Vous êtes sur la page 1sur 355

KEYNES' GENERAL THEORY

Reports of Three Decades

Keynes'
General
Theory
ED IT E D BY

REPORTS

OF
THREE
DECADES

Robert Lekachman
Barnard College, Columhia University

Palgrave Macmillan

ISBN 978-1-349-81809-9
ISBN 978-1-349-81807-5 (eBook)
DOI 10.1007/978-1-349-81807-5

Copyright 1964 by St Martin's Press, Inc.


Softcover reprint of the hardcover 1st edition 1964978-0-333-00112-7

All rights reserved


No part of this book may be reproduced in any form with
out written permission from the publisher
MACMILLAN AND COMPANY LIMITED, London
also Bombay, Calcutta, Madras, Melbourne
THE MACMILLAN COMPANY OF CANADA LIMITED,
ST MARTIN'S PRESS, INC.,

Toronto

New York

Library of Congress Catalog Card Number: 64-11970


Designed by Richard Brodney
Manufactured in the United States of America
E. A. G. ROBINSON'S "John Maynard Keynes 1883-1946" is reprinted by pennission of the Economic Journal and the Royal Economic Society.
w. B. REDDAWAY'S "The General Theory of Employment, Interest,
and Money" is reprinted by pennission from the Economic Record,
vol. 12, no. 22, June 1936.
R. F. HARROD'S "Mr. Keynes and Traditional Theory" is reprinted
from Econometrica, January, 1937, vol. 5, by pennission of The
Econometric Society.
D. G. CHAMPERNOWNE'S "Unemployment, Basic and Monetary: the
Classical Analysis and the Keynesian" is reprinted by pennission
from the Review of Economic Studies, June 1936.
ABBA P. LERNER'S "The General Theory" is reprinted from the
International Labour Review, October 1936, by pennission of the
International Labour Office.
JACOB VINER'S "Mr. Keynes on the Causes of Unemployment" is
reprinted by pennission of the publishers from the Quarterly Journal
of Economics, Cambridge, Mass.: Harvard University Press, Copyright 1936 by the President and Fellows of Harvard College.
GOTTFRIED HABERLER'S "The General Theory Mter Ten Years" is
reprinted from The New Economics by Seymour E. Harris, by permission of Alfred A. Knopf, Inc. Copyright 1947 by Alfred A.
Knopf, Inc.
PAUL M. SWEEZY'S "John Maynard Keynes" is reprinted by pennission from Science and Society, 1946.
PAUL A. SAMUELSON'S "The General Theory" is reprinted from
Econometrica, July 1946, vol. 14, by pennission of The Econometric Society.

Preface
T

HE IMMEDIATE STIMULUS to the compilation of this


volume was chronological. What better way to celebrate the General
Theory's first quarter century than to ask some of the distinguished
economists who responded to Keynesian economics at the doctrine's
inception to place in print their present evaluations of the Keynesian
revolution or "revolution"? Printed side by side, early essays and late
essays are a guide not only to the shifting, the stable, or the developing opinions of each of the nine authors but to the varying
fortunes of Keynesian macroeconomics since 1935.
It is as unnecessary as it would be presumptuous to analyze the
essays which follow. Their writers have spoken understandably for
themselves for many years. However, a word on the organization of
the contents may be in order. Part I is devoted to Professor E. A.
C. Robinson's masterly obituary article in the Economic Journal
and to Mr. Robinson's pointed reminder that the world would be a
different and worse place if Keynes had neither lived nor invented
Keynesian economics. Part II consists of five essays written in
the immediate wake of the General Theory by Reddaway, Harrod,
Champernowne, Lerner, and Viner. These are early attempts to describe the theoretical skeleton of the General Theory, critical estimates of Keynes' originality, and resolute evaluations of the new
doctrines of money and real wages, interest, resource allocation, and
involuntary unemployment. The 1963 essays which complement each
of the early articles follow a number of different intellectual strategies. Harrod states the Keynesian claim to anticipation of the dynamic
theories that Harrod himself has done so much to advance. Reddaway demonstrates how fiscal policy can explicitly be fitted into the
theoretical structure of the General Theory. Champernowne closely
examines some neglected Keynesian links between the present and
the future. Lerner warns against a resurgence of pre-Keynesian attitudes and stresses the contemporary importance of administered
v

vi, Preface
prices. Viner reaffinns his admiration of Keynesian short-run analysis
and his preference for older analyses of the long-run.
The last group of essays, the contents of Part III, mark three
events: the passage of a decade after the publication of the General
Theory, the end of World War II, and the untimely death of Lord
Keynes. It is not surprising that Haberler's, Sweezy's, and Samuelson's early and contemporary judgments vary considerably on such
issues as theoretical originality, analytical usability, and public policy
guidance. Indeed, it is scarcely necessary to say that among them
these nine economists cover most of the available theoretical and
political attitudes toward the New Economics.
The date of publication suggests that the delays which attend the
coordination of contributors, editor, publishers, and printers have
made it necessary to interpret a quarter of a century with a certain
indulgence. Nevertheless, it is much more than a routine courtesy
on my part to express my gratitude to nine distinguished contributors.
ROBERT LEKACHMAN

New York

Biographical Notes
D. G. CHAMPERNOWNE, now Fellow of Trinity College,
Cambridge, initially read mathematics as an undergraduate at Kings
College Cambridge. Coming under J. M. Keynes' inRuence, he
turned to economics and joined W. B. Reddaway as a pupil of
Keynes'. A prize fellow of Kings College in 1937 and Cambridge
lecturer in statistics in 1938, he directed the Oxford Institute of
Statistics from 1945 to 1948 and held the position of Professor of
Statistics at that University between 1948 and 1959. In the years
1945-49 he was a Fellow of Nuffield College, Oxford. Since 1959 he
has been a Fellow of Trinity and Reader in Economics at Cambridge
University.
GOTI'FRIED HABERLER, Professor of Economics at Harvard
and 1963 President of the American Economic Association, was born
in Purkersdorf, near Vienna, in Austria, and came to the United
States in 1936. A student at the University of Vienna, the Handelshochschule in St. Gallen, and the University of London, he also
attended Harvard and other American universities. He has taught
at the University of Vienna and Harvard, been associated with the
Board of Governors of the Federal Reserve System, and been elected
President of the International Economic Association (1950-1951)
and the National Bureau of Economic Research (1956). His extensive publications in international trade and economic Ructuations
include The Theory of International Trade with its Applications to
Commercial Policy, and Prosperity and Depression which since its
publication in 1937 has gone through several editions and been translated into a number of other languages.
SIR ROY HARROD, a don at Christ Church College, Oxford,
since 1922, is a fellow of the British Academy and the recipient of
honorary degrees from the Universities of Poitiers and Aberdeen.
During the Second World War, he was a member of Winston
Churchill's private statistical branch. Sir Roy was a member of the

vii

viii Biographical Notes


United Nation's Subcommission on Employment and Economic Stability (1947-1950), and one of the International Monetary Fund's
research staff (1952-1953). He has written The Life of John Maynard Keynes and a number of other volumes which include the
Trade Cycle, Towards a Dynamic Economics, Policy Against Inflation, and Topical Comment. Sir Roy's most recent work is The
British Economy.
ABBA P. LERNER was born in Bessarabia, Russia, studied
and taught in England, and moved to the United States in 1940.
His career includes spells as a cap machinist and a Hebrew teacher
as well as more orthodox academic employment at Columbia University, Roosevelt University, the New School for Social Research,
Johns Hopkins, and the Hebrew University in Jerusalem. He is now
Professor of Economics at Michigan State University. Winner of
many awards and fellowships, he has been a consultant to the Rand
Corporation and an adviser to the Israeli government. Author of
many technical articles, he has also written a number of influential
books, among them The Economics of Control, The Economics of
Employment, and Essays in Economics Analysis.

w. B. REDDAWAY, Director of the Department of Applied


Economics in Cambridge University, is also editor of the London
and Cambridge Economic Bulletin. A graduate of Kings College,
Cambridge, Mr. Reddaway won Cambridge's Adam Smith prize and
enjoyed as his supervisor John Maynard Keynes. At the time of the
writing of his original essay on the General Theory, he was Research
Fellow at the University of Melbourne. Among his publications are
The Russian Financial System, The Economics of a Declining Population, The Measurement of Production Movements, and The Development of the Indian Economy.
AUSTIN ROBINSON, F.B.A. and C.M.G., is Professor of Economics in Cambridge University. A pupil and subsequently a colleague of Lord Keynes, he is presently joint editor of the Economic
Journal. During the period 1934-1945, he was assistant editor to
Lord Keynes. In succession to Lord Keynes, he has been since 1945
Secretary of the Royal Economic Society. He is the author of the
1947 Economic Journal commemorative article on Lord Keynes
which is included in this volume.

Biographical Notes ix
PAUL A. SAMUELSON, who has been President of the American Economic Association and the Econometric Society, was trained
at Chicago and Harvard. His extensive works range from a bestselling textbook Economics (5th edition and numerous translations
since 1948) to advanced treatises like Foundations of Economic
Analysis (1947) and (with R. Dorfman and R. M. Solow) Linear
Programming and Economic Analysis. At M.LT. since 1940, he has
served as consultant for numerous government and private organizations, lectured extensively, and engaged in financial journalism for
domestic and foreign papers.
PAUL M. SWEEZY is coeditor of the Monthly Review, an
independent socialist journal, and a former member of Harvard's
economics department. A 1938 winner of Harvard's David A. Wells
Prize and a 1945 recipient of the Bronze Star, he served during the
Second World War with the Office of Strategic Services in England,
France, and Germany. In recent years he has been Visiting Professor
of Economics at Cornell and Stanford. His publications include
Monopoly and Competition in the English Coal Trade, The Theory
of Capitalist Development, Socialism, The Present as History, and,
with Leo Huberman, Cuba: Anatomy of a Revolution.
JACOB VINER, Professor Emeritus of Economics at Princeton, was born in Montreal, Canada, and educated at McGill and
Harvard. A naturalized American citizen, he taught at Chicago and
Princeton and served several terms in the American government. A
past president of the American Economic Association, he is the recipient of twelve honorary degrees, a special award from the American Council of Learned Societies, and the American Economic
Association's highest honor, the Francis A. Walker Medal. His inHuential publications include Canada's Balance of International
Indebtedness, Studies in the Theory of International Trade, International Trade and Economic Development, and The Long View
and the Short.

Contents
Preface
Biographical Notes
Introduction
PART ONE

vii

THE MAN/THE THEORY

E. A. G. ROBINSON

John Maynard Keynes 1883-1946 / 1946


Could There Have Been a "General Theory" Without
Keynes / 1963
PART TWO

13
87

THE THIRTIES/THE SIXTIES

W. B. REDDAWAY

The General Theory of Employment, Interest, and


Money / 1936
Keynesian Analysis and a Managed Economy / 1963

99
108

R. F. HARROD

Mr. Keynes and Traditional Theory / 1937


Retrospect on Keynes / 1963

124
139

D. G. CHAMPERNOWNE

Unemployment, Basic and Monetary: the Classical


Analysis and the Keynesian / 1936
Expectations and the Links Between the Economic Future and the Present / 1963

153
174

ABBA P. LERNER

The General Theory / 1936


Keynesian Economics in the Sixties / 1963

203
222

JACOB VINER

Mr. Keynes on the Causes of Unemployment / 1936


Comment on My 1936 Review of Keynes' General
Theory / 1963
xi

235
253

xii . Contents
PART THREE

THE FORTIES/THE SIXTIES

GOTTFRIED HABERLER

The General Theory After Ten Years / 1946


Sixteen Years Later / 1962

269
289

PA UL M. SWEEZY

John Maynard Keynes / 1946


The First Quarter Century / 1963

297
305

P A U L A. SAM U E L SON

The General Theory / 1946


A Brief Survey of Post-Keynesian Developments /
1963

315
331

KEYNES' GENERAL THEORY

Reports of Three Decades

Introduction
IN THE VERY LONG RUN every economic doctrine loses its
capacity to persuade. In 1798, a mere twenty-two years after the
Wealth of Nations, Malthus' Essay on Population substituted demographic pessimism for Smith's institutional optimism. Nineteen years
later David Ricardo completed Malthus' work with a tight theory
of distribution which left small scope for encouragement about the
prospects of the ordinary citizen. When Ricardo's work was done
little remained of Smith's treatise save a dauntless confidence in the
primacy of egotistic economic motives and in the pervasiveness of
the self-adjusting capacities of competitive markets. In nineteenthcentury England economists were few-Malthus himself occupied
the first chair of political economy at Haileybury, and reliable statistics were nearly' as scarce. Even so, two decades appeared a very
respectable run for a major economic conception.
In the twentieth century English-speaking lands lack for neither
economists nor statistics, and reputations are as readily acquired by
destroying theories as by creating them. For a single doctrine still to
be the center of learned controversy after nearly three decades is a
striking testimonial to the powers of its creator and the richness of
his creation. Such is the state of Keynesian economics, for the contributors to this volume seem agreed that there is no contemporary
economist who commands Keynes' position in 1963, much less in
1946, the year of his death. In 1962 when the New Republic published its symposium Time for a Keynes the principal impressions
which the assorted opinions produced were two: we need a Keynes,
and no Keynes is in sight.
A new Keynes would have work to occupy him. Problems of economic development have assumed an urgency that they did not
evince in the mid-1930'S. Although Keynesian-influenced growth
models have certainly been applied to the economic situations of
numerous developing countries, no single model and no single
growth theorist has come close to duplicating the feat of the General
I

Robert Lekachman

Theory or the personal success of its author. As for the advanced


countries, the original simplicities of Keynesian policy prescriptions
have been overtaken by the actual complexities and contradictions of
applying monetary and fiscal techniques to situations simultaneously
subject to inflation and unemployment, or high interest rates and gold
outflows, or falling demand and rising prices.
The dilemmas of recent American economic policy exemplify the
point. President Kennedy's major response to unsatisfactory rates of
unemployment was the 1963 program of tax reduction and tax reform. Now if simple Keynesian fiscal policy were enough, such a
program would dependably stimulate aggregate spending, diminish
unemployment, and restore the economy to some approximation of
full-capacity operation. Yet even Administration spokesmen have
made comparatively cautious claims for the speed and the adequacy
of this policy, and few economists indeed believe that by itself tax
policy is capable of reducing unemployment to tolerable levels. It
is a sign of the times that this tolerable level has itself shown a
secular tendency to rise from 2~-3 per cent to 4-5 per cent. For the
sad, uncomfortable fact may be that just about the tiine that some
version of the simplest Keynesian revelation has at last won the
hearts of businessmen and politicians, the nature of the economic
problem has sufficiently changed to require different remedies and
different theoretical justifications to support them.
What is a Keynesian to make of recent American economic experience? In the five years preceding mid-1963 (which were divided
about equally between Republican and Democratic Administrations)
unemployment averaged 6 per cent. Between 1948 and 1958 the
average duration of unemployment rose from 10.3 to 14-3 weeks.
Between 1958 and 1962, the United States could probably have produced something like an additional $170 billion of goods and services
in existing plants, employing readily available labor. During this
term of years, aggregate demand no doubt was deficient, but it is
less easy to say why it was deficient and unsafe to assign the only
major role in the economic drama to aggregate demand. Not that
explanation has been lacking. Here are no fewer than six:
1. Gold losses primarily to the Common Market countries have
compelled American governments to pursue monetary and fiscal
policies more cautious than domestic circumstances by themselves

Introduction 3
required. For the first time since World War I, the United States
has a balance of payments problem.
2. The apparent intractability of the unemployment problem is
the consequence of an "unusual" spurt in the growth of the labor
force. During the 1950'S the year-to-year increase of the labor force
averaged only 700,000. In 1963 the expansion approximated a million
and a quarter. Worse is yet to follow during the rest of the 1960's.
3. More jobs are lost because of automation than are gained because of economic expansion. Whether automation is or is not a
phenomenon qualitatively different from the incremental technological innovations that have been routine since the Industrial Revolution fascinates the theorists. What is certain is this: in both factories
and offices routine, slightly-skilled jobs are vanishing. It is estimated
that in offices every $5,000 of investment in the new computer devices displaces a clerk. The record is to be read in rising clerical
unemployment: 2.8 per cent in January, 1957, 3.8 per cent in January, 1960, 4.2 per cent in January, 1961, and 4.6 per cent in
January, 1962. No one is certain, but as many as a million office
workers may lose their jobs annually because computers do their
work more efficiently.
4. In the labor market supply and demand are imperfectly adjusted.
Partly as the consequence of automation, partly as the result of continuing shifts in the occupational structure from manufacturing to
the service trades, unskilled and semi-skilled jobs are becoming relatively scarce, and managerial, technical, and professional opportunities
are expanding. The existing distribution of skills does not match the
new job specifications. Displaced clerks make indifferent programmers. Discharged assembly-line workers make poor managers.
Remedies are easier to state than to implement. Older workers
require retraining and possibly relocation. New entrants to the job
market need different, greatly improved forms of vocational education. All workers would benefit from the command of the skills of
word and number, which enhance worker confidence and facilitate
the learning of new skills throughout a working career.
All that has been said on this score applies a fortiori to Negroes
and Puerto Ricans, whose job opportunities have been cruelly limited
by racial discrimination to the very unskilled positions which are now
rapidly disappearing.

4 . Robert Lekachman
5. Then there is the much disputed influence of large corporations
and large unions to which Lerner alludes in his 1963 essay. If in
the concentrated industries unions force wages up more rapidly than
productivity rises, and if managements raise their prices at the breath
of favorable economic breezes and somehow omit to lower them in
economic squalls, then any Administration is compelled to switch
from expansion to restriction well in advance of full employment.
Unless, of course, it is willing to pay the price of inflation. Samuelson speculates in his new essay that price stability in the United
States may be purchasable only at the cost of 5-6 per cent levels of
unemployment. Is such an outcome worse than annual increases in
general prices of 2-3 per cent? The answers are far from unanimous.
6. Finally, there is the possibility, stressed in Sweezy's second
essay, that the United States is undergoing a recurrence of the problems of secular stagnation which were masked but not resolved by
World War II and the consumption boom that followed it. The new
stagnationist doctrine bases itself on income distribution. Studies by
Lampman, Kolko, and Miller, among others, have documented the
proposition that the trend toward diminished inequality of income
halted by the end of World War II.
What is notable about these varied, partially overlapping, explanations of American economic sluggishness is this: with the important
exception of the last one, they have comparatively little to do with
Keynesian diagnoses. The first hypothesis is no more than the venerable gold standard justification of internal deflation. The novelty
of the gold standard prescription in the 1960's is that no nation
actually is on the gold standard and few are willing to argue for the
concomitant wage and price reductions. Explanations 2, 3, and 4 are
related either to "accidental," temporary events like unusual increases
in the labor force, or to frictional impediments which represent labor
market imperfections. Number 5, the popular postwar emphasis upon
cost-push inflation, is founded upon the behavior of exactly those
large economic units whose activities are so imperfectly comprehended within the Keynesian conceptual framework. For by abstracting from the degree of competition, Keynes in effect denied the
importance of variations in market organization to the workings of
his apparatus.
It is the sixth line of argument which, purified of its demography,

1ntroduction 5
is most interesting. If we use secular stagnation as a convenient label
for any persistent tendency of aggregate demand and aggregate
supply to approach equilibrium at less than full employment values,
we can glance at a question which has received only moderate attention in recent years from Keynesians: how much in Keynes supports
the stagnationist judgment of capitalist prospects?
Unquestionably the Keynes who is now in vogue is a safely moderate figure completely devoted to the proposition that the proper
application of intelligent monetary and fiscal policies can maintain
intact the current institutional arrangements of Western capitalism.
Indeed, taken solely as an analytical apparatus applicable to shortrun problems exclusively (as Viner argues in his second essay we
should understand Keynes), Keynesian macroeconomics plausibly
support this conservative characterization.
All the same, there is another Keynes present both in the General
Theory and in much that preceded it. This Keynes is less optimistic,
less capitalist, and more radical. Certainly as early as 1920, Keynes
was already emphasizing the fragility of capitalist arrangements.
Here, for example, is his comment on the delicate balance of forces
which facilitated the social harmonies of pre-World War I Europe:
While there was some continuous improvement in the daily
conditions of life of the mass of the population, Society was so
framed as to throw a great part of the increased income into
the control of the class least likely to consume it.... This remarkable system depended for its growth on a double bluff of
deception. On the one hand, the laboring classes accepted from
ignorance or powerlessness, or were compelled, persuaded or
cajoled by custom, convention, authority and the well-established order of Society into accepting a situation in which they
and Nature and the capitalists were co-operating to produce.
And on the other hand the capitalist classes were allowed to
call the best part of the cake theirs and were theoretically free
to consume it, on the tacit underlying condition that they consume very little of it in practice.1
As Keynes described the operations of this system of "double bluff,"
a good deal depended upon the mystique of the gold standard and
the astute, secretive central bankers who worked its delicate coni

Keynes, J. M., Economic Consequences of the Peace, Macmillan, London, 1920, p. 3

6 . Robert Lekachman
troIs. Much rested upon a complex international division of labor
which required comparatively free trade among the nations and comparative political tranquillity between the major trading partners. The
first World War slaughtered the gold standard, disrupted the international division of labor, created a host of new, protectionist nations,
and dealt a mortal blow to free trade. After the British General Strike,
few could argue that the social harmonies remained any more intact.
Keynes' concerned observation and study of English economic
troubles during the 1920'S led him in 1930 to diagnose England's
economic ailment as over-saving. In part, the difficulty was geriatric:
"England is an old country.... The population will soon cease to
grow. Our habits and institutions keep us, in spite of all claims to
the contrary, a thrifty people."2 Obviously anyone steeped in the
classical economic tradition could not believe thrift hostile to full employment, if only thrift's companion were vigorous investment. Indeed, precisely this combination of high saving and high investment
explained to Keynes' satisfaction the economic successes of 18701914- But by the time Keynes came in the 1930'S to write the General
Theory, he had begun to question the plausibility of high investment. Here is one such passage:
Today and presumably for the future, the schedule of the
marginal efficiency of capital is for a variety of reasons much
lower than it was in the nineteenth century. The acuteness and
the peculiarity of our contemporary problems arises, therefore,
out of the possibility that the average rate of interest which
will allow a reasonable average level of employment is one so
unacceptable to wealth-owners that it cannot be readily established by manipulating the quantity of money.s
What therapy is possible? To conservatives the most acceptable
Keynesian treatment is interest rate variation. But suppose, as Keynes
hypothesized in the quotation, interest rates must be forced down
so low in order to stimulate reluctant investors that the owners of
wealth will preserve rather than part with their resources for rewards
so small. Then the money that central banking authorities can so
easily create will fail to flow into productive investments. And if in
addition interest rates have little effect on new investment at best,
2

Treatise on Money, Macmillan, London, 1930, Vol. II, p. 428.


General Theory of Employment, Interest, and Money, Harcourt, Brace
& World, New York, 1935, p. 309.

Introduction 7
then even direct lending by government agencies at rates of interest
little higher than zero will have only trivial effects upon investment,
income, and employment.
This is familiar enough ground. Once we depart from the analytical skeleton of the General Theory, we can find a good many obiter
dicta which uneasily contemplate the possibilities that extra money
will simply vanish in the liquidity trap, that interest rates cannot
fall to levels low enough to evoke necessary, full-employment investment, and that the marginal efficiency of capital will remain stubbornly low. These Keynesian speculations reflect a mood and a guess
about the shape and position of the relevant functions, rather than
any empirical evidence.
But if by itself monetary policy is for reasons of this kind inadequate what else is there? Keynes' clearest statement of an alternative
policy is to be found near the end of the concluding chapter of the
General Theory:
In some other respects the foregoing theory is moderately conservative in its implications. For whilst it indicates the vital
importance of establishing certain central controls in matters
which are now left in the main to individual initiative, there
are wide fields of activity which are unaffected. The State will
have to exercise a guiding influence on the propensity to consume partly through its scheme of taxation, partly by fixing the
rate of interest, and partly, perhaps, in other ways. Furthermore, it seems unlikely that the influence of banking policy
on the rate of interest will be sufficient by itself to determine
an optimum rate of investment. I conceive, therefore, that a
somewhat comprehensive socialisation of investment will prove
the only means of securing an approximation to full employment; although this need not exclude all manner of compromises and devices by which public authority will co-operate
with private initiative. But beyond this no obvious case is made
out for a system of State Socialism which would embrace most
of the economic life of the community. It is not the ownership of the instruments of production which it is important for
the State to assume. If the State is able to determine the aggregate amount of resources devoted to augmenting the instruments and the basic rate of reward to those who own them, it
will have accomplished all that is necessary. Moreover, the
necessary measures of socialisation can be introduced gradually
and without a break in the general tradition of society.4
4

Ibid., p. 378.

8 . Robert Lekachman
The quiet tone and the temperate words should not conceal a startling message, nothing less than the admission of the virtual bankruptcy of any version of capitalism which endeavors to limp along
as an old-fashioned private enterprise system. Keynes' own "capitalism" must make its peace with a state that exerts a "guiding influence"
on consumption and undertake on its own account a "somewhat
comprehensive socialisation of investment." Since in a closed economic system consumption and investment (private and public) are
the components of total spending, Keynes' capitalism must be translated as a system in which total spending is directly or indirectly
controlled by the state and the state's instrumentalities.
How is it possible to term such a resolution of economic impasse
"moderately conservative in its implications"? In part no doubt because the implied alternative to such policies is not old-style laissezfaire but a "system of State Socialism which would embrace most
of the economic life of the community." The rest of the answer is
inferable from Keynes' affection for much that capitalism achieved
in the past and from his persistent confidence that reasonable men
could combine new economic arrangements with old cultural values.
For few have plead capitalism's case more eloquently:
Let us stop for a moment to remind ourselves what these advantages are. They are partly advantages of efficiency-the
advantages of decentralization and of the play of self-interest.
The advantage to efficiency of the decentralization of decision
and of individual responsibility is even greater, perhaps, than
the nineteenth century supposed; and the reaction against the
appeal to self-interest may have gone too far. But, after all,
individualism, if it can be purged of its defects and abuses, is
the best safeguard of personal liberty in the sense, that, compared with any other system, it greatly widens the field of
personal choice. It is also the best safeguard of the variety of
life, which emerges precisely from the extended field of personal choice, and the loss of which is the greatest of all losses
of the homogeneous or totalitarian state. For this variety preserves the traditions which embody the most secure and successful choices of former generations; it colors the present with the
diversification of its fancy.1>
"Purged of its defects and abuses": there's the crux of the matter.
Can a system still recognizably capitalist endure after defects of inG

Ibid, p. 3 80.

Introduction' 9
come distribution, inadequate investment, excessive saving, and persistent unemployment are removed? How much individualism will
linger if personal incomes are equalized? How much independence
of choice will businessmen retain where government intervenes so
substantially in the investment decisions which are at the heart of
the entrepreneurial process? How much of that variety of life which
Keynes enjoyed for himself as much as he cherished it for others
can survive in the world of the giant organization?
In the perspective of this second, gloomier Keynes, the American
doctrinal situation possesses its ironies. Intelligent businessmen and
alert politicians have swallowed Keynesian techniques of monetary
manipulation and fiscal adjustment just about the time that these
measures give every sign of inadequacy. For increasingly the real
choices of public policy must be made among various versions of
private, publiC, and hybrid socialization. No socialist, President Eisenhower in a startling speech delivered just before the end of his term
in office alerted Americans to the dangers of undue influence from
the combination of military officials, defense contractors, and their
congressional allies who have acquired vested interests in the hardware of national security. Mr. Eisenhower feared that this increasingly effective coalition threatened to "involve the very structure
of our society." Occasionally truth issues from the mouths of ideological babes. For what the General was describing was the emergence in an uncomfortable form of Keynes' highly complex new
"compromises" by which public authority will co-operate with private
initiative." The defense-oriented industries-electronics, missiles, and
airframes-are government enterprises in everything but name and
profit distribution. The Communications Satellite Corporation, a
congressional creation of 1962, is a complicated mixture of the private
socialism of American Telephone and Telegraph and the public
socialism of the federal government. Inadequately, incompletely, and
muddle-headedly, the socialization of investment that Keynes favored
is indeed occurring and under a wide variety of auspices.
These speculations are not designed to deny that the main thrust
of the General Theory is in the direction of monetary and fiscal
policies. Still, it is odd that Keynes' interesting premonitions of our
world receive so little attention and that in the latest stage of Keynesian influence the version of Keynes which attracts the respect-

10

Robert Lekachman

able leaves the world more or less as it is. For everybody is a


Keynesian now: eminent conservative economists admit as much.
But what is badly needed is a theory to arm the vision of the second,
more radical Keynes. For neither Keynes nor anyone else has told
us how to socialize investment, invent wholesome "compromises by
which public authority will co-operate with private initiative," content all the social classes, and preserve at the same time the freedom
and variety of which capitalism at its best was capable. Until a
powerful theory does emerge-here is the task for the new Keyneseconomists will continue to advocate with decreasing relevance the
policies of the conservative Keynes. How Keynes might have laughed
to see the sight. How cheerfully he would have set to work to amend
or discard his own theories in favor of better substitutes.

PART ONE

The Man
The Theory

E. A. G. ROBINSON

John Maynard Keynes


1883-1946*
W B ARB TODAY TOO NEAR to Maynard Keynes to appraise
him dispassionately and to establish with complete detachment just
how high he stands in the company of the great. To measure him
must be the task of our successors, with the truer perspective that
time will give. It must be our task to put on record for them the
qualities which, to his own generation and to his pupils, made the
mind and personality of Keynes something wholly of its own; to attempt to capture and imprison in words the evanescent elements of
his make-up, so that they may picture, as nearly as they can, the
man as we knew him.
But even for a task thus limited, anyone person must feel conscious of his own inadequacies. For more than any man in our times
Keynes was, not indeed in a chameleon sense all things to all men,
but in a truer sense many things to many men. His tastes were
catholic, and he had the gift of achieving a knowledge of many subjects to a level that made him the respected confidant of those that
professed it. With philosophers, mathematicians, historians, bibliophiles, with the critics and exponents of modern painting and of the
ballet, just as truly as with economists, financiers, civil servants and
politicians, he could speak on a ground of equality of knowledge and
understanding. And because of this, what one man saw in Maynard
Keynes was likely to be something different from what another
might see.
2

In those delightful essays in biography which Keynes over many


years contributed to the Economic Journal he showed always a fascination in the stock from which the great economists had sprung. He
drew for us the eccentric Daniel Malthus, the tough evangelical
cashier of the Bank of England who fathered Alfred Marshall, Edge-

Written in 1946 and reprinted as written.

13

14 . E. A. G. Robinson
worth's gentle Irish philosopher father and romantic Spanish mother,
the "yeoman farmers owning their own land back to the sixteenth
century and beyond, turning in the eighteenth century into thrifty
parsons and scholars" that produced the Paley stock from which
carne Mary Marshall. It was natural that his interest should have extended to--perhaps even have grown from-an interest in his own
forebears. As a schoolboy he had delved deeply into the history of
the Keynes family; and of late years he reverted to his interest in
them.
In his own immediate generation Maynard Keynes was completely
a product of Cambridge. His father, born in 1852 and still vigorous
today, carne as an undergraduate to Pembroke College, Cambridge,
with a scholarship in mathematics after completing a full degree
course at University College, London. Deserting mathematics at the
end of his first year, he was Senior Moralist in 1875, held a Fellowship at Pembroke from 1876 to 1882, and was University Lecturer
in Moral Science from 1884 to 191 r. To later generations he was
best known as Secretary of the Local Examinations and Lectures
Syndicate from 1892 to 1910, and thenceforward to 1925 as Registrary of the University. Pembroke elected him to an Honorary Fellowship in 191 r.
In Neville Keynes' day, and much later, the study of economics
lay within the Moral Sciences Tripos, and formed an important part
of it. He had himself been examined in that Tripos by Foxwell and
Jevons. When he took his degree, Sidgwick and Fawcett of the older
generation were in the ascendant, though Marshall, in his first creative period, was working out in Cambridge the system of analysis
which was privately printed in 1879. Neville Keynes lay between
the two generations. His most important work was his Formal Logic,
published in 1884, and, after many editions, still selling a few copies
each year when the remaining stock of sheets was destroyed in the
London blitz. To economists, however, he is better known as the
author of The Scope and Method of Political Economy, a book
which is in essence of the older generation. But he was, not only in
age, but also by association, the intimate of Marshall and the new
generation. Maynard Keynes, in his biographical notice of Marshall,
recorded how his father had watched and deplored, as his pupil and
colleague, "Marshall's obstinate refusal to understand where his

John Maynard Keynes 1883-1946 . 15


special strength and weakness really lay." And Neville Keynes was
very much more than an interested spectator. He was a prominent
member of the Syndicate which in 1902-3 led to the setting up of
the Economics Tripos. He was Chairman from 1908 to 1919 of the
Special Board for Economics and Politics which governed the teaching and examining in the subject, and of which from 1910 to the
outbreak of war Maynard Keynes was secretary. Thus on his father's
side Maynard Keynes had the closest of associations not only with
Cambridge, but with Cambridge economics.
If we go back a generation further on the Keynes' side, we find
that Maynard Keynes' grandfather, John Keynes, came from a typical English middle-class stock, running his own business and taking
an active part in the local affairs of the old cathedral town of Salisbury, which the family had made their home for several generations,
having previously had long associations with Dorset. John Keynes
had succeeded his father as a manufacturer of high-grade brushes,
but handed the factory over to his brother when he found himself in
a position to devote his energies to his hobby, the cultivation of
flowers. This had been a passion from his youth; he used to tell his
friends how when a lad he pawned his watch to buy his first precious
plants. He built up in Salisbury a large and flourishing nursery garden in the days when modem flower-gardening was in its infancy; he
bred and exhibited many new varieties, and sold his prize dahlias
and roses, and later his vines, to the big houses that then abounded
in the west country. He married as his second wife Anna Neville, of
the Essex family of that name, and of that marriage John Neville
Keynes was the only child. The decision to send Neville Keynes to
Cambridge was taken on the advice of Henry Fawcett, the blind
Professor of Political Economy, afterwards Postmaster-General, who
came from a Salisbury family connected by marriage with the family
of Keynes.
Grandfather John Keynes died in 1878, a few weeks after completing his year of office as Mayor of Salisbury, by Victorian standards a well-to-do man. Maynard Keynes never plumbed to his
satisfaction the sources of that wealth. John Keynes had never put
money-making first. He had always been warm-hearted and openhanded. It may be that there was more money in those days to be
made out of horticulture and the laying out of rich men's gardens

16 . E. A. C. Robinson
than in our more austere and egalitarian times. Maynard Keynes
suspected, with what warrant I know not, that there had been an
element of judicious and fortunate speculation in land at the time
that the railways were pushing westwards, followed by good and
skillful Victorian investment of the proceeds.
If we wish to go back still farther on the Keynes' side, the family,
like so many English families, had suffered its vicissitudes. Of Norman origin (the name of Keynes has been traced to one Guillaume
de Cahaignes, who came from a village of that name between Vire
and Bayeux), the family had at one time widespread estates, to
which their name in many cases, as at Horsted-Keynes and MiltonKeynes, still attaches. Included in their estates was Tilton, in Sussex,
which Maynard Keynes many years later acquired as a country
home. In the sixteenth and seventeenth centuries the family, clinging to the Catholic faith, bred Jesuits of ability and distinction;
Father John. Keynes was a Professor of Logic at Liege. But the
Catholic adherence ultimately involved the final loss of the estates,
already much reduced by the failure of male heirs and the marriage
of heiresses into other families. The Keyneses, as recusants, fell on
hard times, until, as in so many other cases, they re-emerged as part
of the middle-class stock of England.
On his mother's side Maynard Keynes was almost equally a child
of Cambridge. Indeed, he often claimed that he was the first son of
a union of a Cambridge Fellow with a member of Newnham.
Whether this is literally true I do not know; certainly Mary Paley,
of an earlier generation at Newnham, had married Alfred Marshall
in 1877, five years before Florence Brown married Neville Keynes,
and there were other such matches, but none, I believe, produced a
son. The year 1882 was a great year for matrimony; the University
Statute which had required Fellows to resign on marriage had that
year been repealed.
His mother sprang from a background as typically British as his
father. Her own father, the Reverend John Brown, D.D., was minister for thirty years of Bunyan's chapel at Bedford. The life that he
wrote in 1885 is still the best, the most authoritative, and the most
readable study of Bunyan. Dr. John Brown was famous in his generation both as preacher and as historian-he wrote a number of
books on the puritan movement and the pilgrim fathers-and was a

John Maynard Keynes 1883-1946 . 17


President of the Congregational Union. His Doctorate in Divinity
was conferred on him by the University of Yale. Of his three daughters and three sons, Florence (Maynard Keynes' mother) was the
eldest. One of her brothers, Walter Langdon Brown, became, in his
last years, a very well-known ngure in Cambridge as Regius Professor of Physics.
Dr. John Brown had started life with few advantages. His father
had been a respected citizen of Bolton-Ie-Moors in Lancashire, but
of no special distinction. John Brown was one of those curious sports,
occasionally known to all of us who teach in universities, where a
family with no recent history of intellectual distinction suddenly
produces a man of great gifts. And beyond Bolton-Ie-Moors, the
family of Browns traces back farther to Scotland, and to the village
from which came one Agnes Brown, who bore and brought up that
odd sport-Robbie Burns. It is amusing to speculate whether the selfsame stock produced two sports-the poet, and the line that beginning with John Brown was to give us Maynard Keynes.
To his mother all biographers of Maynard Keynes will be greatly
indebted. Over the last twenty-nve years of his working life she has
kept full and complete cuttings of all that concerned him, and of his
youth has recorded so much as to make it easy for those who were
not his contemporaries to form a lively picture of him. To her care
and her recollections this note owes more than can be told. In Cambridge, as Justice of the Peace, Alderman and Mayor, Mrs. Keynes
has always been, and still remains, a ngure in her own right, and not
merely, or primarily, as wife and mother.
If one could picture the Keyneses and the Browns bringing their
diverse gifts to endow the infant, one might say that it was the
solid Keyneses who brought a precision of thought, a pleasure in detail, a practicality of outlook, perhaps also a pleasure in good living.
It was the more brilliant Browns who brought the sparkle and the
quick, intuitive leaps of intellect, the almost feminine perception of
the essential and inevitable pattern of things, which enabled Maynard Keynes to outpace in his rapidity of thought all those of us
whose intellectual movements are more pedestrian. But the infant
would himself have been the nrst to remind us that heredity is no
such simple matter. And, indeed, it was the curious conjunction of
contrasting qualities which made Maynard Keynes so essentially un-

IS E. A. G. Robinson
like the common run of mankind. There was, however, one thing in
his make-up on which he would sometimes dwell: he was, improbable as it might seem in one of his intellectual gaiety and nimbleness, utterly and completely English. And sometimes, in days of
stress, that Englishness would unexpectedly peep out.
3

Maynard Keynes was born on June 5, 1883, at 6 Harvey Road,


Cambridge-a tall Victorian Gothic house, which still retains many
of the William Morris influences of the time of his parents' marriage. It stands in a road familiar to many generations of Cambridge
undergraduates as the thoroughfare to Fenner's cricket ground. In
the same road lived the Stanfords, the Macalisters, the Glazebrooks,
the Besants, the Bougheys, the Clays. It was a characteristic Cambridge home of its period, in a characteristic Cambridge University
milieu.
As a young child, Maynard, the eldest of three children (Geoffrey,
surgeon, bibliophile and editor, and Margaret, social worker, pioneer
in the care of the aged, and wife of Professor A. V. Hill, physiologist and for long Secretary of the Royal Society, both survive him),
enjoyed to the full the advantages that such a home could give. He
was quick, intelligent, vivacious, but in no extreme sense an infant
prodigy. Throughout his childhood and school days he was thin and
delicate; sometimes inclined (as many of us who had known him
best in later life had now almost forgotten) to stammer. His pleasures were from an early stage pleasures of the mind, and his playthings ideas rather than toys. He followed the ordinary educational
paths of a boy of his class. He went at the age of eight to St. Faith's
preparatory school in Cambridge. Within a short time he was showing remarkable promise, more particularly in mathematics, and it was
no more than a fulfillment of confident expectations when he won in
1897 a scholarship at Eton-tenth out of twenty elected-a satisfactory, but not exceptional, performance, for which he had had but
little special preparation.
There were two great influences in the early shaping of Maynard
Keynes-his home and Eton. And great as was the influence of the
first, that of Eton was almost certainly the greater. Eton gave him the
confidence and poise to match his zest for knowledge. And Eton,

John Maynard Keynes 1883-1946 19


with its variety and lack of regimentation, was the environment in
which his manifold interests could happily develop. Though at the
end of his time at Eton he was predominantly a mathematician, his
first training was in the classical tradition; he was winning Iambics
Prizes and Division Prizes for Classics, as well as Mathematical
Prizes. Much later, in 1902, he was in the Select for the Newcastle. And when the time came for him to enter for a scholarship at
King's College, Cambridge, his Eton Scholarship, won for mathematics, was converted into an Open Scholarship, partly because the addition of Classics placed him ahead of one or more of ~e Open
Scholars. Throughout his time at Eton, moreover, history had a
fascination for him, and his essays and papers to the Eton Literary
Society on subjects of wide diversity-the character of the Stuarts,
Bernard of Cluny, the Difference between East and West-provided
an eagerly seized outlet for his creative and literary abilities, and
gave the first signs of his future gifts. But it was as a mathematician
that the young Maynard Keynes excelled. All the Mathematical
Prizes that Eton had to offer fell to him. For the two years before he
took the Cambridge scholarship examination he was working primarily, but by no means exclusively, at mathematics. Nevertheless,
in his last two halves, with the Cambridge scholarship behind him,
we find him reverting once more to his classics.
If at Eton Maynard Keynes' pleasures remained largely those of
the mind, he was not in the schoolboy world a mere sap or bookworm. At cricket he was no performer-"if the theory and history
(not the practice) of cricket were included in Trials I would do well
in the subject" was his own verdict. But at the wall-game in his last
year he considerably distinguished himself, despite his light weight,
for College against Oppidans in the annual St. Andrew's Day match.
As an oar he was tried, but, somewhat to his relief, rejected, in Trial
Eights, and rowed in Monarch-a position of dignified retirementin the Procession of Boats in his last year at Eton. His greatest schoolboy triumph-the greatest triumph open to an Etonian-was to be
elected to Pop. Yet, though all the distinctions which Eton could
shower on him were his, the picture of him that we get, particularly
from the letters written to his father by his Tutor, Mr. Gurney Lubbock, whose insight into Maynard Keynes' character and qualities

20

E. A. G. Robinson

was truly prophetic, was of a boy who managed to be little affected


by success:
No amount of marks and prizes seem to spoil the modest,
sensible way in which he accepts success. . . . Such simplicity
and such ability are not often combined.
And again:
I am sure Maynard will go far. He seems to have the power
of being interested in everything, and at the same time he
seems incapable of doing anything in a dilettante manner . . .
but even I, I confess, was fairly dazzled by the actual result. It
is an extraordinary performance. He certainly does command
success to an amazing extent; but then no one ever deserved it
better. His way of accepting it is characteristic; just as quiet,
frank and modest as ever, enhancing all the pleasure his success
gives one.
As one reads the record of Keynes' Eton days one almost forgets
the passage of the years. Yet that is in many ways the most remarkable thing of all. It is an effort to realize that Keynes, who was in
the 1930'S so utterly of the younger generation, so essentially a
Georgian, was being elected to College Pop in the same week in
which he watched the funeral of Queen Victoria.
Maynard Keynes arrived in Cambridge with a reputation. Even if
we make full allowance for the rose spectacles of hindsight, and for
a tendency to endow the undergraduate of yesterday with the glamour
of subsequent achievement, it was a remarkable generation in a remarkable college into which Keynes came in October 1902. The
older generation of fellows included the now fabulous figure of
Oscar Browning, round whose oddities and vanities a complete
Cambridge saga has grown up, and such well-known personalities
as Montague Rhodes James, later Provost of King's and finally of
Eton, W. H. Macaulay, the college tutor of Maynard Keynes' day,
W. R. Sorley, Sidgwick's successor in the Chair of Moral Philosophy, and the logician, W. E. Johnson. Among the younger fellows
were Goldsworthy Lowes Dickinson, J. H. Clapham, then at Leeds,
A. C. Pigou, E. J. Dent and H. O. Meredith. The students of
Keynes' time included J. T. Sheppard (now Provost), Harold Temperley, Oliffe Richmond, Will Spens, Stephen Gaselee, Gerard
Mackworth Young, Dillwyn Knox, Eric Milner-White, R. A. Furness (Maynard Keynes' peculiar intimate) and C. R. Fay, a scholar

John Maynard Keynes 1883-1946 2I


of Keynes' year, to name only a few. E. M. Forster had taken his
degree a year before. And Trinity, only a few minutes away, housed
such as Lytton Strachey, Adrian Stephen and Clive Bell. Duncan
Grant, Strachey's cousin, was a frequent visitor. R. G. Hawtrey, a
Wrangler of 1901, was to be seen in Keynes' room at an odd weekend.
Keynes quickly established in King's the ascendancy that he had
achieved over his contemporaries at Eton. In his third term Mr. Fay
remembers him as the first-year undergraduate spokesman at a large
meeting in the College Hall to protest against religious obligations
at the boys' club which the College was proposing to start in South
London. He soon became friendly, from this and other occasions,
with Lowes Dickinson, who through his life exercised a great influence on him, and with many of the other younger fellows. Then, as
now, there was a tradition of closer intimacy in King's between fellows and undergraduates than prevails in most colleges. And his
Eton associations brought him the friendship of Monty James and
his circle.
Maynard Keynes quickly made a reputation also in wider university circles. To the debates of the University Union Society he
devoted much of his energies. He spoke, for the first time, for a brief
four minutes, in the fourth debate of his first term, in favor of a
motion deploring party government. In the first debate of his second
term he was on the paper as opposer of a motion approving joint
action with Germany in Venezuela. (How many of us today have
even a glimmering of the issues involved?) He was first in the election to the Committee at the end of his third term. He was made
Secretary for the Michaelmas Term, 1904, and succeeded to the
Presidency early in his Vice-President term when H. G. Wood, the
term's President, had to resign on leaving Cambridge.
Those prominent in the Union in Keynes' time included not only
J. T. Sheppard, but also E. S. Montagu, President in Keynes' first
term, D. H. Macgregor, another ex-President, J. K. Mozley, J. E.
McTaggart, the Society's Librarian and Steward, Wilson Harris,
H. G. Wood, Stephen Gaselee, Arnold McNair, Alex Wood, C. K.
Webster, Walter Layton, Gordon Selwyn, Harry Hollond. A. C.
Pigou, yet another ex-President, was still joining in the free-trade debates of the day. William Temple came from Balliol to cross swords
with Keynes on the Government's domestic policy.

22

E. A. C. Robinson

At this time he was a forceful and pungent speaker, a little inclined when excited to stutter. The mellow charm of utterance that
belonged to his later days was not yet his. But the Cambridge Review
describes his speech on retirement from the Presidency, on the
motion "That this House views with impatience the prolonged existence of the present (the Balfour) government" as "in his best
style-cool, logical, and yet full of a regard primarily and above all
for the highest and best moral principles of statesmanship. We lose
no ordinary speaker in Mr. Keynes, but one who raises the tone of
any debate he may take part in"; and the less restrained Granta records that "his speech was cool and well reasoned, yet with just
sufficient passion to carry conviction." The written record is supported by the photograph which hangs on the wall of the Union
Society-a picture of a young man with burning and penetrating
eyes, a thin and angular face, and already the little bushy moustache
that he carried through the rest of his life.
The energies that Maynard Keynes devoted to the Union were,
at least in more distant retrospect, not misdirected. He had come to
Cambridge with a very considerable Eton reputation as a mathematician. But it fairly quickly became apparent that by the exacting
standards of Cambridge he was not in that narrow superlative class
which alone can hope to achieve fame in the field of pure mathematics. In his own year at King's, W. M. Page was his superior.
Keynes' mathematical interests lay rather in the borderland between
mathematics and the philosophical problems which were familiar to
him through his father and his father's friends. To say this is not
to suggest that the young Keynes was idle; he worked seriously, but
not with that all-exclusive devotion which gave to a familiar walk
to the bare outskirts of Cambridge the name of Senior Wrangler'S
Grind-the limit for which a senior wrangler might tear himself
from his exacting labors. It was still the day of mathematical coaching, though Routh and the famous coaches of the nineteenth century were now gone. Maynard Keynes coached with Hobson of
Christ's, Senior Wrangler of 1878 and a famous Cambridge figure,
but possessed of one powerful disadvantage for Keynes: he coached
at nine o'clock, and locked his door against stragglers at ten minutes
past the hour; early rising was never easy, then or later, to Keynes,
and more than once Hobson had to be cut.
Keynes' absorbing interest at this stage of his life was politics. He

John Maynard Keynes 1883-1946 23


had more than a schoolboy knowledge of political theory, and in
1904 he won the Members' Prize with an essay on ''The Political
Doctrines of Edmund Burke." He was fascinated by the problems
of government. In the Union debates all this gained a new reality,
and played a part of great importance in the shaping of Keynes' life
and ambitions. First, it made him look immediately to the Civil
Service rather than to an academic career as a pure mathematician
as an outlet for his activities. Second, it made him approach all problems, both then and later, as problems of government, in the making
of decisions: never, so far as I remember, did Keynes in late life devise an economic tool purely for its own sake rather than to solve
an immediate practical problem in the application to government of
the methods of economic analysis; his absorbing interest in politics
and government made Keynes, in the very best sense of those words,
a political economist.
But it would be wholly wrong to picture the young Keynes as a
solemn and prematurely adult undergraduate dividing his time between the higher Bights of mathematics or mathematical logic and
the solution of the manifold problems of world government. His
contemporaries recollect almost more vividly a lighter side: amusing,
often brilliant, sometimes chaffingly malicious, and always ready to
shock the less sophisticated product of the grammar school and of the
more conventional way of life. The friend who lived above him in
his freshman's year paints him thus:
Hospitable, receptive, intellectually impish, nipping up the
staircase with his latest ;I;>urchase from David, hurrying off at
five o'clock to McTaggarts philosophy lecture at Trinity, almost
as grown up as Gaselee, almost as good a bridge player as Spens,
admittedly the inferior of Furness in Rabelaisian wit, yet in the
sum, because he could do (or say) nearly everything, superior
to them all. If he couldn't find it at King's, there was Trinity
and Lytton Strachey. If it wasn't to be had at Trinity there was
the wide world without. This was the sense in which he was
the counterpart of O. B. (as O. B., of course, spotted) and far
enough from W. D. or the Brookie or the Monty James of
Eton and Kings'.1
lThe o. B. is Oscar Browning. W. D. is Walter Dumford, Fellow
and later Provost of King's. Brookie is A. E. Brooke, Fellow, later
Provost and Ely Professor of Divinity, uncle of Rupert Brooke. David
is the celebrated antiquarian bookseller who for many years kept a
stall in Cambridge market-place.

24 . E. A. C. Robinson
In the event Keynes was twelfth wrangler in a good, but not exceptional, year 0. E. Littlewood, now Rouse Ball Professor of
Mathematics at Cambridge, was 6rst)-a satisfactory result, but a
little disappointing to those who had formed high hopes of his mathematical abilities. This con6rmed his decision not to go on to Part II
of the Mathematics Tripos. For a time he toyed with the idea of
tackling the Economics Tripos, and with that in mind read for a
while pretty widely in the 6eld of economics. But in the end he
made up his mind to take the examination for the Civil Service, and
to postpone for the moment any attempt to write the Fellowship Dissertation, which was, then as later, a necessary preliminary to a
Fellowship at King's.
For most men entry into the Civil Service meant a year with a
London "crammer," usually Wren's. A very few dared the risks of
being an "auto-coach": Keynes, needless to say, ventured. And it is
little exaggeration to say that history has been changed by that decision. For it was while thus engaged in his fourth year at Cambridge
that he 6rst came into formal working contact with Cambridge economics. He decided to take in the Civil Service Examination, besides
mathematics, papers on political science, on ethics, metaphysics and
philosophy, on logic and psychology, and on economics. By way of
preparation for the last he went to some of Marshall's lectures, and
some scrappy notes of the lectures survive, covering a wide variety
of subjects; one is reminded of his comments, written many years
afterwards, on Marshall's methods of lecturing:
I think that the informality of his lectures may have increased
as time went on. Certainly in 1906, when I attended him, it
was impossible to bring away coherent notes.
More important, Keynes wrote, as did others who were going to
the course, the papers which Marshall (as Pigou also in a later generation) was accustomed to set, and Marshall, whose red ink comments
often submerge the author's original black, got a feeling of Keynes'
quality that was later to be important. His comment still survives on
one of Keynes' essays:
This is a very powerful answer. I trust your future career
may be one in which you will not cease to be an economist. I
should be glad if it could be that of an economist.

John Maynard Keynes 1883-1946 . 25


There were two posts only in the Civil Service list of openings
that Maynard Keynes was willing to take: one in the Treasury, the
other in the India Office. Somewhat to his own surprise, and that of
his family, he came out second; he was top in the English Essay, in
Logic and Psychology, in Political Science, second in a rather "Oxford" paper on philosophy, he did poorly in mathematics, and got his
worst mark in economics. On the last his comment is reported to
have been: "the examiners presumably knew less than I did," and
in one sense it was probably true. For he had grown up among the
economists or near economists of the Moral Sciences-his father,
Henry Sidgwick, Foxwell, McTaggart, C. P. Sanger, Sydney Chapman, and Pigou; and in his undergraduate years, which coincided
with the tariff reform campaign, his interest in and contribution to
the Union Debates were on political issues in which economics were
usually involved. What he did not know, or declined to reproduce,
was the set-piece answer which was, perhaps, expected by his examiners. His economics of 1906 were, in fact, mainly self-taught,
and it was only after a spell of service in London that he gave anything like the whole of his mind to economic science. Marshall's
lectures, the only ones which he ever attended as a student, did not
purport to give a systematic exposition of the whole subject and
Keynes never covered the field or submitted to the full discipline of
the Cambridge teaching of Marshall's day in the same thorough
sense in which Pigou or Macgregor or Layton did. To the end of his
days there remained parts of the field into which his interests had
never happened to lead him, and in which he was appreciably less
sure-footed than in those parts that he had made completely his own.
He went to the Civil Service, therefore, with an open mind about
his own future. Already, before the publication of the Civil Service
results, he was at work on a fellowship dissertation on Probability.
But it must be borne in mind that a Prize Fellowship could be held
concurrently with a Civil Service post, and imposed no duties of
residence in Cambridge.
4

In the Civil Service his second place meant that he missed the
Treasury and was appointed to the India Office. He would sometimes in later life describe his days in the India Office as devoted to

26 . E. A. G. Robinson
the study of The Times in the morning and to the writing of his
fellowship dissertation in the afternoon. But this, as anyone who has
known Maynard Keynes' regrettable incapacity for idleness might
suppose, is a caricature with no more than an element of verisimilitude. He was posted first to the Military Department, where he
began work within a few days of the publication of the results, in
October 1906. He found his work interesting, but not particularly
congenial, and from time to time insufficient to fill his day. In March
1907 he was moved into the Revenue, Statistics, and Commerce Department, with little routine but very much to interest him. By the
following July the Permanent Secretary, Sir Arthur Godley, whose
affectionate interest in the young Keynes throws a happy sidelight
on the Civil Service of that day, was drawing the attention of John
Morley, Secretary of State, to a minute by Keynes, and Morley in
turn was asking Godley to tell Maynard Keynes that he regarded it
as "a performance of admirable promise."
Meanwhile Maynard Keynes was making good progress with his
fellowship dissertation. To that he devoted most of his evenings,
and on that account he refused the offer, and additional pay, of a
resident clerkship, which Godley had pressed upon him, even when
Godley, whose interest in his winning the fellowship was manifest,
urged that the duties of a resident clerk were in the main so nominal
as to involve no interruption of his work.
His dissertation was submitted for the Fellowship Election in
March 1908. To the considerable vexation of Maynard Keynes himself and to the surprise of his friends, he was not elected. At this
moment of disappointment, Maynard Keynes received an offer from
Marshall of a Lectureship in Economics with a salary of '100 attached to it. For some years Marshall had found the money for two
lectureships from his own pocket. After his retirement in 1908,
Professor Pigou continued for some years to do the same. Keynes
consulted Lowes Dickinson, and got from him the advice that, if he
came back into residence and revised his dissertation, his election
the next year was almost certain. He decided to burn his boats. Sir
Arthur Godley and his immediate superiors were full of regrets at
lOSing so promising a member of their staff, but full also of understanding. They remained his close friends, and when he was later
studying Indian finance did all in their power to help him.

John Maynard Keynes 1883-1946 27


His short period-a little less than two years-in the India Office
was not wasted. He had learned there a great deal about the way
the machinery of government operated, and in particular to see the
problems of economics from the angle of the administrator, who
must make the decisions, as well as from that of the academic, who
must discuss the principles that must underlie the decisions. He had
formed an abiding interest in the problems of Indian currency and
finance. Above all, he had attained a maturity unusual at the age
of twenty-five.
He left the India Office in July 1908, and returned to Cambridge
in the Michaelmas Term, after a summer devoted to his dissertation.
At the Fellowship Election in March, 1909, the verdict, as had
been expected, was this time favorable, and his position in Cambridge and his college was now assured.
From 1909 to 1915 Keynes was mainly in Cambridge. But throughout those years (as also in his undergraduate and Civil Service
time) he spent much of his vacations in travel, most often with
Duncan Grant or the Stracheys; Scotland and the Orkneys, Italy,
the Pyrenees, Greece and Constantinople, Egypt (to visit his old
friend Furness, who had gone into the Egyptian Civil Service) all
saw him. Many of these journeys were, however, interrupted by illness. Keynes at that time in his life did not enjoy robust health;
early in 1914 he suffered a very severe attack of diphtheria in the
South of France.
Over the earlier years he was working not only at economics, but
also, somewhat intermittently, at mathematics. It was still as a possible successor to Sorley, in the Chair of Moral Philosophy, that
some of his youthful contemporaries pictured him; and while he
was preparing the Treatise on Probability for publication (it finally
appeared only in 1921), and even later, he found in the problems
of mathematical philosophy and his wrestlings with A. N. Whitehead, W. E. Johnson, Bertrand Russell, G. E. Moore of the older
generation, Ludwig Wittgenstein and C. D. Broad of the younger
generation, at once a more severe intellectual exercise and a greater
intellectual satisfaction.
During those six years down to the outbreak of war he was finding his feet as an economist, and making the beginnings of his rapid
ascent to fame. The lecture list of 1908-9 shows him lecturing, as

28 . E. A. C. Robinson
from January, on Money, Credit, and Prices, to a third-year class.
Mr. Fay, who later went to these lectures, has clear recollections of
their clarity of argument, breadth of comprehension and wealth of
historical illustration. Thanks not a little to Keynes (in later years
he used to claim that at that time he was teaching more than half
the economics students in the University), Part II of Marshall's
tripos was working well. Between 1907 and 1914, with no more
than a handful of men and women in each year-there were nine
men and three women in 19 I o-it had produced, among others,
Walter Layton, Hugh Dalton, Frederick Lavington, Harold Wright,
Hubert Henderson, Dennis Robertson, Gerald Shove, Claud Guillebaud, Lynda Grier-a level of talent which, despite the addition of
an elephantine rump, the post-1914 years have never equaled.
Keynes' first economic writing on any large scale, apart from his
lectures, was the essay which won the Adam Smith Prize (after his
election to a fellowship) in 1909, entitled The Method of Index
Numbers. While it is possible to recognize in this essay certain
elements which later appeared in Book II of the Treatise on Money,
the main argument of that brilliant treatment would seem to be of
later date.
His first further advancement as an economist was his appointment as Editor of the Economic Journal in 19II. To all except the
inner circle of Cambridge economists he was still relatively unknown. He had published only one major article in the Journalthat on "Recent Economic Events in India"-and some few reviews, of which that of Irving Fisher's The Purchasing Power of
Money is by far the most substantial and interesting. He owed his
appointment mainly to Marshall's confident recommendation. But
his relative inexperience (he was still only twenty-eight in a society
in which then, as later, grayheads predominated) dictated the appointment of an editorial committee-Ashley, Cannan, Chapman,
Edgeworth-with which Edgeworth, his predecessor, had not been
shackled.
Shackles of any kind were never very effective in imprisoning
Maynard Keynes. Despite his committee, he acted, as was to be
expected, on his own. Some eighteen months before his death, when
he and I were wrestling together in the middle of a wartime Atlantic
crossing with a huge bundle of manuscripts sent to us for publica-

John Maynard Keynes 1883-1946 . 29


tion in this Journal, he told me of the drawer full of manuscripts
which he had inherited from Edgeworth-manuscripts that Edgeworth had neither persuaded himself to accept nor to reject. They
included two manuscripts of the redoubtable Archdeacon Cunningham, who only a few years before had come so near to wrecking
Marshall's tripos. Edgeworth could muster neither the will to print
nor the courage to refuse; Keynes went firmly (then as afterwards)
through that drawer, and the two manuscripts went back-with a
polite little letter of apology-to the venerable and formidable Archdeacon.
In 1913 the Royal Economic Society asked him to take over the
responsibilities of Secretary as well as Editor. The Secretary's main
responsibilities are the collection of subscriptions, the extension of
the membership of the Society, the supervision of all the Society's
expenditure on the printing and publishing of the Journal and the
Society's other publications. When Keynes became Secretary there
were 576 members of the Society; when he handed over that office
to become President there were 4,619 members. To Maynard Keynes,
both as Editor and Secretary, loyally supported in the latter capacity
by Mr. S. J. Buttress, for many years his right hand as head clerk in
the Bursars' office at King's, this great growth was largely due.
In such leisure as remained to him from the preparation of a constantly increasing load of lectures-by the year 1913-14 he was
lecturing twice a week throughout the year on the Principles of
Economics to the third year, as well as twice a week on a variety
of monetary subjects-from his responsibilities for the Journal and
for the Royal Economic Society, and from his college and university
tuition, he was working on problems of Indian currency and finance.
His first interests in this had been formed during his period in the
India Office. His continuing studies, including his first article in the
Economic Journal in 1909 on "Recent Economic Events in India,"
kept him in touch with a number of his former colleagues, and when
the same problems rose to such urgency as to require the appointment of the Royal Commission on Indian Currency and Finance in
1913, Maynard Keynes, whose book, Indian Currency and Finance,
was by now already in type, was first approached to become Secretary,
and finally offered a place on the Commission.
Keynes' association with the Commission had a threefold impor-

30 . E. A. C. Robinson
tance. First, his own contributions to the problems at issue were great
and significant: not only does the Report itself show obvious indications of his influence; it includes a very substantial Memorandum,
specifically attributed to his authorship, on Proposals for the Establishment of a State Bank in India, which won high commendation
from Alfred Marshall. Second, the experience of the Committee
greatly fortified and consolidated what was to prove Keynes' great
strength as an economist-his power to fuse rigid economic thinking
with a great mastery of practical and administrative detail. Third,
and in the event most important of all, the Commission brought him
into the closest touch with a number of persons who were greatly to
affect his subsequent life. The Chairman of the Commission was
Austen Chamberlain; among its members were Sir Robert Chalmers
(later Lord Chalmers), Permanent Secretary of the Treasury, Lord
Cable, Lord Kilbracken, Lord Faber, all three leading figures in the
contemporary world of commerce and finance, and the Secretary
of the Commission was Basil Blackett of the Treasury, then and
for many years a close and valuable friend. to Keynes. From the time
of the Commission he had measured himself against the greater
world and knew his stature; he moved, thenceforward, with the assurance that such knowledge can give.
5

When war came in 1914 it was in a world very different from that
of 1939. The Civil Service of the day was appropriate to a liberal
world in which the main objective of government was not to govern
inadvisedly. The expansions of the social services begun by Asquith's
administration and associated with the name of Lloyd George had
not yet proliferated into the organizational mazes of today. There
had not been planned beforehand any scheme of expansion to meet
the needs of a world war; indeed, many believed that a world war,
as we have twice since known it, was beyond the endurance of the
delicate organization of a modern economic State. The Treasury
itself was a small and very select body of officials, expert not so much
in finance or economics as in the making of economic and financial
decisions based on the best available technical advice. And, if one
may believe only part of the saga told by those who served then in a
junior capacity, many of the senior officials were the ripe products

John Maynard Keynes 1883-1946 31


of an age which produced eccentric genius more richly than our
own, or in which the forces of a scholastic mass production were
less highly successful than today.
A day or two before war broke out Keynes was hurriedly summoned for consultation to the Treasury. The immediate problem
that confronted the Government was whether or not to suspend
specie payments. A Memorandum by Keynes was mainly responsible
for convincing Lloyd George, then Chancellor of the Exchequer,
that it would be best to avoid suspension.
But apart from this early call, Keynes' services remained for some
time unused. He remained in Cambridge, continued to lecture, and
occupied himself with the writing of two articles. The first, 'War
and the Financial System, August 1914," appeared in the Journal
of September 1914, and, making all allowances for the opportunities
of an editor, those twenty-seven pages must have been written within
a very few days after the outbreak of war. While it is in the main a
narrative, it is such a narrative as could only be written by one who
was immediately seized of the fundamental issues and in very close
touch with the effects of events upon particular institutions. The
second article, "The Prospects of Money, November 1914," carried
the story farther, and discussed particularly the immediate outlook of
gold and gold reserves; he was already speculating on whether the
war would lead to some international regulation of exchanges.
It was inevitable that the call should soon come for Keynes' services. In January 1915 he was invited to join the Treasury, to work
with Sir George Paish, who was Adviser to the Chancellor of the
Exchequer and to the Treasury on Financial and Economic Questions. While they were concerned with a very wide range of problems, Keynes' responsibilities lay primarily in the field of external
finance, and particularly inter-Allied finance. The system of Allied
War Loans, the controls over the use of such borrowings, and the
whole relationship of the financing of our Allies to our own internal
problems and to supplies of materials, was developed by Keynes. It
was his system which was later taken over almost bodily by the
Americans when they came into the war. And it was as arch-creator
of this system, designed much more with the immediate purpose of
rationing the several Allies in their claims on a pool of limited resources than to provide a record of mutual obligations, that Keynes

32 . E. A. G. Robinson
violently assailed it when, after the war, the cancellation of these
debts was resisted. For, from his special knowledge of the purposes
and circumstances of the loans he knew well that the accumulations of inter-Allied debts reflected neither the inferior war efforts
of the debtors, nor the greater efforts of the lenders.
But apart from the problems of inter-Allied finance, Keynes was
also concerned, in the first war as in the second, with the wider
problems of external finance, with the problems of obtaining particular currencies-a field in which his dexterity became legendary
-and with the short as well as the longer-term problems of achieving
a balance of payments. In all this Keynes quickly proved himself a
master, and his standing in the Treasury rose rapidly. He was constantly occupied, in the first war as in the second, in major financial
negotiations with our Allies, and accompanied the Prime Minister,
the Chancellor of the Exchequer, or the Governor of the Bank of
England to numerous conferences in Paris and elsewhere; he only
at the last moment was detained from accompanying Lord Kitchener
on his fatal mission to Petrograd. In 1917 he received a C.B. (Companion of the Bath)-an honor seldom given to a temporary member
of the Civil Service; he owed it to the strongest possible pressure from
the Treasury after he had, as was thought, been improperly excluded. And by 1918 he had reached the rank of Acting Principal
Clerk-the contemporary equivalent of what has more recently
been known as Assistant Secretary; in the Treasury hierarchy of the
day only the two Joint Permanent Secretaries were above him. Both
his standing in the Treasury and his particular knowledge of the
details of external finance made him a natural choice as financial
representative of the Treasury at the Peace Conference.
6

His arrival in Paris divides the phase of apprenticeship and academic development from the phase of controversy in which Maynard
Keynes spent the rest of his life. And none of those controversies is
yet sufficiently resolved for us to dogmatize as between right and
wrong. The mere facts of his share in Versailles are quickly told.
Keynes, after conducting the first post-armistice discussions with
the Germans in regard to foodstuffs and other supplies, was in Paris
almost continually from January to early June 1919. He was there

John Maynard Keynes 1883-1946 . 33


as Deputy for the Chancellor of the Exchequer on the Supreme
Economic Council, with full powers to take decisions. He was one
of the British Empire representatives on the Financial Commission.
He was Chairman of the Inter-Allied Financial Delegates in Armistice Negotiations with Germany. He was principal Treasury Representative in Paris. In an advisory capacity he was present at meetings
of the Council of Four-the charge that he was not present at what
he described, made then and repeated more recently, he refuted in
The Times of May 24, 1921. When (to quote his own words in the
preface to The Economic Consequences of the Peace) "it became
clear that hope could no longer be entertained of substantial modification in the draft terms of Peace," he resigned and returned to
Cambridge, and to freedom of speech.
There was, at the time and later, some uncertainty whether Keynes
had broken down in health or resigned. In fact, both are true. In a
letter to his parents on June I, 1919, he wrote:
Partly out of misery for all that's happening, and partly from
prolonged overwork, I gave way last Friday and tooK to my bed
suffering from sheer nervous exhaustion .... My first idea was
to return to England immediately, but General Smuts, with
whom I've been working very intimately for changes in their
damned Treaty, persuaded me that it was my duty to stay on
and be available if necessary for the important discussions of
these present days, declaring that one can only leave the field of
battle dead.... I dragged myself out of bed today to make a
final protest before the Reparation Commission against murdering Vienna, and did achieve some improvement.
Mrs. Millin, in her book General Smuts (Vol. II, pp. 255-7),
has given an account of a letter from Keynes to Smuts written
shortly after the former's return to London, and prints in extenso
Smuts' reply-omitting only the extremely important matter of the
date. May I quote a few sentences?
And now as to the future. I think it would be very advisable
for you as soon as possible to set about writing a clear connected
account of what the financial and economic clauses of the
Treaty actually are and mean, and what their probable results
will be. It should not be too long, as we may want to appeal to
the plain man more than to the well informed or the specialist.

34 .

E. A. G. Robinson

Mrs. Millin's account is that Keynes sent an answer (I quote


her, not him): "intolerable anguish and fury, he explained, had
compelled him to leave Paris. He could do at any time, and speedily,
what General Smuts proposed, for he had it clear in his mind, and
it only needed putting on paper."
Such, according to Mrs. Millin, deriving her knowledge presumably from Smuts and Smuts' papers, was the origin of the book.
Though I know of no evidence, either from Keynes' papers or from
what he has said to me or others, to verify this from Keynes' side,
I have no reasons for doubting Mrs. Millin's account in its essentials.
At the time Keynes clearly regarded General Smuts as one of his few
allies in high places in a fight for sanity. Mrs. Millin suggests that
Smuts would have gone considerably less far in deriding the Big
Four or in debunking the Peace. But maybe while Keynes was
primarily concerned with stating uncompromisingly what he believed to be the essentials of abstract justice, Smuts was more worried about the problems of statesmanship and politics involved in
setting the world straight.
Freedom of speech took first the form of a course of lectures delivered in Cambridge during the Michaelmas term of 1919 entitled
"Economic Aspects of the Peace Treaty." My principal memory of
them is of the dense throng and the fight to find even standing room,
for everyone was prepared to cut anything to hear Keynes; I was
then a classic, and was duly reprimanded by my tutor for surprising
lacunae in my knowledge of Cicero's Letters. But almost equally
vivid is my memory of the burning sense of the world's stupidities
which animated the lecturer. Those lectures appeared in a variant
form at the end of December 1919 as The Economic Consequences
of the Peace, and the world shared our excitement.
The reception of the book was at first surprisingly favorable, even
if one may suspect that many of the reviewers, as well as of the
readers, were more interested in Clemenceau's gray gloves than in
the economic consequences of Versailles. But it was not solely a joy
in political striptease which moved the reviewers. Very many of
them, both in England and in America, showed a far greater capacity
to comprehend the issues and to appreciate the strength of the arguments than one might expect in their successors of a quarter of a
century later.

