set at 20 per cent in 1229 (Riu 1979, p. 137). In 1235, the crown ruled that Christian bankers could lend at 12 per cent, while Jews could

lend at 20 per cent (ibid., p. 138). In 1284, the books of the money changer/banker were accepted as proof that transfers had occurred; at the same

time, various regulations were

placed on

bankers (for example, they had to insure their banks with the government so that the funds of depositors were guaranteed). The first laws concerning fraud and bankruptcy had been passed around

1300, after the first bank

failures occurred

in the late

thirteenth century. The city of Barcelona began to use the private banks as a treasury, depositing funds and relying on the banks to

37

float city debt (at an interest rate of 20 per cent) (ibid., p. 143). Due to continuing insecurity and failure of private banks, Catalonia founded a public municipal bank at the beginning of the fifteenth century - and Valencia and Palma de Majorca soon followed suit (ibid., p. 149). The municipal banks operated as a treasury for the cities, offered saving accounts to private citizens, made loans to cities (typically at five per cent interest), and offered demand deposits with checking privileges to private citizens, communities, and public corporations. Certain agents were required to hold deposits

in the

municipal

banks,

such

as

executors

of wills,

guardians, and court officials. These banks held cash against the deposits of private citizens, but deposits of businesses and others were balanced against loans to the state, official organizations, and

private individuals (ibid., p. 151). Private banks held reserves in the municipal banks and borrowed from them to meet payments. Money orders and checks were used by 1436 to transfer funds between accounts. Between 1441 and 1450 laws were passed to decrease private competition with the municipal banks (ibid., p. 153). For example, private banks were forbidden to offer credit on the basis of bills of exchange. This brief introduction to ancient and premodern financial institutions and to the evolution of banking gives some idea of the range of practices and instruments available before the development of a monetary system with most production for the market and with most individuals involved significantly in market activities. This examination has shown that private credit instruments could substitute for state or commodity money as necessary. Credit in such societies was primarily created to finance trade and had little to do with production. In such pre-capitalist societies, money and credit transferred purchasing power across time, but served primarily to ‘lubricate’ exchange. As I will discuss below, money and credit become much more important as capitalism develops. The example of Aragon, however, serves as an introduction to increasing involvement of the state in financial practices, a subject to which I now turn.

The role of the state in the development of a monetary system

In the ancient societies previously examined, the state could rely

38

on coercion to obtain purchasing power. Property rights were neither well developed nor well respected. While markets existed, the State was not excluded as a regulating force. Indeed, as discussed in Chapter 1 above, the state played a major role in creating markets in the first place, and prices were set by treaties negotiated by states. Gradually, however, with the transition to capitalism, the state came to be excluded from markets as property rights gained recognition (Heilbroner 1985). As I will discuss in this section, the transition to capitalism made it more difficult for

the state to obtain purchasing power through the use of coercion, and so necessitated the development of other means of obtaining finance. Above I discussed the role of the state as the guarantor and enforcer of contracts. However, during the Middle Ages, the state became much more directly involved in the creation of monetary instruments and institutions. With the rise of the modern state came the concommitant rise of an almost insatiable desire for credit - primarily to finance military adventures, although the high consumption levels of the crown cannot be ignored. Italian cities had a long tradition (dating back to Roman times) of ‘farming out’ tax collection to individuals who would retain a portion of the revenues. In a time of unusual need for revenue, the city could ‘anticipate’ its revenue by selling the right to collect future taxes. As early as 1164, the Venetian Republic obtained an advance on future tax revenues (Ehrenberg n.d., p. 47). Because

all the citizens of a republic were held to be individually responsible with unlimited liability for city debts, cities typicalty experienced no difficulty in obtaining such loans. In Venice, the Bank of St George monetized the city debt as early as the mid-fifteenth century, and the Banco del Giro was set up in 1619 specifically to provide finance to the government.” Even as late as the sixteenth century, cities expressly gave creditors the right to the person and property of all citizens should the city default on debt (ibid., p. 33). Indeed, in Genoa, the debts of the city were so liquid

that they circulated as freely as currency, functioning as means of payment (ibid., p. 48). While republics generally were able to borrow without difficulty, monarchies were very limited in their ability to obtain loans. While the crown could also anticipate revenues, its subjects were not normally liable for crown debts. In addition to advances on future

tax revenues, the crown could resort to the sale of offices, the sale of land, net bullion inflows, and debasement of coin. Of these, only

ag

the last two could be a major continuing source of revenue. As I will discuss below, the policy of Bullionism was designed to increase bullion inflows. However, the state mint was limited in its

ability to provide purchasing power to the crown through debasement, because traders became quite adept at discovering

debased coin. Typically, traders would base the value of a coin on its metallic content and not on its nominal face value. That is, these coins had not achieved the status of fiat money, but still

represented commodity money. Traders would also develop their own currencies or credit substitutes (Ehrenberg n.d., p. 31). This will be discussed below. As the conduct of war passed from the hands of the citizenry to hired mercenaries, the ability to raise revenue became increasingly important. During the thirteenth and fourteenth centuries, Italian cities increasingly hired the Condottieri (professional private undertakers) for military purposes. Given the ability of the city republics to obtain advances on revenues, finance of the mercenaries could normally be met through expansion of credit. In general, military spending accounted for one-third to two-thirds of total government spending. In Medici Florence, military spending alone during the period 1421-30 was twice total government revenues (Goldsmith 1987, pp. 164 and 249). The Genoese debt in 1408 ‘was equal to about four times the total volume of Genoese maritime trade that year’, with much of that debt accumulated during a war with Venice (Day 1987, p. 158). As another example, ‘the expenditure of the Spanish Crown in putting down the rebellion in the Netherlands averaged two to three million gold crowns a year, i.e. more than the yearly revenue of the Netherlands Government during the most flourishing trade period’ (Ehrenberg n.d., p. 28). Furthermore, ‘in the sixteenth century there were only twenty-five years, in the seventeenth century only twenty- one years, in which there were no war-like operations on a

large

scale’

(ibid.,

p.

28).

After

1585,

the

Elizabethan

government’s spending on the military accounted for four-fifths of total government spending (Goldsmith 1987, p. 193). Thus, the enormous costs of war became a normal expense which had to be met. While city republics could meet this through credit creation, monarchies had much more difficulty. It was simply impossible for a king to build a war chest large enough to conduct even a small war without borrowing. The result was a never ending cycle of debt. Typically, the crown could only borrow on the basis of 40

anticipation of a specified revenue source. Even then, the crown frequently had to obtain a guarantee from a group of respected

wealthy individuals, from the church, or from a city in order to

obtain an advance, normally made only at extremely high interest

rates - even as high as 42 per cent (Prestwich 1979, p. 87).'4

However, the high interest rates and the continuing needs of war finance forced the crown to default on interest payments and even on principal (Ehrenberg n.d., pp. 39-41). Indeed, the sanctions against ‘usury’ frequently arose because the crown found it expedient to abrogate its own interest commitments (ibid., pp. 34, 43). For example, in England interest charges on loans made to the crown were explicit until 1240, when Henry III forbade usury as a way to reduce his own payments on debt (Prestwich 1979, p. 85). Sanctions against usury could still be avoided by including a large penalty charge for late repayment, and then requiring repayment

in-an unrealistically short period.”

If interest and principal were rarely repaid, why would anyone provide loans to the crown? In some cases, the loans were forced,

but many loans were made voluntarily in return for special treatment. For example, Italian lenders to the English crown received in return for the loans which were rarely repaid: land; exemptions from taxation and jury service; commercial advantages; use of the exchequer’s machinery against clients who defaulted on their obligations;

and

favorable

treatment

in courts

(Prestwich

£9795..091). In spite of these concessions,

even

by the end of the Middle

Ages, most of the crowns of Europe were hindered in their ability to raise revenue or to issue loans.'* There were three primary means through which the state could increase its purchasing power. First, it could make its subjects responsible for state debts. Citizenship, patriotism, and representative government all contributed to this development. These will not be discussed further, as they are beyond the scope of this study. Second, the state could increase its ability to generate revenue. I will discuss

Bullionism, which was a policy designed to increase the flow of

commodity money to the crown. Finally, the state could develop fiat money to replace commodity money. In this case, purchasing power could be increased merely by minting coins. This would require, however,

that state money

become

more

desirable than

private credit money, as otherwise state money would circulate at a discount within the giro. Bullionism was a perfectly natural development encouraged by

41

the crown’s chronic inability to raise revenues and by its need to conduct foreign wars. This partially explains the concern with finding new sources of gold and silver, and the resulting race to colonize the Americas. Alternatively, if a trade surplus could be generated which would lead to a bullion inflow, then the crown could tax a portion of that inflow to meet its military and other payments. Above all, the crown had to prevent export of bullion. Traders, however, used bills of exchange as the primary means of international trade. The crown frowned on the use of such bills because they also reduced the possibility of bullion imports. There was also a belief that bills somehow led to the export of bullion which led to periodic attacks on the Italian bankers. For example, as early as 1340 the English crown tried to force wool exporters to pay a tax in the form of precious metals. The wool exporters responded that Flanders (buyer of the exported wool) forbade the export of bullion so that payment of the tax was impossible, and the tax was repealed. The crown tried again in 1429. It required staplers (wool exporters) to demand full payment in the form of English coin, and to bring one-third to the English mint. However, normal trade had been based on payment (by drapers, the purchasers) of one-third in cash, while the remainder was received in the form of two bills: one payable in six months and the other in twelve.’’ Thus, the entire trade was

based

on

circulation of bills of exchange. The English staplers complained that the ban on the use of credit was destroying the export market,

and the law was finally repealed in the 1470s."

The state in the late premodern period was still outside the giro. It needed a source of spending power to conduct its wars, but because its own debts were not acceptable, it was forced to rely on bullion (commodity money) for payment. Thus, Bullionism was adopted as a way to accumulate commodity money to increase the Sstate’s purchasing power. However, Bullionism keeps the state outside the giro. Private traders are thereby encouraged to continue to use privately created credit money as a way to remain outside the crown’s taxing power.” As long as trade can occur on the basis of private giro money (such as bills of exchange), the state’s ability to raise revenues remains limited. The state could respond by debasing the coin: in 1250, the English pound contained 324 grams of pure silver, but this had fallen to 112 grams by 1600. Similarly, between the years 800 and 1600: the French monetary unit fell from 390 grams to 11 grams; the Milanese

unit fell from

390 grams

42

to 4.9 grams;

and

the

Venetian unit fell from 390 grams to 3.5 grams (Cipolla 1980, p. 201). While it is difficult to obtain general price levels in terms of minted coin, and while there is considerable controversy over the so-called ‘price revolution’, it is ‘generally held that between 1500 and 1620, the average level of prices in the various European countries increased by 300 to 400 percent’ (ibid., p. 227). As long as the state remained outside the giro, its subjects would rightly object to debasement: the bullion they paid to the state was undervalued by the state mint as it attempted to obtain seigniorage profits. In response, prices in terms of minted coin would rise to compensate for debasement. Alternatively, traders could turn to their private giro. For example, in 1619, the Hamburg giro bank developed a unit called the mark banco which was used for transactions within the private giro. This unit fluctuated in value against the state monetary unit (thaler) (Knapp 1924, pp. 144-9). In the early seventeenth century, the ‘Genoese created a perfectly stable accounting unit or moneta di fiera (fair money), the scudo di marco...’ of constant value against gold (Day 1987, p. 148). The inflation which thus resulted from debasement of state currency was due neither to ‘excess demand’ nor to a simple Monetarist effect of ‘too much money’. Prices rose because they were noted not in terms of fiat money, but in terms of a quantity of metal. Only after the replacement of commodity money by fiat money would the state gain purchasing power by ‘printing money’. This was recognized by Keynes, who argued: When...a coin is no more than a quantity of bullion, of which the

stamp may certify the quality and indicate the quantity, [it]...will not circulate except for its bullion value. In this elementary stage the expedient of debasement is not available. It cannot appear, until with the development of contract the conception of a money of account has emerged, and the coins issued by a state have acquired the character of legal tender and enjoy a cours force as the legal discharge of obligations calculated in this money of account. It is at this stage that money, in the sense in which we understand it, makes its entry into human institutions. (Keynes 1982, p. 226)”

The modern state can become a member of the giro by accepting the liabilities of banks in payment on taxes and in other dealings with the state. When the state becomes a part of the giro, the liabilities of the bank become the generally accepted means of payment. At this point, the bank note is not only useful in dealings

43

with the bank and with the bank’s customers, for it is also useful in dealings with the state, the newest member of the giro. This would expand the state’s ability to raise tax revenues. It also makes the bank note acceptable to citizens who were formerly outside the bank giro since they can use the bank notes to meet the demands

of the state. That is, those with obligations to the state become

members of the giro by virtue of their relations to the state - even if they have none with the giro bank. The state’s power can be increased if it can become the apex of the giro. To do this, it must ensure that its liabilities are preferred above those of banks, so it requires convertibility of private debts into state money, over which the state maintains a monopoly.” As I will show below, this is a long process which first requires the development of a central bank and a mono-reserve system. If state fiat money can replace commodity money, then the state can increase its purchasing power by printing money without fear of merchants changing the exchange ratio between state money and giro money or goods. The state gives impetus to the development of banks by accepting their liabilities, but places restrictions on banks by forcing convertibility. Convertibility is not merely a method used to control the issue of notes, but is a method of ensuring the value of state money. The state also restricts banks by eliminating private notes so that only central bank notes will act as currency. This facilitates movement to a system based on demand deposits against which reserves are held at the central bank. Thus, the state is gradually transformed from an impoverished borrower which must rely on private guarantees of its credit worthiness in order to issue debt to the modern capitalist state whose debts are the most preferred and whose guarantees lend credit worthiness to private debt issues. Again, this transition is partially based on the creation of patriotism, the development of republican forms of government, and other (primarily) noneconomic matters which are beyond the scope of this study. However, a brief study of the rise of modern banking in England and of the Bank of England will show how an attempt to solve the chronic financial problems of the state also helped to bring on modern banking. That is, the rise of the modern state and the rise of the modern financial system are inextricably related. Indeed, it would not be far from the truth to argue that the state forced modernization on the economic system (Heilbroner 1985, p. 88). I will first discuss the evolution of the financial system, and then turn to a discussion of the capitalist economic system, in general.

44

The development of modern banking: the case of England The Bank of England was founded in 1694 to provide loans to a Whig government at war with France and in dire financial straits.” The Bank of England was incorporated as a private bank with special privileges: it was created as the sole depository for state balances; it was the only firm incorporated with limited liability for

stockholders; in 1697, Parliament declared that no other bank would receive a State charter; and in 1708, Parliament rewarded

the Bank of England for taking up government bonds by prohibiting any other joint stock bank of more than six partners from issuing notes in England (Bagehot 1927, pp. 90-95; White 1984, p. 38).

Until 1826, there were three other categories of banks: private London banks which issued notes; private note issuing country banks; and joint stock banks which could not issue notes. The special status the Bank of England enjoyed as a note issuing joint stock bank gave it a tremendous advantage during this period. As a joint stock bank, it could raise a large sum of capital (1.2 million pounds) through subscription. In contrast, country banks which issued notes were chronically undercapitalized because they were limited to six partners (White 1984, p. 38). Furthermore, as a limited liability firm, the Bank of England’s owners could protect their fortunes - unlike private bankers and the owners of other joint stock banks, who placed their entire fortunes on the line. It is imperative to understand the importance of note issues. Banks did not begin as depositories of funds. Early banks engaged in three activities: providing loans to the government, changing coin, and facilitating circulation of bills of exchange (Bagehot 1927,

p. 77). Gradually, however, banks moved from the facilitation of bills of exchange to discounting bills by providing bank notes. ‘[N]o nation as yet has arrived at a great system of deposit banking

without going first through the preliminary stage of note issue...’. (ibid., p. 88)” In 1800, bank notes accounted for 50 per cent of M1 in England, while deposits accounted for only 10 per cent.’ Deposit banking arrived much later, after the public had become

accustomed to using bank liabilities as media of exchange, means

of payment, and stores of value (Cameron 1967, p. 8). Thus, by creating this special right of note issue, the state provided an advantage which would help the Bank of England to monopolize note issue in London. Although private banks could legally issue notes in London, their notes had been almost

45

completely displaced by Bank of England notes by the mid-eighteenth century (Bagehot 1927, p. 96). Since checking deposit banking had not yet appeared, this meant the Bank of England ‘became the bank in London’ (ibid., p. 97). Other London banks focused on foreign trade, relations with stock and bill brokers, and relations with the country banks. Between 1797 and 1819, the Bank of England was prohibited from redeeming its notes for bullion. The ‘Bank Restriction’, as it was called, essentially ensured that a run on the Bank of England could not occur. ‘A bank of issue, which need not pay its notes in cash, has a charmed life; it can lend what it wishes, and issue what it likes, with no fear of harm to itself, and with no substantial

check but its own inclination’ (Bagehot 1927, p. 107). (Of course, Bagehot exaggerates, since note issue is always constrained by the willingness of the public to accept notes.) Furthermore ‘since 1797 the public have always expected the Government to help the Bank if necessary’ (ibid., p. 108). Thus, the Great Restriction actually strengthened public confidence in the Bank, rather than reducing it as one might expect suspension of convertibility to do. Bank of England notes were accepted by the government in payment of taxes and customs duties and for subscriptions to government loans (Cameron 1967, p. 22). This helped to increase the desirability of the notes. Parliament also gave the Bank of England’s notes a special status when it allowed country banks to make their notes convertible into Bank of England notes in 1833 (White 1984, p. 39). In this way, Bank of England notes would become as desirable as commodity money since they would replace bullion as the banking system reserve. Later, mere deposits at the Bank of England would serve as the reserve for London banks, while deposits at London banks would serve as the reserve for country banks - as will be discussed below. In order to increase the circulation of Bank of England notes outside London, Parliament also periodically restricted note issue by country banks. For example, Parliament prohibited any bank other than the Bank of England from issuing one pound notes after 1775, and the five pound note was banned in 1777. These restrictions were later temporarily dropped, but the prohibition on the one pound note was restored in 1829 (Wood 1939, p. 38). In 1826, Parliament encouraged the Bank of England to open branches in cities outside London to increase the circulation of its notes (ibid., p. 39). By 1844, the quantity of Bank of England notes circulating in England had risen to equality with that of private

46

notes (ibid., p. 23). A study of the history of Lloyd’s Bank can serve as a useful example of the evolution of private banks during this period. Lloyd’s began as a country bank in 1765, primarily engaged in the iron trade~ (Sayers 1957, p. 5). A decade later, it had about 277 customers, most of whom were craftsmen or tradesmen. Its assets consisted of £45,000 in the form of IOUs, £13,000 of bills, notes, and cash on hand, and £27,000 on deposit in London banks. Its

liabilities consisted of £43,000 in deposits and £33,000 issued in the form of bank notes (ibid., p. 10). The borrowers were typically businessmen, but loans were also made to the gentry (ibid., p. 95). The bank was much more concerned with the length of the loan rather than with the purpose of the loan, preferring short term loans which could be liquidated quickly. The bank was quite

selective in its customers, even in choosing its depositors. That is,

it would refuse to open accounts for small customers (ibid., pp.

96-8). London had been the center of English foreign trade since medieval times, with the bill of exchange drawn on London acting

as the primary international means of payment.” Taxes had to be paid in London. London acted as the clearing house for country bankers (although country bankers in a town or small region would operate a local clearing exchange which would not require intervention by London). Thus, country bankers typically held deposits as a primary reserve in a London city bank. These could be drawn upon for clearing or if the need for cash arose. The country bank would also leave stocks and bonds in London as secondary reserves, against which overdrafts could be drawn (Sayers 1957, p. 109). These reserves in London varied greatly, and were frequently overdrawn. It was not uncommon for a country bank to have an overdraft outstanding equal to fully 10 per cent of the bank’s total balance sheet (ibid., p. 121). Thus, London banks became the depositories for country bank primary and secondary reserves because London was the financial center. London banks frequently had an arrangement by which they would pay 4 per cent interest on country bank deposits, and charge 5 per cent on overdrafts. The normal deposit of reserves required of Country banks was £10,000 - it was not determined by the size of the country bank’s balance sheet (Sayers 1957, p. 35). If country banks wished to hold more ‘liquidity’, they would send more stocks and bonds to their correspondent in London to be held as a secondary reserve against which overdrafts could be arranged

47

(Wood 1939, p. 21). London banks would use the primary reserves

to earn interest, and would use overdrafts as a way to influence the

behavior of country banks (Sayers 1957, p. 124). Country banks would issue notes payable on accounts in correspondent London banks. In this way, the clearing mechanism was enhanced, and the acceptability of notes was actually improved because country bank notes could then freely circulate outside their limited giro, since other country bank giros could always send the notes to London for clearing. In turn, London banks held Bank of England notes as their reserve. After 1833, as noted above, country bank notes were

convertible into Bank of England notes, which made the liabilities of the Bank of England even more desirable as reserves. Gradually, London banks (and even later, country banks) substituted deposits at the Bank of England for notes to be held as reserves (Wood 1939, p. 178). Thus, England developed a one reserve system: all reserves ultimately were held in the form of notes or deposits issued by the Bank of England. The country banks held reserves at London banks, the brokers held reserves at London banks or at the

Bank of England, and the London banks held reserves at the Bank of England. ‘No other bank holds any amount of substantial importance in its own till beyond what is wanted for daily purposes’ (Bagehot 1927, p. 28). The reserves of the Bank of England took the form of bullion or Bank of England notes. The Act of 1844 separated the Bank of England into two departments: the Note Issue Department and the Banking Department. The Note Issue Department issued notes on the basis of government debt and bullion held. Thus, finance was provided to the state when the Bank issued notes against government debt. The Act originally permitted the Bank to issue £14 million of notes against securities, but this was supplemented periodically by new authorizations. By 1903, the Bank was permitted to issue more than £18 million against government securities (Bagehot 1927, p. 24). The Banking Department could not issue notes, but held the deposits of the state and of banks and

brokers. It then purchased government securities, other securities, notes, and coin, operating as a profit seeking bank (Bagehot 1927, pp. 23-5). In 1795, government securities accounted for more than 77 per cent of total assets held by the Bank of England, but by 1869, government securities accounted for one-third of the assets of the Issue Department and for 31 per cent of the assets of the Banking Department (Cameron 1967, p. 20, and Bagehot 1927, pp. 48

24-5). Through the early nineteenth century, the dichotomy between country banks and London banks was strict. The Bank of England controlled note issue in London, while country banks dominated

note issue outside London.” Country bank notes were used mainly

for payment of wages and other small local payments. Until 1826, very few Bank of England notes circulated outside London, so virtually all notes in circulation in the country were those of the country banks. However, the Act of 1844 placed a ceiling on the quantity of private issues, and deposits gradually replaced notes. Furthermore, Bank of England notes began to circulate more widely in the country, reaching SO per cent of total issues by 1844. The Bank of England (wisely) accepted country bank notes (normally drawn on London), so it could remove them from circulation by presenting them for payment (Wood 1939, pp. 17-23). In this way, the Bank of England could enhance its own

circulation. After 1830 the London banks aggressively pursued relations with country banks. Similarly, the Bank of England began to expand its activities outside London during the same period. The Acts of 1826, 1833, and 1844 gave impetus to this by centralizing note issue

and by permitting unlimited joint stock banking.” The Act of 1826 permitted unlimited joint stock banking outside the London area, while the Act of 1833 allowed joint stock banking in London provided that the bank did not issue notes. The Act of 1844 provided for the gradual suppression of note issue by country banks, with note issue to be completely eliminated within twelve

years.” Until the 1820s, it was generally believed that the rate of growth of country bank note issue was controlled by the Bank of England. This control was maintained through first, convertibility of country bank notes into Bank of England notes; and second, balance of

payment flows.” If country banks wanted to maintain the value of their notes, it was widely believed that they would have to agree to redeem them on demand for Bank of England notes. This was believed to require some ratio of country bank notes to Bank of England notes. If country banks issued too many notes, the balance of payments would turn in favor of London, with the country bank notes flowing to London where they would be taken out of

circulation.”

However, by the 1820s it was argued that the rate of growth of country bank notes exceeded the rate of growth of Bank of

49

England notes. This led to the belief that the country banks were thwarting the Bank of England’s attempt at sound monetary policy (Sayers 1957, p. 140). The series of laws restricting bank note issue were

designed

to halt this practice.

After

1826,

the Bank of

England would favor only non-issuing country banks in its discounting policy. That is, it would refuse to discount paper submitted by country banks which issued bank notes. It pressured country banks to call in their own notes and replace them with Bank of England notes. Many country bankers were able to resist this pressure - especially in areas where the population preferred the country bank notes. Banks could also get around restrictions on note issue by creating very short term bills - and hostility in some areas toward the Bank of England would ensure local preference for these over Bank of England notes (ibid., pp. 151-2). These policies at first worked in the favor of the Bank of England, and its notes displaced some country bank notes. However, the development of the bank check was probably far more important in eliminating private country bank notes. While the Acts placed limits on the quantity of notes each bank could issue, actual issues quickly fell below the ceiling (Sayers 1957, p. 151). Thus, restrictions on note issue probably had little effect on

most banks.” On the other hand, the laws favoring joint stock banking had important and lasting effects. First, London banks declined in importance because the joint stock banks in the country would establish correspondence relations, so that the London banks became less important. Mergers of joint stock banks led to the development of national banking networks, which would include a bank in each major city, as well as a position on the London Clearing Bank, by the late nineteenth century (Sayers 1957, p. 248). Second, the limited liability of joint stock banking (permitted by the Acts of 1858 and 1862) led to a movement to force such banks to publish balance sheets (Sayers 1957, p. 224). Even by the midnineteenth century, there were no established rules (or even rules of thumb) regarding the level of reserves or leverage ratios to be maintained by a prudent bank: as late as 1888, the ratio of cash plus short term money market loans to deposits plus notes varied from 4.8 per cent to 32 per cent among banks (ibid., p. 178). Indeed, this ratio would vary from year to year for an individual bank by as much, or more, than it did for banks on average.” Even after published records became required, they meant very little until the twentieth century since banks would engage in ‘window

50

dressing’ just before the publication date.“ However, the movement to force publication of balance sheets would gradually make bank practices more uniform regarding such matters as reserve ratios. Third, the evolution of joint stock banks and the centralization of decision making led to rules of thumb and rules of behavior which also gradually established more uniform banking practices, such as those concerning the quantity of primary and secondary reserves which should be held against bank liabilities. Banks also established more sensible pricing policies. Interest rates on most country bank assets and liabilities had been set by collusion and rough rules of thumb. Even as late as the end of the nineteenth century, price competition was rare (Sayers 1957, p. 160). Integration of financial markets and competition from the growing joint stock banks, however, forced country banks after 1850 to compete with London interest rates. This would gradually increase the control of the Bank of England over financial conditions throughout the nation (ibid., pp. 162-5). As the single repository of reserves (and because its own notes counted as reserves), the Bank of England could cause tight money by raising the price of reserves. That is, it could call in advances made

to London

banks or brokers, or it could raise the interest

rate required in its discount of bills, or it could simply refuse to discount bills altogether. When faced with tight money, London banks would call in overdrafts and force the country banks to sell consols or stocks held by their London correspondent. London banks would help to sell bills for their country bank clients, but normally refused to ‘rediscount’ bills themselves - it was the policy of London banks to avoid putting their own name on a bill to be sold. This tended to cause long term interest rates to rise as the prices of bonds fell (Sayers 1957, pp. 125-7). The Bank of England used open market sales or other methods to institute tight money primarily to increase its reserves, and not primarily to influence credit conditions (Wood 1939, p. 5). When the Bank of England was faced with an external drain of specie (for example, when England ran a trade deficit), it would institute tight money to attract bullion inflows. However, an external drain would frequently lead to an internal drain: bank customers would notice a drain on bank reserves, become worried over the stability of banks, and try to obtain loans (discount bills) before credit was cut off, or would try to withdraw deposits or to redeem notes. This,

of course, would lead to a further drain on the reserves of the

Bank of England. A panic would result whenever the Bank of si)

England acted like it might not provide the reserves needed by the private banks. Indeed, the Bank of England would frequently try to sell securities

in order

to replenish its own

reserves

during

an

internal drain (Bagehot 1927, p. 65). This was the case because the Bank of England frequently saw itself not as a central bank, but as

a profit seeking bank.”

When banks were small and local, a run on deposits or notes could be averted by the declaration of a local wealthy or respected family that the bank’s books were sound (Sayers 1957, p. 208). Local merchants could help if they would agree to continue to accept the bank’s notes as means of payment. On occasion, other banks in the region would step in to save a failing bank. As relations between London and the country banks developed, crises which began in London would spread to the country. If the Bank of England restricted loans, the London banks would often freeze the funds of country banks and call in overdrafts. The country banks would be forced to follow suit. Each such financial crisis would be followed by an unwillingness among banks to accept long term commitments and to prefer discount of bills to overdrafts by

customers” (ibid., p. 212).

Thus, evolution of the banking system to an integrated system of joint stock banks where decision-making was centralized in London, and to a one reserve system which placed considerable power in the hands of the Bank of England, occurred during the nineteenth century through the end of World War I. The Bank of England was originally founded to provide state finance, and was provided with special privileges as a reward. These special privileges made its notes circulate in London to the exclusion of all privately issued notes. Country banks established relations with London banks to facilitate note (and, later, deposit) clearing and to provide services as depositories of reserves. The special status of Bank of England notes, the special relation of the Bank of England with the state,

and legislation giving special rights to the Bank of England all enabled it to become the ultimate depository of reserves for a financial system which pyramided reserves. This special status of the Bank of England allowed its notes to become fiat money and to replace commodity money, primarily because privately issued giro money was made convertible into Bank of England notes. This brought the state into the giro, since State finance was provided by the Bank of England, which purchased government securities. Because London was the center: for international finance, the nominal sterling became the world

a2

currency for use in international trade. Bullionism was displaced by Free Trade because it was no longer needed as a means to provide State finance, and, indeed, would hinder the growth of international

trade with London as the center. The evolution of banks to a centralized financial system based on London increased the complexity of financial relations, increased leverage ratios (through increased pyramiding of reserves), made banks more similar, and led to an economy which was more integrated so that crises would quickly spread. Thus, the actions of the Bank of England would have more impact on banks far from the center, but integration of the financial system also increased the responsibility placed on the Bank of England to prevent crises. Before proceeding, I will briefly summarize the points of the previous three sections. As the economic system evolved to the modern capitalist system, the financial system also underwent great changes. In the early premodern period, most production took place outside the market, so that the monetized sector was small. Goods circulated on the basis of bills of exchange and commodity money, although other forms of giro money were also used in specific regions. Except in the case of city republics, the state was very limited in its ability to issue debt so that fiat money was precluded. Gradually, however, private property rights were extended and the importance of the market increased. This required the development of sophisticated credit facilities - primarily based on note issue by private banks. In order to extend the size of the giros, clearing facilities had to be created. This led to the development of a system which concentrated reserves to facilitate clearing. The earliest central banks (such as the Swedish Riksbank and the Bank of England) were explicitly created to provide state finance (Goodhart 1989b, p. 88). By generating special monopoly advantages,

the central

bank’s

notes

became

fiat money

which

could provide purchasing power to the state. At the same time, a monoreserve system was created - partially due to the advantages granted to the central bank by the state, but also because the financial system already pyramided reserves. As Goodhart (1989b) emphasizes,

this development

was

not intentional,

but was

an

unforeseen and evolutionary development. The central bank only very slowly came to recognize that it could exercise some control over the financial system, because its liabilities had become the primary reserve as a side-effect of various state policies which were designed merely to enhance state purchasing power by supporting

Sp)

the development of fiat money. This transition to the modern financial system which is based on credit money and deposit banking, and in which state purchasing power rests on fiat money, was essential for the development of the modern capitalist state and economic system. I will now turn to a discussion of the reasons why a capitalist system must be based on this sort of financial system.

Money and the logic of capitalism As discussed above, money has existed as a unit of account since

the development of private property. The development of markets long predates the development of capitalism, but the monetized sectors of precapitalist societies were small and frequently required government intervention for development and maintenance. In such societies, credit was primarily responsible for financing inventories in the process of transport, consumer expenditures, and government spending. A wide variety of credit instruments was invented to satisfy these purposes. Foreign trade generally was associated with the most advanced credit forms, primarily due to the impersonal (and, therefore, risky) nature of such trade.

Credit money was the first form of money, created as a unit of account. Markets were later developed, which led to the use of money as a medium of exchange. As discussed in Chapter 1, the money of account came to be measured by commodity money, first in the form of barley, later in the form of known weights of metals, and finally in the form of stamped coins. Once markets had developed and the public had become accustomed to the use of coins, the government could increase its purchasing power through seignorage. In most cases, both government-issued commodity money and privately-issued credit money existed side-by-side in precapitalist societies. Fiat money is, for the most part, an innovation of modern capitalism. | It is interesting to note that once credit money has led to the development of commodity money, credit money ceases to be essential in precapitalist societies. For example, Athenian society and the Roman empire functioned almost entirely on the basis of commodity money, and commodity money was the primary means of payment used in the Middle Ages. Credit money seems to have been used when the supply of commodity money was insufficient, or when credit actually reduced transactions costs. For example,

54

after the fall of the Roman empire, the supply of commodity money generally fell far short of the demand, so various credit instruments were developed to fill the gap. In the case of foreign trade, credit instruments like the bill of exchange eliminated unnecessary carrying costs which would be associated with shipping bullion. In any case, credit in such societies augments the supply of

commodity money.”

The situtation is far different with the development of capitalism, where the logic of the system requires that credit money is the ‘normal’ form of money and that commodity money is the aberrant case. Indeed, it cannot be a coincidence that the modern financial

system developed with the industrial revolution (Goldsmith 1987, p. 4). In a capitalist economy, production occurs not to satisfy needs or to produce goods which can be bartered, but to produce goods for sale in markets to realize money. The social purpose of credit is to provide purchasing power to the capitalist so he may buy the goods and services needed today to produce the goods and services which will be sold tomorrow. If this sale takes place at the expected price, the capitalist realizes sufficient money to repay his debt and retire the credit money. This distinction between a capitalist economy and a noncapitalist economy was recognized by Keynes, who argued that in a capitalist economy, the object of production is to ‘result in a larger net sum of money at the end of the accounting period’ (Keynes 1979, p. 66). He defined a barter economy ‘as one in which the factors of production are rewarded by dividing up in agreed proportions the actual output of their co-operative efforts’ (ibid., p. 66). This economy does not preclude the use of money ‘for transitory convenience’, so ‘it might perhaps be better to call it a real-wage economy, or a co-operative economy as distinct from

an

entrepreneur

economy,

economy’

however,

money

(ibid., p. 67). In the entrepreneur becomes

useful lubricant for exchange.

essential

and not merely a

Production is a continuing process, so the capitalist develops a relation with a banker who will provide credit to fund working capital expenses (including the wage bill, raw materials costs, and trade credit granted to buyers). That is, the short term credit is continually rolled-over to fund working capital expenses. In the early stages of capitalist development, working capital expenses account for the vast majority of expenses, while investment in fixed capital is a very small percentage of national income. During the early period, short term credit advanced by banks frees the

55

capitalist’s profits for the finance of investment and accumulation of (fixed) capital. Thus, credit in early capitalism can be characterized as financing working capital expenses, which allows production to proceed so that profits can be generated to finance capital accumulation (Cameron 1967, pp. 36-9, 52). In the process of financing working capital expenses, Income 1s generated and a market is created for the goods which will be produced. For example, the working capital expenditures will be received by workers, who will spend their income on consumption goods. Growth of this market requires that working capital expenses continually grow. This, in turn, means that the revolving fund of short term credits must continually expand. Accumulation of capital, which increases the capacity to produce goods and services, requires a growing market - otherwise, produced goods cannot be sold at a price sufficient to meet payments on short term credits, to generate profits, and to finance capital accumulation. As Kalecki (1954) has shown, it is the production of investment goods and goods which will be consumed by capitalists that generates aggregate profits. In the early stages of capitalism, short term credit allowed rising expenditures on working capital used to produce such goods. Thus, growth of short term credit generated - aggregate profits, which funded investment and accumulation of capital. At an individual level, the capitalist borrows a sum of money to produce goods which he hopes to sell for a greater sum. In the aggregate, this will be possible only if investment goods and capitalist consumption goods are being produced. However, as I just discussed, investment raises productive capacity so it can be undertaken only if market growth is expected. That is, because profits require investment, and because investment requires expectation of growth, the logic of the capitalist system rests upon

growth and capital accumulation.*

Heilbroner (1985) has also examined the internal logic of capitalism and reached similar conclusions. He argues that the drive to amass capital is the single most important element of capitalism. The form in which the social surplus is accumulated under capitalism is profit, accumulated individually by the owners of capital. Accumulation of wealth by capitalists is not an end in itself, but is the means to further accumulation. As will be discussed in Chapter 4 below, in Marx’s scheme, money enters the

production process as capital to generate more capital in money form (m-c-m’). This continual ‘expansive metamorphosis of capital’ is the essential logic of capitalism (Heilbroner 1985, p. 36).

56

A system which relies upon growth and short term credit to fund working capital expenses cannot rely upon commodity money. If credit were to be constrained by the quantity of commodity money available, there would be no guarantee that this quantity would expand to meet the needs of an economic system which relies upon growth. This is why the capitalist system requires development and

extension of the credit system. However, the logic of a capitalist system also requires that purchasing power is not merely extended willy-nilly to anyone who desires credit. Ideally, credit would be extended to capitalists who will increase productive capacity and increase output of those goods which can be sold at a price sufficient to cover expenses and leave profits to fund capital accumulation. Credit creation can be used to funnel resources into the production of investment goods rather than consumption goods so as to generate a surplus which can be accumulated

(Cameron

1967, p. 292). There is, of course,

no mechanism in a capitalist system to guarantee this result. Individual profit seeking behavior of those who advance credit, however, ensures that loans will normally only be made to those who are extremely likely to repay them. Furthermore, individual firms are forced to produce commodities for sale to obtain money,

since they cannot produce money, but are forced to borrow it. Thus, profit seeking behavior necessarily rations credit, and repayment of credit necessitates orientation toward the market. Financial institutions try to develop close personal relations with borrowers and then apply various rules of thumb (such as collateral requirements, market share of the borrower, credit history, and so

on) when deciding whether to make a loan. The financial institution operates as the expert judge in determining who is likely to repay loans. (Ong 1983, p. 45) Innovations and successful experimentation have gradually expanded the types of borrowers deemed credit worthy. Again, however, there is no guarantee that credit rationing will necessarily generate the right amount of credit and place it in the correct hands. Thus, failure and bankruptcy is a necessary, if infrequent, outcome of a privately motivated capitalist system.

As

fixed

capital

became

increasingly

important

with

the

development of the factory system and as capital became more complex and expensive, investment could not easily be funded out of retained profits. It must be remembered that even these retained profits represented working capital expenses in the investment goods and capitalist consumption goods sectors - funded on the

aby

basis of short term credits. (Again, short term credits fund working capital expenses, and expenditures on investment goods and capitalist consumption goods generate profits.) However, gradually firms shifted from ‘internal’ finance of investment (retained profits) to ‘external’ finance (borrowing), in which short term credits would continue to be used to fund production of the investment goods,

and long term credits would be used to finance a position in the finished capital goods. In the aggregate, it is the short term credit (which finances working capital expenses) which generates the income used to provide the long term credit, which in turn finances the position in the capital asset. Provision of long term credit through bank ‘intermediation’ is necessarily a later development coming after the public has become accustomed to holding and using the short term liabilities of banks - that is, after the ‘banking habit’ has become well developed. As Bagehot argues, and as historical evidence generally shows, banks begin by issuing notes and only later do deposits become important, after the public has experience with ‘bank intermediation’. At this stage, banks can finally ‘mobilize capital’ by accepting deposits as a short term liability and by making long term loans to enable firms to finance capital purchases. Thus, the true order of events shows that orthodoxy clearly has reversed the process through which investment is funded. Banks do not begin as intermediaries which accept deposits of ‘savers’ and then make loans to ‘investors’, for this would assume that the public has already developed the ‘banking habit’. This habit is the end result of public experience with short term bank liabilities, which have been created as banks extend short term credit to finance working capital expenses (Cameron 1967, p. 295). After the public becomes used to receiving bank notes as payment for wages and to using them for consumption expenditures, it will gradually accept bank liabilities as a form of intermediation. Just as the logic of a capitalist system requires that it cannot be constrained by commodity money, its logic also requires that it cannot be constrained by saving. The production of an investment good is financed through short term credits, which generate profit income. In the absence of long term credit, positions in investment goods are financed by this profit income directly. However, the direction of causation is clear: it is the spending on investment goods - financed on the basis of short term credits - which generates the profit income which can be ‘saved’ in the form of final sales of investment goods. The transition to external finance

58

of positions in investment goods does not change the logic of this system: it is still the production of the investment good which generates profit income. However, in this case, profits need not directly fund sales of finished investment goods since long term financial assets can be sold to those receiving profits. Alternatively, such surplus income can be used to buy short term financial assets issued by banks or other financial institutions, which take up the

long term instruments.”

In a well functioning capitalist economy, most money will take the form of credit, as short term loans fund working capital expenses on the basis of note issues or, later, demand

deposits.

The value of the credit money is maintained by the value of the loans. Occasional default and subsequent failure of a bank would, of course, make the public wary of banks and impede the development of the banking habit. In order to placate the public, bank liabilities would often (but not always) be made convertible into some other asset, which would be held in reserve to meet the

demand for redemption. Of course, development of the banking habit and normal operation of the financial system would require that bank liabilities were rarely redeemed for the reserve asset. At first, the reserve asset might consist of gold coins, to which the public had already become accustomed. However, a credit system based on a gold reserve is unworkable - the reserves of an individual bank can never be sufficient to meet a run on a bank. When banks were small and regionalized, a bank run would be stopped not by its gold reserves, but by guarantees of creditworthiness provided by trusted individuals. Thus, a gold reserve would suffice only if the public did not exercise its right to redeem bank liabilities.

A financial system with large, national banks could not rely on individual gold reserves. The first step toward an integrated banki:g system would require an integrated system of reserves. In England, this took the form of a system of country bankers which relied upon London banks to provide reserves as needed. While the London banks could mobilize reserves to meet a local run on reserves, they were still helpless to stop a general run. The development of the Bank of England as the mono-supplier of reserves could potentially provide sufficient reserves (Bank of England notes) to stop any run. However, as long as It viewed itself as a private bank, it would not be willing to do so. Only a central bank which is willing to issue its liabilities without taking into account profitability can fulfill the lender of last resort

bY

function. The logic of the capitalist system requires that credit normally functions as money and that reversion to commodity money 1s an irrational aberration and rare occurence. A run to commodity money causes debt repudiation and destroys the value of accumulated capital assets. Orthodoxy would have us believe that the normal functioning of a capitalist economy requires that spending is related to M1, defined as currency and demand deposits, where the quantity of demand deposits is strictly constrained by reserves provided by deposits of currency or by the central bank. According to orthodoxy, banks learned that they needed to hold only a fraction of a deposit on reserve, and could loan the rest. That is, the logic of the orthodox system can be stated as deposits make reserves, which make loans. In reality, the existence

of demand

deposits, and the relation

between demand deposits and reserves are developments which came after the public had become accustomed to the use of a wide variety of private liabilities. Banks first operated by loaning their own liabilities (notes), which were gradually accepted by the public and later deposited in banks. The logic of a capitalist system is that loans make deposits because the system cannot be constrained by deposits of commodity money. Reserves exist because the system occasionally breaks down so that credit cannot function as money - and not because banks discovered they could loan a portion of commodity money deposited with them. The existing system represents the result of evolution of the financial system away from a wide variety of privately issued credit instruments to a system based on demand deposits backed by the liabilities of a central bank. Indeed, it is instructive to note a steady decline in the velocity of M1 (national income divided by M1 - bank notes plus deposits plus currency) in England from 4.0 to S.2 in 1688-1775 to only 1.6 by the last quarter of the nineteenth century (Cameron 1967, p. 42). This clearly indicates a transition from a system with a wide variety of credit instruments to a system based on deposits of commercial banks, and contradicts the orthodox story in which banks gradually discover the existence of a ‘deposit multiplier’. Credit money is the normal case, and current institutional relations, which may obscure

this fact, are the result of the evolution of the financial system. Finally, as is argued by Heilbroner (1985), the development of capitalism required the gradual recognition by the state of property rights and the gradual exclusion of the state from markets. In 60

ancient Greece

or Rome,

coercion was sufficient to enable the

State to grab part of the surplus. As discussed above, in the premodern period, Bullionist policies and forced loans (and sales of monopoly charters) were gradually abandoned as the modern state emerged with the development of capitalism. However, capitalism requires a state - not only to provide defence and to fulfill the other functions enumerated by Adam Smith and Milton Friedman, but also to provide those necessary inputs which cannot be profitably provided by markets (for example, an educated workforce, transportation systems, communication systems, and so on). This requires, of course, vast revenue-generating facilities. The expansion of the economy increased the size of the tax base, while the development of republican forms of government legitimized the transfer of purchasing power to the state. However, this source of revenue cannot be counted upon as the tax base or willingness to pay taxes might prove insufficient to meet necessary expenditures at certain points of time. Thus, just as firms need a source of credit, the government also needs one. The development of a central bank could provide the government with the means to expand purchasing power essentially at will. As the liabilities of the central bank become the reserve, these liabilities also take the

form of fiat money. Thus, the government can sell bonds to the central bank to increase purchasing power. Just as short term credits issued to finance expansion of investment must increase ‘saving’, the deficit spending of the government also must raise ‘saving’ by an equal amount. At this point, the central bank can sell bonds to be held by ‘savers’ to provide long term finance to the government. This, of course, is not necessary since the central

bank might choose to hold the bonds for the long term. Deficit spending by the government allows accumulation of a portion of society’s surplus by the government. Except at full employment, deficit spending not only increases the nominal value of the surplus, but also generates expansion of the social surplus in real terms. At the same time, government deficit spending generates claims on the government which are held by ‘savers’ or by the central bank. Thus, deficit spending has real and financial impacts on the economy. For this reason, the government can play a role in stabilizing the economy, in addition to its role of providing necessary inputs. The development of the central bank created the possibility of stabilizing the financial system through provision of reserves. At the same time, the creation of the central bank also enhanced the state’s ability to use deficit spending, which

61

provides another stabilization tool.

The third world and endogenous money The knowledge gained from a study of the evolution of financial relations which occurred during the transition from nonmonetary economies io modern capitalist economies can be applied to a study of money and banking in the less developed countries (LDCs) today. I will merely provide a preliminary analysis in this digression which I hope will point the way for further research in this field. The LDC today faces a particular problem which was not faced by any country during the Middle Ages - the presence of a highly developed banking system in a neighboring country.” Indeed, banks in LDCs have typically been set up by a bank from a developed country (DC) (Thirlwall 1974, p. 111). In many cases, the banks were initially created by a colonial power, and forced to maintain 100 per cent reserves in the form of the DC’s currency against any LDC currency issued (ibid., p. 112). In this case, the LDC would only be able to issue currency if it ran a trade surplus with the colonial power (or borrowed DC currency). Clearly, the LDC currency in this case would be commodity money, and not fiat money. Thus, “Bullionist’ policies would be required to provide state finance. This would be particularly true in any state in which taxing ability were not well developed. In the LDC, market activities would be limited, so the tax base would be small. Even if the LDC were to

break relations with the colonial power, it would be unable to convert its currency to fiat money unless it could become part of the domestic giro. However, the existence of a preferred currency would make it difficult for the state currency to enter the giro as fiat money. For example, if the fiat money of the former colonial power is still preferred, the LDC currency will remain commodity money. In this case, debasement will not be possible, because, as discussed above, minting more state money will just lead to price inflation (in terms of state money - not in terms of the giro money). In this case, important prices (particularly future prices) will be quoted in terms of the giro money (currency of the colonial power) - either overtly or covertly (through the exchange rate

which will prevail at the time of payment).*

In this way, the domestic value of the LDC 62

currency will be

determined relative to the value of the DC currency. Bullionist policies (trade surpluses or high interest rates) can be used to try to increase the domestic value of the currency and increase the spending power of the state by increasing the taxable (DC currency) base. However, high interest rates may slow economic development and may not be sufficient to attract financial flows. If wealth owners in the LDC prefer the debts of the DC, ‘capital’

will flow out of the LDC even in the presence of high interest

rates. That is, even at high interest rates, agents in the LDC will

not be able to issue debt to finance spending because the liabilities of the DC are preferable. In this case, the money supply of the LDC cannot be endogenously increased because high ‘liquidity preference’ (that is, preference for DC debts) prevents creation of LDC money. An LDC in this case can obtain finance only by running trade surpluses or by borrowing from the DC. As the world financial system has become more integrated, each country must generate those assets desired in global financial markets, or it will be unable to generate finance and retain capital. For this reason, it is not at all clear that financial development in the LDCs can proceed along the path outlined in the sections above. As discussed above, banks began by issuing notes. However, the typical LDC government has outlawed note issue in an attempt to reserve the right of note issue to the government. As Cameron argues, ‘historically local banks of issue have played an important role not only in habituating the populace to the use of financial instruments and institutions, but also in the development of small-scale local industry and agriculture’ (Cameron 1967, p. 319). It may be much more difficult to develop the ‘banking habit’ through the use of deposit banking. Furthermore, deposit banking is a characteristic of an integrated banking system and has generally been based on larger banks. As I have discussed above, credit rationing by banks is the norm. The disadvantage of a system based on deposit banking rather than on small local banks may be that small borrowers cannot receive credit. This would tend to hinder the development of small scale industry and agriculture. In fact, this does seem to be a characteristic typical of the LDC. The obstacle to development in the LDC is neither ‘lack of saving’ nor inflation caused by ‘too much money’. The LDC needs finance, which requires the development of fiat money, trade surpluses, or foreign loans. Export led growth is possible because trade surpluses increase commodity money and provide finance. 63

Foreign loans also provide commodity money and finance. Finally, the development of fiat money would allow domestically generated growth. As we have seen, the development of the Bank of England led to the creation of government fiat money. LDCs today face a monumental barrier which England did not face because the debts issued by DCs are readily available to the wealth-holders of LDCs. ‘Lack of domestic saving’ can only be rectified if investment can find finance so that saving can increase. This requires either domestic fiat money, domestic credit money, or foreign loans. Inflation occurs because prices are stated in terms of commodity money and can only be reduced through a transition to fiat money as the unit of account.

Conclusions In the orthodox story, barter is replaced by exchange based on commodity money. Banking is developed as some agents offer to act as depositories and then discover that they can loan part of the deposits since a small reserve is sufficient to meet normal withdrawals. Thus, banks evolve as intermediaries between ‘savers’

and ‘investors’ and are constrained by prudent reserve ratios. I have argued, however, that this story conflicts with the historical evidence and is inconsistent with the logic of capitalism. Credit money and commodity money have existed side-by-side in most economies. Banks break free of reserve constraints as soon as they become members of a giro, in which reserves are necessary only if the giro money might leave the giro. It is only during a panic that giro money loses its status as money so that reserves become necessary within the giro. Furthermore, the quantity of reserves held by banks has never been fixed - it varied among banks and over the cycle - and never consisted solely of commodity money. For example, in England, banks would increase their secondary reserves (bills, stocks) rather than their primary reserves to increase liquidity. Investment was never constrained by saving or even by commodity money. The persistent shortage of commodity money led to the creation of credit money and to private giros to provide finance. Thus, money was always endogenously created whenever one agent was willing to spend in excess of income and another was willing to accept the debt of the deficit spender. In fact, the logic of the capitalist system requires that credit grow over time to

64

fund expanding working capital expenses and generate profit income which can be used to fund capital accumulation. Credit money is the key element which frees investment from saving and which makes capital accumulation possible. The stage of development of the banking system is important, particularly for state finance. The state plays a major role in the evolution of the banking system, in increasing the size of the bank giro, and in the suppression of alternative giros. The central bank was created to provide state finance, and its special privileges led to its position of dominance over private banks. The development of the central bank led to the integration of a financial system and to

the

development

of true

state

fiat

money.

However,

this

integration also increased the fragility of the financial system by increasing the tendency for the spread of financial crises throughout the system. Thus, the role of the central bank as a lender of last resort becomes necessary to ensure that bank money remains a substitute for state money. Finally, enhancement of state

purchasing power through the development of fiat money also creates the conditions required for counter-cyclical deficit spending to maintain capital accumulation in the face of declining private spending. When I argue that the money supply was always endogenously determined, however, I do not mean to argue that credit is not quantity constrained. Obviously, the set of economic agents which can issue debt that is deemed acceptable has always been

restricted.” Agents in LDCs are quantity constrained: only certain agents and certain activities can obtain finance from DCs. Furthermore, even agents in the LDC may prefer the debts of

agents in the DC.* Thus, ‘capital’ flows out of the least developed areas because the debts of agents in these areas are not deemed liquid (relative to those of the developed areas - and it is the existence of preferred debts which increases the illiquidity of the debts of the less developed areas). In this case, it is not merely the level of development of the banking system in the LDC which is the problem, but the existence of a developed area with preferred debtors.

65

Notes 1. Chick (1986) has presented a history of the evolution of banking and tied this to economic development and growth. Briefly, banking has evolved through five stages. In stage 1, bank liabilities are not widely used as means of payment, and banks are small and isolated and act as intermediaries

between

savers and

investors,

so that ‘deposits

make loans’. In stage 2, clearing arrangements are developed, which allow for a ‘deposit multiplier’, and bank liabilities circulate as media of exchange,

so that ‘reserves

make

loans’. In stage 3, interbank

lending develops so that individual banks can ‘create money’ through deposit expansion by increasing reflux. In stage 4, central banks operate as lenders of last resort and try to stabilize interest rates, so that banks are no longer reserve constrained and, in fact, ‘loans make

reserves’. In stage 5, banks use liability management to ensure that they can meet all credit-worthy demand for loans, and they actively pursue expansion of their balance sheets. After stage 2 or 3, the central bank can’t control the supply of money. The analysis of this chapter will focus on the evolution of the financial system rather than of banking. The conclusions reached will diverge somewhat from Chick’s. 2. Chick and Dow (1988) have extended the analysis of Chick (1986) and applied it to the study of economic development. They begin with a ‘center’ (or developed country - DC) and a ‘periphery’ (or less developed country - LDC). In stage 1, banks in the DC or center may funnel savings to areas, outside the area where saving occurred, with investment opportunities - so that the saving constraint is not necessarily operative at a local level. In stage 2, the banks in the DC or center are much more able to expand because their liabilities are preferred over those of the banks of the LDC or periphery, which have trouble retaining reserves. In stage 3, regions with shortages of reserves can borrow from other regions, which frees banks from local reserve

constraints.

In stage 4, transfers between

regions

are

no

longer necessary because the central bank operates as a supplier of reserves. Finally, stage 5 has been reached in some DCs, where banks

use liability management so that reserves and the money supply are endogenously determined. Banks in such DCs are able to attract capital flows from LDCs by gradually raising deposit interest rates. This contributes to the retardation of the development of the LDCs, particularly as loan rates of interest rise to compensate for rising deposit rates. 3.

Thomas Tooke formulated what he called ‘the law of reflux’, which

he described: ‘This law operates in bringing back to the issuing banks the amount of their notes, whatever it may be, that is not wanted for the purposes which they are required to serve’ (Tooke 1848, p. 185). Money and banking textbooks term the loss of deposits a ‘clearing

66

drain’, so that recapturing these deposits would be called ‘a reverse clearing drain’. Tobin has called this the ‘deposit-retention function’ (Moore 1988b, p. 53). Rather than adopting either of these terminologies, I will follow Tooke and call this reflux. Tooke seems

to have only had one form of reflux (note redemption) in mind, but I will use the term more broadly. See Munro (1979, pp. 191-5). See Ahmed (1989) for an analysis of modern Islamic banking. The prohibition of interest is not a serious impediment to the development of a variety of financial instruments. See Goldsmith (1987, pp. 1-14). [t must be admitted that evidence of credit arrangements in ancient Roman and Greek societies is sketchy at best. Bottomry loans were made in the maritime trade, with the ship pledged as collateral - the loan would be cancelled if the ship failed to reach its destination. It has been estimated that total bullion reserves in all of Europe amounted to 2000 tons in the period 1325-50, but this fell to only 1000 tons during the period 1450-75 (Day 1987, p. 60). 10. Usury laws were apparently easy to evade. McIntosh reports that there are no recorded court cases involving usury during the period 1383-1600 in the Havering courts. (McIntosh 1988, p. 566) f 1: Popularity of bills of exchange appears to be related to: limitation of liability (compared to a profit/loss compagnia contract); and to the facility of making international payments without requiring shipment of coins (Munro 1979, p. 172). Day (1987, p. 144) argues that bills of exchange were primarily used for foreign trade where such credit arrangements were more efficient. 12) See Goldsmith (1987, pp. 159-69). . The Banco del Giro was established in Venice in 1619 to finance

government spending by buying the government’s bonds and issuing liabilities. These liabilities were declared legal tender, which no one could refuse in payment (Day 1987, p. 153). In Genoa, the government debt was bought by private creditors, who gradually gained substantial control over the government, leading to the development of‘an autonomous organization of state creditors more powerful, respected and durable than the state itself (Day 1987, p. 156). 14. It y interesting to note that government debts were originally only liquid if the government obtained a private guarantee. Today, of course, private debts are liquid if backed by a government guarantee. This transition will be discussed subsequently. 158 Death of the king provided another way out, although his heirs were sometimes held liable because the original creditors required such a stipulation in the contract (no doubt experience taught them that such was necessary) (Ehrenberg n.d., p. 35).

67

16. Powerful kings (such as Louis XI of France) could use forced loans, primarily on the wealthy, to meet extraordinary expenses (Ehrenberg n.d., p. 61). Forced loans rarely carried interest and even the principal often remained unpaid. 17: These bills would then be redeemed after finished woolen goods were sold at fairs in Flanders, and the proceeds loaned to merchants (mercers) in return for bills of exchange drawn on London. The merchants would ship goods to London and retire the bills of exchange after selling the goods at English fairs (Munro (979, °p:

195);

18. ae was again restricted in the 1530s on the belief that importers operating in England exported bullion rather than goods. The Mayor of London,

Richard Gresham,

intervened

on behalf of merchants,

arguing that the cloth trade was disrupted because the merchants required ‘exchaunges an rechaunges’ for their trade, and King Henry VIII was forced to abandon the restrictions (Munro 1979, p. 209). 10% For example, the Medici banks kept their business records in a separate unit of account (Goldsmith 1987, p. 157). Also see Day (1987, pp. 152-3). 20. Foley discusses the German hyperinflation in much the same way. ‘Prices were set in terms of gold marks or pounds or guilder, currencies which retained a close relation to gold values, and paper mark prices were calculated by multiplying gold prices by the current market price of exchange between marks and gold currencies established in foreign exchange markets’ (Foley 1983, p. 16). IAN The state also used deposit banks to monetize the debt, and would

enforce convertibility of bank liabilities into specie and/or declare bank liabilities to be legal tender. See Day (1987, pp. 150-3) for examples in Genoa (1398), Venice (1619), and Genoa (1720). 2a The previous government (Charles II) had destroyed trust in itself by freezing deposits made by the goldsmiths in the State Exchequer and forcing them into bankruptcy. As a result, William HI’s government was unable to obtain any large loans to conduct a war with France. The Bank of England was also created with monopoly power in the hope of obtaining monopoly profits for its owners (Cameron 1967, p. 20). . Bagehot cites the example of The Bank of Dundee, which was founded in 1763 as a note issuing bank, but which had no deposits until 25 years later (Bagehot 1927, p. 83). Even as late as 1865, the banks of France and Germany were primarily note issuing and had relatively few deposits (ibid., p. 84). The record of the Stuckey Bank of Bristol indicates that note issues were three times the quantity of deposits in 1819, but were of about the same size by 1832 (Wood 1939sp 2): 24. Cameron (1967, p. 42) presents the following data for M2 (each component is expressed as a percentage of M2):

68

Year Specie Bank notes

Deposits Other

c

1688-99 50 10

91750') SROe ies:

1775 aoe AIS)

91800-01) 183 191855 2 eee 81ers) s Ue

ea

a

eal

b

b

5.09

241978

38.7

40

50

eh

AO

aL

AS)

Notes: (a) deposits were illiquid and insignicant; (b) deposits are included with bank notes; (c) other is comprised primarily of bills of exchange. See Cameron for sources. 25% Country bankers frequently came out of the wholesale or retail trade, or out of manufacturing. The banks were started to provide clearing facilities, to provide credit, and to provide a local medium of exchange (Cameron 1967, p. 24). 26. As shown in the table in Note 24, bills of exchange were something like four times greater in quantity than the sum of bank notes and demand deposits. 2k Lancashire was an exception to the rule that locally issued notes

circulated outside of London. A series of bank failures in Lancashire had made the public wary of private notes. Thus, bills of exchange and Bank of England notes served as the media of exchange (Cameron 1967, p. 26). 28. During 1825-26 a severe crisis led to the failure of 60 country banks, which led to demands for banking reform. It was believed that a primary cause of failure was the inadequate capital base of private banks. Thus, part of the reform was to allow joint stock banking as a means to expand equity (Cameron 1967, p. 27). 29: However, some notes were still being issued as late as 1921 (Sayers 1957, pp. 139-48). 30. This reflux will be discussed below in Chapter 4 when I examine the Currency School-Banking School controversy. 31. Prior to 1825, it was normally argued that balance of payment flows were governed by prices: if the country banks expanded notes too fast, prices would rise in the country so that purchases would be made in London. This would cause country notes to flow to London,

where they would be extinguished. After 1825, it was argued that in order for country notes to increase faster than London notes, lower interest rates would be required in the country so that country banks could make more loans and issue more

notes. However,

the notes

would then flow to London in search of higher interest rates. Thus, they would again be extinguished. See Wood (1939, p. 175). OZ: For example, the 1844 Act permittted the bank of Stephens Blandy to issue up to £43,271 in the form of notes, but by 1854 it had only £27,000 in notes outstanding. While Lloyds was authorized to issue £38,816, it decided in 1865 to abandon note issue altogether (Sayers 1957, p. 150). See Bagehot (1927, p. 24) for further information on note issues.

69

333 According to Wood (1939, p. 19), the reserves Stuckey’s bank held against notes varied from 5 per cent to 25 per cent, even in ‘normal’ times. The reserves of Becket’s bank (coins plus Bank of England eae notes) were equal to 50 per cent of notes issued. 34. This variability of reserves shows how little relation the liabilities of country banks would bear with Bank of England liabilities. Many of the theorists of the time recognized that Bank of England control was always through price, and not directly through the quantity of reserves, as will be discussed below in Chapter 4. 35: As late as 1797, the Bank of England refused to discount for country banks (Cameron 1967, p. 22). Bagehot recommended ‘What is wanted and what is necessary to stop a panic is to diffuse the impression that, though money may be dear, still money is to be had. If people could be really convinced that they could have money if they wait a day or two...most likely they would cease to run in such a mad way for money... The holders of the Bank reserve ought to lend at once and most freely in an incipient panic, because they fear destruction in the panic’ (Bagehot 1927, p. 64). 36. While government debt became a popular secondary reserve among private banks after World War I, it never played an important role prior to this. However, paper issued by canals, railroads, and foreign projects were important. Beginning in the early nineteenth century, country banks would buy bills of exchange on London to hold for liquidity. Short term loans to stock brokers were also favored as a liquid asset and included as a source of reserves. 373 This doesn’t mean that money was created to replace barter, but that once a money of account has led to the creation of markets, commodity money can suffice. 38. In Kalecki’s equation, profits are identically equal to investment, plus capitalist consumption, plus the government deficit, plus net exports, less saving out of wages. At the individual firm level, of course, growth is dictated by the necessity of remaining ‘competitive’ with other firms in the industry. 39) Thus, credit funds investment and investment generates profits. Profits, in turn, fund long term positions in investment goods. Saving cannot be the source of funding of investment in the short term, as

it is (at best) a result of investment. ‘Saving’ is accrued in the first instance as profit income, which can then be used to finance capital accumulation (‘save’). 40. Of course, the level of development of the financial system is only one of the barriers faced by an LDC. 41. For example, the giro money apparently now used in Brazil for property purchases is the US dollar. 42. If this were not true, it would be difficult to explain why populist groups continually attacked the banking system and advocated ‘easy money’. Credit rationing is a normal feature of all monetary

70

economies. 43. See Chick and Dow (1988) for a similar discussion of the situation in the DC

and

LDC,

although

their analysis

also concerns

relation between the center and the periphery within a DC.

71

the

3.

ENDOGENOUS MONEY VERSUS EXOGENOUS MONEY

Introduction

In the previous two chapters, I defined money and examined its origins, I examined

the evolution

of financial institutions,

and I

discussed the role that money necessarily plays in capitalist economies. Both the history and the logic of capitalism lend support to the endogenous money approach. However, this approach is not the one which has been adopted by orthodoxy. In this chapter, I will further define what I mean by the exogenous and endogenous money approaches; distinguish between two kinds and and contrast the exogenous of exogeneity; compare money has only not that argue approaches; money endogenous been endogenous in the past, but that it cannot be made exogenous; and distinguish between two different endogenous money approaches. In Chapter 8, I will review the evidence in support of the endogenous money approach - the present chapter will be primarily definitional and theoretical.

The endogenous money approach In the endogenous approach to money, money enters the economy during normal economic processes. In contrast, orthodox theory begins with a model of exchange to which money is then added

as

an

exogenous

variable.

However,

money

is neither

exogenous in the statistical sense, nor is it exogenous in the control sense, as will be explained shortly. Thus, money is not a variable which can be added to theory when it becomes convenient, since money enters into the determination of the most important variables which theory attempts to analyze. In Chapter 1, I argued that money is a unit of account which transfers purchasing power from the future to the present. Once created, money serves as a medium of exchange, as a means of payment, and as a store of value. In a capitalist system, the 72

important debts are those which are created as capital assets are privately produced and purchased. These debts involve the creation of money, and money is used to retire the debts, which destroys money. Thus, money is endogenously created as assets are produced and financed, and is endogenously destroyed as positions are liquidated. Money also plays an important role precisely because it is (or can be converted into) the medium used to retire debts: accumulation of money reduces the hardship which arises if at some point payment commitments can’t be met out of income flows, and reduces the probability that undesired debt (or asset sales) will become necessary. As such, accumulation of money becomes a way in which one deals with an uncertain future. Money reserves provide individual security in a world which is to some extent unknowable. Increased uncertainty about the future increases the desire to accumulate such reserves as a buffer against disappointed expectations. However, the attempt to accumulate liquid assets (an increase in liquidity preference) has effects which spread throughout the economy. In particular, the demand for less liquid assets declines as a general run to liquidity occurs. This, in turn, can reduce the price of illiquid assets (including real capital) below production costs, so that real output is affected. In this way, a rise in the demand for liquidity to ward off uncertainty will have real effects. Thus, money cannot be a veil. The endogenous money approach includes three essential propositions, first, loans make deposits, second, deposits make reserves, and third, money demand induces money supply (Lavoie 1985). The first two propositions imply that banks do not passively await deposits so that they can issue loans. With a developed and integrated financial system, as loans are spent, the vast majority of expenditures return as deposits to the banking system (leakages into currency are minimal). Asset and liability management, the Fed funds market, international sources of liquidity, and central bank loans (either at the discount window or in lender of last resort operations) ensure that banks that need reserves are normally able to obtain them.’ As the previous chapter demonstrated, deposit banking is a relatively recent development, anyway, preceded by a long history of the endogenous creation of credit (such as bills of exchange and bank notes) to finance spending. These considerations lead to the third proposition, that the supply of money is a function of the demand for money. This suggests that

13

neither excess money supply, nor excess money demand, makes much sense. As Kaldor (1985) argues, money is supplied because someone wants it. Money supply and money demand are simply different sides of the balance sheet. From the firm’s point of view, money demand is its willingness to go into debt, and money supply is the IOU it issues. Of course, the firm’s IOU is not money unless

someone is willing to accept it. From the bank’s point of view, money demand is indicated by the willingness of the firm to issue an IOU, and money supply is determined by the willingness of the bank to hold that IOU and to issue its own liabilities to finance the purchase of the firm’s IOU. To argue that excess money is possible forces one to invent mechanisms by which unwanted money is injected into the system, such as Friedman’s money helicopters. The supply of money increases only because two parties willingly enter into commitments. Furthermore, money demand (by creditworthy parties) can always be met. However, the demand for liquidity cannot always be met by quantity adjustments, so that price adjustments are required to bring liquidity preference into equality with the supply of liquidity. A rise in money demand may not affect the interest rate, but a rise in liquidity preference is

likely to do so. In Chapters 7 and 9, I will show how financial institutions create money, and how they meet reserve requirements through asset and liability management. This analysis will lead to the proposition that the supply of money is determined primarily by those factors which determine the demand for loans, although not all demand will be met since credit rationing is a normal feature of a capitalist economy. Except in the case of a liquidity crisis generated by a collapse of expectations, the demand for loans is primarily a function of the demand for spending in excess of income. I will argue that this demand, in turn, is primarily a function of the expected income stream which will be generated through ownership of the assets. Thus, the supply of money will be shown to be primarily a function of profit expectations.’ This, of course, is consistent with the discussion of the role money plays in the ‘logic’ of capitalism presented in the previous chapter.