John Maynard Keynes 1883-1946 35


The first powerful counterblast came in a very long and full review in The Times on Monday, January 5, 1920. If the salt of
Keynes' attack on nonsense at Versailles was to be found in the
pungent portraits which he drew of the "Big Four," The Times
returned blow for blow:
How came it, they may ask, that the man who could write
the pages of incisive portraiture, not to say caricature, that fill
the chapter on "The Conference," came to hold the position of
technical adviser to one of the most technical Departments of
State? How, unless his bias had been throughout akin to that
of the conscientious objector, could he place the Allies persistently on the same moral level as Germany in regard to the war?
... Indeed one of the most striking features of Mr. Keynes' book
is the political inexperience, not to say ingenuousness, which it
reveals. Yet he sits in judgment upon and condemns severely, as
statesmen and as men, the French and British Prime Ministers
and the American President. He draws portraits of them in
which only those who know his subjects more intimately can
distinguish the true from the false . . . Mr. Keynes may be a
"clever" economist. He may have been a useful Treasury
official. But in writing this book, he has rendered the Allies a
disservice for which their enemies will, doubtless, be grateful.

The Times having roared, a few-but surprisingly few-jackals


fell in behind. Nevertheless the charge that he was pro-German,
when in reality he was primarily pro-English and pro-World, attached to Keynes.
Among all the contemporary reviews, I find that by D. H. Robertson in the Economic Joumal of March 1920 by far the most penetrating. There are two passages in that review which are in a sense
as prophetic as anything which Keynes himself had written:
The incompatibility of the policies of crushing Germany's
economic life and of tapping her for indemnities is frequently
put forward, by those whose minds have a tilt in that direction,
as an argument for doing neither. But clearly such an argument
is invalid; the policy of the Carthaginian peace is not necessarily
wrong-headed because it may prove disappointing to the givers
of election pledges or even to the holders of just claims. Mr.
Keynes in an illuminating passage hints at the comparative indifference of M. Clemenceau to matters of reparation; and by
nothing which has been said hitherto is the Clemenceau attitude, which Mr. Keynes states with great detachment, proved

36 . E. A. G. Robinson
in itself illogical and absurd. To understand his own opinion,
that it "is not practically right or possible," it is necessary to look
at the general basis of his economic philosophy as set forth in
his Chapter II ("Europe before the War") and in scattered
passages elsewhere throughout the work.
Mr. Robertson was, I am convinced, right in seeing that the book
was rooted in Keynes' philosophical view of the fundamental nature
of the European economy. First, despite its successful weathering of
the war, it was a "delicate organization" (the first words with which
Keynes opens his description). Its restoration was therefore the
easier the less it was mutilated and changed, and the fewer the problems of adaptation enforced upon it. Second, while before 1914
(in Keynes' words) "the interference of frontiers and of tariffs was
reduced to a minimum" and "the various currencies, which were all
maintained on a stable basis in relation to gold and to one another,
facilitated the easy flow of capital and of trade to an extent the full
value of which we only realize now, when we are deprived of its
advantages," and again "over this great area there was an almost
absolute security of property and person," nevertheless the existence
and whereabouts of frontiers are to be regarded as a matter of first
importance. Mr. Robertson's comments go so deep to the roots not
only of the matters then at issue, but also of much of Keynes' subsequent work, that I must quote him in extenso:
Now the startling thing about this analysis of the economic
structure of Europe is that it is in some respects very different
from, and indeed diametrically opposed to, that of pre-war optimistic, free-trade, pacific philosophy, and resembles much
more nearly that upon which, consciously or unconsciously, the
edifices of protectionism, militarism and imperialism are reared
(question-begging words must be used for brevity's sake). To
begin with, political frontiers are not, in Mr. Keynes' view,
economically unimportant, but charged with significance, especially if they be multiplied. The transfer of a province from
one sovereignty to another is not an economic irrelevance; Governments do seek to impose barriers and prohibitions and confiscations of private property, and they can enforce them.
Secondly, the international division of labour between Europe
and the rest of the world is not, as by our fathers, hailed as an
unquestionable dispensation of Providence, but is seen to be
fraught with peril and instability. The 'devil' of Malthus is un-

John Maynard Keynes 1883-1946 37


chained, and reminds us that the mouths are many and the
loaves and the empty spaces few. Generalities about the interdependence of nations are not allowed to veil the prospect of
France and Italy and Northern Europe jostling against one another and against Germany in a rational solicitude to obtain the
lion's share of a coal supply which will not suffice for all. The
difficulty of obtaining an indemnity from Germany in noncompetitive goods becomes not a mere argumentum ad hominen: to cajole and fog ~he su~picious Britis~ II?-anufacturer, but
a SIgnal that the dommant mterests of Bntam and Germany
are competitive rather than co-operative. The two countries are
pictured as competing for limited iron and steel markets, for
inexpansive supplies of cotton and wool, for the exhaustible
sheaves of America and Russia-a very different picture from
the familiar pre-war one of two complementary countries, which
was a picture based on the extent of the trade between them,
and giving insufficient attention to its nature.
How, then, does Mr. Keynes, starting from premises which
would seem to point to the conclusions of a Chamberlain or
even a Cato, arrive at such opposite results? The answer would
seem to lie in two words-organisation and psychology. As regards organisation, his doctrine seems to be this: 'If being a
European is such a chancy business at best, let us not make it
worse by Hying at each other's throats. If coal is short, let us
be reasonable and distribute it properly. If productivity is low
as compared with needs, let us not lower it still further by tariffs
and trade wars, and the destruction, from nationalist motives,
of such organisation as exists.' He objects to the economic provisions of the Treaty because they strike at the existing methods
of organisation of an impoverished Europe, which has no
strength and no present breathing-space to improvise new ones.
Time has, I would suggest, gone far to vindicate Keynes' view
that frontiers are supremely important, and that the problems of
transfer across frontiers are sui generis, essentially different, both in
character and in the possibility of changes in quantity, from the
problems of supply and redistribution within a country. But I think
we should now, many of us, feel less certain of the other underlying
implication of Keynes' thought at this time-that, since the organization of Europe was delicate and resistant to change, it was therefore
best to seek to return to the status of 1913. That would imply that
the real changes of the war, the loss of manpower in all countries,
the loss of foreign assets and markets in the case of Britain, were

38 . E. A. G. Robinson
negligible as compared with the greater forces at work, and did not
make a return to 1913 as impossible of achievement as some other
objective.
The more powerful indictment of Keynes' position has come more
recently, and, as might have been expected, from a Frenchman. In
his book The Carthaginian Peace or the Economic Consequences of
Mr. Keynes, Etienne Mantoux has asked the question that in some
form or other we have all asked ourselves during the war: If our
treatment of Germany in 1919 had been different, could we have
escaped 1939? The gravamen of Mantoux's charge is that Maynard
Keynes' debunking of the peace-makers was the source of all subsequent evil. The peace was to be looked at as a unity. Woodrow Wilson, both in his Fourteen Points and in subsequent negotiations,
made self-determination supreme over the balance of power. Such
a fundamental change was untenable save on the condition of American willingness to underwrite the security of France. When American willingness was in the balance, Keynes made both the peace
and the peace-makers a laughingstock, gave impulse to American
desires to be rid of European entanglements, and effectively destroyed the whole security system. From that moment on, in Mantoux's view, both the re-emergence of Germany and the impossibility
in such an event of saving Wilson's newly created States in Central
Europe were inescapable, as Foch had pointed out in 1919.
One would like to have heard Keynes' own rejoinder to Mantoux,
for the charge was not wholly new, and he must already have considered it in his mind; Winston Churchill, in The World Crisis,
had argued that Keynes' political judgment was in 1919 greatly
inferior to his economic judgment. I cannot pretend to know what
Keynes would have answered. He had certainly said more than once
in my hearing during the last war that the fault of Versailles was
that it had failed either to be sufficiently Carthaginian or sufficiently
liberal. And he would certainly not have accepted any suggestion
that the world that emerged from Versailles was the world that he
was seeking to create. Would a world in which the full rigors of the
reparations clauses were retained in all their unachievable nonsense
have proved a better world? Could the Allies, in Keynes' silence,
have retained a unanimity of passion to destroy Germany economically? Would public opinion in America, Britain, and other countries

John Maynard Keynes 1883-1946 39


have tolerated the human consequences of an attempt to destroy
Germany economically? Would the logic of events not have brought
us sooner or later, and with even worse hatreds engendered meanwhile, to the same destination to which we came? Did Keynes alter
the shape of things, or merely make us face the ineluctable facts a
little earlier?
If we concern ourselves only with matters of timing, it would seem
almost impossible that The Economic Consequences of the Peace
should have been a primary cause of all that followed. The discussions of the Treaty in the Foreign Relations Committee of the
U.S. Senate began late in July 1919. The first and second Reports
of the Foreign Relations Committee, which proposed fundamental
amendments and reservations which would have effectively destroyed
the possibilities of the United States playing the part in Europe
that the French would have wished, were presented in September
and November 1919. Wilson's unsuccessful tour to win support for
the Treaty began early in September, and his illness followed immediately after. Unconditional ratification was defeated in the Senate
in November. All these events preceded the publication of the book.
It played some part, it is true, in the events of February and March
1920, when further attempts to reach agreement came to nothing.
And by then it was already abundantly clear that the United States
would ratify only with far-reaching reservations, if at all; and in all
this it was the Covenant of the League, rather than the muddles of
reparations, which were the predominant consideration.
For an understanding of Maynard Keynes' attitude to these issues
one must, I think, go back to certain fundamentals in his intellectual
make-up. He was-though it is difficult for those who remember
him primarily in the role of Cassandra to bear it in mind-profoundly an optimist and profoundly an idealist. He was an optimist
in the sense that he believed that very many of the world's evils
were remediable, if only the obstacles of human stupidity could be
removed; he hated stupidity not only with the esthetic hate of one
who sees in it a form of ugliness, but with the passionate, emotional
hatred of one who believed that it was depriving the human race of
so much that was most valuable. He was an idealist in the sense that
it mattered to him deeply and immediately what was the fate of
mankind; he was not prepared to tolerate wrong in the present in

40 E. A. G. Robinson
the belief that it might ultimately be mitigated. From beneath a
Georgian skin there peeped out from time to time an almost Victorian
sense of moral purpose and obligation; neither Eton, nor Cambridge,
nor Bloomsbury had obliterated wholly his heritage from generations
of Keyneses and Browns. And in all this Maynard Keynes was distinguished fundamentally, on the one hand from the cynicism that
insensibly permeates the thinking of any but the most pachydermatous of civil servants, on the other hand from the Hippancy which had
affected so many of the Georgians. The civil servant, regarding international negotiation as an ever-continuing process of give and take, is
apt to think his purpose achieved if in winning his prime objectives
he has not yielded too much on secondary objectives: if today's solution is nonsense, it represents, none the less, the best that is achievable today; tomorrow will demonstrate to the world that he was
right and others wrong, and thus tomorrow he can really get right
what the world is not ready to put right today. It was fatally easy so
to regard the problems of peace-making: frontiers are difficult to
change, let us aim first to get the best solution of frontier problems;
reparations are easy to change, if the world is not ripe for a sensible
solution let us have a nonsense solution and leave it to time to get it
right; a thoroughly bad and utterly unworkable plan may even be
preferable for the moment to a moderately bad and almost workable
plan. Keynes would never admit the initial hypothesis of the civil
servant-that public opinion that made policies possible or impossible
was itself immutable and impervious to reason. The dramatic swing
of public opinion, largely the result of Keynes' own writings, was to
demonstrate his rightness in this view.
He was equally removed from the Hippancy of the Georgians. It
would have been easy for him, as for them, to have escaped from the
whole confusion of political muddle-headedness and of time-serving
defeatism by retreating into the exercises of the intellect, into the
arcana of mathematical philosophy, or of the pure theories of economics. It was his passionate care for the fate of his fellow men and
his sense that economics shorn of the underlying realities is void that
forbade such an escape.
And it was this Victorian element of moral purpose which contributed not a little to the inHuence that he exercised, then and later,
over his colleagues and pupils. If one pursued as his devoted adher-

John Maynard Keynes I883-I946 4I


ent the solution of reparations, or currency reform, or full employment, it was because he had fired one, not merely with the thrill of
the intellectual's chase of knowledge, but with a powerful conviction
that thus only could the evils of the world be mitigated. It is a witness to the sense of purpose with which he infected us that, despite
his almost notorious faCility for improvising new solutions, I am not
aware that any who came under his influence ever for a moment suspected him of being himself under the spell of that nonsense syllogism that has so much bemused economics in recent years: I want to
be a great man; Lord Keynes is a great man; Lord Keynes always says
something that appears to be paradoxical nonsense; therefore I must
discover something that is paradoxical nonsense and say it. Truth is
too delicate a fabric to be best produced as a by-product of intellectual
vanity; it is those who care for truth who are most likely to find it.
7
From 1919 onwards Keynes never returned to the full load of
teaching that he had borne in Cambridge before the war. Younger
men, many of them his pupils, were becoming available to carry the
routine lecturing. Keynes' own share was a very limited amount of
lecturing-seldom more than some eight, remarkably inspiring, lectures in anyone year, concerned with problems at which he was
himself working-the supervision, in the Cambridge sense of weekly
essays, of some few of the best undergraduates doing economics at
King's, and his Political Economy Club.
During those early postwar years it was through the Club that
Keynes' influence was widest and most powerful. It was essentially
an undergraduate club. Dons, both economists and others who, like
Richard Braithwaite or Frank Ramsey, were interested in kindred
problems, might come. If Keynes had a visitor, more particularly a
visitor from abroad, he would bring him. But the papers in nine cases
out of ten would be read by undergraduates or young research workers-in those days researchers in economics were few, their distinction from undergraduates unimportant, and the Ph.D. unknown. To
the undergraduate of the early twenties, I can say from experience,
Keynes' club was fascinating but alarming. Fascinating because here
one heard Keynes, a large part of the Faculty, and all the best of one's
rivals discussing in realistic detail all the real and most urgent prob-

42 .

E. A. G. Robinson

lems of the world. Alarming because if one read a paper one was
likely to nnd one's undergraduate efforts (I speak from painful memory) being dissected by a visiting Mr. Hawtrey, destroyed by the full
power of Frank Ramsey's dialectical analysis, and when one had
maintained one's postion to the best of one's ability for some three
hours, Keynes would sum up in friendly but utterly devastating
fashion-I learned a certain sympathy with the prisoner waiting for
the judge's black cap. Alarming also because if it was not one's turn
to read the paper, one must draw a number from the hand of the
Secretary, and take one's turn on the hearthrug to discuss a paper on
a subject about which one might well feel an embarrassing ignorance
in the presence of some of the most critical minds of Europe. But a
wonderful training, because in Keynes' presence there were certain
forms of nonsense that one did not enjoy perpetrating once, and remembered for life not to perpetrate a second time.
Through his Club, Keynes knew intimately right down to his
illness in 1937 all the best of each generation of Cambridge economists, and exercised a more personal influence upon them than anyone else. The very great influence of Professor Pigou on the whole
technique of Cambridge economic thought in our generation was of
a rather different character-exercised less personally and more
through his writings and lectures. And through the Club we insensibly acquired certain elements in Keynes' own approach to the
problems of economics. In the early years his interests were almost
wholly in the practical problems of economic policy. I can remember
very few papers on purely theoretical issues, though we covered a
very wide range of questions. The choice was mainly our own, made
in consultation with the undergraduate Secretary, but our tastes were
in some measure the consequence of his.
My recollections of Keynes in the early twenties are of a formidable, and at moments somewhat intolerant, personality. But we were
his allies contra mundum. We could only be effective allies if we
learned to eschew nonsense. None of us would willingly talk nonsense in his presence. But if he was intolerant of nonsense, he was,
nevertheless, eclectic in his dislikes of nonsense. In the young, and
even more in the not-so-young, he particularly disliked some sorts of
pretentious nonsense. On the other hand, he could sometimes be surprisingly tolerant of honest stupidity. I well remember, two or three

John Maynard Keynes r883-r946 43


years later, when I had begun to teach, murmuring after a meeting a
word of apology for what I had regarded as a somewhat imbecile performance by one of my pupils; Keynes' rejoinder was: "I enjoyed it;
he said exactly what eighty out of every hundred Englishmen are
thinking."
The slight intolerance of those early postwar years disappeared, I
think, rapidly as time went on-it is difficult in retrospect to disentangle what were changes in Keynes himself from what were changes
in the awe that he inspired in the rest of us. I believe it would be
true to say that in those early years he had not yet fully achieved the
ease and confidence which subsequent recognition of his abilities
gave to him. But certainly he seemed to mellow over the later twenties, and became the easy and friendly person that later generations
of undergraduates have known. George Wansbrough, who was a
pupil of Keynes at King's in this later period, gave in The Times of
April 26, 1946, a happy picture of Keynes, seen as a supervisor:
In the middle twenties Lord Keynes was known to the general public chieRy as a merciless critic of the-only too plentiful
-economic errors and follies of the financial and political leaders of the period. Nothing could have been more unlike the
persona than the man as his pupils at Cambridge knew him. It
was to be expected that he should be tolerant of the youthful
shortcomings of his better pupils; but he was not less merciful to
the failings which would have provoked other teachers' fierce
and wounding criticism. Stupidity, pomposity and pretentiousness are no monopoly of the old; but even with undergraduates
who had no claim to intellectual distinction Keynes dealt with
such faults with a kindly restraint which was in marked contrast not only to the practice of other brilliant dons but also to
the scathing savagery which he showed to similar failings on
the part of leading figures. Truly he was a great teacher.
All of those who came in touch with him as undergraduates in
those days will have their own personal memories of his extraordinary
kindness in helping them. I myself have two vivid memories. The
first is of his kindness when I made a first appearance at his Club as
reader of a paper. I was not a King's man, and was feeling considerably overawed in the great man's presence. The second is of the
trouble he took with me, neither his pupil nor of his college, when I
developed a stye just before the Tripos, sending me to the right

44 . E. A. C. Robinson
doctor and the right oculist, and taking infinite pains to see that I
should not be handicapped by it.
But there was, curiously, another side to Keynes. I have said that
he hated stupidity, not only with esthetic but also with a moral
hatred: stupidity prevented the accomplishment of what was best for
the world. And it was for this reason, I think, that Keynes, even to
the end of his life, did not possess the power of living in wholly
amicable disagreement with those whose views on essentials differed
from his own. Many of those who were at one time his close friends
found themselves, because their views and his diverged, for a time at
least his enemies. In many cases the enmity was short-lived. Some of
his earliest opponents-among them Lloyd George and Winston
Churchill-became later his friends and allies. And Keynes was always one who, in retrospect, appreciated redoubtable opponents
above ineffectual allies. Others lingered unhappily in antagonism.
His rare lectures and the Political Economy Club did not exhaust
Keynes' contribution to Cambridge life. Ordinarily in the 1920'S he
spent three days in the week in Cambridge-Saturday, Sunday, and
Monday. But into those three days he managed to crowd as much
work as most men did in a week. He became Second Bursar of King's
in 1919. From 1924 to the end of his life he was First Bursar, and
over those years carried a heavy responsibility for the finances of his
college. The remarkable success with which he handled the college
finances, as also his own, is well known. He left King's with financial
resources to match, almost for the first time, its great architectural and
intellectual heritage.
Apart from his services to King's, he played an important part also
in the affairs of the University, occasionally in matters of university
finance, where his criticism and advice were valued; more continuously on the Faculty Board of Economics and Politics, of which for a
short period he was chairman, and in whose counsels he always
played a leading part. The recent establishment of a Department of
Applied Economics in the University for statistical and other applied
research was largely due to his efforts.
Great as were his services to Cambridge over these years, from
1920 onwards Cambridge could not provide a sufficient outlet for his
activities. The days in each week that were devoted to London
covered a range of activity even greater than the days at Cambridge.

John Maynard Keynes 1883-1946 . 45


He had continued during the years 1914-18 to act as Editor of the
Economic Journal, but when his responsibilities at the Treasury
greatly increased, Edgeworth, as Chairman of the Editorial Board,
helped to carry a share of the load. After the war, in 1919, Edgeworth was formally made joint editor; in 1926 D. H. Macgregor took
over from Edgeworth; and in 1934 I, in turn, this time as assistant
editor, succeeded to Macgregor. The existence of a second editor
somewhat lightened Keynes' responsibilities. Both Macgregor and I
were successively responsible for the reviewing of books and for the
preparation of the somewhat laborious back pages of the Journal.
From the time of his illness, 1937-8, I managed to persuade him to
devolve upon me the whole of the more mechanical tasks of seeing
the Journal through the Press. But as regards the all-important
articles, Keynes, even during his illness, was never for a moment
anything but the effective editor.
As an editor he was always prepared to take infinite trouble if he
thought a manuscript had the germ of something valuable in it. I
know of many cases in which Keynes' help and kindness gave a
young man just the start that was necessary, and produced an article
that did credit both to the author and to the Journal out of material
that many editors would have rejected-not infrequently out of material that other less discerning editors had already rejected. In the
case of articles outside his own special knowledge, he would on
occasion ask the help and advice of his friends, and particularly that
of Clapham and Pigou. But the ultimate decision was always his
own. The process of improving manuscripts cost him much time and
labor. He would often write very long and detailed criticisms and
suggestions to the authors, and in the end it was difficult sometimes
for even the author to know how much of what was printed was his
own and how much the article owed to Keynes' suggestions.
The work of the Economic Journal pursued him wherever he
might be-in Cambridge, in London, in later years at Tilton, even
around the world. He carried large masses of material about with
him, and devoted his odd moments to it. In Cambridge in my own
time the main decisions as to what was to go in, who was to be asked
to write certain things, and the general exchange of information and
views which a divided editorship involves, were apt to be made on
Sunday mornings. Even before his illness he always preferred to do a

46 . E. A. G. Robinson
morning's work in bed before getting up, and my picture will always
be of him sitting propped up in bed by many pillows, either in his
room in King's or later in his Rat in St. Edward's Passage, a bed-table
half across the bed, books, manuscripts, the Sunday papers scattered
about the bed, and on chairs or tables by his bedside. And (here I am
anticipating) after his illness Lydia forms an essential part of the
picture, letting me in, giving me a ration of time, slipping in and out,
and chasing me out when, despite my best endeavors to fulfill her
orders, Maynard had kept me gossiping beyond the permitted hour.
In writing of Keynes as editor I have inevitably run far ahead of
the years. In the 1920'S the Journal filled only a fraction of the time
that he was in London. There were two-or maybe three-other
main calls on his time besides economics: his business interests; writing and speaking on the current economic issues of the day; his literary and other pursuits. May I take them in tum?
Keynes' principal business interest in the early postwar years was
the National Mutual Insurance Company, of which he became
Chairman in 1921; among his fellow directors were his old friends
Walter Layton and O. T. Falk. Both to the problems of insurance
and of investment Keynes devoted a great deal of thought. His theories of investment were neither orthodox nor always acceptable to
more conservative minds. His speeches to the Annual Meeting of the
Company were often used by him as an opportunity for discussion
both of the general economic situation and the particular problems of
investment. Largely under Keynes' inRuence, the National Mutual
led the way in increasing the proportion of equities and industrial
shares generally that it was thought proper for such organizations to
hold. Around Keynes' theories of investment there have from time to
time been acute controversies. As to his Rair for investment there can
be no doubt. The prosperity of the National Mutual, of his college,
of the Royal Economic Society, of his own finances, all bear ample
witness.
At a later date he became for a time director also of the Independent Investment Company, originally founded to exploit certain new
theories of investment which he had helped to develop, and in later
years of the Provincial Insurance Company.
A quite different type of interest was represented by his chairmanship of the Nation and Athenreum weekly newspaper, and later by

John Maynard Keynes 1883-1946 47


his directorship of the combined New Statesman and Nation. At the
Annual Conference of the Liberal Party in 1922 the view had been
put forward that there was need for a first-rate Liberal weekly. The
Nation and Athenreum, despite Massingham's brilliant editing, was
financially in low water. A group of Liberals acquired it; Maynard
Keynes became chairman and Hubert Henderson editor. For some
years Keynes took much more than a paternal interest in it. He contributed not only over his signature, but occasionally (a thing he did
nowhere else) anonymously. And at least on one occasion he took
over the editorship for a short while when Hubert Henderson was on
holiday. The task of identifying Keynes' writings in the old Nation is
not yet complete. It is the more difficult since much that represented
also Keynes' way of thinking at that time is as likely to have come
from Hubert Henderson's pungent pen as from his own.
In later days his relationship to the combined New Statesman and
Nation was much more passive. He was content to leave to Kingsley
Martin almost the complete responsibility, seeking to exercise no
control or censorship, and even at one critical moment just before the
outbreak of war differing in essentials from the editor.
To tum, secondly, to his activities and writing on current economic
issues. The underlying theme of The Economic Consequences of
the Peace had been the urgency and the difficulty of reconstructing
Europe. Over the next years that was the issue to which Keynes was
devoting most of his thought. The problems were many: a satisfactory
reparations settlement; the internal problems of the loads of debt, and
closely related to it the practicable price levels of the different countries; and closely related to the latter, the re-establishment of a satisfactory system of exchange rates, and the levels at which the various
countries could hope to stabilize.
Those were the principal issues. Keynes attacked each of the problems in tum. And over the next few years he was prolific of writing
which, essentially ephemeral in intention, has, nevertheless, a more
lasting interest in the development of economic thought in the field
of monetary and trade problems. Of his books, The Revision of the
Treaty in 1922 carried on and brought up to date his indictment of
reparation policy, and A Tract on Monetary Reform in 1923 was his
first postwar book on the wider issues of the value of money, the exchange rates and the merits of devaluation. Once more Keynes was

48 . E. A. G. Robinson
ahead of his time. Devaluation was as inevitable as a scaling down of
reparations; the political difficulties were immense; in a world which
regards the Hexibilities of Bretton Woods as a modern straitjacket, it
is difficult to recapture in retrospect the sanctity of the pre-I914 rates.
But my recollection is that we who were then his pupils thought of
the Tract as a volume which did little more than conveniently bring
together ideas which he had for some time been disseminating.
The most important outlet for his views on European reconstruction was the series of special supplements of the Manchester Guardian Commercial Supplement, entitled Reconstruction in Europe,
which appeared under Keynes' general editorship from April 20,
1922 to January 4, 1923. They were of great contemporary importance; first, because they provided a forum in which Keynes, Cassel,
and others who were trying to work out the underlying economic
principles were publishing much of their thought; second, becausea triumph for Keynes' editing-the real experts from all over Europe
in the different fields of industry, finance, commerce and the like
were here engaged, if not always in cooperative thinking, at least in
sharing a single cover-for contributors included Prime Ministers,
Finance Ministers, heads of National Banks, heads of great international industrial and commercial concerns, as well as the world
experts in population, transport, exchange operation, whatever it
might be. Keynes' principal contributions to them included essays on
currency stabilization, on purchasing-power parity, on the forward
market in exchanges, on the Geneva Conference, on Russia, on inflation as a method of taxation, on the social consequences of changes
in the value of money, on population, on the possibilities of a settlement of reparations, on speculation in the German mark and a final
conclusion and summary on the underlying principles of European
reconstruction. Thus in many ways the Tract was no more than a
further statement of ideas that he had already here traversed.
Apart from his contributions to the Manchester Guardian Supplements, Keynes was writing or influencing writing elsewhere. He contributed over these years a large number of articles to the Manchester
Guardian on various questions, mainly related in the earlier years to
reparations; he attended the Genoa Conference in 1922, and contributed a series of brilliant articles on it to the Manchester Guardian.
By 1925 by far the greatest outstanding monetary problem of
Europe was the future of the pound. Already, in the Tract on Mone-

John Maynard Keynes 1883-1946 49


tary Reform, Keynes had made clear his view that Britain's right
policy was to aim first at a stable internal price level, eschewing deflation and wage and income reductions at all costs, and to follow a
policy of short-period but not long-period stabilization of exchange
rates (in his scheme of stabilization the Bank of England was to have
a temporarily fixed, but variable, buying price for gold comparable
to its temporarily fixed but variable discount rate). Meanwhile there
was increasing pressure to implement the Report of the Cunliffe
Committee, which had recommended in 1918-19 deflation and a
return to gold at the prewar parity. For many months in the early
part of 1925 the papers were overwhelmed by a spate of articles,
letters and reports of discussions on the wisdom or folly of returning
to the gold standard, with Keynes, or what Keynes had said, was believed to have said, or should not have said, a principal ingredient in
every argument. In his Budget Speech of 1925 Winston Churchill,
by now Chancellor of the Exchequer, announced that the embargo
on the export of gold would not be renewed after its expiry on
December 31, 1925, that meanwhile the Bank of England had been
given a general license to deliver gold for export against any form of
legal tender, as from April 18, and that a bill would be introduced to
place the Bank of England under obligation to sell gold bullion at a
fixed price equivalent to a return to the gold standard at the prewar
parity.
Keynes retorted with The Economic Consequences of Mr. Churchill, in which he depicted the likely consequences to British trade,
employment and industrial relations of a policy which he declared to
represent an improvement of 10 per cent in exchange by means which
involved a reduction of two shillings in the pound of wages and other
costs, and an increase of 1,000 millions in the wealth of rentiers.
In particular he picked out the coal industry as the principal victimto-be. The disaster to that industry in 1925 and the strike of 1926,
even if other causes contributed, provided a strong prima facie confirmation of his diagnosis.
With Britain's return to the gold standard the pattern of European
reconstruction came near completion. True, there were certain major
unsettled problems. In reparations the hard logic of events was gradually, through one adjustment after another, reducing nonsense
towards sense. Inter-allied debts were still unsettled. And in all this
the achievement fell very far short of the ideals that Keynes had set

50 . E. A.

C. Robinson

for himself and the world. Nevertheless, the year 1925 divides broadly
the years of reconstruction from the subsequent phases both of European economic history and of Keynes' personal life.
And I think the year 1925 was significant in Maynard Keynes' own
life in another very important sense. On August 4, 1925, he married
Lydia Lopokova, brought up in the Imperial Ballet School at St.
Petersburg, great dancer and even greater personality-as remarkable
in her world as Maynard in his. No ordinary woman could have provide4 as she did the perfect partner to Maynard. Lydia, herself a
great authority, assured in her judgment in those arts in which her
knowledge and experience were supreme, never attempted to rival
Maynard in those other arts that were completely his. And it was
under Lydia'S hand that the passionate missionary of the twenties
gradually mellowed and grew first into the scholar and creative
thinker of the thirties and then into the philosopher and statesman of
the forties.
8

If Maynard Keynes had died in 1925 it would have been difficult


for those who knew intimately the power and originality of his mind
to have convinced those who had not known him of the full measure
of Keynes' ability. In the twenties we who were his pupils of the
younger generation in Cambridge thought of him, indeed I suspect
that he thought of himself, primarily as one who had in unusual
degree the capacity to apply to the economic problems of the day the
corpus of economic thinking which he had inherited from Marshall.
There is more than a grain of introspection in the words of introduction that he wrote for the small Cambridge Economic Handbooks; in
1922 he was saying:
Before Adam Smith this apparatus of thought scarcely existed. Between his time and this it has been greatly enlarged
and improved. . . . It is not complete yet, but important improvements in its elements are becoming rare. The main task of
the professional economist now consists, either in obtaining a
wide knowledge of relevant facts and exercising skill in the application of economics to them, or, in expounding the elements
of his method in a lucid, accurate, and illuminating way, so
that, through his instruction, the number of those who can
think for themselves may be increased.

John Maynard Keynes 1883-1946 . 51


Above all, we thought of him as the perfect example of the thesis
that in economics what can be thought clearly can be said clearly. In
the twenties, it is true, he had made minor improvements to Cassel's
now unfashionable purchasing-power-parity theory of the exchange.
But his exposition of the factors determining the value of money
owed obvious debts to Marshall and to Irving Fisher.
It is difficult to say precisely where in time begins the development that was ultimately to give Keynes high claims to ,be included
among those who have made the most original contributions to
economic thought. I think it is most convenient, though not necessarily most accurate, to begin from 1925, when the problems of
unemployment began to succeed to the problems of financial reconstruction for priority both in practical urgency and in Keynes'
attention.
More important from the point of view of the development of
Keynes' analytical thinking, the beginning of 1926 saw the publication of D. H. Robertson's Banking Policy and Price Level, at which
he had been working during 1925. That book was the first to bring
home to us in Cambridge (I recognize that we may in important
respects have been insular, isolated, and behind the times) two
things that were fundamental to all the subsequent work on monetary and employment theory: first, the essential distinction between
the act of saving and the act of investment; second, the essential
interconnections between what hitherto we had regarded as monetary
phenomena-variations in the velocity of circulation, or in the desire
to hold money, or in the real value of the current income streamand the process of saving.
The rightness or wrongness of Mr. Robertson's 1926 thinking is
not here at issue; in many respects, I believe, his own thinking has
since then been modified. The important thing is that in the process
of writing it he had forced not only himself, but also Keynes, to reexamine a number of the fundamentals of monetary theory. He has
stressed in his introduction the extent of this co-operative thinking:
I have had so many discussions with Mr. J. M. Keynes on the
subject-matter of Chapters V and VI, and have re-written them
so drastically at his suggestion, that I think neither of us knows
how much of the ideas therein contained is his and how much
is mine.

52 E. A. G. Robinson
Apart from Mr. Robertson, it was probably Wicksell who exercised
most influence on these early stages of Keynes' thought. Keynes was
mainly instrumental in securing the translation of Geldzins und
Giiterpreise into English and its publication by the Royal Economic
Society. In the Treatise he makes considerable use of Wicksell's
ideas, and if he later abandoned them, it was with a consciousness
of having advanced with their aid. But apart from Wicksell his debts
to writers outside the English language were inconsiderable. He
read German indeed, but not with great facility, and he said on one
occasion that he had never comprehended through the medium of
German an idea that was new to him.
While there is little doubt that 1925 had seen the first planting of
the seeds of certain new ideas in Keynes' mind, it was some time before they germinated and bore fruit. The years 1926-8 are in
retrospect, and measured by Keynes' own standards, relatively barren
of academic output, though certainly not of activity or thought. We
find him concerned with many of the aftereffects of the return to
gold-the embargo on foreign lending, the failure, implicit in the
terms on which the general strike and the coal strike were settled, to
implement the declared policy of deflation and wage reduction-as
well as with more general problems of the treatment of monopolies,
and programs of national development. He was an active member
of the group of Liberals which produced in 1928 the Liberal Industrial Inquiry under the title of Britain's Industrial Future, and
several chapters in it were mainly drafted by him. Included in his
other writings in this period are his attempts to shift the Liberal
Party from its traditions of laissez faire and to steer it down a middle
course between doctrinaire socialism and uncritical individualism.
From 1927 onwards, however, Keynes was already working at the
Treatise on Money, which appeared in 1930, and the ideas which he
printed there were progressively developing. Each year from 1926
onwards he gave a short course of eight lectures, usually in the
Michaelmas Term, on the Theory of Money, and year by year the
form and content of his ideas developed. But as late as 1927-8 the
theoretical structure was, so far as I can discover, of an essentially
"velocity of circulation" design, though with important improvements
of his own. Much of the final version of the Treatise on Money is still
comparatively little changed from this earlier pattern; Book V, which

John Maynard Keynes 1883-1946 53


discusses velocities of circulation and distinguishes between savings
deposits and cash deposits, and the velocities of income deposits and
business deposits was being lectured in much the present form as
early as 1927-8. The fundamental equations of the Treatise appeared
for the first time, I understand, in Keynes' lectures of 1928-9. They
were still relatively new and relatively undigested even by the people
in Cambridge in closest touch with his work when the Treatise appeared in 1930.
Over the years from 1927 to 1930 Keynes' ideas were developing,
moreover, under outside influences as well as academic influences.
He was active at the Liberal Summer Schools, where he and others
were seeking to infuse new ideas. In 1928 he was advancing in the
Evening Standard plans to increase employment and prosperity.
More important, in the election campaign of 1929 (the year that we
now regard as the pinnacle of postwar prosperity) he was helping to
draft the Liberal plan to cure unemployment, and his own pamphlet,
written with Hubert Henderson, Can Lloyd George Do It?, is of
much more than ephemeral interest, since it first analyzes, by the
sort of methods that were later developed and improved, the cost to
the community of putting men to work. The publication of a White
Paper by the Conservative Government of the day, aimed to discredit the Liberal proposals, brought Keynes immediately into the
arena, protesting vigorously against what he regarded as a political
misuse of the Civil Service and attacking the arguments of the
Treasury Memorandum.
In the latter part of 1929, moreover, and the first weeks of 1930,
two tasks were imposed on him, both of which were of importance
to his later thinking. In November 1929 he was made a member of
the Macmillan Committee on Finance and Industry, and in January
1930 he was made a member of the newly created Economic Advisory Council. His developing theoretical ideas both influenced and
were influenced by the work of the Macmillan Committee, to whose
deliberations he gave for a time very much of his energies. In particular, it was on the Macmillan Committee that the important
theoretical issue as to how far reductions of money wages might
stimulate employment was first forced on his attention as a practical
issue. The pages of the Volumes of Evidence of that Committee, particularly his exchanges with Professor Pigou and D. H. Robertson

54

E. A. G. Robinson

when they appeared as witnesses, provide what is probably the best


published record of Keynes' thinking in that period. And among the
unpublished papers of the Committee there is understood to exist a
record of Keynes' exposition of his ideas to his fellow members, in
which he followed the main lines that he later published in the
Treatise on Money, but expanded many points in more detail under
cross-examination. This is of such great interest to the history of the
development of economic thought that some form of publication
would surely now be justified.
The Economic Advisory Council also provided a forum for the
exposition and practical application of Keynes' continually developing ideas. And it was later in the same year that he was asked to
advise, as a member of a Committee of the Economic Advisory
Council, on the serious economic depression that Britain shared with
almost the whole world. While the existence of that Report has. been
known, and the lines of its contents were freely rumored at the time,
it has never been published.
It is not easy to relate Keynes' political writings of 1929 to the
Treatise on Money, the preface of which is dated September 14,
1930. I think it would be true to say that very much of the Treatise
was complete before the election period, and that subsequent changes
were small, so that, despite the dates, the Treatise really represents
an earlier stage of a rapidly developing thought.
The Treatise on Money is in retrospect far more difficult than any
of Keynes' earlier works to assess. It contains very much that has
a permanent importance-in particular the brilliant discussion of
index numbers in Book II, which retraverses some of Keynes' earliest
work in 1909. Much also that he wrote on the applications of monetary policy has been little affected by the subsequent changes in detailed analytical thought. The fundamental equations of the Treatise
belonged, however, to a formative stage in Keynes' thinking. The
book will, I suspect, remain of permanent interest because it was here
that the savings and investment nexus made its first appearance in
Keynes' published work. But Keynes was still thinking primarily, as
had the creators of the various velocity of circulation theories, of the
factors which made prices go up and down. The fundamental equations of the Treatise belong to the theory of money. It was only later
that Keynes helped to break down the division which goes far back in

John Maynard Keynes I883-I946 55


the history of economic thought, at any rate in Cambridge, between
the theory of employment and the theory of money. Before the appearance of the Treatise, there were individuals, it is true, who had
seen, particularly in relation to the trade cycle, the limitations of
this dichotomy. Both Robertson, in some of his earliest work, and
Knut Wicksell had stressed this. But because of its existence it was
understandable that Keynes should at that date have visualized too
simple and direct a mechanism whereby changes of investment were
supposed to react on prices, and that in a strictly formal sense some
of his conclusions required the assumption that output was fixed and
not variable.
It so happened that the publication of the Treatise coincided in
time with a remarkable younger generation in Cambridge. R. F.
Kahn had just been made a Fellow of King's; J. E. Meade was spending a year in Trinity before returning to Oxford; they with P. Sraffa,
C. H. P. Gifford, A. F. W. Plumptre, L. Tarshis, Joan Robinson, and
several others of us formed a "circus" which met weekly for the discussion of the Treatise, and R. F. Kahn retailed to Keynes the results
of our deliberations. We managed quickly to seize on the formal
importance of some of the implied assumptions regarding employment and output levels and within a few weeks Keynes had not only
seen the force of our criticisms, as well as those of a number of
others both inside and outside Cambridge, but was dashing ahead of
us in redeploying his ideas in terms of a variable level of employment.
At the same time there was the problem, already hinted at in Professor Pigou's review in the Nation, of how, if saving exceeded investment, equilibrium was re-established-the problem, so to speak,
of where the savings went to, which was only ultimately resolved
when the necessary equality of savings and investment was recognized.
At no moment in his life, I think, did Keynes' greatness of character appear more strongly than at this time. He had just completed
what he might reasonably have regarded as his magnum opus. There
are few authors who would not have been sorely tempted to take up
a position and to use their authority and reputation to maintain it
until dislodged by sheer weight of opposition. Keynes never even
appeared to hesitate. He was off with the rest of us in pursuit of

56 . E. A. G. Robinson
truth with as enthusiastic a zest as if he were demolishing the work
of his worst enemy.
In the years following the publication of the Treatise Keynes
continued his practice of giving eight lectures annually, usually in
the Michaelmas Term. Each year he gave us the development of
his ideas to date. And these lectures became something wholly unlike
anything else that I have ever known in Cambridge lectures. Apart
from the third-year undergraduates, to whom officially the course
was addressed, there were to be found there the whole body of research students, at least half the members of the Faculty, a visiting
Professor or two from America, Australia or where you will, and on
occasion a few spies Cor should they be called in the modern jargon
"members of the fifth column"?) from London, Oxford, or elsewhere.
Gradually year by year the essential features of the General Theory
emerged. Thus, when it was finally published in 1936, Cambridge
at least knew fairly well what to expect.
To the world outside, however, the General Theory of Employment, Interest, and Money came as a shock. Keynes had meant to
shock. An ordinary book, wooing with sweet reason, would not have
fulfilled his purpose. The ideas that happened to be congenial might
have been accepted, the others discarded according to taste. Keynes
was firmly convinced, as he said in his preface, that "if orthodox
economics is at fault, the error is to be found not in the superstructure, which has been erected with great care for logical consistency,
but in a lack of clearness and of generality in the premisses." He had,
as he saw it, to force his fellow economists to decide whether in toto
they were for or in toto against. For if he was right in his analysis,
the basic assumptions of the automatism of the economic system were
at fault. Because he believed it necessary that economists should face
these issues, he deliberately2 heightened differences and emphasized
controversy.
This involved saying things that hurt about men that he had admired-including in their number Alfred Marshall. It involved also
that subsequent discussion was conducted very much in the atmosphere of the revivalist meeting: "Brother, are you saved?" I am not
2

May I draw attention to Keynes' own account of this in the Preface to


the French edition? Vide Economic Journal, April 1943, p. 159 (The
word "prepare" should read "precede.")

John Maynard Keynes 1883-1946 57


sure that this may not have been because Keynes himself had found
it a severe moral struggle to let go of his old liberal faith in the ultimate automatism of the economic system. But now that the Moody
and Sankey has come near to being orthodoxy, some of us can look
back at those years and smile at the odd and ever-changing divisions
between the sheep and the goats. The first to be saved seemed to include not only the pure of heart, but a large admixture of the empty
of head. The sinners were not only those who preferred to remain in
darkness, but many whose intellectual honesty forbade them to commit themselves to the support of conclusions so radically opposed to
all that they had learned and taught without a very thorough process
of sifting, and many others also who felt that the General Theory
contained exaggerations or misrepresentations which might discredit
what they still believed to be true and important in their own and
others' past thinking. The following years provided a most illuminating example of the processes and psychology of conversion, as economists, not only in Cambridge or in England, but all over the world,
adjusted their thinking to the new set of ideas that Keynes had put
before them, some of them going the whole way with him, some even
dashing ahead of him, others accepting in part, and remaining in
part unconvinced.
To the outside world the development of the theories of money
and employment by and around Keynes appeared necessarily as a
spasmodic and episodic process, though rumors of the developing
ideas, passed on by traveling research students, had made known that
the Treatise was not Keynes' final word. To those in Cambridge the
whole process seemed much more continuous, and both the Treatise
and the General Theory appeared much more as instantaneous
photographs of an athlete in action than as carefully posed studio
pictures with every limb perfectly set. Already, before the war,
Keynes was himself moving ahead again, as in his contribution to
the Irving Fisher Festschrift (The Lessons of Monetary Experience),
and was finding himself prepared to accept certain developments or
modifications of his ideas propounded by writers in England, America, and elsewhere. It would be a disaster, and wholly alien to
Keynes' own concept of the academic pursuit of truth, if the effect
first of the interruption of the war and then of Keynes' own death
were to be to establish for the first time a fixed and immutable creed

58 . E. A. C. Robinson
that it is wrong to criticize or change or develop. I can conceive no
memorial to Keynes' teaching that he would more passionately have
assailed.
9

Today we are all asking ourselves the question, "Where does


Keynes stand among the great economists?" I said at the outset that
I believe that is not a question that his own generation can hope to
answer. For Keynes' claims to greatness are so fundamentally different from most of the others whom we regard as the great men of our
profession. From Adam Smith to Alfred Marshall and beyond, the
great generation of liberal economists were building up a magnificent
technical apparatus of thought-they were essentially, though by no
means entirely, tool-makers. We connect with each of the great
names certain new tools that we use daily. Keynes was only in a
limited sense a tool-maker in the way that Alfred Marshall was.
Indeed, until the Treatise, as I have suggested above, he was hardly
a tool-maker at all. His major contribution to tool-making came with
the General Theory. It would, I would suggest, be proper to regard
the whole "savings and investment" nexus as a tool, in the sense that,
in thinking of the problems of underemployment, it is the apparatus
that most of us now use. Many who do not go the whole way with
Keynes-for instance, about the determination of the rate of interest
-accept the general "savings and investment" approach, even if they
do not fully accept the detailed refinements which Keynes introduced
in the General Theory. It is, however, by no means easy to say how
far we owe that tool to Keynes. He himself recognized many
predecessors; he recognized in relation to his own thought the very
great contributions of Professor D. H. Robertson in the early years
and his debts also to the Swedish economists. But I do not think that
it is necessary, if one would advance Keynes' claims to greatness, to
argue that we might not have reached the same destination by other
routes or at a later date; to name only one other, Michal Kalecki was
independently approaching the same goal. I think it would be almost
universally admitted that Keynes' development of the ideas and his
strong support and advocacy greatly advanced the date of their acceptance as a part of the orthodox corpus of economics.
One is on more disputable ground when one turns to the parts of

John Maynard Keynes 1883-1946 59


the General Theory that are more truly original and more completely
Keynes' own work-the more precise relations between saving and
investment, and the theory of the rate of interest. For here there is as
yet by no means so complete agreement. But, as regards the former, I
believe it to be true to say that a considerable majority of economists
would now hold that, in some sense a decline of investment reduces
activity and income progressively down to some level determined by
current investment and the willingness to save. Even if those who
believe that this process cannot be fully explained without resort to
such concepts as hoarding and dishoarding prove to be right, Keynes'
claims to have been the first to see the general pattern of the forces at
work would appear to be indisputable.
It is in relation to Keynes' theory of the rate of interest that dispute
is most vehement. There are many who believe that the differences
that still persist are verbal issues of precise formulation rather than
differences of substance. Time alone can show whether in the main
Keynes was right or wrong in supposing that the determination is
essentially a monetary phenomenon which does not afford any automatic link between saving at full employment and investment at
full employment. If further analysis were to show that there were
some such long-period link, then much of what appeared to our
generation most revolutionary in Keynes' thought might ultimately
be controverted. But even then it is not clear that, in relation to
shorter periods of time, and to monetary institutions as we know
them, Keynes' scepticism, based on the practical experience of our
time, as to the automatic adjustment of the rate of interest to the
desirable level may not prove to be both right and important. No
phrase of Keynes' making has been more often repeated than "in the
long run we are all dead." It was natural to one of Keynes' impatience to try to see whether the slow-working forces of long-term
economics could not, by the aid of human thought, be made less painful without being made less efficacious.
But Keynes should not, I think, be judged solely, or even primarily, as a tool-maker. He has three independent claims to greatness.
First, he relinked the analytical studies of academic economics to
the administrative problems of economic government at a moment
when they were, in Britain at least, tending to drift apart. In the
forties, after the close cooperation of war, we now take such a con-

60 . E. A. G. Robinson
nection for granted. But in the twenties it was far less evident. In his
earlier days Keynes was primarily an applied economist, in the same
sense that Adam Smith was, developing and using the techniques of
economic thought on the practical problems of the day. In Cambridge, at least, he taught our generation how to do this.
Second, we owe to Keynes more than to any of his contemporaries
in England the integration of the analytical and the statistical approaches to economics. While he took little personal part in the
minor refinements of econometrics, he followed closely the work of
the statisticians, and it was his natural inclination to approach any
problem from the angle of measurement of the phenomena. He was
for many years prominent in the affairs of the Royal Statistical
Society, and took a lead in the establishment of the London and
Cambridge Economic Service. My impression is that he owed more
of his own original ideas to the study of other men's figures than to
a study of other men's writings. He had a curious sixth sense about
figures, and would instinctively reject a figure which he felt could
not be right. His gifts of forecasting and his sense of orders of magnitude were sometimes uncanny. But just as he was sceptical of ideas
that could not be verified by measurement, so he was sceptical also
of the adventures of the statisticians into the worlds of correlations
built on insufficient logical foundations.
Third, and most important, Keynes insisted, both with himself
and with his pupils, in making us search out the assumptions which
underlay our argument. I remember him saying at a meeting of his
Club early in the twenties that the most interesting study in relation
to the classical economists was to discover the implicit assumptions
on which they were working. His own later work showed that in
very many cases we were implicitly assuming that more of one thing
was conditional on accepting less of another-a state that was not
universally, or even most frequently, true. It was, I myself firmly believe, a great step forward in economic thought when Keynes insisted
that we should have a general theory-a theory which was valid not
only with full (or near-full) employment, but also with unemployment-and that we should know quite clearly which of the propositions of economics were universally valid, and which were valid only
in conditions in which it might be true that an increase of one activity was possible only at the expense of another activity. In the

John Maynard Keynes 1883-1946 . 61


Cambridge thought of my time I believe that no single forward step
has been so important. It has gone right to the heart of the method of
estimation of the opportunity costs of doing anything. Before, we
were allowing ourselves to believe the paradox that it might be most
profitable to a society to allow resources which might have produced
something to stand idle. We came to see why that paradox was, in
fact, itself untrue.
10
The charge of inconsistency has been leveled against Keynes at
frequent intervals from 1920 onwards. No one can read deeply in the
literature and the newspaper cuttings which give the background to
his life without being impressed by the number of "Keynes' Plans"
and other proposals to deal with this, that or the other which he had
fathered in the past twenty-five years. But if one examines them more
closely, very many of them are designed with the same ultimate aims
in view. Indeed, it is difficult not to be impressed by the consistency
of his main strategic objectives: the full employment of resources;
the achievement of balances of payments for all countries by methods
that would not be inconsistent with full employment; as a means to
this, a system of exchange rates that would combine the short-term
virtues of fixity and predictability with the long-term virtues of flexibility; and, as a means to full employment, low interest rates. These
fundamental objectives will be found as clearly in The Economic
Consequences of the Peace and in the Tract on Monetary Reform as
in his later writings. It would be wrong to suggest that his worries
about employment began in the thirties. The pleas for conquering
unemployment were written in 1929, when we were still unaware
that that year marked the top of what came later to be regarded as a
boom, and were mainly conscious of the large residual volume of unemployment. And for many years before that the problems of employment had been actively in his mind, though less urgent than other
more immediate issues.
A careful study of Keynes will, I believe, show him to have been
remarkably consistent in his strategic objectives, but extraordinarily
fertile in tactical proposals for achieving them. Like a resourceful
tactician, he would probe, try to find the enemy's weak points; if
repulsed he would quickly fall back and regroup and put in another

62 . E. A. C. Robinson
attack elsewhere. But in the end his victory was complete. Full employment of resources has become the national objective; some
would say it has obscured other objectives. Ordered flexibility of exchange rates has become the agreed world system. Low interest rates
have become the official policy to the extent that former advocates
now begin to fear.
Keynes was consistent also, possibly overconsistent, in another
sense. He began life as a Liberal. He belonged by birth, upbringing
and natural abilities to the class for which the carriere ouverte is
the natural aim. He was not a stagnant Liberal; he sought to prevent Liberalism becoming a backward-looking and uncritical worship of laisser faire; he did not in any narrow sense regard the
criterion of the market place as final and decisive, and in an essay
in 1932 in The Political Quarterly he argued for the economically
unsound as the ultimate and distant ideal, to be reached through
the economically sound. But he belonged to the world of reason. He
was possessed of an inherent distaste for -isms and shibboleths and
emotionally reached views. He was never anti-Socialist, but he was
anti-doctrinaire. I do not think that the fashionable school of historians who believe that a glimpse of a man's passbook can explain
the whole of his character will have an easy task with Keynes. He
had made himself rich, it is true, in a Liberal world; but he had
nothing to fear in Socialism; he would have been a perfectly loyal,
happy and financially carefree Governor of the Bank and the Fund,
or whatever might come his way, in a Socialist world or in any other.
His objectives were included in their objectives. His political Liberalism was seldom obtruded in later life. I do not know (and I doubt
if anyone knows) what his final views would have been upon the
present moves towards nationalization. Neither, I suspect, wholly
unfavorable nor wholly favorable. Yet it was natural to him, in
trying to make a framework within which the world economic
system might develop, to think of it in the essentially Liberal terms
of markets and prices and discounts and borrowing rates. And when
the General Theory was interpreted as an argument for closing
national economic systems, or his pupils expressed scepticism whether
interest rates could ever provide a wholly perfect and effective working link between the volume of saving and the volume of investment in full employment, Keynes' fundamental belief in the Liberal

John Maynard Keynes 1883-1946 . 63


economy was apt to nnd itself in conflict with his passionate care
for his objectives. He never, I believe, satisfactorily resolved that
conflict for himself, and he has left it as a conflict for our generation.
II

In 1937 Keynes, who for some years had been in better health
than at any stage in his life, fell seriously ill with a form of heart
trouble difficult to diagnose, but ascribed to some poison in his system. The overwork of many years almost certainly contributed. For
some months he was completely laid up, and reduced his work to a
minimum; included in that minimum was always at least an oversight of the editing of the Economic Journal. While, after the period
of rest, he recovered somewhat, it remained necessary for him to
take careful precautions against excessive strain, and to conserve
his efforts for what was most important. The difficult task of restraining him from the impulses of his own energies and from the importunities of others fell upon his wife. We all learned the importance
of obeying her behests, if he was to carry on; but neither for him
nor for us was it easy. Over the last years before the war he was
very much of an invalid, and one had almost come to think of him
remaining as such for the rest of his days. He cut down permanently
a certain amount of the work that he had been doing in the past. He
resigned his Chairmanship of the National Mutual, and he spent
far less of his time in London, and more of it than in previous years
in Cambridge or at the house that he acquired at Tilton in Sussex,
where he was expanding his farming activities and was still able
to exercise a general supervision.
12

In what has gone before I have sought to portray the Keynes of


the prewar years as we in his own generation knew him. That
picture has inevitably been no more than a partial picture. In particular, I have not attempted to tell in detail of his services to the arts.
From his undergraduate days he had been a collector of books,
and in his later years possessed a very nne library, which he left to
King's. It came to represent a more or less complete collection of the
early editions of authors particularly important in the history of
thought. The philosophers predominated, but it included also a

64 . E. A. C. Robinson
remarkable collection of the work and papers of Isaac Newton, and
of works on alchemy bearing on the study of his manuscripts.
He had a small but interesting collection of modern paintings,
particularly of the French School, and for very long his close friendship with Roger Fry, Duncan Grant, Clive and Vanessa Bell, and
others meant that he was in close touch with the leaders of English
criticism and painting. His rooms in King's had been decorated for
him by Duncan Grant and Vanessa Bell.
His interest in the ballet went back to the early years in London
just after the first war. His more intimate touch with it began after
he first got to know Lydia Lopokova, and developed during the
following years, and with and through her he played a large part
in the growth of ballet in England. From 1932 he was Treasurer of
the Camargo Society, and helped as much as anyone to discover
the happy mean between artistic insolvency and a conventional popularity rewarded by financial success, and it was he also who helped
unobtrusively to achieve, in 1932, the union of the resources of the
Camargo, the Vic-Wells Ballet, the Rambert Ballet and others which
first gave us British performances of distinction. Arnold Haskell was
writing in that year:
I do not suggest that Mr. J. M. Keynes is the new Diaghileff,
he has other work which he might mistakenly regard as more
important, but he has succeeded with tact and energy in centralising the available talent, a gigantic task; and to that extent
the season is very much his.
Many economists the world over will remember, moreover, the performance in 1933 at Covent Garden in honor of the Delegates of
the World Economic Conference, which owed its conception and
organization very much to Maynard Keynes, and equally they will
remember Lydia Lopokova's delicious performance in it.
If in later years the ballet became better able to carryon without
Keynes' help, that was largely a result of his work during those
critical early years. In 1936 there was opened in Cambridge the Arts
Theatre which Keynes built himself and later transferred to a trust
representative both of university and town. He was chairman of
the directors, and while he found the perfect manager in Mr. Norman Higgins, he gave also himself much time to its problems, both
artistic and financial.

John Maynard Keynes 1883-1946 65


It was natural, therefore, that Keynes should be drawn into the
various national organizations that were created to foster and promote the arts. He was made a Trustee of the National Gallery and
was appointed Chairman of the Council for the Encouragement of
Music and the Arts in April 1942.
All these inevitably made calls upon Keynes' time. They represented, in fact, his hobbies, together with his farming plans at Tilton
and on the College estates in Lincolnshire, and, I suspect, his operations on the stock and commodity markets. How far did they interrupt
more serious work? I think, very little. For Keynes had, as I have said
earlier, a marked incapacity for doing nothing, for wasting time, as
most of us waste it, with worthless books and worthless thoughts.
Professor Schumpeter, in the charming and penetrating study of
Keynes that he has contributed to the American Economic Review,
has suggested that Keynes permitted his later work to suffer unduly
from haste and interruption, and that, instead of castigating Marshall
for his slowness, Keynes ought to have learned from him. That is,
I believe, to misunderstand Keynes' capacity for work. He was always
a very rapid worker who thought quickly, and did his best work
quickly. I myself would argue that all his best work was done very
rapidly, at white heat, and that the work which, whether in literary
form or in logic and lucidity of presentation, was below his best
very often represented overmuch work and a process of mental
development between the beginning and the end, as did both the
Treatise and the General Theory. True, in such cases he might at
the end have rewritten completely. But even to one of Keynes'
rapidity that would have been a long, as well as a tedious, task. The
development of ideas in the field of monetary theory was exceedingly rapid in the 1930'S. Would it have been better to postpone
each of those books by a year or eighteen months in order to rewrite?
Was it not better, in the words that he himself wrote of Marshall, "to
be willing to cast his half-baked bread on the waters, to trust to the
efficacy of the co-operation of many minds, and to let the big world
draw from him what sustenance it could"?
The example of Keynes' rapidity of writing that lingers in my
memory is so utterly appropriate that Professor Schumpeter will
forgive me for quoting it. One summer evening not many years
before the war Professor Schumpeter was to dine in King's with

66 . E. A. G. Rohinson
Keynes, and I was bidden to meet him. Shortly before he arrived I
brought Keynes the news that our much beloved Mary Marshall
was thought to be dying, and that it was believed that one thing
alone could save her-a message from us all giving her the call to
live. No one but Keynes could write it, and I besought him to do so.
With a word of apology, Keynes left me to entertain Professor
Schumpeter, and sat down for a moment at a writing table in the
crowded Combination Room. On half a sheet of writing paper, without an erasure, in no more than a couple of minutes, he wrote a
message that for beauty and simplicity could have been equalled by
no man that I know. Mary Marshall always declared that she owed
her recovery to that message. I wonder whether it remains among
her papers.
Nor do I feel certain that Keynes' speed of work was excessive.
When I first started to do research and was finding a beginner's
difficulty in getting my ideas on paper, Keynes urged me to follow
his own rule of life-to try every day to write three pages. He did
not always keep to this rule of life, but he was persuaded (and
surely he was right) that mental constipation was best cured by
regular writing. Richard Braithwaite has confirmed that Keynes told
him that his best work was written at the rate of about a thousand
words a day. Does that represent undue haste? Is the average speed
of fifty to a hundred words a day that most of us seem to achieve
a reflection of our superior scholarship or of inferior industry and
capacity? In measuring the rate of Keynes' work one must remember,
moreover, that from 1920 onwards routine lecturing played no part
in his life. There was practically nothing in Keynes' economic work
that was not ultimately destined, often after very substantial revision,
for publication.
That Keynes compressed much into his life is undeniable. It is
undeniable also that he suffered much from interruptions, though
I doubt whether his opportunities for continuous work were inferior
to those of the ordinary university teacher. He gave himself unstintingly to his friends and, whatever the load upon him, was always welcoming and apparently unhurried. All of us relied upon
him, consulted him, and made no major decision in our lives without
him. He loved to entertain, loved good conversation, to meet people
and to argue with them. Would it have been better if he had been

John Maynard Keynes 1883-1946 . 67


different, more of a hermit, more of an academic, and less a man
of many worlds? I think not. For Keynes surely brought to economics,
as did Ricardo, the critical wisdom of one who saw it in its relations
and applications to the work-a-day world. Certainly to his younger
pupils that was Keynes' compelling strength. And, if we think back,
we realize that for a generation it is Keynes who has first been asking
the questions that we have all been trying to answer-that almost
throughout the period current economic controversy has turned on
whether Keynes was right or Keynes was wrong. At different times
and at different places conviction, or sometimes, I suspect, even
fashion, has dictated that a majority should be his critics or his supporters. But economics has been about Keynes. In a sense we have
all been parasitic upon him. And that is the true measure of his
dominance.
13

When war broke out in 1939 Keynes was still a sick man. None
of us imagined it possible that he should play a part in this war
comparable to the part that he had played in 1914-18. Nevertheless,
within a few weeks of the outbreak of war he had developed ideas
of the scale of the British war potential and of the methods of war
finance which were later to form the basis of the whole operation
of the British war economy. The study of the problems of war potential he published in the Economic Journal of December 1939.
His proposals for financing the war were first read to the Marshall
Society in Cambridge and subsequently published, first in The
Times and later in the form of a pamphlet How to Pay for the War,
which had very wide currency.
This pamphlet had a twofold importance. First, it approached the
problems of war finance not from the orthodox, budgetary, angle,
but from the angle, implicit in all his work on employment problems,
of total income and total demand. Second, it included the proposal
that as a third source of war finance, additional to taxation and voluntary saving, we should call to our aid compulsory saving.
The second of these-the proposal for compulsory saving-naturally provoked discussion, and even antagonism. It was not finally
adopted until the Budget of 1941. But the first-the approach to
war finance from the angle of the national income-involved no

68 . E. A. C. Robinson
similar political problems. Some of us were already in Government
service, and Keynes' ideas were known to us, both directly and
through the paper to the Marshall Society which I, at least, had
heard. When How to Pay for the War was published we quickly
convinced our seniors that it was right to set up a small section in
the Cabinet Office to improve our national income estimates and to
make, with all the resources of government, the calculations which
Erwin Rothbarth had attempted, not wholly successfully because
of the limits of information, to make for Keynes. Within a few weeks
James Meade and Richard Stone were at work, and the first fruits
of Keynes' original idea and of their labors was available in time to
influence the Budget of 194 I.
But by then Keynes, in defiance of all expectations, was already
himself at work in the Treasury. When the Chamberlain administration was succeeded by that of Winston Churchill in summer 1940,
Sir Kingsley Wood became Chancellor of the Exchequer, and to
strengthen the Treasury he appointed Keynes and Lord Catto in
July 1940 as his advisers. They lived in adjacent rooms, and quickly
became so intimate that the door between them seemed to be perpetually open, and the irreverent of the Treasury soon came to know
them as Catto and Doggo.
In the Treasury, Keynes had this time no specific responsibilities.
He was there to help the Chancellor, and in practice to share some of
the heavier burdens with the Second Secretaries, on whom fall the
main responsibilities for the ordinary day-by-day business of the
Treasury. In the earlier years he was much concerned with the problems of internal finance, and there can be little doubt that the budget
speeches of 1941 and 1942 were considerably influenced by Keynes.
But increasingly Keynes became concerned with postwar problems.
The broader problems of internal war finance had by 1942 in the
main been reduced to order. In the British system the general
economic planning of the war was by then conducted almost exclusively in terms of the real resources, in manpower allocations, in
raw materials allocations, in shipping allocations, in building allocations, and the like; the needs of civil consumption were taken into
account at that stage, and the problems of war finance increasingly
became, as Keynes had argued that they should, those of seeing that
taxation and savings mopped up so much as possible of the redundant

John Maynard Keynes 1883-1946 69


purchasing power, and that inflation was thus kept within as narrow
bounds as possible. This is not to belittle these problems-they were,
indeed, tremendous-but others in the Treasury were in competent
control. Keynes' preoccupations increasingly became those of foreseeing and planning the problems of the transition from war to
peace, whenever that might corne.
But his influence was felt throughout the Treasury and was in no
sense confined to the particular preoccupations of the moment. One
of his colleagues in the Treasury has described his labors thus:
Despite the physical frailty from which Keynes was still suffering as a result of his illness, he characteristically took all the
Treasury's financial and economic activities as his province. He
sat in daily at the Treasury for long hours. He saw every important paper that the Treasury produced or that was wandering through the Treasury. Many he put away for reference,
and out of any heap of papers in one of his locked drawers
Keynes could unerringly pull out the one he wanted. He irrupted frequently and unexpectedly into major questions such
as the use of man-power, the necessity for allowing no waste of
resources, the problems of taxation, such as "pay-as-you-earn,"
income-tax credits, and the work on the statistics and presentation of the national income. In his later years he paid particular
attention to the balance of payments.
He read official papers voraciously and always remembered
what he wanted to remember, though he was not incapable of
inventing what he asserted to be a fact. He accepted all these
burdens not only as a good and unselfish colleague, but for
deeper reasons. As is said at an earlier stage of this essay, Keynes
was specially interested in the application of economic theory
to the practical problems of Government. He was increasingly
impressed with the necessity for allowing the whole stream of
official work to flow through his consciousness if he wanted to
produce any economic plan whose structure would resist informed criticism. Although he had been thinking and writing
about schemes like the Clearing Union as far back as I938, the
plan that he produced for public discussion was reinforced by
his knowledge of the actual situation confronting the United
Kingdom. Similarly, his more tentative proposals for international commodity policy derived much from the experience of
bulk purchases during the war and from the anxieties created
by the often uncontrolled rise in commodity prices in many of
the supplying countries.
But it was in his presentation of the case for the American

70 E. A. G. Robinson
Loan that his journeyman's work at the Treasury showed its
full achievement. There was no figure that Keynes used at
Washington whose application to the present and future life of
the English people he could not illustrate from his knowledge
of the practical details of government and administration.
Conferences in Keynes' room at the Treasury were seminar
classes in adult education and his new, but middle-aged, pupils
at the Treasury experienced something of the same sense of
excitement as the younger men who had worked with Keynes
when he was a teacher at Cambridge.
Even on the many journeys he made abroad he took care to
receive and read regularly the most important of the current
papers reaching the Treasury. The feel and touch of things was
to him a large part of the area of government.
In the course of 1940, moreover, the transition of Keynes from
outside critic, outspoken, fearless and therefore disreputable, to inside adviser and maker of policy, respectable and acceptable, had
been advanced an important stage by his election to the Court of
the Bank of England. His own immediate and irreverent comment
was: "I am not sure which of us is being made an honest woman
of-the Old Lady or me." But, with his quick power of absorbing
the atmosphere of an institution so soaked in tradition as the Bank,
the arch-critic was soon its most loyal and sincere defender, and no
appointment, I suspect, ever gave him more real pleasure. The
transition was furthered when Maynard Keynes was raised to the
Peerage in the Birthday Honours in June 1942, becoming Lord
Keynes of Tilton, in the County of Sussex; and the Government,
when he broke a long period of silence in May 1943, gained a powerful spokesman in the House of Lords, who possessed all the special
knowledge of a Minister and none of a Minister's departmental
responsibility. The College of Heralds appended to his Arms the
very appropriate punning motto "me tutore tutus eris." In the following March, the Borough of Cambridge elected him to the dignified medieval sinecure of High Steward, a successor to a great
line which included Sir Francis Bacon, Oliver Cromwell, Lord
Macaulay. And if he had not died, he was to have received the
signal honor of the Order of Merit. Truly had Cassandra become
the Delphic Oracle.
There was one other honor that came to Keynes which he prized,
perhaps, even above all the rest. In 1940 he was elected a Fellow
of Eton as representative of the Head Master, Lower Master, and

John Maynard Keynes 1883-1946 71


Assistant Masters. From the first, at the deliberations of the Provost
and Fellows he made his influence felt. His encyclopedic knowledge
and his capacity to frame at a moment's notice a complicated resolution made him a valued member of their body. Whenever he was
in England he liked to attend their meetings, and to spend the night
in Eton so as to keep touch, as far as he could, with all that was
happening there. Eton had to the end a special place in his heart.
Meanwhile it was inevitable that Keynes, with the double advantage of great inside knowledge of our financial problems and
freedom from routine responsibilities, should be used in some degree
as a kind of financial ambassador. He went to Washington as early
as May 1941 to discuss Lend-Lease and other financial problems
confronting the two countries. But his period of almost continuous
service in this capacity began in 1943, and was primarily, but by
no means exclusively, concerned with the great problems of financial
and economic reconstruction upon which he had been working for
some time past.
Keynes' greatest endeavors in this field, his contributions to Bretton Woods, to the Lend-Lease settlements and to the American
Loan, are still, and will long remain, in the zone of controversy. It
is no part of my purpose to attempt to show where Keynes was or
was not right. Final judgments, both on these issues as a whole and
on Keynes' part in them, must wait, not only on a longer and truer
perspective, but also on the publication of much that still remains
unknown. My purpose is rather to give some more human picture
of Keynes' share in all these gigantic labors.
His first major contribution in wartime to this group of problems
was his share3 in the development of a scheme for an International
S

That it was a share and not sole authorship will be plain to anyone
who knows the methods of Civil Service criticism and collaboration.
But may I quote Keynes himself CHouse of Lords, Official Report,
May I 8, 1943)? "To associate it too closely with a particular name is,
I venture to say, to do it an injustice. It has been the subject of intensive criticism and of progressive amendment and the final result is the
embodiment of the collective wisdom of Whitehall and of experts and
officials throughout the Commonwealth." This is not the only matter
in which outside critics would seem to attach to Lord Keynes or to
other individuals a personal responsibility which is cOlI!pletely alien
to a system in which many civil servants necessarily collaborate and
in which ultimate responsibility must rest with a Minister and with
the Cabinet.

72 . E. A. G. Robinson
Clearing Union (often called the Bancor or the Keynes' Plan), first
published in April 1943 almost simultaneously with an American
variation (known as the Unitas or White Plan).
The objective of the "Keynes' Plan", as of the other also, was to
secure the middle way that he had earlier pursued in the Tract on
Monetary Reform between exchange fixity and leaving exchanges
completely to the short-term forces of the market and of the political
and financial policy of individual countries. It was an essential part
of the Keynes' Plan that the responsibility of relative adjustment,
where the existing exchange and price relations did not provide a
basis for equilibrium, should be laid on creditors as well as debtors,
so that the pressure on the debtor to deBate and contract activity
and incomes should be mitigated.
The first stage of the process of attempting to reach wider international agreement was the examination of these proposals by the
experts, not only of Britain and the United States, but of certain
other countries also. This task, which occupied Keynes for several
months during the later part of 1943 and was finally completed
in the spring of 1944, resulted in a compromise scheme in which
the Keynes Clearing Union had given way to the White conception of a Fund, held in the currencies of the various participants-a
sort of international equalization fund-but in which other elements
of the Keynes Plan, which from the British point of view seemed
essential, had survived. The fruits of these labors were presented to
the British public in a fresh White Paper-A Statement of Principles for an International Monetary Fund-which included both
the Joint Statement agreed by all the experts, and explanatory notes
by the British representatives, making plain, in view of current
British criticism, that the substitution of a fund for a clearing union
now made unnecessary an international currency, that adequate arrangements were included to give elasticity of exchange rates, that
the scarce currency clause might be expected to have a powerful
effect in placing the onus of relative adjustment on creditor as well
as debtor, and finally that the difficulties of transition were provided
for.
It was inevitable that these modified proposals should represent
a compromise which would afford something less than the best
possible for Great Britain individually, were it possible to enforce

John Maynard Keynes 1883-1946 . 73


on the world what suited Britain. But it was of the essence of these
proposals that the combination of flexibility of exchange rates in
conditions which justified flexibility, with rigidity in conditions which
did not justify changes, should involve some loss of complete national
sovereignty in these respects. The gain of limiting the sovereignty
of others was to be bought only by compromise and by limiting our
own. Keynes always believed that sovereignty was retained in the
one matter where it was absolutely nec~ssary, the right to follow
an internal economic policy consistent with full employment. He
defended the proposals with eloquence and vigor in the House of
Lords and effectively trounced those who accused him, the archcritic of gold, of having re-established gold:
Was it not I, when many of to-day's iconoclasts were still
worshippers of the Calf, who wrote that 'Gold is a barbarous
relic'? Am I so faithless, so forgetful, so senile that at the very
moment of the triumph of these ideas, when, with gathering
momentum, Governments, Parliaments, banks, the Press, the
public, and even economists, have at last accepted the new doctrines, I go off to help forge new chains to hold us fast in the
old dungeon?
But it was clear from the first that the proposed scheme touched
on problems of commercial and tariff policy and might affect the
systems of imperial preference and of sterling area relations that had
come to be regarded as the safeguard of the British economic position in the thirties, as well as the tariff systems of the United States
and of the British Dominions. It was natural that many critics should
feel that, so long as future policy in these fields was uncertain, hardand-fast commitments in the field of monetary policy were in some
sense premature. But they were inclined to forget, what Keynes
was never in a position to forget, that by Article 7 of the Mutual
Aid Agreement we had already incurred obligations, at a time when
American aid was vital to us, to make provision for agreed action
to the elimination of all forms of discriminatory treatment in
international commerce and to the reduction of tariffs and other
trade barriers.
While the English critics were inclined to regard the monetary
settlement as something standing by itself, and Article 7 as an in-

74 . E. A. G. Robinson
cident in the past with only academic connection with the projects
currently under discussion, the Americans never shared this view,
and many disillusionments on this side of the Atlantic were due to
the fact that this was not sufficiently appreciated.
The new compromise proposals that emerged from the discussions
of the experts at Washington were referred to a huge International
Conference, representative of forty-five nations, to be convened at
Bretton Woods in New Hampshire. But before it met another event
of moment had occurred in Keynes' public life. In May 1944 the
wartime Coalition Government published the White Paper on Employment Policy which was designed to shape its postwar plans in
this very important field. The White Paper was a triumph for Keynes
in a twofold sense. First, it represented the last stage in the progress
of the general ideas that Keynes had developed in the thirties from
the phase of the solitary voice in the wilderness to the phase of
orthodoxy. Keynes had often complained that it had taken well over
half a century for Adam Smith's ideas to move from the study to
the floor of the House of Commons. His own had made the passage
in less than fifteen years. Seqmd, the White Paper itself was only in
a very minor degree his own work. I suspect that if seventy scholars
spent seventy years trying to discover the authorship of that White
Paper, we should have seventy, or more, different answers-it was
the most completely typical piece of Civil Service collaboration, a
jigsaw of parts of many versions by many hands. But it showed,
more than anything else, how widely Keynes' ideas had now become
dispersed. Keynes would, I suspect, have been the first to urge that
the publication of a White Paper does not solve a problem-that full
employment will come only if people will make it come.
The vast congeries of conflicting personalities which assembled
at Bretton Woods provided for Keynes the greatest test of his public
life, and in the event the greatest triumph. To argue the intensely
complex issues which pervaded the draft agreement in a small meeting of experts was one thing, to ensure that a reasonably coherent
and workable plan should come out of a large polyglot democratic
debating society-there were some 750 persons collected at Bretton
Woods-in which many were highly expert, but not a few were
moved by political, as much as economic or administrative considerations, was something very different. Moreover, since voting rights

John Maynard Keynes I883-I946 . 75


were related to quotas, national prestige was deeply involved in
their fixing.
For a week before the main conference was to meet at Bretton
Woods there were further discussions by the group of experts at
Atlantic City to clear up, so far as possible, doubts, ambiguities and
differences. The main Conference opened at the beginning of July
1944. The problems which confronted it were more than ordinarily
difficult. It was inevitable that the first major conference to deal
with a particular group of postwar economic problems should tend
to stray into far wider issues than were properly on its agendathe more so since the representatives of all countries naturally wished
to have as clear a view as possible of the world for which they were
planning a currency policy. The issues of commercial policy, of postwar settlements of sterling balances, of Britain's particular problems
of the transition, were never far below the surface, but the time was
not yet ripe for their emergence. And all this naturally came back
as a special responsibility upon Keynes.
The proper tasks of the conference were divided between a series
of commissions and committees. The work of these Keynes shared
with his colleagues, Sir Wilfrid Eady, Professor D. H. Robertson,
Professor Lionel Robbins, and N. B. Ronald. His own special responsibility was the Bank, and any success that the Bank may
ultimately achieve will be in no small measure due to Keynes'
appreciation of the limitations of the Fund itself and the importance
of the combination of the two. But, as always, the tasks of coordination and of exchange of views outside the actual meetings exceeded
the labors, great as they were, of the meetings themselves. Before
the end not only Keynes, but other members of the delegation also
were perilously near exhaustion. Keynes himself, despite more than
one minor heart attack, carried on at full pressure, resting when he
could, exhausted often to prostration by the end of the day, but
never remitting.
The progress of any such complex negotiations turns almost as
much on the success of the negotiators in establishing mutual relations of confidence and respect as on the force with which the
intellectual arguments may be advanced. At Bretton Woods, Keynes
was able to earn the personal trust and respect of a great many of
his colleagues, and particularly his intimacy and friendship grew

76 . E. A. C. Robinson
with Mr. Morgenthau, Secretary of the U.S. Treasury, and with
Mr. Harry White of the U.S. Treasury, to whose masterly chairmanship Keynes offered eloquent testimony.
The three weeks of intensive work brought their disappointments
and disillusions as well as their successes. International negotiations
are apt to include horse dealing as well as idealism. But in the end
something emerged which was not too wholly dissimilar from the
objectives of the idealists. In the felicitous speech in which he moved
the acceptance of the "final act," Keynes could sincerely say that
he believed that they had accomplished something in the way of
constructive internationalism. And conscious of his share and of the
way that, despite fatigue and weakness, he had dominated the conference, the delegates paid their tribute by rising and applauding
again and again.
The objective which Keynes had set himself in the Tract on
Monetary Reform of a system of exchange rates which combined
the virtues of short-term fixity with those of long-term flexibility
seemed a stage nearer achievement. But on his return to London
he found that he had to defend himself against those who, now
convinced of the virtues of flexibility, believed that the plan represented a return to the gold standard. The main criticism, however,
voiced by The Times, came from those who believed that Bretton
Woods represented an irrevocable first step towards a commercial
policy which might prove to be unworkable in a postwar world.
These critics could be only partly reassured by Keynes' promise that
there was nothing in the Bretton Woods Agreement itself which
was fundamentally incompatible with some of the commercial measures which they advocated.
Keynes had scarcely returned from Bretton Woods when at very
short notice he was asked again to return to the United States-this
time to share with Sir Robert Sinclair the task of leading the British
delegation which was to discuss with the U.S. authorities the terms
and quantities in which Lend-Lease might be available to Britain
if Germany should have been defeated, but war should still be continuing, as it was then thought that it would for a considerable time,
against Japan.
Since the problems included not only the transitional fiI)ancial
arrangements, but also the British requirements and supplies of mili-

John Maynard Keynes 1883-1946 77


tary, naval, and air equipment, of food, shipping, oil, and other supplies, as well as the whole issue of the limitations upon the British
right to export at the same time that she was receiving Lend-Lease,
Keynes found himself one of a quite substantial team, including a
number of very senior officers of all three services, and of senior
officials of the supply departments and of the Ministry of Production
as well as of the Treasury and the Board of Trade. And this team
was reinforced from the permanent missions in Washington. While
those concerned in the negotiations for military equipment had to
reach agreement with their American counterparts on the more practical problems of actual supplies, and similar detailed agreements
had to be reached in regard to food, raw materials, and other supplies, it fell to Keynes to secure in general terms the approval of
the United States authorities, for transitional arrangements which
might prove agreeable to the United States and practicable to Britain,
whose whole economic future was involved in the re-expansion of
exports, and whose power to live until a balance of payments was
restored was at issue.
This task involved deploying to the United States authorities in
full detail the existing and prospective economic and financial position of Britain, both in written and oral evidence. In this task Keynes'
capacity to master both the general principles and the detail was
once again demonstrated. And, since the negotiators on the American side again included Mr. Henry Morgenthau, and Mr. Harry
White, the mutual respect and confidence established at Bretton
Woods greatly contributed to the progress of the discussions.
But, as at Bretton Woods and in later negotiations, the load upon
Keynes was tremendous. The questions to be argued inevitably covered a far wider range than his own Treasury responsibilities, but
it was obviously impracticable to allow each particular issue to be
argued by a separate expert, even had they been competent to do
it, and it fell to Keynes to master, after the fashion of a barrister,
a series of highly complicated issues both of current practice and
of past history. He did it with a remarkable success. But it involved
once again very long stretches of very intensive work. Any delegation in another capital must find itself, as did that one, a handful
opposed to the army of another country's Civil Service; to hold its

78 . E. A. G. Robinson
own against that army the small handful must rely on its own capacities for rapid labor.
Actual negotiations occupy, moreover, a very small fraction of
the time. The main tasks of the leaders lie in the protracted discussions of policy within the delegation itself, where conflicting interests
must often be harmonized, in the preparation of material for the
next round of discussions, and above all in keeping the responsible
Ministers and Departments in London informed of progress and
agreed to the lines of policy to be proposed. Seen from 3,000 miles
away, the difficulties of persuading London appear sometimes of
almost equal magnitude with those of persuading Washington; but
London is at once the source of support and strength, the safeguard,
and, if necessary, the alibi.
The strain upon the leaders of such a delegation comes not only
from the volume of work-Keynes' own capacity for work proved
again to be terrific-but also from the measure of the responsibilities
involved. The few occasions when Keynes showed serious and obvious signs of strain were all when things appeared for the moment
to be going badly. His wonderful resilience he owed, I believe, in
large measure to his capacity for putting all his cares and worries
temporarily aside. All who shared in them will remember with joy
some of his gay little dinner parties. I myself shall always treasure
vivid recollections of an utterly happy afternoon after a particularly
exhausting morning when Lydia persuaded him to take us to Mount
Vernon, and for three or four hours he forgot everything except
the exquisite perfection of that lovely house, and the shock of beauty
of the view from its terrace, while he engaged in lighthearted
and completely preposterous arguments about George Washington
and modern America.
While that capacity to shed temporarily his cares enabled him, on
this occasion, as on others before and later, to carry through to the
end, we all felt before it came the measure of his physical exhaustion.
Yet he had to go on to further negotiations, scarcely less strenuous,
in Canada. And on the homeward voyage he never ceased to work.
He crossed the Atlantic always by ship when he could, for flying,
even for short distances, was a strain upon his heart which sometimes
prostrated him for several days. The crossings in some measure
provided a short holiday, for he would mix lighter reading (Jane

John Maynard Keynes 1883-1946 . 79


Austen on this occasion) with his work and relax a little. But in
weather in which the rest of us were satisfied merely to exist he lay
propped on his bunk and wrote in his most brilliant vein an account
of all that had been done. I shall always remember one remark of his
on that return voyage: "We are now at the happiest moment of our
lives; we have ceased to worry about the war, and have not yet begun
to worry about the peace." How true it was!
14

The end of the German war and the general election wrought
no change in Keynes' position and responsibilities. He had served
first Sir Kingsley Wood and later Sir John Anderson; Dr. Dalton
now asked him to continue as his Financial Adviser. When the
Japanese war ended unexpectedly in August 1945 Keynes' ambassadorial services were again urgently needed. So rapid an end of
the Japanese war had not been expected, and the process of transition
of the British economy from war to peace had not been carried beyond the stage to which the mission of September 1944 had carried
it. The abrupt cancellation of Lend-Lease, entirely legitimate and
necessary as it may have been, left Britain in acute difficulties.
Keynes was asked to join his old friends Lord Halifax and Mr. R.
H. Brand as a mission to see what credit arrangements could be made
to meet the immediate emergency.
It is no secret-indeed he proclaimed it in the House of Lords
on his return-that Keynes at the outset of the negotiations was
himself convinced that the moral case for very generous treatment
of Britain by the United States was unassailable, and that it should
take the form of an outright gift. He was convinced (as were all of
us) that in terms of equality of sacrifice Britain had borne more
than her share-that she had been in the war for a longer continuous period, that in terms of physical losses of ships and property,
and particularly in terms of the tremendous change from a creditor
to a debtor position, she had suffered as much or more than any
other of the Allies. Keynes started in high hopes that he could convince the American administration also that an outright gift was
not only an act of generosity but equally an act of practical wisdom
designed to rebuild the world's trade and prosperity. This case was
argued in full, and listened to with patience and sympathy. But it

80 . E. A. G. Robinson
was very soon apparent (to quote Keynes' own words to the House
of Lords) "that a primary emphasis on past services and past sacrifices would not be fruitful." The principle of equality of sacrifice
had never been accepted by any of the Allies (ourselves included):
sacrifice itself was of many kinds, and a comparison of it invidious
and insoluable. America was more interested in the future than the
past. Thus the offer which Keynes and his colleagues could secure
was something quite different in, character from what they would
have liked. What they were offered was a loan of '1,100 millions
at 2 per cent, of which about one-seventh was to represent payment
for all outstanding Lend-Lease obligations, with special arrangements
to enable Britain to begin the interest payments only in 1951 and
to escape them in any year if her financial position required it. But
added to this, and at least equally important in any long-term view
of the settlement of war debts, was the complete cancellation of the
greater part of the Lend-Lease account.
It was largely because he felt that in the latter respect the American gesture had been truly handsome, and had received but niggardly acknowledgment in England, that Keynes, on his return, was
able sincerely to say that, disappointed as he and his colleagues
were at their failure to secure the gift for which they had originally
hoped, this represented "an act of unprecedented liberality." But
the great difference between the original hopes and the final outcome made the course of the negotiations exceedingly tortuous. While
Keynes and his colleagues were quickly convinced that their original
ambitions were unrealistic, both official and unofficial views in London were less quickly revised, and Keynes found himself, once again,
persuaded by the Americans, but unable at all readily to persuade
the British. The critics fastened, not so much on the interest and
amortization terms, though in Britain's prospective difficulties these
added responsibilities were certainly matter for concern, as on the
strings attached. The principal of these was the obligation to fulfil
at an early date the requirement, which would have fallen upon us
under the Bretton Woods agreement only after a transition period,
to make other countries' sterling earnings freely convertible. The
critics felt, perhaps wrongly, that Britain was progressively being
committed to a system of international free trade which could be
made a reality only if American tariffs were reduced, and forced

John Maynard Keynes 1883-1946 . 81


to fritter away on other issues the ammunition which should be
reserved for the reduction of the bastion of the Hawley-Smoot tariff.
These conflicts of view, whatever the rights and wrongs of them,
left Keynes between two fires. London, seen from Washington,
seemed sublimely unaware of the realities of America. Keynes himself, long isolated in America, was perhaps insufficiently sensitive
to the cross currents of English opinion. The long drawn out struggle
left its permanent mark upon him. By the end he was weary and
exhausted almost to collapse. If anyone of his wartime efforts may
be said to have hurried the end, it was those days of grueling fight
to interpret to each other the two countries that he loved, and to
secure for Britain the breathing space that she so desperately needed
at a price that she could afford to pay.
Even on the voyage home, Keynes was in no position to relax.
His colleagues, totally exhausted by their labors, were spending most
of the days as well as the nights in sleep. But for Keynes himself the
battle was not yet over. He was thinking all the time of the political
struggle that lay ahead of him when he landed, and of the debate in
the House of Lords for which he hoped to be in time. Every few
hours he would creep along to the radio room to hear, with growing
mortification, the misrepresentations of his settlement which were
appearing in the press and in Parliament, and creep back, with
growing scorn, to his stateroom to sharpen and to polish the phrases
with which he intended to give the lie to his attackers. By the time
that Southampton was reached the speech had been rewritten several
times. But he knew that he had forged a weapon that would silence
his critics and give him the debating triumph of his life.
Thus when Keynes reached England in December 1945, a very
tired man, he found an atmosphere of impotent but rather widespread hostility to the agreement which had been drafted in Washington. True that no one felt confident that we could have done
better. True also that everyone was warweary and doubtful whether
the frightening further austerities which rejection would impose
upon us would be either physically tolerable or politically acceptable.
Nevertheless, there was a feeling, however unjustified, that we had
been outmanceuvred and were becoming too deeply committed to
what might prove to be an unworkable world. For these reasons the
acceptance of the Bretton Woods plan and of the loan itself by

82 . E. A. G. Robinson
Parliament was far from a formality. In the House of Commons,
even though many backbench members of the Labour Party were
uncomfortable, there could be little doubt that, presented by the
Chancellor of the Exchequer as Government policy, the loan would
be accepted and the Bretton Woods Bill passed. But in the House
of Lords the issue was far less certain.
Keynes landed from the Queen Elizabeth on the day on which
the debate in the Lords opened. He reached the House in time to
hear Lord Pethwick-Lawrence put the Government's case, and Lord
Simon, Lord Woolton, and others criticize it severely, though differing as to whether they would or would not reject it. Keynes himself
intervened only momentarily to confirm what the Chancellor of
the Exchequer had said in the House of Commons, that there could
be no misapprehension in the State Department as to the British
view that reduction or elimination of preferences by ourselves could
only be considered in relation to, and in return for, similar reduction
by other countries, and that there could be no question of unilateral
surrender by us.
Keynes opened the resumed discussion the next day. The speech
to which he had devoted so much labor achieved a success as great
as he could have dared to hope. There was no question that it was
the greatest debating triumph of his life. He was able to give the
full American background to the negotiations which made clear the
impossibility of much that earlier speakers had been demanding. He
made abundantly clear his own disappointments. Above all, he was
able to show that the supposed alternatives were a mirage. It was a
tremendous performance, and was acclaimed as such. The Bill was
passed by ninety votes to eight. But many had abstained. And though
Keynes had clearly persuaded not only the House of Lords but also
the country that in existing circumstances nothing better could be
achieved, there remained the strong feeling of discomfort and of
uncertainty whether, in the absence of reductions of tariffs abroad,
all these plans were going to be workable.
It is not yet easy to see Keynes' work for the Loan Agreement
in its true perspective or to fit it into the unexposed sequence of history. It far transcended both in the magnitude of the issues at stake
and in the difficulty and strain of the actual discussions all the earlier
negotiations in which he had taken part. It is fatally easy for the

John Maynard Keynes 1883-1946 83


distant critic, fettered by no limitations of practical politics, to claim
that some undefined or unattainable alternative was better. It remains that Keynes won for England a chance of economic recovery at
an absolutely critical moment. Besides that, all minor considerations
fade into insignificance. It was beyond cavil Keynes' greatest achievement, and as such he saw it.
With the Bretton Woods Bill finally passed, Keynes began to
think of rest. For the previous three years the man whom before
the war we had begun to think of as a lifelong invalid had carried
an almost insupportable burden. Lydia had accompanied him on all
his journeys; he and she treasured an official letter from the Admiralty assuring them that they had incurred their Lordships' displeasure by insisting on her traveling on a transport in wartime. She
guarded him, cared for his health as best she might, and shared with
him the joys and sorrows. But no care could protect him from the
inevitable strains that these missions had involved, or from his own
impulsive energies. Under her eye he would obey the doctor's rules
and climb the stairs, as he must, one step, one breath. But in her
absence he would forget, and pay the penalties.
There remained, however, one further step in relation to Bretton
Woods-a step which he and most of us had regarded as a formality
-the Inaugural Meetings of the Fund and the Bank on both of
which Keynes had been made Britain's Governor. Indeed, it had
been regarded as scarcely necessary at one time that he should himself attend. He was then turning over in his mind whether or not he
should go, and coming down ultimately and on balance in its favor,
since it would give him an opportunity for escaping for a few weeks
and having a short holiday in a warmer climate than an English
winter, and a chance also to renew many happy friendships of Bretton Woods.
But the Savannah Conference did not in practice tum out like
that. There were unexpected last moment differences, and the British
position was made more difficult by the slow progress of the U.S.
Loan in its journey through Congress; final approval came only
after Keynes' death. The superficial disagreement was on the site
for the Fund; but below the surface the issue at stake was whether
the Fund and the Bank should be political or financial instruments.
In the end Washington, the political center, was chosen despite

84 . E. A. G. Robinson
the arguments of the British representatives for New York. Disagreement arose also as to what should be the power of the permanent staff of the institutions as against the powers of the executive
directors who were representatives of the great powers. Again the
decision went in favor of the political power and against the financial.
Because he had gone in no expectation of the differences of opinion
that emerged, Keynes found it the most exhausting conference that
he had attended.
He arrived home from Savannah very tired, and went to Tilton
to rest. On Sunday April 21, 1946, Easter Day, he had a sharp heart
attack. He died within a couple of hours of the onset.
For some years before 1946 both Keynes himself and his friends
had known the possibility, even the likelihood, that if he continued
to work as he was working the end would come one day just as it
did. Despite this knowledge he carried on. This was his contribution
to the war and to the future of his country.
His death left a gap everywhere. In the Treasury, where, despite
his amorphous position, he carried a massive share of the total responsibilities for the country's financial welfare. In the world of
academic economics, where for so many years he called the tune
to which the rest of us had danced. In his College and University,
where his learning and his leadership alike were irreplaceable. In
the world of the ballet and the arts, where his sustaining energies
had kept so much alive that might so easily have died. Not least, in
his own family. But perhaps some day we may learn to say that it
was right that he, like others whom the gods love, should die young.
At sixty-two he was in the plenitude of his powers. That brilliant
mind was still at its best-rapier sharp, leaping always with intuitive
rapidity far ahead of the rest of us. The memory that will remain
is of that mind at its perfection.

15
In the six years from the outbreak of war to its end Maynard
Keynes had subtly changed. At its outbreak he was still fundamentally the academic and the philosopher, with truth and right as
his objectives. He was still the critic, trying to persuade, and content
if sometimes the process of persuasion involved an element of
overemphasis on the component that was undervalued by others. The

John Maynard Keynes 1883-1946 . 85


war brought him into a position of responsibility where action followed close upon advice, and advice needed therefore to be measured. And as his ambassadorial journeys took him to place after
place and problem after problem, he came insensibly to think of
issues more, as does the Civil Servant, in terms of what was practicable; of what could or could not be secured in a given international
set-up. His love for the verities never deserted him, but he was
readier than the young Cassandra had been to believe that a desired
destination could be reached by speeding slowly.
As a Civil Servant he had two God-given weapons-a voice and
a pen. No account of Maynard Keynes would be complete without
a reference to that voice. To many is given the gift of dissuasion:
every word they utter seals the fate of the object they so passionately
desire. To Maynard Keynes was given a voice that could persuade.
One who was by force of circumstances for the moment his opponent said to me after a meeting dominated by Keynes: "The worst
of it is that Keynes' voice can persuade me of anything, however
wrong-headed I believe it to be." That beautiful, musical, resonant
voice, allied to an unparalleled power of lucid exposition and to a
range of vocabulary and a joy in words comparable only to that of
Winston Churchill in his generation, made him a pleasure to listen
to, whether you agreed or disagreed, whether you knew all about
what he was talking about, or nothing about it. He never bored. He
never exhausted. He was never trite.
And allied to the voice, the pen. To the jaded Civil Servant,
wearied by a perpetual surfeit of the cliches that conveyed, by a
process of automatic writing, ~e tired thoughts of tired men to the
tired eyes of more tired men, those occasional official papers that
Keynes wrote were champagne-a sparkle of ideas set out in words
which fizzed into one's memory. Many of them were entirely of the
moment. Some will, I fear, never be able to see the light of day.
But others should surely, after due interval, be published. May we
hope that the Treasury will give thought to such a kindly act of piety?
To the arts of persuasion Keynes added, as a Civil Servant no less
than as an academic, his divine intolerance of nonsense. To his
immediate colleagues, and to those who agreed with him, this intolerance was wholly delightful. To those in other departments his
facility in argument and in demolishing their carefully reasoned

86 . E. A. G. Robinson
cases was often disturbing-the more so as they often felt that it
was not the case but the arguing of it that was at fault. But if the case
was good, reason would quickly prevail. Keynes would often laughingly say that he possessed the shortest memory in the world. And
I believe there was truth in this-he had a wonderful memory for
arguments, but no memory for their authors. If next day you returned to the same problem, you were as likely to find him parading
your arguments of yesterday-if they were good arguments-as his
own. And those of us who knew him learned to retreat when worsted
today and to re-engage tomorrow.
And to these others he added the greatest gift of the Civil Servant,
the power to inspire devotion. The missions which he led were all
happy parties. They involved, as I have said, almost unbelievable
pressure of sustained effort, not only for Keynes himself, but also
for all who were engaged either in responsible or routine work. All
drove themselves, partly because Keynes was driving himself, but
at least as much because he filled them with his own sense of
urgency. He was utterly informal, almost without sense of hierarchy,
approachable by everyone who had a problem on which to brief him,
or a difficulty in his own particular negotiations, extraordinarily quick
to see the point at issue and to fasten on to the problems that were
worrying you, extraordinarily rapid in judgment and decision.
I can with difficulty think of others who were nearly Keynes'
equal in a fastidious forensic ability; of others who approached him
in the power to clothe perfect clarity of thought in a perfect choice
of words; of others, again, who shared his mastery of technical and
administrative necessities; of yet others who had grasped the fundamental economic problems of our age, if not so well, yet not very
markedly less well. But no other man in our age has combined all
these rare qualities so remarkably in one person. It was because he
was all these things and more that Maynard Keynes was utterly
unique.

Could There Have Been a


"General Theory" Without
Keynes?
[1963J
IT

HAS BECOME fashionable in recent years to attempt to


identify in the General Theory elements which had independently
been discovered by Keynes' predecessors, and by implication to belittle Keynes' own contribution and to suggest that the essential
"Keynesian revolution"-whatever may be meant by that-could and
would have occurred without Keynes. I propose to devote my contribution to this volume to an attempt to assess Keynes' own part in
that revolution.
It would, I think, be a tragedy if the Bloomsbury delight in reducing the great Victorians to the stature of the common man-or
more truly to the stature of the very uncommon inhabitant of
Bloomsbury-ended in the supreme, but in my view, equally indefensible achievement of reducing the great iconoclast to the level of
the second-rate economist.
It is not, I want to argue, possible in these days for any economist
to build without any of the bricks used by his predecessors. The
greatness of a Corbusier lies not in an impossible attempt to use
none of the normal materials of building, but in his ability to put
them together differently and to make out of them something essentially new and more appropriate to his own age. The break-away
is not in every detail but in a new total conception. That is what, I
think, Keynes was achieving. And the ants who go round trying to
identify the provenance of the bricks are too near the ground to see
the broad scale of the achievement of the architect.

87

88 . E. A. G. Robinson
I doubt whether Keynes himself knew whence he derived all of
his bricks. He was a rapid and omnivorous rather than a painstaking
and laborious reader. He would sometimes argue that anyone (and I
suspect that meant himself) could extract the essentials from almost
any book in a couple of hours. He remembered vividly the ideas
which he absorbed into his own thinking. But he did not, I think, remember with great certainty whence he had got them. They became
part of his own system of thought and would not, I suspect, have
remained identified with an author. He took over into his own
thinking equally and indiscriminately ideas that he had himself
evolved, ideas that had occurred to him while stimulated by a book
or article he was reading, 1 the ideas of the actual author of the book
or article, ideas put into his head in conversation with his Cambridge
colleagues, his pupils, and those with whom he worked in all the
widespread activities of his life. All these became part of his thinking,
and those of us who lived around him found pleasure and amusement when our own ideas were occasionally retailed to us againretailed, however, with a new vividness that they had acquired in
the process of being adopted.
When he was writing, Keynes did not ordinarily work with a pile
of the writings of his predecessors on a table beside him, for reference
and accurate quotation at every moment. His method was rather to
start with a virgin block of paper and write from his own mind. He
used his authorities where necessary to provide facts, but rarely to
buttress his own analytical thinking. If one turns to the Treatise,
which in many ways was more of a work of economic scholarship
than was the General Theory, Keynes' references are usually either
to give the source of some facts or figures or, more often, to justify
his exposition of some theoretical analysis from which he differs. He
seldom, if ever, tries to identify the first origins of an idea with which
he agrees.
Keynes' interest in his predecessors nearly always, I think, came
after the conception of the first draft of a piece of work, or at earliest
1

I suspect that many of us would share this difficulty. Few of us when


reading relax and accept passively the arguments of the author. Most
of us, with the author as guide, think critically through the problems
the author is examining, and ideas may occur to us either through
agreement, disagreement, or mere embroidery.

A "General Theory" Without Keynes? . 89


when a new piece of work had already taken shape in his mind. His
interest in Malthus' thinking about employment came after his own
ideas were already formulated, if not completely, at least in essentials. When his own work was finished, he took, however, real interest in identifying predecessors. But I doubt whether he himself
could ever have disentangled the ideas that he may have acquired
from other authors from those which he evolved for himself. Nor do
I think he would have regarded it as a. profitable occupation. The
whole concept of private property in ideas, subsequently bred by the
Ph.D. thesis and competition for faculty promotion, was alien both
to him and to most of his Cambridge contemporaries. At that moment we were more excited about what was happening to economics
as a whole than about who might possibly claim responsibility for
which brick.
The elements of the General Theory were clearly many and varied.
Much of the essentials derived originally, I have always felt, from
Dennis Robertson's seminal thinking in Banking Policy and the
Price Level, 2 the first clear distinction known to Cambridge between
the act of saving and the act of investing. The very first steps forward
in 1931 or thereabouts from the Treatise to the General Theory, involving the abandonment of what we christened "the buckets-in-thewell fallacy" of the Treatise with its implied assumptions of constant
output, owed more to the "circus" of bright young pupils3 to which I
referred in the Economic Journal "Obituary" than to Keynes's personal thinking. Elements, moreover, of this move forward were provided by Richard Kahn's work on the multiplier and by what we
irreverently christened in Cambridge "Mr. Meade's relation," which
cumulated the elements of saving arising from an act of investment
to parallel Richard Kahn's cumulation of expenditures and employments, and provided the first hints at the automatic equalization of
In regard to Keynes' contribution to the development of his ideas, see
p. 5 of Banking Policy and the Price Level, editions of 1926 and
1932, and especially the preface to the 1949 edition, pp. xi to xiv.
8 I was wrong, Lorie Tarshis tells me, in making him a member of
the "circus." He arrived in Cambridge a little later. There is no survivin~ written record of the very infonnal arguments in Richard
Kahn s rooms, and others besides those I named may on occasion have
been there. See also R. F. Kahn, "The Financing of Public WorksA Note," Economic Journal, September 1932, pp. 492-5.
2

90 . E. A. G. Robinson
saving and investment. 4 In the process of writing the Treatise, moreover, Keynes had read more than usually deeply in Wicksell's writings. Ralph Hawtrey was, of course, an old friend and continuous
correspondent. From all these sources Keynes undoubtedly drew
perhaps even more than he, or we at the time, knew.
But this is not to belittle his own contribution. For all of us in
Cambridge, it was Keynes who was both the stimulus and the focus
for these various lines of thought. I think, moreover, that he possessed much more clearly than any of the rest of us a complete and
coherent picture of the system that was emerging from all this process
of re-examination. For I have long felt that Keynes' economic thinking was, in reality, intuitive, impressionistic, and in a sense feminine
rather than precise, ordered, and meticulous. 5 He saw how things
must be, and was sometimes prepared to leave others-and particularly Richard Kahn-to work out more formally how his intuitions
happened to be true.
If one is to see Keynes' own contribution one must, I think, take a
step back and ask what was the revolution that has been called after
him. As so often, I believe the principal innovation consisted not so
much in answering as in asking new questions, or in making us all
ask afresh questions to which we wrongly thought we knew the answer. First and foremost Keynes, aided and abetted by those working
most closely with him, made us ask what are the factors that determine the level of output as a whole, not only in a full-employment
long period but also in a short period. To a younger generation of
economists, brought up on Keynesian and post-Keynesian versions of
the theory of employment, it is difficult to imagine a time when
these questions were not central in all economic analysis. But before
1931, while we were inevitably interested in unemployment and its
causes, our attempts to explain it were guilty, as I think a rereading
of Pigou's Theory of Employment will show, either of applying to
the whole of activity propositions that were properly true only of an
4

See numerous references in R. F. Kahn, "The Relation of Home Investment to Employment," Economic Journal, June 1931, pp. 187191.

I owe this suggestion originally to Polly Hill (Mrs. Humphreys),


Keynes' niece and, at the time she made the comment, my assistant
in editing the Economic Journal under Keynes.

A "General Theory" Without Keynes? . 91


infinitesimal sector of all activity, or made assumptions about the
supply curve of labor to all activities that were manifestly unrealistic.
Secondly, as part of this rethinking, it was necessary to rethink the
whole set of propositions traditionally regarded as Say's Law, the determination of the level of effective demand, and, as one element in
it, the relation of the volume of investment to the volume of saving.
In the process of sorting out these problems one found oneself
applying to what had hitherto been regarded as monetary problems
-to be dealt with in terms of the Fisher or Marshall quantity of
money equations-the tools of analysis that were customary in the
theory of value. It was an essential element in the revolution that
the new theories of prices represented an integration of two previously separated theories of money and theories of value. A schizophrenia in which we had all been educated required to be cured.
Thirdly, it became clear that, certainly in terms of the short period
and not impossibly in terms of much longer periods, a fall of the
propensity to consume would contract effective demand, diminish
the incentive to create new capital, and progressively contract the
level of activity until a combination of reduced saving and increased
dissaving, associated with whatever level of investment remained and
possibly an element of investment in unsold stocks, should re-establish an equilibrium at a lower level. This process of contraction
might be averted if the rate of interest were reduced sufficiently not
only to offset any consequential effects on the marginal efficiency of
capital but also to increase investment sufficiently to absorb the resources released from making consumption goods. It was clear, however, that the process of equalization of saving and investment would
be achieved independently of any change in the rate of interest.
Thus one could not invoke in any simple-minded terms a difference
between savings and investment as a force reducing the rate of interest. And in consequence one could not properly assume that, independently of any action by the monetary authorities, a rate of
interest would be established which would restore a condition of full
employment. And thus one could not assume that the system was
automatically self-righting.
This is, of course, the area in which controversy has continued
without pause ever since the publication of the General Theory. It
may, moreover, be right to admit that Keynes' own analysis is prob-

92 . E. A. G. Robinson
ably most vulnerable in his handling of the longer-term determinants
of the rate of interest. But in a more practical sense, I believe the
subject to have become almost irrelevant. The monetary authorities
of the world have now become so steeped in the general thinking of
the Keynesian revolution that, where a reduction of the rate of interest is desirable, and mayor may not in the purest logic be automatic in theory, they in fact tend to bring about the necessary adjustment; in this way the responsive actions of the authorities can be
regarded as having become a part of the theory in the same way that
they traditionally did in the assumed operations of the gold standard.
To all intents and purposes we now act as if the system is not selfrighting and needs assistance in certain circumstances.
The greatest and most important element in the "Keynesian revolution" I have deliberately left to the last. Before 1936 the task of a
government in a depression was, at most, ambiguous. It encouraged
its central bank to take measures designed to protect the currency
reserves; a strong central banker might be prepared to assist colleagues in difficulties. It might engage in a limited scheme of public
works. It might (as in the 1930's) encourage international consultation. But by and large, booms and depressions were accepted as
inevitable, and as uncertainly predictable as their climatic analogies.
An economist's job was to help intelligent businessmen to manage
their businesses in such ways that they suffered a minimum of
damage. From 1936 onwards our whole attitude to boom and depression has changed. We are not always agreed about what exactly
should be done, or about the exact timing of action, but we are
agreed, almost unanimously, that governments can and should take
action along certain specific lines to mitigate a depression or damp
down a boom that is getting out of hand, and that by doing this they
can in large measure avert a major recession.
It is this change of attitude as to what can and cannot be done, and
the effects of this change on the dimensions of booms and recessions,
that represent the greatest economic change of our life time, and
carry the greatest implications for economic policy. All of this has
involved and been made possible by new attitudes of central bankers
and of finance ministers, which in my own country began, indeed,
with Keynes' inspiration of Kingsley Wood's 1941 Budget, but have
been progressively developed through the work and experience of

A "General Theory" Without Keynes? . 93


countless others, in the Treasury, the Bank of England, and the various Commercial Banks.
It would, of course, be ridiculous to credit Keynes individually
with the whole of the so-called Keynesian revolution. On the other
hand I am myself firmly convinced that it would not have been
achieved in anything like a comparable period of time without him.
The change did not consist in a series of separate minor amendments
to an almost perfect pre-existing classical system-amendments each
of which one might accept or reject individually. It consisted in inducing a reluctant body of dedicated but perhaps rather cautious,
critical, and conservative thinkers to abandon a large part of what
they had given their lives to learning and teaching, and to accept, as
one complete (or virtually complete) package, a set of new and
highly debatable propositions and of new ways of handling familiar
problems.
If the ideas embodied in the General Theory had all derived individually from others-a proposition which I would most vigorously
deny-I would still think that Keynes brought about one of the most
astonishing advances in economics, first by building them into a new
and consistent whole, and secondly, and far more remarkably, by
securing their essential acceptance by the great majority of all economists. I can with difficulty be made to believe that some of the lesser
men of the present generation of economists could have worked out
in their completeness the essentials of the Keynesian theories. Nothing will persuade me that they have possessed the qualities which
enabled Keynes to convert the world.
Looking back, I find it difficult to explain just how he did it. One
must, I think, begin by remembering that this was not the first time
in his life that Keynes had made both his countrymen and the world
rethink a major issue. He had done it first in terms of reparations.
The voice in the wilderness in 1919 had become the orthodoxy of a
few years later. He had done it secondly in regard to the gold standard. He was now, for a third time, making us stand our thinking on
its head. He was, moreover, doing this not merely-as any of his contemporaries and most of his successors would have had to do--as a
professional economist rewriting professional economics, but also as
an active belligerent in the current battle to conquer unemployment
at the depth of the most serious depression in history. He possessed

94 . E. A. C. Robinson
an intellectual leadership which no other economist of his countrymen, save possibly Ricardo, could have claimed in the past and certainly no other English economist possesses today. I doubt whether
any more ordinary academic who did not possess this leadership could
have achieved what Keynes did.
When he believed that he had an idea of which the official or
wider world needed to be convinced, he was prepared to take endless
trouble to convince it, by writing, by pressing his views on such committees as the Macmillan Committee or the Economic Advisory
Council. And he could be immensely persuasive. But having said
that, I find it difficult to argue that the General Theory is as well
written, as readable, and as persuasive to the general reader as the
great bulk of Keynes' writings. One rereads the Economic Consequences with as much joy as one read it originally. His more polemical pamphlets have a fire and vitality and a happiness of phrase that
few economists have equaled. Large parts of the General Theory are
sheer labor to read and I suspect that few even of the most ardent
Keynesians refresh themselves regulady at the original spring. Here
I find myself agreeing with some of the comments of Professor Johnson in his essay on "The General Theory after Twenty-five Years."
Indeed I find it strange that Keynes, the great stylist, should be remembered principally by the least well-written of all his books. It is a
paradox that the world seems to like its bibles-if one may think of
Das Kapital as well as of the General Theory-to be so nearly incomprehensible as to require a body of expert theologians to expound
them.
It is, of course, easy to explain how it happened. Keynes never,
I feel sure, regarded the General Theory as immutable law, to be
perpetuated on tablets of stone. His instinct was always to cast his
bread on the waters: to get his present thinking into the hands of
readers before the policies that he was seeking to inHuence were
crystallized. He was a pamphleteer rather than a procrastinating and
perfectionist pedant. I suspect that by the end of the rather long
period of preparation of the General Theory he was more anxious to
get it out than to rewrite it.
But the more fundamental reason why it is difficult to read
is that it is two books in one. It is a positive exposition of a new
set of theories. It is an escape from an old set of theories. When

A "General Theory" Without Keynes? . 95


Professor Johnson asks why it has needed popularization by others
of Keynes' disciples, the answer is surely that the popularizers
have been content to write, for those not encrusted in older theories, a simple exposition of the new theories and have left it at that.
But Keynes himself was not content to leave it to the growth of a
younger generation-to Professor Johnson, his contemporaries and
successors. He had to have it out with the older generation and that
needed a different sort of book. As is clear from his introduction to
the French translation of the General Theory, he felt impelled to
shock and to heighten controversy. In retrospect it is easy to wonder
whether a simple and clear exposition in a different General Theory,
with his more destructive criticisms in the pages of the Journals,
would not have served his purpose. I do not think it would. It was
of the essence that economists had to be forced to make up their
minds whether they were, or were not, prepared to make a break with
their own past thinking.
As I have said elsewhere, it was all very much in the atmosphere of
a revivalist mission. I believe it had to be done. I do not believe that
any smaller man than Keynes could have done all this with the speed
with which it was done. Nor do I think it could have been done by
a series of minute advances by a series of individual and independent
scholars, each making his own contribution. I am persuaded that we
could not have had a Keynesian revolution without Keynes in the
1930'S and that we would have had a very different world in the
195 0 'S.

PART TWO

The Thirties
The Sixties

W. B. REDDAWAY

The General Theory of


Employment, Interest, and
Money
[1936J
IT HAS BECOME TRADITIONAL for reviewers of important
books to begin by saying that it is impossible within the limits of a
single article to do more than consider very briefly a few of the
questions at issue. Such a statement would assuredly be justified in
the present case, for in this book Mr. Keynes attacks one of the
fundamental assumptions which has underlain orthodox theory since
the days of Ricardo. This is the doctrine which used to be expressed
categorically in the phrase "Supply creates its own demand." Later
writers have been more guarded on the subject, and often refrained
from stating it specifically in any form at all. But however it might be
expressed or implied, orthodox theory has continued to be based on
the principle that "what constitutes the means of payment for commodities is simply commodities"! (Mill); from which it follows
(inter alia) that money makes no difference except frictionally, that
consumption is limited by production and not vice versa that general
oversupply is impossible, and that, to quote Professor Pigou, unemployment is due to the fact that "frictional resistances prevent the
appropriate wage adjustments from being made instantaneously."
In place of this Mr. Keynes seeks to substitute a monetary theory of
production according to which unemployment may be due, not to
labor's refusal to accept a lower reward, but to a deficiency of
"effective demand."
Probably the first thing which many people wanted to know on
1

Later writers have, of course, added "services" to commodities; but


that is not the point at issue.

99

100

W. B. Reddaway

the announcement of the new book was its relation to the Treatise
on Money. Mr. Keynes deals with this question in the preface, and
there is little need for me to add anything. There have been important changes in method and terminology, the latter being bound, I
fear, to create much initial confusion. But the underlying principles
are fundamentally the same, though they have been clarified and, in
the process, considerably developed. Those who have accepted the
main thesis will, as soon as they are accustomed to the new terminology, find a more penetrating and more logical exposition of it; those
who still remain unconvinced will be stimulated to greater endeavors
to refute the heresies, now more dangerous than ever as a result of
more forceful expression.
The most fundamental change is a clear recognition that the
thing to be studied is the forces which determine the total volume of
output and employment, rather than the various price levels. Not a
few people, including the present reviewer and probably Mr. Keynes
himself, must have felt that this really was the quaesitum even in the
Treatise, and that the analysis was for that reason rather artificial. In
effect the process was supposed to work in two stages; an increase in
investment, for example, would lead first to higher prices being
realized for an unchanged output of consumption goods, and the
resulting windfall profits then provide an inducement to entrepreneurs to expand output. Mr. Keynes was assailing the notion
that money was a mere counter which could exert at the most a
transitory influence on "real" things like output and employment;
but the old tradition was so strong that he made it work through the
medium of changes in prices, which would only subsequently lead
to changes in output. To quote from the preface of the new book:
"My so-called fundamental equations were an instantaneous picture
taken on the assumption of a given output. They attempted to show
how, assuming the given output, forces could develop which involved
a profit-disequilibrium, and thus required a change in the level of
output. But the dynamic development, as distinct from the instantaneous picture, was left incomplete and extremely confused."
Of all the ideas contained in the Treatise the most familiar is that
associated with "the difference between saving and investment."
This phrase has become a catchword which has been applied, or
rather misapplied, to all manner of problems without any recognition

The General Theory .

101

of the special meanings attached to the terms. The concept depended


essentially on the definition of the entrepreneurs' income (and hence
of their saving); this was taken as being neither the profit actually
realized from current operations, nor the profit which they expected
when they decided on what scale to conduct these current operations,
but in some sense as a "normal" or "standard" profit. As a result saving exceeded investment by the excess of this normal profit over that
actually realized; in effect the entrepreneurs were considered to draw
the difference from a sort of contingency fund before computing their
income, and were correspondingly bound to add any surplus to it.
Their saving was then the difference, positive or negative, between
the income so computed and their expenditure on consumption.
Now this concept has certain advantages where the entrepreneur
is a joint-stock company, which will probably put any abnormal
profits to reserve before declaring a dividend, and may draw on
reserves to maintain a conventional rate. But it is a somewhat artificial notion, and in the new book it has been abandoned, income
being measured by the profit actually realized. There are certain
complications on minor matters, notably the question of depreciation;
but the important point is that total income of the community for
any period is simply the value of the output of that period. Saving is
defined as before as the difference between income and expenditure
on consumption; investment is the part of output which is not consumed. It follows as a matter of simple subtraction that saving and
investment, as now defined, are necessarily always equal; this is inevitably true whether the supply of money is being altered or not,
whether we are at the bottom of a slump or at the height of a boom.
This contains, of course, nothing fundamentally new; it is simply
the expression in the new terms of the proposition in the Treatise
1= S +Q (the value of investment is equal to saving plus windfall
profits). But at first sight it may appear difficult to reconcile with, for
example, the freedom enjoyed by each individual to save as much of
his income as he likes. The solution is, however, quite straightforward. When we think of an individual deciding how much to save
we instinctively take his income as given; his actions are too insignificant to have any reactions on it. But for the whole community
this is, in general, untrue. If, for the moment, we assume that the
amount of investment to be done during any period is fixed, then the

102

W. B. Reddaway

national income must so adjust itself that the community will, in


the aggregate, do just that amount of saving. This may come about in
various ways, depending on the type of economic system in which
we live. To take a simple example, imagine a closed system in which
there is only one entrepreneur, and he decides irrevocably what is to
be produced and sells the whole output of consumption goods to his
employees for whatever it will fetch. In this case the adjustment will
be secured wholly through changes in the entrepreneur's income. If
the public spend the whole of their wages, then he will make a profit
on the sale of consumption goods equal to the expenditure on investment; if the public save a part, then his takings, and so his income
and his saving, are correspondingly reduced.
In this example the adjustment is effected solely by changes in
prices altering the income of the entrepreneur without affecting his
expenditure on consumption. But as a rule there will also be changes
in income due to changes in employment, and these are frequently
responsible for the bulk of the adjustment. An increased desire to
save will not merely lead to entrepreneurs receiving a reduced income, but also in a diminished volume of employment, and the extra
saving of those still employed will be balanced in part by the reduced
(probably negative) saving of those thrown out of work. Another
consequence will probably be an accumulation of stocks, which represents unconsumed output, and, therefore, constitutes investment.
The reader must on no account think of investment as applying
solely to construction of fixed capital; the necessary equality between
saving and investment may not infrequently be secured partly
through changes in the size of stocks.
Put broadly, Mr. Keynes' new analysis may be summed up as
follows. We visualize a series of levels of employment. As we go up
the scale the community's real income increases. Its psychology is
such that consumption will increase with increased income, but by a
smaller amount; the exact relationship is given by a function which
Mr. Keynes calls "the propensity to consume." Hence employers
would make a loss if the whole of the increase in employment were
devoted to production for immediate consumption; the increase will
only be justified if part of it is engaged on producing investment
goods. Given the propensity to consume the level of employment is
determined by the amount of current investment; for it must give

The General Theory I 03


that level of income out of which the community will do this amount
of saving.
It remains, therefore, to find out what determines the amount of
current investment. This, according to Mr. Keynes, depends on the
relation between "the schedule of the marginal efficiency of capital"
and the rate of interest (or more strictly, the complex of rates applicable to debts of different maturities). The first is constructed by
considering the supply price and prospective yield over its whole life
for each possible type of new capital asset. The marginal efficiency
of that type of capital is then defined as being equal to "that rate of
discount which would make the present value of the series of annuities given by the returns expected from the capital asset during its
life just equal to its supply price" (p. 135). Now if there is increased
investment in that type of capital during any period the prospective
yield will fall, and the supply price may rise owing to pressure on the
facilities for producing it. So that we can build up a schedule showing
how much new investment there will have to be in it to reduce its
marginal efficiency to any given figure. This may be done for all
types of capital; the results are then aggregated to give a schedule
showing the total amount of investment there will have to be to reduce the marginal efficiency of capital in general to any figure. Mr.
Keynes quite rightly emphasizes that it is the expected yield and
the current supply price of newly produced assets which are relevant;
since it is obvious that these expectations cannot be quickly verified
or disproved, it follows that the state of confidence is vitally important.
Assuming for the moment that the rate of interest is given, Mr.
Keynes deduces from this schedule what the amount of investment
will be; for investment will be pushed to the point which reduces the
marginal efficiency of each sort of capital to approximate equality
with it. Without disputing the general validity of the method, I feel
that this part of the book needs more extended treatment, particularly
in the matter of risk and the varying estimates made (often for good
reasons) by different people. Mr. Keynes inserts a very helpful
discussion of the two sorts of risk which are relevant. There is firstly
entrepreneur's risk, or the doubt in his mind as to whether he will
actually earn the yield he considers most probable; where he is using
his own money, this is the only relevant one, and presumably (though

104

W. B. Reddaway

Mr. Keynes does not explicitly say so) we deduct an allowance to


cover it from the expected yield in calculating the marginal efficiency
of capital. But there is also another risk where the entrepreneur uses
borrowed capital; the entrepreneur will not invest unless the probable yield exceeds the interest he has to pay by enough to cover his
risk; but the lender will not lend unless the rate he receives exceeds
the "pure" rate by enough to cover the risk of default, voluntary or
involuntary. The allowance for this "lender's risk" is a sort of handicap against the entrepreneur who has to borrow; and in addition
there is the cost of bringing borrower and lender together, which may
be considerable.
Now in constructing his schedule Mr. Keynes seems to take no
account of differences between individuals. He talks of the "prospective yield" as if it had only one possible value, gives no specific instructions as to the treatment of risk, and compares the resulting
marginal efficiency with the pure rate of interest for loans of the
relevant maturity.2 To make the treatment logical we must do the
thing in considerably greater detail. If the asset under consideration
is not readily marketable, then we consider each potential owner individually. We deduct from the yield, as estimated by him, firstly
entrepreneur's risk, and secondly lender's risk and the cost of securing the loan (where applicable); and then we relate this to the
supply price. (Alternatively we can add a capital sum to the supply
price to cover these; this is perhaps better for the "preliminary expenses" involved in securing the loan.) By combining the results for
different individuals, not forgetting that one may invest in several
units, we can build up a schedule which really can be compared
with the rate of interest. Where the asset is one with a free market
there are two additional considerations. The individual who takes an
unusually optimistic view can buy an old unit rather than construct a
new one; and the initiative is frequently taken by promoters, who are
influenced more by what they think the public will think of the
enterprise than in their own views. To state precisely how the schedule of the marginal efficiency of capital is to be constructed is almost
2

This is the procedure followed in the past dealing specifically with the
marginal efficiency of capital (cf. footnote on p. 137). But in a preliminary outline he uses the expression "complex of rates of interest
on loans of various maturities and risks" (p. 28) (my italics).

The General Theory .

105

impossible-there are many complications and mutual inter-actions


which I have not space to describe.a But we need not reject the general idea of a scale showing how much extra investment would be
considered payable if the rate of interest were lowered; it appears in
many theories under such titles as "the demand curve for capita!."
We are left then to find what determines the rate of interest. It is
here that Mr. Keynes differs profoundly from the exponents of the
classical theory. Space is too limited for me to reproduce or criticize
his attack on this doctrine; it would indeed be a waste of time, for I
consider it well-founded, and the next move should obviously come
from the defence. The substance of it is that in constructing their
supply curve of savings the classical economists tacitly assume that
the level of employment is unaffected by decisions as to saving-in
effect that their theory is only applicable to a world possessed of some
mechanism whereby involuntary unemployment is made impossible.
Mr. Keynes' "general" theory is designed to apply whatever the level
of employment, including the classical theory as a special case.
This theory is built round the conception of "liquidity preference"
as explaining man's desire to hold cash. There is firstly the familiar
consideration that people need a certain amount of liquid resources to
assist them in spending their income and carrying on their business.
This demand for money, so to speak, "on current account" arises essentially out of its use as a medium of exchange and depends in the
short period on the level of activity. But money is also to the individual a store of value-a fact which is mentioned at the beginning
of all textbooks on the subject, but seldom fully explored. Moreover,
it has several great advantages as a means of holding wealth. It involves no carrying costs--indeed a modest return can be secured by
holding it in the form of a fixed deposit; it is liquid, and involves no
expenses of acquisition and realization; and in contrast with bonds,
its value cannot depreciate. An owner of wealth has to balance these
8

Thus the schedule is greatly affected by the level of employment, because this reacts on confidence. Investment should perhaps be represented as a function of income CY) and the rate of interest Cr) rather
than of r alone. This appears to involve circular reasoning, because we
started by saying that income depends on investment. But, as shown
later, this is not a valid objection, the factors mutually determining
one another.

I06

W. B. Reddaway

advantages against the increased income which he can get by holding


a bond in deciding what proportion of it he will hold in the form of
cash. The amount will in general increase as the price of bonds rises
(i.e., as the rate of interest falls), largely because of the added risk of
capital loss involved in holding bonds. Thus Mr. Keynes constructs
a second liquidity function which relates the amount of cash which
people will want to hold, so to speak, "on capital account" with the
rate of interest. The total amount of money available (M) is determined by monetary policy (or, in the absence of policy, by the
supply of gold or some other accident), and this must be held by
somebody, either on current account or on capital account. Hence
we have the relation

Now if we know Y, the level of income, we can deduce Ml and


hence M 2 ; and we can then say that the rate of interest must be such
that people will want to hold this amount of money on capital account rather than transfer to bonds. But the object of finding the
rate of interest is to deduce the amount of investment and hence the
level of income. Are we then reasoning in a circle? The answer is,
no, we are merely faced with the inevitable difficulty of trying to
describe a system where the four variables mutually determine one
another. This can best be seen by a sort of mathematical shorthand,
where I stands for investment, S for saving, Y for income and r for
the rate of interest. Our four propositions are represented approximately by

= f (Y)
(2) 1= g (r)4

(I) S

(3) 1= S
(4) M

= Ll (Y) + L2 (r)

These are set down here to show that we have really got enough relationships. Mr. Keynes, quite rightly in my opinion, deprecates the
spurious air of exactness introduced by too much mathematics. But
4

If we accept the argument of the previous footnote, this should be


written I=h CY, r).

The General Theory .

107

in his endeavor to describe the system without this sort of shorthand


he has tended to obscure the fact that the determination is mutual.
This is particularly noticeable in the first paragraph of p. 248, where
he glosses over the fact that the rate of interest depends on M2 and
not on M when dealing with investment.
And now I must make the reviewer's inevitable apology for having
attempted the almost impossible and failed. There are any number
of important things in the book which I have not had space to mention. There is a useful chapter on the vexed question of a reduction
of money wages. This would be more convincing if it took account
of the fact that many non-wage-eamers relate their spending to the
money derived in the form of dividends, etc., from the production of a
previous period. If this money is made more valuable there is a real
reason why their spending should be greater proportion of their income from current operations. And Mr. Keynes concludes with a
chapter on the social implications of the theory; but this is rather like
showing Moses the Promised Land which he can never enter. A
blissful picture is drawn of a society which is supplied with all the
capital it needs and has reduced the rate of interest to zero; but a
large part of the book has been devoted to showing the difficulty of
doing this.
Parts of the book are undoubtedly difficult, but they are relieved
by less technical interludes, such as the magnificent description of
the operations of Wall Street. Moreover, Mr. Keynes has a clarity of
style which saves much confusion. If the propounders of rival theories would only set out what they take as given and what they seek
to find, as he does in Chapter 18, then we should be spared many a
headache. Many will consider that he underrates the dangers of expansionism, but that is surely better than the attitude of people who
forbid the State to increase employment for fear of "distorting the
structure of production," or because of the absence of any corresponding "genuine saving." The logic of the argument would be improved
if "the rate of interest" were not used so frequently to represent "the
cost of raising capital"; particularly in Australia the other elements,
such as quantitative control of credit, are often far more important,
and the rates applicable to different industries and borrowers may
move differently for institutional reasons. But whether or not one ac-

108

W. B. Reddaway

cepts the new doctrine, few will dispute Mr. Keynes's remarks about
the importance of the questions at issue. To take a single example:
"A decreased readiness to spend will be looked on in quite a different
light if, instead of being regarded as a factor which will, ceteris paribus, increase investment, it is seen as a factor which will, ceteris
paribus, diminish employment" (p. 185).

Keynesian Analysis and a


Managed Economy
[1963J

HE EDITOR HAS GIVEN HIS AUTHORS such a vast field from


which to select their subject that I can only make this embarrassing
choice by falling back on a somewhat artificial argument: as I was
included in the list because I reviewed the General Theory in 1936,
therefore I should discuss the ideas which now strike me on rereading
that almost-forgotten review.
This rereading suggested three main themes which I would have
liked to discuss: the incorporation of Government into the analysis;
the static nature of Keynesian analysis; and the factors determining
investment. Unfortunately, a brief trial soon revealed my error in
assuming that all three could be handled in the space allotted, so
that this chapter deals only with the first. The editor has, however,
granted me a dispensation to include a few notes on the last two in
an addendum.
For all three, my ideas spring largely from the little four-equation
"model" set out on page 106 of my review, and it is convenient to
reproduce it here in an alternative (but algebraically equivalent)
form, which puts the emphasis on consumption (a positive act)
rather than saving (often a residual).

Keynesian Analysis and a Managed Economy .

109

(1)
C =f(Y)
(2)
I =h(Y,r)
(3) C+I =Y
(4)
M =L1 (Y)+L s(r)

Perhaps I may repeat Keynes' warning (General Theory, p. 297)


that equations of this kind are deliberate oversimplifications, which
require us to examine provisional conclusions in the light of probable
interactions which are not shown in the equations.
THE TREATMENT OF GOVERNMENT
Macroeconomic analysis cannot dodge the problem of allowing for
the behavior of "Government"-taken in the widest sense to cover
everything from the smallest local authority to the central government, and also the central bank. This immediately raises a logical
problem: are these governmental bodies to be assumed to follow certain rules of behavior based on their own position, in the same way
as is assumed for individuals and companies, or are they assumed to
make "policy" decisions, designed to influence the economy as a
whole (or some section of it), with little or no regard for the consequences to their own position as separate accounting units?
The importance of this issue may be seen by a few examples. Thus
we might well be content to assume that small local authorities work
on the principle of a balanced budget,1 deciding how much they will
spend with reference to the ease of raising the necessary revenue (and
so probably spending more in times of prosperity, because they can
afford to). Such behavior is broadly similar to that of individuals,
and to a first approximation we could then, if we wished, treat the
consumption expenditure (and saving) of "local authorities plus individuals" as a single item, which rises as the national income rises.
(In other words, they could be treated as having a single, consolidated
"propensity to consume.")
With the central government, however, this seems to me a mistake,
even at the level of theoretical analysiS, and certainly anyone concerned with statistical applications in real life finds it almost impos1

Throughout this chapter I use this phrase in its literal sense, to mean
a budget which shows no significant deficit or surplus according to
some convention.

LIO

W. B. Reddaway

sible to work with a single figure for "consumption," to cover government expenditure on everything from armaments to zoological museums along with the ordinary purchases of individuals. Taxation
and transfer payments by governments have to be brought into the
picture in order to pass from the national income to private disposable
income, and private consumption and saving need to be related to
this;2 government expenditure on goods and services is then a positive
element of demand in its own right, broadly-but only broadly-corresponding with the negative effect which taxes have on personal
consumption.
The logical dilemma posed above then takes the form "what are we
to assume about the forces which determine government consumption, taxes, and transfer payments? Can we, even broadly, assume
that the government will always aim at a balanced budget, or some
other relationship between revenue and expenditure?"
In the simple model given in my review this is essentially what
was assumed-and I can invoke the authority of a letter from Keynes
for saying that I did not misrepresent his treatment. On such an assumption one can, as a first approximation, regard total consumption
(government and private together) as related to the national income
through a single "propensity to consume"-as shown in my first
equation. Admittedly this means that the propensity, i.e., the shape
of the function, will change if the government's policy about budgetbalancing changes; but if one regards big changes in such policies as
unlikely, one may accept this awkward feature for the sake of the
simplicity gained by having few variables.

A MORE GENERAL FORMULATION


Whatever may have been wise in 1936, however, I feel no doubt
that nowadays one should not "bury" two of the government's main
instruments of policy-tax rates and budgetary expenditure-as mere
invisible ingredients in a catch-all "propensity to consume," so that
the operations of fiscal policy can only show themselves through a
change in the parameters of a relationship which one would like to
regard as normally stable. The distinction between government consumption and personal consumption is in any case worth making for
2

I do not wish to get involved here with the further important question
of what determines the amount of company profits which is not distributed, and so does not become part of personal incomes.

Keynesian Analysis and a Managed Economy

I I I

its own sake; tax revenue also deserves a place in the list of mutually
dependent variables which we want to study. The literature on the
balanced budget multiplier shows the need to bring out these elements, even if the government did subject itself to rather narrow
behavioral rules by always aiming at a balanced budget.
In effect, then, we need to replace equation (I) by one which
shows consumption as made up of two parts: private (related to
private disposable income) and governmental. We might conveniently adopt the following notation:
P = private consumption
G = Government consumption
t = fraction of the national income taken in taxes
D = transfer payments made by the Government
and then-taking a very simple case which ignores many awkward
complications-we can replace equation (I) by equation (I a)
(Ia) C=P+G=f[Y (I-t) +D]+G
Now the main reason for writing down the equations in my review
was to see that there were the right number to account for the unknowns which have to be explained. This transformation of equation
(I) has introduced three new unknowns-t, D and G-and this immediately raises the question whether the system is determinate. The
answer must be that it is not, unless we attribute to the government
both certain objectives and certain views about how it wishes to attain them. Nor need this be regarded as in any way paradoxical: the
outcome clearly should depend on governmental policy, both as to
ends and to means. The original model left M, the quantity of money,
as essentially determined (in a modern economy) by the authorities,
thus recognizing the role of monetary policy as an influence on
events-whatever the objective of that policy might be: this discussion does no more than put fiscal policy explicitly in the picture in
the same way.
We must not, of course, exaggerate the extent to which the government can exercise its freedom of choice about macroeconomic
policies. For the near future at least, its expenditure (including transfer payments3 ) is likely to be determined within fairly narrow limits
3

Some transfer payments, notably those to the unemployed, are likely


to rise as the national income falls; this is one of the "built-in stabilizers" of the system. One could include this feature explicitly in the
model by another equation showing D as, in part, a (decreasing) func-

112

W. B. Reddaway

by past decisions or by factors which must be allowed to override


arguments based on the consequential impact on the state of the
economy: if more expenditure on defense is considered necessary for
reasons of national safety, any argument that the resultant increase
in demand would turn a fully employed economy into an inflationary
one would be met by a call for counteraction through higher taxes,
rather than by adjustment of defense expenditure to fit in with a
chosen economic policy. There is more freedom to vary taxation and
monetary policy, but even here there are limits to the extent of the
variations, imposed partly by convention (e.g., about what is "tolerable" by way of a budget deficit, or a scale of income tax, or a rate of
interest) and partly by technical considerations.
Nevertheless, while the government's freedom of action is circumscribed in various ways, it is still right to recognize that it exists, both
as to objectives and to means. Moreover, we must assume that the
government's policy will be to react to changes in the other factors
which govern the system, in an endeavor to maintain its objectives, so
far as this can be done by methods which it considers acceptable. In
effect, it has the last word in influencing the state of the economy,
even though there are limits to the power of that word. 4 Consequently it is quite illegitimate to state what the effects of, say, an increase in thrift will be without specifying a great deal about what
reaction from the government is assumed. Vague references to "ceteris paribus" or "assuming a neutral policy on the part of the government" simply will not do: the phrases are open to quite different
tion of Yj this would reSect the fact that it is the rules of the unemployment insurance scheme which are (for the time being) fixed, not
the total outgoings. But such an equation would also have to show a
large part Of D as independent of Y, and subject to governmental decisions about the scale on which all sorts of pensions and other payments are to be made.
4 I am using "the government" in a wide sense, to cover the central
bank as well. It is of course possible that the government (in the narrower sense) and the central bank may not see eye to eye about policy,
and I do not wish to debate who will then have "the last word" in
practice. Similarly there may be divisions of opinion between the
executive and the legislature, or between different departments, which
may also leave "the last word" in a somewhat unsettled place. In
Britain, there is a dictum that "Her Majesty's Government is one and
indivisible," but this may be considered more relevant to constitutional
theory than to everyday experience.

Keynesian Analysis and a Managed Economy . I 13


interpretations, each of which would be defended by some men of
good will as "the natural meaning," and the crucial part of the answer rests (and rightly so) on the choice of assumptions.

AN ILLUSTRATION
Let me illustrate this rather sweeping statement by showing how
one might rewrite the quotation from Keynes about the effects of increased thrift, given at the end of my review, assuming a closed
system.
The effects of a decreased readiness to spend [on the part of
individuals] depend essentially on how the government and
central bank react to it, and indeed on what policy they were
pursuing before. Two groups of possibilities may usefully be
set down, corresponding with two fundamentally different basic
assumptions, but this is far from exhausting the possibilities.
Basic Assumption A. The authorities have been pursuing, with
fair success, a policy of full employment without inflation, and
continue to do so.
On this assumption the level of the national income will be
unaffected by the increase in private thrift, and to ensure this
result the authorities may adopt any of the following measures
(or some combination):
(I) remit taxes or increase transfer payments, so that personal disposable incomes are increased, and personal consumption is maintained at the old level, despite the smaller proportion of these incomes which is spent. In this case personal
saving will be doubly increased, and government saving reduced (possibly becoming negative); total saving, investment,
the quantity of money and the rate of interest can remain
unchanged;
(2) increase government expenditure on goods and services
by the same amount as personal consumption falls, without
raising taxes. If the government expenditure is of the kind
classified as consumption, the increased private thrift (and resultant fall in private consumption) is simply offset by compensating changes in government thrift and consumption; if the
extra expenditure is classified as investment (e.g., making new
roads), extra private saving has financed extra government investment;
(3) pursue an easy money policy so as to raise private investment to match the increase in private saving;
(4) encourage private investment by other devices, notably

I I4

W. B. Reddaway

fiscal concessions-e.g., accelerated depreciation, tax holidays,


depletion allowances, government guarantees, etc.;
(5) encourage personal consumption by making credit easier
to obtain (e.g., by relaxing regulations about installment sales)
and so counteract the movement towards increased thrift.
Basic Assumption B. The authorities' policy is to have a balanced budget and a stable quantity of money, and to leave the
economy to look after itself.5
In this case increased thrift will mean lower consumption
and in consequence the national income and level of employment will fall. There will be a complicated series of further
consequences, which cannot be strictly described in a logical
"chain," because the variables are mutually dependent. Broadly,
however, we may pick out the following main features of the
new position:
(a) The government will have to raise tax rates and/or make
cuts in government expenditure, in order to preserve a balanced
budget, since the fall in the national income will lower the
revenue from existing tax rates (and there will also be a rise in
transfer payments to the unemployed). The resultant further
fall in private disposable income will mean a further fall in
consumption. 6
(b) The rate of interest will be lower, as less money will be
required for current transactions-see equation (4).
(c) Investment will be subject to two conBicting forces: the
lowering of interest rates will tend to stimulate it, but the depressed state of the national income will tend to reduce it.
There is no logical way of saying which force will be the
stronger, but many people may have intuitive ideas on the subject.
(d) Similarly, private saving will be subject to two different
forces: the increase in thriftiness will mean that a larger proThis assumption is intended to reflect, in one flausible manner, the
idea that policy should be assumed to be "neutra ." It is not, of course,
the only possible interpretation: one might, for example, say that a
neutral fiscal policy meant stability in tax rates and expenditure principles, rather than a balanced budget, and that a neutral mon~tary
policy meant stability in the re-discount rate rather than in M.
6 In case it may seem fanciful to younger readers to imagine that governments might react in this way, let me remind them that this is
iust the kind of reaction which was observed in the depression of the
early 1930'S in a number of countries, reflecting the ideas which were
current before "the Keynesian revolution". Thus the British Labour
Government adopted just such a policy in its 193 I Budget, which it

Keynesian Analysis and a Managed Economy

I IS

portion of disposable incomes will be saved, but disposable incomes themselves will fall, both because of a fall in the national
income and because of the rise in tax rates needed to preserve
the balanced budget. (Government savings will, ex hypothesi,
remain at zero, since we are assuming a balanced budget).
This answer is clearly a somewhat cumbrous affair, and it remains
so even if the movements in the main variables are summarized in a
table, such as that shown on page 116. A cynic may ask whether economics is much use as a science when the table shows that, for
example, private saving may rise for two reasons, or rise for one
reason, or remain unaffected, or possibly fall-depending on what
assumption one makes. ''You pay your money and you take your
choice" or even "The Emperor has no clothes on" may suggest
themselves as comments.
Nevertheless, none of the assumptions which we have taken as
possibilities are in any way absurd as interpretations of "ceteris
paribus." It would indeed be possible to add further plausible alternatives which would lead to some variables which at present feature
as moving only in one direction (or remaining unchanged) being
shown to move in the opposite one: thus if the Basic Assumption B
were interpreted to mean keeping interest rates constant, the quantity of money would be shown as falling "in accordance with the
reduced needs of trade." There is nothing for economists to be
ashamed of in all this: economics provides a technique of analysis,
and the problems are not specified until the assumptions are statedand different assumptions make different problems and different
answers.
The big contrast comes, of course, between the two basic assumptions, with Keynes assuming that the second one was the "natural"
assumption to make about the actions of the authorities, so that any
other action (such as is assumed in the A series), would have to be
regarded as a separate factor. In effect, he accused the classical economists of assuming that the economic system contained within itself
presented with a great display of self-righteousness. When the Labour
Government seemed to be weakening in its resolution, and was replaced by a National Government pledged to "sound finance", that
Government received an overwhelming majority in the general election.