Exogenous control of the money supply In this section, I will examine several representative analyses in

which the money supply is taken as an exogenous variable. As I

74

mentioned above, exogeneity may be defined in the control sense or in the statistical sense. I will first discuss exogeneity in the control sense and will show in the next section that exogeneity in the control sense does not imply exogeneity in the statistical sense.* Exogeneity in the control sense means that the central bank determines the stock of money. More specifically, the Fed determines the reserves (or monetary base), which then determine the money supply through the deposit expansion process (or money multiplier). This assumption is pervasive in mainstream literature. For example,

Anderson

and Jordan, in their test of the relative

effectiveness of monetary and fiscal policy, argue that: The monetary base is considered by both the portfolio and the modern quantity theory schools to be a strategic monetary variable. The monetary base is under direct control of the monetary authorities, with major control exerted by the Federal Reserve System.... The money stock is also used as a strategic monetary variable in each of the approaches to stabilization policies.... (Anderson and Jordan 1968, p. 13)

Poole, in his analysis of the optimal choice of monetary policy, assumes ‘throughout this paper [that] the money stock can be set at exactly the desired level [so] the money stock may as well be called an instrument of monetary policy....it is, for example, a straightforward matter to use the approach of this paper to treat the monetary base as an instrument and the money stock as a stochastic function of the monetary base’ (Poole 1970, p. 198). Brunner describes three major conclusions of the approach he named ‘Monetarist’ in 1968: First, monetary impulses are a major factor accounting for variations in output, employment and prices. Second, movements in the money stock are the most reliable measure of the thrust of monetary

impulses.

Third,

the behavior

of the monetary

authorities dominates movements in the money business cycles. (Brunner 1968, p. 9)

stock

over

He concludes that empirical evidence shows that ‘actions of the Federal Reserve are transmitted to economic activity via the

resulting movements in the monetary base and money supply, which initiate the adjustments in relative prices of assets, liabilities, and the production of new assets’ (Brunner 1968, p. 24). Thus, the Fed controls reserves (or the base), which

then

determine the money stock, at least within a narrow range. In 1963,

ibs)

the aggregate Tobin argued that in orthodoxy, ‘determination of ting and volume of bank deposits is just a matter of accoun es by reserv bank of arithmetic: simply divide the available supply 272). p. the required reserve ratio’ (Tobin 1987, this ‘old’ Gradually theory and empirical evidence cast doubt on y of histor the trace will I 1963). in it d view (as Tobin characterize er, Howev r. chapte next the in view ‘new’ the the development of in and oks textbo in living and well and alive still is the ‘old’ view in that empirical studies. For example, Balbach set out to show . desires so it if spite of the Fed’s record, it actually can hit targets would this and Balbach argues the Fed can control the base, enable it to hit monetary targets: It has been widely asserted that, although excessive monetary growth is a cause of inflation, the tight ‘mechanistic’ control of monetary aggregates is infeasible. This argument is based on allegations that the monetary base cannot be controlled, or that the base multiplier is too variable for the central bank to control monetary growth, particularly over short periods of time. This article...demonstrates that the basic constraint on money growth - the monetary base - can be controlled with precision. Nondiscretionary accounts in the Federal Reserves’s balance sheets are much smaller, and vary less, than those which it can

control directly. (Balbach 1981, pp. 11-12)

Meltzer distinguishes between the Fed’s ‘potential for monetary control’ and its ‘actual control’. After reviewing various studies of the degree to which the Fed could control monetary aggregates, he asserts:

Allowing for the estimated standard deviations, it appears reasonable to conclude that there are a number of control procedures capable of holding the average monthly growth of money within 1 percent to 2 percent of the announced target with a high (95 percent) probability. (Meltzer 1982, p. 637)

After

comparing

this

standard

against

the

Fed’s

actual

performance, he concludes:

[A]ctual performance is substantially less than the attainable best. Instead of a 95 percent probability of coming within +/- 1 percent of the annual target, these data suggest that the 95 percent probability region is + /- 3.5 percent... [A]ctual variability is considerably above the computed minimum. We can, however,

76

move toward the attainable minimum found in the studies I have

cited, and in other studies, done within and outside the Federal

Reserve. I propose that we run the experiment - control the stock of money without relying on estimates of the demand for money or interest rates - and evaluate the results. The experience with monetary targets during either the past six years or the past six months is so far from the attainable minimum that we are unlikely to do worse than we have. (ibid., pp. 638-9)

Thus, the Fed can and should establish targets for monetary aggregates and can and should hit them. Only lack of ‘moral fiber’ prevents it from doing so. The rational expectations debate over the effectiveness of monetary policy can be taken as a final example of the use of an exogenous money supply (in the control sense). Typically, a reduced form equation representing a Phillips curve with rationally formed expectations is used, along with a rule adopted by the monetary authorities which determines the quantity of money supplied and an equation which links prices to money. {I]f the money supply rule is known to economic agents and is based on the same information as those agents have (for example, the money supply may be adjusted on the basis of lagged values of prices and output), then the predictable effects of the money supply on prices are embodied in t-1(P)t and monetary policy can affect output only by doing the unexpected. (Pischer1977,p..191)

Thus, if the money supply is exogenously controlled by the Fed, if markets

clear,

and

if agents

are

‘rational’

(in the

rational

expectations sense), then money doesn’t matter. Fischer goes on to show that money could matter if the central bank has superior information than do private agents, or if wages are sticky due to long term contracts. In these cases, the monetary authorities could use counter-cyclical policy to offset random disturbances. In some ways, rational expectations Monetarism could be seen as the ultimate form of neoclassical macroeconomic theory.’ With flexible prices, all markets clear. Equilibrium levels of employment, real output, and relative prices are all determined without

reference to money. If agents are ‘rational’, then money cannot have any real effects. The Fed supplies money and thereby

determines the rate of inflation. An increase in the money supply will only raise nominal incomes sufficiently that the new higher quantity of money is held as desired transactions balances. If an

OE

increase in the money supply has real effects, they are the result of unanticipated (random) changes in the money stock, or irrational behavior, or sticky prices. Thus,

the normal

assumption

made

in orthodoxy

is that the

money supply is exogenously controlled by the central bank. I will argue below that this is not and cannot be made so. First, however, I will examine statistical exogeneity.

Statistical exogeneity There is, of course,

a monumental literature which attempts to test

a multitude of variations on the exogenous money theme. In this section, I will first review empirical work which tests the money-spending relationship.° These are not tests of the exogeneity of money, for they assume the money supply is given exogenously. I will then argue that if money is in fact endogenous (in the statistical sense), most empirical tests of the money-spending relation are biased. Indeed, as Moore states: ‘All models that treat

money

as exogenous

- or virtually

everything

written

in the

monetary, macro, and growth literature - are either misspecified or

incomplete’, including most tests of crowding out, interest rate determination, spending multipliers, and monetary control (Moore 1988b, p. xiv). I will look at tests of the exogeneity of money in Chapter 8. Friedman has contributed a great deal of research in this area. For example, Friedman and Schwartz (1963) argue that the demand for money is a function of permanent aggregate real income per capita (Yp) multiplied by a permanent price level (Pp), so that:

Md = Pp[F(Yp)]

()

The empirical specification then becomes:

M/Pp = a(Yp)"

(2)

where M is the money stock, and a and b are numerical constants

to reflect a proportional relation between changes in the real money stock and permanent income. Note that money demand has been replaced by money stock on the assumption that ‘the amount demanded is always equal to the amount supplied’ (ibid., p. 225). 78

The money stock is given exogenously and the demand for real balances is a function of permanent income. If the money stock is increased exogenously so that people have excess real balances, spending, income, and prices adjust until all of the money stock is held as desired real balances. Friedman and Schwartz find that a transformed version of this equation performs quite well, so ‘for major movements in income, we concluded that there is an extremely strong case for the position that sizable changes in the rate of change in the money stock are a necessary and sufficient condition for sizable changes in the rate of change in money income’ (ibid., p. 235). Judd and Scadding (1982) review the literature concerning the money demand function. It was believed that if a stable money demand

function

could

be found,

this would

indicate

that the

quantity of money is predictably related to a small set of key variables which link money to the real sector. This would then justify the belief that an increase in the money supply results in excess money and forces predictable adjustments until money demand is equilibrated to the money supply. Empirical studies did find stable

money

demand

functions

in the early 1970s, which

placed pressure on the Fed to establish targets for monetary and

credit aggregates.’ However, any empirical specification which must identify a money demand equation separately from a money supply equation, or any specification which posits an exogenously determined money supply when it is in fact endogenously determined, is subject to bias. Even if the Fed is able to control the money supply, this does not warrant the conclusion that the money supply is exogenous in the statistical sense. And if the money supply is not statistically exogenous, then most empirical work in this area suffers from simultaneity bias. Cooley and LeRoy explain: In macroeconomic theory a variable, the level of which is set by government as an implementation of economic policy, is represented analytically as an exogenous variable; that is, as a variable not determined within the model, since the latter explains only the behavior of the private sector. Analysis then centers on the effect of policy changes, represented by shifts in the exogenous policy variable(s) on endogenous variables. The definition of exogeneity relevant for statistical estimation, however, is entirely different: a variable in a regression is statistically exogenous only if it is independent (in the probability sense) of the unobserved explanatory variables. Whether a

19

variable which is exogenous in the former sense is also exogenous in the latter and relevant sense depends on how government policy is in fact conducted; that is, on the nature of the government’s policy rule. (Cooley and Leroy 1981, p. 838)

Desai (1989) has presented a rigorous definition of exogeneity. Let M represent money and X represent a vector of variables such as

income,

prices,

and

interest

rates,

while

Z will

represent

exogenous variables such as technology and tastes. The money supply is then exogenous if M is exogenous with respect to X, although M might still depend on Z. If M is exogenous with respect to current values of the variables in X, but depends on

past values of X, then money may be said to be weakly exogenous. In this case, money might be exogenous in the control sense, but is not statistically exogenous. If M depends on current values of X, then it may not be exogenous even in the control sense. It is very unlikely that the money supply is exogenous in the statistical sense, so most empirical studies are biased. For example, a specification which relates money demand to income and the interest rate will give inconsistent least squares estimates if there is feedback from the error to the interest rate. ‘Since the Federal Reserve chooses the setting of the interest rate partly or largely in order to influence the behavior of the money stock...there appears in fact to be a strong prior presumption for the existence of correlation between the explanatory variables and the error...’ (Cooley and LeRoy 1981, p. 836). Laidler (1977) has argued that if there is a variable which shifts the money supply function without affecting the money demand equation, then reliable estimates of the money demand function can be obtained.

However, Cooley and LeRoy argue that none of

the likely candidates will suffice. For example, reserves certainly enter the money supply function but wouldn’t be expected to directly enter the money demand function. ‘But the condition that supply shifts independently of demand means that reserves must be excluded not only as an explicit explanatory variable on the demand side...but also as an explanatory variable for the unobserved determinants of money demand’ (Cooley and Leroy 1981, p. 837). If the Fed accommodates random shifts in money demand

(which seems likely), then covariance will exist between

reserves and the error in money demand, so that ‘neither the parameters of money demand nor the covariance of the error with reserves is identified...’ (ibid., p. 837). Moore argues that economic theory should take as exogenous

80

those variables whose changes cause changes in endogenous variables. However, since there may exist underlying variables that cause changes in both ‘exogenous’ variables and in ‘endogenous’ variables, empirical tests can never be decisive because they may well be biased. If money is to be taken as a statistically exogenous variable, it must be shown that it does not respond to money demand. This, in turn, requires that monetary authorities (and banks) don’t respond to changes in money demand. Even if the Fed controls the base, and even if the base is Teliably related to

the money supply, neither the base nor money supply can be treated as exogenous variables if the Fed responds to anything which increases the demand for credit. Those who use an exogenous money supply in empirical work must provide a theoretical justification for the absence of responsive monetary policy (Moore 1988b, pp. 144-7). Cooley and LeRoy conclude: ‘there is no obvious way to

formulate models of equilibrium in financial markets in which the demand for money is identified’ (Cooley and LeRoy 1981, p. 840). Thus, there is no way to determine whether an equation is estimating money demand or money supply. Therefore, any equation of the form:

M = aY+bR

(3)

where M is money, Y is income, R is the interest rate, and a, b are

parameters will almost certainly give biased estimates. It is not surprising that the data can be used to support any elasticities for income or interest rates from negative values to positive values, since the equation is inadequately specified. Friedman’s empirical work relating money demand to permanent income and interest rates suffers from this problem, since there is no way to determine whether he has estimated money demand or is money supply functions. His assumption that the money supply fixed by the Fed, and that money demand is brought into equality with money supply through adjustments in nominal income, cannot be verified by his empirical work. Many other possible explanations exist. For example, if nominal income determines money demand,

and money demand then determines money supply (because either the Fed or commercial banks supply money on demand), one would expect a very close relation between nominal income and the money stock. Thus, it would be impossible to distinguish among various explanations of the empirical correlation. 81

Empirical studies which require exogeneity of the money supply are of questionable value. However, this does not invalidate the argument that the Fed could and should control the money supply. While Friedman’s theory that money causes income cannot be verified empirically, his argument that the Fed can and should exogenously set a constant money growth rate rule survives the Cooley and LeRoy critique. Although it is impossible to use empirical data to justify Friedman’s belief that a constant growth rate for the money supply would be associated with stable growth of income, it is still possible to make the normative argument that the Fed should stick to a constant growth rate rule.’ However, I will argue below that such a rule is not feasible. In summary, a high correlation between money and spending (which Friedman reports that he finds) proves nothing regarding causality, since ‘reverse causation’ (that is, ‘spending causes money’) would lead to identical results. If the money supply is endogenous, then any equation similar to Equation 3 above suffers from simultaneity bias. Therefore, the parameter estimates are unreliable and provide little guide for policy regardless of the true state of the world. Mainstream theory faces the following quandry: (1) If the Fed controls the money supply, what rules does it follow? If it follows either accommodative policy, or leans against the wind, then money demand determines money supply, at least to some extent. In this case, money supply and money demand equations cannot be statistically identified, and it is difficult to say much empirically about the role of money. If the Fed’s policy is random, then a money demand equation can be estimated, but it would

be difficult to explain why the Fed

would choose random behavior. (2) If the Fed does control the money

accommodative

policy, could

supply, normally using

it choose

instead

to follow

Friedman’s growth rule? If it did so, what would be the results?

Historical data cannot be used to predict the results if these data are contaminated by a statistically endogenous money supply. (3) If the Fed cannot control the money supply, what does? If the money supply accommodates money demand, then the determinants of money demand must be examined. However, it is impossible to estimate these since there is no known way to identify a money demand function.

82

Endogenous money and central bank behavior For the reasons discussed above, empirical tests of the money-spending relation are likely to be biased because the money supply is not likely to be statistically exogenous. I will first discuss why an endogenous money approach leads to the expectation that money and spending will normally be closely correlated. I will then argue that the Fed is not free to set the money supply at any desired level, so that the money supply is not exogenous in the control sense either. I will argue that the demand for money normally determines the supply, and that the Fed is normally accommodative. While the Fed can influence the rate of growth of the money supply, bank behavior usually enables banks to escape Fed constraints. When a Fed constraint actually becomes operative, the Fed is eventually forced to abandon the constraint and accommodate the demand for money. The quantity theory posits a stable relation between a monetary aggregate and spending. If spending and money are narrowly defined, then it is not surprising that a more-or-less stable relation will emerge. For example, if the monetary aggregate is defined as demand deposits and if spending is defined to exclude all intermediate goods, then ‘money’ and ‘spending’ will be highly correlated, for the following reason. Labor is almost always paid in terms of demand deposits. The capitalist takes out a short term loan (issuing an IOU and receiving a demand deposit) to meet working capital expenses, most of which are labor costs. Thus, the size of this demand deposit is closely related to the quantity of wages paid in any given payment period. The payment of wages conforms to a regular pattern (weekly or monthly). Worker expenditure of income is comprised mainly of running down demand deposits, and this is also on a fairly predictable schedule. Thus, there is a high correlation between demand deposits held by workers and their wages.” Indeed, if wages are regularly spent over the income period, the average demand deposit balance of workers will be equal to one-half the wage bill (Graziani 1987, p. 32). Since wages make up three-fourths of income, there will be a high correlation between ‘money’ and income. Since most wages are consumed and because consumption is the largest part of spending on final goods and services, there will also be a high correlation between ‘spending’ and ‘money’. The relation between these two aggregates will change as wage payment schedules change, as worker spending patterns change, as the distribution between wage

83

income and non-wage income changes, as the number _of intermediate steps in production changes, and as the proportion of wages consumed changes. Such changes are not normally abrupt. However, high inflation and interest rates in the 1970s did lead to a shift out of demand deposits and into other checkable deposits (as will be discussed in Chapter 7 below). Futhermore, the distribution of nominal income between wages and other income did

the

proportion

of income

consumed.

Thus,

changed,

as

assets.’

The other is to begin with a money supply which is

beginning in 1974, money demand equations which forecast money demand based on data concerning various parameters (such as inflation, interest rates, and income) began to overpredict real money balances. This overprediction was confined almost entirely to the demand deposit component of M1. This suggested that the money demand function had shifted downward, and has been attributed to innovations and other ‘shocks’ (Judd and Scadding 1982, pp. 995-6). Orthodoxy refers to this as the ‘missing money episode’, which led to the ‘search for the missing money’. There are two ways to deal with the ‘missing money’ episode. One is to redefine money so as to include a wider variety of

endogenously created as spending occurs. The second option is taken here. The first is unsatisfactory because it will attempt to relate a broader, but still exogenous, monetary aggregate to final spending. By concentrating on final spending, the analysis will miss the important intermediate transactions in which positions in assets are taken or liquidated, in which debts are created and destroyed,

and in which expectations are formed and validated or disappointed. Thus, rather than focusing on the use of an exogenous money supply as orthodoxy does, it is better to look at the creation of money during normal capitalist processes in which an endogenous money supply finances deficit spending. Experience during the 1980s has shown the folly of merely redefining money. Beginning in 1981, orthodox economists found there was far more ‘money’ than their money demand equations predicted.” That is, the income levels and expenditures on final goods and services were too low relative to the quantity of money. Because orthodoxy excludes all but expenditures on final goods and services and because it ignores the asset side of bank balance sheets, it is unable to explain how bank balance sheets could expand at a rapid rate even while production of real goods and services remains sluggish. As Brockway (1988) argues, much of the rapid growth of the money supply during the early 1980s was due

84

to the creation of money as positions in the stock market were

taken on the basis of IOUs.”

Thus, the endogenous approach to money is able to explain a close relation between spending and money, but is also able to explain why this relation is institutionally based and, therefore, why it can change over time. Innovations which generate new monetary assets, changes in institutional practices, shifts in the distribution of income, and changes in leverage ratios can all lead to a shift in the relation between a particular form of money and a particular type of spending. However, the money supply remains endogenously determined. This endogeneity is not dependent upon Fed behavior - the money supply cannot be made exogenous in the control sense. In Chapter 7 on institutions below, I will discuss in more detail how banks are normally able to avoid constraints through a combination of asset and liability management, innovation, and off-balance sheet operations. Banks normally meet the demand for loans, and then obtain the funds required to cover them either through the purchase of funds, through balance sheet management, or through innovation. In those cases when a bank is not able to obtain funds in retail or wholesale markets and is not able to make required adjustments in its balance sheet, it turns to the Fed as a lender of last resort. If the Fed refuses to provide needed reserves at this point, the bank will be forced to liquidate assets. Due to the nature of some

of the bank’s

assets,

there

is a limit to its ability to

liquidate. Furthermore, as balance sheets of different economic units are connected, forced liquidation of assets by one unit can lead to repercussions throughout the system and to a general debt repudiation. Thus, I will argue that the Fed is ultimately not free to refuse to provide reserves through loans at the discount window or through open market operations. Commercial banks typically establish relations with customers to reduce the uncertainty involved in making loans. Once a bank has entered into a relationship with a customer, it has strong incentives to meet the demands of that customer. As Minsky says, Because bankers live in the same expectational climate as profit-seeking bankers will find ways of businessmen, accommodating their customers.... Banks and bankers are not

passive managers of money to lend or to invest; they are in business to maximize profits. Loans, unlike securities, involve a

customer relation, in which banks use private knowledge willingly

given by the borrowing

units. An

85

implicit commitment

for

continuing relations may exist.... (Minsky 1986b, pp. 228-230)

Thus, banks establish relations with customers and provide services

and loans essentially on demand.” Banks then obtain the funds

necessary to meet the drain on reserves as customers utilize their loans. As these loans are an illiquid asset, banks are unlikely to sell them in secondary markets because the large discount required (due to private information involved in the customer-bank relation) would be too great since the bank couldn’t absorb the loss. If loans comprise a large portion of a bank’s assets, then its ability to meet reserve deficiencies by selling assets is limited. In this case, the bank must increase its liabilities in order to obtain reserves.” If the money supply is growing faster than the Fed wishes, it may use open market sales of government securities, raise the discount rate, institute quantity controls at the discount window, or raise

the required reserve ratio. An open market sale or a rise in the required reserve ratio will place direct pressure on bank reserves. Banks which were in equilibrium previously (no excess reserves) will now be forced to increase reserves. Banks can attempt to purchase reserves from the banking system and from overseas banks (Eurodollar loans), to induce customers to switch from deposits with a high required reserve ratio to deposits with a lower reserve

ratio, to sell assets, or to borrow

from

the Fed. In the

aggregate, the only sources of reserves are public currency holdings, overseas funds, reserve economizing liability management, and the Fed. The banking system as a whole cannot liquidate assets (except by selling assets to the Fed, foreigners, or currency hoarders), so it must operate primarily on the liability side. If banks have exhausted all private sources of reserves (including reserve economizing adjustments), then they will be forced to liquidate assets or to turn to the discount window.

However, if the Fed is

trying to constrain reserves, it will not want to open the window. A generalized attempt to liquidate assets or call in loans will lead to a run on deposits, since any sale by one bank will usually lead to a withdrawal of deposits on another (and a loss of reserves). Furthermore, since secondary markets for many assets (particularly loans) are thin, forced sales will result in depreciation

of asset

values. Thus, there is an asymmetry in the Fed’s ability to directly control bank reserves: the Fed may be able to increase reserves (or excess reserves) through open market purchases or to increase excess reserves through a reduction in required reserve ratios, but it may

86

not be able to reduce bank reserves.’* The Fed cannot exogenously decide to lower the rate of growth of the money supply since it must take account of the portfolios of banks. ‘Once banks have reduced their portfolios of marketable securities towards the minimum secondary reserve levels required for liquidity purposes...the central bank has no choice but to make the required reserves available on demand’ (Moore 1985, p. 12). The Fed is not free to use open market sales, rising reserve ratios, or quantity controls at the discount window to reduce bank reserves. However,

the Fed does have more power to set the discount rate.!’ Therefore, while the Fed may not be able to refuse to operate as lender of last resort, it can set the rate it will charge on its loans within some range. Banks which must borrow reserves will be forced to pay the higher discount rate. This will affect the bank’s ability to make new loans, and can be expected to induce a rise in interest rates, generally. As the price of credit rises, the rate of growth of the money supply can be expected to (eventually) fall as customers decide to postpone taking out loans. However, the Fed is limited even in its ability to push up interest rates. A rise in interest rates causes distress which may force the Fed to abandon its tight monetary policy. For example, rising interest rates harm financial institutions which hold illiquid assets that earn

a low rate of return,

but which

have

financed

their

position in these assets with short term liabilities. Since equity is typically very small in relation to the size of assets, most assets are financed on the basis of an expansion of liabilities. The classic example is a saving and loan association which holds long term, fixed rate mortgages financed through issues of shorter term time deposits. In the presence of an interest rate ceiling on deposits, the institution faces disintermediation when the interest rate rises. But even in the absence of interest rate ceilings, the savings and loan may be forced into technical default if it must pay more on its liabilities than it receives on its assets. In this case, the Fed must

either lower the discount rate to prevent default, or must (in conjunction with the federal deposit insurance agency) take over a failed bank. Alternatively, the Fed could allow the system to collapse. As I will argue below, for the past half-century the Fed has chosen to intervene rather than wait to see whether the system would collapse, although it has recently been willing to push a large number of institutions into default. Constraining reserves may cause the rate of interest banks must pay on deposits (and other issued liabilities) to rise relative to the

87

reduce rate they receive on assets. Therefore, tight Fed policy can banks price the raising by supply money the of growth of the rate to ness willing must pay to attract funds and thereby reduce their rates. t interes ted make loans or renew loans at previously negotia rate At the same time, the loans they do make will entail a higher

of interest, which may reduce the quantity of loans demanded (again, lowering the rate of growth of the money supply). Thus, tight Fed policy may reduce the rate of growth of the money supply. However, tight Fed policy does not actually mean that the Fed reduces the supply of reserves. If tight policy causes a run to liquidity, the Fed may have to respond by supplying more reserves than it did when the money supply was growing at a high rate.

Minsky (1986b), Niggle (1986) and Wolfson (1986) present similar evaluations of Fed attempts at monetary constraint. Typically, toward the peak of an expansion, the Fed attempts to constrain the growth of monetary aggregates. Because loan demand initially remains strong, banks find ways to continue to make loans. This involves innovations and reserve economizing liability management, both of which tend to reduce bank liquidity. (That is, the asset-to-reserves-plus-government bonds ratio rises.) Eventually, however, the monetary restraint begins to push up interest rates on acquired liabilities. Some financial institutions find that the return on their assets is not sufficient in the new regime of higher interest rates. Some firms find that their income stream is not sufficient to meet new higher payment commitments (which have increased due to rising interest rates on short term liabilities which must be rolled over). In general, economic units

in short positions (borrowing on a shorter term than the term of assets) such as financial institutions,

speculators, and Ponzi units

are placed in a precarious position.” Various units attempt to liquidate assets to meet commitments. A general run to liquidity creates the conditions for a financial crisis, because, as discussed

above, general liquidation of assets requires massive writing down of asset values and debt repudiation. The Fed averts the crisis when it abandons monetary restraint. As a specific example, in Chapter 8 I will look in detail at the Fed’s ‘experiment in monetarism’ beginning in 1979. The Fed was to ignore interest rates and to target the rate of growth of the money supply. As the Fed squeezed reserves, bank liquidity (reserves to liabilities) fell, as did the liquidity of the public as it was induced to substitute bank liabilities such as certificates of deposit for liquid liabilities such as cash and demand deposits.

88

Eventually, this increased interest rates to record levels and brought the financial system to the brink of collapse. The Fed was forced to abandon its tight money policies to prevent a serious financial crisis from becoming a financial crash. Lower discount rates, greater open market purchases, and intervention as lender of last resort prevented collapse of the financial system. This, combined with exploding government deficits, allowed a recovery.” In conclusion, the money supply is endogenously determined. In a capitalist system, capital accumulation is based on expansion of credit, which allows growth of income and spending, and generates profits. However, endogenous processes tend to increase systemic fragility until a point is reached at which the system itself cannot provide the liquidity needed. (This doesn’t mean that money couldn’t be privately created to meet the demands of credit worthy borrowers, but that liquidity preference is ‘too high’ relative to the supply of liquidity.) This point may be reached sooner if the Fed is pursuing a tight monetary policy, which raises interest rates and makes it more difficult for the banks to meet the demand for liquidity. This may lead to forced sales of assets. As forced liquidation proceeds, this eventually affects an institution which is large enough to jeopardize the integrity of the entire system. At this point, the Fed is forced to intervene as a lender of last resort to provide the liquidity needed by the system. Thus, the Fed abandons its attempt to contro] the rate of growth of the money supply. In any case, during a general run to liquidity, the system itself cannot be liquidated, so liquidity cannot be supplied on demand. Thus, intervention by the Fed is required to halt forced liquidation. Fed control of the money supply is ultimately not possible, although Fed policy certainly can influence the quantity of money privately supplied and the price of credit. During an expansion, tight monetary policy can push up interest rates and thereby increase the fragility of the system. During a downturn, tight monetary policy exacerbates the liquidity crisis. Eventually, accommodative policy will be required to stop the crisis. The money supply can’t be directly controlled by the central bank since it is determined by private decisions to enter into debt commitments to finance spending, which includes financing positions in assets. Thus, the quantity of money is a joint decision of those willing to go into debt and those willing to hold debt. On the surface, it appears that the central bank can control the monetary base. However, the Fed is not able to directly control the

base since maintenance of an orderly financial structure requires

89

that the economy cannot become liquidity constrained. As balance sheets expand, if the Fed constricts the growth of the base, the quantity of money increases relative to the size of the base. That is, liquidity becomes stretched. If economic units for some reason try to liquidate positions (due to an increase in liquidity preference, for example), the Fed must allow the base to increase in order to forestall a crisis. Liquidity preference may rise sooner if the Fed is attempting to constrain liquidity. Financial stability requires that the base is at least partially a function of the money supply.” The orthodox, exogenous money approach would like to make the money supply function as if it were commodity money by having the Fed manipulate reserves to hit money targets. However, as discussed in Chapter 2, commodity money is not consistent with the logic of a capitalist economy. Thus, constraining reserves (and hitting a narrow money target) will not constrain credit creation which is essential to capital accumulation. Such constraints will, however, stretch liquidity, raise interest rates, and make

it more

difficult for the private financial system to accommodate the demand for liquidity which results when the accumulation process breaks down. It is only during a crisis that the capitalist system reverts to commodity money - and this crisis can only be resolved by accommodative action on the part of the central bank to supply fiat money. The alternative is a debt deflation which destroys the value of accumulated capital.

Conclusion

I will conclude this chapter by defining what I mean by the endogenous money approach and by contrasting this with an exogenous money approach and with an alternative endogenous money approach. At the most general level, endogeneity implies that the supply of money is not independent of the demand. As Moore (1988b, p. 7) argues, an exogenous money supply makes sense only in the case of commodity money based on a commodity of which there is a fixed supply, or in the case of government fiat money which is not supplied in response to demand (for example, as part of some stabilization policy). In an economy with credit money, however, strict exogeneity is virtually inconceivable: the money supply should respond at least to some extent to money

demand.”

It might be useful to think of a continuum of possibilities, with

90

strict exogeneity at one extreme and strict endogeneity at the other. In the case of strict exogeneity, the supply of money is fixed - it is absolutely unresponsive to money demand. In interest-money space, the money supply curve would be vertical. This has been called the verticalist’ approach by Moore (1988b). Although this is the way the money supply curve is normally drawn in orthodox textbooks, it is probably not perceived as realistic by many economists. Most would probably admit that a large shift upward of the money demand curve (which would cause a large increase in interest rates) would induce an outward shift of the money supply curve either due to central bank interest rate stabilization policy, or to private initiative as commercial banks offer loans in response to higher interest rates (for example, by reducing excess reserves or by economizing on reserves). It is Moore’s position (which he calls the ‘horizontalist’ approach) that the money supply curve is horizontal at any interest rate determined by Fed policy. Thus, a shift of the money demand curve will raise the equilibrium quantity of money, but not interest rates (unless the Fed chooses to change interest rates). Therefore, in his model it is the interest rate, and not the quantity of money, which is exogenously determined (Moore 1988b, pp. 127-30). The Fed cannot directly control the quantity of money, but can affect it through its interest rate policy. Finally, private banks passively supply credit money at any short term interest rate established by the Fed’s discount and open market policies. They are able to meet any level of credit demand since they can always obtain reserves from the central bank. Thus, Moore’s analysis relies critically on the role of the central bank as the ultimate provider of liquidity, albeit, at a price chosen by the central bank. This approach has at least one obvious weakness: it cannot apply to any economy which does not have an accommodative central bank. I will call Moore’s theory the endogenous money-exogenous interest rate approach. The approach taken here, however, implies an upward sloping money supply curve. This follows the original insights of Minsky (1957), as I will discuss in Chapter 5S below. In this view, the money supply function is a complex interaction of the behavior of private

borrowers, private lenders, and the central bank. The central bank

uses a combination of quantity and price constraints to implement

monetary policy. While it is true that the most important function of the central bank is to act as a lender of last resort to guarantee

financial stability, it is free to use quantity constraints within a

wide range of discretion. Ultimately, however, quantity constraints OF

enhance fragility, inducing instability and forcing abandonment of quantity controls. Similarly, the central bank can use a wide range of discretion in setting the price at which it will provide reserves, which then has some influence over market interest rates. Again,

if the central bank succeeds in pushing market interest rates ‘too high’, it increases fragility to the extent that it must abandon tight policy to restore stability. Similarly, in the approach taken here, banks don’t passively supply money on demand. As Moore correctly argues, banks do try to establish relations with customers, negotiating lines of credit. Thus, to some extent, banks must provide credit on demand. When

expectations are high, banks will increase the size of their balance sheets without requiring rising interest rates. They finance positions in assets (such as loans) primarily through expansion of liabilities. This necessarily entails leveraging equity, and normally is associated with declining cash-to-asset and cash-to-liability ratios, as I will discuss in subsequent chapters. Thus, expansion of balance sheets occurs on the basis of increasing leverage ratios and reducing liquidity of the portfolio. Beyond some point, which is determined primarily by convention, banks will require a higher interest rate on loans to induce them to increase leverage or to reduce liquidity. However, innovations and revisions of rules of thumb shift the money supply curve out. Thus, the money supply curve can properly be thought of as an upward sloping step function.” That is, for a period of time it is horizontal, while it will

be generally upward sloping. Indeed, if the central bank were to operate as Moore believes it does (supplying reserves on demand), the horizontal sections of the money supply curve could be expanded. Therefore, accommodative monetary policy can temporarily make the money supply curve horizontal. Even though banks negotiate lines of credit, this doesn’t mean

credit is simply granted on demand at an exogenously determined interest rate. In good times, banks actively seek out potential borrowers. In bad times, as Wolfson (1986) argues, banks cut off credit to all but established customers. Banks have more discretion over the quantity and price of loans granted than Moore allows. As I discussed above in the previous chapter, and as I will argue in Chapter 6 below, credit rationing is normal and rational behavior.

With optimistic expectations, the money supply curve is nearly

horizontal. It begins to get steeper as an expansion proceeds and as balance sheets become increasingly illiquid. It may become nearly vertical just before the business cycle peak. In bad times,

92

the money supply curve is likely to be vertical: banks become unwilling to meet the demand for loans so the money supply curve may remain steep throughout a downswing. Only when expectations have turned around and liquidity ratios have risen (due to reductions in the size of balance sheets and to easier money policy by the central bank) will the money supply curve become flatter. Returning to the exogenous approach, orthodoxy admits that the money supply curve is not now completely exogenous. Monetarists argue that misguided monetary policy has resulted in a money supply curve which accommodates money demand. However, they argue that the central bank can and should make the money supply curve vertical. Textbook Keynesians, on the other hand, fear that

such policy would generate overly large fluctuations in interest rates. Thus, while they agree that the money supply curve is under the control of the central some they advocate bank, accommodation to money demand. In this case, the money supply curve is exogenously controlled, but is clearly not exogenous with respect to the determinants of money demand. In this chapter, I argued that even if the money supply curve is exogenously controlled by the central bank (as both orthodox schools maintain), this does not mean that money supply can be treated as statistically exogenous. Since both textbook Keynesians and Monetarists admit that in the past the central bank has accommodated money demand, they must also accept that virtually all of the empirical literature on money is misspecified, as Moore argues. I also argued that the central bank cannot implement Monetarist policy. There are two main reasons why a ‘constant growth rate rule’ for money is not feasible. First, as argued above, the rule will have to be abandoned if it is effective in constraining the growth of reserves. As fragility increases, the system becomes unstable and increases the probability of a run to liquidity. If this occurs, the central bank will have to step in and provide reserves to prevent debt repudiation. Second, if the central bank targets a monetary aggregate (for example, M1), profit seeking behavior and private initiative ensure that banks and other financial institutions will circumvent constraints to meet demand for credit from worthy customers. Thus, forms of money not included in the target will grow to accommodate demand. Therefore, quantity controls over demand deposits lead to the creation of other checkable deposits, of large while controls on checkable deposits lead to the creation on constraints and deposit, of certificates denomination asset-to-equity ratios lead to off-balance sheet operations. I will

93

discuss these points in great detail in Chapter 7. Thus, the endogenous approach to money taken here relies on two main principles. First, the money supply curve is not independent of money demand because profit seeking behavior and entrepreneurship ensure that the demands of credit worthy borrowers are normally met. This implies that the money supply is not statistically exogenous, and that the money supply curve cannot be made independent of money demand even if the central bankers have ample ‘moral fiber’. Second, the Great Depression and periodic bouts with financial crisis have shown that the central bank is ultimately responsible for maintaining financial stability: private markets cannot do it on their own. Thus, attempts to control monetary aggregates conflict with the most important function of a central bank, which is to prevent debt deflation.

Notes 1. Banks are rarely quantity constrained, in the sense that they can normally purchase reserves as needed. An individual bank can obtain reserves by borrowing from another bank. In the aggregate, the only net sources of reserves for the domestic banking system as a whole come from deposits of cash by the nonbank public, from foreign depositors, or from the central bank. Given the public’s preference for liquidity and a net inflow of liquidity from foreigners, the Fed

determines the total amount of reserves available. When there is a ‘run to liquidity’, banks cannot meet the demand for liquidity because they also become demanders of reserves to meet the run on deposits. This happens when there is a domestic run on banks, or if foreigners decide to liquidate deposits in US banks. A bank experiencing a run is ‘quantity constrained’. It cannot accept IOUs because its own liabilities are not acceptable. 2. As discussed above in Note 1, only the Fed or foreigners can satisfy a demand for cash hoards. A demand for liquidity cannot be met by private institutions when it takes the form of a demand for hoards of cash. Excess demand for hoards drives up the interest rate, as Keynes

argued. Excess supply of hoards is not possible, since debts could be retired or assets could be purchased (Kaldor 1985). 3. In a crisis, some economic units may be willing to borrow to meet cash commitments. If income in a downturn is insufficient to cover interest payments on outstanding debt, such units may ‘capitalize’ interest: borrow merely to pay interest. Minsky has termed this ‘Ponzi finance’. (See Minsky 1986b, and note 18 below.) Moore argues that if Ponzi finance is possible, agents always choose to borrow too much if there is only a price constraint. Thus, banks must impose quantity

94

constraints (Moore 1988b, p. 49). Empirical tests are often used to show that the money supply is exogenous with respect to spending, or to show that reserves are exogenous with respect to the money supply. However, as I will argue in the next section, these tests are likely to be biased. Even if the money supply is exogenous in the control sense, these tests cannot be used as evidence of the central bank’s ability to control the money supply unless money is also exogenous in the statistical sense. Although many who consider themselves to be neoclassical reject rational expectations, the basis of their objection appears to be due to emphasis on market imperfections such as price rigidities or imperfect knowledge. See Foster (1986b) for a similar discussion. However, by the mid 1970s, these money demand equations overpredicted real money balances. Some researchers assumed that this could be caused by financial innovations, changes in regulations, or high and variable interest rates. Changing the specification of the money demand equation to account for innovations (for example, redefining M1), or for interest rate changes (for example, allowing for ‘ratchet effects’) at best seemed to do only marginally better. For an updated look at empirical estimates of the money demand function, see Stone and Thornton (1987). Ben Friedman (1988), as noted below, examines the unexpected decline in velocity which occurred during the 1980s. As velocity is the inverse of money demand in orthodox theory, this implies a massive increase in money demand which cannot be explained by orthodoxy. Presumably, the Fed’s lack of ‘moral fiber’ caused the rate of growth of the money supply to explode, and for an unexplained reason, money demand also exploded without requiring rapid growth of income or inflation, and without requiring much of a reduction of real interest rates. As Moore (1988b, p. 147) argues, Friedman asserts that the Fed can and should control the money supply, and then uses empirical evidence to argue that changes in the money supply cause changes in income. However, these empirical tests are biased unless the Fed’s behavior in the past has been entirely random. Friedman admits that the Fed has pursued accommodative policy in the past, but formulates empirical specifications in which the money supply must be statistically exogenous. Moore argues: ‘Friedman cannot have it both ways. He cannot consistently argue that the Fed (mistakenly) pursues offsetting and accommodating policies...so that the money stock responds endogenously to changes in money income, and yet simultaneously argue, from the same statistical evidence, that

changes in money income are governed primarily by changes in the money stock, which in turn are derived primarily by exogenous changes in the supply of the high-powered base’ (Moore 1988b, p.

148).

95

10. Moore shows that various empirical tests confirm the hypothesis that the money supply is primarily a function of wages. See Moore (1985 and 1988b, Chapter 7). 11. Judd and Scadding (1982) find that redefining money to include a broader range of assets only marginally improves the performance of their money demand equation. 12: Stone and Thornton (1987) study the ‘velocity puzzle of the 1980s’ and argue that the apparent decline in velocity may be due to misspecification: either M1 or GNP have become bad proxies for money or spending, respectively. They show that the ratio of financial transactions to M1 has not fallen, which seems to indicate that the

shift in velocity has occurred only relative to spending on real goods and services. However, redefining money (for example, using ‘money indices’) does not improve the forecasts of money demand. Stone and Thornton conclude that: ‘Alone, none of these explanations can account for the behavior of M1 velocity. Perhaps, instead, several influences have combined to produce the anomalous velocity behavior that has puzzled many researchers’ (Stone and Thornton 1987, p. 22). Ben Friedman (1988) argues that a fundamental shift in the relations between money and income and between money and inflation occurred after 1980. He shows that no monetary aggregate (M1, M2, M3, or measures of credit) can be used to predict inflation

or income after 1980. He argues that it is not sufficient to explain this as a decrease in velocity, for this is merely a tautology. Experience since 1980 thoroughly discredits the notion that monetary aggregates or even intermediate targets (such as the monetary base or reserves) can serve as guides for monetary policy. He concludes that orthodoxy has not provided an explanation for the fundamental changes which have taken place in the 1980s. Goldfeld (1989) argues that although early attempts to find a specification for money demand appeared to be successful, these successes turned out to be ‘flimsy’. He concludes that the ‘current state of affairs finds the empirical money demand function to be in a bit of disarray’ (Goldfeld 1989, p. 142). 133 Between 1982 and 1987, M2 grew by about $1 trillion (a 48 per cent increase), while M1 grew by $272 billion (57 per cent) and national income grew by about $1 trillion (40 per cent). (All figures are from the Economic Report of the President, 1988.) During this same period, the value of the S&P index of 400 industrials rose by 175 per cent. Brockway (1988) argues that the total increase in the value of stocks between 1982 and 1987 was about $1 trillion, which is (not coincidentally) equal to the increase in M2. When the stock market crashed

in 1987, it eliminated

about

$1 trillion of money

wealth.

Thus, the rate of growth of the money supply should fall drastically in 1988, as it appears to have done. According to the St. Louis Federal Reserve Bank, the compounded annual rate of change of M1

96

between the October 1987 crash and March 1988 reached only 1.1 per cent. (Source: ‘U.S. Financial Data’, The Federal Reserve Bank

of St. Louis, November 3, 1988.) Dow and Saville (1988) argue that a large part of new loans have been made to allow portfolio adjustments, rather than to fund spending on goods and services. Thus, the ‘money supply’ would increase relative to ‘spending’. This helps to explain the large fall of velocity. 14. Again, I must emphasize that this does not mean that all demand for ae will be met. In Chapter 6 I will analyze credit rationing in more etail. 15 . In fact, the average ratio of loans to cash assets is nearly five to one. (See Chapter 7.) The reason the sale of a loan requires a discount below face value is because the buyer normally has not established a relation with the customer who accepted the loan. However, Minsky (1987) argues that loans are increasingly ‘securitized’ for sale in secondary markets. If this tendency continues, then individual banks may be able to rapidly sell loans in secondary markets to obtain reserves. However, if many banks are attempting to sell loans (as would occur during a run to liquidity), such sales could not result in a net source of liquidity to the banking system unless the public gives up currency (which would be highly unlikely), foreigners buy the securitized

loans, or the central

bank

buys the loans. Wojnilower

(1987) argues that the trend to securitization could not occur without implicit central bank guarantees of the securitized instruments. Again, this is consistent with the argument presented in Chapter 2 that the logic of capitalism requires that credit not be liquidated. 16. Moore (1985, pp. 10-11) argues that when the Fed increases bank reserves, the money supply will increase. I believe he has overstated the Fed’s ability to increase the money stock. The money supply will only increase if the banks are faced with prospective loans. Otherwise, excess reserves merely increase. As I show in Chapter 7, excess reserves do vary greatly over time. Baghestani and Mott (1988) show that easy money policy in the mid 1980s led to an increase in excess reserves held by banks. 7. Moore (1988b) argues that the central bank can only use price controls, not quantity controls. The central bank sets the price of reserves through its discount policy. Short term market interest rates are set as a markup over the Fed funds rate, which is determined by the central bank’s discount and open market policies (Moore 1988a). The money supply curve is then horizontal at that market interest rate. I believe Moore has overstated the ability of the Fed to exogenously set the discount rate, has underestimated its ability to use quantity constraints, and has misspecified the money supply function. Each of these points will be discussed below. 18. Minsky has defined a speculative unit as one which has financed a position in an asset which is expected to generate a stream of

97

earnings more than sufficient to retire the debt. However, in the near term, the earnings stream is only sufficient to make interest payments: principle cannot be retired until some future date. A Ponzi unit has financed a position in an asset which does not even generate sufficient income to make interest payments in the near term, so that the interest is capitalized. A rise in interest rates can force a speculative unit to become a Ponzi unit. (See Minsky 1986b.) 19) See Minsky (1986b) for an analysis of the role government deficits can play in fueling an expansion. 20. Moore (1988b) argues that the base is a stable function of the money supply. The Fed chooses the price at which it supplies reserves and then passively supplies them on demand. However, as I will show in Chapter 7, this is probably an oversimplification. The percentage of total bank reserves supplied by the Fed, as well as the percentage of an increase in bank reserves which is supplied by the Fed vary greatly from year to year. In contrast to Moore, I argue that the Fed has much more discretion regarding the quantity of reserves it will supply during ‘normal’ periods. However, like Moore, I believe that when

OM

the stabilityof the financial system is threatened, the Fed must abandon quantity constraints. See the discussion of Moore’s approach in Chapter 5 below. . I will discuss the determination

of interest rates in more

detail in

Chapter 6 below. This is not meant to imply that money supply and demand functions are independent, however. A shift of money demand is likely to be associated with a shift of the money supply function. Thus, these money supply and demand curves are not to be thought of as rigorously defined functions.

98

4

HISTORY OF THE ENDOGENOUS MONEY APPROACH

Brief overview of the history of the approach In the previous chapter, I presented my interpretation of the endogenous money approach, and briefly contrasted it with the currently dominant, exogenous money approach. However, the endogenous money approach did not suddenly appear with Post Keynesians during the past couple of decades. In fact, the origins of the endogenous money approach can be traced back at least 150 years. In this chapter I will very briefly present a history of the endogenous approach to monetary theory. This chapter is not meant to be a comprehensive survey of all those who worked within this approach, but is meant to suggest that the approach is not new. I will argue that while the Classical economists admitted that the relation between money and nominal values was a two way relation, their analysis did not lead to a theory of an endogenous money supply. The Banking School, as represented by Tooke, came much closer to an endogenous approach. Marx essentially adopted an endogenous approach to money, however, his work has not been of major importance in further development of the approach.' Keynes argued in the Treatise on Money and elsewhere that the money supply is endogenously determined, but may have assumed that it was exogenously given in his exposition in the General Theory. An exogenous money supply was adopted in the neoclassical synthesis (the textbook version of Keynes). The quantity theory of money was revived with the development of Monetarism, also with an exogenously determined money supply. Thus, the two major schools (textbook Keynesian and Monetarist) have long accepted an exogenous money supply as common ground, so that debate has been almost entirely over issues which follow from an exogenously determined money supply. In the last section of this chapter, I examine Schumpeter’s theory of the relation among credit, development, and innovation. Development is based on innovations which occur in search of profits, while development is funded on the basis of credit creation.

99

Financial institutions also innovate to reap profits, and in doing so, they find cheaper ways to finance the innovations of nonfinancial institutions. Thus, not only does Schumpeter show how a rise in investment is funded by credit (which is related to Keynes’s argument that investment cannot be constrained by saving), but he also relates credit creation to profit seeking behavior of capitalists. In the next chapter, I will examine

the modern

‘revival’ of an

endogenous money approach, which can be traced to Kaldor, the Radcliffe Committee, and Minsky. I will then examine the work of Tobin, of Gurley and Shaw, and of several Post Keynesians, including Moore and Lavoie.

The Currency School and the Banking School There are many similarities between the Currency School-Banking School controversy and the debate between Monetarists and Keynesians. I will show that the Currency School position is similar to that of Monetarists, while the position of the Banking School is close to that of Keynesians, particularly to that of Post Keynesians. These similarities have been discussed by Cramp (1962, 1970), but he did not present the arguments of the Banking School in great detail. The Currency School included Lord Overstone, Robert Torrens, and George Warde Norman, while the Banking School was represented by Thomas Tooke, John Fullarton, and John Stuart Mill (Cramp 1962). The debate centered on the relation between money and spending, on the ability of the Bank of England to control the quantity of money, and on the relation between the quantity of money and England’s external balance. The debate took place during the 1830s and 1840s, but before analyzing the debate, I will first present some background information. Hume analyzed the relation between prices and the money supply. Like other Classical economists, his analysis focused on the effect of a change in prices on the trade balance. Suppose four-fifths of all the money in Great Britain to be annihilated in one night, and the nation reduced to the same condition, with regard to specie as in the reigns of the Harrys and Edwards, what would be the consequence? Must not the price of all labour and commodities sink in proportion, and everything be sold as cheap as they were in those ages? What nation could then dispute with us in any foreign market, or pretend to navigate or

100

to sell manufactures at the same price, which to us would afford

sufficient profit?

In how little time, therefore, must this bring

back the money which we had lost, and raise us to the level of all the neighbouring nations? Where after we have arrived, we

immediately lose the advantage of the cheapness of labour and commodities; and the farther flowing in of money is stopped by our fulness and repletion. Again, suppose, that all the money of Great Britain were multiplied fivefold in a night, must not the contrary effect follow? Must not all labour and commodities rise to such an exorbitant height, that no neighbouring nations could afford to buy from

us; while their commodities,

on the other

hand, become comparatively so cheap, that...they would be run in upon

us, and our money

flow out; till we fall to a level with

foreigners, and lose that great superiority of riches, which had laid us under such disadvantages?’ (Hume 1898, p. 333)

_ Thus, a decline in the money supply causes prices to fall and generates a trade surplus. This causes a specie inflow until the money stock and prices have returned to their original level. On the other hand, an increase in the money supply raises the price level and results in a trade deficit. Gold flows out of the country until the money supply and price levels are restored to their original levels. Ricardo extended this argument to include paper money. If the quantity of money were to be increased: [These notes] would be sent into every market, and would everywhere raise the price of commodities, till they were absorbed in the general circulation. It is only during the interval of the issue, and their effect on prices, that we should be sensible

of an abundance of money; interest would, during that interval be under its natural level; but as soon as the additional sum of notes

or of money became absorbed in the general circulation, the rate of interest would be as high. (Ricardo 1951, Vol. 3, p. 91)

Hume

and Ricardo

did not offer a detailed

analysis of the

working of the money market mechanism, so that their analyses of

the relation between money and spending were inadequate. There is no presentation showing how the money supply is increased or decreased. If the money supply is exogenously increased, this leads to a trade imbalance so that the money supply is restored at its original level. However, there is no theory regarding how the money supply attained the ‘correct’ level in the first place. Tooke and J.S. Mill did attempt to extend classical theory through an 101

analysis of the role of the banking sector in the economic system.

This led Tooke to reject the position of Hume, Ricardo, and of the

majority of his contemporaries that an increase in money directly leads to an increase in the price level. He argues, ‘the prices of commodities are not liable to be influenced in any perceptible degree by variations in the issues of the Bank, in a convertible

state of the paper’ (Tooke, 1848, Vol. III, p. 245).

Before this conclusion can be examined, it is first necessary to

discuss Tooke’s theory of money and to distinguish it from that of the Currency School. I will first look at Tooke’s argument that the supply of bank notes and other private credit issues is a function of demand. I will then examine Tooke’s distinction between private liabilities and those of the government. Tooke believed that the Currency School reached incorrect conclusions regarding the relation between money and prices because it ignored the essential differences between the ways in which government fiat money and private credit enter the economy. I will next discuss the main controversy of the period, which concerned the proper way to rectify a trade imbalance. The Currency School advocated tight money policy, while Tooke argued that this disrupts credit markets. Finally, I will look at the monetary policy advocated by Tooke, which included policy to stabilize interest rates and to ensure solvency of banks. Tooke and other members of the Banking School distinguished between paper money and paper credit. The term paper money referred to money issued by the government to purchase goods and services. It corresponds to what is now called government fiat money. Paper credit was the term applied to various liabilities issued by banks or issued by firms and accepted by banks. I will discuss this distinction in more detail below. The Banking School argued that paper credit is issued on demand. Bank credit is either created to satisfy the needs of depositors as they draw down deposit accounts to purchase consumption goods, or to facilitate production by allowing ‘circulation of capital’. [A]lI the transactions between dealers and dealers, by which are to be understood all sales from the producer or importer, through all the stages of intermediate processes of manufacture or otherwise to the retail dealer or the exporting merchant, are resolvable into movements or transfers of capital. (Tooke 1959, pp. 35-6)

Bank notes are merely paper credit, as Fullarton argues:

102

Now, what are bank-notes but a form of credit? What are they

but credit parcelled out into small and even sums, for greater convenience in circulation? Each note is simply a transferable acknowledgment of a debt due from the banker to the first recipient of the note, and which he (the banker) promises to pay on demand: the value of which it purports to convey is that which it enables the holder to command by sending it to the bank, from whence it issued, and no other.... (Fullarton 1969, pp. 36-7)

Bank notes are issued at the initiative of the depositor, and cannot be issued at the will of the bank. Bank notes are the debt of the bank, and can only be issued if someone wants to hold bank debt. Thus, bank notes functioned much as demand deposits function today: issued on demand. Therefore, it is difficult to conceive of an ‘excessive’ supply of bank notes, as I will discuss below. These positions of the Banking School are in direct contrast to those of the Currency School. Tooke distinguishes between the Banking School and the Currency School: It was held by most writers of any authority on the subject of the Currency, till within the last few years, that the purposes of a mixed circulation of coin and paper were sufficiently answered, as long as the coin was perfect, and the paper constantly convertible

into coin; and

that the only evils to be guarded

against by regulation, were those attending suspension of payment and insolvency of the banks, a large proportion of which blend an issue of promissory notes with their other business. This, in point of fact, is what is understood in general terms as the banking principle, and is that upon which our system of currency is constructed and conducted. But a new canon of currency has of late been promulgated by persons of no mean authority. According to these authorities, it is not sufficient that the bank notes should be at all times strictly convertible into coin, and that the banks, whether issuing or not issuing, should be solvent; they

consider that a purely metallic circulation (excepting only as regards the convenience and economy of paper), is the type of a perfect currency, and contend that the only sound principle of a mixed currency is that by which the bank notes in circulation should be made to conform to the gold, into which they are convertible, not only in value, but in amount.... A regulation of the issue of bank notes, in conformity with this doctrine, is now

understood to be designated as the Currency principle (Tooke 1959, pp. 1-2).

That is, the Currency School wishes to tightly constrain the issue

103

of bank notes, so that the number issued corresponds to the amount of bullion. The restriction of the issue of bank notes was

not primarily to ensure solvency of banks, but to constrain the amount of paper money issued. The arguments urged in favour of such separation [of the functions of bank note issue from the functions of a bank as a safe depository]

have,

as it should

seem,

made

considerable

impression on the public mind, and schemes founded upon this principle have been strongly pressed on the attention of government, on the ground not only of guarding against the danger of suspension and insolvencies of the issuers, but of imparting more confidence and stability to credit and trade, and of securing greater steadiness in prices, and thus obviating or abating the alternations of feverish excitement and the extreme of depression, which have prevailed under the existing system, and which are imputed to a neglect of the currency principle (Tooke 1959, pp. 2-3).

Thus, the Currency School, like the Monetarists of today, argued that control over the supply of money would tame or eliminate the business cycle, as I will discuss below.

The Banking School-Currency School controversy centered on the correct method of stemming an outflow of bullion. The Currency School accepted the position of Hume and Ricardo cited above that the money supply determines prices. A rise in the supply of money raises prices, causing a trade deficit and an outflow of gold. Thus, the external balance could be restored if the money supply could be controlled (Cramp, 1962, p. 8). Restricting the growth of the money supply would prevent inflation and trade deficits. To understand the controversy, it is necessary to examine first the difference between the two schools regarding the definition of money adopted. At this time, the most important liquid assets included gold coin; ‘Bank of England and country bank notes; bank deposits subject to cheque; and the "promises to pay" of firms and individuals, which might be embodied in a negotiable instrument (bills of exchange) or might consist in mere book entries (trade credit) (Cramp 1962, p. 8). The Banking School emphasized the ability of a wide range of assets to function as money. In particular, bills of exchange were capable of making payments. ‘It was common practice for B to sell goods to A, draw a bill which A accepted and B then discounted with C, who might use it to settle a debt to his creditor D and so on: it might be G or L or P who eventually presented the bill to A 104

(or his banker) for payment’ (Cramp 1962, p. 10). Mill argued that in the early part of an expansion, trade was financed largely by trade credit. Thus, the relation between Bank of England notes and ‘near-monies’, as well as the relation between Bank of England notes and spending, were not believed to be close. The Currency School, however, included as money only Bank of

England notes, specie, and country bank notes. Thus, if issues of these could be controlled, then the quantity of money could be constrained. The Currency School wished to restrain private bank note issues by requiring that banks hold specie or Bank of England notes in an amount equivalent to private bank note issues. Tooke argued that country bank notes are fundamentally different from government note issues (this is the difference between paper credit and paper money alluded to above), and that a country bank note is merely a debt which is not fundamentally different from other private debt issues.° In other words, either money should include

a wide range of liquid assets, or the term should be limited to the debts of the government plus gold. He argued that it made no sense to include one specific private debt, and then to exclude all other private debts from the definition of money. If the definition of money is merely to include those assets which function as a medium

of circulation, then a wide variety of credit instruments

must be included. However, the Banking School argued that private debts should be distinguished from government debt: [W]hen, in the progress of society, credit comes to perform an important part in all mutual dealings, and in the great majority of transactions supersedes the necessity for this interchange of equivalents, an entirely new principle is brought into play, and one governed by distinct laws. Credit becomes then the legitimate substitute for money, but, in all its modifications and phases, it is distinguished by a broad and impassable line from money itself. The real question, then to be considered is, not whether this or that particular form of credit be entitled to the designation of ‘money’, but whether, without a perversion of terms and an outrage of principle, that denomination can be applied to credit in any shape. (Fullarton 1969, p. 36)°

Thus, credit substitutes

for ‘money’,

but is distinguished

from

what is now called fiat money. The Currency School, however confused ‘Bank notes, strictly convertible into coin, with a compulsory and inconvertible paper currency [fiat money]... This erroneous impression arises from a neglect of the consideration of 105

the difference in the manner and purpose of the issue’ (Tooke, Vol. IV, p. 175). Tooke goes on to distinguish government fiat money from private debt issue: It is quite clear that paper created and so paid away by the government, not being returnable to the issuer, will constitute a fresh source of demand, and must be forced into and permeate all the channels of circulation. Accordingly, every fresh issue beyond the point at which former issues had settled in a certain rise of prices and of wages, and a fall of the exchanges, is soon followed by a further rise of commodities and wages, and a fall of the exchanges; the depreciation being in the ratio of the forcibly increased amount of the issues. It will hence appear that the difference between paper money so issued, and bank notes such

as those

of this country

consists,

not

only in the limit

prescribed by their convertibility to the amount of them, but in the mode of issue. (Tooke, Vol. IV, p. 177)

Thus, government paper money is like Friedman’s helicopter money. As I discussed above in Chapter 2, this indicates that the government has not yet become a member of the giro, so its paper money is not yet true fiat money. (By this time, however, Bank of England notes surely were fiat money - Tooke’s description would apply to the debasements which had occurred a century earlier.) However,

Tooke

does

not have

this in mind,

but seems

to be

simply arguing that government demand is always excess demand. On the other hand, bank liabilities are issued on demand because

someone wants to hold or spend them. Bank liabilities do not represent an additional demand because they are merely debt. The quantity [of bank notes], therefore, is an effect and not a cause of demand. A compulsory government paper, on the other hand, while it is in the course of augmentation, acts directly as an

originating cause on prices and incomes, constituting a fresh source of demand in money.... (Tooke, Vol. IV, p. 177)

In summary, the Banking School position was that private credit is issued on demand and substitutes for money (narrowly defined as fiat money) to facilitate both circulation of final goods and to facilitate production by allowing circulation of intermediate goods among producers, importers, and retailers. Private credit cannot directly cause prices to rise because it is issued merely on demand to facilitate circulation, and does not represent an additional monetary demand.’ In contrast, however, government fiat money

106

(paper money) does represent an additional demand.

It differs

from private debt issue because it represents an asset, but not a

true liability of the government which ever needs to be retired.

Private

debt, however,

is retired.

Bank

liabilities

are

issued

to

finance relatively short term assets, and will be retired in a relatively short time. The quantity of bank liabilities which remain in circulation is decided by the holders of bank debt, and not by banks.* Since the quantity of bank liabilities outstanding is demand determined, it can never be excessive. Thus, it can’t directly cause prices to rise. Monetarism, like the Currency School, confuses bank liabilities

with fiat money. Both argue an issue of bank notes (or demand deposits) raises prices once holders have decided their real balances are excessive. However, the Banking School cannot ‘conceive any means whatever by which the circulation could be so augmented...’ (James Wilson, quoted by Tooke, Vol. IV, p. 195). Furthermore, the Banking School rejects the belief that banks could directly affect the quantity of bank notes outstanding, as this would be determined by the rate of ‘reflux’, which is determined by

the holders of bank notes. This law operates

Tooke describes his ‘law of reflux’:

in bringing back to the issuing banks the

amount of their notes, whatever it may be, that is not wanted for

the purposes which they are required to serve. The reflux takes place chiefly in two ways: by payment of the redundant amount to a banker on a deposit account, or by the return of notes in discharge of securities on which advances have been made. A third way is that of a return of the notes to the issuing bank by a demand for coin (Tooke, Vol. IV, p. 186).

His law of reflux helps to further distinguish private debt issues from government fiat money. Private institutions are limited in the quantity of debt they can issue due to reflux, while the government is not ‘because there is no reflux’ (Tooke, Vol IV, p. 186). However, if private institutions issue more debt than the public needs to facilitate circulation, reflux will increase. If the needs of

circulation increase, private debts will be issued to meet these

needs.

Thus,

control

over

bank

notes

alone

would

not control

private debt issues, as other forms of debt would be issued in their place to facilitate circulation. The Currency School responded with the argument that trade credit must finally be retired using gold or Bank of England notes, thus, a stable proportion must exist between ‘near-monies’ and 107

money,

narrowly

defined.

Therefore,

‘to restrict

the

flow

of

stock of monetary demand it was necessary to reduce the whole money the ing liquid assets, and this could be done by squeez supply [narrowly] defined’ (Cramp 1962; piel): y In Tooke’s scheme, this attempt would neither affect the quantit If prices. falling in result nor of credit used to support circulation, to have would these issues, note its d the Bank of England reduce and be made up by expanded circulation of private issues of notes notes then ed, increas not were issues private of y quantit bills. If the and gold previously deposited in banks would be removed to meet the needs of circulation. In this case, the quantity of money available as ‘capital’ (deposits which could be intermediated by banks) would fall so that the quantity of money available as means of circulation could be restored. This would increase the rate of interest, leading to gold inflows. Equilibrium in the external account would be restored when the interest rate was reduced to its previous level and the money supply was increased to meet the needs of circulation and production. Thus,

the

difference

between

the

two

schools

concerns

the

mechanism through which ‘tight money policy’ works: the effect is through prices according to the Currency School, while it is indirect according to the Banking School. The Currency School is clearly the precursor of the modern quantity theory of money: tight monetary policy directly reduces prices. The Banking School position is a forerunner of Keynesian theory: monetary policy works indirectly through the interest rate. The Currency School advocated tight control over bank note issue in the belief that this would stabilize the economy by stabilizing prices. Tooke, however, argued that tight monetary policy actually works through destabilizing credit markets. To examine this argument, it is necessary to return to an examination of the external account and of the proper way to rectify a trade deficit. The Currency School believed that constriction of the money supply (bank notes) would lower prices and restore a favorable trade balance. The Banking School, however, argued that constriction of ‘paper credit’ couldn’t directly affect prices (although constriction of ‘paper money’ could). Tight monetary policy would stem a trade deficit only by raising interest rates. Tooke argued that this policy would be very disruptive of credit markets because large swings in interest rates would be necessary to rectify trade imbalances. However, the Currency School was able to prevail on policy 108

makers, and tight monetary policy was adopted in the Act of 1844, as discussed above in Chapter 2. Tooke argued: As the result of a careful examination of the principle on which the Act of 1844 was founded, and of the experience of its working since the time when it came into operation, I have no hesitation in giving it as my opinion that it is a total, unmitigated, uncompensated,

and, in its consequences,

a lamentable failure.

(Tooke, Vol. IV, p. 402)

The variations, in the rate of interest, during the year 1847, have exceeded in frequency and extent any of which there is to be found an example in the commercial history of this country. (Tooke, Vol. IV, p. 400)

The alternative policy advocated by Tooke included ‘a total abrogation of those provisions of the Acts of 1844 and 1845, which limit the amount of note circulation’ (Tooke, Vol. IV, p. 402). Rather than using tight monetary policy, Tooke advocated the use by the Treasury of bullion outflows whenever a trade deficit occurred. Of course, this would only be a temporary measure, but he believed that it would allow lower volatility of interest rates. In order to free bullion for this purpose,

it was

necessary to allow

expansion of bank notes (and other private debt issues) to meet the needs of circulation. Therefore, restrictions on bank note issue had

to be removed to free bullion for use in the external account. Finally, the proper role for the central bank, according to the Banking School, is to maintain solvency of banks and stable credit

market conditions.’ Tooke argued: [I]t is only through the rate of interest and the state of credit, that the Bank of England can exercise a direct influence on the foreign exchanges.... [T]he greater or less liability to variation in the rate of interest constitutes, in the next degree only to the

preservation of the convertibility of the paper and the solvency of banks, the most important consideration in the regulation of our banking system. (Tooke 1959, p. 124)

These certainly sound like ‘Keynesian’ propositions. The central bank should abandon quantity constraints and concern itself with interest rates and bank solvency. The Banking School is also ancestrally related to the endogenous money approach because it focuses on the use of credit to finance expenditure. In particular,

the Banking School, like the endogenous money approach, rejects 109

the simple notion that tight money policy would directly decrease the quantity of money (broadly defined), and rejects the notion that a decline in the quantity of money would lower prices. The quantity of money privately issued can never be ‘excessive’ because the needs of circulation determine how much money actually circulates. Any quantity of private debt issue above this would increase ‘reflux’ and thereby destroy money. Tight money policy only operates indirectly through interest rates, and, by disrupting credit markets, through aggregate demand.