116

W. B. Reddaway

some force which would produce continuous full employment, and


said that such an assumption was not justified. What I am suggesting
is virtually that-at least since the Keynesian revolution-neither
position should be adopted as "natural" or "automatic": full employment is neither guaranteed by the automatic working of the system,
nor yet an implausible assumption to take as first approximation,
since the government may well make this a high-priority objective.
At least it can justifiably be taken as one of several alternative assumptions, and made explicitly; there is no single thing which can
claim to constitute either a neutral fiscal policy or a neutral monetary policy, and, even if there were, there would be no particular
plausibility in basing one's analysis exclusively on the assumption
that the Government would remain "neutral" in this limited, technical sense.
CHANGES DUE TO INCREASED PRIVATE THRIFTINESS
0= no

change; + = increase; - = decrease; ++ and - - imply


that the movement is caused by more than one force
Item

National Income
Consumption
Personal
Government
Total
Investment
Private
Government
Saving
Private
Government
Interest rates
Quantity of money
Tax rates

Assumption (as defined in the text)


AI

A2'"

---

A3

A4

A5

--

-- -- --

+
oor-

oor+

++

+
-

0
0

0
0
0

-oor-

--

?-

?-

'" Where alternative entries appear, the first assumes that all the additional Government expenditure is classified as consumption, the second
that part is classified as investment.

Keynesian Analysis and a Managed Economy . I 17


NEUTRALITY BETWEEN COMPONENTS
To this last statement there might seem to be one exception.
Granted that governments should not be assumed neutral or indifferent about the level of aggregate demand, and must be assumed
to try to bring it to the level they consider appropriate, is it sensible to
assume "neutrality" between the constituents of that aggregate?
The meaning of this question may perhaps be seen most easily
from an imaginary official statement about the government's reaction
to a recent increase in thriftiness:

It is the policy-indeed the duty-of the government to take


such action, in collaboration with the central bank, as will ensure the maintenance of full employment without inflation. But
in choosing between the methods available for securing this
result, the government's policy is to be strictly neutral between
those which operate through consumption and those which operate through investment. Over the last year the government
has observed that the people of this country have shown a desire
to use an increased proportion of their incomes for accumulating wealth, rather than for immediate consumption. The government has thought it right that this freely expressed preference of the country's citizens should be duly reflected in a
greater national total of investment and a smaller total of consumption. It has therefore resisted suggestions for lowering
taxes or raising government expenditure of a recurrent character: instead, it has encouraged the central bank in its policy of
cheap and easy credit, and-since this seemed unlikely in itself to produce a sufficient stimulus to investment-it has accelerated the much needed roadbuilding program. In consequence, the increased thriftiness of the citizens is duly reflected
in a large increase in their holdings of bank deposits and securities of many kinds (including government obligations), and
the country's real capital assets-whether in Government or
private ownership--have increased by an amount greater than
that recorded for any previous year.
Nevertheless, I do not myself attach any importance to the search
for a criterion of neutrality, even as a basis for exercises in academic
analysis. My reasons for this are numerous, and I instance the following only as examples.
1. It may seem possible to define such a criterion (as in this example) for the government's reaction to a single change, but this
only helps if one assumes that the policy was correctly neutral before

118

W. B. Reddaway

the change. A modem government is necessarily in the business of


regulating the economy through nscal and monetary policy all the
time, so that advocates of neutrality must denne what it is to mean
in relation to all the complexities of a country's economy.
Quite apart from the question whether the criterion has any claim
to be regarded as a guide to "correct" or "desirable" policy, rather
than simply a benchmark in a scheme of academic analysis, this is
extremely difficult to do. "A balanced budget," for example, means
entirely different things according to the conventions adopted in the
U.S.A. and the U .K., and very different pictures are presented in the
latter country by the Treasury (in the nnancial statements) and
the Central Statistical Office (in the national income statistics).
2. There is a fundamental conflict between trying to denne a criterion by reference to ultimate economic consequences on the one
hand or statistical convenience (or superncial appearances) on the
other. Thus "a balanced budget" may seem (if we can agree on the
dennition) a good concept to adopt as a benchmark, corresponding
with zero government saving, and so with the idea of budgetary neutrality between consumption and saving (or investment) as constituents of any given size of national income. Whether or not one
regards such a policy as desirable, at nrst sight this might seem acceptable as a benchmark by reference to which policies are described
or classined.
Unfortunately, however, a little analysis shows this to be quite
incorrect. A given national income and a balanced budget might be
equally consistent with "death duties and a steeply progressive income tax, plus dear money" or "consumption taxes and heavy national insurance contributions, plus cheap money"-the former
policy implying more consumption, the latter more investment. 7 An
economist can hardly be satisned with a criterion of neutrality denned by reference to one part of the picture (government saving),
however easily observed, if that can be brought to zero by methods
which have very different effects on the balance between total
saving (or investment) and total consumption.
7

Similarly a budget balanced at a high level, with much Government


expenditure on (say) armaments and education, will need to be
coupled with dear money or other restraints or investment, if it is to
be consistent with the same level of total demand or national income
as a budget balanced at a low level.

Keynesian Analysis and a Managed Economy . I 19


3. One must also allow for the fact that the government will not in
practice succeed in keeping the economy dead on course. If inflation
or depression has set in, the logical meaning of neutrality must
presumably be as follows: the government's policy is "neutral" if the
criterion will (probably) be fulfilled when the general level of
demand is back at the target figure. The test is not that the measures
adopted by the government to get the economy back on course are
neutral: in a recession the Government may take drastic measures
to stimulate investment, but still be declared to be following a neutral policy, because investment had fallen disproportionately and is
hard to revive.
As a piece of strictly academic analysis this work of classification
may have some intellectual interest, but the scope for misunderstandings is obvious.
4. This brings me to the argument which in a sense carries most
weight for an applied economist. Discussions of "neutrality" in
macroeconomic policy may start as a search for a logical benchmark
to be used in purely academic analyses, but the benchmark is all too
likely to become confused with a guiding light. If it seems easily
understandable-e.g., a balanced budget-there may grow up a
political doctrine that the government should aim for it under all
circumstances, even if this involves imposing taxes to undo the effects
of the automatic stabilizers in a recession, and regardless of whether
the country's normal definition of a balanced budget makes economic
nonsense by including obvious capital items in the current account.
It is difficult enough in practice for the authorities (who may not all
work in perfect harmony) to keep the aggregate level of demand
right, even if they are not constrained in their use of the available
weapons by anything more than the "commitments" mentioned
above, the need to avoid violent stop-and-go impacts on particular
industries, and the general necessity of considering the composition
of demand as well as its total.
This last point has of course far more aspects than I can possibly
handle here. Thus the government of every country must consider
how far taxation is to be used to increase collective consumption at
the expense of private consumption; and the government and citizens
of an underdeveloped country may wish to raise national savings
deliberately by a regular series of budget surpluses, in order that
more investment may be done than would be covered by private

120

W. B. Reddaway

savings. These things are very important: a country needs other


methods for its citizens to express their preferences about what shall
be produced besides individual purchases and sales in the market
place.
Adjustment of aggregate effective demand to keep it at the target
level is also an important function of government, and our macroeconomic analysis must take heed both of the fact that the attempt
will be made and of the varying methods which may be adopted;
but these involve using the same instruments as government must
use in pursuing other objectives, and its use of these instruments
must be judged in relation to all the objectives taken together.

ADDENDUM
NOTES ON INVESTMENT AND TIME
There can be no question that Keynes' analysis was in its essence
a description of short-run static equilibrium, rather than of a process
through time. The equations do not "date" the variables, and the
crucial paragraph on page 102 of my review starts by "visualizing" a
series of levels of employment, and ends by saying that the actual one
must be that which corresponds with the level of investment (because otherwise there would be a disequilibrium). There is no attempt to show how one position develops out of another; if
investment changed to a new rate (and stayed there) we are not
told what the process would be by which employment would move to
the "corresponding" level, or how long it would take.
Many attempts have been made to deal with this obvious illogicality by introducing lags of various kinds-notably a lag between the
earning of income and its impact on consumers' spending, and a lag
between spending by consumers and the "resultant" production of
equivalent goods. These lags provide a link between the position in
the present period and that in one or more subsequent periods, and
so enable one to produce a model of a process through time, provided
that either there are sufficient links to give at least the key figures for
the later periods, or information about the missing items can be fed
in from outside.
In a closed system, two things seem to me to be of outstanding importance for determining the course of events: government actions

Keynesian Analysis and a Managed Economy

121

and investment. I have suggested in the body of the chapter that


the former should be treated as largely determined from outside the
model, but as reacting to all the other factors so as to try to ''have the
last word." There are of course limits to the changes in policy which
can be made effective within a short period, so that a very short-run
projection may perhaps be made "assuming no change in Government policy" without implying that this is particularly likely, and
without giving an alternative one to correspond with other policies;
such a projection is virtually an unconditional forecast, and it may
be carried further forward if it yields results which do not seem
likely to provoke a change in policy. But it seems to me unrealistic to
construct elaborate theoretical models dealing with the development
of an economy over long periods without giving any role to conscious
governmental policy (whether monetary or fiscal): this applies
whether the models are designed to produce cycles or steady growth
or anything else-unless, of course, one then deduces what conscious
policy should be, in order to produce more acceptable results, and refashions the model to include policy as one of the determinants.
Investment
The determinants of investment have been the subject of much
discussion: Keynes' treatment was clearly inadequate as an explanation of a continued flow of expenditure, as opposed to the expenditure which would seem, at some one time, to be worth incurring over
a very ill-defined period (see in particular The General Theory, pp.
136-7). I can do no more here than stress that for short-period
Keynesian analysis it is investment expenditure incurred in the
period which is relevant, and not decisions taken in the period to
embark on new projects (which is what one most naturally "derives" from the current state of the economy or its growth, with due
allowance for expectations and projects already started but not completed), nor yet this period's addition to effective capacity (which is
what raises the possible level of output in the period).
It is perhaps worth emphasizing the length of some of the lags
which are involved. In 1958 the Royal Institute of British Architects
made a very interesting inquiry into the time taken at various stages
of a building operation. Their sample showed the following average

122

W. B. Reddaway

time lapses from receipt of the architect's first instructions:


Sketch plans completed
Planning and other approvals obtained
Tenders out
Tenders returned
Working drawings completed
Building work started
Building work completed

7 months
14
"
15
16
16
20

38

"
"
"
"
"

The average time to the completion of the work, shown as 38 months,


was subject to very big variations between projects and project-classes,
with the average for projects under '10,000 being 18 months and
for those over '1 million 50 months. It includes nothing for the
period in which the client is deciding to approach the architect,
which may be quite long if board discussions about the wisdom of
the project are involved, especially if there are alternative proposals.
On the other hand, for simple house-building an architect might not
be employed and the time would be shorter; and some machines can
be bought "off the peg."
The length of these lags has a convenient result for short-term
analysis: to a reasonable approximation the level of fixed investment
over the next six months or a year can be treated as given "for better
for worse" by past decisions, and only in small degree influenced by
events within that period. We do not have to worry (much) about
the determinants of investment for forecasting what will happen in
this period-though the government has to worry about getting
broadly the right amount of decisions taken, to keep next year's
expenditure at a satisfactory level. 8
8

An explanation of investment decisions is also needed, of course, for


longer-term analyses. This problem is rendered more complex by the
fact that this year's decisions must take heed of schemes already under
construction, which nevertheless do not influence this year's capacity
or the current profitability of the industry: many theoretical models
seem to proceed on the oversimple assumption that all decisions are
irrevocably taken on I January (on the basis of the output of earlier
periods and existing capacity) and fully executed by 3 I st December,
so that the year's decisions equals the year's investment expenditure
equals the addition to effective capacity by the time more decisions are
taken.

Keynesian Analysis and a Managed Economy

123

Other lags
Government policy and investment seem to me, as I said above, to
be the really important factors for tracing the movements of the variables through time. Moreover, each has its own time lags, and sudden
large changes in any determining flow of expenditure are improbable except in quite abnormal circumstances (e.g., war or hyperinflation)'
In view of this I have come to believe that the lags mentioned at
the start of these notes are normally of relatively little quantitative
importance; thus I believe that one's picture of what happens will
normally be as good if one assumes that this week's purchases are
influenced by this week's "accrual" of wages as it would be if one
related the purchases to what was received at the end of last week. 9
Similarly, little is gained by assuming that production of consumer
goods reflects recent sales, rather than a reasonably correct anticipation of what is currently being sold.
Only one lag seems to me long enough to be worth some attention
in the area of "income generation-consumption-production of consumer goods," and that is the lag between the earning of profits by a
company (which we may think of as "centered" at the midpoint of
its financial year) and the receipt of dividends by shareholders: in
the U.K. the average for this is probably nearer to a year than to six
months. The interval before this income influences consumers'
expenditure may well be further increased by a slow reaction on the
part of the shareholder in spending his dividend. Quantitatively,
however, the amount of this income (net of tax) is only about 5 per
cent of disposable personal incomes.
There is an even longer lag before the other ultimate "owner" of
the profits (the Chancellor of the Exchequer) receives his share. But
one may hope that he is sufficiently sophisticated to prevent this lag
from having any significant effect on his actions.
9

Indeed, I would personally say that the use of credit and changes in
cash balances make the week's accrual more plausible as a guide whenever the difference is significant. I have never believed in the picture
of a newly unemployed man spending last week's wages without any
regard to his change in circumstances, or a newly employed one buying
nothing until pay day arrives.

R. F. HARROD

Mr. Keynes and


Traditional Theory
[1937J
k

THIS PAPER I do not propose to ask or answer the question, has Mr. Keynes succeeded in establishing the propositions
which he claims to have established? nor again, what kind of evidence is required to establish or to refute those propositions? I shall
confine myself to a narrower question, namely, what are the propositions which Mr. Keynes claims to have established? And in order to
restrict my subject matter still further, I propose to confine myself to
those propositions, which he claims to have established, that are in
conHict with the theory of value in the form in which it has hitherto
been commonly accepted by most economists. In other words, my
question is what modifications in the generally recognized theory of
value would acceptance of the propositions that Mr. Keynes claims
to have established entail?
In order to clarify the issues involved it may be well to divide
commonly accepted theory into the general theory and its specialized
branches. The general theory consists primarily of a number of functional equations expressing individual preference schedules and a
number of identities, such as that supply must be equal to demand,
and the elucidation of such questions as whether there are as many
equations as there are unknowns and whether the solutions are single
or multiple. The result of these enquiries should make it clear
whether the equilibrium of the system as a whole is stable or unstable
or undetermined, whether there are alternative positions of equilibrium, etc. There may be some clues as to the general form of some of
the functional equations, provided by such principles as the Law of
Diminishing Utility, to use old-fashioned terminology, which may
make it possible to predict the direction of changes in the values of
124

Mr. Keynes and Traditional Theory'

125

the various unknowns due to a given change in one of them. More


precise prediction can only be achieved if or when it becomes possible, as a result of the labors of such investigators as Dr. Schultz, to
write down the actual terms of the functional equations. Within the
corpus of this general theory may be included the formulation of the
market conditions that are required for the realization of some kind of
maximum. Thus if one individual A is indifferent whether he produces commodity X or commodity Y for a certain consideration, and
another individual B prefers X to Y, the maximum is not realized if
the market so operates that A normally produces Y and not X for B.
On this condition the maxim of Free Trade is fairly securely
founded, the more general maxim of laissez faire much less securely
so.
In contrast with the theory of value in this very general form, may
be set the special theories formulated to deal with specific problems
such as interest, profit, joint production, discriminating monopoly,
etc. The normal method used in dealing with these departmental
studies is to assume that certain terms, which appear as variables in
the general system of equations, may be treated as constants for the
special purpose in hand. For instance, in studying the behavior of
duopolistic producers of a given commodity, it may be assumed that
the duopolists can obtain the services of factors of production at rates
the determination of which in the market will not be appreciably
affected by the duopolists' behavior. Such methods constitute short
cuts to the unraveling of particular problems and they are often perfectly legitimate. In the minds of most economists, other than those
who stand, so to speak, at the philosophical end of the economic
array, the conclusions reached by these short-cut methods constitute
the main findings of economic theory.
I may say at once that in my opinion Mr. Keynes' conclusions need
not be deemed to make a vast difference to the general theory, but
that they do make a vast difference to a number of short-cut conclusions of leading importance. Thus to those whom I may perhaps call
without offense the ordinary working economists they ought, if accepted, to appear to constitute quite a revolution. Whether they entail a substantial modification of the more general theory depends on
how that is stated. I need hardly observe that there is no authorized
version. Those whose main interest is in the general theory, may, if

126

R. F. Harrod

they have laid their foundations well and carefully, be able to look
down with a smile of indifference on the fulminations of Mr.
Keynes. Pavillioned upon their Olympian fastness, they are not likely
to show much irritation.
It is convenient to take Mr. Keynes' theory of interest as the starting point of this exposition. In the commonly accepted short-cut
theory there are two unknowns and two equations. The two unknowns are the volume of saving (= the volume of investment) and
the rate of interest. Of the newfangled view, sponsored by some outof-the-way definitions in Mr. Keynes' Treatise on Money, that the
volume of saving may be unequal to the volume of investment, it is
not necessary to say anything, since it has played no part in the
standard short-curt formulations of interest theory (although it has
figured in recent writings concerned with practical monetary problems). The commonly accepted interest theory from the time of the
early classical writers onward entails that saving is always and
necessarily equal to investment.
The two equations in the traditional theory of interest correspond
to the demand and supply schedules relating to a particular commodity. First there is the demand equation:

y=f(x),
y, the marginal productivity of capital, depending on x, the amount
of capital invested per unit of time. So much capital will be invested
that its marginal productivity is equal to the rate of interest; that is,

y=y',

where y' is the rate of interest. Since both the traditional theory and
Mr. Keynes hold that investment is undertaken up to the point at
which the marginal productivity of capital is equal to the rate of interest, y' may be suppressed, and y made to stand for the rate of
interest which is equal to the marginal productivity of capital.
Then there is the supply equation:

x=p(y);
x, the amount which individuals choose to save, which is equal to the
amount of investment, depends on the rate of interest. Thus there
are two unknowns, the rate of interest and the volume of saving, and
sufficient equations to determine them. It is not necessary for the
present purpose to consider controversies concerning the forms of

Mr. Keynes and Traditional Theory'

127

these equations, such as whether a rise in the rate of interest tends to


cause people to save more or less.
This treatment of interest and saving is analogous to that of the
price of a particular article and the amount of it produced. The treatment depends on the short-cut assumption of ceteris paribus. This is
often legitimate in the case of particular commodities, although it is
recognized that in certain cases it is idle not to bring in certain other
variables, for instance the prices of close substitutes. Among the
"other things" which are supposed to be "equal" is the level of income in the community under discussion. In many cases it may be
true that when we are trying to determine how much of a particular
commodity a producer is likely to produce, his decision to produce a
little more or a little less will not have a sufficiently large effect on the
total income of the community to react on the market for his goods in
such a way as to make an appreciable difference to him. This
particular short cut is in that case justified. I suggest that the most
important single point in Mr. Keynes' analysis is the view that it is
illegitimate to assume that the level of income in the community is
independent of the amount of investment decided upon. No results
achieved by the short cut of such an assumption can be of any value.
How does Mr. Keynes' analysis proceed? His first equation is substantially the same as that of the traditional analysis,

y=fCx).
The marginal productivity of capital is a function of the amount of
investment undertaken. The marginal productivity of capital appears
in Mr. Keynes' book under the title of marginal efficiency. It does not
appear that there is a difference of principle here. It is true that Mr.
Keynes makes an exhaustive and interesting analysis of this marginal
efficiency and demonstrates that its value depends on entrepreneurial
expectations. The stress which he lays on expectations is sound, and
constitutes a great improvement in the definition of marginal productivity. This improvement, however, might be incorporated in
traditional theory without entailing important modifications in its
other parts.
When we come to the second equation the level of income must be
introduced as an unknown term, giving
x=Cy, i),
where i is the level of income. The amount of saving depends not

128

R. F. Harrod

only on the rate of interest, but on the level of income in the


community.
It might be thought that to introduce the level of income as an unknown at this point is tantamount to abandoning all attempt to have
a departmental theory of the volume of saving, since the level of total
income appears in all the equations of the general theory and it is
impossible to determine its value without taking all factors into account. This would mean that we should have to leave the ordinary
working economist without any departmental theory of saving and
interest which he could grasp, and to let him flounder in the maze of
n X r X s, etc., equations governing the whole system. Mr. Keynes has,
however, come to the rescue and carved out a new short cut of his
own. In his view the value of the unknown level of income can be
determined in a legitimate and satisfactory manner by the departmental equations relating to saving and interest only. To the legitimacy of this assumption it will be necessary to return presently.
Meanwhile, since there are three unknowns and but two equations in the savings/interest complex, another equation is needed. Before proceeding to that, it may be well to recur to the second
equation,
x==: S?lCy, i).
This may be transposed into the form
i=t{lCx, y).

The level of income depends on the amount of investment C= that of


saving) and the rate of interest. In this form the second equation
shows itself as the doctrine of the multiplier. The multiplier is the
reciprocal of the fraction expressing the proportion of any given income, which, at a given rate of interest, people save. If the value
of the multiplier is known for any given rate of interest and level of
income, the actual level of income can be deduced directly from the
volume of investment. Those to whom the doctrine of multiplier
seems an alien morsel in the corpus of economic doctrine should remember that it is merely a disguised form of the ordinary supply
schedule of free capital, but with the level of income treated as a
variable.
In discussing this doctrine, for the sake of a still shorter cut, Mr.
Keynes is inclined to let the rate of interest drop out of sight. Thus
the equation becomes

Mr. Keynes and Traditional Theory

129

i=1frCx);
the level of income depends on the volume of investment. The justification for this procedure is that whereas the relation of the level of
income to the amount of investment is in the broadest sense knownit may be assumed that people save a larger absolute amount from a
larger income-the relation of the amount which people choose to
save to the rate of interest is a matter of controversy. Moreover in
Mr. Keynes' view the level of income has a more important effect on
the amount which people choose to save than the rate of interest.
However, there is no need to pick a quarrel here. The rate of interest
may be brought back into this part of the picture without affecting
the main argument. The propensity to consume may be regarded as
depending on the rate of interest, although for the sake of brevity and
clarity mention of this need not be insisted on at every point in an
exposition of the doctrine of the multiplier.
What of the third equation? We have

y=xCm),

where m is the quantity of money, a known term, depending on


banking policy. This is the liquidity preference schedule. Probably
i, the level of income, ought to be inserted in this equation, thus:
y=xCm, i),
since the amount of money required for active circulation by consumers and traders depends on the level of income. Ought not the
price level to come in also? That may be taken to be subsumed under
i, the level of income, in a manner that I shall presently explain. The
residue of money, not required for active circulation, is available for
ordinary people who are discouraged by their brokers from immediate investment, and, more important, for firms, who may want
cash for capital extensions or similar purposes within six months or a
year or two, and are unwilling to hold their reserves in the form of
securities to which some risk of depreciation within the prescribed
period is attached. Since the amount of money available for liquid
reserves is strictly limited and cannot be increased by the mere desire on the part of firms to hold more money than that, the prospective
yield of less liquid reserves must be sufficient to confine those who
insist on a money reserve to the amount of money available for that
purpose. The less the amount of money available the higher the rate
of interest will have to be, both because the high rate is a quid pro

130

R. F. Harrod

quo against the risk of depreciation of the capital and also because
the higher the present rate the less probability is there of depreciation within the prescribed period.
It is not necessary to give a final pronouncement on the significance of the liquidity preference equation. It appears that even if
some modification is required in this third equation, which determines the rate of interest, a type of analysis similar in its general
structure to that of Mr. Keynes may be maintained.
We now have three equations to determine the value of the three
unknowns, level of income, volume of saving (= volume of investment), and rate of interest (= marginal productivity of capital).
For the working economist these results may be set out in still
briefer shorthand as follows. The amount of investment (= amount
of saving) depends on the marginal productivity of capital and the
rate of interest; the level of income is connected with the amount of
investment by the multiplier, i.e., by the propensity to consume; and
the rate of interest depends on the desire for liquid reserves and the
amount of spare cash in the community available to satisfy that desire. The amount of this spare cash depends on the policy of the
banks in determining the quantity of their 1.0.U.'s that are outstanding and on the level of income (the higher this, the more money
will be taken away into active circulation).
Thus if the schedules expressing the marginal productivity of
capital, the propensity to consume, and the liquidity preference are
known and the total quantity of money in the system is known also,
the amount of investment, the level of income and the rate of interest may readily be determined.
The next topic for consideration is the legitimacy of the assumption that the level of income may be regarded as determined by the
complex of considerations expressed in the savings/interest equations, rather than by the whole system of equations. In general the
level of activity is traditionally conceived as depending on the preference schedules of the various factors expressing their willingness to
do various amounts of work in return for income, and on the
schedules expressing the relation between the amount of work done
and the income accruing from it (Laws of Returns). In considering
the former schedules we have to take into account all the factors of
production. Now in Mr. Keynes' system the supply of capital has

Mr. Keynes and Traditional Theory

I3I

already been dealt with by the savings/interest equations. For the


supply of risk-bearing, we may provisionally content ourselves with
the elegant device which he provides in his footnote to page 24. He
writes, "by his (the entrepreneur's) expectation of proceeds I
mean, therefore, that expectation of proceeds, which, if it were held
with certainty, would lead to the same behavior as does the bundle of
vague and more various possibilities which actually makes up his
state of expectation when he reaches his decision." Thus considerations affecting the supply of risk-bearing are subsumed in the equations which determine the volume of investment.
There remain the factors other than those covered by the category
of investment. Of these we are only concerned with those the supply
of which can be varied. Thus we are left with those which may
roughly be designated prime factors. What is the nature of their
supply schedule? What is the form of their preference for income in
relation to the work required to obtain it?
In this field Mr. Keynes' argument is vitally dependent on his observation of real conditions. The work/income preference schedule
exerts its power upon the economic system through the terms on
which the prime factors are willing to sell their services. The contracts or bargains of the entrepreneurs with prime factors are normally fixed in money, with no proviso regarding the general level of
prices. In the exceptional cases in which there is such a proviso, it is
none the less usually the case that a rise in prices involves some fall
in real rewards to prime factors and conversely. It is true that in a
time of rising prices the factors may press for a rise in rewards, but,
even if they achieve this, there is still no proviso to safeguard them
against a further rise of prices, and prices may, for all the new
bargains lay down, and indeed are very often in fact observed to, run
on ahead of rewards. Conversely in a time of falling prices. This
gives the supply schedules of the prime factors a very special kind of
indeterminacy which undermines their power to determine the
general level of activity. Mr. Keynes discusses this matter in Book I
and its importance in his logical edifice justifies him in giving it
pride of place.
Consider next the second set of schedules determining the general
level of activity, namely those expressing the relations between the
amount of work done and the income accruing from it (Laws of Re-

132

R. F. Harrod

turns). Since the bargains with prime factors are expressed in money,
the returns due to their employment should be expressed in money
also. But the money value of these returns depends on the level of
prices. The general price level might be regarded as determined by
the Quantity Theory of Money; Mr. Keynes does not so regard it for
reasons which will be explained below. On the contrary he regards
the general price level as completely malleable and determined by
the equations in the general neld without reference to the quantity of
money.
The consequence of the conclusions yielded by the interest/savings equations, if these are accepted, is, that the level of income and
activity is determined. Now suppose the entrepreneurs decide to
produce more than the amount so determined. Owing to a dencient
propensity to consume, they will nnd dencient purchasing power,
and either accumulate stocks or sell at a loss. If they do the former
the accumulation of stocks will constitute an additional (involuntary)
investment on the part of the community, which when added to the
intended investment, makes the total investment of the community
such as to be consistent, in accordance with the interest/savings equations, with the higher level of activity which entrepreneurs are
choosing to indulge in. But such a position is unstable. So long as
stocks are accumulating, they will reduce activity and continue to do
so, until it reaches the point indicated by the interest/savings equations. If on the other hand they sell at a loss, they will be dis-saving;
the propensity to consume will be temporarily raised, so that the
higher level of activity which they are choosing to indulge in becomes
consistent with that required by the interest/savings equations. But
again the position is unstable. The marginal propensity to consume
will not be permanently sustained at an abnormally high ngure by
these means. To avoid losses, entrepreneurs will restrict and continue
to do so, until activity and income are reduced to a level which satisnes the interest/savings equations, with the marginal propensity to
consume normal for that level of income. Converse arguments would
apply in the case of entrepreneurs deciding to produce too little.
Now if the level of activity so determined is indeed the equilibrium level of activity, the price level must be appropriate to it. Let us
suppose that the price of each commodity is determined by the
marginal money cost of production, in the crude way that a tiro might

Mr. Keynes and Traditional Theory Z33


describe erroneously supposing himself to be explaining the true
classical theory of cost of production. If the law of diminishing returns prevailed on balance, as Mr. Keynes supposes that it does anyhow in the short period, the general price level would be expected
to rise with increases of output and to fall with decreases. To make
the matter still more crude and common, suppose prices to vary not
merely in proportion to changes in the number of units of factors
required per unit of output, as output varies, but also in proportion
to changes in rates of reward to the factors. In this case we should
find, as output rose and diminishing returns came into play, that the
rise of prices would just sufficiently exceed the rise of wages, etc., if
any, to cover the increased real marginal cost of production per unit.
Factors might press for a rise of rewards, but though they might gain
on balance in some trades, they would always be beaten by the price
level in the system as a whole.
Now this is precisely what Mr. Keynes supposes actually to happen. It is, however, "subject to the qualification that the equality
(between marginal cost and price) may be disturbed, in accordance
with certain principles, if competition and markets are imperfect"
(p. 5). The objections to this view which upholders of the Quantity
Theory of Money might raise must be considered. But first observe
its relation to the determination of the level of activity.
Take a period within which prime factor bargains do not change.
The supply of each of these in money terms may then be represented
by a horizontal straight line. But if prices vary in proportion to costs
(cost variations including allowance for overtime rates, the employment of less efficient labor, etc.), then the money value of the marginal net product of each factor must be represented by a coincident
horizontal straight line. Therefore on these conditions the two sets of
schedules leave the level of output entirely indeterminate. If the
matter is expressed in real terms both sets of schedules are downward
sloping to the right; they are still coincident. If money rewards to
factors are raised or lowered in response to changes in the level of
employment and prices are adjusted accordingly, the same result
ensues. Thus this complex of equations does not determine the level
of activity; therefore it leaves that level free to be determined by the
savings/interest complex. Q.E.D.
Thus the crux of the matter seems to have shifted to the Quantity

134 . R. F. Harrod
Theory of Money. The essence of the difference between the traditional theory and Mr. Keynes' theory can be put thus: In the traditional theory the supply and demand schedules of all the factors stand
on the same footing; the level of activity is an unknown, but the
price level is determined by the monetary equation. This determination of the price level enables the level of activity to be determined
by the factors' money supply schedules, and by their marginal productivity schedules. In Mr. Keynes' theory the level of activity is
determined by the equations governing the savings/interest complex. In the general field, in which we are now only concerned with
the demand and supply of prime factors, the level of activity is conceived as determined ab extra. It is a known quantity. But the price
level is conceived to be completely malleable. If it were not the system in the general field would be overdetermined. Thus the monetary equation is shorn of its former powers. The level of activity
being a known quantity the price level is determined by the money
cost of production, with suitable modifications for imperfect
competition.
What right has Mr. Keynes to gut the monetary equation in this
way? Has, then, the banking policy no power to influence the situation? Yes, certainly it has. The fact is that the power residing in the
monetary equation has already been used up in Mr. Keynes' system
in the liquidity preference equation and it cannot therefore exert any
direct influence in the general field. To make it do so would be to
use its determining influence twice over. In fact in Mr. Keynes' system all the old pieces reappear, but they appear in different places.
Explanation is necessary. It will be remembered that according to
the liquidity preference equation, the rate of interest is determined
by the desire of people for liquid reserves and the quantity of money
available for that purpose. The quantity of money available for that
purpose is equal to the total quantity of money in existence less that
required for active trade. 1 Now if the quantity required for active
1

In his liquidity preference equation Mr. Keynes includes the demand


for money for whatever purpose, and the quantity of money that appears in it is the total quantity of money in the community. It has
appeared simpler in this part of the exposition to divide this total into
two parts, the amount required for active circulation and the residue,
to define the quantity of money which appears in the liquidity preference equation as that residue, and the demand which the equation

Mr. Keynes and Traditional Theory 135


trade were perfectly indeterminate, as it must be by the Quantity
Theory-for according to that the price level depends on the quantity
of money available for active trade, and therefore it is unknown what
quantity of money any given amount of active trade will absorb---the
residue would be indeterminate also. But if the m in
y=X(m)
is indeterminate, there are too many unknowns in the interest/savings set of equations. Thus it is necessary to the validity of Mr.
Keynes' solution of the problem of investment and interest that the
amount of money available for liquid reserves should be determinate,
and that involves that the price level should be determined otherwise
than by the monetary equation. And so, in Mr. Keynes' system it is.
The matter may be put thus: The savings/interest equations suffice to determine the level of activity, subject to the proviso that the
quantity of money which appears in the liquidity preference equation
is a known quantity; and this will be known if the price level and
therefore the amount absorbed in active trade is known. The equations in the general field suffice to determine the price level, subject
to the proviso that the level of activity is known. Thus there is after
all mutual dependency. The level of activity will be such that so
much money is absorbed in active trade that the amount left over
enables interest to stand at a rate consistent with that level of activity.
The mutual interdependency of the whole system remains, but the
short cuts indispensable to thinking about particular problems, as Mr.
Keynes has carved them out, remain also.
The amount of investment depends on the marginal productivity
of capital and the rate of interest. The level of income and activity is
related to the amount of investment by the multiplier, that is by the
marginal propensity to consume, the price level is related to the level
of activity by the marginal money cost of production (which depends
on the amount of activity undertaken), the amount of money absorbed in active trade depends on the volume of trade and the price
level, the amount of money available for liquid reserves is equal to
the total amount of money in the system less that required for active
expresses as the demand for purposes other than those of active circulation. This re-definition of terms is merely an expository device and
does not imply any departure from Mr. Keynes' essential doctrines.

136 R. F. Harrod
trade, and the rate of interest depends on the amount of money available for liquid reserves and the liquidity preference schedule.
It may be well to do some exercises. Suppose the banks to increase
the total amount of money available by open market operations. The
increment may eventually be divided between active circulation and
liquid reserves. An increase of money available for liquid reserves
will tend to reduce the rate of interest; and so to increase investment.
This will increase the level of income through the multiplier in accordance with the marginal propensity to consume. If the fall in the
rate of interest increases the marginal propensity to consume, the increase of income will be pro tanto greater, but it is not certain that it
does so. The increase of incomes involves an increase of turnover,
and of prices in accordance with the law of diminishing returns. This
involves an increased use of money in active circulation. Thus the
fall in the rate of interest will not be so great as it would be if all the
new money went into liquid reserves. The money will be divided between the two uses, but there is no reason whatever to suppose that
the increments in each use will be in proportion to the amounts of
money previously employed there, as is assumed in a Quantity
Theory using a compendious index number. The comparative size of
the increments will depend on the current elasticity of the liquidity
preference schedule and the current elasticity of the marginal productivity of capital schedule (which involves expectations).
Suppose a fall in rewards to prime factors. The price level will
drop. Money will be released from active circulation for liquid reserves. This will tend to make the rate of interest fall and to react on
the level of investment and activity accordingly. Thus the stimulus
to activity is very indirect and its effectiveness depends on the same
factors as that provided by an increase in the quantity of money. This
is very different from the view that a reduction of rewards will stimulate activity because costs fall while prices are sustained by the
quantity of money remaining the same.
It appears to me that the achievement of Mr. Keynes has been to
consider certain features of traditional theory which were unsatisfactory, because the problems involved tended to be slurred over, and
to reconstruct that theory in a way which resolves the problems. The
principal features so considered are ( I) the assumption that the level
of income could be taken as fixed in the departmental theory of inter-

Mr. Keynes and Traditional Theory 137


est and saving, (2) the peculiar nature of the supply schedules of the
prime factors which arises out of their bargains being fixed in money
without proviso as to the price level, and (3) the failure of monetary
theory to explain how the total stock of money is divided between
liquid reserves and active circulation, or, in other words, the unsatisfactory character of the theory of velocity of circulation.
I stated above that the old pieces in the traditional theory reappear,
but sometimes in new places. It might at first be thought that the
liquidity preference schedule is a new piece, and that therefore either
the new system is overdetermined or the traditional writers must have
been wrong in supposing that their system was determined. But it is
not really a new piece. The old theory presupposed that income
velocity of circulation was somehow determined. But precisely how
was something of a mystery. Thus the old theory assumed that there
was a piece there but did not state exactly what it was. Mr. Keynes'
innovation may thus be regarded as a precise definition of the old
piece.
By placing it where he does, he overcomes a difficulty, which has
been assuming an alarming prominence in recent economic work. In
monetary literature the rate of interest has been treated, and increasingly so, as an influence of vital importance in the monetary situation. But in traditional theory, neither in the general system of
equations nor in the departmental theory of interest does it appear
that the rate of interest is more intimately connected with the
numeraire than the price of any other factor of production. This is a
striking discrepancy. Mr. Keynes introduces the liquidity preference
schedule at a point which makes it a vital link between the general
system of equations and monetary theory. His treatment is in harmony with recent literature in that he justifies the special connection
of the price of this particular factor with monetary problems. It is an
immense advance on recent literature because it removes the discrepancy between the treatment of interest in the two branches of
study.
In my judgment Mr. Keynes has not effected a revolution in fundamental economic theory but a readjustment and a shift of emphasis.
Yet to effect a readjustment in a system, which in its broad outlines,
despite differences of terminology, has received the approval of many
powerful minds, Marshall, Edgeworth, and Pigou, the Austrian

138

R. F. Harrod

School, the School of Lausanne, Wicksell, Pantaleoni, Taussig, and


Clark, to mention but a few, is itself a notable and distinguished
achievement. And in the sphere of departmental economics and short
cuts, which are of great concern for the ordinary working economist,
Mr. Keynes' views constitute a genuine revolution in many fields.
The foregoing account has attempted to expound, not to appraise.
The only criticism of Mr. Keynes which I venture to offer is that his
system is still static. Note has been taken of the fact that at certain
important points, e.g., in his definition of the marginal efficiency of
capital, Mr. Keynes lays great stress on the importance of anticipations in determining the present equilibrium.
But reference to anticipation is not enough to make a theory dynamic. For it is still a static equilibrium which the anticipations along
with other circumstances serve to determine; we are still seeking to
ascertain what amounts of the various commodites and factors of
production will be exchanged or used and what prices will obtain, so
long as the conditions, including anticipations, remain the same. But
in the dynamic theory, as I envisage it, one of the determinands will
be the rate of growth of these amounts. Our question will then be,
what rate of growth can continue to obtain, so long as the various
surrounding circumstances, including the propensity to save, remain
the same?
Saving essentially entails growth, at least in some of the magnitudes under consideration. No theory regarding the equilibrium
amount of saving can be valid, which assumes that within the period
in which equilibrium is established, other things, such as the level of
income, do not grow but remain constant.
I envisage in the future two departments of economic principles.
The first, the static theory, will be elaborated on the assumption that
there is no growth and no saving. The assumption that people spend
the whole of their income will be rigidly maintained. On this basis it
will be possible to evaluate the equilibrium set of prices and quantities of the various commodities and factors, excluding saving. In the
second department, dynamic theory, growth and saving will be taken
into account. Equilibrium theory will be concerned not merely with
what size, but also with what rate of growth of certain magnitudes is
consistent with the surrounding circumstances. There appears to be
no reason why the dynamic principles should not come to be as pre-

Retrospect on Keynes . 139


cisely defined and as rigidly demonstrable as the static principles. The
distinguishing feature of the dynamic theory will not be that it takes
anticipations into account, for those may affect the static equilibrium
also, but that it will embody new terms in its fundamental equations,
rate of growth, acceleration, deceleration, etc. If development proceeds on these lines there will be a close parallel between the statics
and dynamics of economics and mechanics.
But to develop this theme further would take me too far from my
subject.

Retrospect on Keynes
[1963]
KEYNES IS LIKELY TO HAVE a permanent place in the
history of economic thought as being the first person to develop a
fully articulated theory of what we now call macroeconomics. Such
theory as is to be found in classical economics is largely implicit, and
depends on the proposition that capitalists will always strive to keep
their capitals fully employed; the full employment of labor was,
through the wages fund theory, a corollary of this. One could,
alternatively, at later dates, search for an employment theory in the
proposition that there will be a tendency to equate the marginal
disutility of work with the marginal utility of its product. But it was
not then shown how this tendency worked in practice in relation to
the institutional arrangements of actual economies.
Keynes' finding that free enterprise economies did not in all circumstances tend towards an equilibrium position of full employment
is well known. And it appears to have been generally accepted; witness the widespread recognition that deliberate monetary and fiscal
policies are needed, to supplement the automatic workings of a free
enterprise system.
Since the war attention has been increasingly focused upon the

140 . R. F. Harrod
need to develop a dynamic theory. I call attention to the concluding
paragraphs of my address on Keynes to the Econometric Society in
the summer of 1936. "The only criticism of Keynes which I venture
to offer is that his system is still static." I proceeded to give an outline
of how, in my view, a dynamic theory should be developed. It is to
be noted that for the elaboration of any such dynamic theory, a
macroeconomic theory of statics was an indispensable foundation.
The traditional micro theory did not provide the necessary tools.
Thus Keynes may be truly regarded as the father of dynamic
theory. And that, in the long run, will prove to have been his greatest contribution of all.
Mention should also be made of the great impetus that he gave to
the compilation of national economic statistics, both by his encouragement of independent research in this held before the second
World War, and also by his getting this work put in hand in the
British Central Statistical Office during that war. The British still
appear to hold the hrst place in the compilation of national income
statistics; but it has now become common form in all countries that
attempts should be made, even when the basic data are very imperfect, to furnish such statistics. Their provision has proved to be of
the utmost value to economic researchers.

If one is asked for an opinion about how far the specihc doctrines
of Keynes have held their ground, or how far the corpus of economic
doctrine, as generally accepted and taught, is "Keynesian," it is
difficult to answer. It is possible to dwell on only a limited number
of topics here. And it seems expedient to stress certain aspects, the
neglect of which has, in my view, been disappointing.
Keynes has suffered from his own fertility and from his anxiety to
present his latest thoughts to the public as quickly as possible. In
one way that was a virtue; and in the case of the General Theory it
was most fortunate, since almost all the rest of his life was occupied,
hrst by serious illness, and then by his work in the British Treasury
on behalf of the war effort and on postwar planning.
Since his death, students of Keynes have tended to focus their
attention exclusively on the General Theory of Employment, Interest, and Money, to the neglect of his Treatise on Money, which was

Retrospect on Keynes .

141

published a few years earlier. This earlier volume is a much richer


store of Keynesian thinking than the later volume, which is now so
much better known.
The General Theory made important advances, notably the integration of monetary theory with a macroeconomic income theory.
Keynes was able to incorporate the well-known doctrine of the
"multiplier," which he derived from Professor R. F. Kahn. One
would certainly not wish to belittle the General Theory; but many of
the doctrines of the Treatise have passed from view, and much has
been lost thereby. I suspect that, had Keynes had a further span of
good health and free time at his disposal, he would have wished to
write a still more comprehensive work, bringing back into view some
doctrines of the Treatise, which are crowded out of the General
Theory.
It is well known that in the General Theory Keynes insists
strongly on the bookkeeping identity of saving and investment. It has
probably been a good thing that this usage has become well established in economic theory, since it is indispensable for national
income statistics, and it is expedient to have terms used in the same
senses in unofficial writing and in official publications. But according
to the Treatise saving and investment are not necessarily Cor usually)
equal. Keynes was of course fully aware, when he wrote the Treatise,
of the bookkeeping identity; indeed he referred to it several times in
that work. The need is still felt, however, in economic writing for
concepts of saving and investment such that they are not necessarily
equal; and writers fulfill this need by prefixing them with such words
as "ex ante" or "planned." Ex ante or planned investment or saving
are not, however, quite the same concepts as the investment and
saving of the Treatise.
We may let St stand for saving in the sense in which it is used in
the Treatise. St$T, because saving out of excess corporate profits,
which of course must be counted in S when S=T, was excluded by
Keynes from the totality of saving. On the other side, when aggregate
profit is below normal and less is saved from corporate profit than would
have been, if profit had been normal, this shortfall, by comparison,
was reckoned as a "loss," but not deducted from total national "saving" (S,), so that it was possible for St to be greater than T. To put
it in other words, companies were credited with having done the

142 . R. F. Harrod
amount of saving that they would have done, had their profits been
normal, and not with the saving that they actually did, so that the
total saving (St), as reckoned by Keynes, was greater or less than the
bookkeeping saving (S), which was the saving that actually occurred
in the period in question, and had to be equal to investment. I believe
that this idea of the possibility of investment exceeding saving, thus
leading to inflation, or saving exceeding investment, thus leading to
depression, was a generally intelligible one.
To complete the definition, it is plainly necessary to have a definition of "normal profit," for it is upon this that the Keynesian St depends. His definition of normal profit is "that rate of remuneration of
entrepreneurs, which, if they were open to make new bargains with
all the factors of production at the current prevailing rates of earnings, would leave them under no motive either to increase or to
decrease their scale of operations." (It is not necessary, of course, for
equilibrium that all entrepreneurs should be in this position, but
only that those with excess profits relatively to the norm should
balance those with deficient profits).
The above definition of normal profit is static, but it lends itself
easily to a dynamization. Delete the last fifteen words and substitute: "give them a motive for increasing their scale of operations by
no more nor less than the increase in the requirement for their
product that is to be expected when the economy as a whole is growing at its optimum rate".
By using this definition Keynes intends that short-period global
corporate profit should be regarded as a residual in the working
of the system as a whole. This brings macro theory into complete
harmony with micro theory, in which the profit of the particular
firm is a residual in the short period. If global corporate profit is a
residual, then global corporate saving also becomes a residual. While
we may think of this as merely the logical consequence of the definitions, it is really something much more important than that.
Keynes wanted us to think of corporate saving as in fact a residual,
in relation to the real causes operating in the economic process. And
it was precisely because it brought this important causal relation
under the spotlight that Keynes' special definition was valuable.
(I would note in passing the great difference between Keynes'
definition, by which it is not necessary that St=I, and those other
definitions that also yield the possibility of inequalities of the general

Retrospect on Keynes' 143


type: "saving=last year's income minus this year's expenditure."
Definitions of that type may well be useful in certain contexts; I have
never been able to use them fruitfully myself).
It is no doubt difficult to establish a rate of normal profit and
normal company saving statistically. They doubtless change from
time to time, like other fundamental economic determinants. But
by analyzing time series, it should be possible to distinguish a trend
of normal corporate saving from oscillations. Of course we must
not rule out the possibility of a sudden once-over change in the
norm. If we can establish that corporate saving is normally x percent of national income and can be relied upon to cover investment
of a corresponding amount, then the spotlight turns on to the relation between the remaining investment requirements and personal
saving, whether or not the latter is supplemented by government
saving. If total investment requirements, minus that part covered by
normal corporate saving, begins to grow more than personal saving,
then inflationary pressure will arise; and in the opposite case slackness and depression.
In the former case corporate saving will automatically swell up,
as excellently explained in the Treatise, to fill the gap, so that the
bookkeeping identity of S = I is preserved. In the latter case corporate
saving will be depressed below normal. When stress is laid on the
bookkeeping identity S = I, and all saving including corporate saving is lumped together, this vital relation between investment requirements and personal saving is blurred. The most important
doctrine of the Treatise is that it is the relation between investment
requirements and personal saving that determines the tendency to
expansion or contraction, and that the enlargement or contraction of
corporate saving comes in as the residual consequence of the whole
economic process, to balance things out and restore the bookkeeping
identity of S = I. This essential doctrine is lost sight of in the General
Theory. But I do not believe that Keynes intended that. He had
much that was new to say in the General Theory, and one has to
do one thing at a time. As Keynes never wrote that comprehensive
volume postulated above, which might have brought back some
elements of the Treatise into prominence, I believe that it is just
as important for anyone who wants to understand Keynes thoroughly
to read the Treatise as it is to read the General Theory.
In the United Kingdom, personal savings have recently been rising

144 . R.

F. Harrod

strongly relative to investment. The consequence has been a squeeze


on profit. Here we have a striking confirmation of Keynes' theory.
But commentators have not taken up this point. Instead they
talk of profit being squeezed by higher wages. If one asks them
why prices have not recently been rising in proportion to wages, as
they did at a somewhat earlier period, they may refer to a stricter
hold on the money supply in the later period. This latter can doubtless have its effect, by good Keynesian doctrine, through causing
the maintenance of interest rates above the right level and difficulties
of borrowing, and can be of primary importance in relation to the
growth rate achieved and the volume of investment. But as regards
the squeeze on profit (which also probably has an adverse effect
on growth), this can confidently be attributed to the upsurge of
personal saving relatively to investment.
Another valuable element in the Treatise is the sharp conceptual
distinction between two possible causes of a rise (or fall) in prices.
The two causes correspond to what we have now come to call "demand pull" and "cost push," although Keynes did not use those
expressions. The distinction is also present in the General Theory
but less sharply delineated. In the General Theory, when causes
are operating, which we should now classify as of the "demand pull"
variety, their effect is seen in a rise of prices expressed in "wage
units," whereas the other type of cause operates when the rise in
prices is due to a change in the value of the wage unit itself.
In the Treatise in all the equations set out on pages 135-8 (vol.
I), which constitute the kernel of the whole theoretical structure,
there are two separate items on the righthand side which together
determine the price level, as expressed on the lefthand side. One
of the two causes is any change in the level of earnings divided by
the level of output; these earnings do not include excess profit. The
other cause is any excess of investment over savings (in the Treatise
sense). What corresponds to the latter type of cause in the General
Theory is the rise of aggregate effective demand above the critical
level; what corresponds to the former type of cause is a change in
the value of the "wage unit." It is to be noted that these two types
of change capable of causing a rise of prices are causally, as well as
conceptually, distinct. This strikes me as a valuable element in the

Retrospect on Keynes . 145


Treatise formulations. Naturally Keynes has much to say about
interactions between the righthand terms, as when a rise of prices
due to excess investment sets up a demand for higher factor rewards, which may raise prices still further. Keynes laid great stress
on the point that no rise in factor rewards can ever catch up on a
given rise of prices, so long as the separate and independent cause
of rising prices is operating, viz. an excess of investment over saving.
A rise of prices may occur owing to eXCeSS factor rewards, whether
there is an excess of investment over saving or not. When there is an
excess of investment over saving (in the Treatise sense) the accountancy identity of total income and higher priced output is preserved
by the occurrence of additional corporate income of an amount
equal to the excess of investment over saving. I believe that this
conceptual set-up retains its value.
There is another passage specifically devoted to this topic (pages
166-70) where "changes in the rates of efficiency earnings and in
the price levels due to the character of the wages system (including
in this e.g., the powers and activities of trade unions)" are called
"spontaneous," while those "arising from the existence of profits or
losses due to the currency authority permitting or promoting a
disparity between investment and saving" are called "induced." The
rest of this passage suggests strongly that, to ensure overall stability,
we require not only currency management, but also what we have
come to call an 'incomes policy."
In a further passage (p. 208) he makes the same distinction, in relation to the downward movement of prices, by using the expressions
"Income Deflation" and "Profit Deflation."
Very important and central in the General Theory is Keynes'
interest theory. The essentials of this are present also in the Treatise.
I find it very disappointing how little acceptance this theory appears
to have won, especially among those who discuss contemporary
monetary problems.
It has to be admitted that Keynes overplayed his hand. I shall
note three respects.
(I) He stated emphatically that interest was not a reward for
saving, but for parting with liquidity. It cannot be denied that some-

146 . R. F. Harrod
one who does saving, but is unwilling to part with the liquidity
which thereby accrues to him, can earn no interest. But if there are
marginal savers, who just would not save, or would not save so much,
if no interest could be earned, then for these interest is the reward
for saving. They may have no concern about liquidity, and if they
save, no question arises for them whether they should part with
liquidity or not. Their motive for saving may be solely to get the
interest that they obtain thereby, and for them that is the end of
the matter. It is not inconsistent with Keynes' contention that the
rate of interest is always at such a level as to balance the marginal
inconvenience of parting with liquidity, to hold also that one function of interest is to be a motive for saving.
(2) Keynes puts his spotlight on the relations between bonds and
cash; since lack of liquidity is the sole characteristic differentiating
bonds from cash, the payment of interest on bonds must be the compensation for loss of liquidity. But Keynes does not in the General
Theory enlarge his horizon and discuss the mutual relations between
cash, the yield of bonds and the yield of other assets; had he done so,
he might have made his theory more readily acceptable.
(3) The concept of a "natural" rate of interest, which has a kinship with the concept so named by Wicksell, is prominent in the
Treatise, but disappears from the General Theory. This seems to
be a definite loss. It left Keynes exposed to the (independent) criticisms of D. H. Robertson and J. R. Hicks, which may be summarized
in the words used by the latter that he ''left the rate of interest hanging by its own boot straps."
I believe that the reason why Keynes discarded this expression
was that he wanted to avoid a connotation of the word "natural" as
used by economists from Adam Smith onwards, namely, an equilibrium price that the ordinary workings of the economic system tend
to establish in the long run. He wanted to emphasize his view that
there was no such equilibrium towards which the system normally
gravitated, and he feared to arouse misunderstanding, if he used the
word "natural."
Nonetheless, it should be quite obvious to the discriminating
and careful reader that there is in fact a natural rate of interest implicit in the doctrines of the General Theory. In the Treatise the
natural rate is that which keeps investment equal to saving (in

Retrospect on Keynes 147


the Treatise sense). In the General Theory there is clearly a right
rate of interest, namely that which keeps aggregate demand at the
level required to give full employment. If the rate of interest is too
low, there will be an excessive aggregate effective demand, leading
to inHationary pressure, and, if the rate of interest is too high, there
will be insufficient aggregate effective demand. Therefore, there is
a specific rate of interest, which is the "right" one from the point of
view of full employment, and one would have thought it would
have been appropriate to call that right rate the "natural" rate, despite
its possibly misleading historical associations. And be it noted that
the forces which determine what this right rate is are the propensity
to invest and the propensity to save. Thus the doctrine of the "right"
rate, as I have described it, is entirely conformable with traditional
theory, and I believe that, if Keynes had made this concession to
traditional terminology, he would have found much greater acceptance for his views. (I may say in passing that I believe that, when
we come to dynamic theory, we have to discard the notion of a right
rate as being determined by the propensity to invest and the propensity to save.)
What Keynes' doctrine, as thus interpreted, amounts to is that
(I) the actual rate is determined by the demand for and supply of
liquidity, (2) there is a "right" rate which is governed by the
propensity to invest and the propensity to save, and (3) the authorities should so regulate the supply of liquidity as to ensure that
the actual rate is equal to the right rate, it being taken for granted
that the actual rate will never find its way on its own to the "right"
level, unless the authorities do some doctoring. All this seems to
me to fall naturally into line with the traditional theory, although
Keynes' whole treatment of liquidity constitutes a vast advance
upon traditional theory, which, relying on ambiguous concepts of
velocities of circulation of money, was rather weak in all this field.
By refusing to concede enough to tradition and going too far in
certain respects (e.g., (I) above), Keynes has caused what was new
and true and important in his own theory to suffer undue neglect.
I next come to my troubles.
First, those very eminent economists Alfred Marshall, Irving
Fisher, and Gustav Cassel alike hold that the expectation of inflation

I48 . R. F. Harrod
is a cause of high interest rates. This theory does not occur in Keynes
and is inconsistent with his theory of interest. I conceive this to be
a great advance on the part of Keynes. But his view does not appear
to have gained wide acceptance. Recently the rate of interest has
been standing at a much higher level in Britain than before the
war, and this may readily be explained by the fact that the supply
of liquidity has not risen since then nearly so much as the demand
for it. When one points this out, and adds perhaps that one thinks
so high a rate of interest is bad for investment and growth, one is
very often met with a flat rejection. "Oh no, the higher rates of
interest prevailing now are simply due to fears of inflation; nothing
that the authorities could do could alter that."
The idea is that if the natural rate of interest is 3 per cent, and
if a man lends 100 for a year, expecting the general price level to
rise by 2 per cent in that time, he will require 5 per cent, the extra
2 per cent being to compensate him for the loss in the real value
of his principal. But if, following Keynes, we hold that the rate of
interest on bonds is a compensation for their inferior liquidity compared with cash, and if we hold strictly to the rule that the causes
of interest on bonds are to be found solely by comparing bonds with
cash, then the argument will not hold. The lender of the 100
would have the alternative of holding cash; but, if he held cash,
that also would lose 2 per cent in real value during the year. In this
respect there is no difference between bonds and cash, so that the
man who holds a bond instead of cash cannot require this extra 2
per cent. By holding cash, one does not hedge against inflation,
any more than one does by holding bonds. In this respect there is
no difference between them, and therefore the question of whether
inflation or deflation is expected does not enter into their relative
valuation; therefore it has no effect upon the rate of interest.
What is affected by the fear of inflation is the relative valuation
of bonds compared with such other assets-equities, real estate etc.
-as do provide a hedge against inflation. The events in recent years
fully confirm this. The drop in the yield of equities relatively to
the yield of bonds has been caused, in part at least, by the expectation of inflation. That is by no means inconsistent with Keynesian
doctrine.
In my opinion, the lingering hold of the old-fashioned doctrine of

Retrospect on Keynes . 149


Marshall, etc., on the minds of men and the failure to recognize
that it has been demolished by Keynes have been doing great harm.
I am confident that the low growth rate in Britain has been due to
the authorities holding actual interest rates above their natural level.
They may genuinely believe the Marshallian doctrine, according
to which there is nothing they can do to bring interest rates down
-except, of course, to stop inHation-on the ground that public opinion would, owing to its fear of inHation, merely push them up again.
It is therefore most important that this idea should be finally eradicated.
I should suppose that in so highly developed a country as the
United States, very low interest rates would be required to sustain
growth at the potential level of the economy, so that in that country
also the false idea that fears of inHation would prevent interest
rates falling, if more liquidity were provided, can do much harm.
Second, my disappointment in regard to the lack of acceptance of
Keynes' theory of interest covers wider ground. It was said above
that by his theory the actual rates are governed by the demand and
supply of liquidity. While there are doubtless many factors governing demand, including what Keynes called the "speculative motives"
which became very powerful in the Dalton period of 1946-7, the
main determinant of the demand for liquidity is the money value
of the national income. Now since before the war the supply of
liquidity (probably best indicated by the level of "current accounts"
or "demand deposits") has been increased much less than the money
value of the national income. This suffices to explain the high rates
of interest that have been obtaining. The shortfall in the supply of
liquidity has grown much more marked in recent years, and the rise
in interest rates has been correspondingly more marked. Indeed, if
we take the whole course of events since before the war, we find that
Keynes' theory has been confirmed in the most striking way in
Britain, with very few important deviations. During the period of
very low interest rates shortly after the war, the supply of liquidity
was kept very abundant. Then it was reduced relative to national
income by stages, sometimes more quickly and sometimes more
slowly, and interest rates followed a similar pattern.
Opponents of measures for reducing interest rates rely, not only
on the argument about inHation aforementioned, but also on the

150

R. F. Harrod

contention that high interest rates are due to high investment demand. But there is no truth in the view that investment demand is
higher relative to personal saving either in the United States or the
United Kingdom than it was before the war. The facts are exactly
opposite; personal saving has risen relatively to investment. It is at
this point that the distinction between personal saving and corporate
saving becomes so important. If we take total saving, then no argument can arise about the rate of interest in relation to the balance
between saving and investment, since saving is always equal to investment; but, if we concentrate upon personal saving, treating
corporate saving as a residual, then the natural rate of interest should
be lower now in both countries than it was before the war. Actually
rates are higher; but that is solely because insufficient liquidity has
been provided.
.
What the authorities argue is that it would be useless to try to
get interest rates down by providing more liquidity, since the high
rates are due to high investment demand. This attitude is objectionable in two respects, namely (I) because investment demand has
not been not high in relation to personal saving-and it is personal
saving that matters in this context-and (2) because it violates
Keynes' view that interest rates can be brought down by providing
more liquidity whatever investment demand may be.
It is to be noted that personal saving has risen especially strongly
in Britain during the past few years, relatively to investment, and it
is precisely in this period that interest rates have risen most strongly.
But then in this period the reduction in liquidity became especially
intense.
Now it may well be that if Britain and the United States established much higher growth rates, with much larger requirements
for investment to match, and if both provided much more capital
for developing countries, personal savings might prove inadequate
and interest rates above the prewar level might be needed. We just
do not know if this is so or not. But, as things are, we are in a vicious
circle in which the high interest rates are impeding that rate of
growth and investment which alone might make it expedient to
have them. I find the current great neglect of the Keynesian theory
of interest and liquidity depressing.

Retrospect on Keynes r 5 r
There is the problem of internal liquidity and there is the problem
of international liquidity. If there is a shortage of the media of international reserve and settlement, this will tend to lever up those
interest rates to which international capital movements are sensitive,
by mutual competition between countries. This international competition in respect of high interest rates might be a valid reason
for one particular country to hold its own rates up, regardless of
whether such high rates correspond to the "natural rate" as required
to maintain employment and growth in the internal economy. Keynes
reverts again and again to this theme in the Treatise. I have no doubt
that he would hold that, if nothing can be done about the shortage
of international liquidity, it would be better to impose controls over
the international movement of capital, rather than acquiesce in
having domestic interest rates higher than required locally for full
employment and growth. But this view does not appear to be widely
accepted among monetary authorities, and in this respect again I fear
that Keynesian doctrine has not bitten deep.
But cannot something be done about international liquidity shortage? In this respect we need have no doubt what Keynes' views
would be. He would politely, and sardonically, recommend the adoption of his "Clearing Union," as put forward in the Anglo-American
discussions prior to Bretton Woods. And it may be noted that his
proposals on these lines date back much earlier, e.g., to the time of
the World Economic Conference in London (1933) when he recommended the issue of international gold notes. He would doubtless
add that even his Clearing Union scheme as conceived in 1941-2might not be on a sufficiently generous scale to meet modern requirements. He was passionately in favor of an adequate supply of
liquidity, especially if the aim was to have fully multilateral world
trade with the minimum of restrictions. He would certainly not
suppose that the problem could be solved by the minor expedients
recently proposed by Mr. Jacobson and Mr. Roosa.
May I finally put a delicate point. I find myself in a small
minority of my professional colleagues in advocating faute de mieux
--and the mieux is certainly not likely to come to pass-a rise in
the dollar price of gold, and simultaneously increases in the prices
of gold expressed in other currencies also. This proposition is especially repugnant to avant garde economists, I can never understand

152

R. F. Harrod

why. In my own thought gold is the greatest sheet anchor of liberty,


which might otherwise dwindle away under growing regimentation
by an international bureaucracy.
Keynes usually found himself in sympathy with the avant gardists,
although he never allowed his own opinions to be shackled by passing
fashions, e.g., in his sustained view that high profit was a healthy
thing.
It would accordingly be paradoxical in the extreme to suggest that
Keynes would favor a rise in the price of gold, especially as he attacked the old form of gold standard very vehemently and destroyed
many shibboleths connected with it. Despite all this, I believe that
his sense of the importance of an ample supply of liquidity was so
great that, if convinced that anything as radical as his Clearing
Union plan was impracticable, he too would favor a rise in the dollar
price of gold.

D. G. CHAMPERNOWNE

Unemployment, Basic and


Monetary: The Classical
Analysis and the Keynesian
[1936J
I

HE FUNDAMENTAL DIFFERENCE between the General


Theory of Employment put forward by Mr. Keynes and that put
forward by the Classical economists, typified by Professor Pigou, is
that whereas the latter assume that the wage bargains between the
entrepreneurs and the workers determine the real wage, the former
argues that these bargains succeed only in determining the money
wage.
Many people will have found that whereas Mr. Keynes' arguments
in favor of this thesis are convincing when they refer to the effects
of bargaining between entrepreneurs and employees over a short
period of, say, one year, yet these arguments are less convincing
when they refer to the effect of this bargaining over a long stretch
of years. If labor is fundamentally discontented with its real wage
year after year and is unsuccessful in raising it, some violent form
of disequilibrium must arise which will eventually cause unemployment and induce labor to accept its real wage; if labor is so disorganized by unemployment that the competition of the unemployed
continually lowers money wages, a situation must eventually arise
in which the monetary authority takes action to check any resultant
fall in prices, and so makes effective the attempt of the unemployed
to accept a lower real wage.
If this view were correct, it would follow that over a long series
of years the chief determinants of the trend level of real wages and
the trend percentage of unemployment were the bargains between

153

154 D. G. Champernowne
the entrepreneurs and the employees. In this case, we could apply the
whole of the classical analysis to the study of trends of real wages,
unemployment percentage, prices, investment, the rate of interest, etc.
H this view were only partly correct, and it were true only that
wage bargains had some influence in determining the trend level of
real wages and the trend percentage of unemployment, yet a study
based on the classical assumptions might throw some light on the
questions of what will be the trends of real wages, unemployment
percentage, prices, etc. For an application of the classical apparatus
is simpler than repeated application of Keynesian theory to successive
short intervals of time when account has to be taken of the working
out of the demand by labor for a change of real wages, into an
equation between industry's supply and labor's demand for real
wages.
Accordingly, we may proceed to a rationalization of the classical
apparatus for determining the trends of real wages, unemployment
percentage, prices, etc., in view of the criticisms advanced in The
Theory of Unemployment.
2

It was pointed out by Professor Pigou in The Theory of Unemployment that, sometimes, labor is concerned not only with the real
wage which it will receive, but also with the money wage. For instance, although a man would be unwilling to accept a wage rate
of 35S. a week, yet even if the cost of living had risen by 20 per
cent, so that a wage rate of 42S. would represent the same real wage
rate as before, he would be prepared now to accept a money wage
rate of 40S. a week, just because this represents an advance of 5S. a
week on the original offer.
Mr. Keynes argues that this is the general case. He suggests that
labor is always more awake to changes in money wage rates than to
changes in the cost of living, so that labor's policy with regard to the
level of real wages can always be neutralized by changes in the cost
of living. This suggestion constitutes his first wave of attack on the
classical system of analysis.
This argument is entirely convincing when it is confined to immediate effects, and our only escape from the conclusions to which

Unemployment, Basic and Monetary 155


it leads when applied to more distant effects is by way of an appeal
that labor's concern with the money wage rather than the real wage
is a temporary phenomenon.1 This appeal is fabriCated below.
3

It would be ridiculous to suppose that wage earners are really


more interested in their money wage than in their real wage; conversation with a representative wage eamer has convinced me of this,
and I assume that my reader is also convinced of this.
The failure of labor to respond immediately to a rise in the cost
of living by a universal demand of an equal (proportional) rise in
money wages is due perhaps to the following causes:
(i) The existence of a definite contract based on the expectation
of a stationary cost of living;
(ii) Disinclination to incur the inconvenience involved in frequent demands for changes in the wage level;
(iii) Failure to notice immediately the fact that the real wage rate
has fallen;
(iv) The habit of thinking in terms of the price level of some
earlier date.
All these causes will be removed in time. The wage contract will
come up for revision; it will soon be an appreciable time since the
last demand for wage change; eventually the housewife2 and the
agitator will bring it home to the wage earner that his wage now
buys less than before and that money today is not the same thing
as money was six years ago.
When all these causes of delay have been removed, the demand
of the laborer for a certain real wage will become effective at least
to the extent of raising the money wage rate in proportion to the
original increase in the cost of living.
For example, if in 1939 a wage earner is content with 405. a week,
because he has overlooked a 20 per cent rise in the cost of living
during 1938--9, although he would not have accepted 395. a week at
1938 prices, and if during 1939-40 the cost of living does not change,
This was pointed out to me by Professor Pigou and Mr. D. H.
Robertson.
2 This scene was suggested by Mr. D. H. Robertson.
1

156 . D. C. Champernowne
but the wage-earner slowly realizes that owing to the cost of living
having risen 20 per cent during 1938-9 his wage falls short by 20
per cent of the real wage which he had been accustomed to demand,
he is likely to demand a 20 per cent increase in money wage rate.
If this analysis is correct, it follows that if R is the real wage that
labor will eventually try to get when the level of employment is N,
then a state of affairs where employment is N and real wages are less
than R cannot long exist without a rise in money wages taking place.
A similar argument would show that a state of affairs where employment was N and real wages were greater than R would not be
likely long to exist without causing a fall in money wages.
If we could then assume that a rise or fall in money wages must
lead to a rise or fall in real wages, we could assume that real wages
were determined by the supply and demand of labor, through the
bargains between entrepreneurs and employees.
4

Before considering in what sense it is still possible to believe that


a rise or fall in money wages is likely to result eventually in a similar
rise or fall in real wages, we may revise what has been suggested
already, by describing it in terms of different concepts.
According to our theory, the money wage which a laborer demands
today is that money wage which would have given him a certain
definite standard of life (i.e., which would have been a certain definite real wage) at prices ruling at some date in the past.
If the cost of living had been constant for a long time the laborer
would demand that real wage which would ensure him that standard
of life at present prices. If the cost of living has been recently rising,
he will fail to take this into account and will demand a lower real
wage than this: if the cost of living has been recently falling, he will
demand a higher real wage than this. For convenience, we may refer
to the real wage which he would demand if the cost of living had
been stationary as the basic real wage. The basic real wage will obviously be different for different individuals.
Since the cost of living is not in fact always stationary, but rises
and falls, there may at various times be men either in employment,
receiving less than their basic wage, or men out of employment, although they would be able to obtain employment at their basic
real wage.

Unemployment, Basic and Monetary' 157


In a corresponding manner we may state that in any given situation the amount of basic unemployment is the amount of unemployment that there would be in that situation if each man demanded
neither more nor less than his basic real wage.
Then if the cost of living has been rising, we should expect actual
unemployment to be less than basic unemployment, whereas if the
cost of living had been falling we should expect basic unemployment
to be less than actual unemployment. We may call any deficiency
of actual unemployment below basic unemployment "monetary employment," and we may call any excess of actual unemployment over
basic unemployment "monetary unemployment."s On these definitions:
Actual Unemployment=Basic Unemployment-Monetary
Employment
or
Actual Unemployment=Basic Unemployment+Monetary
Unemployment.
Then the unemployment of the monetary-unemployed is due to
the fact that they or their unions have overlooked a recent fall in
the cost of living, whereas the employment of monetary-employed
is due to the fact that they have overlooked a rise in the cost of living.
Insofar as these oversights are likely to be repaired eventually, the
monetary-unemployed are likely to lower the money wage which
they demand, and the monetary-employed are likely to raise the
money-wage-rate which they demand. We may express this by saying that a period of monetary unemployment is likely to cause falling
money wages and that a period of monetary employment is likely to
cause rising money wages. Insofar as we can assume that rising and
falling money wages will respectively cause rising and falling real
wages, we may conclude that a period of monetary employment contains the seeds of its own destruction in the form of a tendency for
real wages to rise, whereas a period of monetary unemployment has
in it the seeds of its own destruction, in the shape of a tendency for
real wages to fall.

The concept of Monetary Unemployment is copied from Keynes'


"Involuntary Unemployment" (General Theory, p. 15), but differs
from that concept.

158

D. G. Champernowne

5
We must now return to the consideration whether it is still possible to believe that in any sense a rise in money wages is likely to
lead to a rise in real wages. For the second wave of Keynes' attack
consists of a convincing demonstration that there is no reason to
expect that a rise in money-wage-rates is more likely to lead to a rise
than to a fall in real wage rates, unless the monetary authority
takes the necessary steps (e.g., raises the rate of interest) to increase
unemployment. He shows convincingly also that a lowering of
money-wage-rates is no more likely to lower than to raise the real
wage rate, unless the monetary authority takes the necessary steps
(e.g., lowers the rate of interest) to decrease unemployment.
There is no obvious Haw in this argument, and it follows that the
demand of labor for a certain real wage can only make itself effective insofar as it influences the attitude of the monetary authority
and its manipulation of the rate of interest.
Consider a position in which there is considerable monetary employment. After a time, money wages must start to rise, and unless
real wages also rise, in which case monetary employment will be
reduced, the cost of living will rise in as great proportion. This must
lead, after a further interval, to another rise in money wages: this
will be accompanied by another rise in prices and followed by yet
another rise in money wages. Under such conditions the bargaining
power of the laborer will become stronger and stronger as he becomes
more and more confident of his ability to raise his money-wage-rate,
and the speed with which he will revise his demands in the face
of increases in the cost of living will become greater the more accustomed he becomes to the danger of his real wage being reduced
by the rise in the cost of living.
We see that a period of monetary employment will be accompanied
not merely by rising money wages and prices, but moreover by money
wages and prices rising at a rapidly increasing rate.
It is difficult to believe that such a process would continue for long
without evoking a very violent protest from those classes of the
community who stood to lose from a rapidly rising price level; considerable difficulty would be experienced in adjusting the rates of
exchange in order to compensate the effects of the increased labor

Unemployment, Basic and Monetary' 159


costs in the export trades; people would become alarmed lest there
should be an inHation "like they had in Germany"; in one way or
another the monetary authority would be forced to put a stop to it.
The effects of the deflationary efforts of the monetary authority
would be to enable labor to get its increase of money wages without
any further increase in the price level but at the cost of considerable
unemployment. Real wages would rise sharply, and unless the
monetary authority had succeeded in applying the brakes very skillfully, not only would monetary employment be eradicated, but it
would be substituted by considerable monetary unemployment.4
In any case, if the monetary authority was to check successfully the
rise in prices, it would have to continue its deflationary policy until
all monetary employment had been eradicated, for so long as monetary employment remained money wages would continue to rise.
There are thus strong grounds for believing that monetary employment will never last very long without there being an intervening period of monetary unemployment. Unfortunately there are
no grounds so strong for believing that monetary unemployment will
never last very long.
However, there are some grounds for believing this: We know
that a period of monetary unemployment will be characterized by
falling money wages and falling prices. This fall is likely to become
accelerated as labor becomes more disorganized by the depression,
and as employers get more desperate and more confident in their
power to cut money wages.
Eventually, it would be imagined, a point must come when the
lunacy of the situation will be realized and the monetary authority
will be urged to put an end to the process of deflation. In this case,
the monetary authority will expand credit until prices start rising
again, and monetary unemployment will be gradually exterminated.
But the operation of this check to monetary unemployment is by
no means so certain as the operation of the check to monetary employment. For there is the significant fact that influential opinion
is far more easily frightened by the thought of an inflation "like
Germany had" than by the prospect of a slump "like they had in
America."
4

Because a check to rising prices checks industrial expansion and leads


to a fall of prices later on.

160

D. G. Champernowne
6

The argument developed in the last section suggests the conclusion


that no period of monetary employment or of monetary unemployment is likely to last for very long. We should expect there to be
alternate periods of monetary employment and of monetary unemployment, so that the actual level of unemployment would oscillate
above and below the level of basic unemployment.
There is some reason to suppose that the periods of monetary
unemployment will be longer than the periods of monetary employment, since the forces tending to bring to an end a period of monetary employment act more swiftly than the forces tending to bring
an end to a period of monetary unemployment.
Yet as a first approximation we may regard the trend value of unemployment as being equal to the level of basic unemployment, since
we may expect the actual level of unemployment to oscillate more
or less regularly about the level of basic unemployment.
If we are content to accept this degree of approximation, it follows
that in order to study the trend value of unemployment, we may
neglect monetary unemployment and monetary employment, and
consider only the movements of the level of basic unemployment. In
order to make this study, we may take over the ordinary tools of
classical analysis, very little modified.
For, by definition, the amount of basic unemployment is that
amount for which the supply price of labor is equal to the demand
price for it, where by the supply price of labor we mean the real wage
which labor would demand (at any given level of unemployment) if
its demands were not warped by any recent changes in the cost of
living. It is on the assumption that the level of real wages and of unemployment are determined by the supply and demand for labor
that the classical analysis is based; this analysis breaks down when
we try to apply it to finding the actual level of unemployment, but
if we neglect monetary unemployment and monetary employment,
and examine only changes in basic unemployment, then the assumption is correct and the analysis is valid.
Insofar as it is true that on the average monetary employment and
monetary unemployment will balance out, we may neglect their
effects also when we are considering the trend values of real wages,

Unemployment, Basic and Monetary

161

prices, the rate of investment, the rate of interest, etc., and make
use of the ordinary classical analysis but slightly modified to discover
the trend movements of these also.

7
In order to render this analysis simple we must make the usual
assumptions.
We shall consider a closed system inwhich there is only one rate
of interest. We shall treat money and labor and capital and product
as though each of these were homogeneous.
Our method of finding the trend values of unemployment percentage, real wages, money wages, prices, the quantity of money, the
rate of investment and the rate of interest will be similar to the classical analysis of the stationary state. It will not be exactly similar, because the equilibrium which we wish to examine is a dynamic
equilibrium in which investment is supposed to be taking place.
We may define dynamic equilibrium to be a state in which the
demand for and supply of labor, the demand for and supply of saving,
and the demand for and supply of money are all balanced. In order
to make this definition precise we must define the six demand and
supply functions more carefully.
By the supply of labor in any given situation we may mean "the
amount of labor which would have been forthcoming in that situation, if the real wage rate, the price level, and any other relevant
conditions had been fairly constant about their present values for
(say) one year."
We may construct analogous definitions of the demand for labor,
the supply and demand of savings and the supply and demand for
money. In each case we suppose the relevant influences to have been
fairly steady for some years.
In particular, we may define the supply of savings to be the amount
of saving which would have been forthcoming if the level of (real)
incomes, the level of real wages, and the rate of interest had all been
fairly steady for some years at their present levels, and the demand
for saving as being the amount of investment that would have taken
place if the rate of interest, the level of employment, and the price
level and the level of real wages had all been at their present levels
for some years.

162

D. G. Champernowne

Then, with these definitions, having described a position of dynamic equilibrium as one in which the supply of labor equals the
demand for labor, the supply of saving equals the demand for saving.
and the supply of money equals the demand for money, we may
proceed to the analysis of the equilibrium position corresponding to
any given situation.
Following the classical tradition and considering only basic unemployment (i.e., assuming the supply and demand for labor to be
brought into equilibrium by movements of the real wage rate), we
find that for dynamic equilibrium, the volume of employment and
the real wage rate are determined by the supply and demand for
labor.
In other words, in any given situation, if at real wage R, the supply of labor Cdefined above) would be N.CR), and the demand for
labor would be N aCR), then the appropriate values of the real wage
R and of the amount of employment N are found from the equation
N=NaCR) = NsCR)
Having found the level of employment and the real wage, we may
deduce from these and from our knowledge of the general situation
the corresponding level of real aggregate income; from this we can
estimate the supply of saving S.Cr) Cdefined above), and the demand for saving SaCr), corresponding to any rate of interest r; we
can then find the position of dynamic equilibrium for Sand r, from
the equation;

S=SaCr)=SsCr)
Since we know now the level of employment, the real wage and
aggregate income and the rate of interest we are in a position to estimate, on the basis of our further knowledge of the general features
of the situation, the real value, measured in wage units, of the
amount of money which the public will require.
Let this real value be H, and let the amount of money supplied
by the monetary authority be M, then we can derive the equilibrium
money-wage-rate, w, from the formula
M=wH5

It should be noted that our assumption that in considering trends


5

The equation M=wH is the same in principle as the quantity equation M=PH, since real wages are "given" and w is simply a convenient index of prices.

Unemployment, Basic and Monetary' 163


only basic unemployment need be taken into account was based on
the supposition that the monetary authority would never allow a
period of rising prices to continue for long, nor allow a period of
falling prices to continue for long, so that the supply of money can
only be considered independent of the demand for it within these
limits. In the broader sense, the rigidity of the money-wage-rate determines the price level and the demand for money determines its
supply. However, granted that monetary policy must not be such
as to allow the price level to alter too much, it is then true that the
supply of money M tells us the money-wage-rate according to the
equation M =wH, provided that we know H from our knowledge
of the amount of employment, the real wage, the aggregate income,
and the rate of interest.
8
The apparatus for determining the equilibrium position corresponding to any given situation can be described more simply by
means of supply and demand diagrams.
R

><"'
Supply and Demand

Supply and Demand

L-_ _ _ _ _ _ N

~Sd

L-_ _ _ _ _
b

Supply and Demand


W

for Money

Ms

~s

"'------~M

The procedure described in detail in the last section can loosely


be set forth as follows:
Ci) The level of real wages and the amount of employment are
determined by the point of intersection of the supply curve and the
demand curve for labor.
(ii) Given the level of real wages and the amount of employment,
and hence, the aggregate income, the amount of saving and the rate
of interest are shown by the point of intersection of the demand
curve and the supply curve of saving.

164 . D. C. Champernowne
(iii) Given the volume of employment, the real wage, aggregate
income, and the rate of interest we can estimate the real value H
of the amount of money demanded; if the money wage is w the
amount of money demanded is wH, so that the demand curve for
money at different wage levels is a straight line passing through the
origin. If we know the amount of money M supplied by the moneM
tary authority, then we can read off the money wage - from the
H
point of intersection of the demand curve and the supply curve.
Thus we read off in turn, the real wage rate and the amount of
employment from the demand curve and supply curve for labor;
having determined these we then find the amount of saving and
the rate of interest from the point of intersection of the demand curve
and the supply curve for saving; having determined these, we then
find the money-wage-rate from the point of intersection of the demand curve and the supply curve for money.
This procedure is only justified if the approximation involved in
assuming that, on the balance over a period of many years, the
effects of monetary employment will cancel the effects of monetary
unemployment, is close; only in this case may we consider the effects
of basic unemployment alone, and act on the assumption that the
amount of unemployment is not affected by the cost of living, but
only by the demand and supply for labor at various real wage rates.
9
The conclusion to which our discussion has led is that if it were
true that in the end real forces such as labor's demand for a certain real wage work themselves out, in spite of temporary disturbances due to monetary phenomena, such as unemployment or extra
employment due to changes in the cost of living, then the classical
analysis would be applicable to the examination of trends of real
wages, etc. It is not, however, suggested that even under these circumstances Mr. Keynes' analysis would be invalidated: it is suggested that it would then predict the same result.
As a sequel to this discussion it may not be inappropriate to consider a formal comparison of the logical structures underlying some
parts of the General Theory of Employment and the classical analy-

Unemployment, Basic and Monetary 165


sis. This discussion does not pretend to be very new or very deep, but
it may serve as a useful mnemonic to those who are slightly confused when comparing the two systems by the difference in the
terminologies used.
There is one problem which both systems of analysis set out to
attack: the problem shortly expressed is to determine the levels of
employment N, Teal wages R, Teal savings S, Tate of interest T, money
wages w, and the quantity of money M, given the general situation,
from the fact that the demand and supply of labor must balance and
so must respectively the demand and supply for saving and for money.
The two systems of analysis are only two of many possible systems each with its different conceivable set of hypotheses: let us
consider the different systems of analysis which could be conceived.
Denoting by the suffixes 8 and rl respectively, the amounts of a
factor supplied and demanded, so that N rl means "the amount of
labor demanded," our problem is to derive values for N, R, S, T, w,
and M from our knowledge of the general situation and from the
six facts:

N=N8=Nrl; S=S8=Srl; M=M,=MtJ

From our knowledge of the general situation, we should ideally be


able to find functions NBCRSrwM), NrlCRSrwM), S,(NRrwM),
SrlCNRrwM), M,CNRSrw), and MrlCNRSrw), where, e.g.,
NBCRSrwM) denotes the value that N, would have now if R, S, T, W,
M had been steady for some time at the values given inside the
brackets of N,(RSrwM).
Having found these functions, we should then be able to find the
values N, R, S, T, W, M from the six equations:
N = N 8 CRSrwM) = NrlCRSrwM)
S =SBCNRrwM)=SrlCNRrwM) ....................... (1)
M =M,(NRSrw) =MrlCNRSrw)
This would be the most general method of finding the position of
dynamic equilibrium corresponding to the given situation; it would
take account of every conceivable cross influence between any pair
of the six variables. It would, for instance, take account of the effects
of a change in the rate of interest on the supply of labor.
Now it is obvious that some effects, e.g., the effect of a change in
the amount of employment on the demand for money, are more important than others, e.g., the effect of a change in the rate of interest

166 D. C. Champemowne
on the supply of labor (ceteris paribus). For many purposes it is more
convenient to use a system of analysis which concentrates attention
on the important effects and ignores the unimportant ones than a
system which indiscriminately considers all of them.
Accordingly, any economist is at liberty to choose a special case of
the general system represented by the six equations (I), and to cross
out some of the symbols within the brackets, and so concentrate attention on the most important relationships between the variables,
e.g., by crossing out the symbol r within the brackets of N. (RSrwM),
we should clear our general system of analysis of any need to consider the reaction of changes in the rate of interest (ceteris paribus)
on the supply of labor.
Both the modified form of the classical system which we described
in sections 6, 7, and 8, and the system of analysis described in
the general theory of employment may be thought of as particular
cases of systems obtained from the set of six equations (I) by erasing
some of the variables within the brackets in such a way as to eliminate consideration from certain unimportant causal relationships, and
to concentrate it on the causes which are economically the most significant.
10

The system described in sections 6, 7 and 8 will be referred


to as the classical system: it may be represented by the six equations:
N =NsCR) =N,lR)
S = S.(Nr) = Sd(Nr) .......................... (2)
M=M.
= Md(Nrw)
and it is important to notice that the symbol tV has been erased from
within the brackets of NsCRSrwM); this means that the system does
not take account of the effect of changes in the cost of living on the
amount of labor which will be supplied at a given real wage and
amounts to a rejection of the possibility of monetary unemployment
or of monetary employment, or to the assumption that all employment is basic unemployment.
The system suggested in the General Theory of Employment will
be called the Keynesian system: it may, I believe, be roughly described by the six equations:
M=M.(r) = Md(rQ')

Unemployment, Basic and Monetary 167


S = S.CNr) = SdCNrQ) ........................ (3)
N =N.CRw) = NdCR)
where Q and Q' are certain influences such as general nervousness,
the state of the news and effects due to the expectation of changes in
the price level, etc.: these influences are suggested as having a very
significant direct effect on the demand for loans Sd and on the demand for money Md'
The six equations given above describe only that part of the
Keynesian analysis which deals with the most significant direct effects; in order to take account of the other indirect effects which are
discussed in the analysis, but on which less stress is laid, we should
have to enlarge the equations to the form:
N =N.CRw) =NdCRw)
S = S.CNRr) = SdCNRrQ)
M=M.Crw) =MdCNrwQ!)
and we should have to make them more comprehensive in order to
take account of every less important indirect economic effect suggested in the book.
But we may concentrate our attention on the difference in emphasis shown by the Keynesian analysis as represented by the equations
(3) and by the version of the classical analysis given by the equations
(2).
11
The most important difference to notice is that whereas the classical system of analysis considers the supply of labor N.CR) as depending only on the real wage R, Keynes considers the supply of
labor NsCRw) to be influenced by the money wage also.
In other words, Keynes takes account of "monetary unemployment," whereas the classical economists assume that all unemployment is basic.
The second point to notice is that Keynes thinks of the supply of
money M.Cr) as being influenced by considerations about the rate
of interest, whereas the classical economists regard the supply of
money as a datum. Thus Keynes' conception of the banks "doing
nothing" would perhaps be that of the banks adjusting the supply of
money in such a way as to keep the rate of interest constant, whereas
the classical economists would regard the banks as doing nothing if

168 . D. G. Champernowne
they kept the quantity of money constant. On the classical analysis
which we have been considering, it would be impossible for the
banks to keep the rate of interest constant, because the rate of interest depends on the supply and demand for saving.
The third point to be noticed is that the classical analysis can only
take account of the forces Q and Q' considered in the Keynesian
scheme by superimposing their effects on an equilibrium position
already found.
Whereas the classical system of analysis was to deduce the levels
of N, R, S, r, M, and w by considering in turn the demand and
supply of labor, saving, and money, the Keynesian system is just the
opposite, namely, to consider in turn the demand curves and supply
curves for money, saving, and labor.
As explained on page 163, the classical scheme is to consider in
turn the three diagrams a, b, c:

>(:".

'?~

L - - - - - -__ N

L-------..s

Supply and Demand

Supply and Demand

(a)

Supply and Demand


for Money

Ms

" - - - - - - -__ M
(e) N,R,S and r given

(b) Nand R given

but the Keynesian scheme is to consider in turn the three diagrams


d, e, f:
Supply and Demand
for Money

~.
Md

(d)

Suppl and Demand


for Saving

(e) M and r given

Supply and Demand


for Labor

~'~'.
Offer

Ss

Ns
Labor's
Demand

(t) M,r,N and S given

Unemployment, Basic and Monetary' 169


12
The contrast between the two techniques may be illustrated by applying them in turn to the problem of determining the effect of an
increase in thriftiness on the part of the people of a country whose
central bank keeps the quantity of money constant.
Let us apply the classical technique. Referring to diagram a, we
find that since neither the supply curve nor the demand curve for
labor will have been affected, the amount of employment N and the
real wage will remain unchanged; hence, aggregate real income will
remain unchanged. Turning to diagram h, we find that there is no
reason why the demand curve for saving should have moved, but that
since aggregate income has not changed, and there has been an increase in thrift, the supply curve of saving will have moved to the
right; hence, the rate of saving will have increased and the rate of
interest will have decreased. Turning to the third diagram c, we find
that the supply of money will have remained the same, but the
demand for money may have risen slightly because the rate of interest
has fallen (so that people will hoard money rather than securities),
so that the demand curve will have moved to the right; hence, the
quantity of money will not change but there may be a slight lowering
of the price level and hence, also (since real wages do not change),
of money wages.
Hence the classical analysis suggests that the results of the increase
of thrift will be: a fall in the rate of interest and an increase in the
rate of saving and a fall in money wages and in the price level, but
no change in real wages, employment or the quantity of money.
Now let us apply the Keynesian technique to the problem. Referring to diagram d, we know that the supply curve of money is a vertical straight line which does not move; similarly, the demand curve
does not move for any direct reason. Hence, the quantity of money
and the rate of interest will not be directly affected by the increased
thriftiness. Turning to diagram e we find that the demand curve for
saving will not be directly affected by the increased thriftiness, but
that the supply curve of saving will be shifted to the right; hence, if
the curves have the slopes showns in diagram e, there will be a fall in
employment and a slight fall in the amount of saving. Turning to
diagram f, we find that the fall in employment is likely to have in-

170

D. G. Champernowne

creased the marginal productivity of labor and the real wage offered
by employers, so the demand curve for labor will have moved to the
right; on the other hand, the fall in employment will have made
labor more prepared to accept a cut in money wages, so that the
supply curve will have been lowered. There will be an increase in
the real wage and perhaps a slight fall in the money wage, and,
hence, a definite fall in the price level.
Hence the direct effects of the increase in thriftiness will be a
decrease in savings and employment and a rise in real wages and a
fall in the price level, with no change in the quantity of money or the
rate of interest.
Certain indirect effects remain to be considered. The supply curve
of money will not be affected: but there will be two indirect effects
on the demand for money. On the one hand, the fall in prices and
employment will decrease the demand for money for business, but,
on the other hand, the psychological effects of depression will be to
increase the demand for money for hoarding. Let us suppose that
the first effect predominates so that the demand curve for money in
diagram d is moved just a little to the left: we see that the rate of
interest will be slightly lowered.
There will be two indirect effects on the demand for saving: the
fall in the price level may lead people to expect a further fall, this
will lower the demand for saving by making investment less profitable; on the other hand, the slight fall in the rate of interest will increase the demand for saving. Let us suppose that the second of
these effects predominates: then the demand curve for saving may
move slightly to the right, so that the fall in saving and in employment may be slightly offset. As a result of this the fall in the price
level and the rise in real wages will also be slightly offset.
We find then that the result of the increased thriftiness will be a
small decrease in the rate of interest, a fall in the rate of saving and
in employment and in prices, and a rise of real wages.
These results may be contrasted with the results predicted by the
classical analysis, namely, a fall in the rate of interest and in prices,
but a rise in the rate of saving and no change in the amount of employment or the rate of real wages.
Let us remind ourselves of the difference in the two techniques.

Unemployment, Basic and Monetary

171

13

The classical analysis uses the equations:


N=NsCR) = NdCR); S=S.CNr)=SdCNr); M=M.=MdCNrw)
whereas the Keynesian analysis uses the equations
N = N .CRw) = N dCRw); S = S.CNr) = SdCNrQ);
M=M.=MdCNrwQ')
in the case where the supply of money M. is fixed.
Hence the Keynesian analysis must give practically the same result
as the classical in the particular case where:
Ci) N.CRw) the supply of labor depends hardly at all on the
money wage w, but almost entirely on the real wage R.
(ii) NdCRw) the amount of labor demanded is independent of
the money wage and depends only on the real wage.
(iii) The factors Q and Q' have no effect on the demand for
money or for saving.
These conditions are, of course, extremely improbable ones, but it
is instructive to see how the Keynesian analysis would under these
circumstances give almost the same result as the classical analysis.
As in the discussion on page 169, we should find that the direct
effects of an increase of thriftiness would be a fall in the rate of
saving and in employment and in money wages and prices, with a
rise in real wages. But now since the supply of labor is hardly sensitive at all to the money wage but only to the real wage, a large fall in
money wage will take place, before labor insists on a high real wage:
hence the supply curve of labor in diagram g will be almost vertical.
Hence the decline in money wages and in prices due to the direct
effects would be very large. We shall see that due to indirect effects
the fall will not be so large.
The indirect effects of the considerable fall in prices and money
wages will now be considerable. The demand for money will be considerably reduced, and since we are assuming the effects Q' to be
absent, the demand curve for money will be considerably shifted to
the left, and there will be a considerable lowering of the rate of interest. This will have a considerable effect in raising the demand for
saving and, hence, of moving the demand curve for saving in diagram
e considerably to the right, since we are assuming the effects Q to be
absent. This will be such as actually to reverse the decline in the rate

172

D. G. Champernowne

of saving and tum it into an increase in saving. The decline in employment will be greatly reduced. We proceed to diagram g, and find
W

Supply and Demand


for Labor

X"'"
Offer

labor's

Demand

Nd

Ns
g

that since the decline in employment has been so much reduced, the
rise in real wages will be very small and the fall in money wages and
in prices will not be very large. Hence, in this case the Keynesian analysis gives the same result as the classical analysis: an increase in the
rate of saving, a fall in the rate of interest, and a slight fall in the
money-wage-Ievel; but the Keynesian analysis predicts also a slight
fall in employment and increase in real wages. The classical analysis
would also have predicted these effects if it had been postulated that
the effect of the fall in prices would be to raise a little the real supply
curve for labor in diagram a.
14

It is not the purpose of the above analysis to suggest that the


Keynesian analysis is merely an elaboration of the classical analysis.
The purpose is to show just how much artificiality must be introduced
into the conditions assumed in order that the Keynesian technique
should lead to the same results as the classical technique: it is to bring
out the point that the Keynesian analysis differs from the classical analysis in the fact that it picks out for emphasis different economic
forces such as the stickiness of money wages, which are evidently
extremely important in the short run. It is also intended to show that
in so far as there is a "real" tendency which must make itself felt
eventually for labor to insist in a certain standard of life, and for
labor's bargaining power to increase as unemployment decreases,
then provided that the monetary authority does not allow labor to be
misled by too long periods of rising or falling cost of living, the "real

Unemployment, Basic and Monetary 173


supply curve of labor" may be a useful concept for estimating the
trend of unemployment and real wages of the rate of interest and of
saving. If there is such a real tendency, then, when considering such
trends, the real wage may be more significant than the money-wage
in determining the trend level of unemployment. The Keynesian
technique is then likely to yield the same results as the classical
technique, and it may be simpler to consider the classical analysis
direct.
This method will be of no avail if outlets for investment are so
scarce or if the employers are so nervous of any increase in the supply
of money that they hoard, and it is impossible to lower the rate of
interest sufficiently to cause sufficient investment to keep prices and
money wages from falling. For in this case the monetary authority
will not be able to prevent a constantly falling cost of living and there
will necessarily be monetary unemployment constantly. It would be
quite inappropriate to make use of the classical analysis in such a
case.
In such a situation the most significant economic fact would be
enormous monetary unemployment. The classical analysis is ill suited
to deal with this because it can only treat it as an indirect effect
caused by a lowering of prices raising the real supply curve of labor.
In such a situation the Keynesian analysis is particularly convenient
because it treats "nervousness" Q and Q' as direct effects on the rate
of interest and on the demand for saving, and regards the demand for
saving as having a direct effect on the real wage, the money wage and
the amount of employment (basic and monetary).
It is only in a situation, if such can exist, where only basic unemployment matters, and where uncertainty and nervousness are not
very important, that the classical analYSis has the advantage over the
Keynesian, because it treats the effect of real wages on the supply of
labor as direct, and the effect of changes in the quantity of money on
the price level as direct. In all other situations the new technique
would seem to be advantageous.

Expectations and the Links


Between the Economic
Future and the Present
[1963J
Thou art no sabbath-drawler of old saws,
Distill'd from some worm-cankered homily;
But spurr'd at heart with fieriest energy
To embattail and to wall about thy cause
With iron-worded proof . . .
. . . I will stand and mark.

Early Sonnet to

J. M. K.

(John Mitchell Kemble)


by Tennyson.

IN CHOOSING A GROUP OF topics out of the multitude


covered in Keynes' General Theory, I have sought one to which
Keynes himself attributed great importance, one which is theoretically interesting and which has not been fully explored already.
I have been attracted to the group concerned with the links between the economic future and the present, because Keynes clearly
believed that it was in the treatment of these links that orthodox
theory was particularly defective. He wrote: "The fact that the
assumptions of the static state often underlie present-day economic
theory, imports into it a large element of unreality. But the introduction of the concepts of user cost and of the marginal efficiency of
capital, as defined above, will have the effect, I think, of bringing it
back to reality, whilst reducing to a minimum the necessary degree of
adaptation. It is by means of the existence of durable equipment that
the economic future is linked to the present."l

1The General Theory of Money, Interest, and Employment, p. 146.

174

The Economic Future and the Present 175


After the General Theory was published Professor Ohlin conducted a controversy with Keynes in the Economic Journal and made
out a strong case that Keynes had chosen to ignore economic theory
emanating from sources ranging from Trinity College, Cambridge, to
Stockholm and Vienna, which took full account of the interactions of
present policy and expectations about the future. He also made some
valid criticisms of Keynes' failure himself always to distinguish
clearly between ex ante and ex post concepts. But this controversy
petered out before either party had shown much appreciation of what
the other was saying. There is still therefore much to be learned by
reading again through Keynes' General Theory and marking particularly those passages that deal with the influence of expectations and
the links between the economic future and the present.
It is on such a revisiting of Keynes' General Theory that the reader
is now invited to embark and not on any consideration of more
recent attempts to interpret and to improve upon his ideas. Failure to
refer to these attempts should not therefore be taken in all cases to
imply disapproval, disparagement, or plain ignorance of them.
The links between the future and the present are the various
means by which the present economic situation can affect expectations about the future, and the various ways in which expectations
about the future affect present economic behavior. Throughout
Keynes' General Theory, the importance of these links is repeatedly
stressed and their influence is a theme that can be traced like a thread
reappearing at each stage of the argument.
It will be unnecessary to remind readers today of the detailed construction of that argument, but for purposes of exposition, it is convenient to regard the discussion as consisting of a sequence of six
stages.
The main object of writing the General Theory may be regarded
as that of refuting the thesis that the cause of heavy unemployment is
the insistence by wage earners on high real wages. In place of that
thesis a correct theory of employment is to be provided. The following stages of the discussion may be distinguished.
I. Marginal prime costs.
Real wages are not determined by the
bargaining of the wage earners: that merely fixes money wages.
Given the level of employment, the true determinant of real wages is

I76 D. C. Champernowne
the schedule of the marginal prime costs of wage goods in terms of
wage units.
2. Propensity to consume. Unemployment is due, not to insistence on excessive real wage rates, but to an inadequate propensity to
consume, and/or to insufficient investment.
3. Schedule of the marginal efficiency of capital. Insufficient investment is due, not to lack of thrift, but to an inadequate schedule
of the marginal efficiency of capital and/or to excessive rates of interest.
4. Schedule of liquidity preference for money. Excessive rates of
interest are due to the high liquidity preference which people have
for holding money.
5. Effects of money wage reductions during a slump. Even if
money wages were flexible, this would not cure unemployment but
would cause violent instability.
6. The properties peculiar to money. The causes of unemployment have been traced to the tendency of the money rate of interest
to be too high. It is certain peculiar properties of money which distinguish it from all other assets and result in the high money rate of
interest being the main obstacle to the abolition of unemployment.
eWe shall insert a further section just before the last section on
the properties peculiar to money: in this we will examine in particular Keynes' treatment of the effects of the present on the expectations
about the future, as distinct from his treatment of the effects of these
expectations. )
I.

The Determination of Marginal Prime Costs.

According to Keynes, expectations about the future played an


important part in determining the pricing policy of employers and
hence the level of real wages and the distribution of income between
wages and profits.
The influence of long-term expectations on pricing policy is allowed for in the General Theory by the introduction of the concept of
marginal user cost. Keynes specifically described user cost as "one of
the links between the present and the future";2 however, his discussion of this concept has received little attention and most economists
would probably agree that this rather elaborate tool of thought was of
II

Ibid., p. 69.

The Economic Future and the Present 177


little use. It may nevertheless be argued that, when carefully interpreted, this concept is relevant both to controversies about the rationale of full-cost pricing and to arguments about the proper pricing
policies for state-controlled enterprises.
Keynes was prepared to accept the view of orthodox economists
that the prices of both wage goods and other goods would be fixed in
such a way that profits were maximized in accordance with the principles equating marginal costs with prices (or marginal revenues).
He therefore assumed that prices are fixed with an eye to these marginal prime costs.
But he considered that orthodox economists made the mistake of
supposing that marginal prime costs consisted only of the marginal
costs of certain factors of production and materials. The ingredient
which he accused them of neglecting was what he calls the marginal
user cost (borrowing a term used by Alfred Marshall). He thought
that they overlooked the element of expectation in the estimation of
marginal prime costs by neglecting to make any allowance for the
effect of the choice of output on the prospects of making profits in
subsequent periods.
The marginal user cost arising from extra output is defined so as
to equal the consequent loss of the capital equipment (at the end of
the period) plus any extra expenditure during the period on the
products of other businesses. It thus includes all extra cost connected
with the long-term prospects.
But to appreciate the importance of this allowance for marginal
user cost, one must remember that by the value of capital equipment
Keynes does not mean its historic cost or i:.s replacement cost; he is
specifically concerned with the value of prospective yields. Thus in
an explanation of how to calculate the user cost involved in the decision to use a particular piece of equipment instead of laying it up,
he uses the following words: "it must be arrived at, therefore, by calculating the discounted value of the additional prospective yield
which would be obtained at some later date if it were not used now."3
Keynes was mainly concerned with the effect of extra output on the
dates at which the equipment would need replacement. But it is clear
that his definition of marginal user cost would cover also any effects
3

Ibid., p. 70

I78 . D. G. Champernowne
on the prospective markets for the firm's goods. These effects might
be favorable through popularizing the various goods provided by the
firm, or unfavorable through spoiling the market by necessitating a
reduction in price. Marginal user cost might also reSect the unfavorable effects on any increase in prices, associated with a reduction
of output, due to encouraging new firms to break into the market.
The effect would be to lower the marginal prime cost of the last unit
of output, and so to discourage any reduction of output.4
Keynes draws two sets of practical conclusions from his consideration of marginal user cost. He argues that towards the end of a
slump, even apart from the direct effects of the beginnings of improvement in employment and money wages, prices are likely to be
increased, because the expectation that some equipment will have to
be renewed much sooner will sharply raise marginal user costS.5 He
argues similarly that organized schemes for scrapping redundant
plant may succeed in raising prices, even if they only succeed in
scrapping half the redundant plant: for they bring the expected date
of the absorption of redundancy nearer, and thus raise marginal user
cost and hence also prices.6
2.

The Psychological Propensity to Consume.

In a closed system, savings and investment, as defined in the


General Theory, must be equal because each when added to consumption gives the same total, namely, income.
A rough indication of Keynes' theory of employment can then be
given as follows. Investment being given, income and employment
will settle at that level where individuals will collectively decide to
save an amount equal to that given investment. This level of employment can be found if we know the psychological propensity to
consume YeN), relating consumption to employment: it is that N
which is obtained by solving the equation
Keynes' insistence that the marginal theorists of his day overlooked
the need to allow for effects on long-term expectations is a little surprising: he may well have been unaware that the undergraduate
lectures on the theory of value given in 1934 and in 1935 by the
late Mr. G. F. Shove in King's College, Cambridge, largely consisted
of a very painstaking examination of the various allowances that
should be made on this count.
5 General Theory, p. 3 02
6 Ibid., p. 7 1

The Economic Future and the Present . 179


Y(N)+D s =0(N)

where 0(N) is the national income corresponding to the employment level N, where Ds is the level of investment and where YeN)
is the consumption function.
This is, however, a very crude indication of the theory since it
suggests that the various terms in the equation are actual or "ex post"
magnitudes. Keynes made it quite clear in Book II of the General
Theory that all three terms relate to what is expected and not to
actual levels.
Thus the propensity to consume is what the community is expected to spend on consumption when employment is N; Ds is what
it is expected to spend on investment: finally 0(N) is the total income whose expectation makes it worth while for employers to employ a total of N men.
This introduction of expectations raises two related questions.
Whose expectations are involved? Need the two sides of the equation
agree, since they may correspond to different expectations by different people?
In fact the expectations are formed by the same set of people on
the two sides of the equation, namely, the various employers of men.
The equation is built up from a set of separate equations each one
relating to a separate firm making consumption goods or investment
goods. In each equation the expectations of the two sides are formed
by the same individual-the employer in the firm concerned. There
can thus be no inconsistency arising from a conflict of expectations
between those forming them on the two sides of the equation.
However, this does not mean that the equation will at every
moment of time be satisfied in the same tautological way as must be
the equation of ex post savings and ex post investment. The fact is
that the symbol N cannot refer to the same magnitude on both sides
of the equation for 0(N): because N involved in YeN), the propensity to consume, is actual ex post employment according to the
published,figures, whereas the N involved in 0(N), the supply function, is determined by the equation, and corresponds to expected
income: in short, the N in the YeN) is yesterday's employment as
known publicly today, and the N in 0(N) is today's employment
resulting from the expectations based on this knowledge.
The day "stands for the shortest interval after which the firm is
free to revise its decision as to how much employment to offer." The

180

D. G. Champernowne

adjustment of employment and income to a change in investment or


to a change in the propensity to consume is therefore not quite immediate. But since the day is a very short time, "it is so to speak, the
minimum effective unit of economic time,"7 the adjustment is speedy.
It is indeed far more speedy than if there were involved the time lag
between two complete rounds of consumption expenditure in the
Keynes-Kahn multiplier process.
Keynes assumes that employers anticipate later expenditure and
immediately produce the consumption goods which they expect
eventually to sell. He thus eliminates the effects of any lag between
the two rounds of consumption and retains only those due to the
interval required before the firm is free to revise its decision as to
how much employment to offer. The treatment of expectations in
this part of the General Theory is thus a key element in enabling
Keynes to regard the adjustment of employment and income to
changes in the rate of investment as being almost immediate.
Keynes is not, however, altogether consistent in his treatment of
the aggregate demand function and the associated propensity to consume. Throughout Book II he consistently treats these functions as
being concerned with expectations: suddenly, however, at the beginning of Book III, he switches from expectations to actual "ex post"
values. On page 89 he has written: "the aggregate demand function
relates any given level of employment to the 'proceeds' which that
level of employment is expected8 to realize. The 'proceeds' are made
up of the sum of two quantities, the sum which will be spent on
consumption when employment is at the given level-and the sum
which will be devoted to investment." But on the following page
and throughout the rest of Book III, he consistently refers to the
function giving the propensity to consume as being "the functional
relationship between Yw, a given level of income in terms of wage
units, and C w the expenditure on consumption out of that level of
income," and in this formulation, expectations are no longer mentioned.
The result of this sudden switch from a function concerned with
expectations to one concerned with actual behavior is to lend a suggestion of greater stability to the theoretical model than can in fact
7
8

Ibid., p. 47.

My italics.

The Economic Future and the Present

181

be claimed so long as the propensity to consume is regarded, as it is


in Book II, as being composed of the sum of a set of separate functions showing how much each firm separately will expect to be
spent on its products as a function of the known amount of employment or of the corresponding known level of total income. For
clearly, when thus defined, the propensity to consume is liable to
all kinds of Huctuations due to the state of confidence about the
prospects of each firm jumping up and down.
It is difficult to make any generalizations about how such a function may be expected to behave, unless one assumes that most firms
will in fact form expectations which are nearly always quite correct.
This would, however, not be characteristic of Keynes' way of thinking. If, on the other hand, one regards the propensity to consume as
being concerned not with expected consumption but with actual
consumption expenditure, then while it is quite possible to make the
kind of generalizations that Keynes does make about the probable
form of this function, it is not reasonable to regard the function,
thus described, as the same one which determines employment by
means of the mechanism which has been discussed in Book II. For
that mechanism quite clearly depends on using a function relating
expectations about consumption expenditure to the actual level of
employment.
It seems probable that Keynes himself largely overlooked the
hiatus in his argument at this point. Had he noticed it, he would
perhaps have defended his procedure on the grounds that expectations about the reactions of a firm's sales to the level of employment
are likely to be based on recent experience of such reactions, with
the result that, for the community as a whole, such expectations are
likely to prove fairly accurate, partly as a result of errors canceling
out. If there were some difference between the two kinds of propensity to consume, that defined ex ante and that defined ex post,
such a difference would be noticeable only at times of pronounced
optimism or pessimism, and that effect he would have treated, in his
customary manner, as an additional reason why waves of such optimism and pessimism might start off large increases or decreases of
economic activity. But, as stated above, it does seem as if he largely
overlooked the omission of this link in his argument.

182

D. G. Champernowne
3. The Marginal Efficiency of Capital.

Keynes' views about the manner in which the rate of investment


depended on the rate of interest are up to a certain point surprisingly
similar to those published six years earlier (1930) by Professor
Irving Fisher. The revolutionary part of Keynes' doctrine concerned
the determination of interest rates; his theories concerning the
schedule of the marginal efficiency of capital were novel mainly in
the stress he laid on the dependence of this schedule on expectations about future changes, and on the frenzies of the speculators,
another novel element was the importance he attached to diminishing
returns in the capital construction industries as an influence limiting the amplitude of the fluctuations in the rate of capital investment. 9
Keynes followed Fisher in supposing that the rate of investment
would be determined by the condition that the rate of return on the
cost of capital would equal the rate of interest with due allowance
for risk and uncertainty. Keynes' schedule of the marginal efficiency
of capital set out what marginal rate of return on cost corresponded
to each (potential) aggregate level of investment: there was only
one possible (actual) level of investment, namely, that which resulted in the marginal rate of return equaling the rate of interest in
the sense explained above. Keynes adopted Fisher's definition of the
marginal rate of return on costlO with due acknowledgment, but referred to this concept as the marginal efficiency of capital.
The following comments of Keynes concerning the role of the
marginal efficiency of capital are entirely in line with the theories
earlier published by Professor Fisher: "It is by reason of the existence
For such diminishing returns ensure that when investment is increased, there will be an immediate rise in the marginal cost of further
investment, and that this rise will be sufficient to match the improvement in expected quasi-rents. For this reason the curve showing the
marginal efficiency of capital falls rapidly in the direction of increasing
investment, and this fact limits the responsiveness of investment to
improved expectations. Keynes relies mainly on this effect to restrain
changes in the rate of investment and to ensure that no small accidental change in the economic climate is likely to set off a cumulative
process of rising expectations and expanding activity. See General
Theory, p. 136.
10 Fisher, The Theory of Interest, p. 159 et seq. General Theory, p. 140.
9

The Economic Future and the Present . 183


of durable equipment that the economic future is linked to the
present." "The expectations of the future should affect the present
through the demand price for durable equipment."ll
When he attacks orthodox economists in the following passage, it
is difficult to suppose that he had in mind Professor Fisher, whose
published theories showed little trace of the alleged mistake. Keynes
argued that it is mainly through the schedule of marginal efficiency
of capital "(much more than through the rate of interest) that the
expectation of the future influences the present. The mistake of regarding the marginal efficiency of capital primarily in terms of the
current yield of capital equipment, which would be correct only in
the static state where there is no changing future to influence the
present, has had the result of breaking the theoretical link between
to-day and tomorrow."12
Keynes laid tremendous stress on the volatility of the schedule of
the marginal efficiency of capital in response to changes in the
climate of opinion. He never for a moment allows the reader to forget that the profits or quasi-rents which determine the schedule are
expected profits and not actual profits. This again raises the interesting question as to whose expectations are involved. The obvious
answer is that it is those of the employer planning the capital investment, but Keynes leans quite far in the direction of the view, apparently held by Professor Fisher, that it is the expectations of the
private investors dealing on the stock exchange. For example, Keynes
writes: "Thus certain classes of investment are governed by the
average expectation of those who deal on the Stock Exchange as
revealed in the price of shares, rather than by the genuine expectations of the professional entrepreneur."13 And he indicates in a footnote his opinion that the importance of the classes of enterprise
where investment was still free of the influence of stock exchange
speculation was rapidly declining. So he led up to his masterly invective: "When the capital development of a country becomes a
by-product of the activities of a casino, the job is likely to be ill-done.
The measure of success attained by Wall Street, regarded as an
institution of which the proper social purpose is to direct new inGeneral Theory, p. 146.
Ibid., p. 145.
13 Ibid., p. 15 1

11
12

184 . D. G. Champernowne
vestment into the most profitable channels in terms of future yield,
cannot be claimed as one of the outstanding triumphs of laissez-faire
capitalism, which is not surprising, if I am right in thinking that
the best brains of Wall Street have been in fact directed towards a
different object."B A few pages later he prophesies that we should
"see the state, which is in a position to calculate the marginal efficiency of capital goods on long views and on the basis of the general
social advantage, taking an ever greater responsibility for directly
organizing investment."15
Modern critics would probably complain that Keynes conceded
too much to orthodox opinion when he allowed that movements in
the rate of interest can exert an important direct influence on the
aggregate rate of investment. Employers are nowadays quoted as saying that in the uncertain situation that faces them, changes in the
cost of investment due to shifts in interest rates are altogether too
insignificant to be taken into account in deciding whether or not
to undertake a particular capital project.
We have already seen that Keynes considered that the influence of
changes in the climate of opinion, by affecting the marginal efficiency
schedule, were in fact likely to have much more influence than any
movement in the rate of interest. Nevertheless he did suppose that
movements in the rate of interest could exert an appreciable effect,
and it is worth considering how he could have defended this view
against criticisms.
He could have pointed out that the rate of interest at which an
employer has to equate the marginal efficiency of his capital investment is in fact the marginal rate at which his institution individually can raise further finance. In times of brisk trade and
golden expectations, an institution will borrow up to a point where
this level is already well above the general market rate of interest. A
sudden hardening of credit, although it has only a small effect on the
general market rate of interest, can evidently drive the effective
marginal rate of interest at which a particular institution can borrow
enough to be able to continue its current level of investment, so very
far, as to encourage the employer at least to postpone some of his
schemes.
14
15

Ibid., p. 159.
Ibid., p. 164.

The Economic Future and the Present 185


Keynes says: "Now it is obvious that the actual rate of current
investment will be pushed to the point where there is no longer any
class of capital-asset of which the marginal efficiency exceeds the
current rate of interest. Hl6 If this is to be strictly true the term
marginal efficiency must relate to a comparison of the actual scheme
of investment undertaken with the next best alternative, which
might be a slight postponement of this actual scheme. This is clearly
the interpretation Fisher intended, for he was very much concerned
with the optimum timing of investment. However, it seems more
likely that Keynes interpreted marginal to refer to the comparison
with not undertaking the marginal project at all. With this interpretation it is no longer necessary that the marginal efficiency of the
capital asset should not exceed the rate of interest: it is merely necessary that there should be some period of years from the present over
which its marginal efficiency does not exceed the rate of interest during that period. If the rate of interest is 5 per cent, an investment
project whose long-term marginal efficiency is 6 per cent will not be
worth undertaking yet if its marginal efficiency, taking all expectations into account, is less than 5 per cent during the first two years.
For it would then follow, from the definition, that it would pay
better to postpone the investment for two yearsP
It seems unlikely that Keynes can have overlooked this point. Indeed, in Chapter 17, on the properties of interest and money, to be
discussed later in this note, Keynes does take account of the effect
of expected changes in the money value of capital assets in discussing
the determination of marginal efficiency. It may well be, then, that
in the earlier discussions he was deliberately ignoring complications
which were not essential to the main point which he there wished to
explain.
4- Schedule of the Liquidity Preference for Money.

At first sight it might seem that in his treatment of liquidity


preference, Keynes was again following Professor Irving Fisher who
16

17

Ibid., p. 13 6.
Since the short-term marginal efficiency of capital may be even more
sensitive to Ructuations in the climate of opinion than the long-term
marginal efficiency, this point may have some importance in explaining
Ructuations in activity.

186 . D. C. Champernowne
in 1930 had written of money "the convenience of surely being able,
without any previous preparation, to dispose of it for any exchange,
in other words its liquidity, is itself a sufficient return upon the
capital which a man seems to keep idle in money form. This
liquidity of our cash balances takes the place of any rate of interest
in the ordinary sense of the word. A man who keeps an average cash
balance of $ 1 00 rather than put his money into a savings bank to
yield him $5' a year, does so because of its liquidity. Its readiness for
use at a moment's notice is, to him, worth at least $5' a year."18
Keynes does not seem to travel far from this when he writes: "A
rate of interest at any time, being the reward for parting with
liquidity, is a measure of the unwillingness of those who possess
money to part with their liquid control over it. The rate of interest
... is the price which equilibrates the desire to hold wealth in the
form of cash with the available quantity of cash."19
But Professor Irving Fisher expressly denied the suggestion that
changes in the quantity of money could have any lasting effect on the
rate of interest, whereas Keynes seized on the conclusion that by
manipulating the quantity of money the authorities could, within
certain limits, control the rate of interest. It is not our concern here to
examine why Professor Fisher thought that any such control must be
transitory, but to examine the subtle considerations concerning expectations about the future which Keynes brings in when analyzing
the motives underlying a man's liquidity preference for holding cash.
Keynes classifies a man's motives for holding cash into the transactions motive, the speculative motive and the precautionary motive.
Expectations might exert some influence on the transactions motive
if he expected a large change in the level of activity of his business,
but the really important effects of changes in expectations and the
climate of opinion will of course be upon his speculative and precautionary motives for holding cash.
A man will have speculative motives for holding cash when he
expects falls in the prices of the alternative forms in which he might
store his wealth: in practice one of the most important classes of
such alternative assets are fixed interest government securities. Similarly, a man will have precautionary motives for holding cash when,
18

19

Fisher, op. cit., p. 216.


General Theory, p. 16 7.

The Economic Future and the Present . 187


even if he does not on balance expect the price of alternative assets
to fall or rise, yet he is uncertain about what their money value may
be at any time when he may suddenly need to sell them for ready
cash. Again, one of the most important classes of such alternative
assets is that of fixed government securities.
Let us then examine in all its subtlety Keynes' treatment of the
inRuence of a man's expectations about the general level of fixed
interest security prices.
At the first level of sophistication, he allows that the man may
expect these prices to rise (or to fall): thus he may be a bull and
wish to hold nearly all his assets in the form of securities, or he may
be a bear, in which case he will wish to hold nearly all his assets in
the form of money.
If all men thought alike the prices of securities would settle at a
level where no considerable change was expected, but Keynes introduces a second level of sophistication by allowing that simultaneously
some men will be bulls and hold securities and other men will be
bears and hold money. Thus the price of securities must be at such
a level that a speculative demand for money by bears is just satisfied by the supply of money available for the speculative purposes.
At the third level of sophistication, Keynes points out that men are
not dogmatic in their bullishness or bearishness but are uncertain in
their own minds. Although they may regard a rise in the security
prices as highly probable they might nevertheless allow that there
was an appreciable risk of a severe fall. Thus although these bulls
may own most of their assets as shares, yet they will own a certain
amount of them as money because of the precautionary motive. Thus
if a man's opinion about how the price of securities is going to move
in the near future may be represented by a probability distribution,
then the quantity of money demanded is a function not merely of the
arithmetic mean of that distribution but of its whole shape as well.
The greater the dispersion about a given mean, the more money he
will wish to hold.
Nor does Keynes ignore the fourth degree of sophistication which
acknowledges that men cannot describe their expectation in the form
of a probability distribution at all but are merely aware that there
are a range of possibilities, some of which appear more plausible
than others; he makes as little use as possible of precise measures of

188 . D. G. Champernowne
expectation and talks instead of such psychological states as bullishness, bearishness, and uncertainty.
All these degrees of sophistication are exemplified in the following
passage, whose understanding incidentally provides an excellent
test of the capacity of university students of economics. "Nevertheless, circumstances can develop in which even a large increase in the
quantity of money may exert a comparatively small influence on the
rate of interest. For a large increase in the quantity of money may
cause so much uncertainty about the future that liquidity-preferences
due to the security-motive may be strengthened; whilst opinion about
the future of the rate of interest may be so unanimous that a small
change in present rates may cause a mass movement into cash. It is
interesting that the stability of the system and its sensitiveness to
changes in the quantity of money should be so dependent on the
existence of a variety of opinion about what is uncertain. Best of all
that we should know the future. But if not, then, if we are to control the activity of the economic system by changing the quantity of
money, it is important that opinions should differ."20
These ideas follow naturally from the considerations which we
have just described. Best of all would be if we all knew the future,
for then there would never be a precautionary motive for holding
cash, and only if the rate of interest was exceptionally low would
there be any speculative motive. Therefore the rate of interest could
easily be kept low enough to ensure full employment and we should
all know that it would remain so.
Now suppose that there is a variety of opinions about the future,
although each individual is certain about his own private opinion: in
this case, the sensitivity of the rate of interest to changes in the
quantity of money will reflect the variety between individuals in
what they expect is going to be the level of the rate of interest in the
near future. If this variety of opinion is large then only a moderate
shift in the speculative demand for money Cat the ruling rate of
interest) is capable of causing quite a large change in the rate of
interest. Under these circumstances, the rate of interest will be responsive to monetary management and this is the explanation of the
paradox that it is desirable that opinions should differ. Moreover, if
20

Ibid., p. 17 2

The Economic Future and the Present . 189


individuals are uncertain in their own minds about the future
interest rate, this might, provided this state of mind persisted, make
the rate of interest even more sensitive to changes in the supply of
money; the explanation of this is that nobody would then have the
arrogance to sell almost all of his securities for cash in response to any
but a large upward shift in the price of those securities.
Suppose, however, that any addition to the money supply added to
the prevailing uncertainty about the future level of interest rates and
security prices. That would reinforce people's readiness to sell their
securities, just because each would now wish to hold money against
the possibility of a fall in security prices, even though he considered
the possibility of a rise to be just as great. In these circumstances, a
large addition to the quantity of money need not cause any rise in
security prices or fall in the long-term rate of interest at all.
The passage discussed above is but one example of the way in
which Keynes was able to make use of his distinction between the
speculative motive and the precautionary motive, and again of that
between variety amongst individuals in their expectations and uncertainty within each individual's own personal expectations.
5. Effects of Cuts in Money Wage Rates.
It is interesting to observe how Keynes considered in advance the
criticisms of any economists who might argue that unemployment
could be cured by a reduction in money wage rates without altering
the quantity of money, on the grounds that this must in some real
sense have the same expansionary effects on economic activity as
would an equal proportionate increase in the quantity of money with
money wage rates held constant.
It was indeed natural that Keynes should anticipate some such
criticism, since Professor Irving Fisher had clearly stated six years
before in his Theory of Interest the view that it was through price
and money wage Hexibility that the rate of interest was inHuenced
towards that level which allowed the rate of investment to settle at
that figure which was consistent with full employment and the savings forthcoming from incomes at their full employment level. It is
remarkable that Keynes considered these criticisms in advance so
fairly and so fully.
Keynes naturally denied that the "real" effects of a money wage

190

D. C. Champernowne

cut with the quantity of money held constant would be the same as
those of an increase in the quantity of money with money wage rates
held constant. The difference, he pointed out, arose from the markedly divergent expectations which would be provoked by the two
developments. In the former situation of falling money wage rates,
people will expect further falls in money wages and prices, but in the
latter situation of an expanding money supply they would be most
unlikely to expect falls in money wages or prices. After wage cuts, the
uncertainty about future monetary expenditure on wage goods, associated with the risk of further wage cuts, both lowers the present
effective demand for these goods and lowers the schedule of the marginal efficiency of capital.
Quite apart from these arguments, it is possible to use Keynes'
method of analysis of expectations to show that under certain admittedly extreme assumptions, the demand for money at a given level
of activity will be almost completely independent of the level of
money wages, although, as we have seen, it will be sensitive to any
expectations of a change in their level.
The gist of this argument is that in slump conditions most capitalists will want to hold a considerable proportion of their assets as
precautionary balances, and the size of these precautionary balances
will be related to the value of their holdings of securities. The value
of fixed interest securities will not be directly affected by the fall in
money wages: that of equities will be lower but only because they
become less attractive to hold. In so far as money held for transaction purposes may be counted as part also of the precautionary
balances, any reduction in the transaction demand due to wage reductions will leave unaffected the total demand for money to be
used for transactions and precautionary motives. On these grounds it
appears that any reduction, due to wage cuts, in the demand for
money will be extremely feeble.
When differences of opinion between bulls and bears are brought
into the picture, the argument is weakened, since the only demand
of the dogmatic bulls for money will be for transaction purposes and
thus their bullishness will be allowed to have greater influence on
security prices when they can reduce their bank balances a little on
account of cuts in money wages and associated price cuts.
The argument on which Keynes himself relied most was that

The Economic Future and the Present .

191

money wage flexibility, far from allowing an orderly return to full


employment, would result in the economy being driven either to a
state of reckless inflation with money wages, prices, effective demand,
and (initially) employment all shooting up, or else to a state of panic
with money wages, prices, employment, and effective demand plunging into a bottomless sink. He regarded the relative stability actually
found in the economic system as being due to the stickiness of
money wages.
Keynes did not of course consider every conceivable manner in
which a reduction of money wage rates might directly or indirectly
stimulate employment. Having discussed seven possible effects of
varying degrees of practical importance he concluded: "This is not a
complete catalogue of all the possible reactions of wage reductions in
the complex real world. But the above cover, I think, those which are
usually the most important."21 One theoretically possible effect which
he did not nnd it worth while to mention was that of an increase in
the propensity to consume resulting from the increased purchasing
power of those assets and balances whose prices are nxed in terms of
money. Some stress has been laid on this theoretical possibility by
recent commentators, but there seems to be no good evidence on
which to doubt Keynes' good judgment in ignoring this effect as
being of no practical importance.22
While accepting Keynes' argument that even if money wage rates
were flexible, this would make the slump worse and not better, one
may still regard some of his strictures on other economists such as
Pigou and Fisher for their attitude to money wage changes as excessive. Keynes complained that "Professor Pigou (with others) has
been accustomed to assume that there is a presumption in favour of
real wages being more stable than money wages. But this could only
be the case if there were a presumption in favour of stability of
employment . . . thus the attribution of relative stability to real
wages is not merely a mistake in fact and experience. It is also a
mistake in logic, if we are supposing that the system in view is stable,
21
22

Ibid., p. 264.
"Empirical research has confirmed that wage adjustment is slow in
depressions and has also shown the 'real balance effect' to be small." H.
G. Johnson, "The General Theory after 25 years". A. E. R. Supp.
J96J, p. J3

192

D. C. Champernowne

in the sense that small changes in the propensity to consume and the
inducement to invest do not produce violent effects on prices."23
To substantiate this criticism of Pigou, it would be necessary to
demonstrate that money wage rates had in fact shown greater stability than real wage rates in most industrial economies. This would be
difficult since although neither real nor money wage rate statistics
normally show substantial falls, money wage rates do, of course, show
far greater rises than do real wage rates, thereby reflecting the
tendency for money wages and prices to rise together in many
periods.
The substantiation of Keynes' criticism of Pigou would also require a far more careful demonstration that from the assumption that
money wage rates are less stable than real wage rates in an upward
(although not necessarily in a downward) direction, it can really be
proved that initially small changes will produce violent effects on
money wages, prices, or employment. Such a logical demonstration
would have to consider more closely than did Keynes in the General
Theory the reactions of the monetary authorities and the effects of
shifts in the distribution of income between high-savers and lowsavers which would follow from the sequence of changes. It is hard to
see how one could demonstrate more than that a gentle rise of prices
and money wage rates might persist for several years if money wage
rates were flexible in an upward direction while real wages were
somewhat sticky in both the upward and downward directions.
We shall return to a further examination of Keynes' view about
the stickiness of money wages in connection with his discussion on
the peculiar properties of money.

sa. Effects of the Present on Expectations


about the Future.
The reader who has persevered this far is likely to have been struck
by the fact that the links between the present and the future seem to
transmit effects predominantly in one direction only. For although
Keynes has so much to say about the effects of expectations about
the future on present economic behavior, he seems to be not nearly
so informative about the causation of these expectations.

The Economic Future and the Present . 193


Keynes pays considerable attention, for example, to the effects
that may follow from abrupt changes in (1) expectations about future movements in interest rates; and (2) the expectations determining the schedule of the marginal efficiency of capital. But in each
case he regards the formation of these expectations as being largely a
matter of psychology and convention. Instead, therefore, of setting up
a detailed theory of the causation of these expectations, he prefers to
discuss the conventions and habits of thought which commonly
govern the formation of these expectations.
For example, he argues that the accepted convention is to expect
the long-term rate of interest to oscillate around the level at which it
has been in the recent past unless one has reason to the contrary.
Since experience confirms that the monetary authorities have some
control over the long-term rate of interest, the main reasons for expecting a change in it are any scraps of evidence concerning the intentions of the monetary authority to change the rate. The other
important factor is the belief based on history that it is normal for
the long-term rate of interest to oscillate only within certain fairly
definite limits.
This is the theory that has led Professor Hicks to say that the rate
of interest is left hanging by its own boot straps. It is not an altogether apt metaphor: the long-term rate of interest, according to
Keynes, will move to where people expect it to move; and provided
the monetary authorities indicate their intentions sufficiently firmly
and back them up with vigorous operations, people will within wide
limits expect their intentions to be achieved and then in fact they
will be achieved. Keynes goes on, however, to give examples of the
limitations to people's credulity in the power of the authorities to
implement their rate-of-interest policy.
In diagnosing the causes of heavy unemployment, Keynes laid
much of the responsibility on the inefficiency with which those
opinions and expectations are formed, which determine the schedule
of the marginal efficiency of capital. These opinions and expectations are concerned with future sales and level of profits, and Keynes
complained that few business men made any serious attempts to
predict these at all far ahead, and that they based such forecasts as
they did make merely on such conventional assumptions as: (1) that
the latest estimate made holds good unless there have been any new

194 . D. C. Champernowne
facts pointing unambiguously to an upward or downward revision;
(2) that where, as is usual, the evidence about the future is scanty,
the most recent experience is the best guide as to what future experience will be.
He concludes that short-period fluctuations in current profitability
are likely to cause unwarrantably violent swings in the marginal
efficiency of capital.
Having emphasized the importance which mere convention must
play in the formation of economic expectation, Keynes characteristically toys with the next stage of sophistication by fancying that at
times the forecasts themselves may well be influenced by speculation
about possible changes in those very conventions which are commonly used in the formation of economic expectations.
6. The Properties Peculiar to Money.
Chapter 17, entitled "The Essential Properties of Interest and
Money," has greater theoretical depth and is more perplexing than
any other in the book. In this paper only the treatment of expectations in this chapter will be analyzed.
The central question of the chapter is: "Why should the volume
of output and employment be more intimately bound up with the
money-rate of interest than with the wheat-rate of interest or the
house-rate of interest?"24
Keynes' answer is given in the following stages, not, however,
necessarily in the exact order in which they will be set down here.
First, he explains what is meant by the wheat-rate of interest, the
house-rate of interest, and so on, and what are the rules governing
the relation between these own-rates of interest of the various assets.
Second, he picks out three characteristics of money-zero (negligible) elasticities of production and substitution, and low carrying costs
-as reasons why it is the rate of interest of money that rules the
roost; he briefly considers other reasons why measures to reduce the
money rate of interest may fail. Third, he discusses why wages are
likely to be, and to be expected to be, more stable in terms of money
rather than of other commodities, and argues that this strengthens the
reason why the money rate of interest shbuld cause all the trouble.
24

Ibid., p. 225.

The Economic Future and the Present . 195


Let us note the points concerning expectations ~s they rise under
these three headings:
I. CONCERNING OWN-RATES OF INTEREST. Keynes is discussing
equilibrium conditions under which for each asset worth producing,
whatever the standard of measurement, the rate of interest will equal
the marginal efficiency of that asset. He therefore allows himself to
use the term "the rate of interest of the commodity" both to mean its
own-rate of interest and to mean its marginal efficiency in terms of
itself as standard. But to avoid any possible confusion we shall use
two distinct terms: by the rate of interest (m.e.) of an asset we shall
mean its marginal efficiency measured in terms of itself as standard.
Keynes defines the (own) rate of interest of an asset to be the
extra fraction of a unit of that asset which, by putting down one unit
now, one would expect to get back with that unit in a year's time.
He explains that the wheat-rate of interest, for example, is less than
the money rate of interest by the percentage excess of the future
price of wheat in twelve months' time over the spot price. The whole
relation between rates of interest of different commodities is thus
linked to the relation between present prices and expected future
prices. In the absence of expectation of price movements all rates of
interest (m.e.) would have to equal the rate of interest (m.e.) of
money; but the possibility of expected price movements complicates
the argument. Investment in a capital asset can continue at a rate
higher than that required merely to equate its rate of interest (m.e.)
with that of money, if its own money price is expected to rise.
Keynes even mentions the theoretical possibility that rates of interest
(m.e.) of other commodities may continuously fall faster than the
money rate of interest (m.e.) if "the present money price of every
commodity other than money tends to fall relatively to its future
expected price." But this last possibility is only mentioned to show
how remote from reality it is.
2. CONCERNING THE ZERO ELASTICITY OF PRODUCTION AND SUB-

In connection
with the very low elasticity of supply of money, Keynes explains that
in the case of an ordinary commodity, if its marginal efficiency schedule is suddenly raised for some reason, it does not follow that its demand price must rise in full proportion since "the elastiCity of its supply would also tend to prevent a high premium on spot over forward
STITUTION AND THE LOW CARRYING COSTS OF MONEY.

196 . D. G. Champernowne
delivery."25 In other words, the spot price by rising only moderately
would become sufficiently far above the expected future price to
raise the rate of interest of the commodity in line with its rate of
interest (m.e.). But in the case of money, if its purchasing power
increases this will not cause an expectation of the mining of money
by private enterprise on a scale which will quickly bring down the
purchasing power of money to its former level again.
Suppose, on the contrary, that the only form of money were silver
coins; that the authorities bought up at a fixed money price all the
silver produced, and minted it and put it straight into circulation.
Then, provided the production of silver was highly price-elastic, even
a slight fall in money wages could quickly restore a tendency towards
full employment. The way this could be brought about is that a fall
in money wages would make profitable a considerable increase in
the scale of silver mining and this would lead to more money being
minted and to the expectation of even more being minted in the
future. Lower rates of interest would result and all investment
would be intensified so that employment would increase until it
became full, at which point money wages would recover to their old
level and the original position of full employment equilibrium would
be restored. It would mean that the prices of other assets, having
quickly fallen with money wages, would be expected to rise again so
that their rates of interest (m.e.) could immediately fall and in advance of the fall of the money rate of interest.
In saying that money has a low elasticity of substitution Keynes
refers to the fact that the services it provides expand along with its
own purchasing power. Such a rise in purchasing power therefore
does not check off the desire to hold on to money to the same extent
as would a rise in the price of some other asset check off the desire
to retain that asset. The only point concerning expectations here is
the obvious one that if there is any uncertainty about the permanence
of the increase in purchasing power, then that will to some extent
check off the demand for holding money.
The low elasticity of substitution of money is given by Keynes as
one reason why large increases in the purchasing power of the money
stock may cause only a moderate reduction in the money rate of
interest.
25

Ibid., p. 235.

The Economic Future and the Present 197


Expectations are again involved in the point that any large increase in the quantity of money may create so much doubt about the
power of the authorities to continue their attempts to hold down the
rate of interest, that people's precautionary motives for holding money
will actually be strengthened and the rate of interest may begin
actually to rise again, because the quantity of money has had to be
increased so much.
We have already noted Keynes' argument that, if it is generally
believed that the long-term rate of interest never falls for long below,
say, 2 per cent, then it can never fall below that level for the reason
that nobody would want to hold securities at a price which was
generally expected to fall.
The third peculiarity of money is its low carrying cost; we shall
examine below Keynes' view that this property of money is partly responsible for people's expectation that money wages will be sticky.
3. CONCERNING THE STICKINESS OF MONEY WAGES. We have seen
what great importance Keynes attached to the point that any fall in
money wages was likely to cause not merely a corresponding fall in
money prices but also the expectation of further falls in both money
wages and money prices. This expectation would lower the schedules
of the marginal efficiencies of capital assets (measured with money
as standard); this would probably more than outweigh the good
effects on investment of any lowering of the rate of interest (m.e.) of
money due to the release of money from transactions duties to satisfy
liquidity demands, and any stimulus to the propensity to consume,
which might result from the lowering of money wages and prices.
For these reasons Keynes argues that any pronounced tendency for
money wages to fall in times of high unemployment and to rise in
times of full employment would cause the economy to be highly
unstable.
One reason that it is so difficult to lower the money rate of interest
is that money is by far the most liquid of all assets. Keynes argues
that money thus attracts so high a liquidity premium largely because
it is a standard in which the future cost of living is expected to be
relatively stable. He urged that this expectation of relative stability
in the future money cost of output might not be entertained with
much confidence if the standard of value were a commodity with a
high elasticity of production.
But it is difficult to see why the mere fact that the standard of

198 . D. G. Champernowne
value is a commodity with a low elasticity of production should lead
one to expect the money cost of wage goods, i.e., the cost of living, to
be stable. The low elasticity of production of the money commodity
would allow large fluctuations in its marginal costs in terms of wage
units, whereas the higher elasticities of production of wage goods
would not allow such large fluctuations in their marginal costs in
terms of wage units. If in fact the fluctuations in the marginal cost
of the money commodity were large in terms of wage units whereas
those in the marginal costs of wage goods were not, there would be
large fluctuations in money wage rates and hence considerable fluctuations in the marginal costs of wage goods in terms of money.
There would then be some grounds for expecting that there will be
considerable fluctuations in the money price of wage goods and the
cost of living because the money commodity has a low elasticity of
production.
In order to reach the opposite conclusion, namely, that the low
elasticity of production of the money commodity would lead one to
expect a stable cost of living, Keynes must have relied heavily on his
assumption that money wage rates are always sticky. He believed
that "the commodity, in terms of which wages are expected to be
most sticky, cannot be one whose elasticity of production is not least,
and for which the excess of carrying-costs over liquidity-premium is
not least."26 The reasons adduced for this are very difficult to discover but seem to be the following. If wages were sticky in terms of,
let us say, coal, then irrespective of the money price of coal, there
would be some unique scale of output of coal, which would equate
its marginal cost (which would be labor costs) with its price; but
there is no reason that this scale of output should be the same as
that which could be sold at that price. Hence stocks of coal would
either pile up or dwindle away until something happened to break
the link between coal prices and wages. For example, the high carrying costs of coal would force down its price when stocks became excessive, or a shortage of coal stocks would force up coal prices.
Whether this is a correct interpretation of Keynes' reasoning is
very uncertain, but it is in any case difficult to feel that Keynes really
established the case for supposing that wages would necessarily be
26

Ibid., p. 23 8.