Marx and endogenous money Marx analyzed the Banking School-Currency School controversy in great detail. He rejected the theories of the Currency School, such as the notion that an increase in the money supply increases prices. However, he accepted some aspects of the theory of the Banking School, which he considered the top monetary theory of that period. He also criticized the Banking School because of the absence of a theory of value (and surplus value) in the approach. Much of his criticism is not directly relevant to the endogenous money approach, so I will only briefly examine Marx’s system and his critique of the Banking School. Marx’s generalized system can be depicted as m-m-c...p...c’-m’-m”’. In the first step (m-m), the money capitalist alienates his money by lending it to the industrial capitalist. In the second step (m-c), the industrial capitalist purchases commodities (for example, labor and raw materials), which are used in the third step (c...p...c’) to produce commodities (c’) with different use values for sale. In the fourth step (c’-m’), the industrial capitalist sells the produced commodities, and where m’>m, the industrial capitalist repays his loan from the money capitalist. In order for interest to be paid, it is necessary that m’”>m. Thus, m” is paid to the money capitalist,

so that if m’>m”, the difference is profit to the industrial capitalist. (Thus, m’ less m” is received by the industrial capitalist as profit, while m” less m is interest received by the money capitalist.) The money capitalist will only loan money (m) if he believes it will be repaid with interest (m”>m). This requires that produced commodities (c’) be sold at a price high enough to enable the industrial capitalist to repay the loan with interest. Of course, the industrial capitalist will not undertake the project unless he believes his net profit (m’ less m’”) will be sufficiently high.

110

Production is undertaken to realize profits through producing goods and services for sale. Ultimately, profits must be realized in money form, rather than in stocks of unsold goods. Thus, capitalist production is necessarily a monetary production: the capitalist must produce goods for the market to earn money to retire debt. Foley (1989) argues that in capitalism, money measures the value of commodities in a universal equivalent. It also functions as a unit of account - as the unit in which promises to pay are measured. Financing determines which agents will be able to carry out planned expenditures. An agent which desires to undertake deficit spending issues his own liability, which entails a promise to pay later in the form of a third party liability which is higher in the ‘debt pyramid’. That is, one’s own liability is issued as one goes into debt, but normally the debt can only be retired through the use of another’s liability. Individuals might retire debts by using bank liabilities, while banks retire debts by using government liabilities. The debtor must operate in the market to obtain these third party liabilities so that debt, whose value is denominated in the universal equivalent (money), may be retired. Of course, capitalist accumulation (which must take the form of accumulation of the means to further accumulation) requires that profits remain after debt is retired. Finally, successful accumulation is measured in terms of the universal equivalent, money.

Marx believed that the supply of credit normally expands to meet the needs of economic activity. He noted that a large portion of loans are created for the purpose of transforming capital in physical form (including inventories) into liquid form until the physical capital can be liquidated (through depreciation or sale) and the credit be retired. The ordinary businessman

discounts, in order to anticipate the

money-form of his capital and thereby to keep his process of reproduction in flow; not in order to expand his business or secure additional capital, but in order to balance the credit he

gives by the credit he receives. (Marx 1972b, p. 525)

Marx argues that most of the credit created is not used in final purchases, but is used in intermediate stages. He quotes Coquelin affirmatively: In every country the majority of the credit transactions takes place in the circle of the industrial relations themselves...the producer of the raw material advances it to the capitalist, who

111

works it up, and receives from him a promise to pay on a certain day. The manufacturer, having completed his share of the work, in his turn advances his product on similar conditions to another

manufacturer, who has to manipulate it farther, and in this way credit extends more and more, from one to the other, down to the consumer....All borrow with one hand and lend with the other,

sometimes money, but more frequently products. In this manner an incessant exchange of credits, combining and crossing in all directions, takes place in the industrial relations. The development of credit consists precisely in the multiplication and growth of these mutual credits, and here is the real seat of its power. (Coquelin, from Du Credit et des Banques dans LIndustrie, 1842, quoted in Marx 1909, p. 472) Thus, most credit takes the form of working capital, as discussed

above in Chapter 2. In the absence of credit, the industrial capitalist is constrained by the quantity of money hoards (either his own or those of monetary capitalists). Marx argued that credit is the creation of industrial capitalists to free them from the necessity of accumulating money hoards or from borrowing from those who do. The creation of financial institutions allows the concentration of money capital which can be used as the basis of credit

extension.”

Production on a large scale and for distant markets throws the total product into the hands of commerce; but it is impossible, that the capital of a nation should be doubled in such a way, that commerce by itself would be able to buy up the entire national product with its own capital and to sell it again. Credit is, therefore, indispensable here. Credit must grow with the growing volume of value in production, and it must grow in the matter of time with the increasing distance of the markets. A mutual interaction takes place here. The development of the process of production extends the credit, and credit leads to an extension of industrial and commercial operations. Looking upon this credit separate from banking credit, it is evident that it grows with an increasing volume of industrial capital itself. (Marx 1909, pp. 564-

5)

The creation of credit allows a small amount of money capital to support a large circulation because money capital is only required when credit must be settled, and only for the amount of net credits. As Marx put it: It is here indisputably true that the 1,000 p.st. [pounds sterling]

I?

which I deposit to-day with A are issued to-morrow and deposited with B. The day after to-morrow it may be issued once more by B and form a deposit with C, and so forth infinitely. The same 1,000 p.st. of money may, therefore, multiply themselves into an absolutely indeterminable sum of deposits by a series of transfers. Hence it is possible that nine-tenths of all deposits in England may have no other existence but that in the entries of the banker’s books, of whom every one stands good for his part of them. (Marx 1909, p. 478)

Bankers are thus able to expand loans to a multiple of their reserves of money capital. The size of this multiple depends on the degree of development and organization of the banking system. Furthermore, the willingness of banks to extend loans also depends on their confidence in the future. Marx ‘stressed banks’ promptness to modify their money-reserve - either by refusing to make new advances or by giving bills drawn on the bank itself, instead of money, in payment to their customer - as soon as the degree of uncertainty as to future events increased’ (Panico 1982, p. 70). Thus, the size of the ‘deposit multiplier’ was variable.

Once credit has been created, it can function like money as a medium of exchange: Credit-money springs directly out of the function of money as a means of payment. Certificates of the debts owing for the purchased commodities circulate for the purpose of transferring those debts to others. On the other hand, to the same extent as

the system of credit is extended, so is the function of money as a means of payment. (Marx 1978, p. 139)

With the development of commerce and of the capitalist mode of production, which has an eye only to the circulation, this natural basis of the credit system is extended, generalised, elaborated. Money serves here on the whole merely as a means of payment, that is to say, commodities are not sold for money, but for a written promise to pay for them at a certain date. We may comprise all these promises to pay for brevity’s sake under the general category of bills of exchange... To the extent that they ultimately balance one another by the compensation of credits since no and debts, they serve absolutely as money, transformation into actual money takes place.... These do not rest on the circulation of money.... (Marx 1909, p. 470) Regarding the role of banks and bank notes, Marx said:

113

The loan is made (we refer here only to the commercial credit in the strict meaning of the term) by discounting bills of exchange, that is, by converting them into money before they come due, and by advances in various forms.... The credit given by a banker may assume various forms, for instance, that of exchanges on other banks, checks on them, opening of credit in the same way, finally, in the case of banks entitled to issue notes, the bank notes of the

bank itself. A bank note is nothing but a draft upon the banker, payable at any time to the bearer, and substituted by the banker for private drafts. This last form of credit appears particularly important and striking to the layman, first, because this form of credit money steps from the mere commercial circulation into the general circulation and serves as money there.... (Marx 1909, p.

474)

Thus, like Tooke, Ma . argues that bank notes are merely another private debt, but one which appears to be different from other private debts because of the degree to which it circulates. Marx disagrees. with the position of Tooke that bank notes cannot represent an additional demand on goods and services, thus, he rejects the strict dichotomy Tooke posited between ‘paper credit and paper money’. Marx argues that a bank loan may lead to an increased demand, just as fiat money would in the scheme laid out by Tooke: ‘the bank may open a credit account for A, in which case this A, a debtor of the bank, appears in the role of an imaginary depositor’ (Marx 1909, p. 540). Thus, the borrower receives an ‘imaginary’ deposit, which can be spent just as if it were money. Tooke’s mistake lies in the implicit assumption that all credit advanced by banks is on the basis of existing ‘capital’ - that is, merely to facilitate circulation.’ However, Tooke was correct in

emphasizing demand

that bank

credits can

rise only in response

to a

for bank credit. In contrast to Tooke, Marx emphasizes

that the quantity of notes issued need not be strictly based upon existing ‘capital’, so can represent an additional demand. Tooke emphasized only the role credit plays in circulating goods which have already been produced, while Marx also emphasized the role credit plays in allowing production to increase. According to Marx, the quantity of money is a function of the price level, velocity, and the quantity of transactions. All of these determinants of the quantity of money are variable. He argues that the theory that an increase in the quantity of money necessarily causes prices to rise is wrong. The quantity of money tends to increase or decrease with the needs of circulation. He says, ‘We have seen how, along with the continual fluctuations in the extent

114

and rapidity of the circulation of commodities and in their prices, the quantity of money current unceasingly ebbs and flows’ (Marx 1978, p. 134). He argues: The degree of the velocity of circulation, in other words, the

number of repetitions of the same function as means of purchase and payment by the same pieces of money in a given period of

time, the mass of simultaneous purchases and sales, or payments,

the sum of the prices of the circulating commodities, finally the balances of payments to be spared in the same period, determine in either case the mass of the circulating money, of currency. Whether the money so serving represents capital or revenue for the paying or receiving party, is immaterial, and does not alter the matter in any way. Its mass is simply determined by its function as a medium of purchase and payment. (Marx 1909, p.

527) Thus, the quantity of money

is not a cause, but an effect, of

spending. Attempting to control prices by reducing the quantity of money will not be effective, because the quantity of money only indirectly affects prices through the cost of production. He says, ‘in fact the reduction of the quantity of gold raises only the rate of interest, whereas an increase in the quantity of gold lowers the rate of interest; and were it not for the fact that the fluctuations of the

rate of interest are taken into account in the determination of costprices, or in the determination of demand and supply, the prices of commodities would be wholly unaffected by them’ (Marx 1909, p. 648). Like Keynes, Marx argues that changes in the quantity of money (even narrowly defined) only indirectly affect prices through changes in interest rates. Marx argued that the demand for credit could be highest during crisis. In a prosperity phase, capitalists can sell their commodities on the basis of bills of exchange. During the expansion, firms are willing to accept these bills since their liquidity is very high. However, during a downturn, firms attempt to discount the bills, so that the demand for money capital (narrow money) rises. According to Marx, ‘In time of crisis, the demand for loan-capital,

and therefore the rate of interest, reaches its maximum’ (Marx 1972a, p. 513). ‘In times of crisis loans are taken up for the purpose of paying, not of buying; for the purpose of winding up

previous transactions, not of beginning new ones’ (Marx 1909, p. 542). Furthermore, in a crisis, ‘the circulation of notes as a means

of purchase is decreasing’ even though ‘their circulation as means bg)

of payment may increase’ and ‘the aggregate amount of the circulation...may remain stable or may even decrease’ (Marx 1909, p. 542). In other words, in a crisis, the aggregate quantity of debts falls as debts are retired and because fewer new debts are taken out. He argues that although the demand for credit will be high in a crisis, that demand won’t necessarily be met. In contrast, high demand for credit in an expansion will be met. Thus, high demand for credit cannot be taken as an indication of high expectations: ‘It is by no means the strong demand for loans...which distinguishes the period of depression from that of prosperity, but the ease with which this demand is satisfied in periods of prosperity, and the difficulties which it meets after a depression has become a fact’ (Marx 1909, p. 532). Marx’s position shares several characteristics with the endogenous money position outlined in previous chapters. During an expansion, the basis of most expenditure is credit expansion. The development of financial institutions reduces the cash reserves which must be held against debt commitments. Debt instruments are normally sufficiently liquid to enable circulation to proceed. In an expansion, credit ‘money’ expands to meet the needs of circulation. The money supply is not exogenously determined, and does not limit production and circulation, since various credit arrangements allow

the creation of ‘near monies’. There is no fixed ratio of ‘near monies’ to reserves (cash, or high powered money in modern parlance). In an expansion, small reserves are required to maintain the liquidity of credit instruments such as bills of exchange, while in a crisis large reserves may be required to meet attempts at liquidation. Because an expansion is based on an increase in the quantity of credit relative to cash, it is impossible to exchange all credit for cash during a downturn. Hence, ‘It is precisely the enourmous [sic] development of the credit system during a period of prosperity, hence also the enourmous development of the demand for loan capital and the readiness with which the supply meets it in such periods, which brings about a shortage of credit during the period of depression’ (Marx 1909, p. 532). The money supply (and even types of credit instruments which function as money) varies endogenously with the business cycle.

Keynes and endogenous money Keynes’s conception of money

is consistent with the definition

116

given in Chapter I above. He occasionally seemed to realize that the quantity of money is endogenously determined, but he apparently adopted an exogenously determined money supply in the General Theory. It has been argued (Foster 1986a, 1986b) that his adoption of an exogenous money supply seriously weakened the arguments in the General Theory.” However, I do not intend to Show how the General Theory could be strengthened by the inclusion of endogenous money. In Chapter 1, I emphasized the role of money as a unit of account, as did Keynes in his writings on ancient currency: An article may be deemed to have some at least of the peculiar characteristics of money 1) if it is regularly used to express certain conventional estimates of value such as religious dues, penalties or prizes, or 2) if it is used as the term in which loans and contracts are expressed or 3) if it is used as the term in which prices are expressed, or 4) if it is used as the habitual medium of exchange. In the first three cases the article in question is the term in a money-of-account, in the fourth case it

is used as actual money. Now for most important social and economic purposes what matters is the money of account; for it is the money of account which is the subject. (Keynes 1982, p. 252)

In the Treatise on Money, Keynes said: ‘Money proper in the full sense of the term can only exist in relation to a money of account...[W]e may elucidate the distinction between money and money of account by saying that the money of account is the description or title and the money is the thing which answers to the description’ (Keynes 1971b, p. 3). He goes on to distinguish among various forms of money which fit the description. Money ‘proper’ is defined as state money plus those private forms of money which are accepted by the state. Thus, as discussed above, when the state becomes a member of the bank giro by accepting payment of bank notes or deposits, bank money achieves a special status, which Keynes calls money ‘proper’. Other forms of money which fulfill the

unit

of account

function

are

merely

‘offers

of contracts,

contracts and acknowledgments of debt’, which exist ‘side by side’ with money proper (ibid., p. 5). Furthermore, ‘for many purposes the acknowledgments of debt are themselves a serviceable substitute for money proper in the settlement of transactions’ (ibid.,

p.5). Keynes

|

|

later criticized the orthodox view of money as a medium 117

of exchange by emphasizing the second protection against an uncertain future.

major role of money:

Why should anyone outside a lunatic asylum wish to use money as a store of wealth? Because, partly on reasonable and partly on instinctive grounds, our desire to hold money as a store of wealth is a barometer of the degree of our distrust of our own calculations and conventions concerning the future... The possession of actual money lulls our disquietude.... (Keynes 1973,

pp. 116-7)

Money has real effects because the premium required to induce people to part with liquidity affects the price of capital assets, which then affects the level of employment and output.” [T]he rate of interest is the reward for parting with liquidity for a specified period.... It is the ‘price’ which equilibrates the desire to hold wealth in the form of cash with the available quantity of cash.... (Keynes 1964, p. 167)

[I]t is the greatest of the own-rates of interest...which rules the roost (because it is the greatest of these rates that the marginal efficiency of a capital-asset must attain if it is to be newly produced); and...there are reasons why it is the money-rate of interest which is often the greatest.... (Keynes 1964, pp. 223-4)

Thus, money plays an important role in Keynes’s theory, where ‘the firm is dealing throughout in terms of sums of money. It has no object in the world except to end up with more money than it started with’ (Keynes 1979, p. 89). In fact, he calls his theory a ‘Theory of a Monetary Economy’, which is: ‘the theory of a system in which changing views about the future are capable of influencing the present situation. For the importance of money essentially flows from its being a link between the present and the future’ (Keynes 1964, p. 293). Optimism about the future lowers the liquidity premium, allowing production to proceed. A given quantity of money ‘provides a revolving fund for a steady flow of activity; but an increased rate of flow needs an increased stock to keep the channels filled’ (Keynes 1973, p. 230). Because a reduction in liquidity preference lowers the premium required to release idle liquid balances, production can expand initially with no increase in the money stock. Once idle balances have been released, the limit to the rise in

production depends on the quantity of finance, the propensity to

118

hoard, and the velocity of circulation. Keynes argues that spending can be constrained by shortage of finance, but never by a shortage of saving. In a monetary economy, saving is not equivalent to finance. Saving has no special efficacy, as compared with consumption in releasing cash and restoring liquidity.... Consumption does just as well; in fact it does

better, since an

increase

in wealth

may

involve a more or less proportionate increase in the inactive demand for cash. A given level of activity and income will involve the same active demand for cash, if the technical conditions governing the time lags are the same, irrespective of the current rate of net investment and saving - irrespective, indeed of whether there is any investment and saving.... There is, therefore,

just as much reason for adding current consumption to the rate of increase of new bank money in reckoning the flow of cash becoming available to provide new ‘finance’, as there is for adding current saving. (Keynes 1973, p. 233)

In other words, consumption is as much a part of loanable funds as is saving, as long as funds spent on consumption find their way into the banking system. Furthermore, since all expenditures are received as income, and because investment goods are unavailable for consumption, that part of income not spent on consumption goods must be equal to expenditures on nonconsumables. The supply and demand of ‘loanable funds’ are necessarily identical, with no adjustment in the interest rate required: Increased investment will always be accompanied by increased saving, but it can never be preceded by it. Dishoarding and credit expansion provides not an alternative to increased saving, but a necessary preparation for it. It is the parent, not the twin, of increased saving. (Keynes 1973, p. 281)

Keynes goes on to argue that while investment can never be constrained by lack of saving, it can be constrained by lack of finance. He argues that firms may attempt to accumulate a cash balance before undertaking investment. He says: [U]nless the banking system is prepared to augment the supply of money, lack of finance may prove an important obstacle to more than a certain amount of investment decisions being on the tapis at the same time. But ‘finance’ has nothing to do with saving. At the ‘financial’ stage of the proceedings no net saving has taken place on anyone’s part, just as there has been no net

119

investment. (Keynes 1937, p. 247)

So by ‘finance motive’, Keynes means a propensity to hoard cash in preparation for funding an investment project. This must be added to other motives for hoarding. The intersection of the demand for hoards and the supply of hoards then determines the interest rate. A rise in investment doesn’t directly affect the interest rate, which is only determined by the propensity to hoard. However, if firms accumulate a hoard in preparation for investing, this does cause interest rates to rise unless someone meets this additional demand for a hoard. There has, therefore, to be a technique to bridge this gap between the time when the decision to invest is taken and the time when the correlative investment and saving actually occur. This service may be provided either by the new issue market or by the banks... If investment is proceeding at a steady rate, the finance (or the commitments to finance) required can be supplied from a revolving fund of a more or less constant amount, one entrepreneur having his finance replenished for the purpose of a projected investment as another exhausts his on paying for his completed investment. But if decisions to invest are (e.g.) increasing, the extra finance involved will constitute an additional demand for money.... (Keynes 1973, pp. 208-09)

To summarize, an increase in investment can be funded by credit (or releases of hoards). As investment occurs, this creates saving. Since most credit advanced by banks to fund investment spending returns to banks (the ‘revolving fund’), no pressure is placed on interest rates. However, if part of the credit advanced by banks ends up as hoards (for example, to satisfy the finance motive), then investment can be constrained by lack of finance. Either banks will have to provide for this extra money which is hoarded, or interest rates will rise to decrease the desire to hoard sufficiently that the existing quantity of hoards satisfies the desire to hoard. There is some confusion in Keynes’s presentation. He correctly argues that saving is a result of investment, so cannot be a net source of funds available to finance investment. He also correctly argues that the interest rate cannot be determined by the intersection of the saving supply and investment demand curves. He argues that it is liquidity preference, or the demand and supply of hoards, which determines

interest rates. When

investment

is

proceeding at a steady pace, a revolving fund of given size is adequate

to finance

that

level

of investment,

120

and

investment

generates an equivalent amount of saving. When investment increases, it may place pressure on interest rates - not because it is necessary to increase the quantity of funds available to finance investment, but to provide the hoards necessary to satisfy a ‘finance motive’. He argues that because investment, itself, can be financed

out of a revolving fund, it does not place pressure on interest rates. Only money ‘sitting still’ (hoards) affects interest rates; money ‘on the wing’ does not. The problem with his analysis arises because he has not adequately explained how the rise of investment is funded. In his model, if investment were to increase without requiring an increase in the hoards necessary to satisfy the ‘finance motive’, investment would be funded with no impact on interest rates. Indeed, Keynes is not even clear whether the increase in

investment would have any impact on the money supply.” Keynes argued several times that the money exogenously fixed. For example, he argued:

supply is not

The banking system has no...direct control over the quantity of money; for it is characteristic of modern systems that the central bank is ready to buy for money at a stipulated rate of discount any quantity of securities of certain approved types. (Keynes 1971c, p. 189) The volume of the paper money, on the other hand, would be consequential,

as it is at present,

on

the state

of trade

and

employment, bank-rate policy and the Treasury bill policy. (197 1a, peels3)

In the Treatise on Money, he recognized that banks create money as they make loans: a bank ‘may itself purchase assets, i.e. add to

its investments, and pay for them, in the first instance at least, by establishing a claim against itself (Keynes 1971b, p. 21). Although banks would then lose reserves during a clearing drain, Keynes recognized that ‘it is evident that there is no limit to the amount of bank money which the banks can safely create provided that they move forward in step’ (ibid., p. 23), since losses through the clearing drain would be compensated by gains. However, he concluded that ‘the rate at which an individual bank creates on its own initiative is subject to rules and limitations’ such as the rate of growth of other banks, rules of thumb and laws concerning reserve ratios, and ‘the aggregate of their reserve resources’ (ibid., pp. 26-7). Thus, the money supply depends on a number of variables and certainly should not be taken as exogenously fixed. 121

Furthermore,

in a very interesting passage, Keynes examined

‘overdraft facilities’, in which

credit is, by nature,

advanced

on

demand. [T]o the extent that the overdraft system is employed and unused overdrafts ignored by the banking system, there is no superimposed pressure resulting from planned activity over and above the pressure resulting from actual activity. In this event the transition from a lower to a higher scale of activity may be accomplished with less pressure on the demand for liquidity and the rate of interest. (Keynes 1973, pp. 222-3) In other words, the existence of overdraft facilities reduces

the

hoards desired to satisfy the finance motive and automatically increases the supply of hoards as long as the banking system accommodates.

This, in turn, allows investment

to proceed with

little impact on interest rates. If Keynes had pursued his analysis of overdraft facilities, he might have adopted an endogenous money approach in which money is supplied on demand. Indeed, he later argued that banks provide the short term finance necessary to allow investment to proceed. As investment generates saving, an increase in investment can be permanently funded out of the

resulting saving.”

[T]he increment of current investment over prior investment (or saving) can only be cared for permanently out of the increment of current saving; and the period during which current savings are kept liquid by their owners must be bridged by an increase in the revolving fund of ‘finance’.... It is the role of the credit system to provide the liquid funds which are required first of all by the entrepreneur during the period before his actual expenditure, and then by the recipients of this expenditure during the period before they have decided how to employ it. (Keynes 1973, pp. 284-5)

While much of the above is consistent with an endogenous money approach, Keynes actually assumed an exogenous money supply in much of his analysis. For example, in the General Theory, he writes ‘the quantity of money as determined by the action of the central bank’ (p. 247), ‘the quantity of money created by the monetary authority’ (p. 205), and makes similar arguments in other places. (See Keynes 1964, pp. 84, 167, 174, 230, 267.) On the other hand, it seems that this assumption was made purely to facilitate the analysis. For example, he argues ‘we can draw the line between

122

"money" and "debts" at whatever point is most convenient for handling a particular problem’ (Keynes 1964, p. 167). He then goes on to say that he will arbitrarily define money as bank deposits. If money were defined to include a wider range of debts, then it would be clear that Keynes would not have accepted an exogenously determined money supply.” Joan Robinson argues that Keynes had to make every possible concession...in order to get a hearing. It would have been much simpler to start by assuming a constant rate of interest and a perfectly elastic supply of money. But then his whole case would have been dismissed as a misunderstanding of the orthodox position. He was obliged to accept the presumptions of his critics in order to explode them from within (Robinson 1971, pp. 81-82).

Thus, it is possible that Keynes merely adopted an exogenous money supply in order to facilitate exposition of various components of his General Theory. Much of his other work seems to recognize that the money supply is endogenously determined, and even the General Theory might be strengthened had Keynes consistently used an endogenous money supply. In Chapter 6 I will discuss how liquidity preference theory strengthens the endogenous money approach. Indeed, endogenous money combined with Keynes’s theories of effective demand and liquidity preference together provide the basis for a Post Keynesian general theory.

Schumpeter on credit and innovation Schumpeter studied the relation between credit and capital. He argued that credit is necessary for economic development because it gives purchasing power to entrepreneurs, enabling them to force a new employment of labor and other resources. Economic change in a captialist society requires that entrepreneurs innovate, and credit allows them to embark on new enterprises. Schumpeter argues that economic development consists of finding new ways to combine labor and natural resources. Great strides in development do not come from ‘adaptive responses’ but from ‘creative responses’, which escape current behavior patterns. The ‘creative response changes social and economic situations for good’ (Schumpeter 1951, p. 217). In a capitalist society, as opposed to a command or communal society, ‘the mechanisms of economic 123

change...pivot on entrepreneurial activity’ (ibid., p. 217). The drive for above-normal profits leads to innovation. Innovation generates ‘capital gain, that is, ...the discounted value of the stream of prospective returns’ (ibid., p. 222). The capitalists which innovate gain a competitive advantage, while ‘the impact of the new product or method spells losses to the "old" firms. The competition of the man with a significantly lower cost curve is, in fact, the really effective competition that in the end revolutionizes the industry’ (ibid., p. 223). In order to innovate, the entrepreneur must have purchasing power so that he can employ labor and resources. Schumpeter argues that credit is not necessary to maintain a circular flow in perfect equilibrium: the production of consumption goods just satisfies the demand for consumption goods, which is determined by the portion of income spent on consumption, while the portion of income not consumed just equals expenditures on the means of production required to produce the consumption goods. However, credit is required to enable production to increase, as the income arising from a circular flow in equilibrium cannot be large enough to enable spending to rise. Credit involves the creation of new purchasing power (Schumpeter 1934, Chapter 3). In a capitalist economy, ‘credit is essentially the creation of purchasing power for the purpose of transferring it to the entrepreneur’ (Schumpeter 1934, p. 107). The creation of this purchasing power is ‘the method by which development is carried out in a system with private property and division of labor’ (ibid., p. 107). Credit transforms the entrepreneur into a debtor to society: he receives resources and labor from society before he has contributed to society. He repays this debt to society only after engaging in production. Schumpeter defines capitalism as a form of economic organization in which ‘the goods necessary for new production are withdrawn from the circular flow by the intervention of purchasing power ‘created ad hoc’ through the creation of credit (ibid., p. 106). Credit is capital, which is ‘nothing but the lever by which the entrepreneur subjects to his control the concrete goods which he needs, nothing but a means of diverting the factors of production to new uses, or of dictating a new direction to production’ (ibid., p. 116). When the entrepreneur uses credit to purchase productive means, he surrenders ‘capital’ as it becomes incomes to those from whom he purchased the means of production. Capital is ‘that sum of means of payment which is available at any moment for

124

transference to entrepreneurs’ (Schumpeter 1934, p. 122). The means of payment are not merely ‘metal money’, but include claims to money which take the place of money (narrowly defined). Banks create money by issuing claims against themselves in the process of granting credit. The aggregate quantity of outstanding credit greatly exceeds the quantity of ‘metal money’. The ‘circulating media...exist in such quantities that they could not possibly all be redeemed at once...’ (ibid., p. 98). Credit is created to give purchasing power to the entrepreneur, but that credit then ‘increases the means of payment’ as it circulates (ibid., p. 99). When an entrepreneur goes to a bank for credit, he can offer securities or the goods he will produce as collateral. In the second case, the credit comes before the collateral actually exists. Thus, in

reality, the quantity of outstanding credit exceeds the quantity which is ‘fully covered’ by securities or goods functioning as collateral. The limit on credit expansion by banks is determined by the necessity that banks make profits. Banks are subject to three constraints, first, they must hold reserves in case the expected production of goods does not occur and the entrepreneur defaults on his debt

(a loan-loss

reserve);

second,

banks

cannot

create

coined money for transactions, so must maintain a reserve to meet the demand for currency; and third, credit inflation can lead to an

outflow of gold (Schumpeter 1934, pp. 112-4). This last point will require some explanation of ‘credit inflation’. Because credit creation gives new purchasing power before new goods are produced, it necessarily generates credit inflation.” Just as when additional gas streams into a vessel the share of t* _ space occupied by each molecule of the previously existing gas 1s diminished by compression, so the inflow of new purchasing power into the economic system will compress the old purchasing power. When the price changes which thus become necessary are completed, any given commodities exchange for the new units of purchasing power on the same terms as for the old, only the units of purchasing power now existing are all smaller than those

existing before and their distribution among individuals has been shifted. This may be called credit inflation. (Schumpeter 1934, p. 109)

However, this credit inflation is normally only temporary: After completing his business - in our conception, therefore, after a period at the end of which his products are on the market and his productive goods used up - he has, if everything has gone

125

according to expectations, enriched the social stream with goods whose total price is greater than the credit received and than the total price of the goods directly and indirectly used up by him. Hence the equivalence between the money and commoaity streams

is more

than restored, the credit inflation more

than

eliminated, the effect upon prices more than compensated for, so that it may be said that there is no credit inflation at all in this case - rather deflation - but only a non-synchronous appearance of purchasing power and of the commodities corresponding to it, of inflation. which temporarily produces the semblance (amount debt his repay now can neur entrepre the ore, Furtherm credited plus interest) at his bank and normally still retain a credit balance (=entrepreneurial profit) that is withdrawn from the purchasing-power fund of the circular flow. Only this profit and interest necessarily remain in circulation; the original bank credit has disappeared... (ibid., pp. 110-11)

Schumpeter argues that the credit inflation is thus temporary as long as the investment projects are successful. In fact, he argues that innovation should lead to a secular fall in the price level as less costly production methods are employed. For the individual bank, the limit to its ability to create credit is

set by the constraint that credit inflation must be temporary. In other words, the credit it grants must be (eventually) retired. Credit inflation, if it were permanent, would ruin a bank (that is, if loans are never repaid). However, if all banks together continually

refinanced credit (making bad loans good by extending debt repayment schedules), permanent credit inflation would result which would not ruin the banks. In this case, aggregate price levels would continually rise, leading to trade deficits and gold outflows. Schumpeter thus argues that some legal restrictions might be required to prevent banks from operating to generate permanent credit inflation (ibid., pp. 114-15). Schumpeter argues that credit is unnecessary to maintain a circular flow, so is unnecessary in an economy which is not growing. Since he relates growth to entrepreneurial innovation, the essential function of credit is to allow development through innovation. He argues that the money market develops with the use of credit to finance innovation, but that the money market is not essential merely to circulate a given flow. And so the requirements of the circular flow are added to the entrepreneur’s demand in the money market on the one hand, and money from the circular flow increases the supply in the

126

money market on the other. Hence we feel in every money-market article the pulse of the circular flow.... But this must not prevent us from distinguishing the transactions in the money market which belong to the circular flow from others. Only the latter are fundamental; the former are added onto them,

and the fact that they appear in the money market at all is Se a consequence of development. (Schumpeter 1934, pp.

Financial institutions are also entrepreneurial organizations which Strive to innovate to generate capital gains. Financial institutions and practices enter our circle of problems in three ways: they are ‘auxiliary and conditioning’; banking may be the object of entrepreneurial activity, that is to say, the introduction of new banking practices may constitute enterprise; and bankers (or other ‘financiers’) may use the means at their command in order to embark upon commercial and industrial enterprises themselves. (Schumpeter 1951, p. 222)

A bank is a business, which makes profits by lending its promises to pay. Schumpeter speculates that three-fourths of a bank’s deposits are merely credits, and ‘there can be no doubt that these circulating media come into being in the process of granting credit and are created especially..for the purpose of granting credit’ (Schumpeter 1934, p. 98). He argues that an entrepreneur borrows what he deposits: bank credits. Like other entrepreneurs, those in the banking industry earn capital gains by innovating, that is, by finding new and less costly ways of granting credit. In credit markets, the entrepreneurs are the demanders and banks are the suppliers. The creation of credit is merely ‘the exchange of present against future purchasing power’ (the borrower obtains purchasing power today by promising to surrender purchasing power in the future) (ibid., p. 125). The main purpose of credit markets is to finance development. The financial institution earns profits out of the interest payments made by entrepreneurs which have borrowed and have successfully undertaken a project. The financial entrepreneur earns capital gains by finding innovative financing methods which provide purchasing power needed by other

entrepreneurs. In summary, credit plays a central role in a capitalist society

because it funds growth and innovation. Banks and other financial institutions provide this credit. Financial institutions are profit seeking organizations, whose only quantity constraints are

127

determined by the necessity that they earn profits. The most important determinant of profits is the necessity that ‘credit inflation’ be temporary: that is, that loans be repaid. Financial institutions which engage in ‘entrepreneurial activity’ earn above-normal profits (‘capital gains’) by innovating. Schumpeter’s analysis is consistent with the endogenous money approach. Banks create credit on demand, and this credit enables expenditure to proceed. Growth of net investment requires the creation of credit (such growth cannot be financed out of prior savings). Banks have an incentive to meet the demands for credit in order to earn profits. Banks which innovate can reduce costs and earn higher profits. Thus, money is created during the normal capitalist processes (including investment) by profit seeking financial institutions (lenders) to meet the demand of profit seeking nonfinancial institutions (borrowers).

Notes 1. This probably says more about familiarity with Marx’s work than it does about the relevance of his work to the endogenous money approach. 2. Mints (1945) and White (1984) have also examined the Banking School-Currency School debate. White identifies a third party to the debates: the Free Banking School. This school displays some aspects of the endogenous money approach. White characterizes each school by its belief concerning the possibility of an overissue of notes. According to the Currency School, any bank can overissue, while the Banking School claims no bank can overissue (due to reflux). According to the Free Banking School, only the central bank can overissue (due to its monopoly status) (White 1984, p. 53). 3. However, Hume at one time seemed to argue that the quantity of money could not affect prices. Tooke said: From the tenour of the questions which were put to me and other witnesses by Mr. Hume, it might be inferred that he was of opinion that bank notes and deposits, as they conferred a power of purchase, were operative as a cause of variations of prices. And it is possible that such was then his opinion; but, not long

before, he seems to have entertained a very different view.... [On] May 12. 1836, Mr. Hume said, ‘With regard to the amount of paper issues, I think I could show that it would be impossible to issue too much paper money, if it were made convertible into gold in every part of the kingdom on demand. Indeed, I believe

128

that great delusion exists in the country with regard to the effect on prices of the currency. My opinion is, that the quantity of money depends on the rise of prices; and that the rise of prices does not depend on the quantity of money. I hold the prevailing doctrine to be extremely erroneous on this point’. (Tooke Vol. IV, p. 174) If Hume had maintained this point of view, he would have sided with the Banking School rather than with the Currency School. See below. See Panico (1982, p. 20). a Tooke recognized that credit money is merely debt: Credit, in its simplest expression, is the confidence which, well or ill-founded,

to entrust

leads a person

another

with a certain

amount of capital, in money or in goods computed at a value in money agreed upon, and in each case payable at the expiration of a fixed term.... [L]enders, if they have occasion for the use of their capital...before the expiration of the term of the bills they hold, are mostly enabled to borrow or to buy on lower terms, by having their own credit strengthened by the names on the bills in addition to their own. A great proportion of the vast transactions of the country among wholesale dealers are carried on by credit... (Tooke 1959, p. 87) Thus, the lender of commodities receives an IOU, and this IOU may

be transferable.

Furthermore,

merely having such IOUs

in hand

makes it easier to obtain credit, that is, to issue one’s own debt.

The term money as used here is synonomous with government fiat money. Tooke is correct to argue that money is an effect, not a cause, of spending. Thus, money is issued on demand. However, as I will argue in the section on Marx below, Tooke incorrectly distinguishes between fiat money and privately issued money. Tooke implicitly assumes that bank notes and other liabilities are issued to allow circulation of existing goods (or are issued when customers deposit government debt and gold). That is, he does not discuss how expansion of credit allows production to proceed. A new issue of money (whether privately issued or issued by government fiat) can represent an additional demand on resources. If the economy is at full employment, a rise in loans issued to finance an increase in investment may result in rising prices as wages are bid up to induce workers to switch jobs. In any case, however, it is not the increase in

the money supply which raises prices, but the rise in demand for resources. Of course, this rise in demand can be met through credit expansion, which is money, broadly defined. There is no need for an increase in the medium of circulation (money narrowly defined) since a rise in velocity is sufficient to allow a given quantity of the medium of circulation to support greater expenditures. Tooke’s error seems to be in his acceptance of the Real Bills Doctrine. (See note 11

129

below. ts ies 2 that the money (broadly defined) which is privately issued cannot be excessive since it is issued on demand. He quotes an interrogation of a banker, Mr. Gurney, by the Committee on the Bank Charter in 1832: Committee: ‘Does it not follow from what you have said, that an over-issue of notes of country bankers cannot easily be effected?’ Mr. Gurney: ‘My belief is that it cannot be effected by any act of the country bankers’ (in Tooke, Vol. IV, p. 236.). On the other hand, Thornton provided a summation of the proper role for the central bank which would have been accepted by the Currency School: To limit the total amount of paper issued, and to resort for this

purpose whenever the temptation to borrow is strong, to some effectual principle of restriction...to afford a slow and cautious

10.

ie

1, 13. 14.

extension of it, as the general trade of the kingdom enlarges itself...and to lean to the side of diminution, in the case of gold going abroad, and of the general exchanges continuing long unfavourable; this seems to be the true policy of the directors of an institution circumstanced like that of the Bank of England (Henry Thornton, quoted in Mints 1945, p. 56). See Panico (1980, 1988), Lianos (n.d.), and Ong (1983) for discussions of the role of credit in Marx. As Moore argues, the Banking School essentially adopted the Real Bills Doctrine. (See Moore 1988b, p. 5) If bank credit is created only to finance goods while they are being produced or transported, the loans would be retired as soon as the goods were sold to final purchasers. This view led the Banking School to argue that credit which is privately created could not be inflationary. See Foster (1986a, 1986b) for an attempt to include endogenous money in Keynes’s general theory. See Chapter 6 below for a detailed treatment of liquidity preference theory. Keynes’s own lack of clarity may have been responsible for the recent debate in the Post Keynesian literature concerning the relation between saving and investment, as will be discussed below. See Asimakopulos (1983, 1985a, 1985b, 1986a, 1986b, 1986c), Davidson

(1986a), Graziani (1984, 1986), Kregel (1984-85, 1986a), Terzi (1986-87), and Wray (1988a). 15. Throughout this article, Keynes emphasized the role of credit in financing an increase in investment. (Keynes 1973, pp. 278-85) 16. In his defense of the General Theory, Keynes argued that he had not adopted a fixed money supply. See Keynes (1973, p. 232). ATE Note that this differs from Tooke’s theory, in which credit is granted to circulate existing goods, so cannot be inflationary.

130

5S

REVIVAL OF THE ENDOGENOUS MONEY APPROACH

Introduction

After Keynes, an exogenous money supply was adopted in mainstream theory, and was incorporated within IS-LM analysis. Monetarists have also adopted an exogenous money supply. Over the years, this position has been challenged at various times. This chapter summarizes several of these challenges, beginning with that of the Radcliffe Committee, in order draw out the endogenous money thread that links the works of the economists analyzed here. I will not discuss every theorist who has contributed to the endogenous approach.

N. Kaldor and the Radcliffe Committee

In 1959, the Radcliffe Committee in Great Britain issued its report: The Working of the Monetary System.’ This report analyzed the

institutional changes which had occurred over the previous decades to determine the impact these had on conventional monetary policy. Specifically, the committee set out to determine whether tight monetary policy and high interest rates could be used to contain high aggregate demand. It was feared that high levels of liquidity which existed after the war (due to forced saving) would lead to excessive spending. Thus, tight money policy might be used to forestall inflationary pressures. The Radcliffe Committee, however, concluded that tight monetary policy would not prevent spending, but would probably cause severe distortions of spending due to high interest rates which would slow investment and reduce technological innovation. The Committee argued: Spending is not limited by the amount of money in existence;...it is related to the amount of money people think they can get hold of, whether by receipts of income (for instance from sales), by disposal of capital assets or by borrowing...spending is not limited

134

by the amount of money in existence...[due to variability of] the velocity of circulation. (Radcliffe Committee 1959, p. 133)

The Committee concluded ‘we cannot find any reason for supposing, or any experience in monetary history indicating, that there is any limit to the velocity of circulation of money’ (ibid., p. 133). Furthermore, ‘the more efficient the financial structure, the

more can the velocity of circulation be stretched without serious inconvenience being caused’ (ibid., p. 133). The emphasis of the Committee on variable velocity may have been due to Kaldor’s testimony.’ Kaldor presented evidence showing that velocity increased between 1955 and 1958, and attributed this directly to tight money policy which had not allowed the money supply to expand as rapidly as spending. He argued that there is no direct impact of money on income, since velocity can expand without limit. Instead, ‘the impact effect of any change in the money supply is not on the level of payments at all, but on the velocity of circulation’ (Radcliffe Committee 1959, p. 146). Thus, tight money would be an ineffective means of controlling inflation, and would only work if interest rates were pushed high enough to induce a deep recession. He argued that a smaller change in interest rates might have no significant impact on spending since the investment-interest rate link was not precise. Furthermore, while a large change in interest rates might affect investment, it would tend to shift spending toward consumption, with adverse consequences for long term growth. Thus, Kaldor and the Radcliffe Committee primarily relied on variable velocity as a critique of the quantity theory of money. Monetary policy would be ineffective because a restricted rate of growth of some monetary aggregate would simply lead to a rise in velocity. Tight monetary policy could indirectly affect spending if interest rates were raised sufficiently. However, by 1970 Kaldor realized that the emphasis on velocity had been misplaced. He began to argue that the money supply itself (and not just velocity) is endogenously determined (Kaldor 1970). Rather than arguing that velocity would adapt to a rise in expenditures, he argued ‘the increase in the supply of money is a consequence of increased loan expenditure’ (Kaldor and Trevithick 1981, p. 5). When

firms utilize

credit

lines, ‘there will be an

automatic increase in the money supply for the simple reason that additional expenditure will swell the bank deposits of the recipient’ (Kaldor and Trevithick 1981, p. 6).

132

Kaldor thus argued that the demand for loans determines the supply of money, because bank lending is limited only by the credit demands of worthy borrowers. He believed that if a stable relation exists between expenditures and the money supply, this is simply proof that expenditures raise the level of the money supply when they are based on the extension of credit (Kaldor 1985, p. 22). Central bank control over the money supply consists of its ability to force interest rates to rise, which reduces expenditures and income. The fall in the money supply is only a result of the fall of

expenditures’ (Kaldor 1985, pp. 22-5).

Furthermore, Kaldor argued that tight money policy would increase interest rates and velocity, and would probably increase the money supply as well. This is because higher interest rates would increase the attractiveness of holding various interest bearing deposits relative to other assets (Kaldor 1985, p. 75). (The attractiveness of bank deposits enables banks to make loans by creating deposits without worrying about a clearing drain as these deposits are exchanged for nonbank liabilities.) As evidence, he pointed to the Monetarist experiments in the United Kingdom and in the United States, in which extremely tight money policy raised interest rates, but, especially in the UK, actually increased the rate

of growth of the money supply (ibid., pp. xv, 75, 97). Monetarists (like Friedman) argued that the failure of the ‘Monetarist experiment’ in both countries should be attributed to the incompetence

of the central

banks, and

not to Monetarist

theory. If only the monetary authorities had been competent, they could have hit money targets and eliminated inflation without inducing recession. Kaldor responded that these targets cannot be hit because tight money induces innovation and economization of cash, and actually increases the ability of banks to attract deposits (at the expense of other financial assets) at the higher interest rates (ibid., pp. xiii, 46). Finally, the central bank must ultimately act as a lender of last resort so that it cannot close its discount window merely because the rate of growth of the money supply exceeds targets without endangering the entire banking system (ibid., p. 25). However, the central bank is free to establish the discount rate and thereby affect the entire term structure of interest rates. ‘The rate of interest...is the one instrument which is entirely under the control of the Government’ (ibid., p. 97). Thus, Kaldor accepts the endogenous money approach. This offers a much stronger criticism of Monetarism than does the original Radcliffe thesis. Friedman believed that the Radcliffe 133

position (and, thus, Keynesianism) relied on the hypothesis of an unstable velocity of money. Thus, when he was able to show stability of velocity, he believed that this clearly contradicts the Keynesian position and lends support to the quantity theory of money.’ However, with an endogenously determined money supply, stability of velocity is also perfectly consistent with Keynesian theory. In Kaldor’s view, spending determines money, but there is also some feedback from money to spending through the interest rate. Kaldor’s position is most consistent with that of Moore’s, to be examined in detail below. Both Kaldor and Moore argue that banks are never quantity constrained because the central bank ultimately accommodates bank demand for reserves. I believe the endogenous money argument can be strengthened by adding some of Minsky’s insights regarding leverage ratios. Thus, because banks are concerned with leverage ratios, there is an operative quantity constraint even though banks could theoretically expand balance sheets without limit. Furthermore,

like Moore, Kaldor argues that the central bank

determines the short term interest rate, which is set by a markup over the discount rate. Kaldor argues: One cannot conceive of the short rate being ‘determined’ in any other way than through the rediscount rate, or the open market policy of the Central Bank. Indeed, it is only through their power to control the whole range of short term interest rates that the monetary authorities can be said to ‘control’ the supply of money in its broader sense. (Kaldor 1980, p. 6)

Like Moore’s ‘horizontalist approach’, Kaldor’s approach relies critically on accommodative central bank policy. Indeed, he argues that before World War I, it was easy for central banks to control the money supply, whose quantity was exogenously determined by gold reserves (Kaldor 1985, p. 102). Endogenous money has existed only since World War I, after which central banks began to accommodate the demand for reserves rather than forcing banks to adapt to varying gold stocks. This is not consistent with the arguments presented in Chapters 1 and 2 above. Furthermore, in Chapter 6, I will offer an alternative theory of interest rates which I believe is consistent with the endogenous money approach, and which does not rely on exogenous determination of interest rates by the central bank.

134

Hyman Minsky and the financial instability hypothesis In Chapter 1, I discussed Minsky’s view of money and showed that it is consistent with the endogenous money approach. In 1957, Minsky related institutional innovation to profit opportunities, and showed how innovation allows business activity to expand even without an expansion of bank reserves. Specifically, he showed how the development of the Fed funds market allows a given level of aggregate reserves to support a greater expansion of deposits, and how repurchase agreements allow a given quantity of demand deposits to support a greater volume of loans (Minsky 1957). Minsky also related innovation to tight monetary policy. If a business expansion causes the monetary authority to fear inflation, it institutes tight monetary policy, which leads to a rise of interest rates. Tightness in the money market doesn’t, however, immediately reduce the demand for loans. Furthermore, as interest rates rise, liquid balances are reduced to take advantage of profit opportunities so that velocity of liquidity increases. However, financial institutions also want to take advantage of the new higher interest rates, and so find new ways of meeting loan demand. Innovations result in a shift of the relation between velocity and interest rates: a higher velocity will be associated with each interest rate. Thus, centrai bank constraint raises interest rates, inducing

innovations which increase velocity. Monetary policy won’t be effective in constraining monetary aggregates unless it decreases aggregate reserves sufficiently to compensate for the rise in velocity which results from innovation (Minsky 1957). Innovations are related to Minsky’s financial instability hypothesis. When innovations are induced by a demand for finance which can’t be met within the existing institutional constraints, systemic liquidity is reduced. ‘(E)very institutional innovation which results in both new ways to finance business and new substitutes for cash assets decreases the liquidity of the economy’ (Minsky 1957, p. 184). Innovations allow an existing quantity of liquidity to support greater expenditures. Innovations allow ‘a pyramiding of liquid assets’ as increasingly illiquid assets replace cash, government bonds, or short term bank debt in portfolios. This increases the instability of the system since failure by only a few large firms (financial or nonfinancial) to meet debt commitments can lead to a cascade of failures as assets are written-down.” Early in a boom, high demand for finance will not raise interest rates much since the supply of finance is very elastic. Banks and

135

and the money market have elastic supply curves, and firms they as long as households will give up liquidity to receive a return are optimistic about the future. But declining liquidity (rising debt to net worth,

or to safe asset,

ratios)

increases

instability

by

exposing assets to the possibility of rapid deflation of value should some units fail to meet commitments. Ultimately, the major limiting factor regarding how far the value of financial assets can fall is the willingness of the central bank to purchase them. Thus, the monetary authorities can halt a debt deflation process by intervening as a lender of last resort to increase the quantity of liquidity by accepting illiquid assets (Minsky 1957, p. 185). The central bank cannot exogenously set the money supply. In an expansion, the public voluntarily reduces its liquidity position, allowing production to expand even if the money supply (narrowly defined) does not expand. However, it is likely that the money supply (even narrowly defined) will expand as innovations permit expansion of loans and demand deposits on a given reserve base. In a crisis, the desired liquidity position may not be reached until the value of assets has fallen so far that employment and income levels have fallen catastrophically. In order to prevent depression, the central bank can supply liquidity (for example, by increasing the reserve base) so that liquidity positions can be restored before asset values and employment bottom-out. The central bank sets a floor to the value of assets by acting as a lender of last resort and/or by directly intervening to guarantee the value of bank assets. Thus, the central bank must ultimately operate as a lender of last resort to prevent debt deflation. After 1957, Minsky gradually developed a sophisticated view of the relations among credit, asset prices, and instability. According to Minsky, rising expectations raise the demand price of capital assets above their supply price, inducing investment.° Some of the investment is financed out of internal profit flows, but much is financed externally by financial institutions. As an expansion proceeds, fueled by investment, profits rise and reward units which have expanded balance sheets. Firms become more willing to commit a larger portion of expected income flows to servicing debt, while banks become more willing to leverage equity as they provide loans and create deposits. Financial fragility increases as income flows and payment commitments become increasingly synchronous. A crisis can occur if finance costs rise, if liquidity preference rises, or if income flows turn out to be less than expected. Endogenous processes tend to ensure that one of these (or all

136

three) will, in fact, occur. As discussed above, Minsky argues that as bank balance sheets expand, there is upward pressure on interest rates because banks are concerned with leverage ratios.’ Furthermore, the central bank often attempts to constrain the growth of reserves at the peak of an expansion, which reduces bank liquidity (the ratio of cash assets to liabilities) and also tends to raise interest rates (Fazzari and Minsky 1984). Rising interest rates increase the portion of income flows which firms must commit to debt payments (firms must roll-over some of the debt since construction projects take real time). As interest rates rise, ‘present value reversals’ occur, which reduce planned investment (and may even cause some projects to be abandoned) (Minsky 1978, p. 16, and 1986b, p. 195). As investment declines, aggregate profit flows decline and make it even more difficult to meet payment

commitments.°

Initially, demand for finance remains inelastic with respect to the interest rate as firms attempt to meet payment commitments through portfolio adjustments. This can lead to rapidly rising interest rates. Firms which cannot meet payment commitments out of income flows or through borrowing must liquidate assets. However, asset liquidation under these circumstances requires devaluation and capital losses. As assets are devalued, and as

falling investment reduces profit flows, debts repudiated.’ In the absence of big government

eventually are countercyclical

deficits and central

debt deflation

bank

intervention,

a general

occurs. However, government deficits place a floor to the level of profit flows, and central bank intervention as a lender of last resort places a floor to the value of bank assets and liabilities (Minsky 1978, p. 17). Minsky does not present a strict dichotomy between ‘banks’ and ‘financial

institutions’,

nor

between

financial

institutions

and

nonfinancial institutions.’ All types of economic institutions engage in balance sheet operations, and any economic institution can create money. As Minsky says, the problem lies in getting it accepted (Minsky 1986b, p. 228). Liabilities are used to finance positions in assets, and represent a ‘money now for money later’ proposition. A wide variety of balance sheet items enables oneto spend now and pay later (Minsky 1986b, p. 228). Certain liabilities of financial institutions also circulate as media of exchange and

means

of payment.’

According

to Minsky,

banks

are

not

intermediaries, but enhance the debts of customers. ‘Banking is not

money lending; to lend, a money lender must have money. The

137

that fundamental banking activity is accepting, that is, guaranteeing some party is creditworthy...’ (ibid., p. 229): It can be seen that Minsky’s insights add significantly to Kaldor’s argument. In an expansion, bankers share optimistic profit expectations with firms. Firms and banks are willing to increase leverage ratios, so that the ratio of debts to liquid assets rises as firms borrow to finance investment and as banks extend finance. Even if the central bank tries to limit the expansion of some monetary aggregate (for example, reserves OF M1), this has little effect in an expansion since firms and banks willingly stretch liquidity. In the early stages of an expansion, balance sheets are expanded even with no rise in interest rates because innovations create new ways to finance positions in assets. When institutional constraints are reached, however, further expansion will require rising interest rates. Rising interest rates, in turn, increase systemic

fragility because a greater portion of income flows must be committed to meeting finance payments. If interest rates rise so high that a few large firms decide to decrease spending, or if a few large firms are unable to meet commitments, expectations will fall. When expectations fall, economic agents revise desired leverage ratios downward. However, general liquidation of positions in assets is not possible without declining asset values

precisely because the expansion was based on leveraged purchases of assets. The central bank can prevent widespread default by increasing liquidity.

James Tobin and the new view

In 1963 in ‘Commercial Banks as Creators of Money’, Tobin distinguished between the ‘old view and the ‘new view’ of banking (Tobin 1987). According to the old view, there is a clear distinction between banks and other types of financial institutions, and between money and other types of financial assets. Only commercial banks can create money, and they do so through the deposit expansion process. Since they possess ‘the widow’s cruse’, they ‘have to be restrained by reserve requirements. Once this is done, determination of the aggregate volume of bank deposits is just a matter of accounting and arithmetic: simply divide the available supply of bank reserves by the required reserve ratio’ (Tobin 1987, p. 273). The new view, on the other hand, focuses ‘on demands for and

138

supplies of the whole spectrum of assets rather than on the quantity and velocity of "money"...’ (ibid., p. 274). In addition, the focus is on the full range of financial institutions. (T)he essential function of financial intermediaries, including commercial banks, is to satisfy simultaneously the portfolio preferences of two types of individuals or firms. On one side are borrowers, who wish to expand their holdings of real assets inventories, residential real estate, productive plant and equipment, etc - beyond the limits of their own net worth. On the other side are lenders, who wish to hold part or all of their net worth in assets of stable money value with negligible risk of default. (ibid., p. 274)

Intermediaries allow borrowers to obtain funds to purchase real assets at a lower interest rate than if they were to borrow directly from savers. Any increase in ‘the amount of financial intermediation in the economy can be expected to be, ceteris paribus, an expansionary influence’ (ibid., p. 275). Commercial banks are special since their liabilities function as means of payment. However, since there is a high degree of substitutability among the liabilities of financial intermediaries, other financial intermediaries also influence real spending. All financial intermediaries are limited in their ability to expand balance sheets. ‘Neither individually nor collectively do commercial banks possess a widow’s cruse’ (ibid., p. 276). Expansion can occur only up to the point where ‘the marginal returns on lending and investing,

account

taken

of the

risks

and

administrative

costs

involved, will not exceed the marginal cost to the banks of attracting and holding additional deposits’ (ibid., p. 277). In the absence of reserve requirements, ‘expansion of credit and deposits by the commercial banking system would be limited by the availability of assets at yields sufficient to compensate banks for the costs of attracting and holding the corresponding deposits’ (ibid., p. 279). But in the presence of required reserve ratios (and particularly if deposit interest rates are controlled by legal ceilings), an expansion of reserves will normally lead to an expansion of loans since ‘the marginal yield of bank loans and investments exceeds the marginal cost of deposits to the banking system’ (ibid.,

paz 19)

The simple deposit multiplier doesn’t necessarily apply to all expansions of reserves. Banks may decide to hold excess reserves if the interest rates on various types of assets aren’t sufficient to 139

in bank reserves may cover the costs of obtaining funds. A rise but this will also induce a rise in the quantity of loans extended, be charged, so will tend to reduce the interest rate which can loaned up or to fully ‘diminish the incentives of banks to keep average nt to hold the borrow reserves, and to make banks conte

higher excess reserves’ (ibid., p. 280). endogenous Much of what Tobin has argued is consistent with the of conflict. money approach. However, there are several points actual ‘that First, Tobin argues that he does not mean to imply ary monet ook textb monetary authorities have any less control than en betwe age’ ‘slipp some authorities’ (ibid., p. 289). While there is of ty quanti the en betwe reserves and liability expansion, and is ity author ary monet the , itures financial assets and real expend es. reserv of on ulati manip gh throu ges able to offset these slippa Thus, by controlling the base, the central bank

can

control

the

es in money supply. Second, although Tobin realizes that chang loans, make to depositor preferences can affect the ability of banks bank of tance impor he does not appear to recognize the full aints. constr ty quanti from liability management in freeing banks And, finally, Tobin does not seem to fully realize that any extension of credit necessarily creates the deposits which the system as a whole requires. Asset and liability management, innovations, and central bank accommodative behavior essentially ensure that a ‘widow’s cruse’ does exist if creditworthy borrowers exist. Tobin’s 1963 paper led to other work which attempted to develop the ‘new view’, as an alternative to the Monetarist view which was rapidly gaining adherents during the 1960s. Gramley and Chase (1965), Karekan (1967), Cacy (1968), and Davis (1968) extended Tobin’s

analysis,

stressing

‘the

role

of assets,

both

real

and

financial, and the relative price mechanism in monetary analysis’”. They also stressed that the money supply is a function of a complex series of relations among banks, the public, and the central bank. Thus, they dismissed the simplistic view that the money supply is determined by the base (set by the central bank) multiplied by a stable ‘money multiplier’. Brunner (1968) attacked the new view on several levels. First, he argued that the new view’s emphasis on ‘a systematic application of economic analysis, in particular an application of relative price theory, to the array of financial intermediaries, their assets and liabilities’ (Brunner 1968, p. 11) did not distinguish it from the Monetarist approach, in which ‘monetary impulses are transmitted to the economy by a relative price process which operates on

140

money, financial assets (and liabilities), real assets, yields on assets and the production of new assets, liabilities and consumables’ (ibid., p. 18). Second, the new view made ‘little attempt at specific Structuring and empirical content’ (ibid., p. 12). The new view abstracted from the observable world to such a great extent that it is nearly empty of form. In fact, Brunner argued that much of the new view cannot be empirically tested, and that empirical evidence contradicts those hypotheses which are testable and which are in disagreement with Monetarist hypotheses. Brunner argued that the new view could be reduced to the hypothesis that the money multiplier is variable and that it could easily be tested by a regression of the money stock on the base. If the money multiplier is stable, then the new view would not be supported. Brunner found that over 90 per cent of the variation in the money stock is explained by changes in the monetary base. Furthermore, he argued that changes in the base can be directly controlled by the central bank. ‘In summary, preliminary investigations yield no support for the contention that the behavior of banks and the public dominates cyclical movements in the money stock’ (ibid., p. 18). He admitted that ‘the existence of a mutual interaction over the shorter-run between money and economic activity..must be fully acknowledged. Yet, this interaction results from the conception guiding policymakers which induces them to accelerate the monetary base whenever pressures o:: interest rates mount...’ (ibid., p. 20). In other words, if the central

bank

did not target interest rates, there would

be little

feedback from economic activity to the money stock. Thus, Brunner was able to incorporate some of the ‘new view’ into Monetarism, and was able to fault the rest because it was

untestable or because it was not empirically supported. The new view was not an effective critique of Monetarism because it did not develop an adequate theory of the way in which money enters the economy. The new view’s emphasis on portfolio adjustments due to changing interest rates was easily incorporated within Monetarism. Because the new view accepted the orthodox view that the base is a determinant of the money supply, Monetarists were able to reduce the new view to the hypothesis that the money multiplier is variable. When Monetarists were able to show a high correlation between the money stock and the base, they believed they had effectively disproven the new view (or, at least, had relegated it to a theory of minor fluctuations in the money multiplier). The results were similar to the Monetarist critique of 141

‘Keynesian’ theory: Monetarists believed that if they could show that the income velocity of money is fairly stable, then the quantity theory of money would be supported, while Keynesian theory

would be disproven.”

However, neither a stable money multiplier nor a stable income velocity of money is inconsistent with the endogenous approach to money. If money is created to finance spending, and if expansion of bank balance sheets requires expansion of reserves, then stable relations between spending and money, and between money and reserves,

will

exist.

Thus,

Brunner’s

empirical

tests

cannot

distinguish between the endogenous approach and Monetarism. Tobin and the other ‘new view’ economists did not go far enough in their development of an alternative to Monetarism because they accepted a direction of causation running from the base to money, and because they accepted central bank control over the base.

John G. Gurley and diversification

E. S. Shaw:

intermediation

and

Gurley and Shaw (1955, 1956, 1960) began with an analysis of the finance process. Spending may be financed internally (through income flows), or externally (debt). External finance includes direct and indirect finance. Direct external finance occurs when a surplus unit (one whose expenditures are less than income) acquires financial claims on debtors. Indirect external finance occurs when surplus units purchase claims on financial intermediaries, which then make loans to the ultimate borrowers (Gurley and Shaw 1956, pp. 258-9). In a closed economy, in the aggregate, income equals expenditure so that net borrowing must be zero. Each private unit which spends more than its income (incurring debt) is offset by a surplus unit (which acquires financial assets). Net financial private wealth must be zero in a closed, private economy (Gurley and Shaw 1960). However, there are two possible sources of financial net wealth:

a government sector and a foreign sector. One country can have a positive net financial worth if it obtains net credits on another. Furthermore, the government sector can issue liabilities which become a source of net wealth for the private sector.’ Gurley and Shaw then distinguished between inside and outside money. Inside money is the creation of the private sector, and must be offset by a private liability. Thus, when

a bank makes

142

a loan, it creates

money which is an asset of the nonbank sector, but also creates a liability of the banking sector. The nonbank sector issues a liability (the IOU), which is held as the asset of the banking sector. Therefore, inside money cannot be a net source of wealth to the private sector. However, government debt (bonds or money) is a net source of wealth for the private sector. The money issued by a lee eae is Outside money (Gurley and Shaw 1960, pp. 2-5). Gurley and Shaw noted that orthodox theory draws a sharp distinction between the ‘monetary system’ (commercial banks and the Fed) and other financial intermediaries.” In orthodox theory, the monetary system issues money, while financial intermediaries merely ‘transfer to investors any part of this money supply that may be deposited with them by savers’ (Gurley and Shaw 1956, p. 260). However, Gurley and Shaw argued that ‘the monetary system is in some significant degree competitive with other financial intermediaries’ (ibid., p. 261). Financial intermediaries issue debt which competes with money issued by the monetary system: ‘their issues of indirect debt displace money...’ (ibid., p. 261). The debt of financial intermediaries functions like money until it becomes necessary to make monetary payments. That is, ‘other financial assets can be accumulated preparatory to money payments, as a precaution against contingencies, or as an alternative to primary securities’ (ibid., p. 261). Since the liabilities of a wide variety of financial institutions act as substitutes, ‘for any level of income, the money supply is indeterminant until one knows the degree of substitutability between money created by banks and financial assets created by other intermediaries’ (ibid., p. 261). When spending by deficit units rises, financial assets and liabilities are created. To the extent that

surplus units are willing to accept ‘near monies’ (liabilities of nonbanks) rather than money (liabilities of the ‘monetary system’), spending can rise with no increase in the money supply (narrowly defined). That is, the velocity of money rises. Gurley and Shaw explain: (1)t is a mistake to suppose that the optimal growth of the money supply is some simple function of trends in income or even of income and debt. The necessary growth in the money supply may be high or low, positive or even negative, depending on the growth of income, the share of spending that is externally financed (especially by long-term securities), the growth in demand by spending units for direct relative to indirect financial

143

assets, and on the development of financial intermediaries whose indirect debt issues are competitive with money. (Gurley and Shaw 1955, p. 531)

If rising spending by deficit units were financed solely by direct finance, surplus units would likely require rising interest rates to induce them to hold increasing quantities of direct debt. This would occur, for example, if there were preference for liquidity since the money supply would be constant, while illiquid forms of assets and liabilities would be rising. Thus, continued expansion of spending might require that financial intermediaries intervene, purchasing the direct debts of deficit units, and issuing indirect debt to surplus units. Eventually, surplus units may acquire all the indirect debt of financial intermediaries desired. In this case, interest rates on these liabilities will have to rise. Alternatively, the

monetary system can intervene, with commercial banks acquiring direct debt and issuing money to satisfy the liquidity preferences of surplus units. Finally, diversified portfolios are desired to reduce risk. In brief, the demand for money, the demand for diversification among other financial assets, offers of differentiated primary securities, and the demand for mixed asset-debt positions are to be explained by a common principle. They are tactics of risk-avoidance by spending units in a society where division of labor between saving and investment creates stocks of claims and counterclaims. (Gurley and Shaw 1960, p. 121)

Because the government creates outside money, monetary policy is not neutral. A rise in outside money raises balances of money. Since this is not offset by a rise in private debt, there is a rise in financial net wealth of the private sector. Furthermore, it results in a rise of (outside) money relative to other financial assets. As surplus units attempt to diversify, the interest rate can be driven down on direct and indirect liabilities of deficit units. This, in turn,

can stimulate expenditure and production. In the absence of government debt, the private sector might reach a point at which further expansion is inhibited because the liquidity premium required to induce surplus units to hold debt is too great. A rise of outside money can alter the premium since liquidity rises." Inside money is not neutral, either. Gurley and Shaw call their approach the gross-money doctrine, in which both inside and outside money matter. They counterpose it to the net-money

144

doctrine, in which only outside money affects real variables. [N]et money doctrine implies that the quantity of inside debt and its counterparts in financial assets have no net effects on the aggregate real demand for money, but this overlooks the desire financial positions by both firms and for diversified consumers...the doctrine distinguishes too sharply between the behavior of private domestic sectors and that of outside sectors. Stockpiling of financial assets is a defensive measure for spending units in a hostile economy, much as stockpiling of weapons is a defensive measure for nations in a hostile world. And in each case diversification of the stock usually pays off. Spending units’ demand for money, as one financial asset among many, depends partly on the types of frictions and uncertainties associated with alternative financial assets. (Gurley and Shaw 1960, p. 188)

Gurley and Shaw’s distinction between inside and outside money has been attacked on two fronts. First, it has been argued that government debt is not outside money. According to Gurley and Shaw, when the government issues bonds to the central bank to finance a rise in government spending, the quantity of outside money rises because the central bank issues liabilities which are offset by government debt, rather than private debt. However, the critique argues that the public should realize that this government debt represents a future claim on private income (in the form of future taxes). Thus, there is no net increase in private net wealth. However, part of the interest on government debt can be paid out of the government’s returns on assets, some of which are the liabilities of foreigners. In addition, the rate of discount applied to future taxes might differ from the rate of discount applied to future interest receipts. In this case, the present value of government debt may be more than the present value of future tax payments. Finally, it is doubtful that the public actually believes that interest will be paid out of tax revenues. This disbelief is not irrational, if one can use the past as some sort of guide to the future.” The second critique of Gurley and Shaw takes the opposite view: all money is outside money.’* Money is characterized as the capitalized value of the returns it yields in services. When banks make loans and create demand deposits, these deposits function as money and provide utility to their holders. This real increment to utility is not offset by a decline in utility of the bank, since it doesn’t give up any money.” Harris summarizes this position: In Pesek and Saving’s analysis, therefore, the creation of demand

145

deposits by the bank giving loans to the private sector does create net wealth. The private sector gains an asset (money) while the banking sector does not increase its liabilities except in a bookkeeping sense, since demand deposits on which no interest is paid are not a genuine liability to the banking sector. (Harris 1981, p. 39)

This argument ignores uncertainty. What the bank has given up is liquidity. By accepting the loan, the bank has increased the ratio of risky assets to equity and liquid reserves. If the loan turns out to have been to a creditworthy customer, then equity is increased, otherwise, the bank suffers a fall in equity. Since the future is

unknowable, it is impossible for the bank to eliminate (or compensate for) uncertainty entirely. Furthermore, by creating demand deposits, the bank has also put its equity at risk since the bank

has

incurred

a potential

monetary

liability.