The Economic Future and the Present . 199


and moreover be expected to be more sticky in terms of money than
in terms of other assets.
It is true that if slight unemployment causes money wages to fall
whereas full employment leads them to rise, then instability is likely
to follow. But suppose that experience leads to a constant expectation of a gradual drift upwards of money wages and money prices.
This need not lead to instability and may on the contrary allow the
commodity rates of interest (m.e.) to fall very low even when the
money rate of interest falls only, let us say, towards 3 per cent. It
does indeed suggest a way out of the difficulties presented by
people's tendency to attach an excessive liquidity preference to the
holding of money: if the wage earners are sufficiently obstinate in a
closed system in steadily pushing up money wages at, say, 3 per cent
per annum, this will enable the rate of interest (m.e.) of capital
assets to fall to zero even if the money rate of interest falls only to
3 per cent. Admittedly, in a country like the United Kingdom, which
relies on selling exports at competitive prices, this policy suffers from
certain disadvantages unless competing countries also adopt it: but
for a closed system such as the world as a whole it would seem to
offer an easier alternative than that of lowering the money rate of
interest towards zero.
It is strange that Keynes did not himself make this point more
clearly since in several passages he comes close to agreeing with it.
For example he sympathizes with "those reformers, who look for a
remedy by creating artificial carrying-costs of money through the device of requiring legal-tender currency to be periodically stamped at
a prescribed cost in order to retain its quality as money."27
He goes so far as to state that "the expectation of a fall in the value
of money stimulates investment, and hence employment generally,
because it raises the schedule of the marginal efficiency of capital"28
and agrees that this is the truth underlying Professor Irving Fisher's
theory of appreciation and interest and his distinction between the
money rate of interest and the real rate of interest. But Keynes follows this up with a page of criticism29 which suggest that he completely failed to appreciate the truth of what Professor Fisher had
Ibid., p. 234.
Ibid., p. 141.
29 Ibid., pp. 14 2 -3.
27
28

200

D. G. Champernowne

in fact said, and he then suggests a "re-writing" of Fisher's theory


which is quite nonsensical.
At the close of this section it is perhaps pertinent to suggest one
other essential characteristic of money, not specifically mentioned by
Keynes, which is yet responsible for many of the economic difficulties
of capitalism. Keynes argues that much of the trouble-causing instability of the demand for money is due to the strength of the
precautionary motives for holding money when uncertainty is prevalent. Now, a main reason that underlies these precautionary motives
is that people want to be reasonably certain about the future purchasing power of their store of wealth measured in money. If they were
concerned with something else (say, its purchasing power in terms
of sugar or houses or its income-yielding power) then general uncertainty would not point so strongly in the direction of holding
money rather than fixed interest securities or capital assets. So one
may add to the essential characteristics of money that it is in terms
of money that people are anxious to be able to foresee the value of
their asset holdings without too much uncertainty.
It seems that Keynes' ideas about the peculiar properties must
have developed further after he had completed the chapter on the
properties of interest and money. For it is perhaps significant that
later in the book Keynes attributes the importance of money to yet
another of its qualities: and he does so repeatedly, and with a
vehemence that transcends anything in the earlier chapter. He has
argued that the division of economics between the Theory of Value
and Distribution, on the one hand, and the Theory of Money, on the
other, is a false division. He suggests that the right division is that
"between the Theory of the Individual Industry or Firm and of the
rewards and the distribution between different uses of a given quantity of resources on the one hand, and the Theory of Output and
Employment as a whole on the other hand."
He then suggests alternatively that the division should be between
the theory of stationary equilibrium and the theory of shifting
equilibrium where changing views about the future are capable of
influencing the present situation. Then three times in slightly different words he stresses the following peculiarity of money. "For the
importance of money essentially flows from its being a link between
the present and the future" ... "we can pass ... to the problems of

The Economic Future and the Present 20 r


the real world in which our previous expectations are liable to disappointment and expectations concerning the future affect what we do
today. It is when we have made this transition that the peculiar
properties of money as a link between the present and the future
must enter into our calculations."3o "Money in its significant attributes is, above all, a subtle device for linking the present to the
future."3l

CONCLUSION
At the end of this perusal of Keynes' various references to these
links between the economic future and the present, one is left in
some perplexity why he should have applied it to three such diverse
concepts as capital equipment, user cost, and money. To the latter
question it is indeed difficult to supply an adequate answer. But to
the question why he kept referring to the links between the economic
future and the present, the answer must be that he considered that in
our society some of these links were defective.
Keynes made it abundantly clear that there was no efficient link
between the decisions of employers how much to invest, and the
decisions of individuals how to divide their incomes between consumption and savings. He expressed his opinion that we should see
the state eventually providing this link in order to prevent heavy
unemployment. But although he did not give such a clear demonstration, he was greatly concerned also with the lack of any proper link
between employers' plans regarding capacity for production of consumption goods in the future and the future decisions of individuals
to spend on consumption goods. He realized that the lack of such a
link would lead to violent swings in expectations and activity.
Economists such as Professor Irving Fisher had written as though
each individual's decision to save now carried with it the decision
when instead to spend the money on consumption goods. Other
economists wrote as though the determination of interest rates and
of prices throughout the future was a simple matter of supply and
demand extended from present markets to a scheme of futuresmarkets covering the years ahead. But this was completely misleadIbid., p. 293. Keynes' italics. I am sorry to find that in 1936 I added
at this point in my copy the marginal note "cranky."
31 Ibid., p. 294.
80

202

D. C. Champernowne

ing. For such futures-markets exist only in rare cases and individuals
do not commit themselves years in advance in respect of their expenditure. Consequently employers have to plan in a thick fog,
enshrouding not only what each other's plans may be, but every
other aspect of the likely state of demand and supply in the distant
future.
It was because of his perception of the inefficiency of this feature
of our method of organizing production that Keynes kept returning
to the theme of the links between the economic future and the
present.

ABBA P. LERNER

The General Theory


[1936J

HE OBJECT OF THIS ARTICLE is to provide as simple as


possible an account of the most important line of argument that runs
through J. M. Keynes' book The General Theory of Employment,
Interest, and Money, so that, except perhaps in some details of
presentation, it contains nothing original. I have endeavored, where
possible, to follow the traditional use of language more closely than
Keynes does, as I have found that this renders the argument both
more intelligible and more acceptable to those who are not familiar
with the oral tradition of Cambridge. While necessarily simplifying
the argument considerably in order to be able to encompass it in an
article of appropriate length, I do not think I have left out anything
fundamental. In discovering what are the points in the argument
or its presentation at which students are liable to jib, I have learned
much from innumerable discussions with economists and students
in London, Cambridge, and Geneva, and of these certainly the most
helpful was Dr. Gottfried Haberler, who has been working towards
similar results along a quite different route. I must add that I would
certainly not have been able to attempt this task were it not for the
time I spent in Cambridge in 1934-5 while Leon Fellow of the
University of London.

Keynes wishes sharply to distinguish his own system from what


he calls the "classical" economics. By that he means the orthodox
body of doctrine, first conceived in fairly complete outline by Ricardo,
and developed by almost all economists of repute from that time
on, both in England and elsewhere, which finds its present culmination in the works of Pigou. Keynes is so keen on making clear the
difference between the classical and his own scheme that he perhaps overemphasizes it, willingly taking this risk in order to be
certain of avoiding the other error, which he considers more danger20 3

204

Abba P. Lerner

ous, of permitting a reader to overlook the revolutionary nature of


the change. He has no patience whatever for the interpreter who
would try to read Keynes' views into the classical writers. This is
not at all-as is frequently suggested-because that would diminish
his claims to originality (in fact I believe he is overgenerous in his
estimate of how near the Mercantilists and the Monetary Cranks
were to his thesis) but because he is convinced that such identincation is made plausible only by obscurity. Keynes therefore complains
that: "Those who are sufficiently steeped in the old point of view
simply cannot believe that I am asking them to step into a new pair
of trousers, and will insist on regarding it as nothing but an embroidered version of the old pair that they have been wearing for
years."l
I have insisted at some length on this because it helps to explain
the extraordinary psychological resistance to Keynes' new argument
that is always displayed by classical economists. (I shall use this
word throughout in Keynes' sense.)
It is this psychological resistance that so frequently leads people
to reject a proposition of Keynes' as a paradox and then to tum uneasily and, almost in the same breath, to scorn it as a platitude. "It's
absolutely wrong"-"we all knew that before!"
The last sentence in Keynes' preface reads: "The difficulty lies,
not in the new ideas, but in escaping from the old ones, which
ramify, for those brought up as most of us have been, into every
comer of our minds." I would like to underline that sentence.
Keynes is concerned with the problem of unemployment. The
classical view is that, in the absence of state interference or other
rigidities, the existence of any unemployment will have the effect
of lowering wages.
This follows immediately from the dennition of unemployment,
for any man who is not in employment but who does not try to
get work at a lower wage is no more considered to be unemployed
than the man who refuses to work overtime or on Sundays. At the
current wage he prefers leisure to employment. He may be idle
but he is not unemployed-at any rate he is not involuntarily unem1 EeN,

Nov. 1931, p. 39 0

The General Theory .

205

ployed. If he really wanted to work, if he were really unemployed,


he would offer himself at a lower wage and this would reduce the
level of wages. Unemployment is incompatible with equilibrium.
The reduction of wages, the argument goes on, will make industrial activity more profitable so that business men will employ
more people. As long as there is any unemployment, wages will
fall; and as long as wages fall, profits rise; and as profits rise, employment increases until all the unemployed are absorbed in industry, and we have equilibrium and no more unemployment.
Unemployment can therefore persist only if the state, or the trade
unions, or some other institution prevents the unemployed from
offering their services at lower wages and so from setting in motion
the automatic mechanism which leads to equilibrium and full employment. What is necessary, therefore, is simply to remove the
rigidity and allow the unemployment to liquidate itself by reducing
wages.
Keynes accepts neither the definition nor the argument. Like
the classical economists, he is concerned only with involuntary
unemployment, but he defines as involuntarily unemployed a man
who would be willing to work at a lower real wage than the current real wage, whether or not he is willing to accept a lower money
wage. If a man is not willing to accept a lower real wage, then he is
voluntarily unemployed, and Keynes does not worry about him
at all. But there are millions of people who on Keynes' definition
are unemployed but who fall outside of the classical definition of
unemployed, and these provide one of the most pressing of modem
social problems. These are willing to work for less than the current
real wage-they would be willing to work for the current money
wage even if the cost of living were to go up a little-yet they cannot find jobs. What determines the number of people in a society
who find themselves in this position? Or, to put the question the
other way round, what determines the number of people who do
find employment? The object of Keynes' book is to indicate the
road leading to the answer to this question.
The classical refusal to consider these men as really involuntarily
unemployed resolves itself into a recipe for finding them employment. They have only to agree to accept lower wages and they will
find work. Keynes objects to this procedure of economists on two

206

Abba P. Lerner

separate grounds. His first objection is on the practical ground of


the uselessness of tendering advice that one knows will not be accepted, even if it is sound advice. It is time for economists who wish
to give statesmen practical advice to realize that money wages are
sticky-that workers will, in fact, refuse to reduce money wages.
But Keynes' main objection consists of a denial of the theory which
is put forward as an excuse for the treatment. If money wages are
reduced, it does not follow that there will be any increase in employment. A general reduction of wages will reduce marginal costs,
and competition between producers will reduce prices of products.
Equilibrium will be reached only when prices have fallen as much
as wages, and it will not pay to employ more men than in the beginning. The workers, who are able to make agreements with their
employers about their money wage, cannot adjust their real wage.
If they could reduce their real wage, more would be employed; but
they can only attempt to reduce their real wage by reducing their
money wage at the existing price level. This, however, only brings
about a proportionate fall in prices so that they are in fact not able
to vary their real wage. That is why their unemployment is involuntary even if they refuse to accept a lower money wage. For that
would not have the desired effect of reducing the real wage and
increasing employment-it would merely remove a certain stability
of prices.
It has hardly been disputed that a cut in money wages, by reducing costs, will have some tendency to reduce prices, but it remains
to be shown why prices should fall prpportionately to the reduction
in money wages so that there is no fall in the real wage and so no
increase in employment in the manufacture of consumption goods.
(Employment in investment industry depends on other factors considered below. For the time being this is taken as given.)
Whether this will be the case or not cannot be decided at all by
looking merely at the effect of the wage cut upon costs. It is necessary also to consider the effect of the wage cut upon demand; whether
directly or whether indirectly through the change in employment
that might be initiated by the first impact of the wage cut. Until
we bring this into the picture, we have not sufficient data to be able
to decide what the result must be.
This has made it possible for one eminent economist to argue that

The General Theory .

207

a cut in money wages will increase employment, and for another


eminent economist to argue that a cut in money wages will not increase employment. The first is able to show that his thesis is
consistent with the cost conditions; for with a larger volume of
employment-with more labor applied to the given productive equipment of society-the marginal productivity of labor is less, marginal
costs are higher relatively to wages; prices (which, with the same
degree of imperfection of competition, must in equilibrium bear
the same ratio to marginal costs) are also higher relatively to wages,
so that the workers by cutting their money wages have been successful in reducing their real wages. The second is also able to show
that his conclusions are consistent with the cost conditions; for if
there is no increase in employment, marginal costs will fall as much
as wages, and prices have to fall in the same proportion as costs, so
that there is no change in real wages. Further, each economist is able
to accuse the other of assuming his conclusions, and then each can
complain of the pot calling the kettle black. So that we have an
infinite regress but no answer to our question.
The necessity of bringing in the demand side is seen even more
clearly if we suppose for a moment that wages are the only item
that enters into marginal costs and that marginal costs are constant.
In this case there is no inverse relation between employment and
real wages. If wages are cut, marginal costs fall in the same proportion as wages whether there is an increase in employment or not.
There will be no fall in real wages, but that tells us nothing about
the volume of employment. To get the answer to our question, we
have to consider the effects of the wage cut on demand, direct as well
as indirect.
The essence of the analysis whereby Keynes obtains the result
that there will be no change in employment comes from a consideration of demand conditions. If there is initially an increase in employment-and, since employers very often think that a wage cut
is a good thing, this impact effect is very likely-the demand conditions will be such as to bring about losses which tend to induce
the entrepreneurs to curtail employment until the previous equilibrium level of employment is restored. Similarly, if the impact effect
is to reduce employment, this will bring about profits which induce
entrepreneurs to raise employment to the previous level.

208

Abba P. Lerner

The losses that accompany an increase in employment in the


manufacture of consumption goods are due to the tendency of people whose income is increased to increase their expenditure by less
than the increase in the outlay on their production, so that there
emerges a net loss. This loss may be mitigated, but not entirely escaped, by the withholding of stocks with the intention of selling
them at a more propitious moment; but this procedure, while
diminishing losses, has the effect of building up superfluous stocks.
The losses and the accumulation of stocks both tend to reduce employment, and these forces must persist and accumulate as long as
employment remains above the equilibrium level. The whole of this
phenomenon is reversed for the case where the initial effect of the
wage cut is to diminish employment.
We must now consider how all this works if items other than
wages enter into marginal costs. Where this is the case, these other
items are payments for the use of productive resources which, in
the short period, are nxed in supply. This is because they accept
whatever they can get, their reward falling relatively to wages until
all those that are of any use whatever are employed.
If, then, wages are reduced, the attempt to substitute labor for
those other productive resources will increase employment and may
reduce the earnings of these resources. As long as these earnings
have not fallen in the same proportion as wages, costs and prices
will not have fallen as much as wages but will have fallen more
than the rewards of the other productive resources. Real wages will
be lower while the real reward to the other productive factors will be
greater. More men will be employed, and the total real income
will be greater; since, with more men employed on the given resources, a greater real product is forthcoming. The aggregate real
income of the other productive resources is increased, since the
quantity employed is unchanged and the real rate of reward is increased. The aggregate real income of labor may be greater or less
than in the beginning, according as the increase in employment is
greater or less than the reduction in the real wage.
As long as this situation remains, prices have not fallen as much
as wages have been reduced; and the workers have been able to
reduce their real wages by reducing their money wages and thus
to increase employment. Such a position cannot be expected to
persist, but contains within itself forces which will still further re-

The General Theory

209

duce the rewards of the factors other than labor until costs and
prices have fallen proportionately to wages, and real wages and
employment are back again at the original level.
In the situation we have just described, total real income is greater
than in the initial position, because more men applied to the same
equipment produce more goods. There is an increase in the total real
costs of the consumption entrepreneurs exactly equal to this increase
in real income (since the incomes of the factors of production are
the costs of the entrepreneurs). Out of this extra income, some will
be saved, so that the total receipts of consumption entrepreneurs
increase (in real terms) less than their outgoings. Entrepreneurs
make losses which cause them to restrict their (output and) demand for productive resources. This goes on as long as more men are
employed than in the initial equilibrium and as long as the real
reward of the productive resources other than labor is greater than
in the initial position. These two phenomena disappear at the same
time, since the tendency to substitute labor for other productive
resources, which led to the increase in employment in the first place,
disappears just at the point where the real reward to the other productive factors has fallen in the same proportion as prices and wages.
A new equilibrium is reached only when employment has gone
back to its original level and the reward of the other resources has
fallen to its old real level. This will only be when their prices have
fallen in the same proportion as wages. As long as these have fallen
only in a smaller proportion than wages, prices will be higher than
before relatively to wages and lower than before relatively to the
reward of the other productive resources, and the disequilibrium
described will continue.
In a longer period, it will be possible to increase or decrease the
supply of productive resources other than labor by varying the application of current factors of production to their manufacture, so
that the above argument, which rests on the fixity of supply of
productive resources other than labor, would not apply. But there
will be no inducement to vary their supply since their price, determined in the longer period by their cost of production, will have
varied in just the same proportion as wages. There is therefore no
point in departing-except as a temporary mistake-from the initial
level of employment.
This does not mean that a reduction of money wages may not

210

Abba P. Lerner

have all sorts of indirect influences which ultimately react on the


level of employment. There will be effects on the demand for money,
on the rate of interest, on entrepreneurs' expectations of future
prices (or rather of the relation of these future prices to present
costs), on the distribution of wealth and spending-all these and
other influences will have an effect on the number of people that
entrepreneurs consider it profitable to employ-but these work in
divergent directions and some of them only after a considerable
interval, so that nothing can be said as to the effect of the sum of
these influences on employment as a result of a reduction in wages
until a complete set of assumptions have been provided as to the
form and strength of these influences. Before we have all this information, we must either assume them to cancel out and say that
there is no effect on employment, or else, if we wish to be more
realistic, we must say that what happens to employment if money
wages are reduced will depend upon other conditions, so that employment might go either up or down. Anything might happen.
There is no simple rule such as the classical economists envisage
relating the level of employment to the money wage.
2

If the level of employment is not affected in any simple way by


the money wage, what is it that does determine the amount of employment? Before answering this question it is useful to contemplate
some very simple equations.
The income of the whole society is earned by the members of
the society in producing either consumption goods or other kinds
of goods. We call these other goods investment goods. This gives
us our first equation. The total income of society (Y) is made up of
the income earned in making consumption goods (C) and the income earned in making investment goods (1). Y = C+I.
Now C, which stands for income earned in making consumption
goods, must also stand for the amount spent on buying consumption goods, since these two are in fact the same thing. (Similarly I
stands also for the amount of money spent on investment goods.)
The aggregate amount of saving in any period (S) is defined as the
excess of aggregate income in the period over the expenditure on
consumption goods. This, the almost universal definition of saving,
gives us our second equation S = Y-C (definition).

The General Theory .

2I I

From these two equations it follows that saving must always be


equal to investment. S = I.
This appears rather peculiar to many people when they first meet
it, since there is obviously no mechanism whereby any individual's
decision to save causes somebody to invest an exactly equal amount.
Mr. Keynes has ineradicably impressed that upon the mind of everyone who has read his Treatise on Money. And of course Keynes was
right in this. Yet there is no paradox.
It is perfectly possible for any individual to save more without
investing more himself. The proposition applies only to aggregate
saving and investment. Neither is it necessary that aggregate investment should increase whenever any individual decides to increase
the amount that he saves. This would be so if an increase in an
individual's saving left unchanged the amount saved by all other
individuals together, so that it always meant an increase in aggregate saving. But we cannot assume that, because the individual
must decrease his expenditure on consumption goods to the extent
that he increases his saving. This diminution in C (if others have
not changed their expenditure on consumption goods) diminishes Y
(by diminishing the income of those who sell consumption goods)
and therefore leaves (Y-C), which by definition is S, the same as
before. Others have saved as much less as he has saved more, so that
aggregate saving is unchanged and equal to the unchanged I. If
there is no change in I, there can be no change in S.
Individuals deciding how much to spend out of their incomes
seem to be able to decide how much to save, and, if we consider
one individual in a large society, this ha~ sense, because the effect
on his own income of an individual's expenditure on consumption
goods can be neglected. But if we take society altogether and neglect
the effect of changes in expenditure on total incomes, we naturally
get into trouble, for we are then making the contradictory assumptions (a) that when people save more they spend less on consumption goods and (b) that the people who sell consumption goods do
not receive any less. And nobody expects to get sensible results by
deduction from contradictory assumptions, not even those who are
most scornful of the canons of "bourgeois" logic.
The classical view that an individual, in deciding to save more,
increases the aggregate amount of saving (S), can be supported
by another argument which does not, at first sight, appear to be

212

Abba P. Lerner

quite as illogical as that just given. We must leave this, however,


until we have examined the classical theory of the determination of
saving and investment and the rate of interest.
A more common sense objection to the proposition that saving
and investment must always and inevitably be equal to each other
is to be found in the query whether the identity of these two cannot be upset by hoarding. In the case of any individual it is clear
that there is no need for his saving to be equal to his investment.
When an individual saves more than he invests he is said to hoard
the difference. Why cannot society do the same? And if society
hoards (or dishoards) will that not make saving greater (or less)
than investment?
We must note more carefully what is meant by hoarding. Our
individual who invested only a part of his saving was left with the
difference in cash. His store of money has increased and it is in fact
this increase in his store of money that is his hoarding. Any individual who saves more than he invests in any period increases his
holding of money by the difference. Any individual who increases
his store of money in any period must have saved more than he invested in the period by just that amount.
The question "Can the society hoard?" means, then, nothing else
than "Can the society increase its store of money?" This will depend
upon whether or not the monetary authority has increased the amount
of money in the society during the period we have been considering.
If the monetary authority does not increase the amount of money,
it is impossible for the society to hoard. If any individual hoards,
other individuals have to dishoard to the same extent, for it is impossible for anybody to increase his store of money without somebody else diminishing his store of money as long as the total store
is unchanged. There cannot therefore be any net hoarding (or dishoarding) by all the members of the society taken together, so that
there cannot for the society be any excess of saving over investment
(or of investment over saving). S=1.
If the monetary authority does increase the amount of money,
then there not merely can be net hoarding by the whole society,
but there must be net hoarding exactly equal to the increase in the
society's holdings of money. This does not mean that there is any
divergence between saving and investment. There is indeed an ex-

The General Theory .

2I3

cess of saving over investment by the individuals who are left with
the extra money that has been put into the society and which must
be in somebody's hands. But this is exactly balanced by the expenditure of money by those individuals who borrowed the extra
money from the monetary authority (the banks). These borrowers
were enabled by the banks to consume or to invest out of borrowed
money that was not part of their income. Insofar as they spend the
money on consumption, this constituted negative saving which has
to be subtracted from the excess saving by the hoarders. The rest
of the borrowed money is invested and provides the investment that
balances the excess saving and shows again the inevitable equality
of saving to investment. We always get back to this really very obvious if not very informative bit of arithmetic. It only appears strange
or suspicious because of the habit of looking at the saving from the
point of view of the individual who has got his income and is wondering whether to save it or not. He is naturally unable to see the
whole social process. Our suspicions should vanish when we realize
that all that the proposition says is that the excess of total income
over income earned in making consumption goods is equal to the
income earned in other ways.
What we have done now is to replace the suspect proposition
that S = I by the even more suspect proposition that it is impossible
for a society to hoard if the banks do not increase the amount of
money. Does this not imply that everything that has been said in
economic discussions about the effects of hoarding is sheer nonsense?
This is, of course, not the case. The trouble arises from a confusion of two meanings of hoarding. When people consider, say,
the deflationary effects of hoarding, they are talking sound and important sense. But if they are to use the word hoarding in the sense
we have used it, so that it indicates an excess of saving over investment, they should speak of the deflationary effects of "attempts to
hoard." These effects are of the utmost importance. They involve
a reduction of prices, of profits, of employment, of incomes, of prosperity generally, and of many concomitants of these. But they do
not involve an increase in hoarding-in our exact sense of increasing the money held-unless the amount of money is increased. It
is only saying the same thing in other words to show that an attempt
by people to save more than they invest will diminish consumption

214

Abba P. Lerner

and incomes and employment, etc., but will never succeed in making
saving greater than investment.
We see then that decisions of income receivers as between spending and saving do not affect the aggregate volume of saving but do
determine the size of both income and consumption. The difference
between them, which is the amount actually saved, is determined
by those who decide the size of I (which is equal to the excess of
income over consumption, because it is that part of income which
is not earned in making consumption goods).
If we have given the size of I, we can say that Y is determined
by the propensity to save. If we suppose that the amount people
save depends only on the size of their income, and that it increases
with the size of income, we can see that income must be at that level
where the amount people wish to save is equal to I.
As long as income is below this level people will wish to save
less than is being invested, i.e., they will want to spend on consumption goods more than is being earned in making consumption goods,
and since these two are identical this means that they will wish
to spend on consumption goods more than they are spending on
consumption goods. This will lead to increased demand and profits
in the manufacture of consumption goods, which will lead to an
expansion of employment and income until this level is reached.
People then wish to save just as much as is being invested, i.e., they
spend on consumption goods an amount that is less than their income by exactly the expenditure on ( = the earnings in the manufacture of) investment goods, i.e., they spend on consumption goods
just as much as is earned in making consumption goods, i.e., just as
much as the cost incurred in making consumption goods. There is
neither profit nor loss but equilibrium. If employment and income
had risen above the level where people wish to save just as much
as is being invested, losses would have emerged to bring incomes
and employment down again to the equilibrium level where people
wish to save just as much as is being invested. S = I. Although there
is no mechanism whereby decisions about saving bring about an
equal value of investment, which is what makes the equation suspicious, because of the long-standing habit of expecting the influences
to work from saving to investment, there is a mechanism whereby

The General Theory .

2I

decisions to invest bring about an equal amount of saving, which is


what makes the equation true. 1= S.
From the expenditure on consumption at this level of income,
we can derive the number of men employed in making consumption goods-for there is a functional relation between this number
of men and expenditure on their product. Similarly, from the expenditure on investment goods, we can derive the number of people
at work in making the investment goods. This gives us the total
number of men employed. This number is determined by the amount
of investment and the propensity to save (or its complement: the
propensity to consume, which is the relationship between income
and consumption). The propensity to consume may also depend
upon other things, such as the rate of interest. These can be brought
in and they fit quite well into the theory, but it is a reasonable simplification to assume that small changes in the rate of interest will
affect different people in opposite directions; and the net effect may
here be neglected.
3
There remains to be considered what determines the rate of investment. It is in the analysis of this that some of the more subtle
and more valuable innovations in the theory are made by Keynes.
Investment consists in the application of productive resources to the
manufacture of capital goods. Capital goods are goods which are
valuable on account of services they are expected to yield in the
future. The efficiency of a capital good, or the rate of return over
cost, as Irving Fisher calls this, is the rate of yield of the capital
good, i.e., it is that rate of discounting the expected future yields
of the capital good which makes the sum of the discounted yields
equal to the cost of making it. For example, if it costs '300 to make
a machine which gives off two services, one in one year's time which
is then worth '220 and one in two years' time which is then worth
'121, the efficiency of this machine is 10 per cent, because, if the
values of the services are discounted at the rate of 10 per cent down
to the present, the sum of their values is '300.
100
('100)2
('220X-='200, '121 X ='100,
110
110

and

'200+'100='3 00.)

The marginal efficiency of any particular type of capital good is

216

Abba P. Lerner

the efficiency of the marginal item of that type of capital good, in the
use where its installation would show the greatest possible efficiency.
The marginal efficiency of capital in general is the highest of the
marginal efficiencies of all capital goods that still remain to be made.
It should be noted that the marginal efficiency of any capital good
is described in the same way (has the same dimensions) as the rate
of interest, so that it can be measured against it. It is a percentage
of so much per annum. But it must on no account be confused with
the rate of interest. The rate of interest is the rate at which money
has to be paid for the privilege of borrowing money; or, from the
point of view of the lender, it is the rate at which one is remunerated
in money for the service of lending money.
There is, however, a certain relationship between the rate of
interest and the marginal efficiency of capital. For it will pay entrepreneurs to borrow money in order to increase the rate of construction of capital goods-which is the rate of investment-as long
as the rate of interest is less than the marginal efficiency of capital.
As the rate of investment increases, the best opportunities for investment are used up, and the marginal efficiency of capital diminishes.
This happens in two ways. As the amount of capital increases, the
expected values of the services of new capital goods fall as these
have to compete with a larger supply of existing capital goods. This
will be a very slow process since the rate at which capital is increased-the output in a short period-is small relatively to the
existing stock of capital goods. But the other way in which the
marginal efficiency falls is operative in the short period. As the rate
of investment increases, the marginal cost of making capital goods
increases, and this immediately tends to reduce the marginal efficiency of capital to the rate of interest. For each rate of interest
there is a corresponding rate of investment. This relationship is the
schedule of the marginal efficiency of capital.
The schedule of the marginal efficiency of capital is sometimes
called the demand curve for savings because the entrepreneurs,
who undertake the investment and have to obtain the funds to finance it, are conceived to obtain them from the savings of individuals
which when summed constitute the "supply" of savings. This is
important in so far as it is brought in to explain the amount of investment that takes place, and upon the amount of investment de-

The General Theory .

2I

pends-as we have seen-the amount of employment which is the


quaesitum of the whole book.
It is clear that the amount of investment undertaken by entrepreneurs in any given position, given the marginal efficiency schedule
of capital, will be determined by the rate of interest. The crux of the
matter lies then in the theory of the determination of the rate of
interest.
According to the classical theory, the rate of interest is given by the
supply and demand schedules for savings. The rate of interest is
the price of savings and that amount of saving and investment comes
about that is indicated by the intersection of these demand and
supply schedules. If the supply of savings is greater than the rate
of investment, the rate of interest will fall so as to bring them into
equilibrium, and vice versa. Savings and investment are brought
into equality with each other in an equilibrium by the movement
of the rate of interest.
This line of reasoning is not merely wrong-it is meaningless.
The equations on page 210 show that savings can never be different
from investment whatever the rate of interest, so that it is nonsense
to say that the rate of interest brings them to equality with each
other. This can be shown in another way. The supply schedule of
savings in this scheme is supposed to be independent of the demand
curve for saving (which is the marginal efficiency schedule of
capital). This means that, given the rate of interest, the amount of
saving is independent of the amount of investment and also of the
size of people's incomes. In fact, of course, it is ridiculous to assume
that this is so, for what happens is that if there is an increase in
investment, incomes increase immediately, so that saving is increased
by exactly the amount that investment is increased. The supply
curve does not keep still. Whatever the point one takes on the demand curve, the supply curve moves to the right or to the left so that
it intersects the demand curve at the point taken.
We can now consider the alternative argument, referred to above,
which is sometimes put forward in defense of the proposition that
any individual, in deciding to save more, thereby increases 5, the
aggregate amount of saving of the whole society. Instead of assuming that when an individual saves more and spends less on consumption goods the seller of consumption goods continues to receive the

218

Abba P. Lerner

same amount as before, so that Y, the aggregate income of society,


is unaffected, it is assumed that whenever an individual decides
to increase his saving by a certain amount, either he or somebody
else always increases investment by the same amount. This increases
the incomes of those engaged in the production of investment goods
by as much as the income of the producers of consumption goods
diminishes, so that Y, the aggregate income, remains the same. C,
the expenditure on consumption goods, has diminished, and (Y-C)
or S has increased as much as the first individual increases his
own saving.
There are two difficulties about this argument. The first is that
there is no satisfactory indication of any mechanism in a monetary
economy whereby the decision to save necessarily carries with it
an instantaneous and equal decision to invest. The second difficulty
is that, if there were some mechanism which did make somebody
decide to invest exactly as much as anybody saved, the classical
explanation of the determination of saving and investment would
be upset in a manner similar to the one we have indicated. For this
would mean that the investment curve-which constitutes the demand curve for the supply of savings-coincided throughout with
the supply curve of savings. At each rate of interest people decide
to save a certain amount (supposing for the moment that the supply
curve of saving is not shifted about by changes in income due to
changes in investment)-and if there is some mechanism whereby
an individual's decision to save calls into being an equal amount of
investment, then saving again equals investment throughout (though
not for the reasons given above). There is only one curve, which
is both the supply curve and the demand curve for savings, so that
the rate of interest remains unexplained.
This argument sometimes takes the form of assuming MV (the
amount of money multiplied by its velocity of circulation) as unchanging. This means that the total amount spent altogether, both
on consumption and on investment, is unchanged, so that if '1 less
is spent on consumption '1 more must be spent on investment.
This assumption is frequently very tacit, and when made explicit
it appears in extremely innocent-looking forms like assuming "other
things remaining the same" or considering what happens "in the
absence of hoarding." This really means that unless something

The General Theory .

2I

special from outside-"hoarding"-intervenes, we may expect MV


to remain constant and that any decision to save will somehow result
in somebody investing an equal amount. This criticism of the illegitimate and sometimes unconscious assumption of a constant MV is
not Keynes' way of dealing with the argument. He usually refuses to
have anything to do with such simple "quantity equations." Dr.
Haberler, however, concentrates on this line of attack, which is only
a more orthodox (and more complicated) route that leads to the
same conclusions as are obtained by Keynes.
There remains unexplained what it is that determines the rate
of interest. The explanation of this is given by Keynes, who derives
it from the inadequate theories of the Mercantilists by an easy
development of a line of thought that had been shut out of economic
theory for over a century. This line of thought has only recently
been coming back into respectable economics under very heavy
disguise in the writings associated with such esoteric concepts as the
"natural rate of interest" and "neutral money."
The rate of interest is what people pay for borrowing money. It is
what people who have money--cash-obtain for lending it to other
people instead of holding it themselves. It is not payment for saving,
for one can save without lending the money saved; and in that case
one does not get any interest payments. On the other hand, one
can lend the money out of what one previously held; and in that
case one gets interest payments without saving. The relevant demand is then the demand to hold money. The supply is simply the
total amount of money that there exists. This demand schedule
Keynes called liquidity preference, and it is the intersection between
the liquidity preference schedule and the supply of money (which
is a perpendicular line if the amount of money is fixed) that gives
the rate of interest upon which the whole thing depends. The higher
the rate of interest the greater the cost-in terms of interest foregone-of holding money and the smaller the amount of money
people will want to hold. Conversely, if there is an increase in the
amount of money, the rate of interest will fall until people want
to hold the larger amount of money. They are induced to want to
hold more money by the fall in the rate of interest, for then, to some
people, the convenience and feeling of security of holding cash can
be satisfied to a greater extent because the cost is less.

220

Abba P. Lerner

Our conclusion is that the amount of employment can be governed


by policy directed toward affecting the amount of investment. This
may be done either by lowering the rate of interest or by direct
investment by the authorities. There may be difficulties for institutional or psychological reasons in reducing the rate of interest to
sufficiently Iowa level to bring about that rate of investment which,
with the existing propensity to consume, is necessary in order to
bring about full employment. It is because of such difficulties that
Keynes thinks that public works are necessary, and may become
more and more necessary as the wealth and capital equipment of
the community increase. For this means that, on the one hand,
people wish to save more out of the larger income corresponding to
full employment while, on the other hand, the accumulation of
capital lowers the marginal efficiency schedule of capital. Equilibrium with full employment is then possible only at lower interest
rates than are practicable unless either (a) investment is increased
by state production of capital goods whose efficiency is less than the
rate of interest or which for any other reason would not be manufactured by private entrepreneurs, or (b) the propensity to save is
diminished--consumption increased-by state expenditure on social
services or by redistribution of income from the rich to the poor, or
by any other means.
The reader may have noticed a considerable similarity between
this last argument and the classical argument that was so vehemently
attacked on pp. 217 and 218. Here in fact an equilibrium is indicated by the intersection of demand and supply curves for savings.
What the argument amounts to is that, if for institutional reasons the
rate of interest cannot be brought down to the level which equates
the supply and demand, the demand curve must be moved to the
right or the supply curve moved to the left until they meet at a level
of the rate of interest that is practicable. But what was impossible in
the classical explanation of the rate of interest is permissible here
because for this argument we were assuming full employment in
order to be able to consider what are the necessary conditions for
that to exist. There is then a given income so that there is a given
supply schedule of savings. The criticism of the classical explanation of the determination of the rate of interest is that its argument
-in assuming a given supply curve of saving-is implicitly assum-

The General Theory'

22I

ing a given degree of employment, namely, full employment. And


it is not useful to consider what determines the amount of employment on the assumption that there is full employment-or even to
discuss the determination of the rate of interest under those conditions without considering whether in fact there is any force which
will bring about full employment.
Keynes' conclusion that the amount of employment has to be
governed by operating on the amount of consumption and investment, via the rate of interest or otherwise, may seem at first sight
to be a very small mouse to emerge from the labor of mountains.
Everybody has known that cheaper money is good for business, and
so is any increase in net investment or expenditure. But except
for occasional lapses from scientific purity to momentary common
sense, the pundits of economic science have been declaring that
people should practice more thrift. There has been a weakening
of this attitude recently-I am not clear to what extent this is due
to the cyclical fluctuations in the attitude of economists and how
much to the influence of Keynes' ideas and some parallel development by J. R. Hicks and the Swedish writers. But we must not
forget that it is not so very long ago that we had Professor Robbins
and Keynes on the wireless, respectively advising the world to save
more and to spend more. And there is still in Milan a World Institute for the Encouragement of Thrift. It will be a long time before
the view that thrift "since it enriches the individual can hardly fail
to benefit the community" is seen to be an important example of
the common logical error of composition. What Keynes has done
is to show that what the ordinary man has often felt in his bones
can be justified by a keener analysis than has so far been applied
to the problem. He has shown further that it is only by working
indirectly on these same determinants that any other remedies can
ever work. Thus, even in the case when a reduction of money wages
increases employment, it does so only in so far as it indirectly reduces
the rate of interest. The direct effect is merely to reduce both
prices and money incomes, leaving the real situation as before. At
the lower price level, people find that they need less money to
carryon their business, so that, if there is no change in the amount
of money, its supply is greater than the demand to hold it, and
the attempt of money holders to lend the spare money to others, or

222

Abba P. Lerner

to buy other assets for money, raises the value of the other assets
and reduces the rate of interest. The reduction of the rate of interest
does the trick by making a larger rate of investment profitable.
Incomes then increase, in accordance with the propensity to consume, until a level of income and employment is reached which
induces people to save at a rate equal to the greater rate of investment. From this it follows that any objections that may be raised
against the dangers inherent in lowering the rate of interest in an
attempt to increase employment apply just as much or as little to
the policy of increasing employment by lowering wages, since that
works only via lowering the interest rate. It is not denied that there
are any dangers, but such as they are, they are inherent in any successful attempt to increase employment. To run away from these
is to refuse to be cured because that will make it possible to become
sick again.
To seek the alleviation of depression by reducing money wages,
rather than by directly reducing the rate of interest or otherwise
encouraging investment or consumption, is to abandon the high road
for a devious, dark, difficult, and unreliable path, for no better reason
than that the dangers that await one at the common destination are
more clearly seen when it is approached by the broad highway.

Keynesian Economics In the


Sixties
[1963J

HE MOST STRIKING DEVELOPMENT in Keynesian economics in the last decade has been the retrogression in its general
understanding and acceptance. Ten years ago it seemed as if the
general theory of employment had almost completed the course
through which all new ideas have to pass. From being considered

Keynesian Economics in the Sixties

223

something wrong and revolutionary it had become something right


and natural and even "what we had known all along." When a
new theory has reached this degree of naturalization there is good
hope of it being applied to practical problems since it is no longer
regarded as "a theory" but as "common sense." Keynesian economics
is, however, now passing through yet another phase. So successfully
has it been disguised as "what we have known all along" that the
ancient fallacies have been able to creep back on the stage, pretending that they are the new "common sense."
In the realm of practical political economy and the public discussion, the balanced budget seems to have regained the reverence
it enjoyed in 1931 when Franklin Delano Roosevelt was promising
that if he were elected President of the United States he would cure
the depression by doing away with Hoover's budget deficit. Unhappiness at the danger of the National Debt crossing the $300billion line is almost as great as were the apprehensions of catastrophe
from its exceeding the $,,-billion limit. Reminders that as a ratio
to the national income the national debt has fallen to less than half
of what it was in 1946 are generally met with incredulity, so thoroughly has the general public been bemused by the anxieties of city
editors (or vice versa).
Even when some genuine common sense breaks through it is
accepted only with pained reluctance. In some recent conferences
groups of economists were mobilized by government departments
to remind groups of more enlightened industrialists that tax cuts
can increase employment and economic growth when these are unduly low, and to suggest that tax cuts might therefore be desirable
even though they would also increase a government deficit already
expected to be quite large. The economists had considerable success.
Most of the businessmen accepted the suggestion. But they seemed
to feel quite guilty about making use of so unholy a remedy, were
very doubtful whether sufficient general acceptance of the suggestion
could be obtained to make it possible for the government to carry
out such a program, and were extremely unhappy about the whole
thing.
In one group in such a conference I attempted to ease the guilt
feelings by pointing out that in agreeing that an increase in an already large deficit could be in the general interest, they had in fact

224

Abba P. Lerner

rejected the balancing of the budget as a principle and were using


it only as a benchmark for indicating how large a divergence from
it was desirable.
In a stationary state the balanced budget point might be considered as something more than a benchmark. Where the population,
the national income, the stock of capital, the wage rate, and the
standard of living were all constant it might seem natural for the
national debt also to be stationary, and that would mean a balanced
budget. But where all of these elements in the economy were rising
it might seem natural for the national debt to keep up with them.
A national debt that just kept pace with, say, Gross National
Product would therefore be a more appropriate benchmark. That
would mean measuring the optimum policy in terms of the departure from maintaining the national debt at a constant proportion
of GNP, and the benchmark would be that size of deficit which
would have the effect of maintaining the constant proportion of
national debt to GNP.
If we accepted the target of a GNP growing at 4 per cent per
annum, its annual increase would be about $24 billion (4 per cent of
$600 billion) and the benchmark deficit would be about $12 billion
(4 per cent of $300 billion). This was in fact just about the expected
deficit in the absence of the tax cuts, and what had been considered
by the conference was the degree to which the unduly low level of
employment and growth (perhaps due precisely to the failure, since
the War, of the national debt to keep up with the rest of the
economy) should be remedied by tax cuts that would make the deficit
exceed the $ 1 2 billion benchmark.
As might have been expected, the businessmen's emotional distress at contemplating government deficits completely prevented
any positive response to this way of making a substantial deficit look
natural-even though the group, in explicitly (if reluctantly) accepting the policy proposal of tax cuts, had implicitly already accepted
the view that the deficit should be made to exceed the new benchmark.
There are many reasons for the retrogression in the public acceptance of the Keynesian theory of economic policy for the maintenance of full employment, but I shall here concentrate on some
of the purely intellectual ones that seem to me to be central to the

Keynesian Economics in the Sixties .

225

issue. The most important of these was a failure of the Keynesians


(probably including Keynes himself) to realize to the full how revolutionary was the analytical basis of the Keynesian revolution.
The so-called classical pre-Keynesian treatment of employment
equilibrium can be represented in its simplest form in Figure I
where the intersection of the supply and demand curves for labor
in real terms determine the volume of employment (at full employment) and the real wage. The vertical arrows indicate the operative
forces that move the economy toward the equilibrium position. At
a wage higher than that shown by the point of intersection (which
is the equilibrium position) the excess of supply of labor over the
demand, the unemployed in the disequilibrium, push the wage down
until the equilibrium point is reached. At a wage lower than equilibrium, the shortage of labor pushes the wage up to the equilibrium
point.
Keynes and the Keynesians pointed out that this argument would
not do--on several levels. The crudest form of the argument consists of the simple transference to the whole economy of the individual employer's experience. A wage cut enables him to sell more
of his product and to provide more employment. This argument
would not do because it depended on the individual employer's competitive advantage in finding his costs cut while his selling price was
unaffected because his competitors did not share in the lower costs
from the wage cut. It was therefore not applicable to a general wage
cut to increase general employment.
A more sophisticated form of the argument avoided dependence
on the individual employer's competitive advantage and claimed
only that even if prices fell together with wages, wiping out any
competitive advantages, there would still be an income effect, the
same total national income would be able to buy more goods and
services and so provide for more employment. This still would not
do because the national income could not be expected to remain
the same.
A third and still more sophisticated argument shifted the ground
again, pointing out that even if incomes fell together with prices
and wages, wiping out the income effect, there would still be a
liquidity effect. There would be a reduction in the demand for
money to be held in connection with the reduced volume of money

226

Abba P. Lerner

transactions. That would result in lower interest rates, more investment, more earnings, more consumption and so more employment.
This argument would still not do-in the first place because the required fall in wages, prices, and incomes could be brought about
only by a long and severe depression, and in the second place because the reduction in demand for money stock would not always
lower interest rates and lower interest rates would not always result
in more investment.
w
p

~------------~N
Fig. 1

I....---~F~---..N

Fig. 2

A fourth and still more sophisticated argument was able to deal


with the second of these two difficulties but only at the cost of aggravating the first. It pointed out that even if the liquidity effect did
not work, there would still be a wealth effect. If wages and prices
and incomes kept on falling, the people who had some stocks of
money or of government bonds would become richer and richer (because of the continuing increase in the value of their money stocks
or government bonds as prices kept on falling) so that they would
increase their spending. The continuing increase in demand would
ultimately restore full employment. But this logically unassailable
fourth argument still would not do. It depends on such a great degree of price reduction that hardly anyone put it forward as a
practical alternative to Keynesian full employment policy. It would
never do to wait for these natural forces to cure depressions. But
the argument does serve to show a certain abstract consistency and
correctness in some interpretations of the pre-Keynesian or "classical"
theory of employment.

Keynesian Economics in the Sixties

227

The state of theory at this point can be represented by Figure 2.


We again have supply and demand curves for labor with the point
of intersection indicating full employment just as in Figure I. But
the interpretation is quite different. What it shows is that if employment is less than full, forces are set in motion that tend to increase the volume of employment. Unemployment induces a fall in
wages and prices which increases in the value of the money stock,
which lowers the rate of interest, which increases investment, income, consumption, and employment (the third argument). Or,
failing that, the fall in prices increases consumption spending by
those made rich by the increase in the value of their stocks of
money and government bonds (the fourth argument). These tendencies being too sluggish for practical purposes, they have to be
helped along by a more direct increase in the value of the money
stock (by creating more money-i.e., by monetary policy) or by a
still more direct increasing of spending by the government or by
citizens benefiting from tax cuts (i.e., by fiscal policy). Both the
natural tendencies and the expansionary monetary and fiscal policy
"aids" that move the economy towards full employment are represented by the arrow on the left of the intersection point of Figure 2.
The arrow on the other side indicates all the same tendencies and
"aids" working in the opposite direction when there is over full employment, inflation (reducing the value of the money stock) and
restrictionary monetary and fiscal "aids" for checking the inflation.
When Figure 2 is drawn without the arrows it looks very much
like Figure I. This similarity helped the Keynesian theory to become accepted as "natural," but when some difficulties developed
with the new theory it also made it easier for the ancient errors
shown up by the Keynesian analysis to move back in.
The difficulties arose when it was found that the use of monetary
and fiscal policy measures simply as "aids" to strengthen the "automatic" equilibrating tendencies of the third and fourth arguments
was not enough. Attempts by expansionary monetary and fiscal
"aids" to increase employment brought about price increases or inflation, before the unemployment was cured. Attempts to cure the
inflation by restrictionary monetary and fiscal "aids" brought about
an increase in unemployment before the price increase was stopped.
The Keynesian policy seemed to be calling for expansionary and for

228

Abba P. Lerner

restrictionary monetary and fiscal policies at the same time to deal


with the unemployment and inflation existing side by side. The
resulting dissatisfaction with the new theory paved the way for a
rehabilitation of the old, under a barrage of ill-informed criticisms
of Keynesianism as elementary economic error, spiced with sedition,
and disowned by Lord Keynes himself on a repentant deathbed.
The Keynesians had laid themselves open to this counter attack by
their failure to look deeply enough at the nature of the weakness in
the "automatic tendencies" to full employment. They had assumed
that the root of the problem was the "downward stickiness" of wages
-i.e., that weakness or sluggishness of the automatic tendencies to
full employment as described in the third and fourth arguments; so
that all could be put right by monetary and fiscal "aids," which, by
quickly increasing (or decreasing) investment and consumption
just when the automatic forces were trying to do this, would be
acting only as amplifiers or servo-mechanisms to increase their power.
But as soon as one asks why it is that wages and prices are "sticky
downward" the obvious answer is that it is because they are not determined by the automatic market forces of supply and demand but
are decreed by individuals or committees who have the power to
countermand the automatic market forces-to stop prices and wages
from falling when there is an excess of supply over demand. The
automatic forces are not weaker than in other markets. They are
inhibited by these men or committees who are wage and price administrators. And if the wage and price administrators have the
power to stop wages and prices from falling when demand exceeds
supply (and the market is telling them to fall) they also have the
power to make wages and prices rise when the market tells them
not to.
The market, i.e., the relationship between supply and demand,
does have some influence on the administrators. When there is great
excess of demand over supply it is difficult for the administrators to
prevent prices from rising, because black markets will emerge
(whether the administrators are government officials attempting to
fight inflation by price control or whether they are automobile manufacturers trying to maintain good will by keeping prices below
what the market will bear). When there is a very great excess of
supply over demand the administrators may be unable to prevent

Keynesian Economics in the Sixties .

229

"cutthroat competition." But the power of the wage and price administrators to raise wages and prices in the face of an excess of
supply over demand, although it is not unlimited, is great enough
to enable them to give us inflation whenever unemployment is much
less than about 7 per cent. And as the economy gradually becomes
richer and as it becomes more inured to this evil the greater becomes
the intensity of unemployment needed to stop prices from rising.
Only when unemployment exceeds this degree of intensity is the
market strong enough to overcome the power of the wage and price
administrators and induce price reductions. Such a situation is shown
in Figure 3. The operating automatic tendency, shown by the
horizontal arrows, is for employment to move not towards full employment at F (where supply is equal to demand) but toward the
price stability level of employment at P where the market forces trying to reduce prices are just able to balance the administrators' efforts
to raise them.
If monetary and fiscal measures are used only as "aids" to provide muscles to bring about quickly what the observable automatic
price movements are "tending" to achieve, what we get is only a
more effective maintenance not of full employment at F but of the
price stability level of employment at P (with about 7 per cent unemployment). The arrows now represent the effects on employment
of expansionary monetary and fiscal measures undertaken when
prices show a falling tendency, and the effects of restrictionary monetary and fiscal measures undertaken when prices show a tendency
to rise. Employment as well as the price level is stabilized, but at so
Iowa level that there emerges pressure for the direction of monetary
and fiscal policies to the different goal of a higher level of employment.
If this pressure succeeds and monetary and fiscal policies are redirected and aimed at raising employment to some point to the right
of P, this results in prices rising. The frustrations from trying to
achieve the conflicting aims of stabilizing the economy at F so as to
have full or high employment while at the same time stabilizing it
at P so as to have no inflation or deflation of prices are the basic reasons for the current retreat from Keynesianism. Meanwhile the U.S.
government seems to be aiming at a compromise point slightly to the
right of P, with unemployment at around 6 per cent and it is nat-

230

Abba P. Lerner

urally encountering some difficulties in restraining the tendency of


prices to rise.
The way out of the dilemma lies only in the artificial establishment of the conditions assumed in the neoclassical models of arguments three and four. There are two ways of doing this. One way is
to replace our modem industrial economy by a perfectly competitive
one in which there are no wage and price administrators but where
all products are sold on perfect markets for whatever they will fetch
in a daily equalization of supply and demand-as in our produce

w
p

w
p

D
I...--p~.-~F--""N
Fig. 3

D
Ol...--P~-~F~-""N
Fig. 4

markets. The other way is to subject the more important wage and
price administrators to regulations that will make the administered
prices move in something like the way prices would move in a perfectly competitive economy. This means that the price administrators would raise the price of a product only when there is an
excess of demand over the available supply and would lower the
price of any product whose available supply significantly exceeded
the demand; they would pay as little attention to profits or losses
as is paid by the forces of demand in a perfectly competitive market. The price level-an average of the prices-would thus be
kept stable. The wage administrators would raise wages in general
at the rate that is compatible with a stable price level, raising any
particular wage by more than this only where labor was significantly
scarcer (unemployment significantly less) than in the economy as
a whole, and by less than this, or not at all, where labor is significantly
more abundant (unemployment significantly greater) than in the
economy as a whole.

Keynesian Economics in the Sixties .

231

Since the first way does not seem feasible and the second is believed to be a very nasty medicine (being easily though wrongly
identified with price control), this diagnosis and prescription is not
likely to restore the popularity of the doctor.
A contributory factor in the decline in the popularity of Keynesianism is a related failure to notice how the Keynesian less-than-full-employment-equilibrium (or disequilibrium, for this is purely a matter
of terminology) does not fit in with the equalization of the prices of
factors with the marginality conditions of neoclassical equilibrium. In
Figure 4 (which is the same as Figure 3 with some qdditional lines
and letters taken from a diagram of Patinkin'sl) the demand curve D
shows how much labor is demanded by employers who, to increase
profits, will want to increase employment whenever the marginal
value product is greater than the wage and to reduce employment
whenever it is less; the supply curve S shows how much labor the
workers are willing to provide at each real wage. Consequently, in
full employment equilibrium (represented by point M) the real
wage is equal to the marginal value product of labor and the marginal
utility of the wage is equal to the marginal disutility of labor. But if
there is less than full employment, as in the case of the price stability
level of employment OP, there is no reason for supposing either of
these equalities to hold. All we can say is that the real wage cannot be
greater than PA because then the demand for labor would be less
than OP, and that it cannot be less than PB because then the supply
of labor would then be less than OP. The real wage could be anything between PB and PA.
Keynes confused the issue by assuming that only the second equality had to be given up-the equality between the marginal utility of
the wage and the marginal disutility of labor-but that the real wage
was still equal to the marginal value product. Such a position would
be indicated by the point A on the D curve. But that shows considerable involuntary unemployment (measured by AN), while
employers are perfectly satisfied with the level of production and
employment, not wishing to employ any more people to produce
any more goods even if they could sell them at the current relationship of wages and prices. One might call this a case of unemployment without depression! One might just as well assume that the
1

Don Patinkin, Money, Interest and Prices, Row Peterson, & Co.
Evanston, Ill. 1957, p. 213.

232

Abba P. Lerner

position would be that indicated by point B on the S curve where


there is no unemployment-everybody who wants to work at the
current real wage is at work-but employers would like to employ an
additional complement of workers BB' at the current wage-price relationship but do not offer any more employment because there are
no customers for what they would produce. This could be called depression without unemployment!
Where the real wage would be (between the limits PB and PA)
will depend on the levels at which the wage and price administrators
had set prices and wages. If they had set them at levels consistent
with the full employment equilibrium position M, and their relative
values had not changed in the course of the adjustment of the economy to the OP level of employment, the economy would be found
at the point K, with just as much depression as unemployment. But
there is no good reason why this should be so. Wages and prices and
the relationship between them are not determined by marginal
equalizations but are free to be settled by bargaining, legislation,
convention, and other institutional forces that have no play in the
full employment equilibrium. At a level of employment other than
OP the price level will be rising or falling and it will be even less
possible to say where the real wage will settle.
The conversation at cross purposes between economists discussing
the marginalist determination of price-wage relationships at M and
at A, and businessmen and labor leaders discussing the institutional
determination of wage price relationships in the area lying between
the supply and demand curves to the left of M, has contributed its
quota to the current turning away of the public from the Keynesian
enlightenment to classical and preclassical superstitions.
A final contribution to the darkening of counsel is provided by the
resurgence of anxiety about the national debt. Since the early history
of economic theory economists have been pointing out that national
debt, unlike personal debt, does not constitute an impoverishment
of future generations because the future generations inherit not only
the obligations of the government to the bondholders but also the
bonds that constitute the credit side of the debt. The future generations are not only the heirs of the debtor, they are also the heirs of
the creditor.
The failure to understand this elementary proposition usually

Keynesian Economics in the Sixties 233


shows itself in a concern about the immorality of burdening future
generations with national debt. A number of ingenious economists,
perhaps because they have grown weary of pointing out that this so
called "burden of the national debt" is just a mistake, have been
seduced by the sport of showing how there can be shown to be a real
''burden of the national debt on future generations" if only one redefines some of the words.
If you define the people over forty (or any other age) as generation a and the people under forty as the generation b, and the government borrows from the over-forties and taxes the under-forties to
repay the over-forties (and the over-forties do not give or bequeathe
their bonds to the under-forties), the national debt does constitute
the shifting of a burden from generation a to generation b.
If you define "burden" not as managing with less but as paying a
tax, then a generation that suffers from malnutrition during a war
or other calamity in which they voluntarily loaned most of their income to the government are shifting the burden onto a future generation that pays taxes that are returned to themselves as interest on
the government bonds they have inherited.
If a generation that inherits government bonds from a previous
generation (together with the government debt counterpart of these
same bonds) feels itself richer on that account (because its members
pay more attention to the bonds they own than to their share of the
national debt), then I would say it is better off since I find it difficult to distinguish between feeling better off and being better off.
The one is just as subjective as the other. If its feeling richer causes
it to save and invest less, then a third generation will inherit less
productive equipment than if the debt had not existed. But since it
will also inherit the bonds, it has at least the same reduction in its
urgency to save as the second generation and will also be better off.
Only when some future generation inherits so much less equipment
that it cancels the benefit from the inherited bonds (and national
debt) will it feel no richer and there will be no further effects on still
more distant generations.
However, if one counts the reduction in inherited productive
equipment as a loss but does not count the inherited government
bonds (together with the inherited national debt) as a benefit, then
the third and subsequent generations appear to suffer a burden.

234 Abba P. Lerner


This, if one can call it a burden, is a burden arising not directly from
the national debt but only from the (uncounted) benefit from national debt, since without the feeling of being richer the second generation would not have reduced its saving. Nevertheless, this has
also been called a burden of national debt.
It is a pity to disturb those who enjoy these acrobatics. But their
activities appear to give respectability to the fundamental error of
seeing the debt as a subtraction from national wealth, they tend to
discredit those economists who are still patiently trying to combat the
superstition, and they have contributed to the public regression from
the Keynesian enlightenment.
However far more important than some economists' inappropriate
marginalism or their fun and games with "the burden of the national
debt on future generations" is the unsolved political problem of
checking the inflationary pressure of administered wages and prices
by means other than an ever increasing level of unemployment.

JACOB VINER

Mr. Keynes on the Causes


of Unemployment
[1936J

HE INDEBTEDNESS OF ECONOMISTS to Mr. Keynes has


been greatly increased by this latest addition to his series of brilliant,
original, and provocative books, whose contribution to our enlightenment will prove, I am sure, to have been even greater in the long
than in the short run. This book deals with almost everything, but
the causes of and the future prospects of unemployment, cyclical
and secular, are its central theme. It brings much new light, but its
display of dialectical skill is so overwhelming that it will have probably more persuasive power than it deserves, and a concentration on
the points where I think I can detect defects in the argument, though
it would be unfair if presented as an appraisal of the merits of the
book as a whole, may be more useful than would a catalogue-which
would have to be long to be complete-of its points of outstanding intellectual achievement.
Written though it is by a stylist of the first order, the book is not
easy to read, to master, or to appraise. An extremely wide range of
problems, none of them simple ones, are dealt with in an unnecessarily small number of pages. Had the book been made longer, the
time required for reading it with a fair degree of understanding would
have been shorter, for the argument often proceeds at breakneck speed
and repeated rereadings are necessary before it can be grasped. The
book, moreover, breaks with traditional modes of approach to its
problems at a number of points-at the greatest possible number of
points, one suspects-and no old term for an old concept is used
when a new one can be coined, and if old terms are used new meanings are generally assigned to them. The definitions provided, moreover, are sometimes of unbelievable complexity. The old-fashioned

235

236 . Jacob Viner


economist must, therefore, struggle not only with new ideas and
new methods of manipulating them, but also with a new language.
There is ample reward, however, for the expenditure of time and
attention necessary for even partial mastery of the argument.
"INVOLUNTARY" UNEMPLOYMENT

Mr. Keynes claims that the "classical"l economists recognized


the possibility only of "frictional" and of "voluntary" unemployment,
and that a vitally important chapter of economic theory remains to
be written about a third class of unemployment, for which there was
no place in the "classical" scheme of things, namely, "involuntary"
unemployment. The concept of "frictional" unemployment relates to
the inevitable loss of time between jobs, and presents no difficulties.
"Voluntary" unemployment is defined as the unemployment "due
to the refusal or inability of a unit of labor ... to accept a reward
corresponding to the value of the product attributable to its marginal
productivity," but is used in such manner as to require the addition
to this definition of the proviso that the money wage offered must
not be below what the laborer regards as a proper minimum rate of
money wages. If laborers refuse available employment at a money
rate below this minimum, or if employed laborers refuse to permit a
prevailing money rate to be lowered and unemployment results for
themselves or for others from this refusal, Keynes would apparently
regard it as "involuntary" unemployment, but deny its possibility or
probability. He defines "involuntary" unemployment as follows:
"Men are involuntarily unemployed if, in the event of a small rise
in the price of wage-goods relatively to the money wage, both the
aggregate supply of labor willing to work for the current moneywage and the aggregate demand for it at that wage would be greater
than the existing volume of employment." Cpo 15). What he seems
to mean by this is that any unemployment which would disappear if
real wages were to be reduced by a rise in the prices of wage goods,
money wages remaining the same or rising in less proportion, but not
falling, would be involuntary. It is with "involuntary" unemployment so understood, its causes and its remedies, that Keynes' analysis
of unemployment is primarily-and almost solely-concerned.
1

Used by him to mean the later economists, such as J. S. Mill, Marshall,


Edgeworth, Pigou, who in the main were adherents of the Ricardian
tradition; a usage which I shall follow here.

Causes of Unemployment 237


In Keynes' classification of unemployment by its causes, unemployment due to downward rigidity of money wages, which for the
"classical" economists was the chief type of cyclical unemployment
and the only important type of secular or persistent unemployment,
therefore finds no place. As will be seen later, it is excluded on the
ground that resistance to reductions in money wage rates generally
does not involve a reduction in the volume of employment and is,
if anything, favorable to employment rather than the reverse. The
omission charged against the "classical" economists is their failure to
note the lesser resistance of labor to reductions in real wages if unassociated with reductions in money wages per se, and their failure to
recognize the existence of a large volume of unemployment for
which the former is an available and practicable remedy, but not the
latter. Keynes' reasoning points obviously to the superiority of inHationary remedies for unemployment over money wage reductions.
In a world organized in accordance with Keynes' specifications there
would be a constant race between the printing press and the business
agents of the trade unions, with the problem of unemployment
largely solved if the printing press could maintain a constant lead
and if only volume of employment, irrespective of quality, is considered important.
The only clash here between Keynes' position and the orthodox
one is in his denial that reduction of money wage rates is a remedy
for unemployment. Keynes even follows the classical doctrine too
closely when he concedes that "with a given organization, equipment and technique, real wages and the volume of output (and hence
of employment) are uniquely correlated, so that, in general, an increase in employment can only occur to the accompaniment of a
decline in the rate of real wages" (p. 17). This conclusion results
from too unqualified an application of law-of-diminishing-returns
analysis, and needs to be modified for cyclical unemployment, as well
as for the possibility that the prices of wage goods and of other goods
may have divergent movements. If a plant geared to work at say 80
per cent of rated capacity is being operated at say only 30 per cent,
both the per capita and the marginal output of labor may well be
lower at the low rate of operations than at the higher rate, the law
of diminishing returns notwithstanding. There is the further empirical consideration that if employers operate in their wage policy in
accordance with marginal cost analysis, it is done only imperfectly

238 . Jacob Viner


and unconsciously, and the level of wages they can be persuaded to
establish is strongly influenced by the profitability of their operations
as a whole, and not solely-if at all-by calculations of the marginal
contributions of labor to output.
Keynes uses the term "full employment" to signify the absence of
any involuntary unemployment (p. 16). He describes it also as the
condition which would prevail "when output has risen to a level at
which the marginal return from a representative unit of the factors
of production has fallen to the minimum figure at which a quantity
of the factors sufficient to produce this output is available" (p. 303).
There are implied here several questionable propositions. The concept of diminishing marginal productivity is generally used in economics in a partial differential sense to indicate the diminishing
increments of output which would result when some particular factor or group of factors was being increased, the remainder of the
working combination being held constant. If all the factors are being
increased simultaneously and in uniform proportions, it requires
some such assumption as that of the general prevalence of external
technical diseconomies from increased production if it is to be accepted that output and return per compound unit of the factors
must be negatively correlated. There is also implied here the assumption that any increase in real wages (money wages remaining
constant, or rising) will result in an increase in the amount of labor
available. If, as widely held opinion since the seventeenth century
has maintained, and as Professor Paul Douglas' recent investigations
for urban labor in the United States appear to confirm, the supply
schedule of labor with respect to real wages is, for part of its range
at least, negatively inclined, the volume of employment could conceivably be much greater when there was "involuntary" unemployment than when there was "full" employment, and Keynes' conditions of "full" employment might be met at an indefinite number of
levels of employment.
"Full" employment rarely occurs, according to Keynes, and the
main immediate responsibility for the persistence of "involuntary"
unemployment lies with the persistence of interest rates at levels too
high to induce employers to bid for all the labor available at the
prevailing money rates of wages. An elaborate and strikingly novel
analysis of the causes determining the level of interest rates leads to

Causes of Unemployment 239


the conclusion that high "liquidity preferences" of savers, an excessive disposition to save, and a low marginal productivity of investment are responsible for the absence of such a relation between the
rates at which savers are willing to lend and the rates at which entrepreneurs are willing to borrow for investment as would result in
an approximation to "full" employment.
Mr. Keynes claims further: (1) that there can be "full" employment only when entrepreneurs make investments sufficient to absorb
any excess of income paid out by entrepreneurs over expenditures on
consumption by income recipients; (2) that the amount of investment entrepreneurs are prepared to make, or their "investment demand for capital," is governed by the relation of their anticipations as
to the yield of additional investment, or what Keynes calls the
"marginal efficiency of capital"2 to the interest rates at which funds
can be borrowed; (3) that the amount which income recipients are
willing to spend of their current income, or their "propensity to consume," a function primarily of the amount of their incomes, 3 determines the quantity of saving; and (4) the rate of interest is determined by (a) "liquidity preferences" and (b) the quantity of cash
available to satisfy such preferences. The quantity of cash is generally assumed to be a constant. I accept most of this as valid in its
general outlines, but I am unable to to accept some of Keynes' account of how these "propensities" operate in practice or his appraisal
of their relative strength.

THE PROPENSITY TO HOARD


Keynes maintains that for centuries back the propensity to save
has been so much stronger than the inducement to invest as to create
a substantial barrier to "full" investment. He finds fault with the
"classical" economists for their alleged neglect of the gulf between
the desire to save and the desire to invest, i.e., for their neglect of
"liquidity preferences." It was a shortcoming of the Ricardian wing
"Anticipated marginal efficiency of capital" would seem to me a more
accurately descriptive label for the concept.
S It is, in my opinion, probably dependent appreciably also on anticipations as to the prospective trend of income, and is surely affected
significantly by amount of accumulated wealth at current valuations
as well as by current income. See infra, "Propensity to Consume," for
further comments on this point.
Z

240

Jacob Viner

of the classical school that in the face of strong criticism they steadfastly adhered to their position that hoarding was so abnormal a
phenomenon as not to constitute a significant contributing factor
to unemployment even during a period of severe deflation. In static
equilibrium analysis, in which perfect price flexibility is assumed and
monetary changes are abstracted from, there is no occasion for consideration of hoarding. In modern monetary theory it is generally
dealt with, with results which in kind are substantially identical with
Keynes', as a factor operating to reduce the "velocity" of money.
There has been, I believe, common agreement among economists that
when price rigidities are important hoarding could present a serious
and continuing problem, and that it is always a significant factor in
the downward phase of a short business cycle. Keynes, however, attaches great importance to it as a barrier to "full" employment at
almost all times, and apparently irrespective of the degree of flexibility of prices.
There are several reasons why "liquidity preferences" loom so
large to Keynes as a source of trouble in the economic process. He
takes it for granted that they are ordinarily so strong for the average
person in control of liquid resources that a substantial interest rate is
required to overcome them; and apparently that they cannot be overcome by any rate of interest if a still higher rate of interest is anticipated in the near future. He assigns to them the role of sole
determinant (given the amount of cash available, which he treats
ordinarily as a constant) of the rate of interest. He believes that the
marginal productivity function of capital and therefore the investment demand for capital have little elasticity. Finally he assumes in
general that nothing can satisfy liquidity preferences except that
"cash" whose quantity is one of the determinants of the interest rate.
We have almost no reliable information about the strength of
liquidity preferences under varying circumstances, and in the absence of statistical information of a genuinely relevant character discussion must be based largely on conjecture. Nevertheless, I venture
to present a series of considerations which, in the aggregate, seem to
warrant the conclusion that Keynes has grossly exaggerated the
extent to which liquidity preferences have operated in the past and
are likely to operate in the future as a barrier to "full" employment.
(a) Keynes stresses the pressure which is exercised by the expec-

Causes of Unemployment

241

tation of a rise in the interest rate on potential purchasers of securities, leading them to postpone their purchases in order to escape
a capital loss. There are, however, in every country large numbers of
investors who have been taught to buy gilt-edge securities on the
basis of their yield to maturity and to disregard the fluctuations in
their day-to-day market values. Even investors of a speculative type
are ordinarily as anxious not to miss a "low" as not to buy too high.
There are many opportunities for investment which are-or seem at
the time to be-of the "now-or-never" type. There is a widely prevalent aversion to the waste of "dead" cash.
(h) Keynes seems to exaggerate the actuarial valuation of postponement of investment during a period of anticipated rise in interest
rates. Rising interest rates are frequently associated with periods of
greater confidence in the security of the investment, as far as payment of principal and interest according to schedule are concerned;
or in the case of equity securities, with periods of more favorable
anticipations of long-run yields. Hence periods of rising interest
rates are often associated with periods of rising rather than falling
prices of securities, especially for equity securities. Keynes seems to
be in error also when he asserts that, abstracting from the risk of default on principal or interest, it will be equally profitable to hoard as
to invest at par in a long-term security paying 4 per cent if the market
interest rate is rising by 0.16 per cent per annum. In the first place,
hoarding and investment in a long-term security are not the only
alternatives. Let it be provisionally granted that hoarding and the
purchase at par of a 4 per cent long-term bond would prove equally
profitable at the end of the first year if the interest rate during that
year had risen by 0.16 per cent. The purchase at the beginning of
the year of a one-year maturity security paying anything over 0.16
per cent would then have been more profitable even if it had to be
exchanged for cash within six months, and even if the short-term
interest rate were also gradually rising by as much as 0.16 per cent
per annum. Secondly, even a purchaser of the long-term 4 per cent security would have been richer at the end of the first year than if he
had hoarded his cash, unless the security were a perpetual bond.
(c) Even if it be granted that liquidity preferences are as strong
ordinarily as Keynes indicates, their operation as a barrier to investment would necessarily be important only if it be assumed (1) that

242

Jacob Viner

liquidity preferences can be satisfied solely by the holding of noninvestment assets, and (2) that the quantity of such assets does not
automatically respond to the demand for them. Keynes takes care of
this second qualification by his assumption that the quantity of
money-in the assumed absence of a positive central monetary control-is constant. Here, indeed, he concedes more than is necessary,
for if liquidity preferences are assumed to be stronger during depressions than during periods of business expansion, then the quantity of money, under such monetary systems as have existed in the
past, varies inversely with the strength of liquidity preferences. But
he does not give adequate consideration to the first qualification.
The satisfaction of liquidity preference on the one hand and of
investment on the other are opposite phenomena only if the range
of assets which can satisfy investment demand corresponds with the
range of assets which can satisfy liquidity preferences, so that it shall
be impossible to satisfy both by the same transaction. If liquidity
preferences can be satisfied by the holding of resources which are not
identical with the "money" whose surrender satisfies investment demand, the satisfaction of the former does not necessarily entail failure
to satisfy the latter. Keynes explains liquidity preference as a wish
to retain one's resources in the form of money. There is no systematic
examination of what is to be included as "money" for this purpose,
but incidentally to his analysis of one particular form of surrender
of liquidity, namely, exchange of money for a debt, he states:
... we can draw the line between "money" and "debts" at
whatever point is most convenient for handling a particular
problem. For example, we can treat as money any command
over general purchasing power which the owner has not parted
with for a period in excess of three months, and as debt what
cannot be recovered for a longer period than this; or we can
substitute for "three months" one month or three days or three
hours or any other period; or we can exclude from money whatever is not legal tender on the spot. It is often convenient in
practice to include in money time-deposits with banks and,
occasionally, even such instruments as (e.g.) treasury bills. As
a rule, I shall ... assume that money is co-extensive with bank
deposits (p. 167, note).