Should

the

depositor decide to hold cash, the bank is required to come up with cash. Since. banks actually hold very little cash relative to the deposits they have issued, they are always exposed to the risk that withdrawals will exceed cash reserves. The public’s gain in liquidity is offset by the bank’s loss. Since utility is immeasurable, and cannot be compared across economic units, it is impossible to state whether there is a net increase in utility. However, it is clear that

the rise in ‘utility-yielding services’ which results when banks create deposits is accompanied by a rise in bank ‘disutility’ since banks are ultimately responsible for providing liquidity on demand. In any case, the Pesek and Saving argument is irrelevant to the major conclusion of Gurley and Shaw, which is that both inside and

outside money matter. The purpose of Gurley and Shaw’s thesis was not to show that only outside money allows real output to increase. Indeed, they consistently argued that there is a high degree of substitutability among various financial assets, so that real spending can rise even if the money supply (whether defined narrowly as government debt, more broadly as government debt plus commercial bank deposits, or even more broadly so as to include deposits at other financial institutions) is held constant. Credit expands to allow some units to spend in excess of income. Of course, expenditure creates income, so the deficit is offset by a surplus. Where surplus units do not wish to hold all the direct debt issues, financial intermediaries intervene with issues of indirect debt. Some of the indirect debt may include money (however defined). As production increases, the demand for money may rise, so that continued expansion with constant interest rates eventually

146

requires expansion of the money supply. However, the linkage between money and income is very imprecise. ‘The same money supply and national income are compatible with various interest rate levels, depending upon the size of non-monetary intermediaries and upon the degree to which their issues are competitive with money’ (Gurley and Shaw 1956, p. 263-264). Gurley and Shaw’s analysis of financial institutions is consistent with the endogenous money approach. Spending is not limited by since monetary financial aggregate defined a narrowly intermediaries intervene with various liabilities which allow spending to rise. Gurley and Shaw’s discussion of portfolio diversification is interesting and adds to the endogenous money approach. If firms had to issue securities directly to surplus units, then an increase in investment might require rising interest rates because economic units would prefer to diversify portfolios. However, financial intermediaries intervene with a wide variety of liabilities so that surplus units can diversify. This allows investment to proceed without causing interest rates to rise. In this way, a given quantity of liquidity (or outside money) can support a wide range of spending levels and inside money.

Basil Moore: the horizontalist approach Moore argues that only when money is metallic money is it exogenously determined. He argues his approach follows that of Wicksell, who constructed a pure credit economy which is consistent with an endogenous money approach. ‘All money is simply someone’s promise to pay. If there is in fact no hard metallic money in which payment can be made, all that can be promised is the convertibility of one promise into another promise...’ (Moore 1985, p. 7).

Bank loans are made primarily at the initiative of the borrower, especially where credit lines have been previously negotiated. Once banks have met credit demands, they hold assets which are typically illiquid. Thus, ‘banks ordinarily cannot increase or reduce their loan portfolios at their initiative’ (Moore 1985, p. 11). Since a reduction in the money supply (for example, through open market sales) requires a reduction of bank assets, ‘the ability of central banks to effectuate an exogenous reduction in the money stock will thus depend critically on the composition of bank portfolios’ (ibid., p. 11). By the 1960s, banks had reduced holdings 147

of cash and government

bonds, making

it difficult for them to

liquidate positions. Due to private information involved in making loans, banks cannot normally sell loans in secondary markets. As a result, ‘the monetary authorities are rendered relatively impotent in their ability to restrict the rate of growth of the money stock’ (ibid., p. 12). Thus, ‘since central banks, as consistent with their paramount supportive role to the financial system, operate to allow the money stock to accommodate increases in the demand for bank credit, monetary aggregates may properly be considered to move endogenously’ (ibid., p. 24). Moore argues that banks are price setters and quantity takers both in retail loan markets and in retail deposit markets. Bank loans are made at the initiative of borrowers at an interest rate set by banks. Banks also set a deposit interest rate, then passively accept deposits at that rate. Since for individual banks, the quantity of loans may exceed the quantity of deposits, banks must rely on wholesale sources of funds, in which banks are price takers. Ultimately, the central bank is the marginal supplier of wholesale funds (at the discount window or in open market purchases). As the marginal supplier, the central bank ultimately determines the wholesale cost of funds. Retail loan rates are then set as a markup over the wholesale cost. The retail deposit rate is a mark-down of the loan rate, to cover costs of administering deposits and to provide profits (Moore 1988b, Chapter 3). According to Moore, the supply of money can be viewed as horizontal at the going interest rate. That rate, in turn, is ultimately determined by the Fed’s discount rate. The Fed is able to control the supply of money only by regulating short term interest rates by determining the price it charges for reserves. The Fed is not able to constrain bank reserves, since banks meet the demand for loans and then must meet legal reserve requirements. ‘Reserves are the result, rather than the cause of movements in deposits’ (Moore 1984, p. 107). The interest rate determined by the Fed can affect the demand

for loans, but once

loans are made,

the Fed must

supply the reserves needed by the banking system. Moore quotes Holmes affirmatively: Since banks have to meet their reserve requirements each week...and since they can do nothing within that week to affect required reserves, that total amount of reserves has to be available to the banking system. The Federal Reserve does have discretion as to how the banks can acquire this predetermined level of needed reserves. The reserves can be supplied from the

148

combination of open market operations and the movement of other reserve factors, or they can come from member bank borrowing at the discount window...the suggestion that open market operations should be used in the short run to prevent a rise in total reserves through member bank borrowing is completely illogical. Within a statement week, the reserves have to be there; and, in one way or another the Federal Reserve will

have to accommodate the need for them. (Holmes 1969, p. 73)

Moore’s presentation is very similar to Kaldor’s: banks are never quantity constrained and the money supply function is horizontal at the going rate of interest (which, in turn, is determined by the central bank). I will argue in the next chapter that this is an oversimplification. At a point in time, the money supply curve slopes upward because banks require rising interest rates to induce them to increase leverage ratios, while innovations cause the curve to shift outward over time. Furthermore, I have stressed that banks

are incapable of supplying liquidity during a collapse of expectations. As banks revise desired leverage ratios downward, they will be unwilling to extend credit on demand. As Wolfson (1986) argues, banks cut off credit from all but established customers. Banks experience difficulty in meeting the demand for liquidity by even established customers because a general run to liquidity makes it difficult for banks to maintain a rate of reflux. As other customers draw down demand deposits to retire debts, or as customers switch from time deposits to liquid deposits, the bank finds that it is unable to retain reserves, so is unable to meet the

demand for credit. Hence, an unwillingness to lend is converted to an inability to lend. Thus, in a recession, the quantity of credit is not demand determined. Moore argues that the central bank passively supplies reserves at the chosen discount rate. This might lead one to expect that the ratio of reserves to loans, or of reserves to the money supply should be relatively stable. As shown in Chapter 7, however, these ratios vary greatly over time. Indeed, the ratio of reserves supplied by the Fed to total required reserves (or increases in required reserves) fluctuates widely as the Fed changes its policy. This seems to indicate that the Fed does not passively supply reserves. When the Fed decides to constrain the growth of reserves, it is likely that banks will increase interest rates as leverage ratios rise. In other words, the markup will change. If the markup is variable, then it does not make sense to draw a horizontal money supply curve.

149

Finally, Moore’s endogenous money approach relies critically on modern financial arrangements and practices. It implies that the money supply would not be endogenous in the absence of central banks, and says nothing regarding the determination of interest rates in the absence of central banks. As I have argued, the money

supply has always been endogenously determined. Indeed, it is more obviously endogenous in the monetary systems which predate the development of a central bank. Chapter 6 presents an alternative to the interest rate theory advanced by Moore and Kaldor, in which the central bank sets the short term interest rate. While the central bank can certainly influence the short term interest rate through its discount policy, I will argue that the central bank cannot ‘set’ the interest rate.

Marc Lavoie: the flow of credit

Lavoie (1984) argues that the major distinction between orthodoxy and the endogenous money approach is that neoclassical theory emphasizes money as a stock, or as an asset in a portfolio, whose value is derived from its usefulness in facilitating exchange of goods which have already been produced. In Post Keynesian theory, however, it is the flow of money which is important. Any flow of production requires a flow of new credit or a renewal of past credit flows. Money cannot be introduced into a system after production has taken place, since money creation has to do with the production process itself. In this scheme, borrowers are the

causal factor and the function of banks is to create money which makes production possible. The money stock merely represents

credits that have not yet been repaid (Lavoie 1984, pp. 771-4).

Lavoie classifies those who adopt the endogenous money approach into four groupings, first, Cambridge Post Keynesians such as Robinson, Kaldor, and Kahn; second, American Post Keynesians such as Davidson, Minsky, Weintraub, Moore, and Wojnilower; third, Circuitistes represented by Schmitt, Barrere, and Parguez; and last, Overdraft economists such as Le Bourva, Berger,

and Weymuller. The latter two schools are primarily French, and little of their work has been published in English” (Lavoie 1985, . pp. 65-6). These four groups accept two ideas. First, loans make deposits; and second, deposits make reserves. Thus, ‘banks do not wait for

the appropriate amount of resources to exist to provide new loans

150

to the public...’ (Lavoie 1985 p. 67-8). Furthermore, as Moore argues, ‘the Federal Reserve System cannot, any more than the Banque de France, control the emission of high-powered money. It can only modify its discount rate or its rate of intervention’ (ibid., p. 70). Both of these ideas have been dealt with sufficiently above. There are several points of disagreement among the four schools, however. Lavoie analyzes two points of disagreement between the Overdraft economists and the others. ‘Overdraft economists believe that there are countries, such as the United States or Germany, where orthodox theory can be applied’, while the endogenous approach applies to countries like France where the central bank accommodates the demand for loans (ibid., p. 71). The other groups, of course, argue that the approach applies to all economies in which sophisticated financial systems and capitalist relations exist. Secondly, the Overdraft economists believe it is possible to have an economy without credit money, and that this may even be preferable. The existence of credit may push interest rates too low to induce saving, so that investment must be financed on the basis of credit. Thus, the Overdraft economists believe that the loanable

funds theory would be applicable to an economy without credit, and that such an economy is possible. This is inconsistent with the endogenous money approach, and is also inconsistent with the Keynesian argument that investment determines saving. Credit is seen as an anomoly by the Overdraft economists. Actually, expansion of investment requires credit. Terzi (1986-87) and Wray (1988a) have shown that saving cannot be a net source of funds for rising investment. Credit is required to fund increases in net investment, which then creates saving. In an economy without privately issued credit money, the government would have to inject credit to allow for net investment. In reality, private institutions provide credit in all developed capitalist economies, and I have argued that the endogenous money approach applies to all such economies. Therefore, I would exclude the Overdraft economists

from the endogenous money approach, which emphasizes the necessity of credit and money. The Circuitistes and both groups of Post Keynesians agree that it is the flow of credit, not the stock of money, which is important, according to Lavoie (Lavoie 1985, p. 74). ‘They recognize that expenditures must eventually be financed by the transfer of funds into demand deposits. But the stock of these funds at any point in

£51

words, time is considered to be a residual’ (ibid., p. 73). In other which credit, of n creatio the s an expansion of spending require e, a exampl For units. ic econom some of ies represents the liabilit which t deposi d deman a s receive and bank a to firm issues an IOU t is is used to finance an increase in spending. The demand deposi the of one as d the liability of the bank and it is normally counte ts deposi d deman of ty components of the money supply. The quanti rate the by ined determ ly outstanding at a point in time is partial at which the firm retires its IOU. Lavoie argues that the important decision takes place when the firm decides to enter into debt in order to increase spending. The firm’s subsequent decision to retire the debt is not considered to be nearly so important. Lavoie focuses on the flow of credit rather than on the stock of some monetary aggregate. Thus, his approach is similar to the endogenous money approach I have outlined in which money is endogenously created as debt commitments are negotiated. His ‘flow of finance’ is equivalent to my definition of money as simply a debt. However, Lavoie goes too far in his assertion that stocks are unimportant. As I will argue in the next chapter, stocks in portfolios are important because they affect interest rates. Indeed, Lavoie’s rejection of the importance of stocks probably accounts for his critique of Minsky’s financial instability hypothesis, which he argues borders on a loanable funds approach.

Notes 1. R. S. Sayers also contributed significantly to the Radcliffe Committee’s report. I will focus on the contributions of Kaldor, whose work played a major role in the revival of the endogenous money approach. See Rousseas (1985a) for a discussion of the Radcliffe Commitee’s report. 2. See Thirwall (1987, p. 294) for a discussion of the importance of Kaldor’s testimony. 3. According to Kaldor, the central bank operates by setting the price of credit through its discount policy. This view is similar to that of Moore, and will be discussed in more detail below.

4. Velocity is not nearly as stable as he claims - see Foster (1986b) and Ben Friedman (1988). 5. See Minsky (1957) and Fisher (1933) for discussions of the debt deflation process. 6. This paragraph summarizes Minsky (1986b, Chapter 8). The demand price of a capital asset is a function of expected yields and the quantity of liquidity. Thus, ‘both a rise in [expected quasi-rents] and

£52

a diminished virtue of liquidity tend to raise the price of capital assets’ (Minsky 1986b, p. 182-3). As the demand price rises above the supply price (which is a function of production and finance costs), investment increases.

‘A rising inelastic demand will lead to a rise in the observed price unless supply is infinitely elastic at the existing price...For a variety of reasons...supply of finance from banks eventually becomes less than infinitely elastic. This means that after favorable conditions for investment are sustained over some time, the cost of financing investment as it is being produced increases’ (Minsky 1986b, p. 195). See a Minsky (1957), (1986b, pp. 196, 214, 227, 239), and (1978, p. 20). As Kalecki (1954) shows, aggregate gross profits are identically equal to the sum of investment, consumption out of profits, the government deficit, and net exports, less saving out of wages. This has been extensively analyzed by Minsky. See Minsky (1986b, p. 214-7) and (1978, p. 17). As noted above in Chapter 4, Marx also argued that credit expansion occurs by leveraging balance sheets, and that because of this, balance sheets cannot be liquidated during a downturn (Marx 1909, p. 532). 10. For a discussion of the banking process in Minsky’s theory, see, for example, the section entitled ‘The Business of Banking’ in Minsky’s chapter, ‘Banking in a Capitalist Economy’in Minsky (1986b). Indeed, much of Minsky’s work from 1957 to the present can be seen as an attempt to develop a sophisticated model of banking. In particular, see Minsky (1957), (1975, Chapter 6), (1982, Chapters 8 and 13), (1986a), (1986b, Chapter 10), and (1987). Hal The extent to which a ‘monetary asset’ will circulate depends on a variety of factors. Demand deposits circulate as a medium of exchange because 1) they have bank equity behind them, 2) banks diversify assets, 3) the government is willing to guarantee that demand deposits will exchange at parity with government debt, and 4) debt commitments are normally written in terms of demand deposits. (Minsky 1986b, p. 231) 12: Federal Reserve Bank of St. Louis Review, August 1968, p. 8. 13. In the General Theory, income is determined by three independent variables: the marginal propensity to consume, the interest rate, and the marginal efficiency of capital. The money supply is then given exogenously. One would not expect a stable relation to exist between money and income since money does not enter directly into the determination of income in Keynes’s model. Thus, velocity should vary. Monetarists believed that if they could show that velocity is stable, then they would have disproven Keynes’s model. See Chick (1983). However, showing that ex post money demand is stable is not the same thing as showing that liquidity preference

is stable. I will

argue in the next chapter that it is liquidity preference, and not

153

money demand, which goes into the determination of interest rates. Showing that velocity is stable merely shows that the relation between money and income is stable, which says nothing about the relative stability of liquidity preference. If money demand is related to planned spending, and if money supply responds to money demand, then one should expect a close relation between spending and money. On the other hand, liquidity preference exerts only an indirect impact on spending, through interest rates and profit expectations. 14. Of course, the net liabilities of the government completely offset the net wealth of the private sector. This will be discussed below. 153 Gurley and Shaw’s terminology is adopted here: money is the debt of the ‘monetary system’ (commercial banks and the central bank), while debt of other financial intermediaries is ‘near money’. 16. This is related to Minsky’s argument that banks and others are concerned with leverage ratios. An increase in the quantity of liquidity (for example, resulting from open market purchases) raises bank willingness to expand balance sheets until desired leverage ratios are restored. Gk See Harris (1981 pp. 35-6) for a discussion of the critique of Gurley and Shaw’s approach. If the current generation is not sufficiently altruistic, then it won’t consider the tax burden on future generations. In this case, it is perfectly rational to ignore current government deficits. 18. See Pesek and Saving (1967). 19: Strictly speaking, this is true only if banks do not pay interest on demand deposits. If they do, however, they are committed to a stream of payment commitments, which means they do lose utility yielding services. In this case, as long as the utility received by depositors exceeds the utility given up by the banks, the banks do create net wealth (Harris 1981). 20. This is similar to Schumpeter’s analysis, explained above in Chapter 4. The position of the Circuitists is most similar to that of Schumpeter. Pa An overdraft economy is one in which lines of credit exist explicitly or implicitly, so that agents may spend more than they have on account at their bank. In this case, the supply of credit is clearly a function of the demand.

154

6

ENDOGENOUS MONEY AND INTEREST RATES

Introduction

Keynes argued that in his theory, the interest rate is determined by liquidity preference as the price of hoards. In general, Post Keynesians still accept the liquidity preference theory of the rate of interest. However, at various points in the previous chapter, I have briefly referred to alternative views of the determination of interest rates. Some of those who have adopted the endogenous approach to money feel that the liquidity preference theory of interest rates cannot be reconciled with an endogenous money approach. In the following sections, I will present the case for a reconciliation of the endogenous money approach with liquidity preference theory, and will examine both the orthodox version of

liquidity preference theory and the markup approach taken by some Post Keynesians.

Liquidity preference in Keynesian and Post Keynesian thought Basil Moore (1988b) argues that Keynes adopted an exogenously determined money supply in the General Theory, even though he had previously recognized in the Treatise that money is merely bank debt whose quantity is endogenously determined. Moore argues that Keynes’s exposition in the General Theory was ‘fatally hampered’ by his determination to vanquish loanable funds theory through liquidity preference theory. He argues that the logic of the General Theory is circular because the demand for money and the level of interest rates vary with income, so that each change of income would change interest rates and therefore affect investment and lead to further changes in income and interest rates (Moore 1988c). Thus, the interest rate is indeterminant in the General Theory, and was made determinant only by Hicks’s IS-LM analysis. According to Moore, if Keynes had only adopted an endogenously determined money supply and exogenously determined interest

155

rates, the arguments

of the General Theory would

have

been

strengthened, and Keynes could have defeated loanable funds theory without relying on the ‘flawed’ liquidity preference theory. Lavoie also rejects liquidity preference theory, and argues that Post Keynesians are confused regarding the determination of the interest rate. If the money supply is truly endogenous, supplied on demand, then a rise in the demand for finance cannot, according

to Lavoie, place pressure on the interest rate. Lavoie argues that Keynes’s liquidity preference theory applies to an exogenously determined money supply, so is incorrect, and ‘Post-Keynesians do not abide by it anymore...’ (Lavoie 1985, p. 76). Lavoie goes on to argue that Minsky’s financial instability hypothesis ‘borders on the lack-of-savings thesis and is quite contrary to both post-Keynesian monetary theory and Keynes’[s] view of the trade cycle’ (ibid., p. TD ,will argue, however, that liquidity preference theory is an essential element of both Keynes’s own system and of the Post Keynesian approach. First, the central role played by liquidity preference theory in the general theories of Keynes and the Post Keynesians is discussed. In the next section, I will show how liquidity preference theory can be incorporated within an endogenous money approach. Moore’s argument that Keynes’s liquidity preference theory is ‘circular’ and Lavoie’s belief that Post Keynesians have rejected liquidity preference theory are perplexing. Kregel (1980) has argued that the interest rate on money measures liquidity preference, which is the differential between forward and spot liquidity. An own rate of interest can be calculated for each commodity, which is the return in terms of the commodity on a loan in that commodity. In other words, it is the amount

of the

commodity which can be bought for forward delivery in terms of a given amount of the same commodity for spot delivery.’ The return of a commodity can also be calculated in terms of money, rather than in terms of the commodity itself. This return can be called the marginal efficiency of the commodity. In equilibrium, the marginal efficiencies of all commodities should be equalized - if the marginal efficiency of one commodity were higher than that of others, demand would induce production until its expected rate of return fell into line with the marginal efficiencies of all other commodities.” Keynes argued: It seems, then, that the rate of interest on money plays a peculiar part in setting a limit to the level of employment, since it sets a

156

standard to which the marginal efficiency of a capital-asset must attain if it is to be newly produced.... For it is the greatest of the own-rates of interest...which rules the roost (because it is the greatest of these rates that the marginal efficiency of a capitalasset must attain if it is to be newly produced); and...there are reasons why it is the money-rate of interest which is often the greatest.... (Keynes 1964, pp. 222-3)

In the General Theory, the marginal efficiency of capital determines the spot price of real assets, while liquidity preference determines the spot price of financial assets. Monetary factors determine the price of debt (or the differential between spot and forward money), which then discounts the expected stream of profits which would arise from the ownership of real assets (that is, the marginal efficiency of capital). Liquidity preference determines the structure of liability prices, which gives the interest rate and simultaneously determines the price of capital assets by discounting expected returns. This affects the decision to hold real assets versus financial assets, so goes into the determination of the level of investment (Kregel 1987). Kregel has argued that liquidity preference theory and the expenditure multiplier are ‘two sides of the same coin’, thus, ‘[t]here can be no difference between Keynes’s insistence that it is the level of income via the multiplier that equates savings and investment and his insistence that it is the degree of liquidity preference which determines the rate of interest’ (Kregel 1988, p. 239). A decline in liquidity preference will lower the interest rate, which raises the demand price of capital assets and causes investment to rise until the marginal efficiencies of all assets fall to equality with the lower interest rate. This is equivalent to arguing that income rises through the multiplier until saving rises to equality with the new higher level of income. Keynes’s theory is a liquidity preference theory of nominal asset prices: a decrease in liquidity preference increases the relative money prices of assets, generating profit opportunities, inducing investment, and raising income through the multiplier. The decrease in liquidity preference changes the relation of spot and future prices for all assets in such a way that production and the multiplier will cause changes in income and saving until the relations between spot and future prices are brought back into line with the new interest rate?

(Kregel 1988).

As I noted above, Keynes rejected the loanable funds theory of interest when he argued that investment determines saving through

157.

changes of income: (T)he initial novelty lies in my maintaining that it is not the rate of interest, but the level of incomes which ensures the equality between saving and investment. (Keynes 1937, p. 250) (N)o amount of actual investment, however great, can exhaust and exceed the supply of savings, which will always exactly keep pace. (Keynes 1937, p. 248)

Indeed, he argued that saving and investment are ‘merely different names for the same phenomenon looked at from different points of view (Keynes 1987, p. 551). Furthermore, the interest rate cannot be determined by loanable funds, since ‘the supply curve of savings and the demand curve for investments have no determinant point of intersection, since they lie along one another...’ (ibid., p. 552). Keynes went on to argue that while investment can never be constrained by lack of saving, it can be constrained by lack of finance. As discussed above in Chapter 4,

however, by finance Keynes did not mean loans. He argued that firms may attempt to accumulate a cash balance before undertaking investment. Since saving can rise only due to investment, this ‘extra finance involved will constitute an additional

demand for money’ if investment is increasing (Keynes 1937, p. 247). Keynes called this the ‘finance motive’ for holding money, and it actually applies to all categories of planned expenditures. That is, money held to satisfy the finance motive is a function of planned expenditures or aggregate demand (Davidson 1965). This will be discussed in greater detail below. Keynes was careful to distinguish his own theory of the determination of interest rates from that of Ohlin, Robertson, and

Hicks, who argued that the interest rate is determined by the intersection of the demand and supply of credit or loans. He argued that their view ‘is exactly the same as the classical doctrine, over again, to the effect that the quantity of saving depends on the rate of interest’ (Keynes 1937, p. 244). Keynes said their view is concerned ‘with changes in the demand for bank borrowing, whereas I am concerned with changes in the demand for money’ (Keynes 1937, p. 245). What Keynes appears to have meant is that the demand and supply of loans in a given circuit cannot place pressure on the interest rate as long as bank loans advanced to finance spending immediately return to banks. Only if bank advances leak into cash 158

hoards would there be pressure on interest rates. Given a degree of liquidity preference of the public and given the willingness of banks to satisfy the demand for hoards, the interest rate will be established without reference to the demand and supply of loans. However, if there is an increase in planned expenditures, this may be accompanied by an increase in hoards to satisfy the finance motive. Therefore, an increase in activity which entails an advance

of loans, some of which will not return immediately to the banking system due to the finance motive, is likely to place upward pressure on interest rates unless banks are willing to accommodate the rising demand for hoards. Thus, as Kregel (1986a) argues, banks play an important role in ensuring that a rise in investment does not force interest rates to rise as demand for hoards increases. Expansion of investment is subject only to a liquidity constraint: the banking system must allow its balance sheet to adjust by satisfying the additional demand for hoards at the prevailing interest rate. Normally, banks will advance credit to meet an increase in the demand for money used to finance investment expenditures with no pressure on interest rates. However, there will be upward pressure on interest rates if first, the liquidity preference of the public rises so that bank credit does not

return

to banks,

and

second,

if banks

will meet

the extra

money demand only at higher interest rates. Keynes argued: [T]oo great a press of uncompleted investment decisions is quite capable of exhausting the available finance, if the banking system is unwilling to increase the supply of money and the supply from existing holders is inelastic. (Keynes 1973, p. 210)

As Chick (1983, p. 240) argues, consumption and investment spending, as well as profit and wage income, primarily remain within the banking system as bank liabilities are used to circulate such flows as a ‘revolving fund’ of finance. The revolving fund itself would finance investment and would place no pressure on interest rates (Keynes 1973, p. 209). An increase in investment (or any other kind of spending) would require a larger revolving fund, and would have to be met by banks. But only an increase in demand for hoards or decrease in willingness to supply hoards would affect interest rates. The supply of liquidity would depend on the terms ‘on which the banking system is ready to become more or less ; unliquid’ (Keynes 1973, p. 219). motive finance the between relation the on focused While Keynes and planned investment, Davidson (1978) has extended the analysis

159

to include other components of aggregate demand. He defines the transactions demand for money as a hoard which is held in anticipation of spending to satisfy the finance motive - and not as money which is held as some proportion of income which has been received. As Keynes had suggested, it is quite possible that an increase in planned government expenditures might be associated with increased hoards if the government borrowed money in anticipation of future spending. Similarly, planned increases in consumption or exports could also increase the transactions demand for money. Furthermore, planned shifts of spending from one component to another might increase transactions demand for money even in the absence of an increase in aggregate demand. For example, if consumers have lower access to credit than do firms, a planned shift from investment spending to consumption spending would increase the need for transactions balances. If transactions demand for money rises and banks are unwilling to satisfy the additional

demand,

interest rates will rise. However,

both Keynes and Davidson suggest that banks might be willing to satisfy rising transactions demand through overdraft facilities. To briefly summarize Keynes’s argument: The demand for hoards (determined by psychological propensities and the finance motive) and the supply of hoards (determined by the willingness of banks and others to provide cash now for cash later) together determine the interest rate. ‘If the banks stand firm, an increased propensity to hoard raises the rate of interest, and thereby lowers the prices of capital assets other than cash, until people give up the idea of selling them or of refraining from buying them in order to increase their hoards’ (Keynes 1937, p. 251). Some Post Keynesians have extended Keynes’s model to broaden the range of assets which can satisfy the desire for liquidity. Thus, ‘hoards’ might include short term liquid assets in addition to cash (Dow and Dow 1989). For example, Mott argues ‘liquidity preference is a theory of the desire to hold short- versus long-term assets’ (Mott 1985-86, p. 230). While a rise in investment must generate sufficient saving to cover it, if liquidity preference is high, households prefer to hold some of their saving in liquid form, rather than purchase long term bonds. Davidson has defined ‘m’ as the marginal propensity to buy long term bonds cut of personal saving (Davidson 1978, p. 299). If m is less than unity, then even though investment generates saving, some of that saving will be held in liquid form so that some of the long term bonds issued to finance investment cannot be sold to households.

160

In this case, a

higher interest rate on long term assets relative to that on short term assets is required to induce households to take up long term bonds issued to finance investment projects (Davidson 1986a). Otherwise, financial institutions will be forced to hold the long term bonds so that households can remain liquid. As Kahn argues, the ‘quantity of money is the means by which the public hold that part of their wealth which is looked after by the banking system (Kahn 1954, p. 238). In other words, the banks can take up the long term instruments and allow the public to hold short term instruments (‘money’). The prices of long term assets (or the long term interest rate) must be at a level which ensures that the public will hold those securities which the banking system will not* (Kahn 1954, p. 238). Thus, if we take ‘m’ as an inverse function of the liquidity premium (the differential between interest rates on short term and long term assets), high liquidity preference of the public means that it would take a prohibitively high liquidity premium to raise m close to unity, so that the banking system is required to remain illiquid in order that the public can hold liquid assets (money). Liquidity preference has also been related to expected profitability: as profit expectations increase, the liquidity premium required to induce households to relinquish hoards falls. In other words, m might be a positive function of profit expectations. For example, Mott argues ‘...the state of liquidity preference is governed by the profitability of business’ (Mott 1985-86, p. 231). Robinson argues that if profit expectations rise, this will cause a rise in investment. If the banking system meets the increased needs of trade by allowing its balance sheet to expand, there is no resulting rise in interest rates. Even if m is less than one, banks

willingly take up the excess long term bonds. In fact, one expects that this would be the case, as bankers should also be subject to the increased optimism. Thus, ‘the improved prospect of profit counts twice over - once in promoting investment at a given cost of borrowing and once in lowering the cost of borrowing’ (Robinson 1979, p. 157). This also explains why money can’t be neutral: as liquidity preference declines and banks willingly meet the demand for credit by expanding their balance sheets, the prices of assets are raised, inducing production through the multiplier until saving has risen to equality with investment and the marginal efficiency of capital has been driven down to equality with interest

rates.

On the other hand, a collapse of profit expectations will probably

161

be associated with a rise of liquidity preference. The public will

want to shift out of long term assets and into liquid assets, while banks will be reluctant to supply short term liabilities and hold long term bonds. Thus, low profit expectations count ‘twice over’: they reduce the incentive to spend on investment goods, and they reduce the willingness of banks to provide credit at the existing interest rate. Thus, we expect that rising liquidity preference and falling profit expectations would be associated with rising interest rates. All of this is consistent with the view of Gurley and Shaw cited above (in Chapter 5) - investment is financed through expansion of credit. There is no reason to expect that an increase in the supply of short term credit should immediately cause interest rates to rise, since this supply is essentially demand determined. However, as firms increase production and income is generated, surplus sectors accumulate net wealth. Gurley and Shaw argue that the surplus units may want to diversify portfolios. Thus, a liquidity premium is required to induce surplus units to hold net wealth in the form of long term bonds. Alternatively, surplus units can hold the liquid liabilities of financial intermediaries, and financial institutions can

take up the long term bonds. In any case, continued growth of (debt financed) investment may eventually push up interest rates due to the diversification motive. Finally, liquidity preference determines the premium required to induce surplus units to hold long term assets. This then sets the standard to which the marginal efficiences of all assets must conform. Thus, liquidity preference goes into the determination of the prices of assets and thereby affects investment and aggregate demand, in general. There is no reason to expect the system to generate full employment because there is no guarantee that the interest rate set by liquidity preference will be low enough to generate sufficient investment.

Flows, stocks, money demand, and liquidity preference In Keynes’s theory, the interest rate is strictly related to stocks. A great deal of confusion has resulted because his critics (and even many of his followers) have failed to distinguish consistently between stocks and flows and between liquidity preference and the demand for credit. Investment and saving are flows, and investment expenditures 162

necessarily

generate

investment flows does

equivalent

saving

flows.

The

not directly affect interest

finance

of

rates, just as

saving flows cannot influence interest rates. Keynes argues:

Credit, in the sense of ‘finance’, looks after a flow of investment.

It is a revolving fund which can be used over and over again... The

same

‘finance’ can

tackle one

investment

after another.

(Keynes 1973, p. 209)

That is, given a revolving fund of a certain size, it can continuously finance a given flow of saving and investment. However, saving and investment flows have effects on stocks. As discussed above, the

decision to increase investment may lead to the accumulation of a hoard of cash prior to expenditure to satisfy a ‘finance motive’. Thus, the decision to invest can increase the stock of hoards if the

banking system accommodates by supplying more cash. If the banking system does not satisfy the desire to hoard, interest rates must rise so that some give up the attempt to hoard. Therefore, the impact which an increase in investment flows has upon stocks can

affect interest rates.”

Investment spending provides income which can be saved. Increases in investment which are deficit financed will at the same time provide long term assets which can be accumulated by savers. However, since we normally expect that the marginal propensity to purchase long term bonds out of increments to personal saving is less than one, portfolio preferences will require a premium to induce the public to hold the long term assets issued by firms to finance positions in the capital assets accumulated as a result of investment spending. Note that the premium is not due to lack of saving - the investment spending has created equivalent saving. The public could save until it is ‘blue in the face’ and this would not eliminate the premium. It is the preference for liquid assets which necessitates a premium to clear the long term bond market. In Chapter 1 above, a distinction was made between money demand and liquidity preference. Money demand was defined as a willingness to issue debt to finance spending. Thus, money demand is linked primarily to expenditure flows.’ Money demand thus defined does not directly impact interest rates. However, an increase in money demand which is met by expansion of bank balance sheets as banks provide credit will affect stocks. In this way, money demand may indirectly affect interest rates, as will be discussed below. Liquidity preference was defined as a desire to exchange illiquid items on a balance sheet for more liquid items,

163

or even to decrease the size of a balance sheet by retiring debt. Liquidity preference is thus related to stocks and directly influences interest rates. In an economy with developed financial institutions, with deposit insurance, and with a central bank which is willing to enter as a lender of last resort, hoards of cash or narrow money are relatively unimportant. Furthermore, lines of credit or overdraft facilities can satisfy the ‘finance motive’ without requiring cash hoards. Thus, Keynes’s notion that the interest rate is determined by the supply and demand of hoards does not strictly apply to our economy. However, if liquidity preference is restated as ‘a theory of the desire to hold skort- versus long-term assets’ then it is consistent with the analysis presented above (Mott 1985-86, p. 230). Banks typically provide firms with short term finance, which is retired when an investment project is complete and long term bonds are floated. Investment spending generates an equivalent amount of ‘savings’, received primarily as bank deposits by surplus units.’ Liquidity preference can be measured as the interest rate differential required to induce surplus units to exchange bank deposits for long term bonds (or as the interest rate differential required to induce banks to purchase long term bonds on the basis of an issue of shorter term liabilities). This measure of liquidity preference should tend to fall as profit expectations rise. Since a rise in profit expectations should tend also to cause a reduction in a bank’s estimate of the risk involved in making loans, this will tend to lower interest rates. Thus, Keynes is quite correct in rejecting any flow determination of the interest rate. The interest rate is determined by stocks. Of course, this does not imply that flows do not have an indirect impact on interest rates, since flows affect stocks. An expansion of bank balance sheets may affect interest rates precisely because banks are concerned with their balance sheet position (a stock). Similarly, if households decide to shift their portfolio to liquid positions, this will also affect interest rates, which must adjust so that households will become satisfied with the existing amount of liquidity. This may not be a smooth adjustment process - it may generate a panic if households realize that liquidity cannot be obtained.” The important point, however, is that an increase in spending, by itself, need not have any impact on interest rates if banks are willing to accommodate by expanding their balance sheets (and taking more illiquid positions) without requiring interest rates to rise. In other words, it is liquidity preference which 164

determines

interest rates, and not money

demand

(as the terms

were defined in Chapter 1). As Dow and Dow (1989) argue, the supply of credit and liquidity preference are interdependent. When liquidity preference is low, banks meet the demand for credit by expanding their balance sheets. As liquidity preference rises, banks become increasingly unwilling to meet the demand for credit, so that credit rationing replaces accommodative behavior. ‘The overall picture, then, is of credit creation accommodating demand when liquidity preference is low, but not when liquidity preference is high’ (Dow and Dow 1989, p. 161). By focusing on the balance sheets of both borrowers and lenders, the relation between liquidity preference and the supply of credit becomes apparent. Keynes, however, tended to strip the argument of institutional detail.? He argued that the interest rate is determined by what people think it will be. The interest rate is partially determined by expectations and conventions - but it is also influenced by competition from nonbank institutions, by the preferences of the public, by central bank actions, and by foreign interest rates. The remainder of this section gradually develops the outlines of a model of the determination of interest rates which incorporates institutional detail and these other considerations. As a first approximation, I will assume that commercial banks provide short term finance - to the investment goods producer - of working capital expenses while investment goods are in the process of production. The investment banks provide long term loans to the investment goods purchaser, whose purchase enables the investment goods producer to repay the commercial bank loans. The investment banks then sell long term bonds to fund long term loans that finance positions in the capital assets which are the result of the investment expenditures. These assets will be used to generate gross profits, out of which interest payments on long term loans are met, and which enable the investment banks to meet

payments on the long term bonds.” The working capital provided by commercial banks (to investment goods producers) will return as deposits to the banking system unless there are leakages into cash hoards, leakages out of the country, or leakages into other types of financial assets (such as

long term bonds).'’ When these banks provide short term finance,

they must ensure that the interest charged on loans is sufficient to cover the costs of issuing liabilities (by attracting deposits through reflux or by borrowing from the public, from other banks, or from 165

the central bank) and to provide a sufficient margin to cover risk of loan default. Short term interest rates will tend to rise as banks provide more working capital (by expanding their balance sheets) if leakages increase or if perceived risks of expanding balance sheets rise. As I discussed above, leakages into cash hoards are small and

rather predictable in countries with modern financial institutions. I will ignore them because normal central bank provision of reserves (for example, as the central bank provides state finance) more than compensates for cash hoards of the public. Leakages out of the country may be substantial in an integrated world economy. In this case, the domestic banks must borrow from foreigners, or the central bank must step in to provide reserves to cover such leakages. Again, I will ignore such leakages to simplify matters. (Furthermore, it is unclear that an increase in working capital would systematically affect leakages abroad.) Finally, those receiving working capital expenditures might choose to buy assets issued by individuals or institutions other than commercial banks. In this case, the interest rate on commercial bank deposits must be

competitive with that on alternative assets. Given the wide range

of short term assets now available in the US, commercial banks

face substantial competition from other institutions. The degree of competition will be an important determinant of the interest rate paid on commercial bank deposits. However, as these are still active balances (received by workers in the investment goods sector), I will assume that rising working capital expenses don’t affect such leakages into nonbank assets - or that those who do sell such nonbank assets return funds to the banking system as deposits. Growing balance sheets may well affect the commercial bank’s perceptions of risk. As a bank provides short term loans and issues liabilities, it leverages equity and cash assets. Prudent leverage ratios are established by a combination of experience and rules of thumb. Innovations and/or revisions of commonly accepted views regarding appropriate levels of leverage allow banks to expand portfolios

without

raising

interest

rates.

Eventually,

however,

financial institutions will come up against the maximum ‘prudent’ leverage ratio. Another innovation, or a revision of standards, will again allow expansion without a rise in interest rates. However, in the meantime, only a rise in the liquidity premium would induce banks to expand loans further. That is, the money supply curve is upward-sloping relative to interest rates in the absence of innovations or revisions of norms of behavior, even though there

166

is no strict quantity constraint on bank lending. The short term interest rate on working capital rises due to flow effects (credit advanced to finance expenditures) on stocks (balance sheet of the bank). It is the ‘liquidity preference’ of the banks which causes short term interest rates to rise. Thus, the price of credit and the liquidity preference of the bank are related through credit flow effects on the balance sheet position of the bank. Once the investment goods have been produced, the investment bank intends to sell long term bonds and issue long term loans so the investment goods purchaser can buy the completed goods and enable the producer to retire the short term finance. Sufficient saving has been generated and received as the deposits of the commercial bank.” If the investment bank can float the long term bonds by selling them to these surplus units, those deposits will flow to the investment bank and then back to the commercial bank, where they will be destroyed as the short term loans are

retired.” The investment bank must charge a high enough interest

rate on the long term loan provided to the investment good purchaser to cover the costs of doing so plus a margin to provide profits and to compensate for perceived risk. Thus, the long term loan rate of interest must exceed the interest rate required to induce the public to give up commercial bank deposits for long term bonds. Again, the investment expenditure may impact long term interest rates due to effects on stocks - the balance sheets of the investment banker and the public. If the investment banker has underestimated the liquidity preference of the public, he will have underestimated the interest rate which will be required on the long term bonds. As the investment good purchaser normally negotiates the long term loan rate of interest before the investment project is undertaken, it is the investment banker who will usually absorb the higher cost if the liquidity preference of the public is higher than anticipated. This is always a risk, because the public’s preference for liquidity certainly might change over the period required to float the bonds. There may be a tendency for investment bankers to raise the rate of interest charged on long term finance as the quantity of finance extended increases. Investment bankers, like commercial bankers,

must be concerned with leverage ratios because repayment of loans is uncertain and because the price at which bonds can be sold is also uncertain. Thus, as the balance sheets of investment bankers

expand, there will be a tendency for long term interest rates to rise

167

as prudent leverage ratios are reached and exceeded. As described above, revisions of rules of thumb or innovative financing methods

may temporarily forestall increases of interest rates. Finally, as the balance

sheets of the public absorb more

long term bonds, the

liquidity premium required to induce the public to exchange short term deposits for long term bonds may also rise. In short, as a first approximation, short term interest rates on loans are a function of the costs incurred by commercial banks as they issue deposits, in competition with alternative liquid assets which might be held by the public, and make loans. The differential between the deposit and loan rates of interest must compensate banks for perceived risk, and is set primarily through custom and rules of thumb. The long term bond rate of interest is then established at a level sufficient to induce the public to hold saving in the form of bonds, while the long term loan rate of interest is set high enough to provide a differential to compensate investment banks for perceived risk - again, primarily determined by rules of thumb. A given level of investment (and saving) should place no pressure on interest rates as long as liquidity preference and rules of thumb don’t change. A rise of investment spending will increase the size of balance sheets and may affect both short term and long term interest rates if perceived risk or liquidity preference rises, or if prudent leverage ratios are exceeded. However, I have ignored actions of the central bank, and have concentrated only on additions to stocks. Additions to portfolios resulting from flows are small in relation to the size of stocks. Thus, decisions to adjust portfolios may have an impact on interest rates in addition to the impact caused by flows. While these considerations complicate the analysis of the determination of interest rates, they do not change the conclusion that any impact flows may have on interest rates must come through impacts on stocks. For example, interest rates tend to rise during an expansion because the Fed frequently decides to push up interest rates in the apparent belief that the expansion will fuel inflation. Thus, the Fed raises the discount rate and, perhaps, attempts to constrain reserves. This directly raises costs incurred by commercial banks. Commercial banks will attempt to find alternative sources of reserves so they can continue to meet the demand of creditworthy borrowers. However, tight money policy can eventually be expected to force up short term interest rates. This need not immediately reduce the rate of growth of balance sheets. Rising short term

168

interest rates may generate expectations of rising long term and short term interest rates and falling bond prices. This may increase the liquidity preference of the public, which now requires higher long term rates to compensate for potential capital losses on bonds. As defaults spread, profit expectations may fall and liquidity preference may rise further. Commercial banks and investment banks revise rules of thumb downward as they take account of rising defaults on loans. Interest rates on loans rise and banks provide credit only to established customers. The demand for loans may be extreme, but this demand is for liquidity required to meet existing payment commitments, and not to expand expenditures. In summary, liquidity preference is measured by the premium required to induce agents to hold illiquid portfolios. Because an increase in the size of a bank’s balance sheet is associated with increasing leverage and declining liquidity, beyond some point banks will raise interest rates as an expansion proceeds. This tendency for interest rates to rise as deficit spending increases can be offset by many factors, including: (1) a shift of depositor preferences toward illiquid positions (that is, liquidity preference falls); (2) intermediation by other (nonbank) financial institutions; (3) central bank provision of reserves through open market purchases and/or low discount rates; (4) optimistic expectations concerning the future (so that liquidity preference falls as profit expectations rise); and (5) financial flows from abroad (which increase the liquidity of bank balance sheets). Thus, liquidity preference theory is not inconsistent with the endogenous money approach. Investment is not constrained by the supply of saving, nor is it constrained by the quantity of money (narrowly defined). As banks are not strictly quantity constrained, their willingness to make loans must be a function of their willingness to take illiquid positions in assets. This willingness will surely be affected by the state of long term expectations regarding profitability. However, it will also be affected by the balance sheet

positions of banks, by portfolio preferences of the public, and by

the policies of the central bank. Liquidity preference theory is not simply a ‘flawed’ expedient adopted in a critique of loanable funds theory. It is a necessary part of Keynes’s general theory. Liquidity preference theory Is closely

tied

to Keynes’s

theories

of effective

demand,

of the

relation between the marginal efficiency of capital and investment, 169

and of the finance motive. Liquidity preference theory is not inconsistent with the endogenous money approach. Indeed, liquidity preference theory strengthens the endogenous money approach, just as the endogenous money approach strengthens Keynes’s general theory.

Liquidity preference and IS-LM models Liquidity preference theory was supposedly incorporated in ‘Keynesian’ IS-LM models which ostensibly utilize liquidity preference theory rather than loanable funds theory. The IS curve is the locus of points showing flow equilibrium in the goods market (where investment equals saving), while the LM curve is the locus of points representing stock equilibrium in the money market.

However, IS-LM models deviate from the approach outlined in the previous sections in at least three ways: they assume a fixed money supply; they misspecify the money demand function; and they assume quick adjustment to equilibrium in both the goods and the

money markets. IS-LM models assume a fixed money supply in order to show that increased spending raises interest rates. While deficit spending is not saving constrained in these models, it is ‘finance constrained’ because excess money demand is generated which can be satisfied only through price adjustments. However, if the money supply is endogenously determined by the willingness of agents to engage in deficit spending, then the money supply would not be fixed as

spending increases. An increase in flows of saving or investment will not necessarily affect the interest rate, which is determined by stocks.’* Furthermore, the interest rate is determined at the

beginning of the income period, before the spending and saving occur - and not simultaneously with flows, as in the IS-LM model. While flows do affect stocks, and may thereby indirectly affect interest rates, this then affects the interest rate which will prevail in the next income period. Therefore, the interest rate cannot be

determined by the intersection of an LM curve with an IS curve. As Davidson (1965) argued, IS-LM models also misspecify the transactions demand for money. In these models, the demand for

money is a function of income, while in Keynes’s model, the demand for money is a function of aggregate demand. Thus, money demand rises when planned expenditures increase to satisfy a ‘finance motive’ for holding money. In this case, both the IS curve

170

and the LM curve would shift each time planned expenditures change (the IS curve shifts as aggregate demand changes, and the LM curve shifts because aggregate demand determines money demand). Similarly, Kregel (1988) argues that a change in liquidity preference will shift the LM curve (unless the money supply completely accommodates) and will also shift the IS curve. The IS curve is drawn for a given set of expectations regarding the marginal efficiencies of assets. If liquidity preference declines, the relative return from capital assets will rise, since the return from

fixed capital relies mainly on profit expectations while the return from liquid assets comes mainly from their liquidity. As liquidity preference falls, the IS curve is likely to shift out. Thus, the economy cannot be dichotomized into real and monetary sectors, and the IS and LM curves are not independent. Minsky (1981) also argues that the IS and LM curves must shift relative to each other as price levels change. Money goes into the determination of the prices of capital and financial assets, so the LM curve has to do with the price system of assets. The prices incorporated

into the IS curve, however,

must be the prices of

current output. This implies either that the two curves are in different planes, or that a change in either price system will cause relative shifts of the curves. The supply price of current output of capital goods (underlying the IS curve) must cover production and finance costs. The demand price of capital assets, however, is determined by the present value of profits to be generated over the life of the assets and must, for a general equilibrium to hold, equal the depreciated historical cost of the capital assets. A change in the quantity of assets, in profit expectations, or in liquidity preference will cause a change in the price systems - forcing a shift of the LM curve relative to the IS. For example, an increase in liquidity preference raises interest rates, lowering the demand price of capital assets as the discounted returns from holding capital assets fall, and causing

a shift of the LM

curve.

At the same

time,

however, an increase in interest rates raises the financing costs of producing current output, which also shifts the IS curve as current output prices rise. If the supply price of capital assets rises above the demand price, investment collapses. In any case, the IS and LM curves are interdependent. This means that the familiar textbook expositions which hold one curve constant while the other is shifted (for example, it is common to increase the money supply while the IS is held constant) are flawed. al

A further confusion arises here because the orthodox version s. does not distinguish between active and inactive money balance and receipt income n betwee gap Money held to bridge the expenditure represents an active balance which will return to the revolving fund during the income period. Given conventions concerning methods of income payment and patterns of expenditure, active money balances will not affect interest rates because withdrawals from the revolving fund will be matched by injections into the fund. If most short term credit is created to finance the wage bill, and if wages are paid in the form of demand deposits which are regularly spent during the income period, then outstanding demand deposits held by workers will average one-half the wage bill. During an income period, all these active balances will return to the revolving fund. At the end of any income period, active money balances will be zero for each agent so that all the

demand deposits will be absorbed into inactive balances, will return to replenish the revolving fund of short term finance, or will be extinguished as short term loans are retired. However, inactive balances held to satisfy the precautionary or speculative demands for money will not necessarily return to the revolving fund during an income period. Thus, the desire to hold inactive stocks of money will influence interest rates. Clearly, transactions balances which are related to income represent active money and should not

be included as a determinant of interest rates. However, balances held as a function of aggregate demand (planned expenditure) as part of a finance motive represent inactive balances and will affect interest rates, as Keynes argued.” The third deviation of the IS-LM model from Keynesian theory regards its equilibrium method. In IS-LM analysis, saving and investment respond rapidly to changes in interest rates. This gives a downward-sloping IS curve.° However, as Dow and Saville (1988) argue, in the short run, neither investment nor saving is very sensitive to interest rates. A moderate increase in interest rates is unlikely to have much impact on investment in the short run during an expansion. In a technologically advanced economy, production of most investment goods is undertaken on order and requires time for completion. A rise in interest rates is not likely to cause firms to abandon projects already in the process of production. Of course, as long as investment is insensitive to short period fluctuations in interest rates, saving will also be insensitive. This does not mean, however, that investment is completely

unresponsive to interest rates. A large increase in interest rates

172

causes a ‘present value reversal’, forcing the marginal efficiency of the capital asset to fall below the interest rate. If the long term interest rate is also pushed above the marginal efficiency of capital, the project may be abandoned. In this case, both investment and saving will fall. However, for moderate changes in short term interest rates, the IS curve would be vertical.

Eichner (1985) also argued that empirical evidence shows that investment is not responsive to interest rates. Furthermore, he argued that the evidence suggests that monetary disequilibrium, rather than equilibrium is the normal case. In other words, there

is strong evidence that the interest rate does not adjust to equilibrate money demand and money supply, and that credit rationing is pervasive. He argued that those who adopt the IS-LM framework have yet to offer an empirically operational definition of money market equilibrium (ibid., p. 185). Given the Cooley and LeRoy critique discussed above (Chapter 3), it seems unlikely that such will ever be developed. Thus, given that the IS and LM curves are not independent, given that the LM curve would likely be nearly horizontal (since money demand is met by money supply), and given that the IS curve is normally vertical, the IS-LM model is not very useful. It neither represents the Keynesian system adequately, nor does it provide a logical system for the determination of income or interest rates. In fact, the IS-LM model appears to be much closer to the old neoclassical model than to the Keynesian model which it is supposed to represent. Indeed, Hicks’s original formulation was designed to present the neoclassical model in a Keynesian perspective - and not to present Keynes’s theory (Kregel 1988). S. C. Tsiang (1956) showed that the textbook version of liquidity preference theory is identical to loanable funds theory, in the sense that the two sets of demand and supply functions, ie., the demand for and the supply of loanable funds, and the demand for money to hold and the stock of money in existence--would determine the same rate of interest in all circumstances, if both sets of demand and supply functions are formulated correctly in the ex ante sense. (Tsiang 1956, p. 539)

In Tsiang’s presentation, the interest rate is determined by the intersection of the money supply with the demand for money, where money demand includes transactions demand as a function of income. He goes on to include money held to satisfy the finance 173

es, however, argued that motive in the transactions demand. Keyn meet expenditures was to the demand for ‘active balances’ used ce motive, and that the finan the distinct from money held to satisfy the liquidity theory of the finance motive ‘is the coping-stone of ng argues that Keynes rate of interest’ (Keynes 1937, p. 667). Tsia ting an exogenous money was merely ‘confused’. However, by adop nd, Tsiang reduces supply and by misspecifying money dema h is ‘identical’ to whic l liquidity preference theory to a mode loanable funds theory. nts money demand In another presentation, Tsiang (1980) prese investment, and ed plann n, mptio as a function of planned consu inflation.” He then other variables such as interest rates and y is identical to ‘proves’ that Keynes’s liquidity preference theor in thrift will decrease loanable funds theory and that an increase g

s ‘everythin interest rates. He argues that his model prove rather Keynes is ‘it and Robertson tried to tell us is quite right’ pp. 472-4). 1980, ng (Tsia ’ wrong and his followers that were in the entirely was s Keyne that is s prove lly However, what Tsiang actua correct. income less Tsiang defines planned saving (‘thrift’) as last period’s that money s planned consumption for this period. He then assume saving, so d planne less demand is a function of planned investment planned in e declin a nt, defined. Holding planned investment consta will thrift) or saving d planne in se consumption (which is an increa when this t agains argued s Keyne er, decrease interest rates. Howev he said ‘If the reduction in consumption posited by Mr Robertson e leaves aggregate income unchanged, there is no reason to suppos tion’ conges that it will reduce the demand for cash or relieve the n (Keynes 1973, p. 232). Indeed, a decline in planned consumptio would

decrease

interest

rates not because

saving increases

but

because planned spending falls. ‘In this case it is certainly not an increase in saving which has relieved the congestion, since there has been no increase in saving’ (ibid., p. 232). Thus, as Keynes argued, if planned spending falls, less cash is demanded to satisfy the finance motive, so that interest rates may fall. It is only by reducing cash hoards in anticipation of spending less that the public can make its plan to increase ‘thriftiness’ known and thereby affect interest rates. Keynes preferred to call this a decline in liquidity preference since the interest rate falls because the public gives up cash and not because saving rises. Tsiang calls this an increase in thriftiness because the public has decided to spend less in the future. However, it is only the release of cash, and not

174

any increase in saving, which would lower interest rates. Indeed, saving would rise only if investment increased - in which case, the

fall of planned consumption would be replaced by increased investment so that the interest rate would not fall. If there were some way for the public to let its plans be known other than by releasing cash, then one wonders why Tsiang has only included planned saving in this period as a determinant of interest rates. Why couldn’t the public let it be known that it will ‘save’ more (spend less) in the next period, or the period after that, or a period far into the future? That is, why wouldn’t ‘planned saving’ throughout all eternity (past and future) be available to ‘finance’ investment? If ‘planned saving’ is a source of loanable funds, why should it be limited to immediate plans? Clearly, this would

reduce

Tsiang’s

model

to timeless

nonsense,

and, in a

timeless model, there would be no room for liquidity preference. Tsiang (1989) has made another attempt to specify a loanable funds model which has room for money. He argues that a loanable funds model should concentrate on flows - where interest rates are determined. The part of the money supply which is diverted to the money market becomes a part of the supply of loanable funds and so can reduce interest rates. Money which is ‘sitting still’ as hoards cannot influence interest rates; only money ‘on the wing’ can do so. Thus, the demand

for loanable funds includes investment, deficit

spending by the government, and any increase in hoards, while the supply of loanable funds includes saving and releases of money hoards (dishoarding). Tsiang is confused on two points. First, as Keynes argued, the public cannot increase or decrease hoards (except as the banking system creates or destroys money). A change in the desire to hoard may change interest rates, but will not affect the quantity of hoards. Second, also as Keynes argued, Tsiang must admit that ‘dishoarding’ in the form of rising consumption is as much a supply of loanable funds as is ‘dishoarding’ in the form of an increase in saving (for example, by purchasing long term bonds). Thus, even on Tsiang’s own terms, his loanable funds model is fatally flawed since all flows must be counted as a supply of loanable funds. Indeed, the supply and demand must be identical: all flows are both supplies of loanable funds and demands for loanable funds.

Tsiang’s attempt to introduce money into a loanable funds model makes it clear that such an hybrid model is neither internally consistent nor can it be reconciled with Keynesian liquidity preference theory. 175

theory is identical Duncan Foley (1975) shows that loanable funds f-period models and to textbook liquidity preference theory in end-o ight. Thus, under fores ct in beginning-of-period models with perfe ity preference liquid of ons many circumstances, orthodox versi the key

ver, theory are identical to loanable funds theory.’* Howe y and the theor rence prefe ity liquid s’s rence between Keyne

diffe flows of finance orthodox version lies in Keynes’s insistence that mists continually cannot affect the interest rate. Orthodox econo loans supplied , return to flows such as saving and investment or itly argued explic s and demanded to finance spending flows. Keyne Tsiang and rates. st intere that none of these can directly influence stocks affect flows such how show Foley modify the exposition to these for ned obtai s result which under and show the conditions this , Again . stocks ting resul the for ned obtai s flows equal result is model s’s Keyne in rate st intere The point. s’s misses Keyne the at ) stock determined by the demand and supply of hoards (a beginning of periods in a world of uncertainty. Foley is unable to ity show the equivalence of loanable funds theory and liquid ct perfe ut witho s riod model of-pe ningpreference theory in begin foresight. This, of course, is the world which Keynes modeled.

A planned increase in expenditures will affect balance sheets before the spending actually occurs. That is, if a firm decides to increase spending, it obtains working capital by issuing a liability to its bank. The bank accepts this debt and issues a loan by creating deposits. Thus, balance sheets adjust before the firm actually undertakes a project. Interest rates are established at the beginning of the period - when the commitments are made. Whether the planned increase in expenditures will raise interest rates depends upon a number of factors, which were discussed in the previous section and which will be clarified in the next section. The critique of orthodox liquidity preference theory is strengthened by an inclusion of the endogenous money approach. If the demand for financial flows is met by an increase in financial flows, then clearly the interest rate must be determined elsewhere. In terms of the IS-LM model, the LM curve must be horizontal. Furthermore, as Davidson concluded, each movement of the IS curve causes the LM curve to shift, so the two curves are not

independent. A more promising approach is to return to Keynes’s theory of effective demand, combined with liquidity preference theory and an endogenous money supply.

176

Endogenous money, expectations theory, and the markup approach Some of those who have adopted the endogenous money approach have replaced liquidity preference theory with a markup theory of interest rates. I have already discussed Kaldor’s belief that the central bank determines the short term interest rate to which a risk premium is added to obtain long term interest rates. Like Moore (1988b), Rousseas (1985b) argues that banks are price setters but quantity takers in retail loan markets. On the other hand, they are quantity setters and price takers in markets for wholesale sources of funds. The interest rate charged on bank loans is a markup over the cost of wholesale liabilities issued by banks. Rousseas refers to Kalecki’s oligopoly theory of the markup in which the degree of oligopoly power attained by a firm governs the markup which can be added to prime costs. Rousseas takes the Fed funds rate as an indication of the prime cost of obtaining funds, and argues that the prime interest rate charged on business loans is a more-or-less

stable markup over the Fed funds rate.” Moore (1988a, 1988b) relates the Fed funds rate to the discount rate set by the Fed. The Fed can set the discount rate within a range (the size of the range depends on the actions of other central banks), and bank arbitrage then keeps all other short term interest rates in close alignment. The interest rates charged by banks are a markup over costs, which include costs of wholesale funds, labor costs, and profit on equity invested in the banks. The long term interest rates are then set by arbitrage to equalize the expected holding period yields, so that the long term interest rate is determined by current expectations of the level of future short term interest rates which will be set through Fed policy. Several objections have been raised to arguments such as these. Robinson (1979, pp. 145-8) criticized the view that long term interest rates are merely determined by the expected value of future short term rates. She argued that it makes just as much sense to argue that the long term rates determine the short term rates. Keynes argued that the long term interest rate is determined

by speculation on the long term interest rate itself, rather than by speculation on the future course of short term interest rates (as Hicks and Kaldor had argued) (Kahn 1954, p. 229). Kahn argues

that expectations about the short term rate will influence the long

term rate only to the extent that they affect expectations about the long term rate (Kahn 1954, p. 235). If tight money policy forces up

EF]

to short term rates, and if agents expect higher short term rates term long the about tions expecta then continue far into the future, rate will probably also shift upward. This, according to Kahn, ‘does not really amount to saying more than that the long-term rate of interest is influenced by expectations of banking policy’ (Kahn 1954, p. 234). Kahn goes on to give an example in which the expectations theory provides an illogical conclusion. Assume banks buy short

term bills from the public and sell long term bonds to the public.”

The short term interest rate will fall, while the long term interest rate will rise. According to the expectations approach, the bond rate rises because the expected future short term rates are rising even though the actual short term rate is falling (Kahn 1954, p. 235). Alternatively, assume that the banks buy long term bonds, causing the long term rate of interest to fall. This fall in the long term rate has nothing to do with expectations about the future short term rate. Indeed, the current short term rate would be rising if the banks are financing their bond purchases by selling short term bills. On the other hand, if banks financed bond purchases by issuing deposits, it is likely that all interest rates would fall as the public found itself in very liquid positions. Clearly, simple expectations theory is not a very useful guide in such cases. Expectations theory also experiences difficulty if decisions can be reversed. As Kahn argues, ‘a decision to go long rather than short, or vice versa, is not indissoluble’ (Kahn 1954, p. 236). The purchase

of a short term bill does not indicate that the buyer will remain in bills. When one goes short, the only relevant bond rate is the one

which will prevail when the short term bill matures. While that expectation will be related to the expected bill rate on the same date, and while the bond rate on that date will also be related to

the bond rate on future dates as well as to expected future bill rates, this does not mean that the long term rate is merely the expected value of future short term rates. The course of the price of a bond is irrelevant even for a bond holder. The only relevant price is that which will reign on the day on which the bond will be sold.”' One is not normally forced to sell bonds when a capital loss would be incurred (Chick 1983, p. 215). If one buys a bond with the intention of selling at the end of a year, this could indicate that he believes that on that particular date, the long term rate on his bond will be above the short term rate which will prevail. This is the position of those who adopt the expectations theory. However, it is also possible that he buys the 178

bond because he believes its rate is greater than the rate likely to prevail one year hence on long term bonds. This was Keynes’s

position.” The

markup

determinant

theory

essentially

takes

Fed

behavior

as

the

of the discount rate, which then determines the fed

funds rate (or some other wholesale interest rate), which is the major component of prime costs.” Given that wholesale cost of funds, banks add a markup to establish the retail loan rate of interest. The money supply curve is then horizontal at that rate of interest. However, I have argued above that the money supply curve is not horizontal at a given rate of interest because banks face uncertainty. Banks are concerned with the ratio of loans to Safe assets such as government bonds plus reserves, and with the ratio of loans to equity. Should borrowers default on loans, liquidation of safe assets permits banks to meet payment commitments temporarily, while equity allows banks to absorb loan losses. The money supply curve is not horizontal even if all created deposits return to the banking system, because banks must be concerned with increasing leverage ratios. Even where additional loans can be made which are no more risky than loans which have been made previously, banks may require higher interest rates to compensate for riskier leverage ratios. In Chapter 2 above, I presented evidence that there was very little price competition in English banking before the twentieth century. Furthermore, banks carefully chose their customers, and

would even refuse deposits if they did not wish to establish relations with a particular individual or firm. Gradually, decision making became somewhat more centralized and routinized. However,

because

great uncertainty must

always exist, it .is still

necessary to establish close relations with customers. Minsky has said that the first rule of banking is that customers lie. That is, customers will try to place their financial position in a favorable light. According to Minsky, the second rule of banking is that bankers lie, too. That is, banks also try to present their position favorably to calm the fears of depositors. Even if neither bankers

nor their borrowers lied, banking is a risky business because the future is uncertain and because money commitments are

necessarily future oriented. For these reasons, banks use a combination of price and quantity constraints to ration credit. A substantial literature has developed to provide explanations for quantity rationing by banks. Calomiris, et. al., (1986) examine the credit market for small farmers, who typically borrow from small, £79

local banks. They find ‘information-intensive localized customer borrowing relationships, rather than impersonal debt and equity

markets, are important in agricultural finance’ (Calomiris, et. al., 1986, p. 443). In such markets, prices do not ration credit. Bankers

fear that if they raise interest rates too high, farmers who are good

credit risks will go elsewhere, while farmers who are bad risks will

remain. Because banks cannot obtain complete information regarding the riskiness of borrowers, an increase in interest rates may actually lead to a loan portfolio with a greater number of potential

defaulters.

Thus,

banks

attempt

to

establish

close

relationships with local farmers to screen potential defaulters. They then attempt to meet the demand for credit by those farmers with whom they have established relations. However, these small local banks have trouble diversifying assets (they typically hold a large percentage of their assets in the form of agricultural loans). When a surprise event temporarily lowers farm income (for example, a fall in farm prices), some farmers are unable to meet payment commitments. If their bank has not sufficiently diversified, it may fail. As local banks fail, farmers become credit constrained because

they do not have access to other credit markets. Blinder and Stiglitz (1983) apply a similar analysis to nonagricultural loans. They argue that capital markets are not perfect: some borrowers may default, and it is difficult or impossible to gain perfect information regarding which ones will. Banks specialize in acquiring information about default risk, but this information is firm-specific and difficult to sell. Banks devise nonprice screening mechanisms because rising interest rates may actually decrease expected returns due to adverse effects on the mix of applicants (less creditworthy applicants are more willing to pay the higher interest rates, particularly if they expect to default).

Banks try to devise contracts with incentives to prevent default. For example, the future availability of credit may be tied to current payment performance. Ties are established between banks and borrowers, so that borrowers are seen by different banks as imperfect substitutes. Thus, loans would also be seen as imperfect

substitutes, creating an obstacle to the development of secondary markets for loans. Similarly, Stiglitz and Weiss (1981) argue that banks use quantity rationing rather than price rationing because higher interest rates would lead to adverse selection in which only the riskier borrowers would accept loans. Furthermore, higher interest rates would encourage borrowers to take on riskier projects with the hope of

180

earning higher gross profits in an effort to meet debt commitments

(adverse incentive). Thus, quantity rations represent maximizing

behavior. Bernanke (1981, 1983) argues that intermediation between borrowers and lenders requires non-trivial market making and information gathering services. Borrowers and lenders develop relations to lower the costs of credit intermediation. He argues that bank failures (such as those which occurred during 1930-33) raise the cost of credit intermediation by eliminating these special relationships. Thus, potential borrowers (whose bank has failed) are credit constrained until they are able to develop a new relationship with another bank. While they are credit constrained, their spending declines, which exacerbates the disruption due to aggregate demand effects. He argues that the Great Depression was propagated by rising costs of credit intermediation, which was a result of bank failures. Finally, Dow and Saville (1988) argue that banks are risk averse and are faced with uncertainty concerning loan defaults. They do not know what interest rate would adequately compensate for risk of default, since such risks are not known with precision. Because banks pool risks, they are partially protected from random loan defaults. However, they can further reduce potential losses by placing quantitative limits on each customer. Although some part of the loan can be made against collateral, the typical loan will be

made against expected future income. Thus, the bank limits the loan amount to some portion of secured property and expected income. The bank will also provide credit only to customers for whom credit worthiness has been established. This assessment cannot be based merely on objective considerations, but must also be subjectively determined by intimate knowledge of the customer. The significance of such considerations is that the interest rate is not set to clear the market. There is always a ‘fringe of unsatisfied borrowers’ - those who receive no credit or less than desired. This means that the money supply curve is not horizontal, and that banks are not ‘price setters and quantity takers’ in retail loan markets. While Moore notes that the typical credit line is only half-utilized, this does not mean

that banks have met all credit

demands. Credit lines are not equally distributed. As Rousseas (1986, p. 65) argues, most unused overdrafts remain unused as they are not where they are most needed. Like Minsky, Dow and Saville argue that bank perceptions of appropriate leverage ratios can change. Higher income levels

181

encourage banks to make loans on the basis of greater debt to-equity-ratios. Revisions of rules of thumb permit greater leverage of prospective income flows. Bank managers follow only general rules, but each tries to keep in step with other managers. Each bank will try to grow at about the same rate as other banks so that deposits are not lost through the clearing mechanism. Overly rapid growth of a particular bank would also raise suspicion about the quality of the bank’s loans. This would raise the cost of issuing liabilities and could lead to a run on certificates of deposit. Each bank will also follow similar pricing schedules - partly due to competitive pressures, but also to avoid raising eyebrows. A bank which pays more for deposits will be thought to be experiencing trouble attracting deposits, while one which charges less on loans will be thought excessively risky. As each bank must maintain the confidence of other banks (which are potential sources of interbank loans), none can deviate substantially from norms of behavior. However,

innovations and revisions of norms

do occur

and can

greatly affect the price and quantity of debt acceptable to banks. These considerations strengthen the argument that credit constraints are normally operative. An individual bank which has been growing more rapidly than others might refuse credit or demand a high interest rate in order to demonstrate that it is prudent. The prospective borrower may then be credit constrained. He may not be able to go to another bank since banks require information about creditworthiness - and this is costly and requires time to obtain. Thus, micro considerations

lead to macro

credit

constraints. This would be particularly true as expectations fall in a recession, during which individual banks have an even greater incentive to avoid growing faster than others. This is because banks which grow faster than average will lose reserves through the clearing mechanism and will face escalating costs of recapturing these as expectations decline and liquidity preference rises, forcing interest rates to rise. Dow and Saville conclude that interest rates and the quantity of loans are primarily determined by norms of behavior (Dow and Saville 1988, p. 61). Of course, speculators influence the interest

rate, each trying to outguess average opinion of the rate which norms will establish. However, because expectations are loosely held and because the constraints on possible interest rates are rather loose, markets like to follow a leader. For this reason, the

actions of the central bank have a larger influence than one would guess by merely looking at the relative size of central bank 182

interventions. If the central bank raises the discount likely to affect expectations and cause other interest If markets come to expect tight money policy toward an expansion, interest rates might rise even before the announces

higher interest rates. Furthermore,

rate, this is rates to rise. the peak of central bank

since expectations

are loosely held, changes in a few key interest rates (such as the CD rate or loan rate set at one of the nine money center banks) will likely have a great influence over expectations. Therefore, even. though outstanding stocks of assets and liabilities are large relative to additions to bank balance sheets during any given period, there is no reason to suppose that actions of banks have little effect on interest rates. For this reason, the liquidity preference of banks will play a role in the determination of expectations held by speculators. Recent changes in central bank behavior and recent innovations have combined to reduce the importance of bank rationing behavior, however.

In the US, the central bank has deregulated

financial markets, reducing geographical, functional, and institutional barriers to entry. This has increased competition for the funds of surplus units and increased competition to fund the spending of deficit units. Financial markets have come to operate more like auction markets and less like markets based on rationing and established relations (De Cecco 1987, p. 6). Increased competition reduced loan rates of interest, expanded secondary markets for long term bonds (and reduced their interest rates), and increased credit availability for those who previously had been

credit constrained.” Wojnilower argues that the trend toward securitization and auction markets would have been quickly reversed in the absence of central bank guarantees (Wojnilower 1987). That is, as soon as one or two spectacular failures led to default on securitized loans, surplus units would have quickly returned to the safe, short term liabilities of commercial banks, and banks would have returned to established relations with customers. However, because the central

bank steps in to prevent default, the new institutional practices are validated so that the new instruments retain their liquidity. The public quite rightly views uninsured certificates of deposit or uninsured money market funds as nearly perfect substitutes for insured demand deposits. As a result, banks must compete on the basis of price, paying high deposit rates of interest and purchasing risky assets in an attempt to earn profits. De Cecco argues that this trend necessarily increases instability and must be reversed. He 183

predicts a return to credit rationing, which was the main feature of the monetary reforms of the 1930s. Moore substitutes an exogenously determined interest rate for an exogenously determined money supply in his endogenous theory of money. He argues that the central bank merely sets a price for reserves and then meets all demand, while banks set a price above this and then meet all demand. Curiously, recent innovations have reduced credit rationing because auction markets have expanded credit availability. On the other hand, innovations and deregulation have actually reduced the central bank’s ability to influence the interest rate, because central bank quantity constraints are less effective in an environment with auction markets and central bank willingness to guarantee virtually all financial instruments. It seems that Moore’s endogenous approach to money requires auction-like markets to ensure that there are no quantity constraints, but requires imposed rationing and barriers to entry to ensure that the central bank can set interest rates. The central bank loses control over interest rates if it guarantees the liquidity of a wide range of financial assets, but such guarantees are necessary for the maintenance of auction-like markets. While the central bank cannot determine ‘the’ interest rate, it can

influence

it. A

lower

discount

competition for other sources

rate

will

tend

to

decrease

of funds, so will place downward

pressure on a variety of rates of interest on sources of bank funds. This may, in turn, lower loan rates of interest if a given differential (loan rate less deposit rate) is maintained. (The degree of competition is one of the factors which determines whether the differential can increase.) If the central bank raises the discount rate, this will tend to increase competition for other sources of funds, so will tend to raise deposit and loan rates of interest. However, the central bank is not free to raise the discount rate to

any desired level, since the deposit rate of interest cannot rise too high relative to the interest rates banks receive on assets, without

threatening the financial system. One of the functions of the central bank must be to ensure that financial institutions can obtain funds at reasonable rates. That is, the interest rate differential must be

sufficient to allow banks to earn the normal rate of profit on bank equity. Restrictive central bank policy can gradually push up both the deposit and loan rates of interest as long as the cost of acquiring funds is not increased too rapidly with respect to the interest rate on bank assets.” The markup approach implies that Keynes was mistaken in his 184

belief that profit seeking behavior in a monetary economy subject to uncertainty necessarily leads to instability. If the interest rate is actually set by the central bank, then there is nothing to prevent the central bank from establishing an interest rate at which full employment will be achieved. If liquidity preference plays no role in the determination of interest rates, then the central bank could

achieve full employment by announcing that the interest rate would be held at some low and constant rate forever. The marginal propensity to purchase long term securities out of personal saving could then be nearly one, for each saver could buy a bond with a maturity equal to the period of the planned saving without worrying about market risk. This is similar to textbook Keynesianism, in which the financial market is removed from the

model. However, this appears to be inconsistent with Keynesianism. The markup approach does not provide a model showing how dynamic forces generate unemployment and financial instability. The essential insights of Keynes (and of Minsky - the financial instability hypothesis) are left out of the markup approach. Keynes had argued that a monetary economy inherently generates an interest rate which is chronically too high to establish a level of effective demand consistent with full employment. In the markup approach, the authorities need only lower interest rates to eliminate unemployment, which is not caused by any intrinsic

features of capitalism, but is caused by stupid policy.” In contrast, Keynes held out little hope that easy money policy would generate low enough interest rates to maintain full employment. The general case, according to Keynes, was a situation in which expected effective demand and expected profits would lead to equalization of the marginal efficiency of investment with the interest rate at a level of investment too low to achieve full employment. Moore reduces this to a special case where the central bank has set the interest rate too high. According to Keynes, uncertainty establishes the interest rate on money because the return to money is determined solely by its liquidity. In contrast, other assets obtain a return due to expected returns generated from income flows and/or capital gains, less carrying costs (such as wastage). As the quantity of any particular asset increases, the expected returns fall so that its marginal efficiency declines. However, Keynes argues that ‘beyond a certain point money’s yield from liquidity does not fall in response to an increase in its quantity to anything approaching the extent to which the yield

185

from other types of assets falls when their quantity is comparably increased’ (Keynes 1964, p. 233). The reasons for this include the high liquidity premium and low carrying cost of money, as well as zero elasticity of production, employment, and substitution of money.”’ In contrast, Moore argues that these characteristics apply only to commodity money, so that the only barrier to reduction of the interest rate is the behavior of the central bank. While it is true that credit money can be created essentially on demand, Keynes is quite correct in emphasizing that liquidity preference cannot be met by an expansion of credit. Moore seems to confuse liquidity preference with money demand. Keynes’s characteristics of ‘money’ apply to liquidity, but not to credit. A rise in liquidity preference is not the same as an increase in money demand, and

a rise in liquidity preference cannot necessarily be met by an increase in quantity - even if the central bank accommodates through easy money policy. As Lavoie and Seccareccia (1988) point out, Keynes had argued that the central bank may not be able to push down interest rates when liquidity preference has risen because speculators are likely to believe that interest rates will eventually rise again. Moore (1988c) points out that easy money policy during the Great Depression pushed the treasury bill rate in the US down to 0.02 per cent, but the long term bond rate remained above 2.5 per cent. He argues that this is a measure of the markup over the discount rate. The discount rate was 5.19 per cent in 1929, fell to 2.6 per cent in 1933, and reached a minimum of 1 per cent by 1939. In contrast, the Baa rate on bonds was 5.60 per cent in 1929, rose to

7.76 per cent in 1933, and fell to 4.96 per cent by 1939. These data seem inconsistent with a markup approach, since the markup increased from 0.41 percentage points in 1929 to 5.16 in 1933. However, this appears to be perfectly consistent with a liquidity preference approach, since a collapse of profit expectations during the depression should have been accompanied by rising liquidity preference. It also confirms Keynes’s belief that ‘easy money’ policy would not be sufficient to push down interest rates sufficiently to achieve full employment. While the central bank can push the discount rate nearly to zero, and while this may reduce the interest rate on some assets (such as government bonds or even Fed funds) close to zero, this does not mean the short term loan rate of interest or the bond rate of interest will decline commensurately. Banks will price and ration credit to customers on the basis of established relation, rules of

186

thumb, and liquidity preference. Liquidity preference will still determine the premium required to induce surplus units to hold bonds. If liquidity preference is inversely related to profit expectations, then low profit expecations cannot be overcome by easy money policy - no matter how easy it is. High liquidity preference, which is expected to reign when aggregate demand is low, will prevent easy money policy from pushing the marginal efficiency of money below the marginal efficiency of capital. Thus, banking behavior, as well as the behavior of the public, will prevent enlightened policy from succeeding.

Conclusions

The central bank sets the discount rate and determines the level of open market purchases to achieve a set of policy objectives. Such objectives might include interest rate targets, bank leverage ratios (reserves to liabilities), rate of growth of bank balance sheets, degree of competition from nonbanks, unemployment rates, exchange rates, and rate of growth of investment. The transmission of central bank actions to the economy is exceedingly complex and subject to great uncertainty because profit seeking bankers and other financial institutions actively subvert policy when the policy seeks to constrain profit maximization. Thus, the discount rate cannot be exogenously set, and cannot exogenously determine short term interest rates. Indeed, it makes no more sense to argue that interest rates are exogenous than it does to argue that the money supply is exogenous. Central bank policy, however, does influence bank behavior and

does go into the determination of costs faced by banks. However, such policy can be swamped by liquidity preference, innovations, competition, and profit expectations. The liquidity preference of banks determines the loan rate of interest in conjunction with the liquidity preference of the nonbank public, which together determine the bond rate of interest. As bonds (and other forms of nonbank finance) and loans are partial substitutes, competition from nonbanks helps to determine the loan rate of interest. The loan rate of interest then goes into the determination of the deposit rate of interest, since it is the differential between these two which largely determines the profitability of banking. Again, nonbank sources of liquid assets compete with banks and help to determine the deposit rate of interest, as does the liquidity

187

preference of the public.

The theory of interest rates which comes out of an endogenous money approach is neither a markup theory nor a theory of the demand and supply of hoards. The interest rate is largely determined by the willingness of banks and other financial institutions to allow their balance sheets to expand. This willingness is determined by rules of thumb, custom, and expectations. The central bank discount policy is also important, but this policy is constrained by the ability of banks to adjust to higher interest rates in the case of tight money policy, and by bank willingness to lower interest rates in the case of loose money policy. Speculators attempt to obtain arbitrage profits among assets of differing maturities. This helps to maintain a ‘term structure’ of interest rates. However, this term structure is subject to variation because there is no objective basis determining what is the proper structure. Thus, each speculator forms expectations in an attempt to outguess average opinion. As Keynes argues, the interest rate is influenced by what people expect it to be. Since there is little objective basis determining interest rates, they are established by rules of thumb

and by norms

of behavior. Thus, an increase in

money demanded to finance rising aggregate demand will not raise interest rates unless people believe it should. That is, if banks believe larger balance sheets would exceed proper leverage ratios, then interest rates must rise to compensate for greater perceived risk. If the public believes that the addition of bonds to portfolios would

increase

risk,

then

bonds

can

be

floated

to

finance

investment only at rising interest rates. In other words, both the money supply and interest rates are endogenously determined.

Notes 1. See Rogers (1989). 2. If the commodity is producible, then its quantity will rise whenever its marginal efficiency exceeds that of other commodities. However, if the commodity is nonproducible, then its price will rise until its marginal efficiency has fallen into line with that of other commodities. Most goods will fall in between: a rise in marginal efficiency will lead to an increase in price and quantity. 3. Moore (1988c) is correct when he argues that a decrease in liquidity preference which increases income is likely to raise money demanded for transactions, so there may be some pressure on interest rates.

188

Keynes dealt with this in Keynes (1973, p. 80). This effect on interest rates will be discussed in more detail below. Dow and Saville (1988) present an alternative argument: as Investment

proceeds,

it raises income

which

will be received

as

deposits. As surplus units attempt to diversify, the interest rate falls as they substitute out of deposits and into long term bonds. However, Dow and Saville have ignored the liquidity premium which would be required unless m equals one. It must be noted, however, that stocks are very much larger than accumulations which occur due to saving and investment flows. This will be discussed below. See Chick (1983, Chapter 9). Money demand is linked primarily to spending flows. However, it is possible to borrow in order to make portfolio adjustments. In this case, money demand might affect interest rates as loans are used to alter balance sheets. Indeed, Dow and Saville (1988, p. 209) argue that much of recent lending has been designed to allow rearrangement of portfolios. According to the theory of the ‘buffer stock’, an increase in bank loans to finance spending leads to excess liquidity, as those experiencing greater income than expected accumulate deposits. Thus, more money is temporarily held than normal against income. Gradually, however, the buffer stock is run down as agents attempt to adjust portfolios by purchasing goods, services and assets. Monetarists emphasize effects on prices of goods and services, while textbook Keynesians emphasize effects on asset prices (or interest rates). See Goodhart (1989a) and Dow and Saville (1988). Goodhart emphasizes that an increase in the demand for credit normally increases the supply, but this doesn’t mean that the demand for money hoards has risen. Thus, the stock of money may be excessive and will lead to purchases of other assets until asset prices change sufficiently that the public becomes satisfied with the existing quantity of hoards. During a run to liquidity, changes in the interest rate will not equilibrate liquidity preference to the quantity of liquidity. Indeed, the types of assets which satisfy liquidity preference during an expansion may not be deemed liquid during a downturn. Thus, not only is liquidity preference determined by expectations, but the liquidity of assets is also subjectively determined, at least in part. Keynes noted that in the

General Theory, ‘whilst it is found that

money enters into the economic scheme in an essential and peculiar manner, technical monetary detail falls into the background’ (Keynes 1964, p. vii). 10. See Davidson (1978, Chapters 12 and 13) for an alternative account. Davidson has underwriters arrange for the sale of long term bonds issued by the firms which purchase the investment goods. If the public won’t buy them, the underwriters are forced to take them up.

189

I have preferred to have the investment banks issue the long term bonds directly as they make long term loans to investment goods purchasers. This was done to simplify the analysis and to emphasize that financial institutions do take long term positions and issue long term liabilities. 11. These arguments apply to the commercial banking system as a whole. An individual bank which increases short term loans to provide for working capital expenses can expect to lose deposits through the clearing mechanism. However, this bank will also gain deposits lost by competitors. If a bank grows at the same pace as its competitors, it will not lose deposits through the clearing mechanism. A bank which grows substantially faster than others will be forced to recapture lost deposits through purchases in wholesale markets (such

as the Fed funds market). This ‘reflux’ need not be instantaneous, for

banks generally have several weeks to meet reserve requirements, as will be discussed in the next chapter. A rapidly growing bank is likely to face higher costs and will have to accept lower gross profit margins (loan rate of interest less wholesale cost of funds). 12; Income received as deposits by surplus units need not represent ‘desired’ saving. This does not mean that anyone is ‘forced’ to save, but merely that some agents have received income which they have not yet spent. I have ignored the multiplier process, in which increased investment is likely to induce increased production in the consumption goods industry. This will increase the wage bill, which will necessitate further advances of short term credit. None of this would change the analysis significantly. See Wray (1988a). 13. As these short term loans are retired, it is most likely that banks will renew the revolving fund of working capital through additional advances. However,

14.

15:

16. Ls 18.

if expectations collapse, banks will allow their

balance sheets to contract. Indeed, Keynes argued in his response to a letter from Hicks that increased investment need not increase interest rates. See Keynes (1973, p. 80). Dow and Dow (1989) define liquidity preference as an ex ante desire to hold a portion of wealth ‘idle’ in liquid form. A rise in the desire to hold assets idle would increase interest rates until wealth owners give up the attempt. They then discuss six ways in which an exogenous increase in the demand for idle balances will raise interest rates, even if the supply of money is endogenously determined. For example in Tsiang’s (1956) ‘proof’ (discussed below) that liquidity preference and loanable funds theories give identical results, the adjustment is instantaneous. Tsiang’s presentation is remarkable for its thorough misunderstanding of Keynes’s argument. For a summary of the literature on the equivalence of the two approaches, see Laurence Harris (1981).

190

19. In fairness to Rousseas (1986), he does not accept Moore’s ‘horizontalist’ approach. Rousseas argues that the central bank does not fully accommodate, and that the interest rate is not completely exogenously-set by the central bank. Tight money policy will raise interest rates, which induces innovations to shift the velocity-interest

20.

21.

POX

23:

rate function. This is similar to Minsky (1957). However, Rousseas does seem to argue that liquidity preference theory should be dropped in favor of a markup approach to interest rates. Why might banks do this? They may have decided they are over-leveraged and want to decrease their holding of long term assets and to substitute short term assets. That is, the liquidity preference of banks has increased. Wojnilower (1987) reports that the average holding period of Treasury bonds of ten years or greater maturity has fallen to 20 days. In 1984, one bank (First Boston Corporation) traded $4.1 trillion of these securities (an amount exceeding the total GNP of the US). In such an environment, what does it mean to claim that arbitrage equalizes the expected holding period yield of short term bills and long term securities? If the holding periods of long term and short term financial assets are approximately the same, does it make any sense to argue that the short term interest rate determines the long term rate? Neither expectations theory nor the markup approach is well supported empirically. Mankiw and Summers (1984) have shown that the evidence is not consistent with the expectations theory of interest rates. Furthermore, the evidence presented by Rousseas (1985b, 1986) does not support the hypothesis that the markup of the prime rate over the Fed funds rate is stable. Laudadio (1987) shows that the differential between the interest rate charged on large loans and that charged on small loans moves countercyclically in ways which would suggest that a greater markup is added to large loans during expansions, but that a smaller markup is added during recessions, while the markup on small loans moves in the reverse direction. The explanation for this is not immediately obvious. Even though it appears that various interest rates closely track the discount rate, this does not show that banks merely use a markup over the discount rate. Because the Fed is not free to set the discount rate wherever it likes, the discount rate can be expected to remain in line with other interest rates. The Fed sets the discount rate to achieve a broad range of objectives, including the portion of reserves it prefers to provide at the discount window relative to the portion it will provide in open market operations. If the Fed uses a higher discount rate, it will then provide fewer borrowed reserves but likely engage in more open market purchases. However, unless thereed wants to discourage all discount window borrowings, it cannot keep the discount rate above market rates.

191

24. It surely cannot be a coincidence that the ‘markup’ (prime rate over

the Fed funds rate) reported by Rousseas (1986, p. 56) dramatically stabilized after the late 1970s. Increased competition in auction-like markets decreased discretionary price-setting by money center banks. oo: Periodically, central banks do attempt to push up interest rates quickly through tight monetary policy as did the Fed in its 1979 experiment in Monetarism, during which the discount rate rose from 7.46 per cent (1978) to 13.42 per cent (1981). The experiment disrupted financial markets and forced the Fed to abandon tight monetary policy. The experiment shows that banks cannot instantly adjust to a higher regime of interest rates by simply adding a markup to the discount rate. It also showed that the Fed cannot quickly and permanently raise the discount rate. However, it does show that rapidly rising interest rates lead to disintermediation and innovation. Eventually, financial institutions can adjust to higher interest rates. The discount rate averaged just over 1 per cent during the 1940s, while it has averaged more than 7 per cent between 1983 and 1987, even though inflation averaged nearly 5.7 per cent per year during the 1940s and only 3.34 per cent during 1983-87. (All data were taken from The Economic

Report of the President, 1988. Inflation is the

rate of increase of the consumer price index.) 20; This, of course, is also the position of Monetarists: stupid bank policy generates inflations and deflations. If only the bank would stick to a rule, the economy would naturally stable prices and full employment. According to the

central central achieve markup

approach, if only the central bank would lower interest rates, full employment would be achieved. 20: Keynes also argues that stickiness of wages in terms of money also enhances money’s liquidity premium. In the concluding chapter of the General Theory, Keynes argues that the authorities should push the

interest rate as low as possible, but believes it is ‘unlikely that the influence of banking policy on the rate of interest will be sufficient by al to determine an optimum rate of investment’ (Keynes 1964, pes78) 28. All data are from The Economic Report of the President, 1989.

192

7

INSTITUTIONAL ANALYSIS

Introduction Above I have argued that banks actively engage in asset and liability management, and that they meet the demand for loans, then seek the funds required to replace reserves as they are withdrawn when the loans are utilized. I have asserted that banks are not reserve constrained because they can obtain reserves from a variety of sources. I will now briefly analyze how the banking system frees itself from reserve constraints. One possible source of error in orthodox monetary analysis arises from the lack of institutional detail in its analysis. If ‘money’ is narrowly defined so that it excludes most liquid assets, then the analysis will exclude much of the process through which positions in assets are taken. If bank analysis of assets and liabilities is arbitrarily restricted to a few over which the central bank apparently has substantial control, then incorrect conclusions will be reached regarding the degree of control the central bank actually has over the wide range of assets and liabilities (overt and covert) commonly held by banks. Finally, if the analysis of the relation between money and spending is restricted to a few liquid assets, then it probably will give misleading results. For these reasons, I will examine banking institutions in detail to show that conventional analysis is wrong. In particular, the view that banks passively accept deposits and then loan an amount equivalent to excess reserves ignores how banks actively manage assets and liabilities and innovate to increase profit potential. Banks take and finance positions in a wide variety of on- and off-balance sheet items. Innovations are a key ingredient, as are asset and liability management, which allow banks to escape quantity constraints. As Minsky (1986b) has characterized banking, it is a game in which the central bank is at a decided disadvantage. Banks innovate and find ways around constraints, placed by central bankers, to increase profits. Finally, an analysis that incorporates institutional detail

193

demonstrates that the relation between ‘money’ and spending is a very fluid one. The Monetarist story, in which money dropped by helicopters induces spending, cannot apply to an economy with highly developed financial relations. Monetarist policy, which tries to control spending through control of some monetary aggregate, is based on a flawed view of the way in which spending in a modern economy is financed. Firms make commitments to a payment stream before they have obtained means of payment. A well functioning capitalist economy requires that means of payment can be acquired as necessary at reasonable terms to enable commitments, particularly when units to meet economic expectations are disappointed. When expectations are high, firms willingly enter into commitments which set up a stream of future payments. The firm’s IOU, backed by a credit line at a bank, is sufficient to enable it to

receive goods and services today on the basis of a promise to deliver means of payment (cash or demand deposits) on a future date. Even with no expansion of the quantity of means of payment today, the quantity of ‘spending’ (goods and services delivered today in exchange for contractual commitments) increases. If the central bank attempts to control spending by constraining the growth of highly liquid assets, then this actually results in an increased volume of contractual commitments relative to the volume of highly liquid assets. That is, liquidity is stretched. Continued central bank restraint may eventually push up the terms on which liquidity can be purchased today (in exchange for payment of liquidity in the future) so high that default on commitments becomes necessary. In the absence of central bank intervention to provide reserves, this can set off a debt deflation

crisis. In the orthodox story, banks are intermediaries which discover

they need hold only a portion of each deposit on reserve, and any excess above this can be loaned. In the endogenous approach, however, banks, like other firms, take positions in assets by issuing liabilities. Unlike the liabilities of most firms, however, the

liabilities of banks function as a medium of exchange and means of payment. Thus, when a bank issues a demand deposit as it takes a position in an IOU, the demand deposit will be immediately used as a medium of exchange and cause a loss of reserves through the clearing drain. This loss can be covered by operating either on the asset side or on the liability side. The assets held by a bank generate

a flow of income, much

194

of it in the form of reserves,

which can be used to meet the clearing drain. Alternatively, the bank might liquidate assets to obtain reserves. Finally, the bank can operate on the liability side by issuing new liabilities to obtain reserves (that is, cause reflux). Of course, all of these options are also available to a nonbank

firm which is required to meet payment commitments on its liabilities. The major difference between a nonbank firm and a bank concerns the percentage of liabilities with an uncertain period to maturity. The nonbank firm typically knows the date on which most cash payments are due - so that most of the uncertainty involves the cash income to be generated on the asset side. However, in the case of the bank, a large portion of its liabilities

can be called in by creditors at short notice. Thus, not only does the bank face uncertainty regarding its cash flow from assets, but also regarding its cash payments due creditors. For this reason, banks typically maintain primary reserves of cash and central bank liabilities, as well as secondary reserves (government bonds and other highly liquid assets), which can be used to meet a clearing drain or liquidation of the bank’s liabilities by creditors. As will be discussed, banks have also developed techniques for ensuring reflux by issuing liabilities. In this way, banks can avoid forced sale of assets at deflated prices. Thus, the difference between banks and other types of firms is in degree rather than in kind: both take positions in assets to generate cash flows by issuing liabilities which entail cash commitments, and both have developed techniques to meet cash commitments when these exceed cash flows. One of these methods is to hold liquid assets, including cash and other liabilities, since a liquid position can provide protection in an uncertain world. Thus, this chapter examines both sides of a bank’s balance sheet

and discusses off-balance sheet items to show how banks take positions in assets and how they help nonbanks to do the same. Both bank sources of funds and bank uses of funds are examined in turn.

Reserve requirements Before looking at bank sources of funds, it may be instructive to examine the process by which reserve requirements are calculated. Then I will show how banks meet these reserve requirements. Since the Garn-St. Germain Depository Institutions Act of 1982,

195

all depository institutions (member and non-member) are subject to reserve requirements. At the beginning of 1989, these requirements were: A. On transactions accounts: i) 3 percent of the first $41.5 million on deposit, ii) 12 percent on total above $41.5 million. B. On nonpersonal time deposits: i) reserves equal to 3 percent 11) of deposits with an original maturity of less than 18 months, maturity. month 18 than greater of deposits on t requiremen no C. On personal time and saving deposits: no requirement. D. On loans from foreign banks or branches abroad: 3 percent of the amount borrowed. E. The first $3.4 million of reservable liabilities is exempt from requirements. (Source: Annual Report of the Board of Governors of the Federal Reserve System, 1988, p. 225)

Member banks may hold required reserves as vault cash or as deposits at the Fed, while non-members are free to hold reserves outside the Fed in institutions which ‘pass through’ these reserves to the Fed. Required reserves are calculated based only on net deposits, thus cash items in the process of collection and demand balances due from domestic banks are excluded from the calculations. Banks are now subject to ‘contemporaneous reserve accounting’, which means that reserves must be based on deposits more or less concurrently held. Reserves are not actually calculated on a contemporaneous basis, however. Required reserves are calculated on a daily-average basis over two two-week periods: the first two-week period applies to time deposits and the second to demand deposits, beginning two weeks after the close of the computation period for time deposits. Thus, the base period (on which reserve requirements are calculated) runs about six weeks, and the settlement period (over which average reserves are calculated) also lasts about six weeks, with over-lapping of the two periods consisting of all but the first two days and the final two days. The vault cash that is counted as part of the required reserves is based on the average holding during the computation base period for time deposits, while the deposits at the Fed which are counted are those which are held about two weeks later. While the calculation process is quite complicated, the result of the process is that the bank comes to the end of the base period with a calculation of its average time deposits (held a month ago) and

196

of its average demand deposits (held over the previous two weeks). It can then figure the reserves required against these deposits. It also calculates the average vault cash holdings (again, held nearly a month previously), and reserves held at the Fed (over the previous 12 days). It then calculates how much must be held on reserve at the Fed over the next two days in order to meet reserve requirements based on deposits already held over the base period, which was the preceding six weeks. If the bank finds that it is short of required reserves, it must hold sufficient reserves on these final two days to raise the average to the required level. If it finds itself with excess reserves, it is free to use them in the most profitable manner. The Fed allows a 2 per cent margin of error: if the bank is deficient, it can carry the 2 per cent deficiency over to the next period; if it has excess reserves, up

to 2 per cent of excess reserves can be applied to the next period. Contemporaneous reserve accounting replaced lagged reserve accounting in 1984. Under lagged reserve accounting, the bank knew what level of reserves would be required for two weeks prior to the deadline for achieving the required level. This time gap has now been reduced to two days, which presumably makes it more difficult for the bank to hit the reserve target. (In fact, after 1984, excess reserves increased substantially, perhaps due to increased difficulty in forecasting required reserves.) The Fed felt that this would give it more control over the money supply. However, this is a matter of controversy. As Moore (1984) has argued, contemporaneous reserve accounting will not free the Fed from acting as a lender of last resort. In fact, contemporaneous reserve accounting may actually increase the endogeneity of reserves: on the calculation day, the deposit holdings are history and banks must come up with reserves to hold for the final two days to meet requirements. The endogeneity of reserves and the money supply will be discussed in more detail in Chapter 8. Orthodox economists typically argue that only excess reserves can be used to meet an unexpected liquidation of bank liabilities or another reserve drain. This is false. Primary reserves (central bank liabilities and cash) allow a bank immediately to meet liquidation demands by depositors and other creditors, while secondary reserves (such as government bonds) allow a bank to obtain reserves quickly through asset sales or by borrowing. Thus, the existence of primary and secondary reserves buys time for a bank so that it is not forced to sell assets quickly whenever creditors demand liquidity. Secondary reserves and other liquid

197

assets can be pledged to allow a bank to borrow reserves from other banks, from nonbanks,

and at the discount window.

Such

short term borrowing buys even more time for a bank - which can then attempt to ‘operate on liabilities’ (for example, sell certificates of deposit), rather than sell assets at a loss, to retire short term

liabilities. Thus, primary reserves operate as a ‘first defense’, secondary reserves and assets which are eligible collateral operate as a ‘second defense’, and ‘operating on liabilities’ acts as a ‘third defense’ against forced sale of assets. Whether primary reserves are required or excess is immaterial. Once the bank has successfully defended its asset position, it must then operate in the financial system to ensure that it meets legal reserve requirements. Normally, the bank will have several weeks in which to come up with required reserves due to the complicated method used to calculate reserve requirements. In any case, however, a bank which is faced with attempts at liquidation by its customers can turn to the Fed as a lender of last resort.

Fed funds market The development and growth of the Fed funds market more than thirty years ago provided an efficient mechanism whereby banks with excess reserves could supply them to deficient banks and thereby turn idle funds into an interest earning asset. Minsky (1957) showed how the Fed funds market enables a given quantity of reserves to support more deposits in the aggregate by reducing aggregate excess reserves. In particular, the Fed funds market enables banks with retail lending opportunities to make loans and then to purchase funds from banks which have exhausted their retail lending opportunities. For example, a small town bank may often find itself with excess reserves, but without a profitable loan

outlet at the retail level. It may then provide reserves in the Fed funds market to one of the money center banks which does have retail lending opportunities. Table 7.1 in the appendix to this chapter shows that aggregate excess reserves declined absolutely and relatively during the 1960s, from about 0.35 per cent of M1 in 1959 to a low of 0.06 per cent of M1 in 1977. Column 9 of Table 7.7 shows that excess reserves as a percentage of required reserves also reached a minimum in 1977. Aggregate excess reserves remained very low until 1984,

198

when they began to rise absolutely and relatively to other monetary aggregates (probably due to the switch to contemporaneous reserve accounting).

Fed funds sold is shown in Table 7.10, Column

4,

while Fed funds bought is shown in Table 7.11, Column 6. Volumes of Fed funds sold and bought increased rapidly during the 1970s, as is shown in Table 7.9, Rows 12 and 13. The growth of Fed funds exceeded that of most monetary aggregates. Table 7.12 shows that Fed funds sold grew as a percentage of total bank assets to nearly 5 per cent by the mid 1980s, while Fed funds bought grew to more

than 9 per cent of total liabilities (see Table 7.13). As a source of reserves, the Fed funds market is very important. In 1986, total Fed funds purchased reached $248 billion, while the total net increase in the Fed’s open market account (‘open market purchases’) was only $30 billion, and reserves borrowed at the discount window

only reached $827 million. (See Table 7.7.) As Table 7.5 shows, small banks typically are suppliers, while larger banks are demanders, in the Fed funds market. For example,

in 1980, Fed funds sold accounted for 10.5 per cent of small bank assets (those with assets of less than $5 million), and for only 2.7 per cent of large bank assets (those with assets greater than $5 billion). On the other hand, Fed funds purchased as a per cent of liabilities amounted to 0.2 per cent for the small banks and to 13.4 percent for the largest banks. Of course, the Fed funds market is not simply a source of reserves for reserve deficient banks, as it is also a source of generalized funds (or loanable funds, using the orthodox

terminology). This function is often missed by textbooks, which repeat the old adage that ‘individual banks can only loan their excess reserves, while money is created by the system as a whole’. However,

the existence

of the Fed

funds

market

ensures

that

individual banks can create money. As textbooks present it, a bank which loans more than its excess reserves will be in trouble as soon as the loans are spent since the withdrawal of the associated demand deposits will leave the bank with negative excess reserves. Thus, individual banks can never lend more than their excess reserves. However, if a bank can quickly recapture reserves lost through interbank settlements by turning to the Fed funds market,

then it can effectively create money and act as if the ‘deposit multiplier’ operates at the individual level. Since the withdrawn

demand deposits will (for the most part) end up as deposits elsewhere, the Fed funds market enables the originating bank to reflux need not be these deposits. Although recapture 199

instantaneous since banks typically have several with required reserves, the Fed funds market instantaneous reflux as an overnight source of The Fed funds market alone cannot free the a whole

from reserve

constraints, even

weeks to come up does allow almost funds. banking system as

though created

demand

deposits can always be recaptured through the Fed funds market. The reason is that the creation of new demand deposits also raises reserve requirements. The Fed funds market is not a net source of reserves for the system as a whole. (However, it is a net source of funds for the commercial banks: in 1986, Fed funds purchased

exceeded those sold by insured commercial banks by $109 billion.)

The Fed funds market, however, increases reflux at an individual

level since banks which create deposits can recapture them, and it enables a given quantity of reserves to support a greater quantity of deposits by reducing aggregate excess reserves.

Repurchase agreements In the mid-1950s, repurchase agreements were used by government bond dealers as a way to borrow funds. The bond house would sell government bonds to non-financial corporations, which would use idle demand deposits to purchase the bonds. The corporations would then receive interest, while the bond dealers were able to

borrow at rates below the bank loan rate of interest. To reduce risk to the corporation, the bond sale would be accompanied by a repurchase agreement, so that the corporation would be able to restore its liquidity position quickly.’ Repurchase agreements have rapidly become an important source of funds

for banks,

as Table

7.2, Column

4 shows.

Overnight

repurchase agreements plus overnight Eurodollars now equal about one-quarter of the volume of demand deposits. Banks sell securities (mainly government bonds) and agree to repurchase them in a short time (often the next day) at a higher price. The bank uses repurchase agreements to borrow funds, with the securities used as collateral. These funds can be used to raise the average reserve holdings of the bank. When the bank sells securities overnight to one of its own depositors, its average deposit liabilities are reduced. Thus, not only do repurchase agreements provide a source of funds (at the individual bank level) to meet reserve requirements, but they also can reduce the deposit base on which required reserves are calculated.

200

For example, if half of the bank’s demand deposits are normally those of corporations, the bank could potentially reduce its demand deposit liabilities by 25 per cent if each firm could be induced to hold a repurchase agreement rather than a demand deposit for half of the base period. If a bank sells bonds overnight to noncustomers, the bank’s average reserve holdings would be credited. Clearly, the existence of repurchase agreements can significantly reduce reserve requirements and provide an important individual source of reserves. In the aggregate, repurchase agreements economize on reserves, but are not a net source of reserves unless

corporations give up cash. Thus, although aggregate reserves do not increase when corporations reduce demand deposits and hold government bonds through repurchase agreements, this does increase the ability of the banking system to issue loans by economizing on reserves.

Certificates of deposit The negotiable certificate of deposit was developed in 1961 to provide an interest earning asset which corporations could purchase to replace idle funds in demand deposits (which could not earn interest). These negotiable certificates of deposit are usually issued in denominations of $100,000 or more, with maturities to suit the

purchaser, and have well-developed secondary markets. Thus, certificates of deposit allow corporations to earn interest without sacrificing much liquidity (Ritter and Silber 1986). The certificates of deposit enabled banks to avoid disintermediation which would result because the interest rate on demand deposits was restricted to zero in an era in which the market rate of interest was rising. Certificates of deposit became even more popular after interest rate ceilings on large certificates of deposit were removed in 1973. As Table 7.3, Column 4 shows,

certificates of deposit have become an important source of funds for banks (included with other large time deposits.) Table 7.9 shows that large time deposits have experienced very high rates of growth since 1960 (they grew by nearly 1000 per cent between 1960 and 1965). Large size, negotiable certificates of deposit issued by insured commercial banks grew from zero in 1960 to $26 billion in 1970 and to $300 billion in 1983 (Ritter and Silber 1986, Chapter 8, Table 3). By comparison, demand deposit liabilities of insured commercial banks in 1983 equaled $389 billion (see Table 7.11).

201

Like repurchase agreements, certificates of deposit economize on reserves if customers can be induced to exchange demand deposits for them. When a bank sells a certificate of deposit to a non-customer,

it obtains reserves. The required reserve

ratio on

certificates of deposit is lower than that on demand deposits except for the smallest of banks (those with less than $41.5 million in demand deposits). Thus, a bank which issues a certificate of deposit with a denomination of $100,000 will have its required reserves increase by a much smaller amount compared to the issue of a demand deposit. Furthermore, a bank which induces one of its customers to exchange a demand deposit for a certificate of deposit will create excess reserves. This means that a bank with a reserve deficiency may be able to lower its reserve requirement by selling certificates of deposit to its own

customers,

and can increase its

reserves by selling certificates of deposit to non-customers. In the aggregate, certificates of deposit act as a net source of reserves to the extent that the public substitutes certificates of deposit for demand deposits, and act as a net source of reserves when sold for cash or to foreigners.

Borrowing from foreign sources As Table 7.8 shows, US banks borrow from foreign banks and

from foreign branches of domestic banks. Liabilities to foreigners grew rapidly after 1973, and particularly so after 1980. Time deposits owed to foreigners (including certificates of deposit) grew rapidly in the 1980s, as domestic banks tapped international sources of liquidity. By 1987, total deposits in foreign offices of US

banks reached $342 billion. The international interbank market functions in the international system much as the Fed funds market operates in the national system. Loans from foreign sources allow individual US banks to purchase reserves. Since reserves must be held against loans from foreign banks, a US bank with a $100 reserve deficiency will have to purchase more than $100 from foreign sources to make up the deficiency. In the aggregate, loans from foreign sources are a potential net source of reserves for the US banking system. Whether they are actually a net source depends, of course, on the size of leakages out of the US banking system and into foreign banks. As Table 7.8 shows, US short term liabilities to foreigners (including those to overseas branches of US banks) greatly exceed

202

US short term claims on foreigners. Thus, foreign sources are today a net source of reserves for the US banking system. The international interbank market also operates like the Fed funds market by allowing a country’s banking system to recapture leakages.

Thus, the international market makes ‘funds available

quickly and efficiently to banks which had lending opportunities...where funds are withdrawn from one part of the system

and

redeposited

elsewhere,

the

interbank

market

is a

natural channel whereby the funds can be transferred back to the original part of the system...’ (Bank for International Settlements

1983, pp. 11-12).*

Discount window and open market operations The Fed discount window is an important source of reserves. The Federal Reserve Act of 1913 made discounting the most important feature of the Fed. The Fed was to provide liquidity as needed to prevent the runs on banks that were a common occurrence during the nineteenth century. By injecting liquidity, the Fed could prevent a financial panic from spreading. In 1980, the Banking Act provided that all depository institutions subject to the Fed’s reserve requirements would have access to the discount window (Ritter and Silber 1986, pp. 209-10).

Unlike some other central banks, the Fed uses quantity constraints in addition to rationing reserves through the discount rate. The quantity constraints are necessary because the Fed normally keeps the discount rate below the market price of other sources of bank funds. It is usually argued that the Fed will turn away frequent borrowers, so that banks are discouraged from borrowing except in emergency situations. A bank faced with a reserve deficit must weigh the lower price of reserves at the discount window against the possibility that the Fed will become upset at the prospect of making an additional loan to the bank. According to Meek: The Federal Reserve serves as the lender of last resort on those occasions when the bank finds itself suddenly short because of unexpected transactions, or the trader cannot find funds in sufficient volume late in the day before the Fed’s wires close. Or the bank’s officers may have concluded that they can risk borrowing at the discount window because of their limited recourse to it in the recent past. (Meek 1982, p. 53)

203

Many loans made However, this may overstate the case somewhat. made whena loans, rary tempo at the discount window are not example, in For cash. of short bank ‘suddenly’ finds itself made at the discount November of 1988, 81 per cent of the loans so that ‘seasonal and window were classified as ‘extended credit’,

cent of the total. adjustment credit’ accounted for less than 20 per 44 per cent of ged Between 1985 and 1987, extended credit avera ed when a grant is total discount window loans.’ Extended credit and not to es, reserv ning bank experiences chronic problems obtai meet a temporary deficiency. borrowed Table 7.7, Column 1, shows the quantity of reserves increased ed borrow y quantit annual e averag from the Fed. The to about 1960s the of half first the during million $251 from about 7.9, Row $1473 million during the first half of the 1980s (see Table ed borrow , general In 17). However, annual variation is high.

reserves are of the same order of magnitude as (or somewhat larger than) aggregate excess reserves. Tables 7.14 and 7.15 show open market transactions of the Fed, Total gross outright purchases of treasury bills doubled between 1970 and 1986. However, the Fed increasingly used matched transactions and repurchase agreements after the mid-1970s. For

example, matched transactions grew from a volume of about $12

billion (equivalent to outright gross purchases) in 1970 to nearly $1000 billion in 1986 (40 times the volume of outright purchases).” The Fed uses outright purchases when it projects that commercial bank needs for reserves will last for a period of several weeks. It uses repurchase agreements and matched transactions when it projects only a temporary shortage or surplus of reserves. These temporary transactions are believed to have less impact on securities dealers: [A repurchase agreement] injects reserves temporarily, without affecting the interest rate risks market participants have to bear, it provides financing of existing dealer positions for a few days, rather than requiring a change in positions. (Meek 1982, p. 120)

Table 7.15 gives the net change in government securities held by the Fed and the total net change in open market transactions. The total net change in open market transactions averaged about $6339 million per year during the first half of the 1970s, and increased to an annual average of $10,745 million during the first half of the 1980s (see Table 7.9, Row

18). Annual variation in open market

204

transactions is high, but open market transactions are typically a much more important source of reserves than are discount window loans (loans at the discount window average about 10 per cent of net open market purchases). Column 3 of Table 7.7 gives the total reserves lent at the discount window plus the total net change in open market transactions. This can be taken as a measure of the Fed’s supply of reserves to the banking system. (The net change in open market transactions actually measures the increase in reserves supplied through open market operations, while loans at the discount window measure the level of reserves supplied through loans.) Variation of this combined measure appears to be somewhat below that of either borrowed reserves or open market transactions taken separately. However, the Fed greatly increased its supply of reserves during 1985 and 1986. Column 5 of Table 7.7 gives the ratio of required reserves to the total reserves supplied by the Fed, expressed as a percentage. With the exception of 1972 and 1980, the Fed typically supplied an amount equal to between 30 and 50 per cent of total required reserves. Over the period 1970-86, the Fed supplied an amount equal on average to 32 per cent of required reserves, but this rose sharply to over SO per cent in 1985 and 1986. Column 6 shows the increase in required reserves over the quantity required in the previous year. This gives an indication of the additional reserves required which banks must obtain. Column 7 shows the ratio of the Fed’s supply of reserves to the increase in required reserves over the previous year. On average, the Fed supplies about five times the additional reserves needed by the system.

However,

1972, 1980, 1984, and 1986 were years in

which the Fed supplied substantially fewer reserves relative to the increase in required reserves. This was probably due to very tight monetary policy in 1972 and 1980, to very fast growth of required reserves in 1986, and to a combination of moderately tight monetary policy and fast growth of reserves in 1984. Obviously, required reserves cannot account

for the use of all

reserves supplied by the Fed. Other factors which absorb reserve funds include currency in circulation, treasury cash holdings, deposits at the Fed (other than required reserves), and other Fed liabilities and capital. In addition to loans at the discount window and Fed open market purchases, other factors supplying reserve funds include float, other federal reserve assets (bank premises and other assets), gold stock, special drawing rights certificate accounts,

and treasury currency outstanding. Of the factors absorbing reserve

205

funds, currency in circulation is by far the largest component, accounting for 81 per cent in 1988. In contrast, member bank reserves held at the Fed accounted for only 12 per cent. In general, Fed holdings of US treasury and federal agency securities are equal to currency in circulation - so Fed open market purchases supply reserve funds equal to reserve funds which are lost to currency in circulation. The only other large sources of reserve funds which are growing are treasury currency and other Fed reserve assets, which together are about equal in size to member bank reserves. Since currency in circulation is many times greater than required reserves, a small shift of public preferences away from currency could potentially lead to a large increase in bank reserves. Table 7.7 can be used to examine Fed policy during the 1980s. In 1980 the Fed pursued tight monetary policy to fight inflation. Open market purchases fell to a level of less than 60 per cent of those of the previous year and the Fed raised the discount rate by one and a half percentage points over the 1979 rate (and to nearly double the 1978 rate). Loans at the discount window actually rose in spite of the rising discount rate, probably due to the fall in open market purchases. The Fed supplied only 20 per cent of the system’s required reserves, or about 3.5 times the increase in required reserves. In 1981 the Fed again raised the discount rate,

but more than doubled its open market purchases. Loans at the discount window fell, but the Fed supplied 31 per cent of required reserves and 6.7 times the increase in required reserves. The rate of growth of required reserves fell sharply in 1981, partly due to a slumping economy. In 1982 the Fed lowered the discount rate and continued to supply about 31 per cent of required reserves - almost wholly through open market purchases. In 1983 the Fed increased its open market purchases by 10 per cent and again lowered the discount

rate.

However,

loans

at the

discount

window

didn’t

increase until 1984, when the Fed reduced its open market purchases by 20 per cent. Finally, 1985 and 1986 were years with high rates of growth of required reserves and unprecedented volumes of open market purchases. In summary,

the Fed does have some

discretion regarding the

percentage of required reserves it will supply and the form in which it will supply them. When required reserves are growing rapidly, the Fed can force banks to the discount window by reducing open market purchases. Fewer open market purchases and high discount rates will force banks to obtain reserves from

206

alternative and (presumably) costly sources (for example, by inducing the public to give up cash). However, the Fed’s influence can be curtailed by bank actions. For example, in 1975 the Fed increased its open market purchases and cut the discount rate by almost SO per cent. During 1975 the Fed supplied an amount of reserves nearly 70 times the increase in required reserves. However, the rate of growth of M1 remained at a sluggish 4.9 per cent (the average rate of growth of M1 over the decade of the 1970s was 8 per cent). Thus, extremely loose monetary policy did not encourage banks to expand their balance sheets. As another example, in 1972 the Fed reduced its open market purchases by about 97 per cent. Banks turned to the discount window and borrowed about 8 times as much as they had the previous year. However, the Fed still only supplied about 6 per cent of the system’s required reserves, or about 70 per cent of the increase in required reserves. Clearly, this was a year of tight monetary policy. But M1 grew at a rate of 9.2 per cent, the rate of growth of required reserves increased from a rate of 6.9 per cent in 1971 to 9.7 per cent in 1972, and excess reserves increased by 56 percent over the level of excess reserves which existed in 1971. Thus, extremely tight money policy did not constrain reserves or the growth of bank balance sheets. This indicates the Fed plays an active role, using a combination of price and quantity constraints, and contradicts the position of Moore, who argues that the Fed uses only price constraints and passively supplies reserves on demand. However, it also contradicts the position of Monetarists, who view the banks as the passive agents. In reality, the banks and the Fed are active players, although the banks may have more to lose. Thus, it is not surprising that they seem able to subvert Fed attempts at control.

Summary of sources of funds Table 7.11 summarizes the liabilities of the FDIC-insured commercial banking system, while Table 7.13 shows the sources of funds by percentage of the total. As can be readily observed, substantial changes in the system’s balance sheet have occurred over the past few decades. Demand deposits have declined in importance, while time deposits (such as certificates of deposit) have become much more important as a percentage of total deposits.

Furthermore,

other

borrowed

207

funds

(from

foreign

branches, affiliates, and so on) have also become important as a source of funds. For example, foreign office deposits (included in total deposits) totaled about $322 billion in 1985, or about 12 per cent of total liabilities. Fed funds have also increased greatly as a percentage of total liabilities, while bank equity has declined. Demand deposits declined in importance as banks economized on reserves by encouraging customers to switch to deposits with lower reserve ratios. Customers were willing to switch to other types of deposits as interest rates rose and as alternative checkable deposits were created. Innovations such as the Fed funds market and large certificates of deposits permitted a given quantity of domestic reserves to support a greater quantity of demand deposits by redistributing domestic reserves more efficiently and by enabling domestic banks to borrow from foreigners. Banks reduced equity as a percentage of liabilities in order to increase profitability: they issued more liabilities in order to purchase more earning assets. Finally, banks increasingly moved to off-balance sheet liabilities to increase fee income and profitability. This will be discussed below. These changes in the sources of bank funds are consistent with the endogenous money view, in which banks meet loan demand and then purchase funds by issuing liabilities. As Table 7.13 shows, only about 17 per cent of a bank’s liabilities now consist of demand deposits, while purchased funds (largely from wholesale markets) have replaced demand deposits as the primary source of funds. In the textbook view, however, banks passively accept deposits and then loan excess reserves. While the data presented in this chapter do not the textbook view, they certainly are consistent with the endogenous money view and cast doubt on orthodoxy. The evidence also shows how quickly innovations can change the entire financial structure from one which relied heavily on demand deposits to one which relies, for the most part, on purchased funds.

Analysis of uses of funds The following sections very briefly analyze the uses of bank funds. Table 7.10 presents a summary of bank assets, while Table 7.12 shows use of funds by percentage of total assets. I will show that banks have reduced cash (non-earning) assets and holdings of government bonds relative to other assets in order to increase profitability. As Minsky (1986b) has characterized it, banks have shifted from operating on assets to operating on 208

liabilities.

Before 1960, a typical bank held large amounts of cash

assets and government bonds, which were liquidated when the bank wanted to make loans. Government bonds were held as a safe, secondary reserve, while cash was held as a primary reserve. Gradually, however, banks began to issue new liabilities rather than liquidating secondary reserves to finance positions in assets. As a result, cash assets and government bonds declined in importance and the quantity of loans increased relative to such liquid assets. At the same time, innovations and improved liability management ensured that banks approached by credit-worthy customers would be able to purchase funds in wholesale markets or induce shifts in customer portfolios so that loans could be made. Thus, banks do not passively await deposits which create excess reserves so that they can make loans. Finally, banks moved to off-balance sheet operations in which they make markets for the debt of nonbanks. This enabled banks to increase fee income without reducing their equity-to-asset ratios, and was partially a response to ‘problem’ energy, real estate, and third world loans.

Reserves As Table 7.12 shows, cash assets (including reserves) have rapidly declined in importance as a use of bank funds. This is primarily due to a shift in the sources of bank funds: a shift from deposit sources which require reserves to non-deposit sources and deposit sources with lower required reserve ratios (liability management). Required reserves declined as a percentage of M1 from 9.2 per cent in 1960 to 7.2 per cent by 1980, and as a percentage of M2 from 4.2 per cent in 1960 to 1.9 per cent by 1980. Cash assets (including reserves) declined from 20 per cent of total assets in 1961 to 16 per cent by 1970 and 12 per cent by 1987. In addition, the development of the fed funds market reduced aggregate excess reserves: from 0.35 per cent of M1 in 1959 to 0.07 per cent of M1 by 1981. (However, excess reserves have begun to rise significantly since 1984, to 0.14 per cent of M1 in 1987. As noted above, this

may be due to the switch to contemporaneous reserve accounting.) As discussed above, each year the Fed generally supplies a significant portion of reserves required by the system (on average, about a third of required reserves) through net open market purchases and loans at the discount window, although this portion declines when the Fed institutes tight monetary policy or when

209

required reserves are growing rapidly. In any case, banks manage portfolios, innovate, and enter into contingent contracts to ensure that reserves can be generated to meet requirements. Portfolio management and innovations have been discussed above. Contingent contracts would include lines of credit arranged with correspondent banks or with overseas branches which ensure future availability of funds should liquidity become necessary. This will be discussed below in the section on off-balance sheet operations. Finally, if all else fails, the central bank must be the ultimate

provider of liquidity. As discussed above, stability of the financial system requires that the Fed will enter as a lender of last resort since it is impossible for the private banking system in_the aggregate to liquidate assets to obtain reserves (except to the extent that it sells assets to the public for cash or obtains reserves from foreigners).

Government. bonds The commercial banking system emerged from World War II with large holdings of government bonds. When a bank was approached by a customer with a request for a loan, it would typically sell a government bond and use the funds to make the loan (asset management). (See Minsky, 1986b.) Government bonds serve as ‘secondary reserves’ since they may be quickly sold with little transaction cost in well developed secondary markets. During the 1960s, banks replaced asset management with liability management so that government bonds declined in relative importance as a secondary reserve, since a bank needing reserves would simply purchase them. As Table 7.6 shows, government bonds have declined in importance. For example, government bonds as a percentage of total loans and securities held by banks has decreased from 31 per cent in 1960 to about 15 per cent today. This ratio gives an indication of the riskiness of bank portfolios. Government bonds fell from a total of 24 per cent of assets to 10.6

per cent between 1959 and 1980. Loans

Tables 7.10 and 7.12 show that loans are now the major assets of the commercial banking system. The ratio of loans to demand

210

deposits has increased from 77 per cent in 1961 to 347 per cent in 1987, and the ratio of loans to total domestic deposits has increased from 51 per cent in 1961 to 90 per cent by 1987. The composition of loans includes short term loans to firms, mortgages (residential and commercial), consumer debt, student loans, and even longer term loans to firms. Table 7.9 shows that the rate of growth of loans has exceeded the rate of growth of required reserves, securities, M1, and various other aggregates.

A recent innovation is the packaging of loans for sale in secondary markets. Actually, secondary markets for loans have existed for some time, but these markets have traditionally been thin. Increasingly, however, ‘bundles of untraded (and individually untradeable) claims such as mortgages, car loans, and student loans are converted into securities and traded’ (Cooper 1986, p. 3). Cooper reports that in the United Kingdom, sales _ of mortgage-backed securities increased from $11 billion in 1975 to $148 billion by 1985. Sales of securitized ‘Euro Commercial Paper’,

which did not exist in 1975, were valued at $35 billion by 1985. In the US, securitized mortgages now account for about one-half of all new mortgages, while the sale of securitized loans in Europe in 1985 was about twice as large as the total value of syndicated Eurobank loans (Mayer 1986). Securitization can increase the liquidity of bank assets, including loans, by creating a secondary market. Minsky (1987) argues that the trend to securitization reflects a much more fundamental change in the financial system.’ Securitization reflects a change in the weight of market and bank funding capabilities: market funding capabilities have increased relative to the funding abilities of banks and depository financial intermediaries. Securitization is in part a lagged response to monetarism. The fighting of inflation by constraining monetary growth opened opportunities for non-banking _ financing techniques. The monetarist way of fighting inflation, which preceeded the 1979 ‘practical monetarism’ of Volcker, puts banks at a competetive disadvantage in terms of the short term growth of their ability to fund assets. Furthermoré by opening interest rate wedges monetary constraint provides profit opportunities for innovative financing techniques. (Minsky 1987, p. 2)

Minsky notes that securitization in the US began in the thrift industry because Monetarist policies which raised interest rates made it difficult for the thrifts to obtain funds. Securitization zAL

allowed them io continue to initiate mortgages by creating a method of obtaining funding. Of course, securitization has now spread to other financial assets, with mutual and pension funds and foreign institutions constituting a ready pool of buyers. Finally, Minsky goes on to argue that securitization further demonstrates the proposition that spending cannot be constrained through quantity constraints placed on certain classes of financial institutions: ‘Securitization implies that there is no limit to bank initiative in creating credits for 1) there is no recourse to bank capital and 2) the credits do not absorb high powered money’ (Minsky 1987, p. 4). Securitization can be characterized as ‘the development of techniques to "enhance credits" without accepting contingent liabilities or the investment of pure equity funds’ (Minsky 1987, p. 4). Securitization, in other words, arises to finance activity without entailing need of the medium of exchange or of any other explicit bank liability. The phenomenal rise of high yield (‘junk’) bonds further demonstrates increasing competitiveness in financial markets. Only about 800 of the 23,000 largest corporations can issue investment-

grade bonds (Wise 1990, p. 7). Before the development of junk bonds, the other corporations were constrained in their ability to

obtain funding outside banks. In 1976, only $15 billion in junk bonds were issued, while this rapidly grew to $200 billion today, or about 20 per cent of the total corporate bond market (ibid., pp. 7-8). In spite of the popular opinion, most junk bonds are not issued to finance mergers or leveraged buy-outs. In fact, a survey of industrial borrowers found that junk bond issuers typically had higher rates of growth of capital spending than did non-issuers, and only three per cent of issuers used the proceeds for mergers or leveraged buy-outs (Jefferis 1990, p. 3). There has, thus, been a

substantial shift away from bank lending and toward bond issues. Between 1980 and 1982, 57 per cent of new corporate credit came from banks, while 36 per cent came from bonds. However, by 1986-88, bank loans accounted for 15 per cent of new credit, while

bonds accounted for 61 per cent (ibid., p. 3). This does not mean that banks have been totally excluded from provision of credit to corporations, however. As will be discussed in the next section, banks play a large role in arranging bond placements in return for fee income. Banks and thrifts are also outright purchasers of junk bonds. Federally insured thrifts hold about 8 per cent of outstanding junk bonds, with a few thrifts holding as much as a third of their assets in this form. Indeed, 212

perhaps buy-out that the 30 per

57 per cent of the financing of the typical leveraged comes from commercial bank loans (Greider 1989). Given cumulative default rate on junk bonds seems to be about cent, bank and thrift involvement in these may prove

troublesome (Jefferis 1990, p. 5). As will be discussed in the next chapter, Wojnilower argues that the shift away from direct bank funding in regulated markets with credit rationing to a system in which credit is supplied through auction-like markets could not have occurred in the absence of government guarantees. The shift to nonbank funding might be attributed to three developments. First, the financial position of banks and thrifts deteriorated due to problem loans. Equity was destroyed through loan losses, and the Fed placed pressure on banks and thrifts to increase equity ratios. Second, innovations in technology and in financial practices made it easier and more acceptable for corporations to obtain funding outside banks. Revisions of rules of thumb regarding acceptable leverage ratios, for example, made it possible for firms to issue more debt, and direct lenders loosened standards to obtain higher returns. Finally, deregulation made it possible for ‘nonbank banks’ to compete with banks in areas which formerly had been restricted.

Off-balance sheet commitments Banks frequently engage in ‘commitments to finance which do not show up as actual funds lent or borrowed’ (Minsky, 1986b, p. 229). Banks are increasingly involved in making markets for nonbank credit through underwriting activities, market related advisory functions, and introductory services (Goodhart 1986). As such activities do not show up on balance sheets, they are not subject to reserve requirements. In 1980, bank loans made up 37 per cent of the total increase in debt issued by non-financial corporations, while this figure declined to only 18 per cent by 1985 (Hindle 1986, p. 18). As the importance of loans has declined, banks have moved increasingly into market making activities. As a result, by 1986, on-balance sheet commitments of the seven largest US banks totaled less than $550 billion, while off- balance sheet commitments

reached $1.4 trillion (Stevenson 1987, p. 2). A line of credit is an example of an off-balance sheet operation. Moore (1984) reports that half of the credit granted by banks is in

the form of lines of credit, and about 50 per cent of the lines of

213,

to credit are typically utilized. Firms often use lines of credit of use The paper. increase the marketability of commercial (see 1960s late the in rapidly unsecured commercial paper grew

(in response Table 7.16, Column 1). As interest rates rose in 1970

to tight monetary policy), the Penn-Central Railroad was forced to default on its commercial paper. This led to a run on commercial paper, which was halted when the Fed encouraged commercial banks to refinance the corporations experiencing a run on their paper. The Fed used open market purchases to increase bank reserves, and opened the discount window

to banks which were

refinancing commercial paper. The commercial paper market was thus saved by Fed intervention. Since 1970, it has been standard practice for corporations which issue commercial paper to obtain a line of credit sufficient to retire their commercial paper if

needed.”

Not

surprisingly,

there

has been

an

explosion

in

commercial paper issued after the late 1970s, much of which is guaranteed by lines of credit. Table 7.17 shows commitments and contingencies of FDIC insured banks. For example, standby letters of credit were $170 billion, commitments to make or purchase loans were $572 billion,

and commitments to purchase foreign currency and dollar exchange

were $891 billion in 1986. In 1987, commitments to purchase foreign currency and dollar exchange reached $1.5 trillion. In contrast, demand deposits in 1986 were only $511 billion. Clearly, off-balance sheet commitments are relatively large. These figures probably vastly understate the total value of off-balance sheet commitments,

which

by nature

are

difficult

to

discover

and

measure. While the credit lines (such as those associated with commercial paper) do not appear on the balance sheets of banks, they are a covert liability. If a firm utilizes its line of credit, the bank must

provide credit at the request of the firm. The bank must then come up with the funds required. Although reserves are not required against credit lines which have not been utilized, the existence of credit lines creates the possibility that banks may be forced to come up with substantial quantities of reserves at short notice. Since the decision to use a line of credit rests in the hands of the firm, the extension

of credit when a line is utilized

must

be

demand determined." Because commercial banks which have extended lines of credit must be able to obtain reserves quickly, they have an incentive to prearrange their own credit lines. Thus, a commercial bank might 214

arrange for a credit line with foreign banks, with parent holding companies,

or with

affiliated

banks.

If a commercial

bank

is

suddenly subject to an expansion in loan demand (from firms with prearranged lines of credit), it can use its credit line to obtain

funds.”

The danger of off-balance sheet items arises because their full impact may not be readily apparent. A bank which grants a line of credit may not be fully aware of the off-balance sheet arrangements which the customer has already made. A credit line makes the issuer responsible at least indirectly for all such arrangements the customer has made. A pyramiding of lines of credit increases the probability that a great number of them might be called upon at the same

time.

Thus, off-balance sheet items

may increase liquidity of a financial asset (as in the case of commercial paper backed by a credit line), but actually increase fragility of the financial system as a whole. Ultimately, the Fed, operating as a lender of last resort, is responsible for off-balance sheet items just as it is responsible for the covert items on a bank’s balance sheet. The new trend in securitization is for banks to move from simply making markets for nonbank debt instruments to becoming major issuers and purchasers of securities. Banks have become the major purchasers of short term paper, and have been heavy buyers of long term paper. They repackage loans for sale, or sell loans outright. The major sellers of loans are US money center banks, and the loans sold are usually obligations of top quality domestic commercial borrowers. Sales generate fee income on off-balance sheet operations and help to maintain customer relations.” A variety of financial instruments, including futures, forward rate agreements,

options,

swaps,

and

zero-coupon

bonds,

have

increasingly become important.” Banks are involved in making

markets for many of these instruments, although they will not show up on the balance sheets of banks. These allow banks and their customers ‘to adjust various forms of risk, or to take advantage of certain fiscal or other comparative anomalies in financial markets, without in most cases leading to any bank balance sheet expansion...’ (Goodhart 1986, p. 86). Innovation has, in some cases, increased liquidity associated with certain previously illiquid assets. For example, the creation of formal futures markets increases liquidity of these instruments and reduces the cost of trading. Sellers are able to hedge risks of price uncertainty. However, interdependence of on-balance sheet and

215

lead off-balance sheet commitments, or covariance of shocks can

to the transmission of shocks throughout the financial system.

Individual participants are undoubtedly able to reduce their risk of by the use of new hedging instruments, but the effectiveness these collectively still depends on a discharge of contracts which might prove difficult if the system as a whole received a severe shock. (Rose 1986, p. 37) [T]he new instruments transfer price or market risk from one economic agent to another, but do not eliminate that risk...they create new credit exposures, and thereby increase the ways in which the default of one borrower can adversely affect others. (Bank for International Settlements 1986, p. 3)

Credit risk is probably becoming more concentrated among fewer financial institutions (Bank for International Settlements 1986, p. 3). Furthermore, the new instruments may even encourage riskier portfolios: Although the new financial instruments and markets allow borrowers to hedge, indeed immunize, their position more closely than in the past, they also allow agents, whether consciously or inadvertently, to adopt even riskier positions, and in forms that may be hard to observe, monitor and limit. (Goodhart 1986, p. 99) Off-balance sheet operations have first, led to increased difficulty

for banks, customers, international exposure;

and the Fed to measure risk and second, increased the role played by

non-bank financial institutions; and third, meant that a change in

a bank’s reported assets may not reflect the change in a borrower’s liabilities (Bank for International Settlements 1986, p. 7). Risks inherent in various financial instruments include market risk, credit risk, market liquidity risk, settlement risk, country and

transfer risks, and so on. The question is whether the various risks involved have been appropriately priced. That is, are the average

profit margins sufficient to cover the potential losses both in the short and the long run? Bankers must price risk based on statistics and rules of thumb. The problem is that in a world of rapid innovation, there simply isn’t sufficient experience to come up with statistics or even rules of thumb. There is some evidence that the risks are not appropriately priced. For example, the BIS (Bank for International Settlements) points out that NIFs (note issuance 216

facilities) are priced at about half the price of commercial paper, even though the risks are similar. New entrants and lack of patents on most innovations push prices down so that in the short run, risks may be systematically underpriced. Even in the long run, risks may be underpriced because of the inability to foresee long run events (BIS 1986, pp. 198-201). The BIS goes on to argue that financial transactions, ‘whether traditional or innovative, will as before be

priced mainly on the basis of current perceptions of risk and the immediate supply and demand pressures for the transactions’ (BIS 1986, p. 202). Furthermore, innovative financial instruments should contain an

extra margin in their prices to account for unexpected events and subsequent losses. However, any institution which tries to incorporate this margin will not be able to compete in the short run. The evidence is, according to the BIS, that the margin on new

instruments is actually less (BIS 1986, p. 202). Off-balance sheet operations which unbundle market and credit tisk tend to spread risk, but actually may increase aggregate risk. While unbundling won’t increase market risk (it must sum to zero in the aggregate), credit risk increases in direct proportion to the volume of financial contracts outstanding. Unbundling makes parties more interdependent, linked through at least one financial institution. Thus, the rapid growth of off-balance sheet operations to spread price risk actually increases credit risk, and the extent to which risk is rising is not easily determined, since these operations

are not generally made public (BIS 1986, p. 204). Finally, if unbundling risk has concentrated price risk in the hands of a few banks (which the BIS believes is probable), then unbundling has not spread price risk. In this case, an unexpected change in the interest rate or the exchange rate may have a heavy impact on a few institutions which have accepted uncovered positions in options, futures markets, or spot markets. In summary, off-balance sheet operations allow non-bank credit to finance expenditures, which stretches liquidity because more money (broadly defined) is issued on a relatively smaller liquid base. Thus, the rate of growth of M1 or of M2 doesn’t give a good indication of the rate of growth of money, debt, spending, or future payment commitments. Furthermore, to the extent that off-balance sheet commitments increase in importance, the use of the prime interest rate or of other published interest rates as an indication of the cost of credit can give misleading results. The price of off-balance sheet commitments is set using rules of thumb and in

217

consideration of the pressures of supply and demand, and may be difficult to measure using published interest rates.