If everything which satisfies liquidity preference is to be included


as money, then money must be broadly defined so as to include not

Causes of Unemployment 243


only demand deposits and time deposits, but also short-tenn securities,
any other assets which are readily marketable without serious risk of
loss through depreciation of value, and even the command over
credit from banks or others. But the conversion of newly acquired
cash into any other form of asset either involves investment directly
or transfers the decision as between hoarding and investment to a
banker or other intennediary between the original saver and the ultimate borrower for investment. If the banker permits his investments
to remain constant while his cash reserves are increasing, or if he
maintains the same cash reserves for idle as for active demand deposits, or for time deposits as for demand deposits, or for deposits as
for banknotes in circulation, then the propensity to hoard which
manifests itself in the maintenance of idle bank deposits does operate
to check investment, but only with the connivance and support of the
banking mechanism.
It may be objected that even if liquidity-preferences operate only,
or in the main, to check purchases of long-tenn securities, they still
operate as a check to investment; because the latter is and must be
largely in durable goods, or in assets far removed from the stage of
the consumers' goods. But the relation between the period of investment intended by the saver and that intended, or in fact resulting,
by the borrowing entrepreneur is not a simple one of necessary
equality. It is highly flexible and approaches to free variability at the
discretion of the borrower. Every money market has an elaborate
machinery for transmuting short-tenn loans into long-term investments and long-tenn loans into short-tenn investments, to suit the
convenience of original lenders and ultimate borrowers. The typical
entrepreneur will shift from long-tenn to short-tenn borrowing, or
vice versa, even though the time period involved in the particular
operation is unchanged, or (as often) unknowable in advance. He
may also be able to shift from long-tenn to short-tenn investment
if the interest rate at which the latter can be financed is much lower
than that at which he can conduct admittedly long-tenn borrowing. If
savers have a 5 per cent per annum preference for cash over investment in Ie-year bonds but only a
per cent preference for cash over
time deposits or short-tenn securities, and if entrepreneurs want
funds for 10 years and are unwilling to incur the sacrifice of their
own liquidity which would be involved in the attempt to finance Ie-

244 Jacob Viner


year operations with, say, 3-month borrowings, middlemen will step
in who are prepared to lend on long-term funds which they have
borrowed on short-term. The modern money market is fortunately
equipped to some extent with procedures for satisfying liquidity
preferences without providing genuine liquidity.
Cd) The propensity to hoard exercises its inHuence as a restraint
on investment through its tendency to raise interest rates. But in
what seems to me the most vulnerable part of his analysis, his explanation of the determination of the rate of interest, Keynes assigns to the desire for cash for hoarding purposes a grossly exaggerated
importance.
Keynes denies the validity of the "classical" doctrine that interest is
the reward for saving and is directly determined by the supply
schedule of savings with respect to the interest rate and the investment demand schedule for capital, and his exposition leaves the impression that the interest rate is not dependent to any important
extent on these two factors. He denies that interest is the "reward"
for saving on the ground that, if a man hoards his savings in cash,
he earns no interest, though he saves just as much as before Cpo 167),
and claims that, on the contrary, it is the reward for surrender of
liquidity. By analogous reasoning he could deny that wages are the
reward for labor, or that profit is the reward for risk-taking, because
labor is sometimes done without anticipation or realization of a return, and men who assume financial risks have been known to incur
losses as a result instead of profits. Without saving there can be no
liquidity to surrender. The saver who has no concern about liquidity
gets the same reward as the person who saved with liquidity as his
initial objective but is persuaded by the interest rate to lend; and the
return is granted for loans irrespective whether it is reluctance to
postpone consumption or reluctance to surrender liquidity which
keeps the supply of funds for investment down to the level at which
borrowers are willing to pay the prevailing rate of interest for it. The
rate of interest is the return for saving without liquidity.
Keynes explains the rate of interest as determined by the schedule
of liquidity preferences and the available quantity of money, the
prevailing rate of interest being simply that price for the sacrifice of
liquidity at which the desire to hold cash is equated with the quantity of available cash Cpo 167). The rate of interest determines the

Causes of Unemployment 245


amount of investment, given the investment demand for capital; but
a change in the investment demand for capital will not affect the
interest rate "if nothing has happened to the state of liquidity preference and the quantity of money." (See especially the figure on p.
180, and the text on p. 181).
There have been previous attempts to discover a basis on which
the interest rate could be held to be determined independently of
the demand for capital, the level of wages, and other important elements in the economy, but the growing recognition of the basic interdependence of all the important economic variables has led to
widespread scepticism that any such attempt could succeed. In
Keynes' present attempt the fatal flaw is, to repeat, the exaggerated
importance attributed to hoarding. In his discussion of liquidity
preferences Keynes distinguishes between the desire for cash for use
in the current transaction of personal and business exchanges, and
the desire for cash as a security against loss from unsuccessful investment. As I have already argued, the latter consideration should
not operate as a barrier to short-term investment, and while it may
induce a high long-term interest rate, it will be compensated for in
part by a shift of borrowing to the short-term market. The pattern of
behavior of the desire for transaction liquidity is probably very largely
the inverse of that of security liquidity, or hoarding proper. As D. H.
Robertson points out in his contribution to this symposium, the transactions desire for cash is for cash to be used and not for cash to be
held unused. It must therefore vary positively with the volume of investment, of income, and of expenditures for consumption. In so far
as it consists of demand for cash from entrepreneurs for business uses,
it is but a reflection of their investment demand for capital. In so far
as it is a demand for cash from consumers who are living beyond
their current income, it is the demand for consumption loans of
older theory. Whatever its origin, demand for cash for transaction
purposes is, dollar for dollar, of equal influence on the rate of interest
as demand for cash for hoarding purposes. The demand for capital
and the propensity to save (which is the reciprocal of the propensity
to consume) are thus restored-though, I admit, in somewhat modified and improved fashion-to their traditional roles as determinants
of the rate of interest.
While (to repeat again) relevant statistical information is scarce,

246 . Jacob Viner


what we do know about the holders of cash balances in the United
States points strongly to the importance of the transactions motive
for liquidity and to the relative insignificance in ordinary times of
hoarding. It is the corporations, institutions, and governments that
hold at all times the bulk of the cash balances, especially if savings
deposits are excluded as constituting investments rather than cash.
Moreover I suspect (I know of no data on the question) that at least
in prosperous times the savers-those who add each year to their
estates-who are supposed by Keynes to be a source of so much
trouble because of their hoarding propensities, typically hold in
cash a smaller percentage of their incomes, let alone of their total
resources, than do the spenders. The former have investment habits,
and abhor idle cash as nature abhors a vacuum. The latter hold cash
until the bills come in for settlement. It would at least be interesting
to know whether these are facts or fancies.
The importance of the transactions demand for cash makes it easy
to explain a whole series of historical phenomena which do not fit
into Keynes' theory. Because the demand for cash for business use
varies positively with the investment demand for capital, and the
demand for cash for personal use varies positively with the level of
income and of expenditures for consumption, there is no need for
treating as a perplexing puzzle the facts that business is active when
interest rates are high and slack when interest rates are low, and
that the quantity of money and the interest rate are historically correlated positively rather than negatively. There is an important
stabilizing influence, moreover, in these circumstances. During a depression entrepreneurs and spenders release some of the cash to
supply the demand of hoarders for security, and during an expansion
of business the absorption of cash by business and by spenders, serving as it does to raise the interest rate, keeps the expansion from going
beyond bounds; or, Keynes would say, from even approaching reasonable bounds.

MONEY WAGE FLEXIBILITY AND


VOLUME OF EMPLOYMENT
Keynes expresses sweeping dissent with the "classical" doctrine
that money wage rigidity is a major cause both of cyclical and of
secular unemployment, although he freely grants that in general in-

Causes of Unemployment 247


creased employment must mean lower real wages. He maintains that
labor strongly resists money wage reductions but takes reductions in
real wages much more calmly, and therefore that even if money wage
reductions were logically a remedy for unemployment they would
not be a practicable one. His view is that a lowering of money wage
rates, unless it proceeded simultaneously and uniformly all along the
line, would chiefly alter the relative rates of wages of different labor
groups. It would not be likely to increase the aggregate volume of
employment of labor, and on the balance of probabilities would be
more likely to reduce it. He does not discuss the effects on employment which would result from pressure from labor for increases in
money wages, or from increases of money wages made voluntarily on
the part of employers, whether for humanitarian reasons or because
of belief that high wages mean prosperity or in response to public
opinion.
Keynes presents his own position mainly in terms of a criticism of
a theory which he imputes to the "classical" economists, according to
which a reduction of money wages and a simultaneous corresponding
reduction in prices would increase employment because the same
volume of monetary expenditures would purchase a greater physical
output of commodities. He easily demolishes this by pointing out
that, if money wages paid out were to fall in amount and investment
by entrepreneurs (measured in wage units) did not increase, the
amount of money income available for expenditures would fall to
an equivalent extent. His discussion of the effects of the wage reduction on the volume of investment is mainly in terms of its influence on the expectations of entrepreneurs as to the future trend of
wages, and he concedes that if entrepreneurs are led to expect further
changes to be in an upward direction its effect will be favorable. He
urges, however, that "it would be much better that wages should be
rigidly fixed and deemed incapable of material changes than that
depressions should be accompanied by a gradual downward tendency
of money-wages" (p. 265).
This does not meet the argument for wage reduction-or rather
money cost reduction 4--during a depression which I had understood
4

From the point of view of effect on output, the reduction of any part
of variable costs is dollar for dollar of the same importance as the reduction of any other part of such costs, and it is only as against re-

248 . Jacob Viner


to be the prevalent one in recent years. In this other doctrine, factor
prices are to be reduced, but not, or not in the same degree, the
prices of consumers' goods. In Keynes' analysis perfect and active
competition is assumed, and prices are supposed to fall immediately
and in full proportion to the fall in marginal variable5 costs. If this
occurred, and output remained the same, prices per unit would fall
in greater absolute amount than would average variable costs,6 and
even more, if current labor cost were a negligible element in the
fixed costs, than would average aggregate costs. The profit status of
entrepreneurs would then be less favorable than before. What I
understand to be the current doctrine is different. It looks to wage
reductions during a depression to restore profit margins, thus to
restore the investment morale of entrepreneurs and to give them
again a credit status which will enable them to finance any investment they may wish to make. It relies upon the occurrence of a lag
between the reduction in wage rates and a response in reduced
volume of sales at the previous prices, during which interval entrepreneurs find prices to be higher than marginal costs and extensions of output therefore profitable, provided buyers can be found
for the increased output. Increase in expenditures to restore depleted inventories and to replace inefficient equipment is relied upon
to increase pay rolls sufficiently to provide the incomes with which
the increased output can be bought, and the gain in employmentand in security of employment for those previously employed-is
duction of outstanding fixed costs, to the extent that they also do not
consist of labor costs, that there is anything to be said for reduction
of labor costs in preference to other costs. But from the point of view
of the effect on the employment of labor, the reduction of labor cost
is more favorable than the reduction to an equivalent amount of any
other cost, because it will tend to lead to a substitution of labor for
other factors, though it will not be as favorable as the reduction of
both or a fortiori of all costs simultaneously and in the same proportions.
II Keynes distinguishes between "factor costs" and "user costs," the two
combined comprising "prime" costs. By user costs he means the
amounts paid out to other entrepreneurs for purchases from them and
sacrifices incurred (extra wear and tear presumably) in employing
equipment instead of leaving it idle. He claims that economists have
generally equated supply Erice with marginal factor cost, ignoring
user cost, whereas it should be equated with prime cost. I see no point
in the distinction between purchases from entrepreneurs and direct

Causes of Unemployment 249


expected to release for expenditure the emergency reserves of the
wage earning class. On the assumption that a large part of an
entrepreneur's expenditures are ordinarily of the postponable class in
the sense that they can be deferred without forcing a reduction of
the scheduled rate of current output, even though not without increasing the current cost of production, and on the further assumption that operations at a loss are conducive to the postponement of
every expenditure not essential for current operation, the supporters
of this doctrine maintain that recovery of a profit margin can lead
for a time to an increase in entrepreneurs' expenditures many times
the increase in their net income, or, alternatively, the reduction in
their net loss. They do not contend that this is certain to occur, but
on the ground that the chief factor in governing the action of entrepreneurs with respect to postponable expenditures is the current
profit status of their operations as compared to their immediately
preceding experience, they say that it is a reasonable probability.
Where external pressure on prices in the face of rigid costs has been
an important factor in the depression, they also expect a favorable
influence on the volume of employment from the effect of a wage
reduction on profits and therefore on the volume of postponable
expenditures, rather than from its effect on prices. While Keynes'
analysis provides materials for strengthening this doctrine at a number of points, I cannot find in it any refutation of its general validity.

PROPENSITY TO CONSUME
Mr. Keynes himself tells us that the functional relationships of
the various economic variables are more complex in fact than is
formally recognized in his analysis. Simplification of this sort is inevitable, if analysis is to proceed at all. In the case, however, of
Keynes' "propensity to consume" function, it seems to me that the
purchases of the services of the factors. What is the point in distinguishing between the cost of coal to a steel mill according as it is
bought from an outside mine or produced in its own collieries? Where
is the line to be drawn between entrepreneurs and "factors"? I am
sceptical as to whether any economists have, explicitly or by implication, excluded cost of purchased materials or depreciation of equipment through use from the costs supposed to determine supply price.
6 Because marginal costs would fall in the same proportion as average
variable costs but would be greater in amount per unit than average
variable costs.

250

Jacob Viner

simplification has been carried further than is necessary to prevent


the analysis from becoming entangled in its own complexities, and
further than is permissible if the concept is to be used fruitfully in
the analysis of the short cycle.
Keynes explains the propensity to consume as a functional relationship between the amount of consumption measured in money
wage units and the amount of income similarly measured. On the
assumption that income in terms of money wage units corresponds
substantially in its variations with the variations in level of employment, it is concluded that income, consumption, and level of employment are related to each other in a simple pattern. Writing
C w for amount of consumption in wage units and Yw for income in
wage units, and accepting as a close approximation that Yw is a
unique function of the level of employment, he states the propensity
to consume function as: Cw=X(Yw) (p. 90).
Keynes lists a number of factors, (p. 96) "subjective" and "objective," which might affect the value of x, Yw remaining constant,
but he assumes in general that the "subjective" factors remain constant, at least over short periods, and that, given Yw, X depends only
on changes in the "objective" factors, which in the aggregate he
takes to be of minor importance as compared to changes in Yw'
Several "objective" factors which he does not appear to have taken
into account seem important enough in the short cycle to be deserving
at least of mention.
Keynes believes that, apart from the effect of a change in the wage
unit on the distribution of income between entrepreneurs and rentiers, who might have different propensities to consume, he has
made adequate allowance in his formula for changes in expenditure
resulting from changes in the wage unit by measuring both consumption and income in wage units. This disregards the possibility
that, for short periods at least, the distinction which Keynes' makes
in his supply function of labor between the response of labor (I) to
changes in real wages accompanied by corresponding changes in
money wages, and (2) to changes in real wages resulting from the
changes in the prices of wage goods, money wages remaining the
same, may have a parallel in the propensity to consume function.
The response of consumption to a reduction in real income may be,
for a time, substantially different if the reduction takes the form of a

Causes of Unemployment

25I

decrease in money income, prices remaining the same, from what it


would be if money income remained the same but prices increased.
Mr. Keynes claims that in general rich countries are worse off than
poor countries with respect to avoidance of "involuntary" unemployment because of the lesser propensities to consume in the former
than in the latter, and thus the greater potential importance of
hoarding. Since I would contend that over long periods, given a
Bexible price system, the propensity to consume will affect the rate
of capital accumulation rather than the volume of employment, I will
confine myself to a consideration of the comparative situation of the
rich and poor countries with respect to the short cycle. The possession of large accumulated resources should operate to level out the
rate of consumption in the face of Buctuations in income, and therefore to check both the downward and the upward phases of the
cycle. Corresponding to the charges against the entrepreneur's budget
which are fixed in aggregate monetary amount regardless of current
output, there are in the ordinary consumer's budget items of monetary expenditure which are fixed for a time, very much regardless of
changes in his money income as far as reductions therein are concerned, and which tend to be increased only as the result of careful
deliberation in response to anticipation of a change of some duration
in the individual's economic status. Aside from the probability that
such fixed charges are ordinarily a greater proportion of the expenditures of the rich than of the poor, the poor in times of severe depression have a partial means of escape from them, in the form of
defaults, to which those with resources subject to levy cannot resort.
What this amounts to is that C w should be treated as a function not
only of Yw, but also of the amount of accumulated resources measured
in wage units held by the individual. In so far as the possession of
resources operates in the manner suggested here, wealth becomes a
stabilizing rather than a disturbing factor. The explanation of the
apparently indisputable fact that the cyclical disturbances are more
severe in rich than in poor countries would then have to be sought
elsewhere than in the differences between rich and poor in propensities to consume. My own guess is that it is to be sought largely
in the differences between the cyclical behavior of rich and poor
with respect to the disposition of the income which they do not
spend. The rich hoard only during depressions and dishoard for in-

252

Jacob Viner

vestment during prosperity, whereas the poor hoard some of their


emergency reserves during prosperity and dishoard during depression.
Mr. Keynes says that a fundamental psychological law, upon
which we have a right to depend both on a priori grounds and on the
w

.
. t hat dC
. .I.e.,
0 f expenence
baSIS
IS
dYw IS
pOSItIve
an d Iess t han umty;

that in terms of wage units consumption varies in the same direction


as income, but in smaller absolute amount than income (p. 96). This
seems altogether reasonable. It leaves unanswered, however, a question of some interest: does Cw ever, except perhaps under war
conditions, exceed Yw? Since the community excess of Yw over
Cw constitutes new investment, if Cw never exceeded Yw there would
be continuous, though fluctuating, accumulation of capital resources,
even through the depths of depression. Mr. Keynes apparently must
believe that for the world as a whole the CW's must often and substantially exceed the Yw'S, for he holds that in spite of "several
millenia of steady individual saving" the world is poor in accumulated capital assets.7 But what evidence there is seems to indicate
that, if any acceptable mode of measuring physical amount of capital
could be found and applied, it would show that the western world
has been getting wealthier fairly steadily during say the past century and a half, not only in terms of aggregate resources but per
capita, in spite of a three- or four-fold increase of population.
In connection with the propensity to consume concept, as with
most of Keynes' concepts, the question arises in my mind how these
concepts would have to be restated in order to provide specifications
for the construction of statistical series by which his conclusions as to
the nature and mode of behavior through time of the various functions could be inductively tested, and I regret that no suggestions of
this sort are provided in this book. I am disposed to support Mr.
Robertson in his claim that concepts expressed in more "monetary"
terms, and expressions for the relationships between variables
7

"That the world after several millennia of steady individual saving,


is so poor as it is in accumulated capital-assets, is to be explained, in
my opinion, neither by the improvident propensities of mankind, nor
even by the destruction of war, but by the high liquidity-premiums
formerly attaching to the ownership of land and now attaching to
money" (p. 242).

Comment on My 1936 Review' 253


which make specific allowance for time lags instead of assigning uniform time units to all the variables, have for purposes of a priori
analysis some points of superiority over Keynes' "propensity" concepts expressed in terms of a single time unit. For purposes of
inductive verification, assuming that the statistical data available will
ever be in a form relevant to the answer of important questions, it
seems obvious to me that the analysis would have to be extensively
restated in terms of directions and degrees of time lags.

Comment on My 1936
Review of Keynes'

General 'Theory
[1963J

My

1936 REVIEW OF KEYNES' General Theory, here revived from the dead, no doubt had limitations of which I was then
unaware and continue to be unaware. It also had, however, some
deliberate limitations which were imposed on it by the circumstances
of its publication. It was written as part of a symposium under the
editorship of Professor Taussig. He encouraged me to include in
my contribution a paragraph or two appraising the importance and
quality of the book as a whole, but asked me aside from this to concentrate on a discussion of Keynes' theory of the causes of changes
in the volume of employment. I interpreted Keynes' theory as in
fact, whatever its intent, a theory only of the short-run determinants
in changes in employment, and kept my comments within that
limited framework. I consequently refrained from any attempt to
assess the contribution of the General Theory to policy formulation,
or to long-run analysis, or its serviceability as a foundation for

254

Jacob Viner

long-run forecasting. I would not have denied that it contained, or


probably contained, materials which could be profitably embodied,
perhaps after appropriate adaptation, in long-run theorizing, but I
would have conceded that I had only hazy notions of how that task
could be performed and that its performance would require new
insights and the invention of new analytical tools. I refrained also
from discussing the various considerations which keep a theory of
employment from being automatically a satisfactory theory of output.
Another deliberate limitation of my review was its abstention from
any discussion of Keynes' ventures in the General Theory into
Dogmengeschichte, although it was this phase of his book which I
felt myself best qualified to appraise and regarded as most vulnerable. As a historian of thought in areas in which he was emotionally
involved as a protagonist and prophet, Keynes seemed to me to be
seriously lacking in the unexciting but essential qualities for the
intellectual historian of objectivity and of judiciousness. Even when
he was engaged in selecting those upon whom to bestow laurels for
having in some degree anticipated his discoveries, his selection
seemed to me then, and still seems to me now that I have acquired
more knowledge of the older literature, often to have been random
when not eccentric. I would nevertheless have agreed with him on
what I take to be his major complaint against the main line of
"orthodox" English economics from Ricardo to Marshall and Pigou
that it failed miserably as an explanation of the causes of (nonsecular) fluctuations in output and employment, although I would have
seen their failure more in their not making any serious attempt at
such explanation than in the erroneous character of their product.
I had myself been trained in that tradition, but by 1930 or 1931 I had
realized that to an appreciable extent it gave no light, or misleading
light, on the nature and origins of major economic fluctuations. I
took it for granted, however, that the unsuitability of its standard
analytical procedures for analysis of "depression" and "boom" was
the result not of general stupidity or perverse bias on their part but
of lack of profeSSional interest in and of dedication to short-run
analysis, a lack even more conspicuous in two other great schools of
economic theory, the Austrian School and the Lausanne School.
Keynes, however, was writing in a white heat of revolt from the
English classical tradition, and treated its neglect of or deliberate

Comment on My 1936 Review 255


abstraction from short-run phenomena as deliberate and stupid
denial of their existence, while himself abstaining from acknowledging or exploring the possibility that in the short-run phenomena there
were inherent forces which tended to produce quite different longrun than short-run consequences.
Almost the only references to my review that I know of were by
Keynes himself and by two avowed Keynesians. All of them, of
course, took issue with me on some major points, but none of them,
I am glad to say, seemed to regard my review as hostile on the whole,
in fact or intent. As best I can remember, I found the General
Theory dazzling in its brilliance, and, if interpreted as primarily
and admittedly a contribution to short-run theorizing, a veritable
landmark, in both its positive and negative or critical aspects, in
the history of our discipline. If I failed to make this clear, it was the
consequence of inadequacy of exposition, not of intent. Myappreciation of the quality and originality of its analysis as short-run analysis
has grown since rather than shrunk, as I have struggled with its
difficulties and as my understanding of it has improved thanks to
the efforts of its many dedicated exegetes and perhaps also to my
own later efforts, with the assistance of my students, to master its
intricacies of thought and peculiarities of vocabulary. Within the
boundaries of short-run analysis, and especially of short-run depression analysis, I regard the claim made for its having achieved a
"Keynesian revolution" in economics as a permissible manifestation
of an enthusiasm for which there is substantial justification, subject,
however, to possible rival claims on the part of the "Swedish School"
which I am not qualified to assess. On its adequacy as long-run
analysis, and especially on its denial that (in a truly competitive
economy) there exist powerful automatic forces which in the long
run, if not counteracted by perverse governmental intervention, will
restore "equilibrium," I still remain sceptical at least, with some
propensity to be hostile. It may be that these long-run equilibrating
forces operate with painful slowness. Real historical time, moreover,
does not exclude those short-period disturbances which traditional
theory tended to overlook or to segregate in a pound of abstraction
whose characteristics it did not closely or penetratingly scrutinize.
A year or two of the Great Depression sufficed to convince me,
"orthodox" theorist though I was by training and temperament, that

256 . Jacob Viner


a substantial shift of emphasis in economic theory to short-period
analysis was called for, and that the greater the degree of crisis the
greater was the relative importance of the forces whose impact was
predominantly short-run in character. Probably also by 1936, and
certainly later, there was no reluctance on my part to concede that
even analysis on strictly long-run assumptions had much to gain from
inclusion of considerations such as Keynes was emphasizing. Nevertheless, without denying that "in the long-run we will all be dead," I
still find "after us the deluge" an unsatisfactory guiding maxim for
mankind in general or for economic theory in particular. What we
should concede, with acknowledgments to Keynes, is that in counting
our putative long-run blessings we should not disregard as trivial the
short-run woes that they may have involved merely because they
were, or may have been, present for a period substantially short of
infinity, a procedure which is tolerable, if anywhere, only in theological discourse.
Within the field of short-run theory, Keynes-and his Cambridge
associates-has great claims to originality. Many of the elements in
his system were by no means new, or even of recent origin, as others
have demonstrated, and as Keynes was even over-ready to concede
provided it was not the "classical school" on whose behalf the anticipation was claimed. I will shortly cite a few instances of such anticipation, but will confine myself to out-of-the-way texts that have
perhaps escaped notice hitherto.
From the seventeenth century on, there was an almost continuous
stream of expositions of the view that unemployment and sluggish
trade were the consequences of the failure of purchasing power to
keep step with the expansion of productive power. Keynes found
to his delight that the literature of mercantilism was saturated with
this idea; it persisted into the nineteenth century and did not need
the prodding of Keynes to continue to Hourish in the twentieth
century outside the realm of orthodox economic theory. One of the
many curious items in this literature was a treatise of twelve hundred
pages on this single theme which paused for breath only for paragraphing and thus constituted what was probably the longest chapter by far on record. 1
1

David Laurie, New System of Finance, Glasgow, 181,.

Comment on My 1936 Review' 257


Among the ideas of narrower scope on which Keynes had anticipators, in some instances a host of them, were: the explanation of
interest as the reward for surrendering (cash) liquidity;2 the dependence of employment and output on the sum of consumption
expenditure and investment expenditure; a loan-fund theory of the
determination of the interest rate; the absence of inBationary pressure
as long as there are involuntarily-unemployed resources;3 and even
the necessary equivalence of (realized or effective) investment with
(realized or effective) saving, so that the outcome of increased effort
to save may be less rather than more saving. 4
Even, however, if few of the bricks Keynes used to build his theoretical structure were completely novel, he refined and remoulded
them to more precise and rigorous specifications. More important,
Keynes was strikingly original in the skill and comprehensiveness
with which he combined these separate ideas or theorems into a
complex and coherent system.
Keynes gave to his integrated system the label "General Theory"
because he thought its claims to originality lay largely in its demonstration of the possibility of "equilibrium" at any level of employment, with equilibrium at "full employment" only a special and not
very probable case. This is a legitimate use of the term "general,"
but it should not be confused with its meaning as used in Walrasian
theorizing, which is "general" in the sense that the Walrasian purports to account for all the significant interrelations between variables or functions. Even in this latter sense, however, Keynes'system,
"Since the Statute against usurie [13 Eliz. c. 8 (1571)] did tolerate
ten upon the hundred everie man hath found such a sweetness in
that usurial gaine that smalle store of money is hoorded up." (From
a 1594 manuscript quoted by H. M. Robertson, Aspects of the Rise of
Economic Individualism, Cambridge, Eng., 1933, p. 198).
8 E.g., Hugo Bilgram, Involuntary Idleness, Philadelphia, 1889.
4 Mter a discussion of "the propensity of man to lay up in store or to
consume:" "Savings do not become capital, unless they are employed
reproductively; and it is the difficulty of finding modes of so applying
them, not the strong inclination of man to spend all that he can obtain, that opposes a bar to the rapid accumulation of capital. Any
plan, therefore, of increasing the capital of a country by an artificial
diminution of the consumption, proceeds upon a supposition of very
dubious truth. You may, by such means, diminish the amount of the
unproductive consumption of the country, but you will not necessarily increase its productive consumption. The more probable result
2

258 Jacob Viner


as a short-run system, was in some, though not all, respects, a more
"general" system than any other earlier short-run system. I find it a
good working rule to test the degree of generality, in the Walrasian
sense, of an economic system by counting the ratio of variables to
constants in the system. On this, as on other grounds as well, neither
Walras's system as a long-run system nor Keynes' system as a shortrun one is as "general" as it could ideally be made to be. In my review of the General Theory, most of my criticisms related to this
issue, although I did not make this point explicit. The General
Theory relies too heavily for its conclusions on the treatment of
variables as constants, that is, on rigidities, and on the treatment of
structures, or complex matrices, such as the wage structure, the price
structure, the rate of interest structure, each marked by complicated
internal as well as external interrelations, as if they were simple,
single-layer, concepts. It disturbs me, therefore, when I think I see
"Keynesians," and especially Keynesian econometricians and geometricians, endeavor to develop and improve upon the contribution of
the General Theory by stripping it of some of the Walrasian-type of
generality it does possess, instead of making it more Walrasian in its
analytic character. I think I have grounds for believing that I am here
closer to Keynes' beliefs, at least as they were in 1946, than are some
of his avowed disciples, although in the conversations I had with him
in that year neither of us made any use of this specific way of
formulating the analytical issue.
One manifestation of Keynes' reliance on too restricted a set of
variables in the General Theory was his treatment of the rate of
interest as the product of the relationship of the demand for
"money" to the stock of "money," with "money" defined or identified, in the General Theory, and in later writings, as "cash," as
"hoards," as "idle balances," or expressly left to be defined by the
reader to his taste. I found fault with this in my review as gross
oversimplification, but I should have given my criticism a wider
will be, either that the amount of annual production will be lessened,
or that a proportion of the unproductive consumption will be shifted
from one class of commodities to another [presumably because one
man's saving will be offset by another man's dissaving]." An anonymous review of James Mill's Elements of Political Economy, The
British Review, and London Critical Journal, xix ( I 822), 158.

Comment on My 1936 Review 259


scope, more range, although perhaps no more weight or intensity,
than I did. The flow of funds to borrowers is restrained not only by
the interest charged but also by "rationing" or limiting availability.
In England the grant of a line of credit, in the United States a commitment to lend, is from the point of view of "liquidity" a close
equivalent to an actual loan, but neither gets into statistics of the
stock of "money," or into the diagrams. In the United States the
"compensatory balances" which banks. often require borrowers to
maintain are spurious parts of the statistical "money" stock when they
exceed the maximum working balances which borrowers would
maintain if they were free to choose. Probably everywhere in some
degree banks control their volume of lending by rationing and
changes in "availability" of loans as well as by changes in discount
rates, and in England before the 1850'S and in Canada before
World War I the maintenance of discount rates by law or custom
at substantially constant levels was an additional factor, which receives no attention in the General Theory, weakening the connection between variations in the stock of money and variations in the
interest-rate structure.
In his reply to the Quarterly Journal of Economics symposium on
the General Theory, Keynes dealt with me gently and generously,
and took sharp issue with me on only two closely related points: my
complaint that he had grossly overemphasized "hoarding" as a major
element in "liquidity preference" and thus as an influence on the rate
of interest, and my claim that "in modern monetary theory the
propensity to hoard is generally dealt with, with results which in
kind are substantially identical with Keynes', as a factor operating
to reduce the 'velocity' of money."5
To take first the question of the role of "liquidity preference" as a
force operating to create, or sustain, or raise the level of, the interestrate structure, Keynes in his reply to me wrote:
Again, when Professor Viner points out that most people
invest their savings at the best rate of interest they can get and
asks for statistics to justify the importance I attach to liquiditypreference, he is overlooking the point that it is the marginal
M. Keynes, "The General Theory of Employment," The Quarterly
Journal of Economics, LI (1937), 209-225. The direct references to
me are on pp. 2 I 0-2 I 2 and p. 2 I 9, note.

6 ].

260

Jacob Viner

potential hoarder who has to be satisfied by the rate of interest,


so as to bring the desire for actual hoards within the narrow
limits of the cash available for hoarding [po 2II].
When Professor Viner charges me with assigning to liquidity-preference "a grossly exaggerated importance," he must
mean that I exaggerate its instability and its elasticity. But if he
is right, a small decline in money-income would lead, as stated
above, to a large fall in the rate of interest. I claim that experience indicates the contrary. [po 219, note].
"Liquidity preference" was Keynes' term, and when I charged him
with exaggerating its importance, I had reference to the meaning he
gave it in many passages in the General Theory, which is also the
meaning he gives it in the two passages from his reply to me which
I quote above, namely, a preference for "hoards," for money deliberately to be kept idle, as the consequence of the "propensity to hoard."
Keynes was surely also using "liquidity preference" in this narrow
sense when he said: ''The rate of interest obviously measures ... the
premium which has to be offered to induce people to hold their
wealth in some form other than hoarded money. . . . The rate of
interest is the factor which adjusts at the margin the demand for
hoards to the supply of hoards" (p. 216). I did not then deny, and
I would not deny now, that "liquidity preference" broadly conceived, so that it included preference for "hoards" only as one element, and probably a very minor one, in the liquidity complex, was
an important determinant or "regulator" of the interest-rate structure. Keynes in his reply did state "nor is my argument affected by
the admitted fact that different types of assets satisfy the desire for
liquidity in different degrees," but he must have meant, if he was
being relevant, that his argument as to the importance of the propensity to hoard did not need qualification because of this capacity
of other assets than money to satisfy "liquidity preference."
The passages I have quoted from Keynes' reply have continued to
puzzle me, but as I was preparing this paper I at last saw what I now
think Keynes meant by them, why he thought them an adequate
reply to me, and how he had originally reached his theory of the
importance of the propensity to hoard as the causal factor regulating, with the stock of money, the rate of interest. It all rests, I fear,
on a basic misunderstanding or misapplication of marginal analysis

Comment on My z936 Review 26z


of the Marshallian kind, in a pattern of error which has occasionally
been perpetrated in the textbooks and with which I have been
familiar for more years than I can remember.
In a statement of the processes of equilibration in a competitive
market, the conditions of (or at) equilibrium are not to be confused
with the causal forces producing that equilibrium. It is not the marginal contribution to supply, or to demand, of an individual, or of a
minor sector of the supplying agents or the demanding agents, or
even of all the supplying or demanding agents in the aggregate,
which determines what the equilibrium level of price shall be, but
the whole supply functions and demand functions. If a distinguishable sector of demand (or of supply) is quantitatively a minor
fraction, under any likely circumstances, of the aggregate demand
(or supply), it cannot be a major or an important determinant of
the equilibrium level of price. The whole equilibrium complex, once
established, will determine for that sector what quantity it will take,
or provide, on the basis of equality between the equilibrium price
and that sector's marginal demand (or supply) price. It will, however, be the equilibrium price determined by the market as a whole
which will (predominantly) regulate where the margin is located
for the sector in question, and not the other way around. For a
homogeneous commodity in a well-ordered competitive market, it
will not make sense, even theoretically, to talk of equilibrium between the demand of a particular sector of aggregate demand (e.g.,
the demand for money to hoard) and the supply of a particular
sector of aggregate supply (e.g., the supply of money for hoarding
purposes). In almost every respect possible, Keynes' theory of the role
of "the propensity to hoard" as regulator of the interest rate contradicts these propositions. In especially sharp contradiction is Keynes'
emphasis, in his reply to me, on the importance of the degree of
elasticity and stability of the demand for money to hoard, if such
demand is, at almost any conceivable range of interest rates, small
in relation to the total demand for money. This is where the need
for "statistics" that I referred to and that Keynes rejected enters into
the picture.
In the only comment on this interchange between Keynes and myself that I can recall seeing in print, Lawrence R. Klein reports me
as not agreeing "with the theory of liquidity-preference because he

262

Jacob Viner

[that is, I] thought that the transaction motive has as much influence
on the rate of interest as the speculative motive." What I said went
further than this: the hoarding motive was a minor force as compared to all the other motives taken together which influenced the
demand for money. Klein thought that this question had been answered "by the events of recent years." The answer, I presume, was
supplied by the statistical record of what went on in those years, and
Klein found it unnecessary to state so obvious a fact as that that
answer confirmed Keynes' position. 6 I am unfamiliar with that
record and cannot therefore contest any inferences drawn from it,
but I at least welcome the implied agreement with me that a "propensity to hoard" theory of the determination of the level of the
interest-rate structure needs supporting quantitative data.
With respect to the existence of liquid assets other than money, of
"near-money," of "half-hoards,"7 possessing in some degree "liquidity
preference" satisfying-power, aside from the question of their role in
the determination of interest rates I would continue to press the point
that the accumulation of such assets constitutes simultaneously both
a satisfaction of liquidity preference and investment, so that the two
are not, as they are in the case of "hoards," functional opposites. I
would now add that there is no fixed or stable scale of liquidity for
different classes of assets, and that depending on circumstances, as,
for example, the existence of fear of extreme inflation or of exchange
depreCiation, or the loss of confidence in the solvency of banks, the
"propensity to hoard" may become a propensity to hoard anything
except domestic money or claims to such money.
Keynes' disapproval of my attachment to the "velocity" concept as
a tool of analysis, with which Klein heartily expressed agreement, is
certainly supported by the record since the publication of the General
Theory; velocity analysis seems to have been totally abandoned by
all except one minority group of monetary theorists. I am impressed,
however, by the quality of the products of this group. I note also
that in his one endeavor to define the "money" which plays so
Lawrence R. Klein, The Keynesian Revolution, New York, 1950,
pp. 101-102.
1 Thomas Gisborne, "Accumulations of Capital," Quarterly Review,
LXXII ( I 847), 206-23 I, an article which contains much of interest
for the history of liquidity preference theorizing, makes pertinent use
of this tenn.

Comment on My 1936 Review 263


fundamental a role in his "propensity to hoard" theory of interest,
Keynes was not able to dispense with "velocity" as a criterion of
what was and what was not "money" for the purposes of this theory.
It is conceivable to me that if he had ever tried to locate precisely
the line between money and not-money, "velocity" would have
turned out to be almost the sole criterion.
The General Theory had policy objectives very much in mind, but
in the main concentrated on providing theoretical foundations for
the formulation of policy and the selection of policy instruments
with respect to the problem of unemployment. To keep within my
assignment, and also because my entanglement with the U.S.
Treasury at the time, a time of some internal controversy within
the Administration with reference to fiscal and monetary poliCY,
inhibited me from public discussion of related policy issues, I also
refrained from discussing policy except for a warning that there was
an inflationary bias in Keynes' thinking. Aside from details, however, I would have accepted the policy implications of the General
Theory in the monetary and fiscal field were they presented as relating to short-run or cyclical fluctuations in employment but were not
extended without further analysis to long-sustained or chronic unemployment. Seymour Harris has said of Keynes, with reference, I
believe, to the early thirties:
He preached what has become a commonplace since he
wrote: Government should spend more and tax less in depression; and spend less and tax more in boom. These simple
truths were discoveries of Keynes's which had to be repeated
and repeated hundreds of times before they made the required
impression. s
This formula may have been a discovery of Keynes, but I used it
at least as early as the summer of 1931, and I don't think I derived it
from Keynes, with whose journalistic writings I then had little
acquaintance. The idea was then a commonplace in my academic
surroundings of the time, and I cannot recall that any of my Chicago
colleagues would have dissented, or that they needed to learn it from
Keynes, or from me. In any case, a talk I gave at the Institute of
8

Seymour Harris, John Maynard Keynes, Economist and Policy Maker,


New York, 1955, p. 149. I do not know whether this formula is a
close paraphrase of something Keynes wrote.

264 Jacob Viner


Politics at Williamstown on August 14, 1931, was reported in the
next day's New York Times as follows:
Tax heavily, spend lightly, redeem debts, are sound Treasury
principles dJ.uing a period of dangerously rapid business expansion: tax lightly, spend heavily, borrow, are equally sound
Treasury principles during a period of acute depression ....
Really sound Treasury policy in a world in which business
depression is a recurrent phenomenon should be a function of
the state of business conditions and should be conducted so as to
contribute to the smoothing out of business Huctuations rather
than so as either to accentuate them or to insulate itself from
them regardless of the repercussions on business in general.
Two conservative journals which commented on this passage
(The Journal of Commerce, August 17, 1931; The Financial World,
September 9, 1931) agreed with it on principle, but warned that in
practice governments would happily conform to the rule during
depression but would disregard it during boom, so that its long-run
outcome would be inHation. There is much in the historical record
since then to justify this warning, but I would then and later have
been willing to take the risk, provided the rule was not extended
to chronic unemployment, as I believe Keynes would have extended
it. There is nowhere in the General Theory any warning about the
evil or the danger of inHation, and Keynes there defines "true" inHation in such a manner that the historical circumstances which
would justify taking anti-inHationary action could exist only in the
rarest of circumstances; not even, say, in the Brazil of 1962 when the
price level rose by over 50 per cent. "When a further increase in
the quantity of effective demand produces no further increase in
output and entirely spends itself on an increase in the cost-unit fully
proportionate to the increase in effective demand, we have reached a
condition which might be appropriately designated as one of true inHation" (p. 303). The advocacy this year (1963) by the Administration at Washington of a deliberate increase in the budgetary deficit
at a time of increasing output and employment, with the implied
promise that it will not result in inHation, is hard for me to understand unless the Administration has this Keynesian definition of
"true inHation" up its sleeve, ready to be produced if it should be
charged with inHationary tendencies.

Comment on My 1936 Review' 265


The prophet and the politician, in their professional roles, generally find it necessary to abandon a rhetoric which is disciplined and
solid but liable to be painfully unpersuasive and to adopt one which,
if it has elements of wildness in it, at least promises to be effective as
persuasion. The issue of sincerity is, with respect to them, not worth
raising, since what men hear themselves repeating they soon come
to believe. Keynes, I am sure, often interpreted his role as that of
the prophet and the politician, and it was in such a period, I presume,
that he once wrote: "Words ought to be a little wild, for they are
the assault of thoughts on the unthinking." I do not challenge this
for prophets and politicians, for they have special occupational
license to promote their objectives free from stodgy inhibitions on
the exercise of all their rhetorical resources for persuasion. I believe
in the virtues of professional division of labor, however, and I am
troubled, therefore, when economists adopt the role and the tactics
of the prophet or the politician, especially when there is any ground
for suspicion that what is involved is false prophecy.
In the early 1940'S, "Keynesians" were forecasting that the cessation of hostilities would be followed by a disastrous degree of depression and of unemployment unless governments deliberately engaged in massive deficit financing. In a series of conversations I had
with Keynes in March 1946, shortly before his death, I tried to tease
him about this, but found him invulnerable. He told me that he had
just met with a group of younger economists in Washington who
regarded themselves as his followers, and who claimed that they had
on hand econometric demonstrations based on Keynesian logic that
disastrous deRation was imminent. Keynes himself, however, was
optimistic and disowned any responsibility for their gloomy prognostications; he especially objected to their reliance on restricted and
mechanical manipulations of a few statistical series, rather than making a broad survey of the significant factors and using judgment in
assaying their importance and the nature of their impacts. His last
publication, which was in galley for the Economic Journal when he
died, startled his disciples by its optimistic tone, and there was serious
consideration of the desirability of suppressing it, on the ground, I
gathered, that it manifested some weakening, in his last days, of his
analytical powers, and that to publish it posthumously would be to
participate in perpetrating an unwarranted blemish on his record.

266 . Jacob Viner


The article was not suppressed, and it survives as one among many
other testimonies of the supreme level of statesmanship, of balanced
vision, of dedication, and of endeavor to reconcile differences in conclusions when there was no difference as to basic objectives and
values, on which he ended his brilliant career. The Keynes of that
period was to me then, and continues to be, a heroic figure, and my
admiration of him as he was then, of his personality and his intellect,
was unalloyed by the qualifications which I felt, and still feel, about
the Keynes of the General Theory, and even more, about the journalistic and polemical Keynes of the 1920'S and early 1930'S.

PART THREE

The Forties
The Sixties

GOTTFRIED HABERLER

Cfhe General Cfheory After


Ten Years
[1946J
1

I SHALL CONPINB MYSELF in this essay to the purely scientific content of The General Theory of Employment, Interest, and
Money, the most famous of Keynes' economic works, whose tenth
anniversary unhappily coincided with the death of its author. In the
light of ten years of intense and voluminous discussion, what remains of the Keynesian revolution, of the New Economics? What
will be the verdict of a historian of economic thought one hundred
years hence? There is no doubt Keynes stirred the stale economic frog
pond to its depth. He has kept economists in a state of agitation for
the last ten years, and probably for many years to come. The brilliance of his style, the versatility, flexibility, incredible quickness,
and fecundity of his mind, the many-sidedness of his intellectual
interests, the sharpness of his wit, in one word the fullness of his
personality, was bound to fascinate scores of people in and outside
the economic profession. Only a dullard or narrow-minded fanatic
could fail to be moved to admiration by Keynes' genius. But the
novelty and validity of the propositions which constitute his system
are a different matter altogether-quite independent of the challenging way in which he pronounced them, of the psychological stimulus
afforded by his bold attack on widely accepted modes of thought, of
much needed change in emphasis which we owe to his book, and of
the wisdom Cor unwisdom) of his policy recommendations. Apart
from a few observations on alleged policy implications of the General
Theory at the end of this paper, we shall be concerned exclusively
with the logical content of the system.

270

Gottfried Haberler
2

The tremendous appeal of the General Theory to theoretically


minded economists has been attributed by many to the (alleged) fact
that it uses for the first time in the history of economic thought a
general equilibrium approach in easily manageable, macroscopic
(aggregative) terms. There is no doubt, in my opinion, that this
made the theory very attractive, especially because such a system
lends itself easily to refinement and dynamization. But we can
safely assume that the concrete content and the policy recommendations which Keynes and others deduced from his system had even
more to do with its persuasiveness (even for his theoretically minded
followers) than its theoretical beauty and simplicity.
The use of aggregative systems of general equilibrium is by no
means new. All business cycle theories run in macroscopic terms. It
is true that most of the earlier business cycle theories are incompletely
stated, the number of explicitly stated relations is frequently not equal
to the number of unknowns, the structure of the system is such that
it is unstable (or does not oscillate, which is bad for a business cycle
theory). But even before the appearance of Keynes' General Theory,
the work of econometricians, notably Frischl and Tinbergen,2 had
done much to clarify these issues and had set higher standards of
formal completeness and precision. In fact, these early models, or
models of models, were superior to Keynes' system in scientific
workmanship because they made a clear distinction between statics
and dynamiCS, while Keynes' system is entirely static, as is well
known (although it lends itself to dynamization).3 Moreover, they
were tentative, experimental, hypothetical, not yet frozen into a
dogmatic pattern. This made them politically neutral, which, toSee his famous contribution to the Cassel Festschrift, "Propagation
and Impulse Problems," 1933.
2 For example, in "Suggestions on Quantitative Business Cycle Theory,"
in BO, 1935.
3 What is strictly static in the General Theory is the theoretical skeleton as precisely stated in several places in the book (e.g., p. 245 et
seq., or p. 280 et seq.) and later fonnalized by Lange, Meade, and
others. The text surrounding the theoretical statements in the General
Theory contains, of course, many dynamic considerations. The frequent use made of the expectation concept shows the dynamic intent.
But the dynamic elements are not incorporated into the theory. All
the functions stated are strictly static.

The General Theory After Ten Years 27l


gether with the fact that they were expressed in mathematical
terms, made them decidedly less accessible and less attractive than
the Keynesian system. But there is no doubt that Keynes gave a
tremendous impetus to model building, static as well as dynamic.
3

Let us look now into the content of the system. We shall first examine the individual relationships ("functions" or "propensities") of
which it is composed, and then the working of the system as a whole.
Little need be said about the marginal efficiency of capital or
demand schedule for capital, because here Keynes follows conventional lines. Investment is a decreasing function of the rate of
interest. In the post-Keynesian, Keynes-inspired literature, it has
been more and more questioned whether the rate of interest is really
such an important factor; in other words, the view has gained ground
that the demand curve for capital may be fairly inelastic with respect
to the rate of interest. But this is not the position of the General
Theory, at least not of its theoretical skeleton, although Keynes in
obiter dicta and policy recommendations frequently accepted openly
or by implication the theory of lacking investment opportunities.
The liquidity preference theory of the rate of interest appeared
very unorthodox and novel in 1936. The ensuing discussion has made
it clear, however, that the only innovation is the assumed relationship
between the rate of interest and hoarding, i.e., money held for
speculative purposes (M2 ) or idle deposits. (Assuming that the velocity of circulation of money, of M I , remains the same, or if it too
varies with the rate of interest, the proposition implies that the
velocity of the total money stock (MI +M 2 ) also is positively correlated with the rate of interest.)4 The older monetary theory as4

The proposition was clearly foreshadowed in the earlier ("classical")


literature. See, e.g., Lavington, English Capital Market (1921), p.
30. "The quantity of resources which [an individual] holds in the
form of money will be such that the unit of resources which is just,
and only just, worth while holding in this form yields him a return
of convenience and security equal to the yield of satisfaction derived
from the marginal unit spent on consumables, and equal also to the
net rate of interest."
See also Pigou, "The Exchange-Value of Legal-Tender Money" in
Essays in Applied Economics (1922), pp. 179-81. In his later, postKeynesian writings Pigou always makes a specific assumption with
respect to the policy followed by the ban king system. In what he

272

Gottfried Haberler

sumed (more or less explicitly) that the demand for hoards is


inelastic with respect to the rate of interest. Keynes assumed it to
be elastic. The reasons given for this are two: (I) Hoarding is the
cheaper (i.e., its opportunity cost is the lower), the lower the rate of
interest; (2) the lower the rate of interest, the smaller the likelihood
that it will go still lower and the greater the chance that it will rise
again. 5
The older theory was probably more realistic on this point. At any
rate, cyclical and other shifts of the liquidity preference schedule are
undoubtedly much more significant than its alleged positive slope.
A change in the rate of interest of a few per cent, other things being
equal, is hardly an important factor in determining the volume of
hoards. The latter is determined primarily by other factors s,uch as
price expectations, general pessimism, temporary lack of investment
opportunities, and so on.6 It is true that some writers, e.g., Kalecki
and James Tobin, have managed to compute beautiful correlations
between the rate of interest, on the one hand, and the volume of idle
deposits, on the other. But the reason is that both are (or until now
were) the joint effect of the same cause, of the business cycle. It is
quite easy, however, to imagine future ups and downs of business
without any significant changes in interest rates. I venture to predict
that in such cases we shall still find idle deposits rising in the downcalls the "normal" case the banks act in such a way as to allow the
quantity of money to rise and fall with the rate of interest. (See,
e.g., Equilibrium and Employment, p. 61.)
This latter-day Pigovian approach, institutional in nature, seems
to me more realistic than the Keynesian liquidity preference theory.
The latter is clearly a direct descendant of the penetratingly classical
Cambridge type of quantity equation (as Hicks pointed out in his
paper, "Mr. Keynes and the Classics," Ee, Vol. 5, 1937), and suffers
from the same weakness as its parent concept, viz., excessive utilization of a marginalistic psychology in a field where a frankly institutionalistic analysis is much more fruitful.
~ We may, perhaps, say in Hicksian terminology: The lower the rate of
interest, the smaller the elasticity of expectation of future rates.
6 I do not deny that hoarding and changes in the velocity of circulation have been much neglected in the literature, and that it is a mistake (of omission rather than commission) to regard these phenomena as datil (or as occasional disturbances) instead of explaining them
systematically. The point is that the level of the rate of interest as
such is a comparatively unimportant factor.
Expectations of changes in interest rates are, however, a different
matter. But the state of expectation is a complicated matter, and no

The General Theory After Ten Years 273


swing and falling in the upswing, which would prove that the correlation between hoards and interest rates does not indicate a causal
relationship in the sense that people hoard more when a fall in the
rate of interest makes it cheaper and vice versa.
Other propositions frequently associated with Keynes' interest
theory--e.g., those concerning the connection between short- and
long-term rates and the alleged floor, well above zero, below which
the rate of interest cannot fall-were frequently discussed in the
pre-Keynesian literature. 7 But Keynes certainly improved the analysis
and utilized those theorems effectively by putting them into the
broader context of a general equilibrium system.
The theory of liquidity stands in great need of further elaboration.
It will be necessary to distinguish a larger number of different types
of assets than just money and real goods, or money, securities, and
real goods. The different types of assets have to be arranged according to their liquidity, with cash on one end of the scale, certain
types of finished goods on the other end, and loans, bonds, equities,
raws material, etc., in between. Much work had been done along that
line before the appearance of the General Theory 8 (and more has
been done since publication of the volume), and Keynes himself
contributed important elements for a comprehensive theory, especially in his Treatise on Money. But these refinements, indispensable
though they are for a useful application of the theory to reality, were
not incorporated, and were not easy to incorporate, into the body of
the General Theory-a fact which should be kept well in mind by
those who try to find empirical support for the liquidity preference
theorem of the General Theory. 9
simple fonnula, such as the one suggested in the preceding footnote,
can do justice to its complexity.
Professor W. Fellner in his elaborate and searching investigation
of the subject reaches the conclusion "that the elasticity of liquidity
provisions with respect to interest rate is not likely to be high"
(Monetary Policy and Full Employment, Berkeley, 1946, p. 200).
1 E.g., in I. Fisher, The Theory of Interest (in the third approximation
of his theory), or Karin Kock, A Study of Interest Rates (London,
1929). See, especially, Chapter VII, "Short and Long Rates of Interest."
8 Cf., for example, Hicks "Gleichgewicht und Konjunktur" in Zeitschrift fur Nationa16konomie, Vol. 4, Vienna, 1933.
9 In all attempts at verification, the liquidity preference theory is applied to the choice between (a) cash (including bank notes and deposits) and (b) the next item on the scale, viz., shortest-tenn securi-

274 . Gottfried H aberler


We now turn to the consumption function. The idea that saving
depends on the level of income-other things such as the rate of
interest being equal-is an old one. Suffice to recall the fact that
scores of writers made the point that inequality of the income distribution is necessary or desirable to guarantee a sufficient supply of
capital, because the bulk of saving comes from the higher income
brackets. Keynes' great contribution was that he strongly emphasized
the income factor and used it much more systematically in the
analysis of economic change than had ever been done before. It is
true that the consumption function has often been overworked by
Keynes and his followers; it has been too rigidly formulated and too
inflexibly applied to short- as well as long-run problems without
allowing for all the necessary qualifications, such as secular shifts,
cyclical fluctuations, the influence of capital gains, and other factors.
But on the whole the change in emphasis toward income was needed
and beneficial. The strong and exceedingly fruitful accent on income
effects, which has become more and more noticeable in recent years
in all branches of economics, such as price and demand analysis,
international trade, etc., is largely due to Keynes. The same is true of
the multiplier technique, whose usefulness should not be doubted,
despite the crudity with which it is often used.
4

Let us turn now to the interaction of the various parts and the
working of the system as a whole. Even if it were true that all the
materials and tools used by Keynes had been known and used before
and that he did not improve them-is it not true that with their help
he constructed an entirely new theoretical structure?
His demonstration that unemployment is possible in equilibrium,
ties, in other words, between (a) money and (b) near-money (i.e.,
money's closest substitute). For that very limited choice (i.e., the decision whether to hold one's idle funds in cash or short-term securities) the short-term rate of interest may indeed be an important
factor. But that choice is an unimportant detail as far as expenditures
on goods and the volume of output and employment are concerned.
And any empirical regularities found with respect to this detail cannot be regarded as a verification of the liquidity preference theorem in
a rougher model which does not distinguish a whole scale of different
assets with small gradations in liquidity, but only two or three types
of assets.

The General Theory After Ten Years . 275


and his analysis of the factors determining the size and changes of
employment and unemployment, are generally regarded as Keynes'
most important theoretical discovery. The originality and importance
of this conclusion remains unimpaired, it will be said, even if it can
be demonstrated that it is derived entirely from well-known premises,
just as the work of a great artist remains great even if he uses well
known tools and techniques.
According to a widely held view, which can be described as a sort
of simplified, popular Keynesianism,10 the possibility of underemployment equilibrium has been denied by the "classical" school and
demonstrated by Keynes. The matter, however, is not so simple as
that. This becomes quite clear if we reflect upon the intricate and
crucial question concerning the role of wage (and price) rigidity in
the Keynesian system. Keynes assumes that (money) wages are rigid
downward. If this assumption, which is certainly not entirely unrealistic, is rigidly adhered to, most of his conclusions follow: Underemployment equilibrium is then possible; an increase in the propensity to consume will then reduce unemployment and a decrease in
the propensity to consume will produce unemployment (except if,
as many classical writers assumed, the demand for idle funds, the
liquidity preference proper, is wholly inelastic with respect to the
rate of interest). But all this is entirely in accord with pre-Keynesian
theory, although these conclusions certainly had not been generally
realized and sufficiently emphasized before the appearance of the

General Theory.
If flexible wages-"thoroughgoing competition between wage earners" (in Pigou's words)-are assumed, the situation is radically
changedY Obviously, under-employment equilibrium with flexible
wages is impossible-wages and prices must then fall continuously,
Unfortunately, there is much of this oversimplified version in the
General Theory itself, especially in the three summarizing chapters in
Book I. A sociology of the formation of scientific schools will attribute much importance to this fact. It helped to crystallize a compact group of followers by repelling and annoying some readers and
attracting others.
11 The crucial importance of wage rigidity in the Keynesian system has
been emphasized by many critics, most systematically perhaps by
Franco Modigliani in his remarkable article "Liquidity Preference
and the Theory of Interest and Money," Ee, January, 1944.
10

276 . Gottfried Haberler


which can hardly occur without further consequences and cannot
well be described as an equilibrium position. 12 This is the weak spot
of the Keynesian system which is usually slurred over by the Keynesians.
As in many other cases, two different attempts to deal with this
problem can be found in the General Theory. The first one, which
belongs to what I called the oversimplified, popular version of
Keynesianism, is stated early in the book Cpo I I et seq.), and has
been too readily accepted by friend and foe. It simply says that
when money wages fall, prices too will fall to the same extent;
therefore real wages will remain unchanged, and since "an increase
in employment can only occur to the accompaniment of a decline
in the real rate of wages"13 Cpo 17), employment and unemployment
will remain the same.
This solution is obviously unsatisfactory and should not be regarded as Keynes' last word. This becomes clear if we consider the
solution consistent with the system as a whole which can be found
in Chapter 19. There it is pointed out that a reduction in money
wages will usually influence employment, but in an indirect fashion,
through its repercussions upon the propensity to consume, efficiency
A logical possibility would, of course, be that all money expressions
(prices, wages, money values) fall continuously, while the real magnitudes including employment remain the same. That would be the
implication of the assumption that the Keynesian relations remain
unchanged in real terms in the face of such a situation. But this case
is surely too unrealistic to be seriously contemplated.
13 Professor Hansen objects to my quoting this passage because it "fails
to include the very important conditions which must be assumed to
make the statement [as quoted from Keynes] true, namely, no change
in 'organization, equipment and technique'; in other words, no change
in productivity. Moreover, Keynes (March, 1939, EJ) explicitly repudiated the notion that employment must increase by or through
a lowering of real wages and a movement along a declining so-called
general demand curve for labor. In his view employment is increased
by raising effective demand, thereby causing an upward shift in the
demand curve for labor." (RES, Vol. 28, Nov., 1946, p. 185. See
also Chap. XII.)
It is true Keynes did qualify his statement by the clause "with
given organization, equipment, and technique" (p. 17). But in the
present context the qualification is irrelevant. For in the short run
(and the problem under discussion is essentially a short-run problem)
Keynes always assumes "organization, equipment, and technique"
constant. In EJ, March, 1939, Keynes took issue with Dunlop's and

12

The General Theory After Ten Years . 277


of capital, or the rate of interest. The last-mentioned route, via the
interest rate, is the one most thoroughly explored by Keynes and
the Keynesians. As wages and prices are allowed to fall, money is
released from the transactions sphere, interest rates fall, and full
employment is eventually restored by a stimulation of investment.
This amounts to giving up the idea of under-employment equilibrium
under a regime of Hexible prices and wages except in two limiting
cases: Full employment may be prevented from being reached via
this route, (a) if the liquidity trap prevents a fall in the rate of interest-that is to say, if the liquidity preference schedule is infinitely
elastic, i.e., if people are willing to hoard unlimited amounts of
money at a positive rate of interest-or (bY if investment is quite
insensitive to a fall in the interest rate. Keynes himself regarded
both these situations not as actually existing but as future possibilities.
But what if we do regard them as actually existing-which as a
short-run proposition, allowing for dynamic disturbances through
unfavorable expectations, etc., would be by no means absurd? We
would still not have established a stable underemployment equilibrium, for wages and prices would still continue to fall. The truth
is that what would happen in this case cannot be told within the
Tarshis' criticism; he there was very reluctant to give up his generalization. "I still hold," he said, "to the main structure of the argument,
and believe that it needs to be amended rather than discarded" (p.
40). He tried to reconcile Dunlop's and Tarshis' findings with his
theory without dropping the assumption of constant organization,
equipment, and technique.
It is, of course, true that according to Keynes "employment is increased by raising effective demand," but he thought (with certain
tentative qualifications as enumerated in the quoted article) that, by
a rise in effective demand, prices are necessarily raised more and faster
than money wages, and that therefore a rise in effective demand
is always associated (in the short run) with a fall in real wage rates.
Keynes' reluctance to drop this hypothesis is understandable because a change of view would have required far-reaching modifications of his whole theoretical structure. He, after all, had emphasized that he was "not disputing this vital fact which the classical
economists have (rightly) asserted as indefeasible" (General Theory,
p. 17). He had argued emphatically that, if workers could effectively
bargain about real wages rather than merely about money wages,
unemployment could always be eliminated by wage bargains at
lower wages. The disputed proposition is, thus, deeply embedded in
Keynes' theory.
I personally always felt that Keynes' dogmatic insistence on the

278 . Gottfried Haberler


Keynesian framework, and Keynes himself would have been the
last one to stick to it through thick and thin. 14 We must assume that
some of the Keynesian schedules would shift. The most obvious
hypothesis would seem to be that the consumption function will
shift upward, because of the accumulation of liquid reserves. 15 For
we must assume, it seems to me, that consumption is not only a
function of income but also of wealth (and liquid wealth in particular) and of other factors which we need not discuss here and
which are in fact indicated in the General Theory (cf. Chs. 8 and
9). A similar argument would seem to hold for the investment function.
Such extensions and modifications of the Keynesian system are
entirely in keeping with Keynes' own injunction against dogmatically
treating any such system as rigid and sacrosanct. He warns us that
the determinant relations and magnitudes of his own system (i.e.,
the three propensities, the quantity of money, and the wage unit)
are "complex, and that each is capable of being affected by prospective changes in the other";16 he says that only "sometimes" (meaning,
obviously, in certain context and over limited ranges) can they be
regarded as "ultimate independent variables."17
It should be clear, however, that even with these modifications
proposition in question was due to the excessively static nature of his
theory. If Keynes had incorporated swings of optimism and pessimism in his theory, he would have had no difficulties in admitting
that an expansion can raise not only money but also real wages (even
in the short run, i.e., with organization, equipment, and technique
unchanged). The plain fact is that Keynes' theory is not only more
static but in several respects also more "classical" than, for example,
Pigou's Industrial Fluctuations, where it had been pointed out that
"the upper halves of trade cycles have, on the whole, been associated
with higher rates of real wages than the lower halves." (1929 edition, p. 238.)
14 See the following paragraph, and footnotes 16 and 17 below.
15 If wages and prices fall, the real value of the money stock will increase beyond all limits. I called attention to this fact and its probable effect on consumption in the first edition of my Prosperity and
Depression (1937) without then using the term "propensity to consume." Pigou has since stressed it repeatedly. Kalecki in his brief
note, "Professor Pigou on the Classical Stationary State-A Comment" (EJ, April, 1944, p. 131), in principle conceded the argument that a rise in the real value of the money stock will act as a
stabilizer in a period of falling prices. He makes, however, the point
that this argument applies only to gold and bank notes which are not
issued by the banks through making loans or through purchases of

The General Theory After Ten Years . 279


the theory is still much too rough for direct application and must
be further elaborated and supplemented before it can be used, even
in a tentative fashion, for the explanation of reality. In the short
run, dynamic repercussions (unfavorable expectations, disturbances
caused by bankruptcies and credit crises, etc.) must be taken into
consideration. Pigou was probably right when he insisted that in a
cyclical depression negative wages and prices frequently would be
necessary to prevent unemployment altogether or to eliminate it
quickly once it appeared. 1s The situation in the long run is radically
different. Unfavorable expectations and credit crises do not last
forever, disturbances caused by bankruptcies disappear, and the assumption of an infinitely elastic liquidity preference and entirely
inelastic marginal efficiency of capital schedule is hardly tenable as
a long-run proposition. But most economists will agree that it is not
only politically easier but also economically more desirable, in the
long run as well as in the short, to bring about the saturation of the
economy with liquid funds (if required) by increasing the quantity
of money rather than by raising its value through a fall in prices.
The reasons for and against that proposition (such as rigidity of
long-term money contracts, avoidance of industrial disputes, and unprivate securities; for in the case of bank money issued against loans
and private securities, the rise in the real value of money is canceled
by the rise in the real value of the corresponding bank assets (loans
and securities) which are liabilities of the public. (The net worth
of the public is, therefore, not increased by the fall in the price level.)
However, as long as there is money which is not issued against
private evidences of indebtedness, Kalecki's argument is invalid from
the theoretical point of view, because money wages and prices (note:
real wages need not fall) can always fall sufficiently to raise the real
value of gold money to any level necessary, however small the (dollar) value of gold or gold certificates in circulation.
From a practical point of view, however (i.e., taking account of
frictions, disturbances through expectations, etc., which are assumed
away in the pure model), Kalecki's argument is important. But I
need not go into that, because I believe (and I think this is also
Pigou's view; cf. the Preface to his Lapses from Full Employment)
that the model under consideration is much too simplified to be
useful for practical application. (See next paragraph in the text.) It
should be observed that the simplifications are essentially the
Keynesian ones.
16 General Theory,
p. 184. What is said of prospective changes
naturally holds also of the actual level and actual changes.
17 Ibid., pp. 246-47. Cf. also p. 297.
18 See, e.g., his Industrial Fluctuations, 2nd ed. 1929, p. 225.

280

Gottfried Haberler

just and undesirable changes in the income distribution, etc.) are


the same ones that were discussed extensively in the literature on
money throughout the nineteenth century and later in connection
with the problem of whether in a progressive economy it is better
to let prices fall or to keep them stable. 19 Therefore, the question
ought not to constitute an issue between Keynesians and non-Keynesians.
One last word on this important subject. There is nothing in the
Keynesian theory to exclude a more direct inBuence of wage reductions on employment. We stated above that according to Keynes
this inBuence works via repercussions upon the consumption function, marginal efficiency of capital, and the liquidity preference (the
rate of interest). In the preceding pages, we discussed the last route.
But it is clearly possible that consumption and investment might be
affected more directly by a reduction in wages. A reduction in the
cost of certain consumption or investment goods may well stimulate
demand for them, and for consumption and investment as a whole.
Is it not possible that more roads, houses, hospitals, will be built
when construction cost is reduced, or that the demand for certain
private consumption goods will rise when their price falls?19 a Assume,
to make it quite simple, that the elasticity of demand for some of
The older literature which dealt with these questions under various
guises and in outmoded terminologies is extensively reviewed in C. M.
Walsh, The Fundamental Problems of Monetary Science (New York,
1903). In the present case the argument for increasing the quantity
of money and holding the price level constant is, of course, much
stronger than in the historical case mentioned, because in the present
case all prices (including factor prices) would have to fall, while in
the other case it was for the most part a question of keeping factor
prices stable and letting product prices fall vs. keeping product prices
stable and letting factor prices rise. But the point is that many of the
arguments used there are relevant for the present case too.
19_ It can hardly be denied that it is possible to raise construction cost
of houses, etc. (to mention a much discussed case) to such an extent
that the demand for houses is seriously restricted. It obviously follows from this proposition that a reduction of such cost (brought
about by elimination of monopolistic and restrictive practices on the
part of labor and contractors, etc.) may stimulate demand for homes
(investment). It is generally assumed that cost reducing innovations
(e.g., prefabrication of houses) can stimulate investment. Why
should then a reduction of labor costs not be capable of bringing
about the same result?
19

The General Theory After Ten Years . 28 I


those things, and therefore indirectly for labor, is unity.2o Then the
wage bill remains unchanged and there are no adverse effects through
a fall in consumption demand of the workers. Then employment
will clearly rise. In Keynesian language we shall have to say that
the marginal efficiency of capital schedule or the consumption function has gone up (which one depending upon whether the newly
produced goods or installations are regarded as consumption or investment goods), and that it is this shift which has brought about
the increase in employment.
One may, of course, be more or less optimistic or pessimistic concerning such favorable direct inBuences. 21 Keynes' theory certainly
does not exclude them.