Conclusions

The purpose of this chapter has been to provide institutional detail and to dispel any remaining suspicion that a clearly definable and exploitable relation between ‘money’ and ‘spending’ exists. Asset and liability management, innovation, and off-balance sheet operations ensure that any relation between a monetary aggregate (for example, M1) and a spending aggregate (for example, GNP) is institutionally based and probably temporary. Modern financial relations break the links between demand deposits and spending, between any monetary aggregate and spending, and between overt bank liabilities and spending. Only final settlements need be made in terms of a universal means of payment. Positions in assets can be taken, and commitments to streams of payments can be taken which do not involve creation of demand deposits or other means of payment until some specified (or unspecified) future payment date. That payment may or may not be made - whether it is or not is irrelevant to the decision made today. The decision made today involves expectations and a wide variety of overt or covert, certain or uncertain, contingent or non-contingent portfolio adjustments. Asset and liability management, securitization, lines of credit, and futures contracts increase the liquidity of the individual. Given a small amount of cash, his spending power may be a large multiple if he can obtain a credit line. He can enter into commitments

which involve delivery of cash, goods, or services

today for cash tomorrow. For the system as a whole, such modern financial arrangements enable the ratios of loans to reserves, of demand deposits to reserves, of loans to demand deposits, and of payment commitments to cash flows to increase. But while these increase individual liquidity, they do not necessarily increase systemic liquidity. That is, as bank liability to reserve ratios rise, as M2 rises relative to M1, and as leverage ratios rise, the ability of the existing liquidity in the aggregate to allow liquidation of assets on a large scale falls. Aggregate liquidity simply cannot cover outstanding commitments: liquidity in the aggregate falls. In any modern capitalist economy, the meaning of ‘money’ is blurred. Ultimately, the liquidity of the system must lie in the hands of the central bank because endogenous processes will

218

‘stretch’ liquidity beyond the ability of the system to liquidate (exchange assets for the medium of exchange) at an exchange ratio which allows commitments to be fulfilled. It is the central bank’s willingness to intervene to guarantee parity of liquid assets with the medium of exchange which ultimately determines how much of the value of aggregate assets will have to be written-off during a debt deflation. The orthodox view of money, of spending, and of banking is flawed, partially due to the limited number of assets and liabilities

included in the analysis. The real world is much more complicated. Control of one (or two, or three) monetary aggregates won’t control spending, just as control over reserves won’t control money. Monetarist policy attempts to control spending by controlling liquidity. The policy is fundamentally flawed, because spending in an economy with developed capitalist relations stretches liquidity. Monetarists would have the central bank squeeze liquidity just as normal processes are already squeezing liquidity. Monetarist policy controls spending by inducing crisis. To restore order, the central bank must intervene to provide with its left hand the liquidity it has previously denied with its right. Even in the absence of a central bank, the money supply is endogenously determined. In previous chapters, I showed how a wide variety of credit instruments has been used to finance spending. The development of a central bank and adoption of various constraints contributed to the evolution of the financial system to the modern system based primarily on commercial banks. Central bank willingness to ensure commercial bank solvency has eliminated wide scale debt deflation (although smaller, regionalized failures have been permitted). Monetary policy appropriate to a commodity money system, but inappropriate to a capitalist, monetary economy, has contributed to a shift away from commercial banks and toward auction-like money markets. This appears to have reduced systemic liquidity and increased fragility, as evidenced in the US by rapidly rising numbers of cases of failures and of central bank interventions as lender of last resort. In the final chapter, I will discuss policy which is more sensible.

Notes 1. This discussion follows Ritter and Silber (1986, pp. 138-42). 2. ‘A commercial bank can lend up to the amount of its excess reserves, and no more.... While a single commercial bank can lend (and create

Zag

demand deposits) only up to the amount of its excess reserves, the banking system can create demand deposits up to a multiple of an original injection of excess reserves’ (Ritter and Silber 1986, pp. 37-40). See Minsky (1957). aes As argued by Minsky (1986b), the expansion of non-member banks, particularly overseas branches of US banks, has eliminated any effective reserve constraint on US banks. This is because nonmember banks are allowed to hold reserves as deposits in member banks. The number of non-member banks grew from 6,900 in 1960 to 8,914 in 1976, and from a total of 51 per cent (1960) of all US banks to 61 per cent (1976). Since member bank deposits total $619 billion (1976), these could support deposits of $6190 billion in nonmember banks subject to a hypothetical reserve requirement of 10 per cent. In 1976, actual deposits of non-member banks only came to $258 billion, so potential reserves would allow deposits in nonmember banks to expand almost without limit (Minsky 1986b, p. 245). Data are from Economic Report of the President, 1989, Table B-69.

Under a repurchase agreement (Repo), the Fed buys a security with an agreement that the seller will repurchase it on a specific date, so that reserves are temporarily injected into the system. Under a matched transaction (Reverse Repo), the Fed sells a security but simultaneously agrees to repurchase it on a specific date, so that reserves are temporarily removed from the system. Instruments eligible for such operations include US government securities, obligations that are direct obligations of (or fully guaranteed by) any agency of the US government, and prime bankers’ acceptances (Meek 1982797 121): Data are from Annual

Report, Board of Governors of the Federal

Reserve System, 1988. Data are from Federal Deposit Insurance Corporation Annual Report, various years. In the next chapter, I will analyze Wojnilower’s view on innovation and the development of auction markets. See Wojnilower (1987) for an account which differs from Minsky’s. 10. See Minsky (1986b, pp. 91-2) for an analysis of this run on commercial paper. it Above I mentioned the explosion in the commercial paper market since the late 1970s. Banks have increasingly become engaged in market-making activities for alternative credit sources. Thus, banks

provide credit lines which back commercial paper used by firms to raise funds. Such off-balance sheet commitments will not show up unless the firm is forced to use the credit line. ‘In December 1986, standby letters of credit issued by insured commercial banks totaled

almost $900 billion, and commitments on interest rate and currency

220

swaps amounted to $376 billion...the contingent liabilities just listed amount to approximately eight times bank capital’ (Newsletter, WW Norton and Company, Incorporated, Volume 4, January 1988, p. =) Note: these figures differ from those presented in Table 7.17, which may be due to the difficulty of measuring off-balance sheet operations. 12. Table 7.8 shows a huge increase in ‘other’ liabilities to foreigners. A substantial portion of this includes liabilities to own foreign offices and borrowing under repurchase agreements. 13. Bank for International Settlements (BIS 1986, pp. 135-7). . Four recent innovations are: a. Note Issuance Facility (NIF): A bank underwrites a stream of short term notes over a medium term. If investors lose confidence in the borrower,

the underwriter will have to take up the notes.

Banks act as both underwriters and as borrowers in this market. Banks borrowing generally issue short term CDs. Most NIF paper is placed with small banks, while underwriters are usually investment banks, or merchant banking arms of commercial banks. NIFs account

for about $75 billion (world-wide) in 1985 (BIS 1986, pp. 19-35). b. Currency and interest rate swaps: Two counterparties agree to exchange streams of payments over time according to a predetermined rule. At first, banks just arranged swaps to earn fees, but now intermediaries act as counterparties by entering into two offsetting swaps (or even without an off-setting swap). The swap exposes the financial institution to price risk (change in interest rate or foreign exchange rate) and credit risk (counterparty fails to perform). There aren’t any authoritative accounting standards to cover swaps. Swaps do not normally appear on balance sheets or in public financial statements. The BIS estimates that by 1985, 100-150 billion dollars in outstanding interest rate swaps existed, with a lesser amount in currency swaps (BIS 1986, pp. 37-51). c. Foreign Currency and interest rate options: An option is a contract conveying the right to buy or sell a financial instrument at a fixed price before or on a certain future date. Most traders hedge, but are not able to assure full protection. The only way to completely hedge an option is through the purchase of an identical (but opposite) option (BIS 1986, p. 77). d. Forward rate agreements: Two parties agree to an interest rate, and at maturity they settle by paying/receiving only the difference between the actual interest rate and the contract interest rate. These agreements are used mainly by banks, and don’t expand balance sheets. They involve credit risk because of the lack of margin requirements or exchange backing. In 1985, these averaged $7 billion per month. Forward rate agreements usually involve the dollar, and London

is the major trading center, with New York second. These

allow banks to decrease the size of their interbank books, decreasing

221

interest rate risk without inflating balance sheets. This improves apparent capital-to-asset ratios, and increases the measured rate of return on assets (BIS 1986, p. 121). 13. These considerations cast doubt on a simple markup approach to interest rates, as used by Moore (1988a) or Rousseas (1985b). In Chapter 6 above, I have offered an alternative to the markup approach to interest rates. Competitive pressures lead to innovations which provide new ways to generate finance, including a wide variety of off- balance sheet commitments. Prices of these commitments are set by rules of thumb and by competitive pressures. Banks cannot simply add a markup over the Fed funds rate (or some other wholesale cost of funds), as they must be concerned with competition from other potential suppliers of similar commitments. As these commitments don’t necessarily entail that the bank actually provide explicit bank liabilities at the present time, the wholesale cost of funds may not be an appropriate measure of the cost of providing commitments.

Furthermore, as the BIS argues, there is no indication

that the price charged on such commitments actually covers the cost (or potential cost).

222

Appendix

List of tables

al bee Th, 7.4 ies

Monetary Aggregates Selected Liquid Assets Domestically Held Selected Liquid Assets Domestically Held Ratios of Earnings and Expenses of Banks Selected Assets and Liabilities (by percent) by Bank Size

7.6 1 7.8

Government

1) 7.10

7-11 iw 7:13 7.14 7.15 7.16 iL]

Debt, Loans, and Other Securities

Reserves and Open Market Operations

224 225 226 22], 228 230 Zoe

Short Term Liabilities to, Short Term

Claims on Foreigners Growth of Selected Items Selected Assets of Insured Commercial Banks Selected Liabilities of Insured Commercial Banks Selected Assets of Insured Banks, By Percent Selected Liabilities of Insured Banks, By Percent Federal Reserve Open Market Transactions Summary of Federal Reserve Open Market Transactions Commercial Paper and Banker’s Acceptances Commitments and Contingencies of Commercial Banks

223

232 233, 234 233 236 230 238 239 240 241

Table 7.1 Monetary aggregates i (M1)

Z (M2)

141

298 312 336 363 393 425 459

147 155 162 170 174 185 199 206 217 231

3 (dM1)

4 (dM2)

;

524 ;

266 278

5 (Req Res)

6 (XR)

13,046 12,979 13,564 13,840 14,220 14,781 15,308 15,388 16,748 17,589 18,005 18,932 20,233 22,190 23,178 24,454 24,580 25,112 26,221 27,453 28,529 30,294 31,757 33,820 35,546 39,057 44,999 54,798 57,693 59,936

506

Notes: Column 1 is M1 and Column 2 is M2, both measured as averages of daily figures in billions of dollars, seasonally adjusted. Columns 3 and 4 are the growth rates of M1 and M2, respectively, measured as the percentage change from the previous year. Column 5 is required reserves (adjusted for changes in reserve requirements) measured as averages of daily figures in millions of dollars, seasonally adjusted. Column 6 is excess reserves, measured in the same manner as required reserves. All data are derived from the Economic Report of the President, 1988 (and 1989 for the years 1987-88), Tables B-67 and B-69. Note that figures for 1988 are projected (from November 1988).

224

tablew/2 Selected liquid assets domestically held n wn

ofa sooococco

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Notes: All figures are averages of daily figures, in seasonally adjusted billions of dollars (except columns 4, 5, and 6, which are not seasonally adjusted). Column 1 presents currency, Column 2 presents demand deposits, and Column 3 presents other checkable deposits. Column 4 presents overnight repurchase agreements (net), plus overnight Eurodollars. Column 5 presents general purpose and broker/dealer money market mutual fund balances, while Column 6 presents institution-only money market mutual fund balances. All data are derived from The Economic Report of the President, 1988 (and 1989 for the years 1987-88), Table B-68. (Figures for 1988 are projected from November 1988.)

jae)

Table 7.3 Selected liquid assets domestically held

ee) te) fr) Te te al a a )

NW Ro

2a) 35 6.8 7.9 8.2

Notes: All data are averages of daily figures, billions of dollars, and seasonally adjusted (except columns 1, 5, and 6, which are not seasonally adjusted). Column 1 presents money market deposit accounts, Column 2 shows savings deposits, Column 3 presents small denomination time deposits (less than $100,000), and Column 4 presents large denomination time deposits (greater than $100,000). Column S presents term repurchase agreements, while Column 6 presents term Eurodollars. All data are derived from same source as Table

Tas

226

Table 7.4 Ratios of earnings and expenses of insured commercial banks for selected items (amounts per $100 of current operating earnings)

Interest on

Interest on

gov't bonds _ dividents & securities

Income on

Fed

Interest on

loans

funds

savings and time deposits

na

9.75 10.71 10.63

22.01 22.04 20.91

$.43 5.63 :

; : :

18.57

;

‘

na na

11.14

na na na na

olumn 1 presents interest on

federal

government

bonds, while interest

on state and local government bonds is included in Column 2 except during the years 1969-80, when it is included in Column 1. Interest on other securities is included in Column 2 until 1981, after which it appears in Column 1. Column 3 includes interest and fees on loans (including Fed funds sold until 1981). Column 4 presents interest on Fed funds sold (after 1981). Column 5 presents interest on savings and time deposits. After 1981, interest on foreign-held deposits is included in Column 5. Data for years 1952-1980 were taken from the Federal Deposit Insurance Corporation Annual Report, various years. Data for years after 1980 were derived from Federal Reserve Bulletin, July, 1987, p. 551.

221,

Table 7.5 (by percentage) of insured liabilities and assets Selected commercial operating in the United States and other areas,

banks grouped by amount of assets December 31,

1971/December 31, 1980 1 All

BANKS WITH ASSETS OF: (in millions) 2 3 4 5 6 7 2-5 5-10 10-25 25-50 50-100 S5001000

8 1000

A) Cash

1971 1980

iS)

is

iS

I)

Ties

anSy

ise) 13.9

B) Gov Bond

1971 1980

12S

27-95

23:4

191

G:8, ee LSS

LO :4 ee re 11.6

31 43 COSTS

41 RL

39 GR

3.5 OE

3.2 49

33 5.8

47:49

48'S

49/9

SSOSNS 522

C) Fed Funds (sold)

1971 |—67TOED

D) Other Loan

1971 1980

S13"

F468

E) -Commer Loan

1971

13S

G5/geeeS

See 0.4:

i) Se 4-4

aS 2G sO

10 me al

1980 F) -- Instlmnt Loan

1971 1980

Soe

1971 1980

41.09

40'5 S32

H) Time Deposit

1971 1980

43°2)

48/49550:3) 54.9 61.8

I) Fed Funds (Pur)

1971

sky

6.9.0

LIABILITIES G) Demand Deposit

1980

J) Equity

1971 1980

Cm LO4 a

50395193 Sarasa

On 2/2)

7-9

379s

951.5950:75

7

5.1

Seeste aes

5034045

372

(Hal O2maO3

0.8

6.9

8.8 119

G Sa CRAIG

lee al G9 10.0

48

Notes: For each asset, the first row presents percentages (as percent of total assets) for 1971, while the second row presents percentages for 1980. Columns present data by size of bank as measured by assets: 1=All banks; 2=Assets of

228

$2-5 million (for 1980, all banks with assets of less than $5 million); 3= Assets

of $5-10 million; 4=Assets of $10-25 million; 5=Assets of $25-50 million; 6=Assets of $50-100 million; 7=Assets of $500-1000 million; and Column

8=Assets greater than $1 billion (for 1980, assets greater than $5 billion). Row A=Cash and due from depository institutions; Row B=US Treasury securities and obligations of other US government agencies and corporations; Row C=Federal funds sold and securities purchased under agreements to resell; Row D=Net loans (excludes losses and fed funds loaned); Row E=Commercial and industrial loans; Row F=Installment loans to individuals; Row G= Demand deposit liabilities; Row H=Time deposit liabilities; Row I=Federal funds purchased and securities sold under agreements to repurchase; and Row J=Equity capital. Sums of assets and/or liabilities will not equal 100% due to rounding error, and due to exclusion of some (smaller) assets and liabilities. Note: commercial and installment loans are included in Row D. All data are from the Federal Deposit Insurance Corporation Annual Reports, 1971 and

1980.

229

' Table 7.6 ial commerc by held s securitie other and loans, Government debt, banks Year(s)

1929 1933 1939

Gov't Sec ($ Billion) Sul) 10.3 2371

1941-50*

—

1955-59* 1960

-61.0

1966-70* 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 Average 1981-88*

— 64.9 89.0 88.2 86.3 116.7 136.3 136.6 137.6 144.3 170.6 179.3 201.8 259.4 260.6 271.4 309.9 335.0 356.8

Loan + Sec ($ Billion) 58.6 42.1 54.6

a — 199.5

Govt Sec/Loan +Sec (Percent) 9.8 24.5 42.3

61.1 40.6 30.6

_

15.4

516.6 SIPAS) 647.9 713.9 745.3 804.9 891.9 10143 1136.1 1239.0 1307.2 1401.0 1553.0 1722.6 1908.0 2089.8 2233.0 2398.1

12.6 15.5 13.6 12.1 15.7 16.9 1S\.3} 13.6 1257) 13.8 13.7 14.4 16.7 Set 14.2 14.8 15.0 14.9 14.9

Notes: Data are derived from The Federal Reserve Bulletin, various years. Figures for 1988 are for November. Column 1 shows total US government securities held by commercial banks (previous to 1960, all banks) at the end of December (previous to 1939, end of June) in billions of dollars. Column 2 shows total loans and securities (including government securities) held by commercial banks (previous to 1960, all banks) on the same date, in billions of dollars. Column 3 is Column 1 divided by Column 2, expressed as a percentage. *Indicates a simple average for indicated period.

230

Table 7.7 Reserves and Fed open market operations (millions of dollars)

BR 1970 332 1971 126 1972 1050 1973 1298 1974. 727 1975 130 1976 53 1977 569 1978-9 868 1979 1473 1980 1690 1981 §.636, 1982 634 1983 774 1984 3186 1985 1318 1986 827 1987 777

OMP

BR+ OMP 4982 5314 8866 8992 272 1322 9227 10,525 6149 6921 8539 8669 9833 9886 7143 7712 6951 7819 7693 9166 4497 6187 9175. 9811 9773 10,407 10987 11,761 8414 11,600 21621 22,939 30211 31,038 9961 10,738 16069

RR

(3/4),

18932 20233 22,190 23,178 24454 24580 25,112 26,221 27,453 28529 30,294 31,757. 33820 35,546 39,057 44999 54,798 57,693

281 444 60 45.4 283 353 394 294 285 321 204 30.9 308 331 297 510 566 186

~dRR_

(3/6)

7-57 1301 69 1957 0.7 988 10.7 12% 75 126 68.8 532 18.6 1109 70 1232 63 107% 85 1765 3.5 9.1463. 6.7 2063 5.0 172 68 3511 33 5942 39 9799 32 2895 3.7 2243

(XR)

(8/4)

248 «13 182 09 284 «13 303 «13 25911 26690 iit 27S A 190 0.7 233 (0.8 4421.5 Stee 7 320 1.0 500 1.5 561 1.6 852 1.5 1057 23 1368 2.5 1029 18

Notes: Column 1 presents reserves borrowed from the Fed by all depository institutions. Column 2 shows total net change in system open market account (see Table 7.15 below). Column 3 is the sum of Columns 1 and 2: borrowed reserves plus net change in open market account. Column 4 is required reserves (adjusted for changes in reserve requirements) measured as averages of daily figures in millions of dollars, seasonally adjusted. Column 5 is Column 3 divided by Column 4 and multiplicd by 100, which gives a measure of the percentage of required reserves supplied by the Fed. Column 6 shows change in required reserves from previous period. Column 7 is column 3 divided by Column 6, and gives a measure of the amount of required reserves which is supplied by the Fed relative to the increase in required reserves. Column 8 is excess reserves, measured in the same manner as required reserves. Column 9 is Column 8 divided by Column 4, expressed as a percentage, and gives the percentage of reserves which are excess reserves. All data are derived from the Economic Report of the President, 1988 (1989 for years 1987-88), except Column 2, which is taken from various years of the Federal Reserve Bulletin (except years 1987-88, which are taken from Annual Report of the Board of Governors of the Federal Reserve System, 1988). Note that figures for 1988 are projected (from November 1988).

zoe

Table 7.8 and short term claims on foreigners to liabilities term Short foreigners reported by banks in the US, billions of dollars, end of period, payable in dollars LIABILITIES

CLAIMS 4

5

6

Tbill

other

total

8 Acc

9 memo

5.6 4.0

43

83

Tel

9.1

yi

Ae)

na

ey

We

135 seul

Bl

tel

3.1

4.0

na na

ills).

ees

Sil)

TENS)

SY)

40

43

na

Sf)

ay?

SO

LS:4

na

e207,

77

42

BO)

na

beh

eT

iis)

hie

na

eyie\

Boys)

YA

RA

EI

na

40.7

394

69.2

182

124

na

489

441

79.9

20.0

14.2

na

Stk}

ee

BY

651

na _ 18.0

(Ss)

7 loans

486

1020

SON

LO8:4 mee O2)

553

393

e159 een

ae:

349

55:69

166.8

991739

Seales

82

307

687)

-219:8)

1762"

“nay

3727,

7642009

167.9

Sena

37

69.1

2278

1608

na

285

90.3

292.4

1584

na

25.6

meal

272 294

Notes: Data were obtained from the Federal Reserve Bulletin, various years. Short term liabilities and claims are those with a maturity of one year or less. Before 1978, claims include claims of US banks on foreign branches, but afterward, only claims on unaffiliated foreigners are included. Column 1=total liabilities to foreigners; Column 2=demand deposits due foreigners (excludes bank liabilities to own foreign offices); Column 3=time deposits due foreigners (excludes liabilities to own foreign offices); Column 4=US treasury bills held by foreigners; Column 5=other liabilities to foreigners (includes liabilities to own foreign offices, other negotiable instruments, and borrowing under repurchase agreements); Column 6=total claims on foreigners; Column 7=total loans to foreigners; Column 8=total acceptances; Column 9= foreign office deposits of US banks.

232

iablea7.9 Growth of selected items (percentage change over period)* ITEM

PERIOD:

3 Req’d Reserve

4 5 6 7 8

Currency Demand Deposit Time Dep. small Time Dep. large Other Checkable

1960-65 19.71 47.12 17.94 25.61 17.70 76.00 960.00 --

1965-70 27.65 36.82 23.67 35.54 24.81 337.97 113.21 0

1970-75. 34.10 62.90 29.83 50.00 28.92 123.63 186.95 800.00

1975-80 43.30 59.63 23.25 58.13 23.83 115.66 100.31 3288.89

1980-85 50.36 57.38 48.54 46.19 2.64 21.34 67.86 484.59

9 Total Bank Loans

68.44

67.31

96.18

10 -Com. and Indust.

43.94

60.83

104.21 35.46

37.84

42.36

12 Fed Funds Sold

11 Securities

133.75

87.43

89.87

13 Fed Funds Pur

205.88

153.85

68.18

14 ST Liab. to For.

132.84

WANG At

111.74

15 ST Claim on For.

397.03

64.74

94.43

53.49

63.36

56.48

611

797

1473

6339

7443

10,745

Notes: Calculated from Tables 7.1, 7.2, 7.3, 7.7, 7.8, 7.10, and 7.11 in this appendix. *Figures for Borrowed Reserves and Open Market Purchases represent annual average over the designated period (not growth during the period). **Growth over period 1967-70. Rows present: 1=M1; 2=M2; 3=Required reserves of all depository institutions; 4= Currency held by public; 5=Demand deposit liabilities of all depository institutions; 6,7=Time deposits of all depository institutions, small (less than $100,00) and large (greater than $100,00), respectively; 8= Other checkable deposit liabilities of depository institutions; 9= Loans by insured commercial banks in the US; 10=Commercial and industrial loans by insured commercial banks in the US; 11=Securities held by insured commercial banks in the US; 12,13=Fed funds sold and bought

(respectively) by insured commercial banks in the US; liabilities and claims (respectively) to/from foreigners; of insured commercial banks in the US; 17=Reserves depository institutions; 18=Total net change in system

Zoo

14,15=Short term 16=Total equity capital borrowed from Fed by open market account.

Table 7.10 ial banks in the United States commerc insured of assets Selected * and other areas, December call dates (billions of dollars) 4 fixed

8 other

45.2

Sh)

1.5

;

88.3

6.0

46

98.2

6.7

63

108.4

8.1

ell

91

7.9

6 C&I

2 Cash

1 Total

280

56.2

455

715

505

83.3

:

531

89.3

;

:

576

93.0

1222

640

98.7

118.4

10.3

8.7

129.0

175.9

15.6

34.0

130.2

178.6

16.7

41.8

160.4

197.1

18.3

50.5

2232

20.6

63.8

192.4

257.7

22.6

65.2

202.0

282.9

25

91.9

82S

454.5

30.4

105.3

334.3

504.1

34.0

120.1

341.8

524.9

36.7

= 120.7

323.6

$65.2

38.4

152.4

340.6

$77.7

40.6

159.2

37/92

600.9

42.6

153.7

358.3

589.7

449

= 550.7

178.3

:

:

Notes: *After 1981, foreign offices of domestic banks are included. 1=Total assets; Column 2=Cash and due from depository institutions, Column 3=Securities; Column 4=Fed funds sold and securities purchased under agreements to resell; Column 5=Total loans and leases (gross); Column 6=Commercial and industrial loans; Column 7= Bank premises and fixed assets; Column 8=all other assets (excludes intangible assets, lease financing receivables, and other real estate owned). Columns will not sum to the total due to exclusion of some minor items. Note that Column 5 includes Column 6. All data were obtained from Federal Deposit Insurance Corporation Annual Report, various issues.

234

Table 7.11 Selected liabilities of insured commercial banks in the United

States (states and other areas), December call dates, in billions of dollars*

1

2

3

Total

Deposit

DD

Notes

a,

248

165

0.5

22

396

210

0.6

34

435

229

ip

37

437

240

3.4

483

247

2.6

539

262

9 Equ

135

781

831

8 Other

46

334

Sal

929

6.7

1016

400

22.8

1095

432

27.9

1192

432

Pais

1509

384

45.4

1706

371

55)!

1842

55.0

1962

69.5

2118

95.4

2284

107.7

2335

121.0

NOTES: *After 1981, foreign offices of domestic banks are included. Foreign office deposits are included in Column 2, but are excluded from Columns 3-5. Column 1=Total liabilities, limited-life preferred stock, and equity capital; Column 2=Total deposits (includes Columns 3-5) Column 3=Total demand deposits; Column 4=Total savings deposits; Column 5=Total time deposits; Column 6= Federal funds purchased and securities sold under agreements to repurchase; Column 7= Demand notes issued to the US Treasury and other liabilities for borrowed money; Column 8=AIll other liabilities (excludes mortgage indebtedness, acceptances outstanding, and subordinated notes and debentures); Column 9= Total equity capital. Note that some minor items are excluded, so columns will not sum to the total. All data were obtained from the Federal Deposit Insurance Corporation Annual Report, various years.

235

Table 7.12 banks in the United States commercial Selected assets of insured

and

other areas, by percentage of total assets

NOTES: Column 1=Cash and due from depository institutions; Column 2=Securities; Column 3=Fed funds sold and securities purchased under

agreements to resell; Column 4=Total loans and leases (gross); Column 5=Commercial and industrial loans; Column 6=Bank premises and fixed assets; Column 7=all other assets (excludes intangible assets, lease financing receivables and other real estate owned). All data were derived from Table 7.10

above.

236

Table 7.13 Selected liabilities of insured commercial banks in the United

States, percentages of total liabilities 1

2

6

7.

8

Deposit

DD

Notes

Other

Equity

88.6

58.9

0.2

1.8

eo,

87.0

46.2

0.1

2.9

eS

86.1

45.3

:

0.2

3.6

3

82.3

45.2

:

0.6

4.0

is

: :

83.9

42.9

s

j

0.5

3.6

US

84.2

40.9

;

5

0.2

Dell

v3

83.2

34.2

:

0.5

sy

7.0

82.2

33.0

2

0.5

elf

Tall

81.6

33.3

:

0.6

742)

6.9

79.8

31.4

:

1.8

3.7

6.8

WES)

30.7

A

2.0

4.7

6.9

77.5

28.1

:

1.8

46

7.0

78.3

:

20

6.1

6.9

7718

:

2.9

6.0

6.8

78.7

:

acd

all

6.9

78.2

F

77.6 AI

:

778

:

32

2.6

7.0

2

40

2.4

7.0

:

41

fa?

6.9

:

46

2.6

6.8

Notes: Column 1=Total deposits; Column 2=Total demand deposits; Column 3=Total savings deposits; Column 4=Total time deposits; Column 5= Federal funds purchased and securities sold under agreements to repurchase; Column 6=Demand notes issued to the U.S. Treasury and other liabilities for borrowed money; Column 7=All other liabilities (excludes mortgage indebtedness, acceptances outstanding, and subordinated notes and debentures); Column 8=Total equity capital. For 1981 and later years, foreign office deposits were excluded from total liabilities to calculate percentages in Columns 2-8 so that derived figures would be comparable to previous years. All data were derived from Table 7.11 above, except that figures for foreign office deposits were obtained from Federal Deposit Insurance Corporation Annual Report, various issues.

Pe |

Table 7.14 Federal Reserve open market transactions (millions of dollars) Total Outright (All Maturities) Gross Purchase

1 12,362 12,515 10,142 18,121 13,537 21,313 19,707 20,898 24,591 22,325 12,232 16,690 19,870 22,540 23,776 26,499 24,078

Matched Transactions

Gross Sale

Redemption

2 5214 3642 6467

3

5599 8639 7241 13,725 6855 7331 6769 8369 3420 8857 4218 2502

Gross Sale

4 12,177 16,205 23,319 45,780 64,229 151,205 196,078 425,214 511,126 627,350 674,000 589,312 543,804 578,591 808,986 866,175 927,997

Notes: See Table 7.15 for sources.

238

Gross Purchase

5 12177 16,205 23,319 45,780 62,801 152,132 196,579 423,841 510,854 624,192 675,496 589,647 543,173 576,908 810,432 865,968 927,247

Repurchase Agreements Gross Purchase

Gross Sale

6 33,859 44,741 31,108 —-74,755 71,333 140,311 232,891 178,683 151,618 107,051 113,902 —-79,920 130,774 105,971

7 33,859 43,519 32,228 74,795 70,947 139,538 ~——-230,355 —-180,535 152,436 106,968 —-113,040 78,733 —-130,286 108,291

127,933 134,253 170,431

—«127,690 132,351 160,268

Table 7.15 Summary of Federal Reserve open market transactions (Millions of dollars) Net change Government Securities

Total net Change

4988

4982 8866

9626 8358 12,631 8908 20,477 29,989

Notes: All data in Tables 7.14 and 7.15 were obtained from Federal Reserve Bulletin, various years, except for years 1987-88, which were obtained from Annual Report of the Board of Governors of the Federal Reserve System, 1988.

239

Table 7.16 Commercial paper and banker’s acceptances (millions of dollars)

1

2

Commercial and Finance Company Paper

Acceptances Outstanding, Held by Accepting Banks

3202

319

4497

662

4686

1272

6000

1153

6747

1291

8361

1671

9058

1223

13,279

1198

16,535

1906

20,497

1544

31,709

1567

31,765

2694

32,126

3480

34,721 41,073

2837

49,070 48,459 53,025 65,112 83,665 112,803 124,374 165,829 166,670 188,304 237,586 300,899 331,016

Note: All figures were obtained from the Federal Reserve Bulletin, various years.

240

Table 7.17 Commitments and contingencies of FDIC-insured commercial banks, domestic and foreign offices

Commitments

to make or

(millions of dollars 1985

1986

1987

$43,248

572,079

613,767

purchase loans Futures and Forward Contracts -Commitments to Purchase -Commitments to Sell

122,668 137,562

When-issued Securities -Commitments to Purchase -Commitments to Sell

Standby Contracts and Options -Obligations to Purchase -Obligations to Sell

27,797 11,780

48,945 16,436

735,209

890,908

1,501,937

Standby letters of credit* -To US addressees -to non-US addressees

175,047 134,895 38,225

169,746 na na

169,891 na na

Commercial letters of credit

28,441

Commitments to purchase foreign currency and dollar exchange

30,594

Participations in acceptances conveyed to others by the bank

9.

Participations in acceptances acquired by the bank

10. Securities borrowed

11. Securities lent

Memo: notional value of outstanding interest rate swaps

715,635

Notes: *Includes foreign office guarantees. All data were obtained from Table RC-7, Federal Deposit Insurance Corporation Annual Report, 1985, 1986, and 1987.

241

8

REVIEW OF EVIDENCE SUPPORTING THE ENDOGENOUS MONEY APPROACH

Introduction

In this chapter, I will briefly review the evidence in support of the endogenous money approach. This evidence will be classified in four groups: 1) institutional innovations which imply an endogenous money supply; 2) informal analysis of the Fed’s role; 3) formal empirical tests which support the endogenous approach; and 4) a brief summary of the typical processes during a business cycle, which lends support to the endogenous approach presented in the sections above.

Institutional innovation

Chapter 7 showed how banks switched from operating on liabilities to operating on assets. When a bank is approached by a creditworthy customer, it makes a loan and then purchases funds to replenish reserves. Sources of reserves include the Fed funds market, repurchase agreements, certificates of deposit, borrowing from foreign banks, or borrowing at the discount window. I have shown that banks have come to rely increasingly on purchased funds as a source of reserves, and on loans as a use of funds. Indeed, demand deposits plus saving deposits are now only 69 per cent of the total loans made by commercial banks (versus 116 per cent in 1970). This tends to indicate that banks do not passively accept deposits and then purchase earning assets. Firms have increasingly negotiated lines of credit which can be used on demand. While this does not prove that banks meet loan demand and then purchase funds, it does show that banks could be forced

to behave in this manner. Finally, I have presented evidence showing the extent of off-balance sheet commitments, which entail spending or commitments to spend, but which do not show up on balance sheets, so do not count as part of the money supply as 242

normally defined. As mentioned in Chapter 7, the off-balance sheet commitments of the seven largest US banks are nearly three times the size of balance sheet commitments (Stevenson 1987, p. 2). Thus, the trend regarding bank balance sheets is consistent with the hypothesis that the supply of loans is a function of the demand for loans. I have shown how banks have greatly reduced reserves relative to liabilities, and have reduced secondary reserves (government bonds) relative to assets. Thus, a given amount of reserves supports a greater expansion of bank liabilities. I have also argued that when banks make loans they create deposits. If deposit holders can be induced to accept bank liabilities with no reserve requirement, the deposits created by the banking system return to replenish reserves. In other words, even if reserves do not increase, banks

do have a ‘widow’s cruse’ as long as the public is willing to hold financial system liabilities which do not increase legal reserve requirements. Therefore, there is no simple relation between reserves and bank liabilities. This indicates that banks are not reserve constrained, in the sense that they could expand loans if faced by additional creditworthy customers. The data presented above also show that the public has, in fact, been willing to increase its holding of bank liabilities which do not entail legal reserves. In the absence of reserve requirements, there cannot be a strict quantity constraint regarding expansion of bank assets and liabilities. This does not mean, of course, that financial

institutions actually will become infinitely large. Banks must make loans to customers who will almost never default because bank equity plus loan loss reserves are very small relative to total assets. Thus, banks are not able to absorb many defaults on loans. Furthermore, banks are constrained by a finite number of creditworthy customers who are willing to borrow loans. Their customers must weigh expected profits against costs, including finance costs, so that their loan demand is finite. Financial institutions are also constrained by rules of thumb regarding prudent leverage ratios and levels of liquidity which are ‘safe’. An individual bank cannot stray too far from accepted norms of behavior without raising suspicion and risking loss of sources of borrowed funds. As Minsky (1957) argues, these rules of thumb change over time, so that a run of good times may cause an upward revision which allows expansion of balance sheets even with no increase in aggregate reserves. The public, too, has been willing to trade liquid cash and demand deposits for less liquid assets, and has increased its 243

willingness to finance positions by issuing debt. Thus, household and business income flows have become more leveraged just as banks have leveraged their own capital stock and liquid assets. It has frequently been argued that the pace of innovation has been increased over the past two decades as banks attempted to subvert central bank constraints, particularly as nominal interest rates increased, due to inflation and tight monetary policy. Wojnilower (1987) provides an alternative explanation. First, he argues that one cannot ignore the substantial technological innovations which reduced transactions costs and which enabled nonbank financial institutions to provide a wide range of services to customers without requiring maintenance of expensive retail offices. The impact of this was briefly discussed in the previous chapter. Second, the regulations put in place in the US primarily during the 1930s eliminated cut-throat competition by creating geographic and functional barriers. This provided sufficient profits that protected commercial banks and thrifts could ration credit and enjoy business with established customers without fear of encroachment. However, technological innovation and changes in the attitudes of regulators gradually led to new entrants and severe competition. This provided an incentive to expand by providing credit to those who previously had been denied credit - because their projects were considered too risky, because their leverage ratios were too high, or merely due to lack of credit history. That is, rules of thumb regarding who is a creditworthy borrower were revised, so that credit markets become

more

than being based on customer-bank relations.

auction-like rather

The combination of elimination of interest rate ceilings, increased

competition for funds, and tight money policy pushed up the cost of issuing liabilities. In a system based on intimate relations, high interest rates would have generated non-price rationing because a bank would refuse to lend to customers if it believed the high loan rate would increase the probability of default. Thus, in a system based on credit rationing, rising interest rates would be met by stricter lending standards.

A ‘credit crunch’

could

occur,

where

prospective borrowers would be turned down. This would then slow down spending and help to end an inflationary boom. Wojnilower (1983) argues that such crunches were ‘wholesome’. Innovations and deregulation have allowed the development of auction-like markets where borrowers are constrained only by interest rates (if they are refused credit at a bank, they turn to junk

244

bonds). Wojnilower (1985) argues that a credit system which is only subject to a price constraint is fundamentally unstable. This is because the demand for credit is extremely inelastic with respect to interest rates. As an example, he points to the December 1980 to October 1981 period, during which the prime rate stayed between 18 and 20 per cent while business loans grew at record rates. The development of the floating rate loan actually decreased the elasticity of credit demand, since a high current short term rate will not reduce demand for credit if it is believed that interest rates will soon fall. The higher the short term rate, the greater is the

belief that it will fall. Furthermore, rising short term rates may even have a perverse effect as firms must borrow more to roll-over short term loans. In such an environment, tight money policy which is designed to raise interest rates can actually decrease the demand for loans only by inducing a crisis. In a system based on credit rationing, however, as interest rates rise, borrowers increasingly face quantity constraints. A speculative boom can be stopped earlier as banks cut off credit. Today, however, with quantity constraints essentially eliminated, a speculative boom can only be halted after widespread default threatens so that profit expectations fall and liquidity preference rises. In an environment without government insurance and guarantees, Wojnilower argues most of these innovations and revisions of standards would have been quickly abandoned as risks became apparent. However, given the central bank’s overt or covert promises to guarantee financial instruments, financial insititutions correctly perceived that they did not bear risks. Thus, rather than seeing innovation as a response to regulatory constraints, Wojnilower sees it as a response to deregulation in an environment in which the government guarantees that failure is impossible. If we take a longer run view, we see that financial institutions, like other types of firms, do innovate in search of profits. However,

we have also seen that there is a dialectic between government intervention and financial innovation. As we saw in Chapter 2, the widespread use of demand deposits was spurred by government constraints on note issue. The government constraint was placed on banks in an attempt to provide the Bank of England with monopoly power over note issue - which was to provide purchasing power to the state. However, by giving Bank of England notes (and, eventually, Bank of England deposits) a special status, the position of banks was ultimately strengthened because central bank

245

intervention as lender of last resort prevents bank failure and makes bank liabilities perfectly safe. More recently, we have seen a series of innovations designed to enable firms to escape constraints on lending which had resulted from government regulations, disintermediation, or institutional barriers. For example, a combination of tight money policy and interest rate ceilings on deposits led to disintermediation in 1969 (Wolfson 1986). Firms turned to the commercial paper market, which grew rapidly in 1970. After Penn Central was forced to default on its commercial paper, the Fed was forced to intervene to protect the commercial paper market. Since 1970, it has been standard practice for firms to obtain a back up line of credit on commercial paper. As another example, the real estate and investment trusts (REITs) were developed to provide loans to the riskiest part of the real estate market because banks refused to make such loans. The REITs, however, were backed up by lines of credit, so that when they failed, banks were forced to step in. As Wolfson discusses, each new innovation increased the ability of the system to provide credit. This stretched liquidity ratios and leveraged balance sheets until some agents were forced to default on the new credit instruments. The innovations were eventually validated in most cases by Fed intervention to prevent debt repudiation. In the US, Fed guarantee (whether de jure or de facto) against default has spread from small bank deposits to ‘most deposits in excess of the federal $100,000 per account

insurance limit, small

State-insured deposits, much commercial paper, repurchase agreements against Treasury securities, customers’ assets held by stock exchange firms, pension liabilities and some life insurance liabilities’, and so on (Wojnilower 1987, pp. 24-5). This has rightly led the public to believe that many types of financial instruments are liquid and safe. In turn, this has decreased the use of demand deposits and other commercial bank liabilities and led to the development of a wide variety of financial instruments which satisfy the public’s demand for liquidity. At the same time, government guarantees have increased the willingness of firms to sell goods, services, and assets on the basis of a wide variety of credit instruments. This, too, has decreased the share of financial business which shows up on the books of commercial banks.’ Therefore,

we

have

seen

a massive

shift out of the banking

system and into impersonal money markets. This has been associated with declining liquidity, increasing leverage ratios, and

246

rising potential for instability. Fed intervention to prevent defaults has" necessarily increased. Innovations have thus changed the environment and will force changes in the way the government intervenes. Wojnilower predicts a return to a financial system based on customer relations. He argues that auction markets do not work well in financial markets, and that the ‘present turbulence in the

US financial structure should make us more appreciative - as we ought to have been all along - of how well the system of 20 years ago worked’ (Wojnilower 1987, p. 26). In summary, trends regarding balance sheet positions taken by firms, households, and banks lend support to the endogenous money approach: even if the aggregate quantity of liquidity (reserves, high powered

money,

or whatever)

does not increase,

asset and liability management and innovations allow balance sheets to expand. If money is seen simply as the debt of banks (or as one of the components of the liabilities of various financial institutions), then the quantity of money clearly is not constrained by the aggregate stock of liquidity. This does not mean that the central bank has no control over the financial system, but that its

control does not operate through the base and a simple ‘money multiplier’.

Federal Reserve Bank behavior As shown above, the textbook version of the deposit expansion process begins with an injection of reserves by the Fed. In fact, as Moore argues: ‘Central Bankers insist that in the short run, money-

stock creation is a joint result of a complex interaction among households,

business

corporations,

financial

institutions,

the

Treasury and the central bank’ (Moore 1985, p. 25). Until 1979, the Fed targeted interest rates such as the Fed funds rate, so provided reserves as needed to maintain stability in financial markets. Although the Fed supposedly abandoned interest targets in 1979, it still did not, in fact, control

reserve growth.

Table 7.7 in the Appendix to Chapter 7 showed that in 1980, open market purchases fell by $3196 million below the level attained in 1979, while loans at the discount window increased by $217 million. As a result, the Fed supplied only 20 per cent of required reserves in 1980,

compared

to 32.1

per cent

in 1979.

However,

total

required reserves still increased by $1765 million between 1979 and 1980, and excess reserves actually increased by $72 million. Thus,

247

even in the face of tight monetary policy, banks were able to find needed reserves, and even increased excess reserves. Apparently, attempts to constrain reserves can be circumvented to some extent. The Council of Economic Advisors argued that lagged reserve accounting which existed until 1984 forced the Fed to ‘validate portfolio and loan decisions already made by the commercial banks.... The CEA concluded that at present the Fed is a passive supplier of reserves...’ (Moore 1984, p. 104). Moore reports that a number of studies had shown that Fed defensive operations had increased in number after the introduction of lagged reserve accounting in 1968 (Moore 1984, p. 108). To remedy this situation, it was recommended that the Fed switch to contemporaneus reserve accounting, which was done in 1984.

As Moore predicted, the switch to contemporaneous reserve accounting has not, in fact, increased the Fed’s ability to control reserves. Even the Fed has argued that the textbook view of the deposit expansion process is mistaken. At any given point, bank reserves depend on the volume of deposits outstanding and the consequent need for required reserves. If the growth in private deposits and reserve demand appear to be more rapid than desired, the Desk holds back on the provision of nonborrowed reserves. This forces banks to seek out sources of reserves, and, on the margin, to turn to the system

discount window. In the first instance...the Desk’s action does not reduce the flow of total reserves; it only changes the mix between nonborrowed and borrowed reserves. The sequence of relationships in this process is clear. The Desk holds back on the provision of nonborrowed reserves, forcing banks into debt at the discount window. This raises money market rates. Higher interest rates lead the public to economize on deposits, and demands for reserves

are

lowered.

In the last analysis,

while

the reserve

tightening process starts with the Desk holding back on the provision of nonborrowed reserves, the actual attainment of slower growth in total reserves and the aggregates reflects a lagged response of money demand to higher interest rates. (Federal Reserve Open Market Committee 1975, p. 31, cited in Moore 1984, p. 107)

If this really is the way the Fed behaves, then the switch to contemporaneous reserve accounting would not affect the Fed’s ability to control the money supply. The method of control would still entail price restraint, not quantity constraints. The data presented in Table 7.7 in the Appendix to Chapter 7 do

248

seem to support this view. Typically, when the Fed decreases open market purchases, it finds that loans at the discount window increase, and vice versa (although not on a dollar-for-dollar basis). Furthermore, as is shown in Columns 5 and 7, the percentage of required reserves (or of the change in required reserves) supplied by the Fed varies a great deal. This seems to indicate that Fed attempts to constrain reserves are not fully effective. In the endogenous money view, when the Fed decreases open market purchases, banks increase competition for other sources of reserves.

They tap international sources, the Fed funds market, and idle cash balances of firms, and they engage in reserve economizing behavior. This will typically raise interest rates. If banks still are in a deficient reserve position, they must turn to the Fed’s discount window. Thornton (1988) argues that targeting borrowed reserves, as the Fed has done since 1982, essentially means the Fed must supply reserves on demand. If money demand rises, banks will bid up the Fed funds rate as they try to meet the demand. As the Fed funds rate rises, banks turn increasingly to the discount window, which raises borrowed reserves. If the Fed has set a borrowed reserves target, it then must use open market purchases to increase bank reserves so that the Fed funds rate falls sufficiently that banks turn away from the discount window. Therefore, using a borrowed reserves target forces the Fed to supply reserves on demand, so that ‘all shifts in the demand for money are accommodated by corresponding shifts in money supply if the [borrowed reserves target] remains unchanged’ (Thornton 1988, p. 34). Thornton concludes ‘the borrowings procedure is a poor method for achieving money stock control’ (ibid., p. 43). As Moore argues, even when the Fed has adopted monetary targets, it specifies multiple targets, describes the target paths as cones, and shifts the cones up each year if actual paths exceed targets (Moore 1988b, p. 81). In this way, the Fed allows ample room for error and for hiding its true objectives. In spite of this, Moore notes that the Fed (like all central banks) characteristically overshoots targets. Central bankers in various countries with developed financial institutions have declared that they cannot control the money supply through manupulation of the quantity of reserves. Moore quotes a number of central bankers, all of whom argue that the orthodox view that the central bank determines the quantity of money by control over the base is simply incorrect. (See Moore 249

1988b, pp. 88-91.) Instead, they argue that they operate through short term interest rates. Finally, I have argued, particularly relying on the work of Minsky and Moore, that the Fed must ultimately operate as a lender of last resort. As a specific example, I will look at the Fed’s ‘experiment in Monetarism’ beginning in 1979. According to Fazzari and Minsky, ‘this shift was toward "practical" monetarism - that is, the Federal Reserve henceforth was to use the quantity and rate of change of some monetary aggregate as its proximate target for policy’ (Fazzari and Minsky 1984, p. 100). The Fed planned to tightly constrain the rate of growth of the money supply to stop the run from the dollar and to fight inflation. The Fed was to ignore interest rates and concentrate on slow growth of

monetary aggregates.”

The rate of growth of M1 fell from 7.1 per cent in 1979 to 6.4 per cent by 1981, while the interest rate on commercial paper rose from 10.91 per cent in 1979 to 14.76 per cent in 1981. (During the same period, the Fed raised the discount rate from 10.28 per cent to 13.42 per cent.)’ In 1980, the economy went into a recession, recovering somewhat in 1981. In 1982, the economy experienced a severe recession: the rate of growth of real GNP was -2.5 per cent, the unemployment rate exploded to 9.7 per cent, and net fixed nonresidential investment in constant (1982) dollars fell to half the 1979 level. At the same time, the number of bank failures nearly quadrupled during 1982 (from an average of 11 per year between 1975-81 to 42 during 1982), and the trade balance actually turned negative. (It is interesting to note that the tight monetary policy actually was associated with an increase in inflation from 8.9 per cent in 1979 to 9.7 per cent in 1981. Inflation fell only after the

Fed abandoned tight money policy.)*

The Fed was forced to abandon its tight money policies to prevent a serious financial crisis from becoming a financial crash. [T]he Federal Reserve felt forced to put on its ‘lender-of-lastresort’ hat. It abandoned the experiment with practical monetarism. The 1982 lender of last resort intervention took the form both of refinancing threatened domestic and international institutions, and of providing reserves to the financial system. As a result, interest rates fell rapidly and the growth of the monetary aggregates accelerated. Within six months of the Federal Reserve’s abandonment of monetarism there began what appears, at this writing, to be a robust recovery. (Fazzari and Minsky 1984,

250

p. 102)

The rate of growth of the money supply (M1) reached 9.5 per cent in 1983, fell to 5.8 per cent in 1984, and then began its spectacular climb from 12.5 per cent in 1985 to 16.5 per cent in 1986. Interest rates steadily fell (except for the slight rise in 1984) through 1986. The economy experienced one of its longest periods of expansion, which continues in 1990. (Again, note that the explosion in the rate of growth of M1 is associated with a fall of inflation to only 2.6 per cent by 1986.) In summary, tight monetary policy restricted the growth of reserves and caused banks to economize on reserves. While the money supply at first continued to grow rapidly, this meant that liquidity became stretched. Eventually, the rate of growth of the money supply slowed, and some economic units found they were unable to obtain funds at affordable terms. The Fed was eventually forced to intervene to prevent collapse by engaging in loose monetary policy and supplying liquidity. Lower discount rates, greater open market purchases, and intervention as lender of last resort prevented collapse of the financial system. This, combined with exploding government deficits, allowed a recovery.’ As Fazzari and Minsky (1984) argue, the Fed’s responsibilities for maintaining stability of the financial system conflict with its atempts to control the growth of monetary aggregates. If the Fed attempts to constrain the growth of reserves by raising interest rates, this increases fragility and forces the Fed to eventually intervene and provide the reserves needed to sustain the system. Giordano (1987) has analyzed the Fed’s response to six financial crises which have occurred over the past 20 years: the 1970 Penn Central crisis, the Franklin National Bank bankruptcy in 1974, the silver crisis of 1980, the failure of Drysdale Government Securities and the foreign debt crisis of 1982, and the Continental Illinois Bank failure of 1984. Giordano identifies three stages in the Fed’s response to such crises: Stage I is characterized by pumping liquidity into the system; Stage II is comprised of ‘watchful waiting’; and Stage III is characterized by large increases in liquidity and declining interest rates. Giordano explains: Stage I of the Federal Reserve’s response to a financial crisis might be deemed the panic prevention phase. Liquidity is quickly pumped into the banking system in order to calm financial market jitters, lower interest rates, and hopefully prevent the shock from spreading and precipitating a more dangerous

e\

systemic crisis. This infusion is accomplished by permanently or temporarily purchasing securities in the open market... This extra provision of reserves beyond the system’s usual requirements allows banks to reduce their dependence on large negotiable time deposits and nondeposit sources of funds, such as repurchase agreements and commercial paper.... [L]iquidity normally increases 1% within three weeks of the initial financial shock. (Liquidity is defined here as the ratio of nonborrowed reserves...to required reserves.) (Giordano 1987, pp. 1-3)

This action by the Fed ‘has always proven sufficient to contain the crisis and generate a relatively quick drop in short-term interest rates” (ibid., p. 3). Stage I lasts two or three weeks, after which the

Fed ‘stops pumping liquidity into the system and begins watching for crisis fallout’ (ibid., p. 4). During Stage II, liquidity actually falls and interest rates begin to rise. If necessary, the Fed then moves into Stage III, ‘the recession containment period. This stage is marked by accelerated liquidity provision and interest rate declines’ (ibid., p. 6). After the stock market

crash of October

1987, the Fed moved

quickly to Stage I to prevent the crisis from spreading. The Fed announced ‘its readiness to serve as a source of liquidity to support the economic and financial system’ (ibid., p. 7). The Fed then ‘pumped in more liquidity than in Stage I of other crises’ (ibid., p. 7). Interest rates fell, and the Fed moved quickly into Stage II. Liquidity began to fall and interest rates rose slightly. However, the Fed appears to have adopted a monetary policy which is looser than it characteristically adopts during Stage II, and ‘Stage III could be a long time in coming, and may not come at all without interest rates first rising again or other shoes dropping’ (ibid., p. 10). In summary, when a financial crisis occurs, the Fed provides liquidity to prevent the crisis from spreading. Easy money policy is not sufficient to stop a recession (in five of the crises examined by Giordano, the recession deepened even after the Fed intervention), but none of these recessions escalated into a full scale debt deflation. Apparently, quick Fed intervention is sufficient to prevent a recession from becoming a depression through debt repudiation. As discussed above, the large countercyclical deficits probably also helped to stabilize the economy. This is consistent with the view of Goodhart (1989b), who argues that central banks have always used interest rate targets rather than quantity targets. The reason, according to Goodhart, is quite simple: the primary purpose of the central bank is to protect the

202

financial system. This requires that debt deflation be prevented, which necessitates that banks avoid asset sales when there is a run to liquidity. The central bank does this by ensuring that banks can obtain reserves - but at a price. As argued in Chapter 2 above, this function of the central bank becomes normal practice only after it recognizes that it is a central bank. For example, the Bank of England did use quantity targets and would implement tight money to stop a drain of reserves, and thereby actually accentuated crises. It is only after the central bank realizes that it is a central bank

that it will behave in the manner described by Goodhart. A central bank which operates as a central bank must eventually supply reserves if the integrity of the system is threatened. As discussed in Chapter 2, the necessity of lender of last resort operations is actually increased as the financial system develops a monoreserve structure.

However,

the central

bank

can,

and

does, use

tight

money policy which stretches liquidity until a crisis is induced. At this point, the central bank must accommodate so that it ultimately cannot constrain reserves.

Formal empirical tests As argued above, econometric tests which take the supply of money to be exogenously determined are likely to be biased because the money supply is unlikely to be statistically exogenous with respect to money demand and other macroeconomic variables. There are several types of empirical tests which can be used to test for the endogeneity of the money supply, but the results of these may also be open to question. These ‘causality’ tests, as Moore (1988b) argues, can never be taken as proof of causality.’ In the final analysis, only theory can determine whether a variable can be taken as exogenous. Two types of causality tests are, first, tests to determine whether causality runs from reserves to money, and second, tests which attempt to determine whether causality runs from money to spending. If the endogenous approach is correct, money should cause reserves, while spending should cause money.

Money causes reserves

In orthodox theory, an empirical test of the stability of the money

253

multiplier would test an equation of the sort:

(1)

M = mR,

where M is money supply, m is the money multiplier, and R is reserves.

Alternatively, the base could be used in place of reserves,

or deposits could be used in place of the money supply. The money multiplier is a function of the required reserve ratio, public preferences regarding demand deposits versus cash and time deposits, and bank preferences regarding excess reserves. AS reserves are supposedly under control of the Fed, causation runs from reserves to the money supply. This equation, however, represents an equilibrium relationship, or a reduced form. If the endogenous money approach is correct, then causation actually runs from the money supply to reserves:

R = (1/m)M.

(2)

Thus, as Moore argues: In spite of the fact that, ex post, the multiplier formulation will always appear stable, there is in actuality no simple direct causal link between the supply of nonborrowed reserves and the money stock. The causal dynamics of monetary control actually works through borrowed and excess reserves and short-term interest rates. Yet these variables are entirely suppressed in the multiplier (reduced form) relationship. (Moore 1984, p. 107)

Several studies have used Granger causality tests to determine whether causation runs from reserves to money supply, or vice versa. Feige and McGee (1977) test whether reserves cause money before and after the imposition of lagged reserve accounting in 1968. They find that prior to 1968, ‘the relationship between reserves and money was instantaneous’, and the ‘possibility of one-way causality from M to R should not be dismissed’ (Feige and McGee 1977, p. 547). After the imposition of lagged reserve accounting, the ‘money supply is exogenous with respect to total reserves’, and ‘money causes reserves’ (ibid., p. 547). The evidence indicates that ‘the monetary authority "leans with" rather than "against the wind" (ibid., p. 547). They conclude that their findings cast serious doubt on the Fed’s ability to control reserves and the money supply, and on any empirical work which estimates money supply equations on the basis of exogenous reserves.

254

Lombra and Torto test whether ‘changes in the monetary base can result from "defensive' OMO (open market operations)’ (Lombra and Torto 1973, p. 52). They review several empirical studies, which find that between 70 and 90 per cent of open market operations were ‘defensive’ in nature (ibid. p. 50). They regressed the change in the Fed’s holding of government securities on the changes in sources of the base (gold stock, float, treasury currency outstanding, minus treasury cash holdings, treasury deposits, foreign deposits, and other deposits and accounts), changes in loans to member banks, changes in required reserves, and changes in currency. They find that the evidence supports the hypothesis that the Fed’s behavior is primarily defensive. They also regress changes in the monetary base on a derived variable which represents changes in the base due to Fed defensive behavior. They find that 91 per cent of the variation in the monetary base is explained by defensive operations. Moore reports that the results of several of his empirical studies show that causation runs from the money supply to reserves or the base (Moore 1985). Using four different tests of causation for the period 1973-81, he found ‘the evidence overwhelmingly supported the position that...unidirectional causality ran from each of the four different monetary aggregates to the monetary base...’ (Moore 1985, p. 17). In another study, Moore analyzed the relation among changes in reserves, in the base, and in Fed open market transactions for the period October 1979 to December 1983 (when the Fed was ostensibly targeting monetary aggregates). He found that the Fed’s open market operations explained none of the change in total reserves, and only 5 per cent of the change in the base over the period. Indeed, the coefficients imply that the Fed must engage in $1000 of open market operations to change the

base by $1 (Moore 1988b, p. 98). Moore also regressed the base on various macroeconomic variables, such as the unemployment rate, the foreign exchange rate, the three month treasury bill rate, the wholesale price index,

average hourly earnings, average weekly compensation, and GNP. He found that wages are the most important explanatory variable for the base, implying that the base cannot be treated as an exogenous variable (Moore 1988b, p. 106). In still another analysis, Moore employed four different causality tests to analyze the relation among loans, monetary aggregates, and the base for the years 1974-1980. He found that in all but one test, causation ran from the monetary aggregates to the base (ibid, p. 163). 255

Forman, Groves, and Eichner (1985) also attempted to determine whether reserves are exogenous or endogenous. Like Lombra and Torto,

they

tested

whether

the

Fed

provides

reserves

in

an

accommodative manner. They regressed changes in the Fed’s holdings of government securities on three variables: the net change in the Fed’s assets, change in currency held by the public, and the change in required reserves. A decline in the Fed’s assets, a rise in currency held by the public, and a rise in required reserves will increase the banking system’s need for reserves. If the Fed provides reserves through an open market purchase in response to one of these changes, then the Fed is accommodating the system’s need for reserves. The authors used quarterly data for the period 1953-1978. They found ‘for the period as a whole, approximately four-fifths of the quarterly change in the Fed’s holdings of government securities is accounted for by these three explanatory variables’ (Forman, et. al., 1985, p. 33). Baghestani and Mott argue that even if the Fed could control reserves, this won’t control the money supply because banks change behavior in order either to economize on reserves in the case of tight money policy, or to hold excess reserves in the case of easy money policy. They show the switch to a reserves operating procedure, which has since been

modified

but

not

completely

abandoned,

has

failed

to

overcome the short-run endogeneity or demand-determination of money which the Fed had experienced prior to October 1979 because the banking system has changed its behavior with respect to the link between reserves and money supply under the newer procedure (Baghestani and Mott 1988, p. 485).

They found any federal reserves as instrument experiment

that before 1979, ‘changes in the demand for money at funds rate were automatically met by increases in necessary for the Fed to maintain its federal funds rate setting’ (ibid. p. 487). During the tight money of 1979-82, however, the Fed no longer supplied

reserves on demand. Instead, an increase in demand for reserves

was allowed to cause the Fed funds rate to rise. This change in operating procedure led to greater volatility of interest rates and ‘created instability in the link between money supply and reserves and therefore greater volatility in the short-run fluctuations of M1 growth’ (ibid., p. 488). After 1982, the Fed targeted borrowed reserves, which ‘leads the Fed to supply reserves in response to changes in the demand for money in a manner similar to that of

256

maintaining a federal funds rate setting’ (ibid., p. 490). Thus, the Fed abandoned the experiment in Monetarism and returned to the accommodative policy of supplying reserves on demand.’ As a result, greater stability of interest rates returned. If the findings of the various studies cited above are correct, then reserves are endogenously determined. This finding is consistent with the endogenous money approach: banks finance positions in assets, then obtain the reserves required. If the banks are unable

to find reserves through private sources, then the Fed must provide reserves. In this case, the Fed’s behavior is also endogenously determined. Not only do these findings lend support to the endogenous money approach, but they also cast doubt on any empirical work which starts with the assumption of exogenous reserves. Spending causes money

According to the quantity theory of money, causation runs from a change in money to a change in spending and income. The money supply is taken as exogenously determined. Monetarists find a close correlation between changes in the quantity of money and changes in nominal income. The relationship can be represented as:

M = kPY,

(3)

where M is exogenously given, k is constant (or at least stable) and represents the inverse of the income velocity of money, P is the aggregate price level, and Y is real income. While Friedman admits that there is some possible feedback from a change in nominal income to a change in money, he argues that the primary influence is from money to nominal income (Friedman 1982b, p. 32). Of course, simply showing that changes in money are correlated with changes in nominal income proves nothing regarding causation. Even if it is shown that changes in money precede changes in nominal income, this still says nothing about causation. For example, if banks meet loan demand by issuing demand deposits, which are then drawn down as firms make planned expenditures, then it is expected expenditure which determines the

money supply.’

In 1972, Sims demonstrated that ‘the money stock could be taken as exogenous in GNP on money-stock distributed-lag regressions...’ (Sims 1980, p. 251). Using formal causality tests, he found that

257

money was Granger-causally prior to income, and that there was no evidence of feedback from income to money (that is, a change in money causes a change in income, but a change in income does not cause a change in money). However, in 1980, he found that if

other variables were included in the tests, money was no longer strongly prior. Specifically, when four variables (interest rate, price level, money stock, and industrial production) are included, ‘only

4 per cent of the variance of production is accounted for by money innovations’ (Sims 1980, p. 252). He concluded: Thus...some of the observed comovements of industrial production and money stock are attributed to common responses to surprise changes in the interest rate. With this shift in attribution, surprise changes in the money stock are left with a very small role in explaining production variance in the postwar period. (ibid., p. 253) The observed responses of money stock to the interest shocks could simply be the tail following the dog: non-monetary economic developments raise interest rates, then push production down; and the demand for money declines smoothly in response.... (ibid., p. 256)

Sims (1983) also tested the rational expectations hypothesis that unanticipated changes in the money supply cause changes in real output. Again, he found that money is not causally prior if interest rates are included in the tests. In fact, he argued that there is little evidence to suggest that unanticipated changes in the money supply cause changes in real output: Despite the need for further work in this direction, the weight of evidence against the rational expectations monetarist view of the origin of the business cycle seems quite strong. There is little consistency across countries or time periods in the relation of money volatility or of money surprises to production. In all countries, money stock and output move smoothly together, money slightly in the lead, after interest rate surprises. That it is precisely these smooth, predictable joint movements in production and money which account for most of the correlation of those two variables seems to contradict the rational expectations monetarist position. (Sims 1983, p. 232)

Moore reviewed several empirical studies and found ‘in both the United States and the United Kingdom, the evidence indicates that

258

annual changes in the volume of bank intermediation are determined primarily by the quantity of bank lending...’ (Moore 1985, p. 19). Thus, loans determine

deposits. He also reviewed

several studies of the determinants of bank lending. His hypothesis is that banks have little control over the quantity of loans extended, since unused credit lines exceed the total quantity of demand deposits. Thus, loans are demand determined. He regressed bank

loans to commercial and industrial corporations (CICs) on variables such as the wage bill, materials costs, corporate tax payments, and stockbuilding costs. His findings include: This equation succeeded in explaining from one-half to two-thirds of the total variation in commercial and industrial borrowing in the United States and the United Kingdom countries.... These coefficients suggest that bank borrowing by CICs increases typically on a one-for- one basis with their requirements for additional working-capital.... (Moore 1985, p. 22) In a later study, Moore again used four different causality tests to determine whether bank lending ‘causes’ monetary aggregates. He found that the tests strongly support the hypothesis that causality tuns from loans to the monetary aggregates (Moore 1988b, p. 163). In conclusion, there are many empirical tests that have confirmed the ‘reverse causation’ arguments that money causes reserves and spending causes money. I have not attempted to present summaries of those studies which confirm the orthodox notions that reserves cause money and money causes spending. Admittedly, these are numerous, and are probably familiar to most readers. However, as I have argued above, simple empirical tests cannot be used as the deciding factor in determining exogeneity.

Money and the business cycle Sims’s empirical work casts doubt on the Monetarist theory of the business cycle whereby an unanticipated shock to the money supply temporarily causes real effects on spending, while anticipated shocks only raise the price level. In fact, an analysis of the business cycle shows that the endogenous approach to money is consistent with many of the features of a stylized cycle. Several analyses of the role that money plays in the business cycle are now summarized. Earlier in this chapter, I presented an analysis of recent 259

experience during the experiment with Monetarism, which brought the financial system to the brink of collapse. The Fed was forced to intervene to alleviate the crisis with lower discount rates, greater

open market purchases, and intervention as lender of last resort. Of course, one failed experiment does not prove the case. Wolfson (1986) examines each of the recent ‘credit crunches’ (1966, 1970, 1974, 1980, and 1982) to develop a theory which broadly explains financial crises which have occurred since 1966. He summarizes the model as follows: 1. Because of increased reliance on debt during the expansion phase of the business cycle, the corporations’ financial condition increasingly deteriorated. 2. As profits declined, the corporations experienced increasing difficulty in meeting debt payment requirements and other fixed commitments such as those due to involuntary investment. 3. Corporations borrowed to try to obtain the funds they needed to meet debt payment requirements and involuntary investment commitments; they relied primarily upon bank loans, and, for large corporations, also on commercial paper borrowing. 4. The banks suffered losses on their business loans, which adversely affected their own financial condition, and were restricted by tight monetary policy; they tightened their lending policies, but at the same time attempted to meet the necessitous loan demands from their long-standing customers. 5. They did so by decreasing the growth rate of nonbusiness loans in relation to business loans, by restricting the business loans they extended to new customers...and by relying on purchased funds, particularly large negotiable certificates of deposit. 6. A surprise event either the sudden imposition of an institutional constraint or an unexpected bankruptcy (or the threat of one) - disrupted these financing patterns. 7. The disruption of financing patterns initiated a financial crisis - a sudden, intense demand for money. 8. The urgent demand for money came either from corporations or

banks

whose

sources

of funds

were

disrupted,

or

from

investors who feared the loss of their funds. 9. The immediate crisis was resolved by lender-of-last-resort operations of the central bank, the Federal Reserve Board. (Wolfson 1986, p. 131)

Wolfson’s model is generally consistent with the endogenous money approach. Money is created as banks make loans. During an expansion, firms rely increasingly on loans to finance positions in assets. However, money plays another important role: it is used

to retire debts. If income flows become insufficient to meet debt commitments, borrowing can temporarily tide one over until

income rises. However, if income flows do not eventually return to

260

a sufficient level, borrowers may not be able to continue to refinance positions in assets. Forced liquidation of assets can result in falling asset prices and bankruptcy. Because balance sheets are interlocking, failure of debtors affects the financial position of creditors. Banks are doubly affected by forced liquidation: their assets lose value and they experience a deposit drain at the same time. They then cut off credit to all but established customers. This, however, will make matters worse since unsatisfied needs for

liquidity can then only be met through further asset liquidation. According to Wolfson, the rate of profit falls just before the cyclical peak because debt payment obligations increase faster than the ability to meet payments, leaving less of the gross capital income as net profit to firms. Firms then attempt to reduce discretionary expenditures. However, many expenditures (such as those

on

overhead,

involuntary

inventory

accumulation,

debt

payments, and new plant and equipment already contracted) cannot be reduced. Thus, firms engage in necessitous borrowing. Financial institutions

try to meet

the needs of established

customers,

but

reduce credit to new customers. The central bank often decides to attempt to reduce inflationary pressures near the peak through tight monetary policy. In order to meet the needs of established customers, banks rely on purchased funds

(certificates

of

deposit,

Eurodollars,

Fed

funds,

and

repurchase agreements) with lower reserve requirements but higher costs. These funds are also uninsured and therefore more volatile. The financial system therefore becomes vulnerable to a surprise event. For example, if a large firm experiences difficulty meeting commitments, its banker must provide the credit. However, the

banker must rely on purchased funds, the supplies of which can quickly dry up if confidence in the bank’s assets is shaken. A financial crisis can therefore occur due to the failure of a large institution, or if the central bank applies an institutional constraint. A financial crisis is manifested as a sudden, intense demand

for

liquidity (not necessarily credit) - an attempt to liquidate assets in order to obtain liquidity. In normal times, credit functions as money, but in a crisis only the medium of exchange and means of payment are accepted. ‘A crisis involves the forced sale of assets, intensified efforts to borrow, a disruption of financial markets...this process is abrupt, not gradual...it often involves an element of panic’ (Wolfson 1986, p. 130). The crisis is resolved when the intense demand for liquidity is relieved - usually through central bank intervention as a lender of last resort.

261

Wolfson’s account is consistent with that of Giordano (1987) discussed above. The Fed intervenes to halt the spread of the crisis by providing liquidity in Stage I to relieve the intense demand for liquidity. As the recession continues, the Fed moves into Stage II], during which it continues to provide liquidity needed to prevent attempts at liquidation of assets. Because of the Fed’s intervention, the recession does not generate a debt deflation. Wojnilower (1980), in his exhaustive survey of recent business cycles, also presents a theory of the cycle which is consistent with the endogenous money approach. He argues that banks use liability management to accommodate the demand for loans. As an expansion proceeds, the demand for loans to finance investment rises quickly, but banks are able to meet the demand. Even when the central bank uses tight money policy to raise interest rates, the demand for loans continues to rise because it is interest-inelastic. As banks are forced to pay more for reserves, they simply raise the loan rate of interest, but continue to meet the inelastic demand.

Toward the peak, the system becomes extremely unstable, growing out of control. However, the rising price of credit will not hamper growth, and neither are banks quantity constrained. Eventually a credit crunch occurs which abruptly restricts the expansion of credit and results in a financial crisis: “cyclically significant retardations or reductions in credit and aggregate demand occur only when there is an interruption in the supply of credit - a “credit crunch" (Wojnilower 1980, p. 278). Credit crunches are caused by ‘regulatory rigidities, such as ceilings on the interest rate, or the emergence of serious default problems in major institutions or markets’ (ibid., p. 278). After each credit crunch, the central bank and private financial institutions ‘reshaped the financial structure so as to prevent the recurrence of that particular form of credit supply interruption’ (ibid., p. 278). Thus, the following expansion is able to proceed for a longer period, and to reach a higher level before another credit crunch occurs. Wojnilower’s analysis differs from Wolfson’s (Wojnilower emphasizes the ‘crunch’, while Wolfson emphasizes the ‘intense demand for liquidity’), and he argues ‘the growth of credit is therefore essentially supply-determined - if not always, then at least at those times that are cyclically important...’ (Wojnilower 1980, p. 277). He argues that because the demand for credit is normally interest-inelastic, the supply of credit determines the quantity issued. However, he argues that institutional arrangements such as credit lines, creation of bank holding companies to issue

262

unregulated commercial paper, Eurodollar borrowings, and the development of the certificate of deposit market all helped to ensure that credit could be supplied on demand. Thus, his approach

approach.”

is generally

consistent

with

the

endogenous

money

The endogenous money approach is also important in Minsky’s (1985, 1986a, 1986b) view of the cycle, as discussed above. Briefly, according to Minsky’s financial instability hypothesis, an expansion sets off a speculative boom which generates financial fragility, setting the stage for a financial crisis. In an expansion, firms increase debt to finance investment. As investment rises, aggregate profit levels rise and raise confidence. Over a string of good years, firms are encouraged to take on greater debt relative to expected income flows and to increase the quantity of short term debt relative to long term debt. At first the supply of credit is fairly elastic, rising to meet demand. Eventually, however, higher interest rates are required to induce financial institutions to provide credit - unless a new innovation temporarily postpones the interest rate hike. As interest rates rise, the cash commitments of firms (which are

rolling-over debt, or which have floating rate loans) increase. Some firms may be unable to meet commitments on the basis of income flows, so are forced to liquidate assets, reduce discretionary spending, and refinance debt. Liquidation reduces the value of assets, causing technical insolvency for some firms - which increases the unwillingness of banks to provide credit. The reduction of discretionary spending reduces aggregate demand (or at least its rate of growth), which further reduces profit flows to firms and increases problems experienced in meeting cash commitments. Due to the interlocking nature of balance sheets in a developed capitalist economy, failure by a few large firms leads to technical insolvency of their banks and can rapidly degenerate into a financial

crisis

as

a

snowball

of defaults

results.

Thus,

the

expansion leads to the development over time of liability structures which can not be validated. A recession results which deflates debt and leads to a revaluation of acceptable liability structures.

Conclusion

Is the endogenous money view correct? It is difficult to use empirical evidence to determine whether it is because simple

263

empirical tests are likely to be biased. However, various pieces of evidence were summarized which are consistent with the endogenous money approach. Bankers and central bankers often behave as if the money supply is endogenously determined. Bank balance sheets (and off-balance sheet commitments) become increasingly complicated in ways which appear to support the hypothesis that they are not constrained by reserves supplied by the Fed. The behavior of credit over the business cycle is consistent with the endogenous money approach. The way the Fed behaves during financial crises is also consistent with the hypothesis that the Fed recognizes that maintenance of stable financial markets requires that banks cannot become liquidity constrained. The absence of depressions and financial breakdowns during the last four decades seems to suggest that the Fed does know how to prevent a financial crisis from becoming a generalized debt repudiation. While the Fed does appear occasionally to lapse into a Monetarist stupor, so far it has always been able to regain its wits in time to prevent wide scale debt deflation.

Notes 1. For example, Visa and Mastercard guarantee the credit of consumers so that stores will accept consumer debt. In the absence of such credit cards, much of this business would have gone through banks. 2. According to the quantity theory of money, the velocity of money is stable and the economy naturally tends toward full employment. Thus, changes in the quantity of money cause changes in the price level. In the equation of exchange, MV=PQ, where M is the money supply, V is the velocity of money, P is the price level, and Q is real output, if Q tends to grow at some fixed rate (say, 2.5 per cent) and if V is fixed (or grows at some constant rate), then changes in the quantity of money should be closely associated with changes in the price level (inflation). The following table shows that recent data are not consistent with the quantity theory of money and that the 1980s have been unkind to Monetarists. It is apparent that changes in M1 are not closely correlated with inflation. Furthermore,

neither the

growth of GNP nor the velocity of money appear to be stable. Of course, none of this implies anything about causation.

264

Year 1979 1980 1981 1982 1983 1984 1985 1986 1987 OO eer 025 1906-2 5953.65068 930.129 2.9 CP ges 8 0559 1 GA. 39 6 13,7453.2 992.66 43.0 OM eile 0 5en OA we 8 Oa 80 ais ol? 0165.03.1 GViee sues) 2) 4.72.0 247 9-63 9-110: 28

Notes: dQ is rate of growth of real gross national product, dP is the rate of change of the GNP deflator, and dM is the rate of growth of M1. These data were obtained from The Economic Report of the President, February 1988. dV is derived from the following calculation: dV = dQ+dP-dM. All data are from the Economic Report of the President, 1988.

Of course, Monetarists might argue that lags exist between changes in money and changes in prices. Thus, the inflation in 1979-81 was a result of monetary policy during the 1970s. Tight monetary policy during 1979-81 then caused inflation to subside during the mid 1980s. However, it is difficult to see why inflation has remained very low from 1986, since the money supply has been growing rapidly since 1983 (with the exception of 1984, during which it grew at a rate of ‘only’ 5.8 per cent). See Minsky 1986b for an analysis of the role deficits can play in allowing an expansion. Interestingly, each of these crises occurred during a recession, with the exception of the Continental Illinois Bank run. Thus, five of the crises occurred after the Fed had already begun to ease policy. ‘Because the recessions were far from completion when the crisis hit, however, it was the financial shock and subsequent tremors that

moved the Fed to ultimately accelerate liquidity injections to prevent the crisis from turning the recession into something worse’ (Giordano 1987, p. 6). While the Fed’s action was sufficient to contain the financial crisis, ‘it is probably fair to say that the financial shocks aggravated the slump to some degree’, since the recessions ‘deepened following all of these financial crises’ (ibid., p. 6). Moore warns that ‘causality tests’ are not tests of ‘causality’. The Granger-Sims test simply determines whether an optimal prediction of variable x can be improved by inclusion of variable y, and whether an optimal prediction of y can be improved by inclusion of variable x. If it turns out that inclusion of y helps in predicting x, but x does not help in predicting y, we say that y ‘causes’ x. (Moore 1988b, p. 151 ate Fed normally supplies reserves essentially on demand, but Baghestani and Mott have shown that the Fed can use quantity constraints within some range - that is - until it induces crisis. When this happens, the Fed must abandon quantity constraints and return to accommodative policy. This is consistent with the endogenous money approach developed in previous chapters, although it is not consistent with Moore’s assumption that the Fed passively supplies reserves on demand.

265

9. See Tobin (1987, p. 497). 10. However, Wojnilower’s analysis implies an upward sloping money supply curve, so would not be consistent with Moore’s endogenous money approach, in which the money supply curve is horizontal.

266

9

THE BALANCE TO MONEY

SHEET APPROACH

Introduction

In this chapter, some simple models are used to summarize and clarify the endogenous approach to money. Specifically, I will show that it is possible for banks to advance credit to cover spending and then to recapture deposits. Thus, banks need not passively wait for deposits which create excess reserves. I will also show, using a simple model, that credit can finance investment expenditures and that these generate sufficient aggregate profits to enable the debt to be retired. Thus, it is not necessary for the central bank to provide excess reserves in order for banks to make loans. This chapter clarifies what velocity means in a model of money which is endogenously created, and shows the role that deficit spending plays in allowing economic growth.

A simple model In this section, simple balance sheets are used to demonstrate some of the propositions concerning an endogenous money supply. Minsky (1986b) has characterized money as simply a balance sheet operation. Following his suggestion, I will show how spending can be financed through expansion of the balance sheets of the banking

system.’ Assume that the public may hold its wealth in one of five forms: goods and services (GS), demand deposits (DD), government bonds (GB), time deposits (TD), and long term private securities (PB).? The central bank requires that commercial banks hold required reserves of ten per cent against demand deposits, but no reserves against time deposits. In this simple model, there is no currency, but it will be incorporated within the model below.’ This model assumes that the public essentially has a ‘stock’ of services which it sells to firms. This stock is assumed to be supplied on demand. The public merely trades its services for a demand deposit. When such a sale occurs, I will only look at the end of the

267

period to determine the quantity of demand deposits the public has accumulated, which equals the value of the services sold. This approach is taken so that I will not need to present income statements

in addition to balance

sheet statements.

Thus, I will

only present the end-of-period balance sheet for each sector. For the public, the end-of-period demand deposits held represent the accumulated income (‘wages’) received for the sale of a stream of labor services. When the public spends these demand deposits on consumption goods (in a section below), I again merely focus on the balance sheet statement for the end of the period. Thus, the

worker now holds consumption goods rather than demand deposits. It will be seen that if part of the public provides services to the investment goods sector, then the aggregate value of the demand deposits held by the public is less than the aggregate income. That is, profits are realized, and are equivalent to investment. This is consistent with Kalecki’s model, and also with Marx’s model

in

which surplus value is generated in production. I will also measure profits as the accumulated end-of-period demand deposits held by firms. For the purposes of these simple models, I believe that this approach is adequate. Assume that commercial

banks are free to hold reserves (RR),

short term and long term private securities (CP, PB), and government debt (GB) as assets. Assume that banks do not hold excess reserves, and that all reserves are held at the central bank

(Fed). In the simple model, net worth will be assumed to be zero initially for all economic units except the public.* Assume that only five economic sectors exist: government, central bank, commercial

banks, public, and firms. The public will consist of workers whose marginal propensity to consume out of income is one, while firms consist of capitalists, whose marginal propensity to consume out of income is zero. These restrictions can be dropped as the model is extended, but are retained here to simplify the analysis. I will initially assume that capital asset holdings are zero, but will later introduce investment and capital accumulation. Balance sheets for each sector can be formulated as follows: Commercial

Bank

Federal Reserve

Asset

Liability

Asset

Liability

RR GB

1D) h “TRIBY

GB

VRIES ‘DDE

PB

*

(GIP

4

*

268

Government

Firms

Asset

Liability

Asset

Liability

Asset — Liability

DDg GS

* a

DD TD GS

188) (le *

DD TD GB

GB

Public

* NW “4 ~

*

*

PB

*

*

*

GS

*

Firms issue securities (PB, CP) in order to hold financial assets (DD, TD) or to purchase goods and services from the public and from other firms (GS). The government issues securities (GB) to purchase goods and services from the public, or to hold demand deposits at the Fed (DDg). Banks purchase financial assets issued by firms (CP, PB) or by the government (GB). They issue liabilities (TD, DD) to finance the purchase of these assets. Firms use the goods and services purchased from the public in the production process to generate a flow of income over time. If all goes well, this income flow will be sufficient to allow retirement of debt issued to purchase the goods and services in the first place, and to generate profit flows which can be accumulated as net worth. Starting from a point of equilibrium, assume that firms decide to purchase some services (GS) from the public by issuing short term

debt to commercial banks.’ Commercial Bank

Firm

Asset

Asset

Liability

aC\P

SIDI)

Z

Public Liability

Asset

Liability

eed

-GS DID)

*NW *

SHES

bi

Because banks must hold required reserves equal to 10 per cent of demand deposits, this cannot be a position of equilibrium. However, banks can return to equilibrium by inducing the public to exchange its holding of demand deposits for interest earning time deposits (which do not have a legal reserve requirement). This may require an increase in the interest rate on time deposits to induce the change in the portfolio of the public. Thus, the following adjustment could move banks to a position of equilibrium:

269

Commercial Bank

Firm

Asset

Asset

Liability

(GP Ss

dd) )

Ce

Public

Liability

Asset

@?

=D SOND)

+ 4

Liability

Daw

Generally, firms prefer to issue longer term liabilities if the assets they hold are long term. Thus, the firm in this example would like to sell a long term security (PB) so that it can retire the short term security it originally issued (CP). Thus: Commercial Bank

Firm

Asset

Asset

-CP

Liability

Se -ED *

In this example,

GS

Public Liability

“27 *

-=CP +PB

the public is induced

Asset

Liability

=D) erls}

NW

to draw down

its time

deposit in order to purchase the long term securities issued by the firm. Of course, the firm must be willing to pay a higher interest rate than the banking system pays on time deposits, if the public prefers time deposits, and/or incurs a penalty when it shifts funds from time deposits to private securities. Furthermore, this example has ignored the intermediate step, in which the public first exchanges time deposits for demand deposits and uses these to purchase the long term securities so that the firm is able to retire the short term securities. If this process is not instantaneous, the banking system will temporarily experience a rise of required reserves. If reserve requirements are calculated on the basis of average holdings of demand deposits, then the banking system will be forced to increase its reserves. This will be discussed below. These simple balance sheets have shown that firms are able to expand expenditures by first issuing short term liabilities. If the banking system is willing to purchase these liabilities, it can issue demand deposits to the firms, which are then transferred to the public in exchange for services rendered. Equilibrium for the banking system is restored when the public is induced to exchange its demand deposits for time deposts. Firms then retire their short term debt on the basis of a new issue of long term debt. In the example above, the public bought the long term debt, but the results would not be significantly different if the long term debt were sold to banks.

270

This example shows that it is possible for credit creation to be the basis for expenditure. No exogenously supplied money was needed in these examples to allow expenditures to proceed and balance sheets to expand. However, I have ignored required reserves, to which I will now turn.

Reserve requirements Reserves are required on demand deposits which are temporarily created before the banking system is able to induce the public to exchange demand deposits for time deposits. Actually, reserves are calculated on the basis of average demand deposits (and short term business saving deposits) held over the relevant period. The actual quantity of reserves required would thus be a function of: 1) the reserve requirement ratio; 2) the quantity of demand deposits created; and 3) the average length of time the demand deposits exist. The faster banks can induce the public to give up demand deposits, the lower the quantity of reserves which will be required. In the examples examined above, required reserves will increase to some extent due to the rise in demand deposits temporarily issued. As has been argued extensively in earlier chapters, banks can meet such reserve requirements by borrowing at the discount window, by purchasing excess reserves in the Fed funds market, by borrowing from overseas banks, or through various reserve economizing activities.

Simple numerical example Retaining all the assumptions above, assume the relevant period over which reserve requirements are calculated is four (S-day) weeks. Assume that a firm wishes to make a $1000 purchase of services from the public, initially financed as above by an issue of short term liabilities of $1000 to the banking system. Commercial Bank Asset

Liability

+1000 CP * +1000 DD 2

Firm

Public Liability

Asset

+1000 GS * +1000 CP *

Zi.

Asset

Liability

-1000GS * NW +1000 DD *

Assume the firm holds a demand deposit for one day before it is able to purchase the services from the public, and that it takes four

days for the banks to induce the public to give up demand deposits and substitute time deposits for them. Assuming a 20 day (four five-day weeks) period, required reserves would rise by (0.10)(0.25)(1000) = $25. If it is further assumed that firms sell long term securities in order to retire the short term securities, and that

this process requires that the public exchange time deposits for demand

deposits, which are then transferred to firms so that on

average, demand deposits of $1000 are held for an additional day, then required reserves will rise by (0.10)(0.30)(1000) = $30. Thus, the final balance sheets will look as follows: Commercial Bank

Firm

Asset

Asset

Liability

+30 RR* +30 BR be ;

+1000 GS

Public

Liability

* +1000 PB ~

Asset

-1000 GS +1000 PB

Liability

* NW *

where BR stands for reserves the commercial bank has borrowed to meet the reserves required on the basis of the demand deposits temporarily issued. Of course, interest payments have been ignored, the balance sheet of the bank excludes any increase in net worth. The bank obviously would not have undertaken this operation without receiving interest, since it must pay interest on the reserves borrowed to meet required reserves. Required reserves are similar to a tax: they increase the cost of issuing liabilities. I will allow for interest payments below. What are the conditions which would enable this process to

proceed? First, the firm must believe that purchase of the services

from the public will generate sufficient income flows to retire the debt (including interest) and provide some expected rate of profit. The bank will engage in the transactions above only where the interest payments it receives on the short term debt (CP) exceeds the interest it pays on the time deposits temporarily held by the public, plus the interest it pays on the borrowed reserves. The public sells its services if the price offered is sufficient to induce it to hold financial assets rather than real assets (capacity to provide services). It is then induced to exchange demand deposits for time deposits, and these for long term securities by successively higher interest rates. Thus, in order for the process to take place, it necessary that

212

(1) {d(TD)+f(BR)}

< r(CP)

Condition for banks

(2) P > r(CP) + r(PB) + GS

Condition for firms

(3) r(1+PB) > d(1+TD)

Condition for public

> DD > uGS

The first condition simply states that the cost of providing finance is composed of the interest paid on time deposits plus the interest paid on borrowed reserves, and must be less than the interest received on commercial paper. The second condition requires that total returns to the firm (P) must exceed total costs, which include interest payments on short term and long term securities, plus the initial outlay on services purchased. Finally, the last condition States that the interest plus principal received on the long term securities must be greater than the interest and time deposit, which must

be greater

than the demand

deposit, whose

value

to the

public must exceed the value of its services sold to firms. Note that this process cannot be initiated unless the value of the services provided by the public (GS) is perceived asymetrically by firms and by the public (so that uuGS is assumed to be nonrestrictive in these models.

A model with production and consumption In this section, the model is expanded to allow for production and consumption. I retain the assumptions above, but now assume that the firm purchases services from the public, which are used to produce goods. Some produced goods are sold to the public for consumption, while others are used as intermediate goods in the

273

production process. For simplicity, it will be assumed that two firms exist, one of which produces only intermediate goods (investment goods), while the other only produces consumption goods for sale to the public. The first firm will be called the investment sector firm, whose output will be termed investment goods (I). The second firm will be called the consumption sector firm, and its output will be consumption goods (C). For simplicity, it will be assumed that the production of each $1 of consumption goods requires an input of $0.90 in labor services, and an initial $0.10 in investment (capital) goods. Investment goods depreciate at a rate of 10 per cent per year. Assume the consumption goods firm decides to produce $1000 of consumption goods. This requires an outlay of $900 for services from the public, and $100 in intermediate goods provided by the investment sector. To simplify matters, assume the investment sector firm uses only the services of the public to produce investment goods. Commercial Bank

Investment Sector

Public

Asset

Asset

Asset

+CP 100* *

Liability

+DD

100

+GS

Liability

100

* be

+CP 100

+DD 100 -GS 100

Liability

* NW *

The consumption goods sector now purchases the investment good from the investment goods sector, issuing a long term private bond. Commercial Bank

Investment Sector

Asset

Liability

Asset

+PB 100 * +CP 900 * -CP 100 * = ie a

+DD

-1100

900

Liability *

-CP 100

*

Lata

ef Stay A= — .seassavecesens

193} i

(Note: balance sheet changes are shown above the dotted line, while final positions are shown below the line.) The firm also purchases $900 in services from the public to be used in the production of consumption goods (C), financed through the issue of short term private debt to banks. (Note: I will temporarily

274

ignore profits in the investment sector.) Consumption Sector

Public

Asset

Asset

Liability

+1100* +PB 100 +GS 900* +CP 900 ie green a

Liability

-GS 900 * +DD 900 * Bsc aera Sees

The consumption sector now produces the consumption goods, using the $900 in services purchased from the public and the $100 investment good. Only 10 per cent of the investment good is depreciated in the production process, so that $90 worth of the investment good will be left at the end of the period. The public spends its entire $1000 demand deposit (accumulated ‘wages’) on the consumption goods. (The public’s demand deposit represents a sale of $100 in services to the investment sector, and $900 in services to the consumption sector.) The consumption goods sector retires all of its short term debt, and 10 per cent of its long term debt (the term of the long term debt is assumed to equal the economic life of the investment good). Commercial Bank Asset

Consumption Sector

Public

Liability

Asset

Liability

Asset

ae

Sh

aa,

Lye s *

-GS 900 +C 1000

-CP 900* -DD 910 -PB 10 * “4 see sess a x e

* *

-C 1000 +DD 1000 -DD 910 ee ec I 90

¥ ig ‘

+NW 90

2 * -CP 900 * -PB 10 es :a ie *

Liability

+C 1000 * -DD 1000 * q bane

NW 90

*

The net result can be summarized as follows: the public has provided $1000 in services - $100 in labor to the investment sector, and $900 in labor to the consumption sector. It has used its income

(accumulated

as

a demand

deposit)

to purchase

$1000

in

consumption goods. The investment sector produced an investment good using labor services of $100, financed through the sale of short term debt to the banking system. This short term debt 1s retired when the consumption sector firm purchases the investment

212

good. This purchase is made on the basis of an issue of long term bonds to the banking system, assumed to have a term of 10 years. The consumption sector uses the investment good and $900 in labor services to produce consumption goods which are sold to the public. Upon the sale of the consumption goods, the consumption sector is able to retire all of its short term debt (issued to cover the cost of labor services) and meet its annual commitment on long term debt ($10). The consumption sector is left with a demand deposit of $90, and with $90 of the investment good (depreciation was $10). At this point, the consumption goods sector could retire the entire remaining long term debt. Alternatively, part of the demand deposit ($90) could be used to provide interest payments to banks and/or profits to the investment sector (which have been ignored). Note that the net worth of the firm has increased by $90, which is

equal to the net investment (gross investment less depreciation). In this case, net profits also equal net investment: $90. If the consumption goods firm decides to liquidate its debt (using the demand deposit to retire the long term bond), then its net profit exists only in illiquid form (the capital good). I have assumed that the consumption sector retires its short term debt of $900. Actually, there is no need for the firm to retire this debt, and it is unlikely that it would. The consumption sector firm has issued short term debt to hire labor so that the capital good it has purchased can be operated over a number of years to generate an income flow. Thus, in each subsequent period, it is likely that the short term debt will be rolled over as a revolving wages fund. As long as the firm remains creditworthy and makes interest payments on the short term debt, the bank will allow it to renew this ‘revolving fund of finance’ each period. Similarly, the investment sector firm may utilize a revolving fund of finance since it, too, is a going concern. Thus, we would expect to find, over time, a close relation among the short term debt issued by the firm (commercial paper in this case), the demand deposits issued by the bank which enable the firm to purchase services from the public, wages paid to the public, and consumption goods purchases. Since the velocity of demand deposits paid as wages is assumed to be one in my model, since the marginal propensity to consume out of

worker income is one, since workers are paid once each period and

they spend their wages once each period, and since capitalists (firms) do not consume, wages=commercial paper = demand deposits=consumption. Of course, a change in any of the

276

assumptions would change these relationships. The point is that the production of an investment good has resulted in income which is not spent on consumption goods, just as it has resulted in the production of goods which are not consumed directly. Investment goods are consumed indirectly, and only gradually as they depreciate. Production of investment goods allows firms to accumulate these non-consumable goods to be used in further production. At the same time, production of investment goods produces income (accumulated in demand deposits) which is not spent on consumption. This income is then available as gross profits (or ‘saving’), which can be used to retire the debt, or to engage in further production. Although I have not referred explicitly to time in these models, the results would not be significantly different if time were incorporated into the analysis. In a model with time, wages as a stream of income would not be equivalent to the accumulation of demand deposits, profit income would not be equivalent to demand deposits accumulated by firms, and so on. In other words, a smaller amount of ‘money’ (demand deposits) can support greater flows of wage and profit income. Adding a time dimension allows velocity to diverge from one. Once time has been admitted, balance sheets

differ from income accounts in these models. Clearly, credit expansion which allows production of investment goods to proceed generates the funds which are necessary to allow retirement of the debt. Prior accumulation of funds (saving) which are then intermediated through banks is unnecessary for the investment process. Every expenditure of a demand deposit which

has been created through expansion of a bank’s balance sheet results in a receipt. Neither is exogenous money necessary, since endogenously created credit is sufficient to allow production to increase. Thus, in an expansion, simple expansion of private balance sheets suffices. Below I will discuss what happens when liquidity preference rises. Briefly, because the expansion has occurred on the basis of credit, if economic units now decide they do not want to hold privately issued credit, the system will collapse. For example, returning to the last balance sheet, assume that the bank decides to ‘liquidate’ the long term bond in order to increase cash reserves. The only ‘cash’ which exists in the system is the demand deposit liability of the bank. Thus, attempting to sell the long term bond will result in a run on bank deposits. There are two possible constraints to growth in this model:

277

finance and labor services.° I have assumed an infinite supply of labor services, so this is not an operative constraint. The other constraint is finance, which is limited only by the willingness of economic agents to accept private liabilities. Neither saving nor exogenously supplied money is necessary. Certainly exogenous money (for example, dropped from helicopters) would also enable the system to expand. However, saving would not. As Terzi (1986-87) has shown, prior saving cannot be the basis of an expansion in an integrated market economy. This is due to the obvious fact that in an economy where production is for the market, income and expenditure must be equal. Any attempt to save income must mean a decline in expenditures, and therefore,

in income. If spending is income constrained, then it cannot grow. Only in a non-market economy can saving be a source of further expansion. In an economy in which a large portion of production is not for the market, thus, where monetary production is not the basis of the economy, saving can be the basis for investment. This

‘primitive accumulation’ can occur because balance sheets are not highly interconnected. A farmer can reduce the portion of his produce which he consumes in order to increase his stock of seed corn so that he can produce more in subsequent years. Thus, in a non-market economy, reduction of consumption can allow expansion. However, as production increasingly occurs not to directly satisfy consumption needs, but to be sold on the market, saving becomes less important as a source of investment. In fact, once a monetary economy develops and saving takes the form of monetary savings, saving cannot be a net source of funds for investment.’ However, as will be discussed below, helicopter drops

of exogenous money can be accumulated without invoking the ‘paradox of thrift’.

as hoards of saving,

Extensions of this model

Several simplifications have been made above to facilitate analysis. For example, reserve requirements have been ignored. As argued above, the quantity of reserves required will depend on the length of time needed for each of the transactions involving demand deposits. Banks will have to obtain reserves or engage in reserve economizing liability management to meet legal reserve requirements. Second, I have ignored a) leakages of demand deposits into 278

currency, and b) saving out of ‘wages’. Leakages into currency reduce bank reserves, so increase the quantity of reserves banks will have to obtain. For example, if the consumption sector firm in the final step decides to exchange its $90 in demand deposits for currency, the banking system will have to come up with $90 in reserves. However, the effects on the system are much greater if the public chooses to hoard some of its income. If the public decides to ‘save’ $90 of its income, then consumption expenditures are only $910. Now the consumption sector firm does not receive any profits. That is, it can just cover required payments on short term debt ($900) and long term debt ($10) using its accumulated demand deposits. If the public holds its saving in the banking system in the form of demand deposits, then the banking system is not affected: the $90 in demand deposits are held by the public rather than by consumption sector firms. Thus, these models confirm Keynes’s claim that it is irrelevant to the banking system whether the public consumes or saves - that is, that consumption is as much a source of loanable funds as is saving. However, if the

public holds currency, then the banking system is again deficient in reserves. Thus, Keynes’s proposition that only hoarding, and not consumption, can reduce the quantity of funds available to banks is correct. If saving exceeds $90, then the consumption sector firm is unable to cover production costs (wages plus payment on long term debt). This would force the firm to roll over debt commitments, which might affect production plans. Third, I have assumed that the term of the long term debt equals the life of the investment good. Of course, this is not necessarily the case. However, aggregate ‘profits’ (accumulated as demand deposits) are sufficient to retire the entire long term debt at the end of the first period. Clearly, the term of the private bond is irrelevant. The point is that sufficient income is generated to retire all debt. However, if the model were expanded to include a large number of firms, income may not accrue to the firm which has issued the debt. The firm which has purchased the investment good will likely only capture a portion of the aggregate ‘profits’, so is not actually able to retire all the debt in the first period. This firm expects to use the investment good in production over a number of years, receiving a portion of the aggregate profits each year so that the debt can gradually be retired. However, in the aggregate, profits do equal the net credit outstanding, so that all debts

could

be

retired.

As

mentioned

above,

this

is neither

necessary nor likely since firms will probably maintain a ‘revolving 279

fund of finance’, so that short term debts will be rolled over to

| maintain a ‘wages fund’. I have arbitrarily assumed that the investment good depreciates at a rate of 10 per cent per period. If the investment good fully depreciates in one period, then capital is not accumulated. However, even in this case, aggregate income is sufficient to retire aggregate debt. Since it would be unrealistic to assume that one consumption goods firm could recapture the entire gross profit of one period, however, debt retirement at an individual level would

be quite difficult. But if the economy were expanded to include a great number of firms, each of which is carrying on production by purchasing investment goods and labor services, it is easy to imagine that each individual firm can (but may not) capture sufficient gross profits to retire debts.

Expectations, uncertainty, and investment finance The simple models presented above give some insight into the investment finance process. Investment funded through credit expansion generates sufficient income at the aggregate level to retire debt. Credit expansion also generates sufficient deposits at the aggregate level to enable banks to capture funds to replace those being withdrawn as loans advanced are utilized by borrowers. However, these models are too simple and mechanistic. Even though aggregate income and aggregate deposits are sufficient, there is no guarantee that individual firms and banks are able to capture funds needed. Individual firms must capture a share of the aggregate income in order to cover costs, repay debt, and obtain a normal profit share. This requires that all income is spent so that it returns to the circular flow. Individual banks must capture deposits to replace reserves as demand deposits are drawn down. This requires that all created funds remain within the system (so that no demand deposits are converted into currency). An investment project is undertaken with the expectation that future income flows will enable the future stream of payment commitments to be met. However, aggregate profits (and saving) depend on aggregate investment. For example, assume that in the second period, the consumption sector firm uses the existing capital assets to produce consumption goods, again purchasing $900 in labor services. (There is no net investment. In fact, investment is negative this period due to depreciation.)

280

Commercial Bank

Consumption Sector

Public

Asset Liability

Asset

_Liability

Asset Liability

+GS 900* DD *

+CP 810

+DD 900 * -GS 900 *

+CP 810* *

CP 810*

PB 90 *

+DD 810

DD 900

GS 900* CP 810 19

* PB9

*

*

NW

DD 900 * NW *

*

90

I have assumed the consumption sector has used its $90 demand deposit which remained from the previous period, and issued short term debt of $810 to purchase $900 in labor services. Now the consumption sector produces consumption goods, which causes the Sia good which was purchased last period to depreciate another Commercial Bank

Consumption Sector

Public

Asset

Liability

Asset

Liability

Asset

~ i se

-GS 900 -I 10 +C 1000

* : *

of

-C 900 +DD 900 *

-CP 810 * -PB 10 * ee

-DD 820

-DD 820

eer

Se I 80

*

(@ 100

+NW90

= ‘ * +C 900* -DD 900*

* =

-CP 810 -PB10 ee

*

NW 180

*

Liability

bs : per ay a * *

The consumption sector sells $900 in consumption goods to the public, receiving a deposit of $900. It then retires the short term debt ($810), and makes a payment on its long term debt ($10), leaving it with a demand deposit of $80 and a capital good worth $80. Again, the demand deposit is sufficient to retire all the long term debt. Note, however, that no net profits have been accumulated: the consumption sector holds a demand deposit of $80 at the end of this period, while it began the period with a deposit of $90 (carried over from the previous period). The consumption sector is also left with unsold consumption goods worth $100. Thus, although net worth has increased to $180, the

firm holds $100 of this in unsold inventories. (In other words, gross ‘saving’ is equal to $100 - the unsold inventories.) Excluding

281

unsold inventories, net wealth has declined by $10, and net investment is -$10, which is equivalent to aggregate profits (excluding undesired inventory accumulation). Thus, aggregate profits depend on investment. As Minsky (1986b) argues, investment this period is undertaken only if investment is expected to occur next period. Otherwise, aggregate profits in the next period would not be positive. In Kalecki’s simple model, aggregate profits equal investment plus capitalist consumption (out of profits). In this example, since net investment (excluding inventories) was negative, profits can be positive only if capitalists consume the unsold inventories. Thus, by drawing down the $80 demand deposit (or by going into debt), capitalists could sell part of the inventory of consumption goods to themselves, and thereby realize profits. Similarly, banks will individually create credit only if they expect that they will either retain deposits or attract new deposits. Although sufficient deposits are created in the aggregate to cover every extension of credit, individual banks must ensure that their

credit extension is covered by reflux or that they are able to purchase funds in wholesale markets (for example, federal funds, wholesale certificates of deposit, or at the discount window). Leakages of deposits into other banks can be returned via interbank loans, while leakages of deposits into currency (or, in an open economy, out of the country) must be made up through central bank injection of reserves or by borrowing from foreign banks. In Keynesian theory, investment is primarily a function of profit expectations. As the future is, to some extent, unknowable, these

expectations are subject to wide fluctuations. An individual firm cannot know whether its investment project will generate the stream of returns necessary to meet the stream of payment commitments undertaken to finance the project. The individual firm must be concerned with the demand for the product it plans to produce. This demand is at least partially a function of the level of aggregate demand, which determines aggregate income. Thus, expected aggregate demand may also enter into the firm’s investment decision. Aggregate demand, in turn, is a function of investment and other autonomous expenditures. Investment proceeds when firms are sufficiently confident about the future. Investment, in turn, generates the aggregate income necessary to enable investment projects to succeed. Investment, however,

proceeds on the basis of expectations about the future. Expected

success, not success itself, is the necessary condition. That is, since

282

credit can fund investment, it is not necessary for firms to accumulate funds (for example, profits on past successful projects) before proceeding with a project. Neither do banks need to intermediate ‘loanable funds’ in order to advance finance. If banks share the profit expectations of prospective borrowers, they can create credit to allow the project to proceed. If the project proves successful, the credit can be retired. However, banks do need to be concerned with the liquidity preference of the public. If the public prefers to hold highly liquid assets (for example, demand deposits), banks may have to pay a higher interest rate to attract deposits into less liquid forms (such as time deposits which are less volatile and which do not carry a reserve requirement). Banks must ensure that the interest rate they charge on the loans they make will cover the cost of retaining deposits over the term of the loans, with a margin to cover default risk and to provide a normal return on bank equity. An increase in liquidity preference can be related to a decline in profit expectations: firms are not willing to invest, which leads to a decline in aggregate demand and profits, and which makes it difficult to meet payment commitments. An increase in liquidity preference can also be related to the public’s preferences for liquid deposits and currency, so that banks must obtain reserves to meet reserve requirements on demand deposits or to meet a drain on deposits (resulting from a conversion of demand deposits into currency). If the Fed does not intervene (by providing reserves), banks may try to liquidate assets to obtain liquid reserves. However, this is impossible in the aggregate, and may generate a general run to liquidity.

Conclusions Short term credit finances the wage bill, which flows back to consumption sector firms and replenishes the wages fund or allows retirement of short term bank debt. Wages paid in the investment sector generate profits received in the consumption sector (which can be divided among consumption sector firms, investment sector firms, and banks). Accumulation of investment goods is matched by equivalent accumulation of financial assets - which may take the form of long term bonds held by the public. Alternatively, banks

might accumulate the long term bonds and allow firms (or the public) to hold demand deposits instead. This confirms Kahn’s 283

(1954) observation that the ‘money supply’ is partially determined by the quantity of bonds which the banking system is required to hold. At any point in time, the quantity of demand deposits outstanding is determined by the short term and long term credit which has not yet been retired (as Lavoie 1985 has argued). Exogenous money dropped from helicopters will also allow economic growth. However, in any modern monetary economy, most money is privately created rather than exogenously dropped. In any case, it is clear that while endogenous and exogenous money allow the system to expand, saving cannot do so. It is the willingness to increase spending which allows growth and which generates saving. Finally, in a pure credit model, spending in each period determines income, and spending on investment goods determines saving. The quantity of saving in the previous period cannot directly affect either spending or saving in the present period. Indeed, saving can be transferred from the previous period to this period only if some debts are not retired. While the quantity of accumulated saving is always sufficient in the aggregate to retire all debts, saving flows are only made possible by accumulation

elsewhere of debts.’

However, in a world of helicopter money, exogenous injections of money will generate equivalent quantities of ‘savings’ in the form of hoards - which are not offset by private debts. In this case, saving can be accumulated through time even without inside debts. In other words, in a credit economy, deficits are necessary to generate saving, while in a world of helicopter money, deficits are not prerequisite to saving. In a credit economy, a constant level of spending requires a constant level of finance in the form of a revolving fund to cover the wage bill. The revolving fund will have to be continually renewed by banks. An increase in spending requires an increase in the quantity of finance advanced by banks and will then lead to a larger revolving fund which can be continually renewed. Notes 1. Boulding also recognized the usefulness of a balance sheet approach to money and spending. His early texts (the first edition appeared in 1941) used balance sheets to show how loans make deposits. He also ato that all banks, acting together, could increase the money supply:

284

A many-bank system is rather like a number of balloons tied together with a string; if one balloon tries to get away from the others the strings will bring it back, but all the balloons can rise together without difficulty. The ‘string’ in the case of a banking system is the loss of cash reserves. One bank expanding loans disproportionately will lose reserves to the other banks, but if all banks expand together they will all lose reserves to each other, which means of course that no bank loses reserves on balance except to the public (Boulding 1966, p. 107). 2. Ihave arbitrarily assumed the public does not purchase commercial paper. The public’s portfolio is comprised of liquid assets with good secondary markets and long term private bonds which are illiquid. 3. All assets are valued in dollars or wage units. I am assuming throughout that there is no inflation. Inflation would complicate the analysis. If wages increase commensurately with general price increases, then it would become easier for firms to repay debts. If wages increase slower than prices generally, then firms may not be able to sell all consumption goods produced. These problems will be ignored by assuming no inflation. Since I am also assuming that the public can supply labor services on demand, labor bottlenecks will not occur. Thus, the assumption of no inflation is not unrealistic for these simple models. 4.

5.

6.

7. 8.

It is, of course, unrealistic to assume that firms and banks initially

begin with zero net worth. The net worth of a firm is an important determinant of its ability to obtain credit. The net worth of a bank is an important factor which makes bank liabilities (such as demand deposits) liquid. Other factors contributing to the liquidity of bank liabilities include portfolio diversification, deposit insurance, and the willingness of the central bank to operate as a lender of last resort. Equilibrium is used merely to indicate a position in which the balance sheet held is the one desired by banks and firms. In the case of the bank, reserves equal required reserves in equilibrium. Equilibrium in the sense used here has nothing to do with full employment. In fact, I have essentially assumed there is an infinite supply of labor services available. A third constraint occurs if the marginal propensity to consume out of wages declines below unity. This possibility will be discussed in the next section. See Terzi (1986-87) for a proof of this statement using flow-of-funds accounts. See also Wray (1988a). Foster argues: Aggregate income cannot be transferred from one time period to

another. This means that one generation cannot implement or diminish the income of future generations by expending less or more than its own means. The traditional concern with this issue

arises out of the palpable fact that individuals can hoard, and

285

they can incur debts beyond their accumulation of assets. These opportunities to individuals arise out of the fact that an individual can, and often does, buy more than he sells or sell more than he

buys. But this opportunity is not available to the entire economy...sales and purchases are necessarily equal.... There is no ‘time lag’ here. The community at large cannot ‘save money’; it can save only by investing, and its savings are constituted by that investment.... [T]he equation between aggregate sales and aggregate purchases forces the instantaneity of the equation between saving and investment (Foster 1981, pp. 967-8).

286

10

CONCLUSIONS AND POLICY IMPLICATIONS

Summary of the endogenous money approach Money came into existence once private property developed. With the development of private property came the possibility of individual insecurity and the possibility of loans of private reserves. Money thus was created as a unit of account when private reserves were loaned. Furthermore, once a large class of propertyless workers was created, production for the market became possible.

Money as a unit of account became increasingly important because production for the market required a large wage fund. Thus, there was simultaneous development of propertyless workers, consumers, a market economy, and money used as a medium of exchange as the economy moved toward a capitalist economy based on money. In a capitalist economy, money is created endogenously as private debts are incurred and as individuals and firms purchase financial and capital assets with the expectation of making profits. The purchase of assets frequently entails debt creation since one can finance positions in assets by issuing debt. As long as one can convince creditors that he is creditworthy, he is not required to first accumulate

credits

in order

to purchase

assets. Thus, one

obtains assets today on the basis of a promise to pay tomorrow. Such promises represent money in its function as a unit of account. A debt is merely a balance sheet entry: ‘money’ (or purchasing power) today for more ‘money’ tomorrow. Some such promises also

circulate as media of exchange.’ Much analysis has been concentrated on demand deposits functioning as a medium of exchange. There are specific reasons (discussed previously) why they are so important as a medium of exchange, but they are also important as a unit of account: they are created as banks make loans. Furthermore, other private debts could function as media of exchange. For example, bills of exchange certainly fulfilled that function a century ago. Today a variety of private

debts

could

function

287

as media

of exchange,

including demand deposits, commercial paper, and government bonds. However, cash (the liability of the government) and demand deposits are the primary media of exchange, for reasons to be discussed below. Demand deposits are often the terms in which final settlements are made. That is, they are used as means of payment. The use of a demand deposit enables one to retire debt commitments and to destroy money. When a firm retires its own debt (destroying money), it may use the debt of a bank (the demand deposit). However, there is no reason why only demand deposits can perform the role of circulating to enable third parties (that is, parties unrelated to the original transaction in which the demand deposit debt was created) to settle accounts. Again, the bill of exchange is a good example of a private debt which circulated widely as it was accepted in payment. However, as will be discussed in the next section, demand deposits do currently play a special role as means of payment in our economy. Orthodoxy concentrates on the use of money (once it has been created) to satisfy transactions, precautionary, and speculative demands. While it is true that money is used for these purposes, if one focuses solely on the use of an exogenously given quantity of money, one ignores the essential feature of a monetary economy: money is created to transfer purchasing power across time. In the absence of this feature, economic growth and capitalist accumulation would be impossible unless exogenous money were fortuitously dropped into the economy (and into the right hands) as needed. The ability to incur deficits gives capitalists the purchasing power which enables economic growth and capital accumulation to proceed and which organizes and gives logic to the capitalist system. The purpose of capitalist production is to end up with more money that you started with, and the purpose of money in such a system is to provide purchasing power to those who will engage in capitalist production. Of course, once money is created in this process, it then fulfills a wide variety of subsidiary functions. The difference between money demand and liquidity preference has also been discussed in previous chapters. Money demand was defined as a willingness to expand one’s balance sheet. Thus, a demand for money means one is willing to go into debt in order to ‘buy now and pay later’. For this reason, money demand is not independent of money supply, since the demand for money by creditworthy borrowers will normally be met by an increase in the money supply. This does not mean that all demand for credit is

288

met. At the individual bank level, loans will be given only to those who are extremely likely to repay them. At an aggregate level, the logic of capitalism requires that credit primarily functions to place spending power into the hands of those who will accumulate capital. Thus, the logic of the entire system also requires credit rationing. One of the problems facing a private enterprise system, however, is that there is no guarantee that credit extension either

at the individual level or at the aggregate level is of exactly the ‘correct’ amount or that credit finds its way into the ‘correct’ hands. As discussed above, an increase in the supply of money is often associated with an increase in the quantity of the medium of exchange and in the quantity of the means of payment, but this is not essential. I defined liquidity preference as a desire to hold liquid assets, measured by the premium required to get one to hold an asset with less liquidity. As such, an increase in money demand is certainly not equivalent to a rise of liquidity preference. Normally, the two would be inversely related, since an increase in money demand indicates a willingness to go into debt, while an increase in liquidity preference would be associated with an attempt to retire debt and reduce the size of balance sheets. Similarly, an increase in money demand would likely be associated with an increase in profit expectations, while an increase in liquidity preference would be associated with falling profit expectations. An endogenous money supply was defined as one which is not independent of money demand. As profit expectations rise and the demand for money to finance positions in assets rises, the supply of money also normally increases because financial institutions (such as commercial banks) share in the optimism. When the money supply increases, the base also usually increases due to legal reserve requirements. However, even if the reserves don’t increase (for example, because the central bank is using tight money policy), banks can economize on reserves so that they are able to accommodate money demand. Given a supply of creditworthy borrowers, the quantity of credit which banks are willing to grant relative to the quantity of safe assets or equity held by a bank is determined primarily by conventions and rules of thumb. Until prudent leverage ratios are reached, banks will expand their loan portfolios with little upward pressure on interest rates. Eventually, however, there is upward pressure on interest rates as banks and other economic units become concerned with leverage ratios. That is, as liquidity becomes stretched, a greater liquidity premium is

289

required to induce economic units to continue to expand balance sheets. Rising interest rates, in turn, increase fragility and help to turn a robust expansion into a downturn. Thus, even if the money supply is endogenously determined, this does not mean that the interest rate must be exogenously determined. As loan portfolios expand beyond a prudent ratio relative to safe assets, banks will require rising interest rates to induce

them

to make

additional

loans.

Therefore,

short

term

interest rates are endogenously determined by the demand for credit and by rules of thumb regarding appropriate bank leverage ratios. Short term interest rates are not completely market determined, however, since much credit is granted only to established customers and credit rationing is the normal case. In addition, the actions of the public and of the central bank go into the determination of short term interest rates, since central bank

actions can affect the price and quantity of reserves, and since public preferences go into the determination of the costs incurred by banks when they take positions in short term assets such as loans. The long term interest rate, in turn, is determined primarily by the premium required to induce the public and banks to hold illiquid assets rather than liquid assets. Finally, speculators can influence both the short term and long term interest rates as they arbitrage in an attempt to outguess average opinion. As auction markets replace close relations between banks and borrowers, speculation and expectations will play a larger role in the determination of short term and long term interest rates actually paid. In this volume, I traced the endogenous money thread which has appeared over the past two centuries in a number of monetary theories. I have also presented evidence which tends to support the endogenous money approach. I have argued that various empirical tests support the hypothesis that income and spending determine the money supply and that reserves are determined by money. This casts doubt on the exogenous money approach, in which reserves determine money, which then (at least according to Monetarists) determines spending. I have also argued that an analysis of business cycles tends to support the endogenous money approach. I have extensively analyzed financial institutions and shown that their behavior is also consistent with that approach. Finally, I have presented simple models which show that expansion of balance sheets in a manner consistent with the theory of endogenous money is at least possible. Thus, the history of the

290

origins of money and banking, formal empirical tests, informal empirical analysis of business cycles, informal analysis of financial institutions, informal analysis of Fed behavior, and simple modeling all give support to the endogenous money approach. Finally, I examined the evolution of financial institutions and the role the state played in the development of a modern financial system. In the early period, the primary reason for state intervention was to provide purchasing power to the state. In the earliest period, this was done most successfully through forced loans or by making the citizens responsible for state debt. Later, however, the state set up a central bank to provide state finance.

The power of the central bank to do so was enhanced by providing special powers which would ultimately make the liabilities of the central bank more desirable than privately issued liabilities. Gradually, a mono-reserve system was created in which many privately issued liabilities were made convertible into central bank liabilities, so that the central bank liabilities would be held on reserve and would be used for interbank clearing. By varying the

quantity and price of central bank liabilities which would be used as reserves, the central bank gained some control over the financial system, and by entering as lender of last resort, the central bank could help stabilize the financial system. Thus, the present financial system is the result of a long process of evolution, and a historical analysis is necessary to avoid attempting to make general theory based solely on an analysis of current institutions. As will be discussed in the next section, the theoretical analysis underlying orthodox policy prescriptions makes precisely this mistake.

Policy implications Demand deposits have attained a special status in our economy because of the special role commercial banks have come to play. Some of the liabilities of commercial banks are insured by the government (FDIC), and the rest of the liabilities are at least implicitly insured because of the Fed’s willingness to enter as a This special relationship has tended to lender of last resort. obscure the fact that demand deposits are merely private debts with

some

degree

of government

insurance.

Like

other

debts,

demand deposits are created when an economic agent (a commercial bank) wishes to purchase an asset today (such as an IOU of a firm) in exchange for a promise to pay (a demand Zot

deposit). Because of the special status of commercial banks and their liabilities, this demand deposit in some sense takes on a life of its own as it is passed from agent A to agent B to agent C, and sO on, as a medium of exchange or as a means of payment. The demand deposit finally ceases to exist when it is returned to the issuing bank either to retire the original IOU or in exchange for bank reserves (liabilities of the central bank or the government). Because the government has tended to develop a special relationship with commercial banks, it has also traditionally attempted to establish special regulations which apply to commercial banks (and to other financial institutions in varying degrees). One such regulation has been to require that commercial banks hold reserves against certain types of liabilities. This requirement serves two functions: it makes those assets which qualify as reserves highly desirable (which initially provided a source of finance to the government), and it helps to reduce riskiness of bank portfolios. Reserves provide a safe asset which can be used to meet an unexpected ‘liquidation’ of bank liabilities (which occurs when depositors exchange deposits for currency or when another bank receives the deposits, forcing a movement of reserves). Furthermore, if those assets which count as reserves can be increased when a bank faces liquidation, default can be avoided. Thus, central bank provision of reserves through lender of last resort operations enhances financial stability. Reserves also act as an indirect tax on banks, raising the cost of

financing positions in assets through expansion of liabilities. It has come to be believed by some that reserve requirements enable the Fed to limit the ability of commercial banks to expand their balance sheets. There are two reasons why this is believed to be desirable: it will decrease risk, and it might limit the quantity of

credit granted. If the Fed implicitly guarantees the liabilities of

banks, it wants to have some control over the size of bank balance

sheets in order to indirectly control the number of assets the bank can purchase by issuing liabilities. This desire is based on the belief that in the absence of quantity controls, the banks will purchase too many (and excessively risky) assets. Thus, reserves raise the cost of issuing liabilities which might set a constraint on the number of assets which can be profitably purchased. Similarly, legal reserve ratios might limit the expansion of demand deposits and so constrain spending.” Perhaps more importantly, however, it was hoped that central bank control over the quantity of credit would somehow ration credit only to the worthy - those who would use it

292

to accumulate capital. If banks were free to grant credit to anyone who wanted it, some unworthy borrowers would use credit in nonproductive ways (such as for consumption) and cause inflation. Thus, central bank control over the quantity of money might eliminate inflationary pressures. However, banks are profit seeking firms. As such, they have an incentive

to lower

the cost

of obtaining

funds,

as well

as an

incentive to purchase earning assets. In order to lower the cost of issuing liabilities, banks found ways to induce customers to hold liabilities which did not entail required reserves. Banks created new types of liabilities to circumvent reserve requirements and various interest rate ceilings. Innovations such as the Fed funds market ensured that banks which needed reserves would be able to obtain them. By ‘operating on liabilities’, banks were able to expand their balance sheets and purchase earning assets. As the ratio of earning assets to bank equity rises, the profit-to-equity ratio can also rise. Indeed, deregulation and central bank constraints raised interest rates and created a moral hazard - banks made increasingly risky loans to cover rising costs of issuing liabilities. Rising competition from nonbanks and tight money policy forced banks to lower standards and increase rates of growth in an attempt to ‘grow their way to profitability’. As the Fed (recently) came to realize that reserve requirements did not function as an effective constraint on the size of bank balance sheets, it came increasingly to rely on bank equity as a quantity constraint. Thus, the Fed pressured banks to increase their equity-to-asset ratios. Banks complied with this during the early 1980s. At the same time, banks innovated in order to get around such quantity constraints and increasingly engaged in off-balance sheet operations to increase fee income so that profit-to-equity ratios could be maintained without compromising asset-to-equity ratios. As I said before, central bankers play a game in which they are at a disadvantage. Their goal is to use reserve constraints or equity ratios to limit the size of the liabilities of the commercial banks in order to control the quantity of assets commercial banks purchase. With the current deregulation mania which reigns in the United States, the difficult job of the Fed is made impossible. Deregulation has reduced direct controls over banking, and Monetarism has substituted indirect controls over the size of bank liabilities. Bankers have too much at stake to allow the Fed to win the game. Of course, to the extent that the Fed is able to make life difficult

293

for commercial bankers, other financial institutions willingly enter the game. If these financial institutions face lower costs, because Fed regulations exclude them or because of technological advantages, then they are able to take business away from commercial banks. If commercial banks are prevented from supplying finance (that is, from accepting the IOUs of economic agents which want goods, services or assets today in exchange for a promise to pay in the future), then other financial institutions will. Money is created as profit seeking agents incur debts today as they take positions in assets (real or financial). Fed control over a few types of liabilities or a few types of financial institutions is not an effective constraint on the quantity of debts incurred in the aggregate. Indeed, attempts to control the rate of growth of liabilities of banks had virtually no effect on the rate of growth of aggregate debt in the US during the past decade. The central bank’s role is made more difficult to the extent that it adopts incorrect theory. Simplistic and silly theories concerning what money is, how it enters the economy, and the role it plays in the economy have complicated the Fed’s job. Orthodox theory says that the Fed controls the money supply. So the Fed adopts target rates of growth for various monetary aggregates and wastes effort trying to hit them. Orthodoxy says that money causes inflation, so the Fed periodically destabilizes the economy trying to fight inflation by attempting to control monetary aggregates. Orthodoxy generally narrowly defines money to include the highly liquid assets included in M1 or M2, that is, some of the liabilities of the government plus the highly liquid liabilities of banks. Orthodoxy gives a special role to demand deposits (and, to some extent, to other checkable deposits) something akin to the role it gives to currency. Demand deposits are supposedly closely related to spending and/or prices, while other private liabilities are relegated to secondary, even inconsequential roles. However, the

special role played by demand deposits as media of exchange arises primarily because the Fed has previously displayed its willingness to guarantee the creditworthiness of banks. Federally guaranteed insurance, Fed discount policy, and Fed willingness to enter as a lender of last resort all help to give a special role to banks, which, in turn, gives a special role to demand deposits. Thus, Fed policy is vitally important in creating the distinguishing characteristics of demand deposits as media of exchange. Indeed, apparent Fed willingness to guarantee not only demand deposits, but all liabilities of banks (and of a variety of other financial institutions) has 294

contributed to the proliferation of a wide variety of liabilities with a great deal of substitutability for demand deposits. While orthodox theory begins with the special status of demand deposits to develop a theory of the relation between these demand deposits and variables such as money and prices, its policy would actually destroy the special status of demand deposits. Part of the reason demand deposits function as media of exchange is because the Fed does not allow banks to become liquidity constrained. By supplying liquidity as necessary, the Fed ensures that bank liabilities do not become devalued (with respect to the liabilities of the government), which is necessary to ensure that demand deposits circulate as media of exchange. The liabilities of other types of firms may be accepted at certain points in time as money by specific agents. However, because these liabilities are potentially subject to devaluation, they do not circulate widely as media of exchange. These liabilities can fluctuate in value to the extent that their issuers are not able to obtain insurance against all foreseeable and unforeseeable losses. If the Fed were to allow banks to become liquidity constrained, then their status would become more similar to that of other types of private firms: they could fail. Failure by a firm means it defaults on part of its debt commitments. In the case of a bank, this could mean that the value

of demand deposits issued by the bank would fall below the value of government liabilities. To the extent that this becomes a possibility, demand deposits would lose their advantage over other types of private liabilities as media of exchange. Thus, orthodox policy includes as money demand deposits because these circulate as a medium of exchange. Orthodoxy finds a high correlation between spending and this definition of money, or between prices and this definition of money. To constrain prices or spending, it would try to constrain the rate of growth of money, so defined. In order to constrain the rate of growth of money, it would constrain bank reserves. However, as banks are profit seeking firms, they find ways to increase their liabilities which don’t entail

increases

in reserve

requirements.

In other words,

tight

money policy succeeds in raising leverage ratios and stretching liquidity ratios. Eventually, this will push up interest rates and may cause a revision of expectations regarding appropriate leverage ratios and future profitability. This, in turn, will likely be associated with a rise in liquidity preference. Because the demand deposits issued by banks do circulate as a medium of exchange, when liquidity preference rises, this will be manifested as an attempt to 295

exchange liabilities (including non-demand banks) for demand

deposit liabilities of

deposits or cash. In either case, banks must

acquire more reserves. If the Fed is constraining the growth of reserves, the bank must liquidate assets. However, general liquidation of assets in an economy with interlocking balance sheets is not possible unless asset values are written-down because balance sheets have been leveraged. As asset values fall, debts must be repudiated. Thus, by allowing banks to become liquidity constrained, the Fed allows banks to fail. If banks can

fail, demand

deposits

lose their special

status. Therefore,

orthodox policy destroys the special status of demand deposits upon which orthodox theory is based. Furthermore, orthodox policy would destroy the relation between demand deposits and spending and/or prices which comes out of orthodox theory. In the process of destroying the special status of demand deposits, orthodox policy would also generate financial collapse. Alternatively, the central bank can abandon tight money and put on its lender of last resort hat. This then prevents failure and validates those financial practices which led to crisis in the tight money regime. An understanding of how money is endogenously created and of the role money plays in the economy should help the Fed to formulate sensible policy. Tooke understood the proper role of a central bank because he had some understanding of the way in which money enters the economy. [T]he greater or less liability to variation in the rate of interest constitutes, in the next degree only to the preservation of the convertibility of the paper and the solvency of banks, the most important consideration in the regulation of our banking system (Tooke Vol. IV, p. 400).

Nowhere does Tooke mention quantity constraints on the liabilities issued by banks. The role of the central bank should be first, to maintain parity between bank liabilities and government issued money, second, to ensure solvency of banks, and third, to maintain

stable credit markets and interest rates. The Fed can ensure parity between federal reserve notes and private liabilities by standing ready to purchase these liabilities (exchange its notes for the liabilities). The Fed can ensure solvency of banks through its willingness to loan to banks in trouble, or to

negotiate liquidation of the liabilities of a failing bank. The problem is that if the Fed guarantees that banks won’t fail and that their liabilities are as good as government liabilities, then neither

296

the holders of bank liabilities nor the bankers themselves have any incentive to police the soundness of bank portfolios. Thus, a bank has an incentive to purchase even risky assets on the basis of issuing liabilities since it knows that if the assets prove to be worthless, the Fed will purchase them at face value, anyway, so that the bank remains solvent. In the US we now have what might be the worst of all possible worlds: deregulation of banking with implicit guarantees of virtually all financial liabilities issued by banks, thrifts, and other favored

financial institutions, combined with periodic attempts at reserve constraints. This has tended to push up the cost of issuing liabilities and so encouraged banks to take risky positions. The orthodox solution is to allow ‘market discipline’ to ensure that banks police their own balance sheets, that is, to combine deregulation with reserve constraints and elimination of government guarantees. Historical experience with such a policy has shown that it does not work satisfactorily. Market discipline means periodic financial catastrophies as debts are repudiated during runs to liquidity. Fed willingness to enter to prevent debt repudiation has worked for more than four decades to prevent a re-occurrence of a 1930s-style depression. However, periodic re-emergence of Monetarist-induced Fed policies have increased Fed instability. Furthermore, willingness to intervene has, itself, tended to increase instability by eliminating ‘market discipline’. The endogenous approach to money points the way to a combination of policies which could help to minimize the instability which is inherent to a capitalist economy. The Fed must continue to intervene to prevent bank failures. In fact, I would argue that the Fed should expand its role as a provider of liquidity. Currently, the Fed selectively doles out reserves in the belief that this will constrain the growth of bank balance sheets. What it actually does is increase leverage ratios: balance sheets expand on the basis of a smaller liquid base. Rather than remaining a lender of last resort, the Fed should become a willing lender of first resort. However, since the Fed cannot allow failure, it must police banks.

Quantity constraints on liabilities do not work, so the Fed must police the assets of banks. Thus, rather than attempting to constrain the rate of growth of bank liabilities (‘money’) as Monetarists do, attention should be focused on the quality of bank assets. Minsky (1986b) advocates that the Fed open the discount window widely to all institutions, but that it strictly limit the types of assets which would be accepted for discount. In this way, the

297

Fed could control the assets of borrowing institutions. By setting a sufficiently low discount rate, banks which played the Fed’s game

would be rewarded. Furthermore, the Fed can refuse to validate

undesired financial innovations either by prohibiting them outright or by refusing at the discount window any bank which utilizes unfavored practices. This has worked in the past - the Bank of England effectively eliminated private note issues through similar practices. Rousseas (1986) discusses the use of selective credit controls. For example, the central bank might favor banks which lend to certain classes of borrowers (such as certain industrial sectors). Indeed, the Federal Reserve System was initially set up to include the possibility of regional differences in discount rates as a method of favoring chosen regions. Rousseas advocates alternative reserve requirement systems. For example, reserves might be based on the velocity of turnover of various types of bank deposits. In light of the discussion above, however, it is doubtful whether this would

have much proposal’,

effect. Rousseas’s however,

are

more

arguments convincing.

for an ‘asset reserve Under

this

scheme,

reserves would be based on bank assets, with higher reserve requirements on riskier assets. This shifts the central bank focus from ‘money’ to the structure of assets. Banks could still choose to make risky loans, but would be penalized by high required reserve ratios. Unfortunately, the asset reserve proposal suffers from two defects. First, higher reserve ratios on risky assets will increase costs and cause interest rates to rise. However, if the demand for

credit is interest inelastic (as Wojnilower believes), then this might not discourage risky loans unless the Fed were willing to raise the reserve ratios on such loans to extremely high levels. Second, banks would still be able to escape such requirements by engaging in offbalance sheet operations. Prohibitively high reserve ratios might, however, raise the fees banks would charge for back-up lines of credit, but it is not clear that this would effectively reduce the credit which firms could obtain outside banks. Furthermore, as the

BIS argues, banks appear to systematically underprice new instruments, so it might take a long time for banks to raise fees on credit lines after the Fed had switched to the asset reserve system. Thus, Fed regulations would probably have to expand to cover both off-balance sheet operations and nonbank credit instruments to have much impact. Wojnilower, as discussed above, advocates a return to a regulated

298

system based on credit rationing and low interest rates. Low interest rates represent a subsidy for firms and homebuyers, encouraging fast economic growth. (Wojnilower 1983) Low interest rates substitute for inflation as a way to encourage investment by keeping real debt burdens low. Quantity constraints entailed in credit rationing work better to inhibit speculation than do price constraints. Thus, Wojnilower advocates government regulation and intervention to eliminate auction-like markets and to return to a system based on close customer relations. Keynes, as is well known, advocated ‘euthanasia of the rentier’ by

gradually lowering interest rates. However, the Fed has periodically raised interest rates in futile attempts to implement Monetarist policy. A proper understanding of the way in which money enters the economy shows the folly of trying to fight inflation through reserve constraints. Stable and low interest rates would increase the stability of the economy by making credit available on reasonable terms. Quantity constraints only constrain the rate of growth of the money supply by increasing instability sufficiently that expectations fall and spending declines. An economy with low and stable interest rates probably will periodically experience inflation. However, less disruptive anti-inflation policies exist.? Rather than destabilizing financial markets through tight money policy, inflation can be fought through counter-cyclical government surpluses (budget surplus in inflationary periods and budget deficit in recessions), through tax incentive programs (‘TIP’, or tax rebates to those who accept lower wage or price increases), or through direct price controls. Monetarist policy tries to fight inflation by restricting credit across the board. More sensible credit policy would be to try to get credit into the hands of those who will use it to increase productive capacity. An analysis of inflation is beyond the scope of this volume, except to note that growth of the money supply is a result, not a cause, of inflation. While it is true that inflation can be fought if the Fed is willing to disrupt financial markets sufficiently, the cost of this policy exceeds the benefits, and

the costs of alternative policies to constrain inflation are lower. Low and stable interest rates and ensured availability of credit would help to stabilize the economy by making it easier to plan. As capital assets have become increasingly expensive and time consuming to build, the importance of stable interest rates and credit availability has increased. During the construction of plant and equipment, a firm normally rolls over short term finance until the construction is complete. If interest rates rise or if credit

299

becomes difficult to obtain, the firm is faced with the possibility that refinancing its position will be on unfavorable terms. To the extent that the Fed helps to stabilize credit markets, planning for the finance of such projects is facilitated. If, as Keynes argued, the interest rate sets the standard which the expected marginal efficiencies of all assets must attain, then a lower interest rate will tend to make production of a greater range of assets feasible. This would tend to raise employment levels and general living standards. If a high interest rate constrains investment, then lower interest rates might encourage investment

and growth of productivity.’

If, as I have argued above, the interest rate is endogenously determined, then achieving low and stable interest rates will be easier said than done. Easy money policy will succeed in lowering the discount rate and the interest rate on government bonds. However, whether this will decrease the loan rate of interest will depend on the state of liquidity preference. As Keynes argued, it is extremely unlikely that easy money policy will push down interest rates much when liquidity preference is high. In this case, monetary policy alone will do little to stimulate the economy. The government can, however, attempt to raise profit expectations and lower liquidity preference through the use of deficit spending. As Post Keynesians argue, even enlightened fiscal and monetary policy will not tame the business cycle. Since the capitalist economy is endogenously unstable, fine-tuning is impossible. However, enlightened policy can do much better than misguided policy has done over the past two decades. Easy money policy and countercyclical deficit spending will not eliminate cycles, but will help to place floors and ceilings on them. Furthermore, as Keynes argued, enlightened policy can shift the long run growth path up so that the short run cycles fluctuate about a path which is closer to full employment (Harcourt 1986). Perhaps this is the best that policy can do. Clearly, much of the analysis presented in this volume applies to the US, with limited applicability to other countries. Most of the analysis probably applies to other developed countries with strong currencies,

such

as Japan,

West

Germany,

and

Great

Britain.

However, the case in less developed countries and undeveloped countries is probably much different and low interest rates in a third world country might lead to disastrous capital flight. The money supply in such countries might be exogenously determined by its trade imbalance. I only very briefly analyzed the situation of

300

such countries, and argued that the endogenous money approach probably cannot fully apply to an LDC in the presence of DCs. Particularly in a very unstable country, it is unlikely that domestic promises to pay are sufficient to obtain goods and services. In that case, only cold cash or even the debt of the DC will do. Thus, the

position of much of the world during ‘normal’ times may correspond to the position of a developed economy during a liquidity crisis: money cannot be created because liquidity preference (for example, the preference for DC currency) is too high. Therefore, the money supply is exogenous in these cases. This does not mean that I accept the position of the French ‘Overdraft’ economists (discussed in Chapter 5) who argue that the endogenous money approach only applies to certain economies with highly developed financial institutions and accommodative central bank behavior.’ I have argued that the endogenous money approach helps us to understand the nature and origins of money and monetary economies. Since the development of a monetary economy, the quantity of money has always been endogenously determined.

Even

in the

absence

of a central

bank,

private

financial institutions have created liabilities to enable spending to proceed. Central banks were created to provide state finance and later to help stabilize the financial system precisely because the private liabilities endogenously created during expansions stop functioning as money during the following depressions.

Notes 1. A medium of exchange is merely a debt which circulates to facilitate exchange without entailing the creation of an additional debt. For example, I buy a shirt by drawing down my demand deposit at my bank. The shirt retailer accepts the debt of my bank (the demand deposit), so that I incur no debt in this transaction. Thus, there is no money created in this case, since the debt of my bank has merely changed hands so that I can obtain a shirt today. Alternatively, I could have obtained the shirt by issuing an IOU.to the retailer, if he were willing to accept my IOU. In this case, money would have been created, since I receive the shirt today in exchange for a promise to pay in the future. ; 2. Similarly, interest rate ceilings on demand deposits were designed to limit the size of bank balance sheets by reducing price competition for deposits. It was also believed that interest rate ceilings would remove the incentive to buy overly risky assets (with a high return) since the ceilings set a maximum price on liabilities.

301

3: For example, Post Keynesians have long advocated incomes policy as an alternative to austerity to fight inflation. See Rousseas (1986). 4. This could make the US more competitive in the world market. Of course, low domestic interest rates would tend to encourage US banks to borrow in the US to buy foreign financial assets. However, the Fed could counter this somewhat by only accepting for discount US assets. Lower interest rates would also devalue the dollar and might improve the trade balance. Moore often seems to come close to the position of the Overdraft economists. His endogenous approach to money relies heavily on central bank provision of reserves on demand. It is not clear whether he would apply his approach to an economy without an accommodative central bank.

302

BIBLIOGRAPHY

Ackley, Gardner

(1957), ‘Liquidity Preference

and Loanable

Funds

Theories of Interest’, The American Economic Review, vol. 47, no. p. 662. 5, September,

Ahmed, Shaghil (1989), ‘Islamic Banking and Finance: A Review Essay’, Journal of Monetary Economics, vol. 24, no. 1, p. 157. Anderson, Leonall C., and Jerry L. Jordan (1968),‘Monetary and Fiscal

Actions:

A

Stabilization’,

Test

of Their

Relative

Importance

in Economic

Federal Reserve Bank of St. Louis Monthly Review,

November, p. 11. Arestis, Philip and

Thanos

Skouras

(eds)

(1985),

Postkeynesian

Economic Theory: A Challenge to Neo Classical Economics, Sharpe, New York.

M.E.

Asimakopulos, A. (1983), ‘Kalecki and Keynes on Finance, Investment, and Saving’, Cambridge Journal of Economics, vol. 7, p. 221.

Asimakopulos, A. (1985a), ‘Finance, Saving and Investment in Keynes’s Economics: A Comment’, Cambridge Journal of Economics, vol. 9, no. 4, December, p. 405.

Asimakopulos, A. (1985b), ‘The Role of Finance in Keynes’s General Theory’, Economic Notes, vol. 14, no. 3, p. 5.

Asimakopulos, A. (1986a), ‘Finance, Investment, and Saving: A Reply to Terzi’, Cambridge Journal of Economics, vol. 10, no. 1, March, p. 81.

Asimakopulos, A. (1986b), ‘Finance, Liquidity, Saving, and Investment’, Journal of Post Keynesian Economics, vol. 9, no. 1, Fall, p. 79.

Asimakopulos, A. (1986c), ‘Richardson on Asimakopulos on Finance: A Reply’, Cambridge Journal of Economics, vol. 10, no. 2, p. 199. Bagehot, Walter (1927), Lombard Street: A Description of the Money Market, John Murray, London.

Baghestani, Hamid, and Tracy Mott (1988), “The Money Supply Process Under Alternative Federal Reserve Operating Procedures: An Empirical Examination’, Southern Economic Journal, vol. 55, no. 2,

October, p. 485. Balbach, Anatol B. (1981), ‘How

Controllable

is Money

Growth?’,

Federal Reserve Bank of St. Louis Review, vol. 63, no. 4, April, p. 5.

Bank for International Settlements (1983), ‘The International Interbank Market: a Descriptive Study’, BIS Economic Papers, Basle, No. 8, July. Bank for International Settlements (1986), Recent Innovations in International Banking, BIS, Basle, April.

Bergier, Jean-Francois (1979),‘From the Fifteenth Century in Italy to the Sixteenth Century in Germany: A New Banking Concept?’ in The Dawn

of Modern

Banking,

Center for Medieval

303

and

Renaissance

Studies,

University

of California,

Los Angeles,

London, Yale University Press, p. 105. Bernanke, Ben S. (1981), ‘Bankruptcy,

Liquidity,

New

Haven

and

and

Recession’,

American Economic Association Papers and Proceedings, vol. 71, no.

2, May, p. 155. NS Bernanke, Ben S. (1983), ‘Nonmonetary Effects of the Financial Crisis in the Propagation of the Great Depression’, Review, vol. 73, no. 3, June, p. 257.

American Economic

Bhaduri, Amit (1986), Macroeconomics: The Dynamics of Commodity Production, Macmillan Education Ltd., London. Bhaduri, Amit, and Josef Steindl (1983), ‘The Rise of Monetarism as a Social Doctrine’, Thames Papers in Political Economy, Autumn.

Blinder,

Alan

S. and

Joseph

E. Stiglitz

(1983),

‘Money,

Credit

Instruments, and Economic Activity’, American Economic Association Papers and Proceedings, vol. 73, no. 2, May, p. 297.

Blomquist, Thomas W. (1979), ‘The Dawn of Banking in an Italian Commune: Thirteenth Century Lucca’, in The Dawn of Modern Banking, Center for Medieval and Renaissance Studies, University of California, Los Angeles, New Haven and London, Yale University Press; Pass

Boulding, Kenneth E. (1966), Economic Analysis, Macroeconomics, Harper & Row, New York. Braudel,

Fernand

(1973),

Capitalism

and

Material

Volume

Life

II:

1400-1800,

Harper Colophon Books, Harper & Row, New York. Braudel, Fernand (1981), The Structures of Everyday Life: The Limits of the Possible, Harper & Row, New York. Braudel, Fernand (1982), The Wheels of Commerce:

Civilization and

Capitalism, 15th-18th Century, Harper & Row, New York. Brockway, George P. (1988), ‘What Cost Us Was the Bull Market Not the Crash’, Challenge, vol. 31, no. 5, September/October, p. 55. Brunner, Karl (1968), ‘The Role of Money and Monetary Policy’, Federal Reserve Bank of St. Louis Review, vol. 50. no. 7, July, p. 9. Cacy, J.A. (1968), ‘Alternative Approaches to the Analysis of the Financial

Structure’,

Monthly

Review,

Federal

Kansas City, March 1968, p. 3.

Reserve

Bank

of

Calomiris, Charles W., R. Glenn Hubbard, and James H. Stock (1986), ‘The Farm Debt Crisis and Public Policy’, Brookings Papers on Economic Activity, vol. 2, p. 441.

Cameron,

Rondo

(ed)

(1967),

Banking

in the

Early

Stages

of

Industrialization: A Study in Comparative Economic History, Oxford

University Press, New York. Caskey, J. and Fazzari, S. (1986),‘Macroeconomics and Credit Markets’, Journal of Economic Issues, vol. 20, no. 2, p. 421.

Chick, V. (1983), Macroeconomics After Keynes, Philip Allan, London. Chick, V. (1985), ‘Keynesians, Monetarists, and Keynes: the End of the Debate or a Beginning?’, in Arestis, Philip and Thanos Skouras (eds),

304

Post Key nesian Economic Theory: A Challenge to Neo Economics, M.E. Sharpe, Armonk, New York, p. 79.

Classical

Chick, V. (1986),‘The Evolution of the Banking System and the Theory of Saving,

Investment

and

Interest’,

Department

of Economics,

University College of London, Discussion Paper No. 86-01. Chick, V. and Sheila Dow (1988), ‘A Post-Keynesian Perspective on the Relation Between Banking and Regional Development’, Thames Papers in Political Economy, Spring. Cipolla, Carlo M. (1980), Before The Industrial Revolution: European Society and Economy, York.

1000-1700,

W.W.

Norton

& Company,

New

Cooley, Thomas F., and Stephen F. LeRoy (1981), ‘Identification and Estimation of Money Demand’, The American Economic Review vol. 71, no. 5, December, p. 825. Cooper, Ian (1986), ‘Innovations: New Market Instruments’, Review of Economic Policy, vol. 2, no. 4, Winter, p. 18.

Oxford

Cramp, A.B. (1962), ‘Two Views on Money’, Lloyds Bank Review, July, pal.

Cramp, A.B. (1970), ‘Does Money Matter?’, Lloyds Bank Review, October, p. 23. Cramp, A.B. (1989), ‘Liquidity’, in Eatwell, John, Murray Milgate, and Peter Newman (eds), The New Palgrave: Money, New York and London, W.W. Norton, p. 185.

Davidson, Paul (1965), ‘Keynes’s Finance Motive’, Oxford Economic Papers, Davidson, Davidson, Journal

vol. 17, no. 1, March, p. 47. Paul (1978), Money and the Real World, Macmillan, London. Paul. (1986a), ‘Finance, Funding, Saving, and Investment’, of Post Keynesian Economics, vol. 9, no. 1, Fall, p. 101.

Davidson, Paul (1986b), ‘A Post Keynesian View of Theories and Causes for High Real Interest Rates’, Thames Papers in Political Economy,

Spring. Davidson, Paul and Sidney Weintraub (1973), ‘Money as Cause and Effect’, The Economic Journal, December, vol. 83, p. 1117.

Davis, Richard G. (1968), ‘The Role of the Money Supply in Business Cycles’, Monthly Review, Federal Reserve Bank of New York, vol. 50,

no. 4, April, p. 63. Day, John (1987), The Medieval Market Economy, Basil Blackwell, New York, New York. De Cecco, Marcello, (1987), ‘Financial Innovations and Monetary Theory’, in De Cecco, Marcello (ed) Changing Money, Basil Blackwell,

Oxford, p. 1. DeVroey, Michel

(1984),

‘Inflation:

a

Non-monetarist

Monetary

Economics, vol. 8, Pp. 381. dge Journal of Interpretation’, Cambri

Desai, Meghnad (1989), ‘Endogenous and Exogenous Money’, in Eatwell, : The New John, Murray Milgate, and Peter Newman (eds), Palgrave 146. p. Norton, W.W. London, and York Money, New

305

Dillard, Dudley (1948), The Economics of John Maynard Keynes: The Theory of a Monetary Economy, Prentice Hall, New York.

Dillard, Dudley (1980), ‘A Monetary Theory of Production: Keynes and the Institutionalists’, Economic Journal of Issues, vol. 14, no. 2, June, a 2o0% Dillard, Dudley (1982), ‘Rewriting the Principles of Economics’, Journal of Economic Issues, vol. 16, no. 2, June, p. 577.

Dillard, Dudley (1984), ‘Keynes and Marx: A Centennial Appraisal’, Journal of Post Keynesian Economics, vol. 6, no. 3, Spring, p. 421. Dillard, Dudley (1986), ‘The Institutional Principle of the Principles of Economics’, Journal of Economic Issues, vol. 20, no. 2, p. 355. Dobb, Maurice (1963), Studies in the Development of Capitalism, International Publishers, New York.

Dow, Alexander C., and Sheila C. Dow (1989), ‘Endogenous Money Creation and Idle Balances’, in Pheby, John (ed) New Directions in Post-Keynesian Economics, Aldershot, Edward Elgar, p. 147.

Dow, J.C.R., and Saville, 1D. (1988), A Critique of Monetary Policy: Theory and British Experience, Clarendon Press, Oxford.

Dow,

Shiela

(1985),

Macroeconomic

Thought:

A

Methodological

Approach, Basil Blackwell Ltd, Oxford.

Dugger, William M. (1984), ‘The Nature of Capital Accumulation and Technological Progress in the Modern Economy’, Journal of Economic Issues, vol. 18, no. 3, September, p. 799.

Ehrenberg, Richard (no date), Capital and Finance in the Age of the Renaissance:

A

Study

of the

Fuggers_and_

Harcourt, Brace & Company, New York. Eichner, Alfred S. (1985), Toward a New

Their

Connections,

Economics:

Essays

_in

Post-Keynesian and Institutionalist Theory, Armonk, New York, M.E.

Sharpe. Eichner, Alfred S. and J.A. Kregel (1975), ‘An Essay on Post-Keynesian Theory: A New Paradigm in Economics’, Literature, vol. 13, no. 4, December, p1293:

Journal

of Economic

Eisenberg, Meyer (1987), ‘The Current Status of the Regulation of Financial Services and Products in the United States: Developments and Trends’, in De Cecco, Marcello (ed.), Changing Money, Basil Blackwell, Oxford, p. 28.

Evans, Gary (1984), ‘The Evolution of Financial Institutions and the Effectiveness of Modern Monetary Policies’, Journal of Economic Issues, vol. 18, no. 2, June, p. 439.

Fazzari, Steve, and Hyman P. Minsky (1984), ‘Domestic Monetary Policy: If Not Monetarism, What?’, Journal of Economic Issues, vol. 18, no. 1, March, p. 101.

Feige, Edgar L. and Robert McGee (1977), ‘Money Supply Control and Lagged Reserve Accounting’, Journal of Money, Credit and Banking, vol. 9, no. 4, November, p. 536.

Fischer, Stanley (1977), ‘Long-Term Contracts, Rational Expecta tions,

306

and the Optimal Money Supply Rule’, Journal of Political Economy,

vol. 85, no. 1, p. 191.

Fisher, Irving (1933), ‘The Debt-Deflation Theory of Great Depressions’, Econometrica,

vol. 1, October, p. 337.

Foley, Duncan K. (1975), ‘On Two Specifications of Asset Equilibrium

in Macroeconomic Models’, Journal of Political Economy, vol. 83, no.

2, April, p. 303. Foley, Duncan K. (1983), ‘On Marx’s Theory of Money’, Social Concept, vol. 1, no. 1, May, p. S. Foley, D. (1989), ‘Money in Economic Activity’, in Eatwell,

John, Murray

Milgate, and Peter Newman (eds), The New Palgrave: Money, New York and London, W.W. Norton, p. 248. Forman, Leonard, Miles Groves, and A. S. Eichner (1985), ‘The Demand

Curve for Money Further Considered’, in Marc Jarsulic (ed), Money and

Macro _ Policy,

Boston-Dordrecht-Lancaster,

Kluwer-Nijhoff

Publishing, p. 29. Foster, J. Fagg (1981), ‘The Reality of the Present and the Challenge of the Future’, Journal of Economic Issues, vol. 15, no. 4, December, p.

963. Foster, Gladys Parker (1986a), ‘The Endogeneity of Money and Keynes’s General

Theory’,

Journal

of Economic

Issues,

vol.

20,

no.

4,

December, p. 953. Foster, Gladys Parker (1986b), The Endogeneity of Money and Keynes’s General Theory, Dissertation, University of Colorado.

Friedman, Benjamin M. (1988), ‘Lessons on Monetary Policy from the 1980’s’, Journal of Economic Prespectives, vol. 2, no. 3, Summer, p. 51.

Friedman, Milton (1953), Essays in Positive Economics, University of Chicago Press, Chicago. Friedman, Milton (1956), ‘The Quantity Theory of Money - A Restatement’, in Studies in the Quantity Theory of Money, University

of Chicago Press, Chicago.

Friedman, Milton (1969), The Optimal Quantity of Money and Other Essays, Aldine, Chicago.

Friedman,

Milton

(1982a),

Capitalism _and Freedom,

University

of

Chicago Press, Chicago.

Milton (1982b), Monetary Trends inthe United States and Friedman, the United Kingdom: Their Relation to Income, Prices, and Interest Rates, 1867-1975, University of Chicago Press, Chicago. Friedman, Milton and Anna J. Schwartz (1963), A Monetary History of the United States, 1867-1960, Princeton University Press, Princeton. Fullarton, John (1969), On the Regulation of Currencies: Being an Examination of the Principles on which it is Proposed to Restrict, within Certain Limits, the Future Issues on Credit of The Bank of England and of the other Banking Establishments Throughout the

Country, Augustus M. Kelley, New York.

Gart,

Alan

(1989), An

Analysis

of the New

307

;

,

Financial

it

Institutions:

Changing Technologies, Financial Structures, Distribution Systems, and Deregulation, Quorum Books, New York.

Giordano, Robert M. (1987), ‘The Federal Reserve’s Response to the Stock Market Crash’, Economic Research, Goldman Sachs Economic

Research Group, December 7. Goldfeld, Stephen M. (1989), ‘Demand for Money: Empirical Studies’, in Eatwell, John, Murray Milgate, and Peter Newman (eds), The New

Palgrave: Money, New York and London, W.W. Norton, p. 131. Goldsmith, Raymond W. (1987), Premodern Financial Systems: Historical

Comparative

Study,

Cambridge

University

A

Press,

Cambridge. Goodhart, Charles (1986), ‘Financial Innovation and Monetary Control’, Oxford Review of Economic Policy, vol. 2, no. 4, Winter, p. 79. Goodhart, Charles (1989a), ‘Has Moore Become too Horizontal?’,

Journal of Post Keynesian Economics, vol. 12, no. 1, Fall, p. 29.

Goodhart, Charles (1989b), ‘Central Banking’, in Eatwell, John, Murray Milgate, and Peter Newman

(eds), Palgrave: The New Money, New

York and London, W.W. Norton, p. 88. Gramley, Lyle, and Samuel Chase (1965), ‘Time Deposits in Monetary Analysis’, Federal Reserve Bulletin, vol. 51, no. 10, October, p. 1380. Graziani, Augusto (1984), ‘The Debate on Keynes’ Finance Motive’, Economic Notes, vol. 13, no. 1, 1984, p. 5.

Graziani, Augusto (1986), ‘Keynes’ Finance Motive: A Reply’, Economic Notes, vol. 15, no. 1, 1986, p. 5.

Graziani, Augusto

(1987), ‘Keynes’ Finance

Motive’,

Economies

et

Societes, no. 9, p. 23. Greider, William (1989), The Trouble With Money: A Prescription for America’s Financial Fever, Whittle Direct Books.

Gurley, John G., and E. S. Shaw (1955), ‘Financial Aspects of Economic Development’, The American September, p. 515.

Gurley, John G., and E.S. Shaw

Economic

Review, vol. NEV

eno

(1956), ‘Financial Institutions

and

Interrelationships’, Journal of Finance, vol. 11,, May, p2257, Gurley, John G., and E.S. Shaw (1960), Money in a Theory of Finance,

The Brookings Institution, Washington D.C. Hamilton, James D. (1985), ‘Monetary Factors in the Great Depression’, ee 1986).

of Virginia mimeo,

November

19, (Revised

March

22,

Harcourt, G.C. (1985), Keynes and His Contemporaries, Macmillan, London. Harcourt, G.C. (1986), ‘Making Socialism in Your Own Country’ , in Hamouda, O.F. (ed), Controversies in Political Economy: Selected Essays of G.C. Harcourt, New York, New York University Press, p. 250. Harris, Donald J. (1978), Capital Accumulation and Income Distribution,

Stanford University Press, Stanford.

308

ee) Laurence (1981), Monetary Theory, McGraw-Hill Company, New ork. Heilbroner, Robert (1985), The Nature and Logic of Capitalism, New

York and London, W.W. Norton and Company.

Heilbroner, Robert L., and James K. Galbraith (1987), Understanding Macroeconomics, Eighth Edition, Prentice-Hall, Inc., Englewood Cliffs,

New Jersey. Heinsohn, Gunnar, and Steiger, Otto (1983a), ‘Money, Productivity and Uncertainty in Capitalism and Socialism’, Metroeconomica, vol. 33, no. 1-3, Spring, p. 41. Heinsohn, Gunnar, and Steiger, Otto (1983b), ‘Private Property, Debts

and Interest Or: The Origin of Money and the Rise and Fall of Monetary Economies’, Studi Economici, vol. 21, p. 3. Heinsohn, Gunnar, and Steiger, Otto (1984), ‘Marx and Keynes: Private Property and Money’, Monnaie et Production, I.S.M.E.A., vol. 1, p. 37. Heinsohn, Gunnar, and Otto Steiger (1985), ‘Technical Progress and Monetary Production: An Explanation’, Monnaie et Production, I.S.M.E.A., vol. 2, p. 85. Heinsohn, Gunnar, and Steiger, Otto (1987), ‘Private Ownership and the Foundations of Monetary Theory’, Economies et Societes, no. 9, p.

220 Heinsohn, Gunnar, and Steiger, Otto (1989), ‘The Veil of Barter: The Solution to "The Task of Obtaining Representations of an Economy in which Money is Essential", in Kregel, J.A. (ed), Inflation and Income Distribution in Capitalist Crisis: Essays in Memory of Sidney Weintraub, New York University Press, Washington Square, New York. Hester, Donald D. (1987), ‘On the Empirical Detection of Financial Innovation’,

in De

Cecco,

Marcello

(ed)

Changing Money,

Basil

Blackwell, Oxford, pp. 77-111. Hindle, Tim (1986), ‘Topsy-Turvy: A Survey of Corporate Finance’, The Economist, June 7.

Holmes, Alan R. (1969), ‘Operational Constraints on the Stabilization of

Money

Supply Growth’,

in Controlling Monetary Aggregates,

Proceedings of the Monetary Conference, sponsored by the Federal Reserve Bank of Boston, June.

Howrey, Philip E. and Saul H. Hymans (1978), ‘The Measurement and Determination of Loanable Funds Saving’, Brookings Papers on Economic Activity, vol. 3, p. 655. Hume, David (1898), ‘Of the Balance of Trade’, in Green, T.H. and T.H. Grose (eds), Essays, Moral, Political and Literary, Volume I,

Longmans, Green, and Company, London, p. 330. 7 Jarsulic, Marc (ed) (1985), Money and Macro Policy, Kluwer-Nijhoff Publishing, Boston-Dorcrecht-Lancaster. Jefferis, Richard H., Jr. (1990), ‘The High-Yield Debt Market: 1980-1990’,

Economic

Commentary,

309

Federal

Reserve

Bank

of

Cleveland, April, p. 1.

Judd, John P. and John L. Scadding (1982), ‘The Search for a Stable Money Demand Function: A Survey of the Post-1973 Literature’, Journal of Economic Literature, vol. 20, September, p. 993. Kahn, R.F. (1954), ‘Some Notes on Liquidity Preference’, The

Manchester School of Economic and Social Studies, vol. 22, no. 3,

September, pp. 229-257. Kaldor, N. (1970), ‘The New Monetarism’, Lloyds Bank Review, July, p. se Kaldor, N. (1980), Essays on Economic Stability and Growth, Duckworth, London. Kaldor, N. (1985), The Scourge of Monetarism, Second edition, Oxford

University Press, London. Kaldor, N., and J. Trevithick (1981), ‘A Keynesian Perspective on Money’, Lloyds Bank Review, January, p. 1. Kalecki, M. (1954), Theory of Economic Dynamics, George Allen Unwin, Ltd., London.

Kane, Edward J., and Burton G. Malkiel (1965), ‘Bank Portfolio Allocation, Deposit Variability, and the Availability Doctrine’, Quarterly Journal of Economics, vol. 79, no. 1, February, p. 113.

Karekan, John H. (1967), ‘Commercial Banks and the Supply of Money: A Market Determined Demand Deposit Rate’, Federal Reserve Bulletin, vol. 53, no. 10, October, p. 1699.

Kaufman, Herbert M. and Raymond E. Lombra (1980), ‘The Demand for Excess Reserves, Liability Management, and the Money Supply Process’, Economic Inquiry, vol. 18, October, p. 555.

Keynes, J. M. (1937), ‘Alternative Theories of the Rate of Interest’, Economic Journal, vol. 47, June, p. 241.

Keynes, J.M. (1964), The General Theory of Employment, Interest and Money, Harcourt Brace Jovanovich, New York and London.

Keynes, J.M. (1971a), The Collected Writings of John Maynard Keynes, Volume

IV, Donald

Moggridge

(ed), The

MacMillan

Press,

Ltd,

London and Basingstoke. Keynes, J.M. (1971b), The Collected Writings, Volume V, MacMillan, London. Keynes, J.M. (1971c), The Collected Writings, Volume VI, MacMillan, London.

Keynes, J.M. (1973), The Collected Writings, Volume XIV, MacMillan, London. Keynes, J. M. (1979), The Collected Writings, Volume XXIX, MacMillan, London. Keynes, J.M. (1981), The Collected Writings, Volume XIX, MacMillan, London. Keynes, J.M. (1982), The Collected Writings, Volume XXVII, Macmillan, London.

Keynes, J. M. (1987), The Collected Writings, Volume XIII, Paperback

310

Edition, MacMillan, London.

Knapp, Georg Friedrich (1924), The State Theory of Money, Macmillan and Co., Limited, London.

Kohn, Meir (1986), ‘Monetary Analysis, the Equilibrium Method, and Keynes’s "General Theory", Journal of Political Economy, December, vol. 94, no. 6, p. 1191.

Kregel, J.A. (1973), The Reconstruction of Political Economy, Wiley and Sons, New York.

Kregel, J.A. (1980), ‘Expectations and Rationality within a Capitalist Framework’, Paper presented at the American Economic Association meetings, Denver, 1980.

Kregel,

J.A.

(ed)

(1983),

Distribution,

Effective

Demand

and

International Relations, Macmillan, London.

Kregel, J.A. (1984-85), ‘Constraints on the Expansion of Output and Employment: Real or Monetary’, Journal of Post Keynesian Economics, vol. 7, no. 2, Winter, p. 39.

Kregel, J.A. (1985), ‘Post-Keynesian Distribution Theory in Relation to Growth and Technical Progress’, Economie Appliquee, vol. 37, no. 2,

Pasi: Kregel, J.A.

(1986a),

‘A Note

on

Finance,

Liquidity,

Saving,

and

Investment’, Journal of Post Keynesian Economics, vol. 9, no. 1, Fall,

padi: Kregel, J.-A. (1986b), ‘Shylock and Hamlet, or Are There Bulls and Bears in the Circuit’, Economies et Societes, Serie M.P. no. 3, p. 11. Kregel, J.A. (1987), ‘Irving Fisher, Great-grandparent of the General

Theory: Money, Rate of Return Over Cost and Efficiency of Capital’, Mimeo, Johns Hopkins University, Bologna Center,

Kregel, J.A. (1988), ‘The Multiplier and Liquidity Preference: Two Sides of the Theory of Effective Demand’, in Barrere, Alain (ed), The Foundations of Keynesian Analysis: Proceedings of a Conference held at the University of Paris I-Pantheon-Sorbonne, St. Martin’s Press,

New York, pp. 231-250. Laidler, David (1977), The Demand for Money: Theories and Evidence, Harper and Row, New York. Laudadio, Leonard (1987), ‘Commercial Banks: Market Structure and Short-term Interest Rates’, Journal of Post Keynesian Economics, vol. 9, no. 4, Summer, p. 632.

Lavoie, Marc (1984), ‘The Endogenous Flow of Credit and the Post Keynesian Theory of Money’, Journal of Economic Issues, vol. 18, no. 3, September, p. 771.

Lavoie,

Marc

Overdraft

(1985), Economics,

‘Credit

wg

and

Money:

and Post Keynesian

The

Dynamic

Economics’,

Circuit,

in Jarsulic,

Marc (ed), Money and Macro Policy, Kluwer-Nijhoff Publishing, Boston-Dordrecht-Lancaster, p. 63.

Lavoie, Marc and M. Seccareccia (1988), ‘Money, Interest and Rentiers: The Twilight of Rentier Capitalism in Keynes’s General Theory’, in

3H

Hamouda, O.F., and John N. Smithin (eds), Keynes and Public Policy

After Fifty Years, Volume II: Theories and Method, New York

' University Press, Washington Square, New York, p. 145. Levine, David (1983), ‘Two Options for the Theory of Money’, Social Concept, vol. 1, no. 1, May, p. 20. Lianos, Theodore P. (no date), ‘Marx on the Determination of the Rate

of Interest’, Unpublished Manuscript. Lombra, Raymond E., and Raymond G. Torto (1973), ‘Federal Reserve ‘Defensive’ Behavior and the Reverse Causation Argument’, Southern Economic Journal, July, p. 47. Lopez, Robert S. (1979), ‘The Dawn of Medieval Banking’, The Dawn of Modern Banking, Center for Medieval and Renaissance Studies,

University of California, Los Angeles, New Haven and London, Yale University Press, 1979, p. 1. Malinowski, Bronislaw (1932), Argonauts of the Western Pacific: An Account of Native Enterprise and Adventure in the Archipelagoes of Melanesian New Guinea, George Routledge & Sons, Ltd., London. Mankiw, N.G., and Summers, L. (1984), ‘Do Long-Term Interest Rates Overreact to Short-Term Interest Rates?’, Brookings Papers on Economic Activity, No. 1, p. 223. Marx, Karl (1909), Capital, Volume HII, Charles H. Kerr and Company, Chicago. Marx, Karl (1971), Capital, 3 Volumes, Progress Publishers, Moscow. Marx, Karl (1972a), Capital, Volume III, Lawrence and Wishart, London. Marx, Karl (1972b), Theories of Surplus Value, Part 3, Lawrence and Wishart, London.

Marx, Karl (1978), Capital, Volume I, Progress Publishers, Moscow. Mayer, Colin (1986), ‘The Assessment: Financial Innovation: Curse or Blessing’, Oxford Review of Economic Policy, vol. 2, no. 4, Winter, p. i

McIntosh, Marjorie K. (1988), ‘Money Lending on the Periphery of London, 1300-1600’, Albion, vol. 20, no. 4, Winter, p. 557. Meek, Paul (1982), U.S. Monetary Policy and Financial Markets, Federal Reserve Bank of New York, New York. Meltzer, Allan H. (1982), ‘Comment on Federal Reserve Control of the Money Stock’, Journal of Money, Credit and Banking, vol. 14, no. 4, Part 2, November, p. 632.

Minsky, Hyman (1957), ‘Central Banking and Money Market Changes’, Quarterly Journal of Economics, vol. 71, no. 2, May, p. 171.

Minsky, Hyman (1975), John Maynard Keynes, Columbia

University

Press, New York.

Minsky,

Hyman

(1978),

Restatement’,

Thames

‘The

Financial

Papers

in

Instability

Political

Hypothesis:

Economy,

Polytechnic. Minsky, Hyman P. (1981), ‘James Tobin’s Asset_Accumulation Economic Activity:

A

Thames

and

A Review Article’, Eastern Economic J ournal, vol.

312

7, no. 3-4, July-October, p. 199.

Minsky, Hyman (1982), Can "It" Happen Again?, M.E. Sharpe, Inc., Armonk, New York. Minsky, Hyman (1983), ‘Financial Structure: Indebtedness and Credit’,

Washington University Working Paper 59, July. Minsky, Hyman (1985), ‘The Evolution of Financial Institutions and the Performance of the Economy’, Washington University Working Paper 86, December.

Minsky, Hyman (1986a), ‘Evolution of Financial Institutions’, Journal of Economic Issues, vol. 20, no. 2, p. 345.

Minsky, Hyman (1986b), Stabilizing an Unstable Economy, Yale University Press, New Haven and London. Minsky, Hyman (1987), ‘Securitization’, Mimeo, Washington University, September. Mints, Lloyd W. (1945), A History of Banking Theory in Great Britain and the United States, University of Chicago Press, Chicago, Illinois. Miskimin, Harry A. (1979), ‘The Impact of Credit on Sixteenth Century English Industry’, in The Dawn of Modern Banking, Center for Medieval and Renaissance Studies, University of California, Los Angeles, New Haven and London, Yale University Press, p. 275. Moore, Basil J. (1979), ‘The Endogenous Money Stock’, Journal of Post Keynesian Economics, vol. 2, no. 1, Fall, p. 49.

Moore, Basil J. (1984), ‘Contemporaneous Reserve Accounting: Can Reserves be Quantity-Constrained?’, Journal of Post Keynesian Economics, vol. 7, no. 1, Fall, p. 103.

Moore, Basil J. (1985), ‘Wages, Bank Lending, and the Endogeneity of Credit

Money’,

in Jarsulic, Marc

(ed),

Money and Macro Policy,

Kluwer- Nijhoff Publishing, Boston-Dordrecht-Lancaster, p. 1. Moore, Basil J. (1986), ‘How Credit Drives the Money Supply: The Significance of Institutional Developments’, Journal of Economic Issues, vol. 20, no. 2, June, p. 443. Moore, Basil J. (1988a), ‘The Endogenous Money Supply’, Journal of Post Keynesian Economics, vol. 10, no. 3, p. 372.

Moore, Basil J. Macroeconomics Cambridge.

(1988b), Horizontalists_and_ Verticalists: The of Credit Money, Cambridge University Press,

Moore, B.J. (1988c), ‘Keynes’s Treatment of Interest’, in Hamouda, O.F.,

and John N. Smithin (eds), Keynes and Public Policy After Fifty Years, Volume II: Theories and Method, New York University Press,

Washington Square, New York, p. 121. Mott, Tracy (1985-86), ‘Towards a Post-Keynesian

Formulation

of

Liquidity Preference’, Journal of Post Keynesian Economics, vol. 8, no.

Dapaeiz: Mis John H. (1979), ‘Bullionism and the Bill of Exchange in England, 1272-1663: A Study in Monetary Management and Popular Prejudice’, in

The

Dawn

of

Modern

Banking

313

Center

for

Medieval

and

Renaissance

Studies, University of California,

Los Angeles,

New

Haven and London, Yale University Press, p. 169. Niggle, Christopher (1986), ‘The Cyclical Behavior of US Nonfinancial Corporation’s Financial Ratios and Minsky’s Financial Instability Hypothesis’, Unpublished Manuscript, March. Norton Newsletter (1988), ‘Newsletter’, W.W. Norton and Company, Inc., Vol. 4, January. : Ong, Nai-Pew (1983), ‘The Logic of Marx’s Theory of Money’, Social Concept, vol. 1, no. 1, May, p. 30.

Panico, C. (1980),

‘Marx’s Analysis of the Relationship Between the

Rate of Interest and the Rate of Profits’, Cambridge Economics, vol. 4, no. 4, p. 363.

Journal

of

Panico, C. (1982), ‘Interest and Profit (a Study in Marx’s Writings)’, Unpublished Manuscript, March. Panico, C. (1988), Interest_and Profit in the Theories Distribution, St. Martin’s Press, New York.

Pesek, B.P. and T.R. Saving (1967), Money, Wealth Theory,

of Value

and

and Economic

MacMillan, New York.

Pigou, A.C. (1949), The Veil of Money, MacMillan, London. Polanyi, Karl (1957), The Great Transformation, Rinehart Company, Inc., New York.

&

Polanyi, Karl (1971), ‘Aristotle Discovers the Economy’, in Polanyi, Karl,

Conrad M. Arensberg, and Harry W. Pearson, (eds) Trade _and Market in the Early Empires, Regnery Company, Chicago, p. 64. Poole, William (1970), ‘Optimal Choice of Monetary Policy Instruments in a Simple Stochastic Macro Model’, Quarterly Journal of Economics, May, p. 197.

Prestwich, Michael (1979), ‘Italian Merchants in Late Thirteenth and Early Fourteenth Century England’, in The Dawn of Modern Banking

Center

for

Medieval

and

Renaissance

Studies,

University

of

California, Los Angeles, New Haven and London, Yale University Press, p. 77.

Radcliffe Committee (1959), ‘Committee Monetary System’, HMSO, London.

on

the Working

of the

Ricardo, D. (1951), Works and Correspondence, Volume III, Sraffa, P..,

and M. Dobb (eds), Cambridge University Press, Cambridge. Ritter, Lawrence S., and William L. Silber (1986), Principles of Money, Banking and Financial Markets, Basic Books, Inc., New York.

Riu, Manuel (1979), ‘Banking and Society in Late Medieval and Early Modern

Aragon’,

in The

Dawn

of Modern

Banking,

Center

for

Medieval and Renaissance Studies, University of California, Los Angeles, New Haven and London, Yale University Press, p. 131.

Robinson, Joan (1951), Collected Economic Papers, Basil Blackwell, Oxford. Robinson, Joan (1971), Economic Heresies, Basic Books, New York.

Robinson, Joan (1979), The Generalisation of the General Theory and

314

Other Essays, St. Martin’s Press, New York.

Robinson, Joan (1980), Collected Economic Papers, Volume Four, The MIT Press, Cambridge, MA. Rogers, Colin (1989), Money, Interest_and Capital: A Study in the Foundations of Monetary Theory, Cambridge University Press, Cambridge, New York. Rose, Harold (1986), ‘Change in Financial Intermediation in the UK’, Oxford Review of Economic Policy, vol. 2, no. 4, Winter, p. 18. Rose, John T. (1973), Federal Reserve System Attrition, Ph.D.

Dissertation, Washington University, St. Louis. Rousseas, Stephen (1985a), ‘Financial Innovation and Control of the Money Supply: The Radcliffe Report Revisited’, in Jarsulic, Marc (ed), Money and Macro Policy, Kluwer-Nijhoff Publishing, Boston-Dordrecht-Lancaster, p. 47. Rousseas, Stephen (1985b), ‘A Markup Theory of Bank Loan Rates’, Journal of Post Keynesian Economics, vol. 8, no. 1, Fall, p. 135.

Rousseas,

Stephen

(1986),

Post

Keynesian

Monetary

Economics,

Armonk, New York, M.E. Sharpe. Samuelson, Paul (1973), Economics, McGraw-Hill, Ninth Edition, New York. Sayers, R.S. (1957), Lloyds Bank in the History of English Banking, The Clarendon Press, Oxford.

Schumpeter, J. A. (1934), The Theory of Economic Development: An Inquiry into Profits, Capital, Credit, Interest, and the Business Cycle,

Harvard University Press, Cambridge, MA. Schumpeter, J.A. (1951), ‘The Creative Response in Economic History’, in Clemence, Richard V. (ed), Essays on Economic Topics of J. A. Schumpeter, Kennikat Press, Port Washington, NY.

Sims, Christopher (1980), ‘Comparison of Interwar and Postwar Business Cycles: Monetarism Reconsidered’, American Economic Association Papers and Proceedings, vol. 70, no. 2, May, p. 250.

Sims, Christopher

(1983), ‘Is There

a Monetary

Business

Cycle?’,

American Economic Association Papers and Proceedings, vol. 73, no. 2, May, p. 228.

Stiglitz, J. E., and Weiss, A. (1981), ‘Credit Rationing in Markets with Imperfect Information’, American Economic Review. Stevenson, Merril (1987), ‘The Risk Game: A Survey of International Banking’, The Economist, March 21. Stone, G. C., and D. L. Thornton (1987), ‘Solving the 1980’s Velocity Puzzle: A Progress Report’, Federal Reserve Bank of St. Louis Review, vol. 69, no. 7, August/September, p. 5. Terzi, Andrea (1986-87), ‘The Independence of Finance from Saving: a Flow-of-Funds Interpretation’,

Journal of Post Keynesian Economics,

vol. 9, no. 2, Winter, p. 188.

Thirlwall, A.P. (1974), Inflation, Saving and Growth Economies, St. Martin’s Press, New York.

315

in Developing

Thirlwall, Anthony P. (1987), Nicholas Kaldor, New York University Press, Washington Square, New York.

Thornton,

Daniel

L.

(1988),

‘The

Procedure: Theory and Evidence’,

Borrowed-Reserves

Operating

Federal Reserve Bank of St. Louis

Review, vol. 70, no. 1, January/February, p. 30. Tobin, James (1982), ‘Inflation’, in Greenwald, D. (ed), Encyclopedia of Economics, New York.

Tobin, James (1987), Essays in Economics, Volume I: Macroeconomics, The MIT Press, Cambridge, Massachusetts.

Tooke, Thomas (and Newmarch) (1848), History of Prices and the State of the Circulation From 1792 to 1856, Volumes III and IV, Adelphi Press, New York. Tooke, Thomas (1959), An Inquiry into the Currency Principle: The

Connection

of the Currency with Prices and the Expediency of a

Separation of Issue from Banking, The London School of Economics and Political Science, Houghton Street, Aldwych, London.

Tsiang, S.C. (1956), ‘Liquidity Preference and Loanable Funds Theories, Multiplier and Velocity Analyses: A Synthesis’, The American Economic Review, vol. 46, no. 4, September, p. 539.

Tsiang,

S.C.

(1980),

‘Keynes’s

Demand

for Liquidity,

Robertson’s

Loanable Funds Theory, and Friedman’s Monetarism’, The Quarterly Journal of Economics, vol. 94, no. 3, May, p. 467. Tsiang, S.C. (1989), ‘Loanable Funds’, in Eatwell, John, Murray Milgate,

and Peter Newman (eds), The New Palgrave: Money, New York and London, W.W. Norton, p. 190. Udovitch,

Abraham

L. (1979),

‘Bankers

without

Banks:

Commerce,

Banking, and Society in the Islamic World of the Middle Ages’, The Dawn of Modern Banking, Center for Medieval and Renaissance Studies, University of California, Los Angeles, New Haven and

London: Yale University Press, p. 255. Weintraub, S. (1961), Classical Keynesianism, Monetary Theory, and the Price Level, Philadelphia. White, Lawrence H. (1984), Free Banking in Britain: Theory, Experience, and Debate 1800-1845, Cambridge University Press, Cambridge. Wise, Tim (1990), ‘Junk Bond Overdose: Can Corporate America Break the Habit?’, Dollars and Sense, no. 154, March, p. 6.

Withers,

Hartley

(1926),

The

Meaning

of Money,

Albemarle, Street, W.

John

Murray,

Wojnilower, A.M. (1980), ‘The Central Role of Credit Crunches in Recent Financial History’, Brookings Papers on Economic Activity, Nor2apazs7:

Wojnilower, Albert M. (1983), ‘Transmuting Profits into Interest: or How to Free

Financial

Laurence

H.

Solutions,

and

Markets

and

(ed), Improving Consequences,

Bankrupt

Money Boston,

Kluwer-Nijhoff Publishing, p. 179.

316

Stock The

Business’,

in Meyer,

Control: Hague,

and

Problems, London,

Wojnilower, Albert M. (1985), ‘Private Credit Demand, Supply, and Crunches - How Different are the 1980’s?’, American Economic Review, vol. 75, no. 2, May, p. 351.

Wojnilower, Albert M. (1987), ‘Financial Change in the United States’, in De Cecco, Marcello (ed), Changing Money, Basil Blackwell, Oxford.

Wolfson, Martin H. (1986), Financial Crises: Understanding the Postwar U.S. Experience, M.E. Sharpe, Armonk, New York.

Wood, Elmer (1939), English Theories of Central Banking Control 1819-1858, Harvard University Press, Cambridge. Wray, Larry Randall (1988a), ‘Profit Expectations and the InvestmentSaving Relation’, Journal of Post Keynesian Economics, vol. 11, no. 1,

Fall, p. 131. Wray, Larry Randall (1988b), ‘Institutional Innovation, Endogenous Money, and the Deposit Expansion Process’, Paper Presented at the Second Annual West Coast Social Science Association Conference of the National Social Science Association, Newport Beach, March.

Wray, Larry Randall (1988c), ‘The Effects of Government Deficit Spending’, University of Denver Working Paper. Wray, Larry Randall (1988d), ‘Endogenous Money and _ the Determination of Interest Rates and Profits’, Paper Presented at the International Workshop on Problems in Income _ Distribution, Gatlinburg, Tennessee, June. Wray, Larry Randall (1988e), ‘The Monetary Explanation of Distribution - A Critique of Pivetti’, Political Economy: Studies in the Surplus Approach, vol. 4, no. 2, p. 269.

Wray, Larry Randall

(1989a), ‘A Keynesian Theory of Banking: A

Comment on Dymski’, Journal of Post Keynesian Economics, vol. 12,

no:,4, Fall p. 152. Wray, L. Randall (1989b), ‘A Keynesian Presentation of the Relations Among Government Deficits, Investment, Saving, and Growth’, Journal of Economic Issues, vol. 23, no. 4, December, p. 977.

Zarnowitz, V. (1985), ‘Recent Work on Business Cycles in Historical Perspective: A Review of Theories and Evidence’, Economic Literature, vol. 23, no. 2, June, p. 523.

317

Journal of

tt

wae (> teal«.

—_

Ca

oases,

obits

al

teatMe

¢ decennlliruign seingy “ye

MA.

erties

te

op-ganrs

et

7

oy

ULAR

eee

i)

:

evi

Se

.

rte yore

:

; fh

4

i.

z

Bs

s

wis

tis

Ol

ATEN

ows

ny

e

‘ee

Tt

e

104.

=]

id

|

he

1

ya

: om

é

>4

(=

7

els

OG

:

a4

s'& i

6

Lo

a"

misa~—z

: ij

a

,

=f)

oe. a

Oe

ree

tina

eyes

FF

tome ”

ay

a

ne)

a

hae

sce,

Ros

:

(eae

ee soe —

©

.

«fal

sen :

a}

oA,

a

-* : a= —

ia

INDEX

Anderson, L. 75 Asset and liability management 73, 74, 85, 140, 193, 218, 247 Assets 2, 11-20, 36, 59, 60, 73, 84-89, 92, 104, 105, 118, 135-137, 139, 143, 157, 160-166, 169, 179, 183, 195, 198, 208-218, 242-244, 287, 289, 290, 292-294, 296-299 financial assets 11, 36, 59, 133, 136, 138, 140-146, 157, 165, 212, 283 real assets 36, 139, 141, 157

Balbach, A. 76 Bank 12, 15, 17, 18, 22, 25-27, 33-36, 38, 39, 44-54, 58-62, 64-66, 85-88, 197 94097 102. 105; 1071018 113.91 1494121 125-128, 135-140, 146-149, 153, 159, 165-169, 177, 180-184, 190, 193-221, 242-245, 261, 282-284, 291-295 bank assets 51, 136, 137, 147, 184, 199, 208, 211, 243, 297, 298 bank credit 102, 114, 126, 130, 148, 159, 217 bank deposits 47, 76, 104, 123, 132, 133, 138, 146, 164, 166, 167, 220, 246, 277, 298 bank lending 11, 133, 166, 212, 259 bank liabilities 15, 45, 51, 58, 59, 66, 68, 88, 106, 107, 111, 159, 195, 197, 218, 222, 243, 246, 285, 291-293, 295-297 bankidloans” 12; 73; 139, 147, 148, 158, 165, 177, 189, 212, 213, 259, 260 bank notes 25, 27, 28, 34, 44, 45, 47-50, 58, 60, 69, 73, 102-109, 113, 14117128, 129 bank reserves 52, 76, 86, 87, 97, 98, 135, 138, 140, 148, 206, 214, 248, 249, 278, 292, 295 history of banks 25-27, 33-36, 38, 39, 44-54, 58-62, 64-66 Bank of England 25, 44-53, 60, 64, 68-70, 100, 104-109, 130, 245, 253, 298

Banking Principle

103

Banking School 69, 99, 100, 102-110, 128-130 Barter economy 3-6, 21, 55 Base 40, 61-63, 69, 75, 76, 81, 89, 90, 95, 96, 98, 136, 140-142, 196, 197, 201, 217, 247, 250, 254-256, 289, 297 Bills of exchange 26, 27, 29, 33, 34, 36, 38, 42, 45, 53, 67-70, 73, 104, 113-116, 287 Borrowed reserves

191, 204, 205, 248, 249, 257, 272, 273

319

Business cycle

17, 92, 104, 116, 242, 258-260, 264, 300

Capital 10-13, 17, 37, 45, 55-60, 89, 102, 108, 111-116, 118, 124, 129, 136, 156, 157, 161, 165, 166, 205, 289, 299 capital accumulation 2, 10, 13, 56, 57, 65, 70, 89, 90, 268, 288 capital assets 18, 58, 152, 172, 268

capital flight 300 capital gains 12, 127, 128, 185 marginal efficiency of capital 153, 157, 161, 169, 172, 187

physical capital real capital 73

111

working capital 21, 22, 37, 55-59, 65, 83, 112, 176, 190 Capitalism 2, 9, 38, 39, 54-57, 61, 64, 72, 74, 97, 111, 124, 185, 289 logic of capitalism 54, 56, 57, 64, 72, 97, 289 Cash 14, 17, 26, 32, 42, 46, 92, 97, 116, 118, 120, 135, 146, 158, 160, 163-166, 174, 175, 194, 196, 204, 208-209, 218, 243 Cash flow 195 Central bank 52-54, 60-62, 65, 75-78, 83, 87, 89, 90-94, 97, 109, 122, 128, 133, 134, 136, 137, 141, 148, 149, 168, 177, 182, 184-187, 192, 194, 219, 244, 245, 247, 249, 253, 292-294, 296, 298 central bank control 75-78, 83-90, 91-94, 109, 133, 142, 247-253, 292

293 creation of central banks 45-54, 60-62 discount window

73, 85-87, 133, 148, 149, 191, 198, 199, 203-207,

209, 214, 242, 247-249, 271, 282, 297, 298 Certificate of deposit 201, 202, 263 Chick) )V9.6657 1,153" 15991786189 Clearing drain 67, 121, 133, 194, 195 Clearing mechanism 35, 48, 182, 190

Commercial bank 17, 146, 165-167, 204, 213, 215, 219, 246, 268-272, 274, 275, 281, 291 Commodity money 4, 9, 12, 24, 27, 29, 31, 38, 40-44, 46, 53-55, 57, 58, 60, 62, 64, 70, 90, 186, 219 Contemporaneous reserve accounting 196, 197, 199, 209, 248 Convertibility 25, 27, 44, 46, 49, 68, 106, 109, 147 Cooley, T. 79-82, 173

Co-operative economy

55

Cramp, A.B. 16, 17, 22, 100, 104, 105, 108 Credit 7, 8, 10-13, 17, 28, 31-42, 44, 53, 54-61, 89, 102-106, 112- 114, 119 Se1242127. 14621478 150: 152, 159, 163, 165-167, 194, 212, 243, 244-246, 259-263, 282-284, 287-290 credit money 12, 2, 29, 31, 41, 42, 54, 55, 59, 61, 64, 65, 90, 91, 114, 129, 151, 186 credit rationing 24, 57, 63, 70, 74, 92, 97, 165, 173, 179- 184, 213, 244, 245, 289, 290, 299 flow of credit 150- 152

320

>

Currency School

69, 100, 102-105, 107-110, 128-130

Davidson, P. 18, 22, 130, 150, 158-161, 170, 176, 189 Debt 6, 11, 13-16, 27, 37, 39, 44, 65, 73, 88-90, 103, 111, 114, 116, 117, 124-126, 129, 137, 143-146, 157, 163, 179-182, 194, 260-264, 275-277, 279-283, 287-289, 291 debt deflation 90, 94, 136, 137, 152, 194, 219, 252, 253, 262, 264 debt repudiation 60, 85, 88, 93, 246, 252, 264, 297

foreign debt

251

government debt 15, 22, 36, 39-44, 67, 70, 105, 129, 143-146, 153, 268 Deficit 143, 144, 146, 163, 183 deficit spending 13, 61, 62, 65, 84, 111, 169, 170, 175, 267, 300 government deficit 61, 70, 153, 299 trade deficit 51, 101, 104, 108, 109

Deposit demand 103, 197, 254,

deposits 12, 14-17, 35, 38, 44, 59, 60, 69, 83, 84, 89, 93, 107, 135, 136, 145, 146, 149, 151-154, 172, 183, 194, 196, 199-202, 207, 208, 211, 214, 218, 220, 242, 243, 245, 246, 257, 259, 267-273, 276-280, 283-285, 287, 288, 291, 292, 294,

295, 296, 301

deposit banking 25, 45, 46, 54, 63, 73

time deposits 22, 87, 149, 196, 197, 201, 202, 207, 252, 254, 267, 269, 270-273, 283 Desai, M. 80 Discount window 73, 85-87, 133, 148, 149, 191, 198, 199, 203-207, 209, 214, 242, 247-249, 271, 282, 297, 298

Disintermediation 87, 192, 201, 246 Dow, J.C. RB. and 1D: Saville97,172, Dow, S. 66, 71, 160, 164, 165, 190

Economic growth

Economy barter economy

181, 182, 188; 189

13, 267, 284, 288, 299

3-6, 21, 55

capitalist economy 2, 10, 11, 13, 16, 20, 21, 24, 55, 59, 60, 74, 90,

12475153, 1945 718, 263, 287, 297) 300 co-operative economy 55 entrepreneur economy 55 monetary economy

1, 4, 8, 13, 21, 24, 29, 118, 119, 184, 185, 219,

278, 284, 288, 301 Endogeneity 24, 28, 85, 90, 91, 197, 253, 256 Endogenous money

1, 2, 13, 19, 22, 23, 28, 29, 62, 72, 73, 82-84, 90,

91, 99, 100, 109, 110,-116, 117, 122, 123, 128, 130, 131, 133, 134,

135, 140, 147, 150-152, 155, 156, 169, 176, 187, 208, 242, 247, 249, 254, 257, 260, 262-267, 287, 289-291, 301 definition of endogenous money 1, 2, 72-74, 83-94, 287-291

B24

Eurodollar 86, 263 Exogeneity 72, 75, 78-80, 82, 90, 91, 259 Exogenous money 1, 72, 77, 78, 81, 84, 90, 99, 117, 122, 123, 131, 174, 277, 278, 284, 288, 290 definition of exogenous money 1, 74-82, 90, 91, 294-296 Exports 70, 153, 160

Federal Deposit Insurance Corporation

15, 220

Federal funds 256, 257, 282 Federal funds rate 256, 257 Federal Reserve Bank 15, 75-77, 79-83, 85-91, 148, 149, 168, 177-179, 196-298, 203-207, 213, 216, 246-252, 254-257, 291-300 Fed policy 88, 89, 91, 177, 206, 294 Fiat money 12, 22, 25, 27, 31, 40, 41, 43, 44, 53, 54, 61, 62, 64, 65, 90, 102, 105-107, 114, 129 Finance 10-13, 19, 56, 58, 59, 61-65, 73, 118-122, 126, 135-138, 142, 144, 158-160, 162-170, 187, 212, 213, 280, 282-284, 289, 294 state finance 36, 39-45, 52, 53, 62, 65, 166, 291, 301 Finance motive 120-122, 158-160, 163, 164, 169, 170, 172-174 Financial crisis 52, 88, 89, 94, 250-252, 260-265 Financial instability hypothesis 135, 152, 156, 185, 263 Financial institution 36, 57, 127, 217, 221 Financial instruments 11, 25, 30, 36, 63, 67, 184, 215-217, 245, 246 Financial intermediaries 139, 140, 142-144, 146, 147, 154, 162, 211 Financing

5, 11, 22, 32, 54, 56, 89, 111, 127, 130, 153, 1675 Tags

204, 211, 213, 260, 292 hedgew 15; 215,2162221 Ponzi 88, 94, 98 speculative 28, 97, 98, 172, 245, 263, 288

Fiscal policy 75

Franklin National Bank 251 Friedman, B. 95, 96, 152 Friedman, M. 3, 20, 61, 74, 78, 79, 81, 82, 95, 10641339257 General Theory 192, 291

21, 99, 117, 122, 123, 130, 153, 155-157, 169, 170, 189

Gurley, J.G., and E.S. Shaw 100, 142-147, 154, 162 Horizontalist 91, 134, 147, 190

Inflation 43, 62, 64, 76, 77, 84, 95, 96, 104, 125, 126, 128, 1329133) 135, 168, 174, 192, 206, 211, 244, 250, 251, 264, 265, 285, 293; 294, 299, 302 Interest rate 28, 51, 74, 81, 87, 91, 92, 94, 97, 119, 120, 132, 134, 135, 137,139,148; 156-162, 164-167, 170, 172- 179, 181-188, 217, 244, 250, 258, 262, 263, 290, 293, 300, 301

Pa:

bd

determination of interest rates 98, 134, 149, 150, 153, 155, 158, 162-188, 290 long term 161, 172, 177, 178, 290 markup theory of determination of interest rates 155, 176, 184-186 191, 192, 222 short term 91, 134, 150, 166, 177, 178, 191 Investment 1, 11, 16, 56-59, 61, 64-66, 70, 100, 119-122, 126, 128-132, 136-138, 144, 147, 151, 153, 155, 157-176, 185, 187-190, 192, 212, 221, 246, 250, 260, 262, 263, 267, 268, 274, 275-280, 282-284, 286,

299, 300 marginal efficiency of investment 185 IS-LM

analysis

1, 131, 155, 170, 172, 173, 176

Kahn, R.F. 150, 161, 177, 178, 283 Kaldor, N. 74, 94, 100, 131-134, 138, 149, 150, 152, 177 Kalecki, M. 56, 70, 153, 177, 268, 282 Keynes, J.M. 6, 7, 11, 12, 19-21, 27, 43, 55, 94, 99, 100, 115-123, 130, 131, 153, 155-160, 162-165, 169, 170, 172-178, 184-186, 188, 189, 190, 192, 279, 299, 300 Keynesian economics 3, 155, 170-173, 185 Keynesian theory 108, 134, 142, 150, 172, 282

Post Keynesian theory of interest rates 155-157, 160-161 Post Keynesian theory of money 10-13 Kregel, \J_Ac) 130) 156; 157; 159, 171, 173 Lavoie, M. 11, 73, 100, 150-152, 156, 186, 284 Lender of last resort

15, 60, 65, 73, 85, 87, 89, 91, 133, 136, 137, 164,

197, 198, 203, 210, 215, 219, 246, 250, 251, 253, 260, 262, 285, 291, 292, 294, 296, 297

Liquidity 7, 16, 17, 22, 48, 73, 88-90, 97, 115, 118, 119, 122, 135-138, 146, 149, 159, 160, 167, 169, 185, 186, 189, 194, 215, 216-219, 243, 246, 251-253, 261, 265, 289, 295-297

definition of liquidity 16-18 demand for liquidity 73, 74, 89, 90, 94, 122, 149, 246, 261, 262 Liquidity preference 1, 16-20, 24, 63, 73, 74, 89, 90, 118, 120, 123, 130, 136, 153-165, 167-171, 173-177, 182, 183, 185-191, 245, 277, 283, 288, 289, 295, 300, 301

liquidity preference and money demand

16, 74, 163, 164, 288, 289

Loanable funds 22, 119, 151, 152, 155-158, 169, 170, 173-176, 190, 199, 279, 283 Loans 6, 10, 12, 16, 27, 45, 57-61, 64, 66, 73, 74, 85-88, 97, 111, 113, 121, 126, 165-169, 179-183, 199, 202, 204-206, 209-215, 218, 259, 287, 289-291 bank loans 12, 73, 139, 147, 148, 158, 165, 177, 189, 212, 213, 259, 260 demand for loans 66, 74, 85, 93, 116, 133, 135, 148, 151, 169, 193,

328

?

243, 245, 262 loans in Islamic society 28-30 loans in the Middle Ages 34-38 loan rate of interest

167, 168, 179, 186, 187, 190, 200, 262, 300

Marx, K. 57, 99, 110-116, 128-130, 153, 268 Meek, P. 203, 204, 220 Minsky, H.P. 11, 16, 85, 86, 88, 91, 94, 97, 98, 100, 134-138, 150, 152-154, 156, 171, 179, 181, 185, 191, 193, 198, 208, 210-213, 220, 24S 2505-251 203, 209) 2672 S227, Monetarism 18, 77, 88, 99, 107, 133, 141, 142, 192, 211, 250, 257, 260, 293 Monetarist 1, 43, 75, 93, 99, 133, 140, 141, 194, 211, 219, 258, 259, 264,297,299 Monetary policy 50, 75, 77, 81, 87, 89, 91-93, 96, 102, 108, 109, 131, 132, 135, 144, 192, 205-207, 210, 214, 219, 244, 248, 250-252, 260, 261, 265, 300 Monetary system

32, 38, 39, 131, 143, 144, 154

Money definition of money

demand for money

11-16, 24, 104, 105, 152, 173, 295

(see money demand)

origins of money 1-10, 24, 291, 301 quantity of money 11, 14-16, 18, 19, 77, 79, 84, 89-91, 100, 101,

105, 108, 110, 112, 114, 115) 117) 148, 121, 122) 128 129 Moe 169, 247, 249, 257, 264, 288, 293, 301 supply of money (see money supply) Money demand 1, 11, 16-20, 22, 23, 73, 74, 77-83, 84, 90, 91, 93-96, 98, 115, 120, 153-155, 158-160, 162-164, 170, 173, 174, 186, 189, 248, 249, 253, 256-258, 260, 288, 289, 295 Money supply ik 2 16-19, 63, 65, 66, 73-75, 77, 80, 82, 84, 89, 90-94, 104, 119, 123, 132-134, 148, 159, 166, 170, 171, 173-176, 187, 188 190, 197, 219, 248-251, 253-259, 288- 290 Moore, BJ. 12, 16, 21-23, 67, 78, 81, 87, 90-98, 100, 130, 134, 147-150, 152, 155, 156, 177, 181, 183- 186, 138; 190; 1978207 21 3Nooe 247-250, 253-255, 259, 265, 266, 302

Neoclassical theory

150

Off-balance sheet operation 85, 93, 193, 195, 208-210, 213-218, 220, 221, 242, 243, 264, 293, 298 Ohlin, B. 158 Overdrafts eon Sao 4 P48, 508 528 228 St Phillips curve

77

Pigou, A.C. 3, 21 Ponzi finance 88, 94, 98

324

>

Post Keynesian theory

10-13, 150, 155-157, 160, 161

Post Keynesians 99, 100, 150, 151, 155, 156, 160, 300, 302 Prices 5, 6, 20, 43, 51, 62, 68, 78, 91, 97, 98, 100-102, 114, 118, 125, 126, 148, 157, 171, 177-184, 215-217, 257-259, 264, 299

price constraint 94, 245 Profit

56-59, 70, 85, 110, 126, 127, 135-138, 161, 162, 263, 276, 277,

280, 282, 293-295 profit expectations 74, 138, 154, 161, 162, 164, 169, 171, 186, 187, 245, 282, 283, 289, 300 profit rate 184, 261, 272 profit seeking 24, 49, 52, 57, 93, 94, 100, 127, 128, 184, 187, 293-295 Quantity constraint 134, 166, 243, 293 Quantity theory 16, 75, 83, 99, 108, 132, 134, 142, 257, 264 Radcliffe Committee 100, 131, 132, 152 Recession 132, 133, 149, 182, 250, 252, 262, 263, 265 Reflux 25, 66, 67, 69, 107, 110, 128, 149, 165, 190, 195, 200, 282 Repurchase agreement 200, 201, 220 Ricardo, D. 101, 102, 104 Robertson, D.H. 158, 174 Robinson yal 23. 150m oie 177, Rousseas, S. 152, 177, 181, 190, 191, 222, 298, 302 Samuelson, P. 3, 4 Saving 58, 61, 64-66, 119-122, 130, 144-146, 151, 157, 158, 160-163, 167-170, 172-176, 185, 189, 190, 277-280, 284, 286 saving and investment 130, 144, 158, 163, 172, 176, 189, 286 Sayers, R.S. 47, 48, 50-52, 69, 152 Schumpeter, J. 10, 99, 100, 123-128, 154 Securitization 97, 183, 211, 212, 215, 218 Speculation 4, 177, 290, 299 Surplus 42, 56, 57, 59, 61, 62, 101, 110, 142-144, 146, 147, 162, 164, 167, 183, 186, 189, 190, 204, 268, 273, 299

Taxes 28, 31, 32, 39, 43, 46, 47, 61, 145 Tobin, J. 67, 76, 100, 138, 140, 142, 265 Tooke, T. 66, 67, 99-109, 114, 128-130, 296 Transactions demand 159, 160, 170, 173 Treatise on Money 6, 99, 117, 121, 155

Uncertainty 7, 11, 19, 20, 73, 85, 113, 146, 176, 179, 181, 184, 185, 187, 195, 216, 280 Unemployment 185, 187, 250, 255 Velocity

15, 60, 95-97, 114, 115, 119, 129, 132-135, 139, 142, 143, 152,

325

153, 191, 257,264, 267,'276; 2777298 Verticalist 91 Volcker, P. 211

Wage bill 16, 22, 55, 83, 172, 190, 259, 283, 284 Wages 49, 58, 70, 77, 83, 84, 96, 106, 129, 153, 172, 192, 255, 268, 275279, 283, 285 Wealth 8, 9, 12, 17, 21, 57, 63, 64, 96, 118, 119, 142-146, 154, 161, 162, 190, 267, 282 Weintraub, S. 21, 150 Wray, L.R. 130, 151, 190, 285

326

~

DATE DUE are

fo 2 1999

FAP

DEMCO,

ZzaAn

INC. 38-2931

Matit?

CLIFTON M. MILLER LIBRARY WASHINGTON COLLEGE CHESTERTOWN, MARYLAND 21626