5
The gist of the foregoing discussion may be brieBy restated from
a different point of view or rather (for it amounts to nothing more)
in terms of a different economic jargon. I take Paul Sweezy's brilliant obituary note on Keynes as my text. 22 Sweezy regards as the
basis of the Keynesian system, and of Keynes' criticism of classical
economics, the "Bat rejection and denial of what has come to be
known as Say's Law of Markets which, despite all assertions to the
contrary by orthodox apologists, did run like a red thread through
the entire body of classical and neoclassical theory. It is almost impossible to exaggerate either the hold which Say's Law exercised
on professional economists or its importance as an obstacle to realistic
analysis. The Keynesian attacks, though they appear to be directed
If the elasticity of demand is not unity, we get a much more complicated situation, which cannot be discussed here. But much of the
argument could be adapted to fit that case.
21 It is true, Keynes calls such influences "roundabout repercussions"
(p. 257) and criticizes older writers for assuming a "direct" effect
of wage reductions on employment. But, as I pointed out in my
Prosperity and Depression (2nd or later editions, p. 241), what to
call direct or indirect is a purely terminological question. The most
direct effect imaginable Keynes calls "roundabout" because, by definition of the terms, it must imply a change in the propensity to consume or in the marginal efficiency of capital.
22 Science and Society, Vol. X, 1946, pp. 396-406. See also Part One
above.
20

282

Gottfried Haberler

against a variety of specific theories, all fall to the ground if the


validity of Say's Law is assumed."23
What is the content of Say's Law? After the early statements of
the Law by the old classical writers, the subject has become so confused by criticism and defense that neoclassical writers only rarely
make use of, or allusion to, it. But I think that a careful perusal of
Ricardo's formulation (which is quoted by Sweezy) should make
it clear what the original meaning of Say's Law was. The passage
reads as follows: "No man produces but with a view to consume or
sell, and he never sells but with an intention to purchase some other
commodity which may be useful to him, or which may contribute
to future production. By producing then, he necessarily becomes
either the consumer of his own goods, or the purchaser and consumer of the goods of some other person.... Productions are always
bought by productions, or by services; money is only the medium
by which the exchange is effected."24
The meaning of this original formulation of this law seems to me
quite clear: It states that income received is always spent on consumption or investment; in other words, money is never hoarded,
the money or expenditure stream, MV (in some sense), remains
constant or, in still other terminology, money remains "neutral."
23
M

Ibid., pp. 4 00- 1


Principles of Political Economy (Gonner ed.), pp. 273 and 275. The
following quotation from Say clearly conveys the same meaning: " .. ,
a product is no sooner created than it, from that instant, affords a
market for other products to the full extent of its own value. When
the producer has put the finishing hand to his product, he is most
anxious to sell it immediately, lest its value should vanish in his
hands. Nor is he less anxious to dispose of the money he may get for
it; for the value of the money- is also perishable. But the only way of
getting rid of money is in the purchase of some product or other.
Thus the mere circumstance of the creation of one product immediately opens a vent for other products." Jean-B. Say, (Treatise on
Political Economy, Prinsep edition, Boston, 1921.) In later editions,
Say obscured and attenuated the original meaning more and more
through his attempts to meet criticisms by Malthus and Sismondi. He
was forced to redefine the terms until the whole proposition became
an empty tautology. See, for a brief account, P. N. Rosenstein-Rodan,
"A Co-ordination of the Theories of Money and Price," Economica,
1936, VP . 268-9; and H. Neisser, "General Overproduction: A Study
of Says Law of Markets," JPE, Vol. 42, 1934, reprinted with revisions in Readings in Business Cycle Theory (1944), pp. 385 et seq.

The General Theory After Ten Years . 283


(Note how clearly the last sentence in Ricardo's passage foreshadows
what a hundred years later became known as "neutral money.")
If this straightforward, monetary meaning of the law is firmly
kept in mind (which is not easy because of the hocus-pocus accumulated over the years in later classical and anticlassical writings
on the subject) two conclusions are obvious. First, Say's Law does
not hold in reality; every depression is a proof to the contrary. Second, hardly any neoclassical economist who ever wrote on money
or the business cycle thought that Say's Law did hold in reality.
The major theme of their theories of money, interest, and the business cycle, is to analyze the causes and consequences of changes
in the "intrinsic" or "extrinsic" value of money, of deviations of the
money rate of interest from the equilibrium rate, and of other "aberrations from monetary neutrality," which are all different expressions
for deviations of reality from the ideal state as postulated in Say's
Law.
A few neoclassical writers, rather naively, attributed such deviations entirely to the wickedness or incompetence of those in charge
of monetary policy, but many, and as time went on more and more,
of them realized that these deviations are deeply rooted in the structure of the capitalist system and cannot be easily prevented or cured
by slight changes in monetary policy. Some recent neoclassical writers like Hicks and Rosenstein25 went so far as to deny the compatibility of money and static equilibrium altogether.
Our conclusion, thus, is that there is no place and no need for
Say's Law in modem economic theory and that it has been completely abandoned by neoclassical economists in their actual theoretical and practical work on money and the business cycle. That
should be clear to anyone who is interested in living science (theoretical as well as realistic) and knows how to distinguish it from
verbal squabbles and historical reminiscences in which economists
so often indulge. The question must still be asked, however, why
Say's Law was more often silently dropped rather than openly
repudiated. Why did some older writers (especially Say and J. S.
25

Even Hayek should be mentioned here. This becomes clear if we


reflect that the extremely complicated nature of a monetary system
which is neutral ill his sense makes the existence of neutral money
in practice utterly impossible.

284 . Gottfried Haberler


Mill), after having been forced to emasculate the law and to make
it tautological, still pay lip service to it?
Liberal prejudices, the inability to rid oneself entirely of the assumption of a pre-established harmony of interests, were undoubtedly
a factor, but it would be a bit too crude and naive to rely on this
factor. 26 There is a perfectly good scientific explanation (as against
a superficial explanation in terms of ideological prejudices) for the
lingering doubt concerning Say's Law, the reluctance of some to
repudiate it openly and the occasional attempts to uphold it in some
rarefied (nonmonetary) form. 26 > The reason is the difficulty, upon
which I commented above, of reconciling a competitive system with
the existence of unemployment. This difficulty has, as we have
shown, not been solved by Keynes.
Summing up, we may say there was no need for Keynes to rid
neoclassical economics of Say's Law in the original, straightforward
sense, for it had been completely abandoned long ago. Keynes was
unable, on the other hand, to solve the riddle of how to reconcile
competition and unemployment which is at the root of some remaining qualms about the matter in the mind of some writers.
6

We thus reach the conclusion that, as far as the logical content


of Keynes' theory goes, i.e., apart from his judgment of the typical
shape of the various functions and of concrete situations and apart
from policy recommendations, no revolution has taken place; the
General Theory marks a milestone, albeit a conspicuous one, but
not a break or a new beginning in the development of economic
theory. The impression to the contrary stems from two sources. The
Very sophisticated writers whom it would be utterly absurd to accuse as capitalist or orthodox apologetics (especially inasmuch as they
are often on the other side of the fence as far as their political convictions are concerned) have been attracted by the intricacies of the
problem and have refrained from rejecting Say's Law out of hand.
Cf., for example, the articles by Neisser and Rosenstein-Rodan mentioned above and some of the literature there quoted.
26> Something like the following formulation is probably in the back of
the minds of many writers: Any amount of money expenditures,
however small, can buy any volume of goods :>fFered for sale, provided prices are flexible and are low enough. This is obviously an
arithmetic truism which cannot be denied, but is not very useful.
26

The General Theory After Ten Years . 285


first is excessive and untenable claims made by Keynes and his followers (and accepted too readily at their face value by many of his
critics)-daims which are based on an oversimplification of the
Keynesian system itself27 as well as misrepresentations and misinterpretations of the "classical" doctrine. 28
The second source is differences in policy recommendations. However, if the preceding analysis is correct, differences about policy
cannot logically be explained by basic theoretical disagreement but
must be explained by different judgments concerning concrete situations, administrative efficiency, the possibility of rational policy making and, perhaps most important, by different attitudes concerning
the broad issues of government intervention and central planning
versus laissez faire. It follows from our analysis that specific policy
recommendations derivable from the Keynesian system are not at all
revolutionary. They are in fact very conservative. Laissez faire liberals, like Michael Polanyi,29 who wish to conserve free enterprise
and freedom of consumer choice, are entirely justified in their enthusiastic acceptance of the Keynesian doctrine. 30
A few words of justification are needed, because fairly radical
proposals for equalizing income distribution, and for direct control
For example, the proposition that in the Keynesian system the rate
of interest is independent of the marginal efficiency of capital and
the propensity to save. Or the misconceptions concerning the role of
the assumption of rigid wages.
28 For example, the proposition, which is closely connected with the
misconception of the role of wage rigidity in the Keynesian system,
that there is no room for "involuntary" unemployment in the "classical" system. Another misconception is the view that classical economics assumed that an act of saving always brings about a corresponding
act of investment, while in the Keynesian system the two types of
decisions are independent of each other, although aggregate saving
and investment are equalized ex-post by appropriate changes in income. In reality, the neo-classical literature, especially its Wicksellian
branch, stressed the fact that new saving may fail to induce new
investments, with a consequent fall in money income and usually
also in real income.
29 See his Full Employment and Free Trade (Cambridge, 1945).
so As a chemist, Mr. Polanyi can be pardoned for overlooking the fact
that his conclusions could have been derived from economic principles widely accepted before the appearance of the General Theory
(which, of course, does not mean that those conclusions were generally accepted before Keynes).

21

286 . Gottfried H aberler


of investment and the location of industry, have been made under
the Keynesian Hag, by Beveridge and his group, for example.
In fact, as far as policy recommendations are concerned, we may
distinguish two wings of the Keynesian School, a radical, interventionist or even socialistic one to which many of the younger Keynesians belong, and a liberal wing represented by John Jewkes,
Polanyi, McCord Wright, and A. P. Lerner (and many others who
do not count as Keynesians because, although acknowledging their
debt to Keynes, they do not believe that the continuity of development of economic thought has been interrupted by the appearance
of the General Theory). There are good reasons to believe that at
the bottom of his heart Keynes himself belonged to the liberal wing
of his school, especially in later years when, after what must have
looked to him a victorious battle for the acceptance of his views, he
regained some perspective. Even during the years immediately after
the appearance of the General Theory, when he was carried away
by his enthusiasm, he never went all the way in accepting socialism
or even anything like Beveridge's radical proposal, although in the
heat of the battle against hostile critics he said things that seem to
give comfort to his radical followers.
But whatever his real attitude was, my point is that the radical
schemes hitched to the Keynesian bandwagon have nothing to do,
logically speaking, with the General Theory. From the point of view
of the General Theory, what is needed to prevent mass unemployment is monetary policy and, at the most, a mild form of fiscal policy.
Monetary policy would be sufficient, in most cases at least, if the
monetary authorities were prepared to extend the scope of their
operations, as Keynes proposed, to purchases and sales of long dated
securities or possibly equities. If fiscal policy is required, it need not
imply increased government expenditures and extended government
activities; it could be of the milder, less interventionist form of varying revenues and thus, when necessary, creating a deficit by tax
reduction instead of by public works.
I do not wish to say, nor did Keynes ever claim, that such policies
would insure literally full employment all the time (much less that
they would cure all economic ills and injustices). It can be argued,
and I am sure Keynes would have agreed (perhaps he actually said
so somewhere), that for quick results in a cyclical depression well-

The General Theory After Ten Years 287


directed increases in government expenditure (public works) would
be needed in addition to tax remissions. This does not follow, however, from the General Theory but from supplementary assumptions
about labor mobility and the distribution of productive resources
among industries and localities, compared with the distribution of
aggregate expenditure among types of goods and services. Let us
not forget that the General Theory runs in broad, aggregative terms
and is therefore precluded from dealing, and is not designed to
deal, with sectional unemployment, which is the result of faulty
allocation of resources or of shifts in demand. It is meant to deal
only with general mass unemployment resulting from a deficiency
in aggregate effective demand (deflation). Its author clearly assumed
that all other problems would take care of themselves, if only aggregate effective demand was kept on an even keel or raised when
necessary, for example when the wage and price level is pushed up
by monopolistic and restrictive policies of aggressive trade unions or
other pressure groups.
This is certainly a much too optimistic view. Keynes and most
Keynesians (especially Beveridge) underestimate, it seems to me,
the possible magnitude of frictional unemployment (people on the
way from one job to the other) and structural unemployment (unemployed workers in special depressed areas and industries) which,
unlike general (i.e., well-dispersed) unemployment, cannot be cured
by merely manipulating aggregate demand. They fail to realize, or
at least to realize fully, the enormous difficulties, or almost impossibility in the kind of "free society" as we today know it in the
western world, to restrain labor monopolies from pushing up wages
and thus forcing a rise in prices whenever full employment is approached or even long before that point, in consequence of which
unemployment becomes necessary to prevent inflation.31 Socialist
economists like Professors Myrdal32 and Pigou33 have seen this problem much more clearly than the Keynesians.
I do not say that historically all depressions have come about for this
reason. I only say that if it were possible to stabilize aggregate demand and to prevent depressions arising from other causes, the
factor mentioned in the text would make it very hard to maintain
full employment for some length of time.
82Cf. Monetary Equilibrium (19p9), esp. pp. 143-147 and 155-156.
88 Lapses from Full Employment (1945), passim. See also the illuminating review of Pigou's book by Professor Hicks, EJ, Dec., 1945, pp.
81

288 . Gottfried H aberler


But all this is a matter of judgment about the operation of certain
social forces. Although crucially important, it does not involve the
principles of the General Theory.

7
What has been said in these pages is not intended to detract from
Keynes' claim to subjective originality or to belittle his many genuine and ingenious innovations, both in substance and emphasis, or
to play down the obvious fact that the General Theory has exerted
a tremendously stimulating influence on economic thinking. Not
only did Keynes inspire a large and growing group of enthusiastic
and highly competent followers, especially among the younger generation of economists, but he also spurred on to clarifying and creative work many of those who at first received the General Theory
with suspicion and skepticism. Keynes forced them to think through
things which they used to leave in an ambiguous twilight, and to
draw from accepted premises conclusions of which they were unaware or which they left discreetly unexpressed. A classical treatise
like Pigou's monumental work, Equilibrium and Employment (which
is so much more general than the General Theory that the latter by
comparison appears as a very special case),34 would never have been
written without the Keynesian challenge, although it is not in contradiction to, but rather constitutes a clarification of, Pigou's own
pre-Keynesian, "classical" position.

398-40 I. Professor Hicks seems substantially to accept Pigou's conclusions, although he finds them "sour."
S4 The superiority of Pigou's great work has been recognized by so
Keynesian a critic as N. Kaldor (EJ, December, 1941). It is a pity
that another work of outstanding originality and scholarship which
was stimulated by Keynes' challenge, viz. A. W. Marget's The
Theory of General Prices (1938-42), has not yet exerted the influence which it should have.

Sixteen Years Later


[1962*J
I T IS NOW 26 years since the appearance of the General
Theory in 1936, and Keynes' ideas continue to influence strongly
economic theory and policy. Not only does every book on money,
business cycles, and related problems refer to and take issue with
Keynes' views, but whole volumes and special articles still appear
that are devoted to a restatement, appraisal or refutation of the New
Economics. Sympathetic reappraisals have recently been attempted
by Harry Johnson and D. McCord Wright. 1 Of recent critical works
Jet me mention Henry Hazlitt's two massive volumes: The Critics
of Keynesian Economics2 and The Failure of the New Economics:
An Analysis of the Keynesian Fallacies. s Of earlier books two stand
out prominently: Pigou's magistral lectures, Keynes's 'General Theory': A Retrospective View, an eminently fair and generous reappraisal written with a warm personal touch,4 and A. H. Hansen's
>(.

Written in October 1962.

H.G. Johnson, "The 'General Theory' After 25 Years." American


Economic Review, May 1961, pp. 1-17. D. McCord Wright, The
Keynesian System, Fordham, New York, 1962.
2 Van Nostrand, New York, 1960. This is a very useful collection
of old and new essays dealing critically with Kej'nesian issues. The
contributions range from J.-B. Say and J. S. Mill to Mises, Knight,
Hayek and Ropke.
S Van Nostrand, New York, 1959.
4 St Martin's, New York, 1953. It will be recalled that Pigou's first
reaction to The General Theory was very negative (review article
in Economica, London, 1936, l?.p. 1 1 5-132). He evidently was greatly
annoyed by "what seemed to lhim] misrepresentations of things that
other people had written" (p. I). In his reappraisal all bitterness
is gone and Pigou confesses that in his original review article of the
General Theory he had "failed to grasp its significance and did not
assign to Keynes the credit due for it" (p. 65). See below what
Pigou in his reappraisal regarded as the basic contribution of the
General Theory.
1

290

Gottfried Haberler

equally incisive, very skillful and immensely popular Guide to


Keynes. 5
There can be no question, then, that Keynes' ideas still have a
strong appeal and fascination, not only for those who are primarily
interested in economic policy and social reform, but also for firstrate theoretical minds. 6 I must confess, however, that after rereading
my article on the General Theory of sixteen years ago (see p. 269
above), I see no reason to change the views expressed there concerning both the subjective and objective originality and the limitations of the General Theory.
To explain my own view, I think I can do no better than to quote
from Pigou's reappraisal. "The kernel of Keynes' contribution to
economic thinking is to be found," Pigou says, "in a short passage
on page 246 of the General Theory" (p. 65). This passage reads as
follows:
Thus we can sometimes regard our ultimate independent
variables as consisting of (I) the three fundamental psychological factors, namely, the psychological propensity to consume, the psychological attitude to liquidity and the psychological expectation of future yield from capital assets, (2) the
wage-unit as determined by the bargains reached between employers and employed, and (3) the quantity of money as
determined by the action of the central bank; so that, if we take
as given the factors specified above, these variables determine
the national income (or dividend) and the quantity of employment.
Pigou continues: "Nobody before him, so far as I know, had brought
all the relevant factors, real and monetary at once, together in a
5

McGraw-Hill, New York, 1953.


I do not know, however, of any theorist who has embraced Keynes'
theory for its own sake-that is to say, a theorist who either takes no
interest in questions of policy and reform (if there still exist such
individuals), or one who lacks sympathy for the Jolicies and reform measures Keynes stood for. Professor Jewkes, . he can be described as a Keynesian, could possibly be said to fall into the latter
category and thus constitute an exception that proves the rule. I
should perhaps add that in my opinion it is logically quite possible
to deduce from Keynes' theory conservative and laissez faire policy
conclusions if the appropriate assumptions are made about the concrete shape of some of Keynes' functions and about administrative
and political feasibility, etc.

Sixteen Years Later 29 I


single scheme, through which their interplay could be coherently
investigated. His doing this does not, to my mind, constitute a revolution. Only if we accepted the myth-as I regard it-that earlier
economists ignored the part played by money, and, even when discussing Huctuations in employment, tacitly assumed that there
weren't any, would that word be appropriate. I should say rather,
that, in setting out and developing his fundamental conception,
Keynes made a very important, original and valuable addition to the
armoury of economic analysis" (pp. 65-66).
In the Keynesian passage quoted above, one of the "ultimate and
independent variables"-"exogenous variables" we would say today
-is the wage unit. This is the assumption of rigid wages. Keynes'
analytic contribution consists largely in working out the implications
of that assumption. It is now almost generally recognized that the
Keynesian theoretical system proper (apart from the discussions of
related matters and of the hints and asides that can be found in profusion in the General Theory) depends on the assumption of wage
rigidity. If that assumption is not made, the Keynesian system simply
breaks down or, to. put it differently, it loses its distinctive and differentiating quality which sets it apart from what is loosely called
the "classical" system.sa
It is true that later in the book (see Chs. 18 and 19) Keynes
extensively discusses changes in money wages, especially what will
happen if unemployment drives down money wages. In the earlier
parts of the book the oversimplified theory is put forward that when
money wages decline prices will fall exactly as much as wages and
hence unemployment can not be cured by making money wages
Hexible, in other words by introducing competition in the labor
market. In Chapter 19, in contrast, it is conceded that full employment would be restored via a fall in interest rates ("Keynes effect")
-except in the unlikely case that the liquidity schedule was en6<

Following Hicks, many writers see the difference between the Keynesian and the "classical" system in the shape of the liquidity functionthe "classical" writers assuming that it is inelastic with respect to the
rate of interest; in other words, that the velocity of circulation of
money is constant (or an exogenous variable). This does not seem to
me a very important difference and it is not correct that nobody before
Keynes had suspected that the velocity of circulation of money may be
influenced, among other things, by the rate of interest.

292

Gottfried Haberler

tirely elastic ("liquidity trap") or investment unresponsive to falling


interest rates. From the comparative static point of view, the Pigou
effect definitively disposes of the possibility that the liquidity trap
or inelastic investment schedules prevent the attainment of full
employment.
Needless to add that the Pigou effect was not meant to be an easy
policy prescription7 and that in the short run friction and expectational complications (vicious spirals of deflation, waves of pessimism,
expectation of falling prices and wages) may confuse things and
prevent the quick restoration of full employment. Some "classical"
writers realized before Keynes that it was not easy to make wages
flexible in the short run and that even if it could be done it would
not immediately eliminate all cyclical unemployment. Let me recall
that Pigou in his Industrial Fluctuations had said that in a depression
it might require "negative wages" to restore full employment quickly.
There can be no doubt, however, that the theory of the "Keynes
effect," the "Pigou effect," or more generally the "wealth-saving relation," constitute a permanent enrichment to our analytical apparatus: this we largely owe to the stimulus provided by the General
Theory. Keynes forced the classical writers to rethink, restate and to
refine their theories. 8
It should be observed, however, that Keynes carefully qualified
his theory (at least in the passage quoted above)9 by saying that the
variables mentioned can only "sometimes" be taken as independent.
He clearly wanted to restrict his theory to short-term analysis and
he surely would have strongly disapproved of using it for long-run
analysis.
See my "The Pigou Effect Once More" in Journal of Political Economy, June 1952, reprinted in my Prosperity and Depression, Harvard,
Cambridge 1958, Appendix II.
B It should be clear, however, that quite possibly wage reductions influence employment directly (rather than via the Keynes or Pigou
effect). This simple truth has been obscured by the fact that in the
Keynesian system the most direct effect that can be imagined-e.g.,
a rise in demand for, and employment of, domestic servants when
their wage falls-has to be described as being operative via a change
in the propensity to consume (or via a change in the marginal efficiency of capital if the services in question were those of engineers
and not those of household help).
9 The General Theory abounds, however, with unqualified statements
-hence everybody can find what he wants.
7

Sixteen Years Later 293


The sin of applying the short-run Keynesian theory to long-run
problems has been Hagrantly committed in the Harrod-Domar models
of secular growth. Surely in the long run Keynes' "ultimate independent variables" can no longer be taken as independent. In fact,
many of the growth models are even more restrictive than the Keynesian system because they postulate not only constant consumption
and investment functions, but constant Cor only exogenously changing) capital-output and consumption-income ratios. Growth models
have been gradually purged of their unnatural and unrealistic
rigidity and instability by Duesenberry, Kaldor, Smithies, Solow,
and others. Growth theory has, thus, slowly and laboriously worked
its way back to the "classical" position-a process which has reached
its high point, so far, in Meade's Neo-Classical Theory of Growth. 10
As far as money wages are concerned, even if they are completely
rigid in the long run, prices need not be rigid, at least not in a progressive economy. Increasing productivity would permit a gradual
lowering of the price level prOvided money wages rise less fast than
output per worker. Only if to wage rigidity downward is added an
inexorable wage push upward do we really get into a nasty dilemma,
also in the long run.
But whatever one believes to be actually the case, one thing is
clear and should have been clear from the beginning. As soon as
we assume wage rigidity and wage push, in the short or long run,
the main difference between Keynes and the classics disappears, and
with it vanishes what many Keynesians regard as Keynes' greatest
achievement-the demonstration of the possibility of an underemployment equilibrium. There is no difficulty in "classical theory"
with any amount of involuntary unemployment, if wages are rigid
and the rigid Hoor is pushed high enough.
What Lerner said in his brilliant contribution to the great debate
on the "General Theory After 25 Years" at the A.E.A. meeting in
1960 is quite true: "almost all of the 'revisions' of Keynes [that have
10

The statement that this quasi-Hegelian process-thesis, antithesis,


synthesis-has largely vindicated the "classical" thesis, because if
properly understood it turns out to be substantially identical with
the emerging synthesis, does not change the fact that the process
has been very productive of clarification and refinement. The ways
of the human mind are often devious and roundabout.

294 . Gottfried Haberler


been proposed] are to be found as implications and hints in
[the General Theory]." But is it not equally true that almost all of
Keynes' innovations (insofar as they are valid) can be found as
implications or hints (or sometimes quite explicitly) in pre-Keynesian
and post-Keynesian "classical" writings?
I do not agree with Lerner that the General Theory would necessarily have become three times as long, if Keynes had "spelled out
all hints and implications." After all, Pigou's Employment and
Equilibrium,ll a really general theory which comprises the General
Theory as a special case, is much shorter than Keynes' great book.
But Lerner is certainly right that the General Theory would have
been much less influential if the hints had been fully worked out,
if, in other words, the book had offered a really general theory. It
surely would not have been as successful as it actually was had it
been built on existing foundations and had it done justice to earlier
writers; had its author refrained from setting up a caricature of
"the classical economics" as a straw man to be knocked down; in
other words, had he written a scholarly, well-balanced treatise instead
of providing an ad hoc, makeshift theory serving as underpinning
for a combination of a policy tract, a passionate call for economic
reforms, and an impassioned indictment of orthodoxy.
2

Turning to the impact of the General Theory on economic policy,


it is certainly true that economic policy everywhere is very different
from what it was in the I920'S and I 930's; furthermore, it cannot
be denied that the change that has taken place can be broadly described as a shift in the Keynesian direction. There is much more
concern with employment, less hesitation to use deficit financing
11

Pigou relates (op. cit. p. 65) that "when I wrote my Employment and
Equilibrium in 1942 and again when I revised it recently, I had not
read the General Theory for some time and did not realize how
closely my systems of equations conform with the scheme of his
analysis."
It is, however, quite safe to say, it seems to me, that Pigou would
not have written his Employment and Equilibrium without the
Keynesian challenge. But it is equally clear that the new book, far
from contradicting classical theory, constitutes a clarification and
elaboration of Pigou's own pre-Keynesian "classical" position-if I
may repeat what I said in the earlier article.

Sixteen Years Later . 295


and little heed is paid to the canons of "sound nnance." This is still
true despite certain latter-day reactions-the "rediscovery of money
and monetary policy" and what E. Lundberg has called the "antiKeynesian counter-revolution."
I doubt very much, however, whether all that would have been
much different if Keynes had never written the General Theory.
My reasons are the following: A number of countries had already
changed their policy in the "Keynesian" direction before the appearance of the General Theory and others changed it demonstrably
without having been influenced by Keynes. Australia and Sweden
are examples of the former and Nazi Germany of the latter category.
In all these countries (including Germany) high-quality scientinc
rationalizations of expansionary policies had appeared which were
not influenced by Keynes. No single one of these writings can compare with Keynes in breadth, originality, elegance, brilliance, and
none had the sustained influence of the General Theory. But their
cumulative impact has profoundly changed the climate of opinion
and brought about the shift in economic policy. In all developed industrial countries policies of economic recovery, stabilization, and
growth have been much more successful after the second World
War than after the nrst. But it is difficult to attribute this to the
spread of Keynesian thinking. It so happens that none of the economists and economic statesmen who were largely responsible for the
assorted postwar economic miracles can be called a Keynesian: not
Camille Gutt in Belgium, nor Luigi Einaudi in Italy, nor Ludwig
Erhard in Germany, nor Reinhard Kamitz in Austria, nor Jacques
Rueff in France. The greatest economic miracle of all, the Japanese,
seems to have been performed by conservative Japanese governments
and statesmen with the help of some ultraconservative American
advisers, while the numerous Keynesians and Marxo-Keynesians had
to look on in impotent opposition.
3
I have argued that the Keynesians greatly exaggerate the originality and validity of the General Theory as well as the influence which
Keynes has exerted on the great changes that have occurred in economic policy during the last twenty-nve years. Let me emphasize
again, as I did in 1946, that I do not wish in the least to belittle

296 Gottfried Haberler


Keynes' claim to greatness and subjective originality. As I see it, what
is true of other sciences, among them the queen of sciences, mathematics, is also true of economics, pure and applied: the accumulated
mass of knowledge is so enormous that it has become impossible for
any single man, however great a genius he may be, to bring about a
real revolution. It does, of course, require men of genius to increase
the stock of knowledge, but the contribution of any single one is
small compared with the existing stock. Even the greatest stand on
the shoulders of those who went before, and in mathematics12 it has
happened many times that great discoveries firmly associated with
great names were later found to have been discovered earlier, but
were then ignored because the time simply was not ripe. The time
surely was overripe and the ground fully prepared for the message
of the General Theory. Let us admire without stint the great genius
but avoid untenable exaggerations. Hero worship is nowhere less
appropriate than in science.

12

Compare on all this N. A. Court, Mathematics in Fun and in


Earnest (Mentor, New York, 1961). Especially Chapter 1 I, "Mathematics and Genius," pp. 75-84.

PAUL M. SWEEZY

John Maynard Keynes


[1946]
LORD

KEYNES,

who died at the age of sixty-two on April

2.1, 1946, was unquestionably the most famous and controversial

of contemporary economists. Moreover, like the great figures of the


classical school-Adam Smith, David Ricardo, and John Stuart Milll
-he was no narrow specialist working in the seclusion of an academic ivory tower. Both as critic and as participant, he played a very
important and certainly a unique role in the public life of Britain in
the period of the two world wars; as a patron of the arts, he was a
power in the cultural life of his country; as head of a great insurance
company and as Bursar of King's College, Cambridge, he proved
that the economic theorist can be a highly successful businessman;
while his noneconomic writings range from the standard (literary as
opposed to mathematical) Treatise on Probability to the incisive Essays in Biography. Keynes was, in short, one of the most brilliant and
versatile geniuses of our time; and one can be sure that his place in
history-not only doctrinal economic history-will be a subject of
discussion and controversy for an indefinite period to come. It would
be presumptuous at this early date to attempt anything in the way of
definitive judgments, and in writing this brief communication I am
far from entertaining any such intentions. I think it should be possible nevertheless to set out some of the factors in Keynes' work and
in his influence on others which will have to be taken into account
in any evaluation of the man, present or future.
In order to understand Keynes one must first understand where
1

Keynes himself used the term "classical economists" to include the


subjective value theorists-especially Marshall and his followers in the
Cambridge group-of the late nineteenth and twentieth centuries. For
reasons which should be clarified by the subsequent discussion, this
practice seems to me to be misleading. It is preferable to regard lohn
Stuart Mill as the last of the classical economists and to labe the
Marshallians the "neoclassical" school.
2.97

298 Paul M. Sweezy


he stood in relation to other economists and schools of economic
thought; for, as we shall see, it was what might be called an accident
of location which accounts for much of the influence as well as for
many of the shortcomings of his work. Modern economics-the
economics of industrial capitalism and of the world market-had its
origins in the later decades of the seventeenth century. During the
next hundred and fifty years England was the home of the most important advances on both the industrial and the theoretical fronts;
and by the time of Ricardo (1772-1823), English political economy
enjoyed a degree of authority and prestige throughout the western
world which has never been equaled before or since. In the second
half of the nineteenth century, the unity of the classical tradition was
broken; what had been a single trunk with only minor offshoots
divided itself into two great branches, each with its own subbranches,
which have on the whole been growing apart ever since. These two
branches may be called the socialist or Marxian and the neoclassical
respectively. To vary our metaphor, each can and does claim to be
the legitimate child of classical political economy, but it must be
said that for brothers they have had remarkably little to do with one
another. This striking fact is due to a variety of reasons: for one
thing, the two schools have diverged in their manner of selecting and
discarding elements of the classical theory; for another, they have
(openly in the case of Marxism, under cover of a pretended scientific neutrality in the case of neoclassicism) become intellectual
weapons on opposite sides of a bitter class struggle; and, finally,
Marxism-partly no doubt as a result of the historical accident of
Marx's own German nationality-took root on the continent of Europe but failed for many years to win a significant following in the
English-speaking world. Thus the two schools, despite their common
origin, became intellectually, politically, and geographically estranged. Such contacts as they had, which were almost entirely outside Britain and the United States, were the contacts of battle and
produced intolerance rather than understanding.
When Keynes took up the study of economics about the turn of
the century, neoclassicism was in undisputed possession of the field
in the English-speaking countries; dissent was regarded as a sign of
incompetence or depravity. Keynes himself accepted the prevailing
doctrines unquestioningly and soon came to be rated as a brilliant

John Maynard Keynes' 299


but essentially orthodox representative of the neoclassical school.
There is no evidence that he was ever seriously inB.uenced by conBicting or incompatible intellectual trends. He borrowed occasionally
from foreign authors,2 and when his own ideas had finally taken
shape he was generous in giving credit for having anticipated them
to a long line of heretics and dissenters; but these were essentially
adventitious elements in Keynes' thought. By training he was astrict
neoclassicist, and he never really felt at home except in argument
with his neoclassical colleagues. In fact, one would be perfectly
justified in saying that Keynes is both the most important and the
most illustrious product of the neoclassical school.
This points, I think, to the true nature of Keynes' achievement.
His mission was to reform neoclassical economics, to bring it back
into contact with the real world from which it had wandered farther
and farther since the break with the classical tradition in the nineteenth century; and it was precisely because he was one of them and
not an outsider that Keynes could exercise such a profound inB.uence on his colleagues. The very same reasons, however, account
for the fact that, as we shall see below, Keynes could never transcend
the limitations of the neoclassical approach which conceives of economic life in abstraction from its historical setting and hence is inherently incapable of providing a scientific guide to social action.
Keynes' magnum opus, called The General Theory of Employment, Interest, and Money (1936) opens with an attack on what he
calls orthodox economics-neoclassical economics, in the terminology of this article-and sustains it almost continuously to the end.
The gist of this Keynesian criticism can be summed up simply as a
Hat rejection and denial of what has come to be known as Say's Law
of Markets3 which, despite all assertions to the contrary by orthodox
For example, the concept of a "natural rate of interest" which plays
an important part in A Treatise on Money (1930) was taken from
the Swedish economist Knut Wicksell (1851-1926). Wicksell himself, however, was essentially a neoclassicist.
S Say's Law in effect denies that there can ever be a shortage of demand in relation to production. Ricardo expressed it as follows: "No
man produces but with a view to consume or sell, and he never
sells out with an intention to purchase some other commodity which
may be useful to him, or which may contribute to future production.
By producing then, he necessarily becomes either the consumer of
his own goods, or the purchaser and consumer of the goods of some
2

300

Paul M. Sweezy

apologists, did run like a red thread through the entire body of classical and neoclassical theory. It is almost impossible to exaggerate
either the hold which Say's Law exercised on professional economists or its importance as an obstacle to realistic analysis. The
Keynesian attacks, though they appear to be directed against a variety
of specific theories, all fall to the ground if the validity of Say's Law
is assumed.
Having once got hold of the essential truth that Say's Law is a
fraud and a delusion, Keynes was obliged to search the neoclassical
theoretical structure from top to bottom to separate those propositions
which depend upon it from those which are valid regardless of its
truth or falsity. The result of this search, as it appears in the General
Theory, is almost incomprehensible to anyone but an adept in neoclassical economics. As Keynes himself says in the Preface, "the
composition of this book has been for the author a long struggle of
escape, and so must the reading of it be for most readers if the author's assault upon them is to be successful"-obviously implying
that he expects the readers to have the same type of training and the
same general background as his own. And then he adds, with refreshing candor, "the ideas which are here expressed so laboriously
are extremely simple and should be obvious. The difficulty lies, not
in the new ideas, but in escaping from the old ones, which ramify, for
those brought up as most of us have been, into every comer of our
minds."
Keynes undoubtedly exaggemtes the simplicity of his own contribution-it is noteworthy that pride in theoretical virtuosity was
utterly foreign to his nature-but I think that almost all teachers
will agree that it is easier to get his essential ideas across to a beginner than to a student who has already been steeped in the doctrines of the neoclassical school. Historians fifty years from now may
record that Keynes' greatest achievement was the liberation of AngloAmerican economics from a tyrannical dogma, and they may even
conclude that this was essentially a work of negation unmatched by
comparable positive achievements. Even, however, if Keynes were
other person. . . . Productions are always bought by productions, or
by services; money is only the medium by which the exchange is
effected." Principles of Political Economy (Gonner ed.), p. 273
and 275.

John Maynard Keynes

301

to receive credit for nothing else (which is most unlikely) his title to
fame would be secure. He opened up new vistas and new pathways
to a whole generation of economists; he will justly share the credit
for their accomplishments. 4
I have tried to show that the opportunity to which Keynes responded was essentially a crisis in traditional economics, a crisis
which was both accentuated and laid bare by the Great Depression.
He was able to demonstrate that his fellow economists, by their unthinking acceptance of Say's Law, were in effect asserting the impossibility of what was actually happening. 5 From this starting point
he was able to go on to a penetrating analysis of the capitalist economy which shows that depression and unemployment far from being
impossible, are the norms to which that economy tends, and which
explodes once and for all the myth of a harmony between private and
public interests which was the cornerstone of nineteenth century
liberalism. But Keynes stopped here in his critique of existing society.
Our troubles, he believed, are due to a failure of intelligence and
not to the breakdown of a social system; "the problem of want and
poverty and the economic struggle between classes and nations,"
he wrote in 193 I, "is nothing but a frightful muddle, a transitory
and unnecessary muddle."6
That Keynes held this view was, of course, no accident. He could
reject Say's Law and the economic conclusions based on it because
Probably only those who (like the present writer) were trained in
the academic tradition of economic thinking in the period before
1936 can fully ap'preciate the sense of liberation and the intellectual
stimulus which the General Theory immediately produced among
younger teachers and students in all the leading British and American
universities.
6 Apologists for the orthodox view are always ready with quotations to
prove that economists were never such fools as this would imply.
Keynes' answer, I think, is correct and convincing: "Contemporary
thought," he wrote, "is still deeply steeped in the notion that if
people do not spend their money in one way they will spend it in
another. Post-war economists seldom, indeed, succeed in maintaining
this standpoint consistently; for their thought to-day is too much
permeated with the contrary tendency and with facts of experience
too obviously inconsistent with their former view. But they have not
drawn sufficiently far-reaching consequences; and have not revised
their fundamental theory." General Theory, p. 20.
6 Essays in Persuasion, p. vii.
4

302

Paul M. Sweezy

he thought they were largely responsible for the muddle; but it


never occurred to him to question, still less to try to escape from, the
broader philosophical and social tradition in which he was reared.
The major unspoken premise of that tradition is that capitalism is the
only possible form of civilized society. Hence Keynes, exactly like
the economists he criticized, never viewed the system as a whole;
never studied the economy in its historical setting; never appreciated
the interconnectedness of economic phenomena on the one hand and
technological, political, and cultural phenomena on the other. Moreover, he was apparently quite ignorant of the fact that there was a
serious body of economic thought, as closely related to the classical
school as the doctrines on which he himself was brought up, which
attempted to do these things. In Keynes' eyes, Marx inhabited a
theoretical underworld along with such dubious characters as Silvio
Gesell and Major Douglas;7 and there is no evidence that he ever
thought of any of Marx's followers as anything but propagandists
and agitators.
This is not the place for a review of Marxian economics. 8 I raise
the issue only in order to show that the school of thought to which
Keynes belongs is rather isolated and one-sided, that some of his most
important discoveries were taken for granted by socialist economists
at least a generation before Keynes began to write, and that many of
the most vital problems of the capitalist system are completely ignored in the General Theory. Marx rejected Say's Law from the outset;9 already before 1900 his followers were carrying on a spirited debate among themselves not only on the subject of periodic crises but
also on the question whether capitalism could be expected to run into
a period of permanent or chronic depression. 1o Keynes ignores technological change and technological unemployment, problems which
figure as an integral part of the Marxian theoretical structure. Keynes
7 General

Theory, p. 3 2
I have tried to provide such a review in The Theory of Capitalist
Development (1942).
9 Marx remarked, in connection with the passage from Ricardo quoted
in note 3 above, that "this is the childish babbling of a Say, but unworthy of Ricardo." Theorien -aber den Mehrwert, Vol. 1 I, pt. 2,
P277
10 See The Theory of Capitalist Development, Ch. IX: "The Breakdown
Controversy."
8

John Maynard Keynes 303


treats unemployment as a symptom of a technical fault in the
capitalist mechanism, while Marx regards it as the indispensable
means by which capitalists maintain their control over the labor market. Keynes completely ignores the problems of monopoly, its distorting effect on the distribution of income and the utilization of
resources, the huge parasitic apparatus of distribution and advertising
which it foists upon the economy. A socialist can only blink his eyes
in astonishment when he reads that there is "no reason to suppose
that the existing system seriously misemploys the factors of production
which are in use. . . When 9,000,000 men are employed out of
10,000,000 willing and able to work, there is no evidence that the
labor of these 9,000,000 men is misdirected."l1 Many other examples
of the insularity and comparative narrowness of the Keynesian approach could be cited. But perhaps most striking of all is Keynes'
habit of treating the state as a deus ex machina to be invoked whenever his human actors, behaving according to the rules of the capitalist game, get themselves into a dilemma from which there is apparently no escape. Naturally, this Olympian interventionist resolves
everything in a manner satisfactory to the author and presumably to
the audience. The only trouble is---as every Marxist knows---that the
state is not a god but one of the actors who has a part to play just like
all the other actors.
Nothing that has been said should be taken as belittling the importance of Keynes' work. Moreover, there has been no intention to
imply that Marxists "know it all" and have nothing to learn from
Keynes and his followers. I have no doubt that Keynes is the greatest
British (or American) economist since Ricardo, and I think the work
of his school sheds a flood of light on the functioning of the capitalist
economy. I think there is a great deal in Marx-especially in the unfinished later volumes of Capital and in the TheO'rien iiber den
llGeneral Theory, p. 379. It is only fair to point out that Keynes'
neglect of monopoly is not characteristic of present-day academic economics. It remains true, however, that the neoclassical treatment of
the subject overconcentrates on the problems of the individual firm
and has not done very much to relate monopoly to the functioning
of the economy as a whole. In the latter field it would be hard to
name a book even today which rivals Das Finanzkapital written by
the Marxist economist Rudolf Hilferding in the first decade of the
present century.

304 . Paul M. Sweezy


Mehrwert--which takes on a new meaning and fits into its proper
place when read in the light of the Keynesian contributions. Moreover, at least in Britain and the United States, the Keynesians are as
a group far better trained and equipped technically (for instance in
the very important sphere of gathering and interpreting statistical
data) than Marxist economists;12 and as matters stand now there is
no doubt which group can learn more from the other.
But while it is right to recognize the great importance of Keynes, it
it no less essential to recognize his shortcomings. They are for the
most part the shortcomings of bourgeois thought in general: the unwillingness to view the economy as an integral part of a social whole;
the inability to see the present as history, to understand that the
disasters and catastrophes amidst which we live are not simply a
"frightful muddle" but are the direct and inevitable product of a
social system which has exhausted its creative powers, but whose
beneficiaries are determined to hang on regardless of the cost. Keynes
himself, of course, could never have recognized, let alone transcended, the limitations of the society and the class of which he was
so thoroughly a part. But the same cannot be said of many of his
followers. They did not grow up in the complacent atmosphere of
Victorian England. They were born into a world of war, and depression, and fascism. Some, no doubt, treading in the footsteps of the
master, will seek to preserve their comforting liberal illusions as long
as humanly possible. Some, in all probability, will range themselves
on the side of the existing order and will sell their skill as economists
to the highest bidder. But still others, while retaining what is valid
and sound in Keynes, will take their place in the growing ranks of
those who realize that patching up the present system is not enough,
that only a profound change in the structure of social relations can
set the stage for a new advance in the material and cultural conditions on the human race.
This last group, I think, will inevitably be attracted to Marxism
as the only genuine and comprehensive science of history and society. Perhaps the clearest indication that this is so is to be found
in Joan Robinson's little book An Essay on Marxian Economics published in England early in the war. Mrs. Robinson, a member of
12

How few there are who really deserve the name!

The First Quarter Century' 305


the inner Keynesian circle, is one of perhaps half a dozen topflight
British economic theorists. Marxists will not be able to agree with
everything she says, but they will find in her a sympathetic critic
ready and anxious to discuss problems with them in a sober and
scientific spirit. Can it be pure accident that one of the most prominent followers of Keynes should be the author of the first honest
work on Marxism ever to be written by a non-Marxist British
economist?

The First QIarter Century


[1963J
APART FROM MINOR MA'ITERS of formulation and emphasis, I am prepared to stand by what I wrote a decade and a half
ago about Keynes and Keynesian economics. In particular, I still
believe that his greatest achievements were freeing economics from
the tyranny of Say's Law and exploding the myth of capitalism as
a self-adjusting system which reconciles private and public interests.
In this connection, however, I would like to amend what I wrote
earlier, not because the amendment is of any particular importance
in itself but because it provides a logical introduction to the remarks
which constitute the substance of what follows.
Starting from the rejection of Say's Law, I wrote, Keynes "was
able to go on to a penetrating analysis of the capitalist economy which
shows that depression and unemployment, far from being impossible,
are the norms to which the economy tends." This now seems to me
to be a misleading statement of the case. Keynes in fact did not
"show" anything of the kind, at least not in the sense of furnishing
a logical, consistent, and convincing set of reasons why it should be
so. He believed that the inducement to invest was, so to speak,
naturally weak and that the structure of the system was such that a
weak inducement to invest would result in depression and unem-

306 . Paul M. Sweezy


ployment. His analysis was focused on this structure, and he gave
reasons (essentially downward rigidity of wage and interest rates)
why it could not generate full employment in the absence of a
strong inducement to invest. But he had very little indeed to say
about why the inducement to invest should be chronically weak,
and this, after all, lies at the heart of the matter.
It is not hard to infer from remarks scattered here and there
throughout the General Theory that Keynes believed that at any
given time the demand for capital is narrowly limited and hence
that the maintenance of a strong inducement to invest depends
almost entirely on what are often called exogenous factors. Perhaps
the most explicit statement of this thesis occurs on page 307:
During the nineteenth century, the growth of population and
of invention, the opening up of new lands, the state of confidence and the frequency of war over the average of (say) each
decade seem to have been sufficient, taken in conjunction with
the propensity to consume, to establish a schedule of the marginal efficiency of capital which allowed a reasonably satisfactory average level of employment to be compatible with a
rate of interest high enough to be psychologically acceptable to
wealth-owners.
So much for the nineteenth century. What of the twentieth? On
the following page, we read: "To-day and presumably for the future
the schedule of the marginal efficiency of capital is, for a variety
of reasons, much lower than it was in the nineteenth century." One
might expect to find this followed by a discussion of the "variety of
reasons," but if so one would be disappointed. Keynes turns immediately to the downward stickiness of the rate of interest and
forgets all about the Bagging marginal efficiency of capital. Nor
does a careful reading of the rest of the General Theory, or of his
other writings either for that matter, yield any but casual and often
peripheral remarks bearing on the subject. 1
About all one can say, then, is that Keynes seems to have been of
the opinion that "the growth of population and of invention, the
1

Such a canvassing of Keynes' writings for light on the long-run determinants of the inducement to invest is reported on by Alan
Sweezy, "Declining Investment Opportunity," in S. E. Harris, ed.,
The New Economics, Knopf, New York, 1947, pp. 425 ff.

The First Quarter Century . 307


opening up of new lands, the state of confidence and the frequency
of war," taken all together, had lost much of their efficacy as stimulants to the demand for capital. This is not much of a theory, and
it is even less impressive when one looks, in even the most superficial way, at the historical record. On a world scale, population
growth has speeded up rather than the reverse, and in an economic
sense there has been no lack of underdeveloped lands. Nor can
one very well complain about a shortage of wars during the twentieth century. That leaves the state of confidence, which is obviously a reflection rather than a cause of the underlying situation.
The truth would seem to be that the "vision" of capitalism as a
system always in imminent danger of falling into a state of stagnation-a vision which permeates, and in a sense even dominates,
the General Theory-is based more on intuition and generalizing
from British experience of the 1920'S and early 1930'S than on any
serious analysis of the factors affecting the inducement to invest.
Among Keynes' followers, only Hansen undertook to develop a
theory of the tendency to stagnation. This theory turns out, however,
to be little more than an elaboration, in terms of U. S. experience,
of Keynes' casual remark quoted above. According to Hansen, the
stagnation of the 1930'S was caused by declining investment opportunities uncompensated by a corresponding decline in the propensity to save. Declining investment opportunities, in turn, he traced
to a slowing up of the rate of population growth, the closing of the
frontier, and (more tentatively) an increasing capital-saving bias in
the newer technologies. This is not the place for a detailed discussion
of these various factors and their relation to the inducement to invest: I will only say that while I found Hansen's theory persuasive
when it was first put forward, subsequent criticisms and developments seem to me to have convincingly refuted it. It is the growth
of purchasing power, not population, that counts, and the relationship between the two is anything but clear. Further, the experience
of the war and immediate postwar periods suggests that population
is, economically speaking, more a dependent than an independent
variable. No one has ever succeeded in demonstrating a relation between the closing of the frontier in the 1890'S (or whenever it is
supposed to have been closed) and the demand for capital in the
1930'S. And it is anyone's guess whether either the character of

308 . Paul M. Sweezy


technological change or its effect on investment is of the kind postulated in Hansen's theory.
Thus neither Keynes nor his followers really provided an explanation of the stagnation of the 1930'S. Their whole analysis was
directed to showing that if the inducement to invest is weak, the
capitalist system, left to itself, will stagnate. But they had nothing
very enlightening to say about why it should be (or remain) weaker
than it had been in the past. Nor were other attempts at an explanation any more successful. Perhaps the most elaborate was that
of Schumpeter who saw the allegedly anti-capitalist politics of the
New Deal as the villain of the piece, a view that -was devastatingly
refuted by one of Schumpeter's own students and warmest admirers. 2
It is, I believe, no exaggeration to say that while Keynes and his
followers had done a lot to clarify the mechanics of the capitalist
system, neither they nor anyone else in the economics profession
made any real progress in solving the mystery of the extraordinary
depth and duration of the Great Depression. When World War II
came along, they gave up trying.
And yet the problem was still there and still crying out for a solution. Furthermore, clues to a promising line of attack were not
lacking. Several years before the publication of the General Theory,
bourgeois social science had effected two complementary and almost simultaneous breakthroughs in the study of capitalism. One,
associated with the names of Sraffa, Chamberlin, and Joan Robinson, was a great stride forward in the theory of noncompetitive
markets. The other, associated with the names of Berle and Means,
was the demonstration of the extent to which the U.S. economy
(and, by easy inference, others in a comparable stage of development) was dominated by huge corporations that by no stretch of the
imagination could be assumed to behave like the entrepreneurs of
classical and neoclassical theory. In retrospect, it seems odd that these
two developments did not suggest, almost at once, the need for farreaching revisions in macroeconomic theory, which in all its versions was still based on the assumption of universal competition.
I'm not sure why this didn't happen. Perhaps the explanation is
the strong grip of a deeply rooted intellectual tradition, essentially
2

Arthur Smithies, "The American Economy in the Thirties," Ameri-

can Economic Review, May 1946.

The First Quarter Century' 309


the same explanation that Keynes offered for the continued acceptance of Say's Law long after its contradiction by everyday experience should have been obvious to everyone.3 In any case, it
didn't happen, not even in Cambridge where the noncompetitive
market theories originated. In particular, the General Theory shows
absolutely no trace of "monopolistic thinking," and its tremendous
impact and prestige doubtless served as an additional bulwark of
the traditional assumption of universal competition. My own opinion now, many years later, is that it was the stubborn commitment
to this assumption-the refusal to contemplate, let alone work out
the implications of, the possibility that monopoly (or oligopoly) had
become the norm-which accounts for the failure of Keynesians
and non-Keynesians alike to make any progress in developing a
satisfactory theory of what happened during the 1930'S.
To be sure, there was one notable exception. Michal Kalecki not
only "discovered the General Theory independently,"4 he was also
the first to include what he called the "degree of monopoly" in his
overall model of the economy. Kalecki showed that an increase in
the degree of monopoly will cause a relative shift from wages to
profits and hence will act to slow down the long-run rate of increase
in output. 5 He did not, however, stress the point, apparently considering it to be of relatively minor importance.6
It was left for Josef Steindl, an Austrian economist who spent
the Hitler period at Oxford, to follow up Kalecki's tentative lead
by constructing a consistent theory of stagnation around the monopolization process which had been so dramatically documented by
Berle and Means. There is, of course, plenty of room for agreement
or disagreement with Steindl, but at any rate one would have exI am precluded from suggesting ideological prejudice as a major
factor since I do not recall that the need for a reconstruction of
macro theory in monopolistic terms occurred to any of us at the
time, whether we were on the left or on the right or in the center.
4 Joan Robinson, Economic Philosophy, London, 1962, p. 93.
5 Cf. his Theory of Economic Dynamics, London, 1954, p. 161. This
is essentially a new and revised edition of two books written more
than a decade earlier.
6 Kalecki believed that the primary cause of the slowing up of the longrun rate of growth was "a decline in the intensity of innovations"
(ibid.). This is evidently related to the Keynes-Hansen theory discussed above and is subject to the same criticisms.

310

Paul M. Sweezy

pected that his work (Maturity and Stagnation in American Capitalism, Oxford, 1952) would have been greeted with intense interest
as (1) the first full-scale welding together of the two great theoretical advances of the recent past-the theory of noncompetitive
markets on the micro level and the theory of income and employment
on the macro level-and (2) the only serious effort since Hansen
and Schumpeter to solve the mystery of the thirties. In fact, however, nothing of the sort happened. Steindl's book was greeted with
resounding silence; and I would guess that today, ten years after its
publication, a majority of the U.S. economics profession doesn't even
know of its existence. It seems quite safe to say that in the whole
history of economic thought there is not another instance of so
important a work being so completely neglected. Only Paul Baran
gave it its due, and the Marxist orientation of his Political Economy
of Growth (Monthly Review, New York, 1957) acted, as always,
to insure that this extremely important work would get the same
reception from the profession that Steindl's had already received.
Why was Steindl's Maturity and Stagnation treated this way? In
one sense, the answer is obvious. It was published in 1952, in the
middle of the Korean War boom and after more than a decade of
almost continuous full or near-full employment. To bourgeois
economists, stagnation seemed like ancient history, and they just
weren't interested.
But of course this is only the superficial aspect of the matter. What
really needs explaining is why economists are so completely dominated by the moment, how it could happen that a problem which
had been so hotly debated within the lifetime of all the economists
then alive could have been so quickly forgotten, why a serious attempt to advance their own theories should fail to arouse their purely
scientific interest.
I suggest that the explanation must be sought along two main
lines. First, there is the natural aversion of the ideological champions
of capitalism to a theory with profoundly disturbing implications for
the stability and even viability of the system. This aversion was, of
course, nothing new, and it played a big part in determining attitudes toward Keynes and the General Theory. But during the
1930'S it was impossible to deny or ignore capitalism's troubles. In the
early 1950'S, it was easy.

The First Quarter Century . 3 I

Second and more basic, there is the deeply ingrained un- and
anti-historical core alike of classical political economy and of neoclassical economics. Advanced capitalism, as it existed in Britain
in the nineteenth century and later in a handful of European and
North American countries, was and is looked upon as the end
product of economic evolution. The focus of economic theory is
this system's tendencies to equilibrium (or disequilibrium) and its
short-run Huctuations. No genuine trends-in Marxian terminology
no "laws of motion"-are conceded to exist, still less subjected to
analysis. Keynes remained strictly within this tradition, the only
difference being that he made no effort to dress up capitalist equilibria as ideal or desirable states; and Keynes' followers have also remained within it. Steindl's work, in contrast, is basically an attempt
to work out a theory of the evolution of advanced (monopoly)
capitalism. As such it is entirely alien to the whole orthodox tradition, and this in spite of the fact that Steindl's conceptual framework, like Kalecki's, is couched in terms made familiar by the
Keynesians on the one hand and the noncompetitive market theorists
on the other. The truth is that the orthodox economists, including
the Keynesians, were not equipped to understand what Steindl was
trying to do and had no standards by which to judge it. 1
It might be objected that the vogue of economic development in
the postwar period contradicts the foregoing analysis. With academic
curricula stuffed with courses on development, and textbooks and
treatises on the subject adorning every publisher's catalogue, how
can it be said that the established profession is not interested in historic trends, "laws of motion," and so on? The answer is simple. The
way bourgeois economics treats development is a connrmation of the
analysis, not a refutation. The subject is always (so far as I am
7

The enthusiastic welcome accorded to Steindl's work by a Marxist


like Baran can be explained by a similar line of reasoning. What
really defines a school of economic thought is the problems it focuses
on rather than the particular conceptual apparatus it employs. Thus
Marx used the classical concepts of value, profit, rent, etc., with
relatively unimportant modifications, though no one would think of
calling him a classical economist. These facts have been obscured in
recent times, at least among Marxists, by a tendency, particularly
strong in the Stalin period, to judge orthodoxy by what may be
called the external aspects of a theorist's work rather than by its substance.

3I 2

Paul M. Sweezy

aware) interpreted to mean the development of underdeveloped


(i.e., precapitalist or partially formed capitalist) societies. The questions asked revolve around possible and/or desirable methods of
getting them on the road to development, i.e., on the road to becoming advanced capitalist societies. The further they progress along
this road, the more they will resemble the Western European and
North American societies and the more their economies will behave
accordingly. Somewhere along the way, they will be able to throw
away the textbooks of economic development and settle down to a
steady fare of the latest edition of Samuelson's Economics. All of
this is merely a modernized version of classical political economy's
concern with how society progresses from a rude state of nature
peopled by deer and beaver hunters to its final perfected form in
the bourgeois paradise of nineteenth-century England. In neither
case is any serious attention paid to real historical processes, and in
both cases the final outcome is the negation of development.
Keynes not only did nothing to overcome this profoundly antihistorical character of received economic theory; on the contrary,
his example and prestige did much to strengthen it.S It is therefore
really no cause for surprise that Keynes' followers were incapable
of transcending his strictly limited vision of capitalist reality. Nor,
considering the usual relation between master and epigone, can
we say that it is a cause of surprise that in vital respects they took
a long step back from the position occupied by Keynes. He was
deeply convinced that capitalism is not a system of economic harmonics and that its gravest failing was precisely its chronic tendency
to depression and stagnation. His followers, with one or two honorable exceptions, cheerfully abandoned this disturbing and at the
same time challenging view and gave themselves over to comforting
speculations about full-employment equilibria, warranted growth
rates, and similar fancies.
Now that stagnation is once again with us, at least in the most
advanced and monopolistic of the capitalist countries, now that its
abeyance during the forties and early fifties can be clearly seen to
have been the result of war, the aftermath of war, and preparation
8

This is excellently symbolized by the way his remark, so contemptuous of history, that "in the long run we are all dead" is quoted ad
nauseam as a piece of profound wisdom.

The First Quarter Century 3 I 3


for war-now that these things are being driven home by practical experience to even the dullest observer of the current economic
scene, the stage would seem to be set for bourgeois economics to
attempt a new step forward. And indeed the need is clearly beginning to be recognized, as evidenced, for example, by the publication
of a special issue entitled "Time for a Keynes" by the New Republic, organ of the Establishment liberals of the Kennedy era.
Whether the need is also beginning to be met is unfortunately more
doubtful. The New Republic's contributors include some of the
most distinguished names in the U.S. economics profession, and yet,
as Professor Hansen says in his "Comment of a Keynesian" (pp.
28-29), they "do not suggest even remotely the emergence of a new
economics." It may not be out of place to suggest that before this
could happen, bourgeois economics would first have to recover the
ground lost since Keynes-to recapture his vision of capitalism, not
as the best of all possible worlds but as a system of profound contradictions and deeply rooted self-destructive tendencies.
It may also not be out of place to ask whether this is at all possible
in the epoch of the competition of the systems, when the new socialist
society is actually challenging capitalism's world leadership. Bourgeois economics, I fear, has irrevocably committed itself to what
Marx called "the bad conscience and evil intent of apologetic."9 If
I am right, Keynes may turn out to be its last great representative,
and further scientific progress will have to come from the socialist
camp (though not necessarily from the socialist countries).
In conclusion, I would like to single out a very minor aspect of
the General Theory as deserving of special emphasis and commendation a quarter of a century later. "Too large a proportion of recent
'mathematical economics,'" Keynes wrote, "are mere concoctions, as
imprecise as the initial assumptions they rest on, which allow the
author to lose sight of the complexities and interdependencies of
the real world in a maze of pretentious and unhelpful symbols."lo
If this was true in 1936, how much more so is it today when the
capacity to generate symbols is becoming the officially recognized
criterion of a "good" economist. If it be said that a mathematical
ignoramus like myself has no right to entertain or express such views,
Vol. I, Preface to Second Edition.
General Theory, p. 29 8.

9 Capital,
10

314 . Paul M. Sweezy


my answer is simple. Let the enthusiasts tell us what new and important steps forward in the understanding of capitalism have resulted from the mass production of symbols in recent years. Until
they do, I shall continue to think that Keynes was fully justified in
dismissing what he somewhat contemptuously referred to as the
"pseudo-mathematical method."ll Only it is now necessary to add
that what could perhaps have been considered a harmless fad twentyfive years ago has today taken on the dimensions of a disease threatening the health of the entire economics profession.
It is perhaps necessary to add that I intend no blanket criticism
of the use of mathematics in economics, nor, I feel sure, did Keynes.
Properly used, mathematics can be very valuable, especially as a
check on results arrived at by nonmathematical reasoning. Still less
am I attempting to disparage the important work that has been done
and is being done toward developing mathematical planning and
programming techniques. These techniques promise to be enormously valuable to planned socialist societies and undoubtedly have
some, if much more limited, applicability to capitalist enterprises and
governments. I would only add in this connection that how-to-do-it
work of this sort, however useful and even indispensable, should not
be confused with the scientific endeavor to understand the modus
operandi of human societies.
Whatever else may be said of Keynes, he was thoroughly committed to that endeavor. Younger economists (and other social scientists) today would do well to heed his example and his advice.

ll1bid., pp. 275, 297.

PAUL A. SAMUELSON

The General Theory


[1946]
T

HE DEATH of Lord Keynes will undoubtedly afford the


occasion for numerous attempts to appraise the character of the man
and his contribution to economic thought. The personal details of
his life and antecedents very properly receive notice elsewhere in
this volume.
It is perhaps not too soon to venture upon a brief and tentative
appraisal of Keynes' lasting impact upon the development of modem
economic analysis. And it is all the more fitting to do so now that
his major work has just completed the first decade of its very long life.

THE IMPACT OF THE


GENERAL THEORY
I have always considered it a priceless advantage to have been
born as an economist prior to 1936 and to have received a thorough
grounding in classical economics. It is quite impossible for modem
students to realize the full effect of what has been advisably called
"The Keynesian Revolution"l upon those of us brought up in the
orthodox tradition. What beginners today often regard as trite and
obvious was to us puzzling, novel, and heretical.
To have been born as an economist before 1936 was a boonyes. But not to have been born too long before!
Bliss was it in that dawn to be alive,
But to be young was very heaven!
The General Theory caught most economists under the age of
thirty-five with the unexpected virulence of a disease first attacking
and decimating an isolated tribe of south sea islanders. Economists
1

lowe much in what follows to discussions with my former student,


Dr. Lawrence R. Klein, whose rewarding study published by Macmillan Company bears the above title.

31 5

316 . Paul A. Samuelson


beyond fifty turned out to be quite immune to the ailment. With
time, most economists in-between began to run the fever, often
without knowing or admitting their condition.
I must confess that my own first reaction to the General Theory
was not at all like that of Keats on first looking into Chapman's
Homer. No silent watcher, I, upon a peak in Darien. My rebellion
against its pretensions would have been complete, except for an
uneasy realization that I did not at all understand what it was about.
And I think I am giving away no secrets when I solemnly averupon the basis of vivid personal recollection-that no one else in
Cambridge, Massachusetts, really knew what it was about for some
twelve to eighteen months after its publication. Indeed, until the
appearance of the mathematical models of Meade, Lange, Hicks,
and Harrod, there is reason to believe that Keynes himself did not
truly understand his own analysis.
Fashion always plays an important role in economic science; new
concepts become the mode and then are passe. A cynic might even
be tempted to speculate as to whether academic discussion is itself
equilibrating: whether assertion, reply, and rejoinder do not represent an oscillating divergent series, in which-to quote Frank
Knight's characterization of sociology-"bad talk drives out good."
In this case, gradually and against heavy resistance, the realization grew that the new analysis of effective demand associated with
the General Theory was not to prove such a passing fad, that here
indeed was part of "the wave of the future." This impression was
confirmed by the rapidity with which English economists, other than
those at Cambridge, took up the new gospel: e.g., Harrod, Meade,
and others, at Oxford; and, still more surprisingly, the young blades
at the London School, like Kaldor, Lerner, and Hicks, who threw
off their Hayekian garments and joined in the swim.
In this country it was pretty much the same story. Obviously,
exactly the same words cannot be used to describe the analysis of
income determination of, say, Lange, Hart, Harris, Ellis, Hansen,
Bissell, Haberler, Slichter, J. M. Clark, or myself. And yet the
Keynesian taint is unmistakably there upon every one of us.
Instead of burning out like a fad, today, ten years after its birth,
the General Theory is still gaining adherents and appears to be in
business to stay. Many economists who are most vehement in criticism of the specific Keynesian policies-which must always be care-

The General Theory 3 17


fully distinguished from the scientific analysis associated with his
name-will never again be the same after passing through his hands. 2
It has been wisely said that only in terms of a modem theory of
effective demand can one understand and defend the so-called
classical theory of unemployment. It is perhaps not without additional significance, in appraising the long-run prospects of the
Keynesian theories, that no individual, having once embraced the
modem analysis, has-as far as I am aware-later returned to
the older theories. And in universities, where graduate students are
exposed to the old and new income analyses, I am told that it is often
only too clear which way the wind blows.
Finally, and perhaps most important from the long-run standpoint, the Keynesian analysis has begun to filter down into the elementary textbooks; and, as everybody knows, once an idea gets into
these, however bad it may be, it becomes practically immortal.
THE GENERAL THEORY
Thus far, I have been discussing the new doctrines without regard
to their content or merits, as if they were a religion and nothing
else. True, we find a gospel, a scriptures, a prophet, disciples, apostles,
epigoni, and even a duality; and if there is no apostolic succession,
there is at least an apostolic benediction. But by now the joke has
worn thin, and it is in any case irrelevant.
The modem saving-investment theory of income determination
did not directly displace the old latent belief in Say's Law of Markets
(according to which only "frictions" could give rise to unemployment and overproduction). Events of the years following 1929 destroyed the previous economic synthesis. The economists' belief in
the orthodox synthesis was not overthrown, but had simply atrophied:
it was not as though one's soul had faced a showdown as to the existence of the deity and that faith was unthroned, or even that one
had awakened in the morning to find that belief had Hown away
in the night; rather it was realized with a sense of belated recognition that one no longer had faith, that one had been living without
faith for a long time, and that what, after all, was the difference?
The nature of the world did not suddenly change on a black
2

For a striking example of the effect of the Keynesian analysis upon a


great classical thinker, compare the fructiferous recent writings of
Professor Pigou with his earlier Theory of Unemployment.

318 . Paul A. Samuelson


October day in 1929 so that a new theory became mandatory. Even
in their day, the older theories were incomplete and inadequate: in
1815, in 1844, 1893, and 1920. I venture to believe that the eighteenth and nineteenth centuries take on a new aspect when looked
back upon from the modern perspective, that a new dimension has
been added to the rereading of the mercantilists, Thornton, Malthus,
Ricardo, Tooke, David Wells, Marshall, and Wicksell.
Of course, the Great Depression of the thirties was not the first
to reveal the untenability of the classical synthesiS. The classical
philosophy always had its ups and downs along with the great swings
of business activity. Each time it had come back. But now for the
first time, it was confronted by a competing system-a well-reasoned
body of thought containing among other things as many equations
as unknowns; in short, like itself, a synthesis; and one which could
swallow the classical system as a special case.
A new system, that is what requires emphasis. Classical economics
could withstand isolated criticism. Theorists can always resist facts;
for facts are hard to establish and are always changing anyway, and
ceteris paribus can be made to absorb a good deal of punishment.
Inevitably, at the earliest opportunity, the mind slips back into the
old grooves of thought, since analysis is utterly impossible without
a frame of reference, a way of thinking about things, or, in short,
a theory.3
Herein lies the secret of the General Theory. It is a badly written book, poorly organized; any layman who, beguiled by the author's previous reputation, bought the book was cheated of his five
shillings. It is not well suited for classroom use. 3> It is arrogant, badtempered, polemical, and not overly generous in its acknowledgments. It abounds in mares' nests or confusions: involuntary unemployment, wage units, the equality of savings and investment, the
timing of the multiplier, interactions of marginal efficiency upon
This tendency holds true of everybody, including the businessman
and the politician, the only difference being that practical men think
in terms of highly simplified (and often contradictory) theories. It
even holds true of a literary economist who would tremble at the sight
of a mathematical symbol.
8> The dual and confused theory of Keynes and his followers concerning the "equality of savings and investment" unfortunately ruled out
the possibility of a redagogically clear exposition of the theory in
terms of schedules 0 savings and investment determining income.
8

The General. Theory . 3 19


the rate of interest, forced savings, own rates of interest, and many
others. In it the Keynesian system stands out indistinctly, as if the
author were hardly aware of its existence or cognizant of its properties; and certainly he is at his worst when expounding its relations
to its predecessors. Flashes of insight and intuition intersperse tedious algebra. An awkward definition suddenly gives way to an unforgettable cadenza. When finally mastered, its analysis is found
to be obvious and at the same time new. In short, it is a work of
genius.
It is not unlikely that future historians of economic thought will
conclude that the very obscurity and polemical character of the
General Theory ultimately served to maximize its long-run influence. Possibly such an analyst will place it in the first rank of theoretical classics, along with the work of Smith, Cournot, and Walras.
Certainly, these four books together encompass most of what is vital
in the field of economic theory; and only the first is by any standards
easy reading or even accessible to the intelligent layman.
In any case, it bears repeating that the General Theory is an
obscure book, so that would-be anti-Keynesians must assume their
position largely on credit unless they are willing to put in a great
deal of work and run the risk of seduction in the process. The General Theory seems the random notes over a period of years of a
gifted man who in his youth gained the whip hand over his publishers by virtue of the acclaim and fortune resulting from the success
of his Economic Consequences of the Peace.
Like Joyce's Finnegan's Wake, the General Theory is much in
need of a companion volume providing a "skeleton key" and guide
to its contents: warning the young and innocent away from Book
I (especially the difficult Chapter 3) and on to Books III, IV, and
VI. Certainly in its present state, the book does not get itself read
from one year to another even by the sympathetic teacher and
scholar.
Too much regret should not be attached to the fact that all hope
must now be abandoned of an improved second edition, since it is
the first edition which would in any case have assumed the stature
of a classic. We may still paste into our copies of the General Theory
certain subsequent Keynesian additions, most particularly the famous chapter in How to Pay for the War which first outlined the
modern theory of the inflationary process.

320

Paul A. Samuelson

This last item helps to dispose of the fallacious belief that


Keynesian economics is good "depression economics" and only that.
Actually, the Keynesian system is indispensable to an understanding of conditions of over-effective demand and secular exhilaration;
so much so that one anti-Keynesian has argued in print that only in
times of a great war boom do such concepts as the marginal propensity to consume have validity. Perhaps, therefore, it would be more
nearly correct to aver the reverse: that certain economists are Keynesian fellow travelers only in boom times, falling off the band wagon
in depression.
If time permitted, it would be instructive to contrast the analysis
of inflation during the Napoleonic and first World War periods
with that of the recent War and correlate this with Keynes' influence.
Thus, the "inflationary gap" concept, recently so popular, seems
to have been first used around the spring of 1941 in a speech by
the British Chancellor of the Exchequer, a speech thought to have
been the product of Keynes himself.'
No author can complete a survey of Keynesian economics without
indulging in that favorite indoor guessing game: wherein lies the
essential contribution of the General Theory and its distinguishing
characteristic from the classical writings? Some consider its novelty
to lie in the treatment of the demand for money, in its liquidity
preference emphasis. Others single out the treatment of expectations.
I cannot agree. According to recent trends of thought, the interest
rate is less important than Keynes himself believed; therefore, liquidity preference (which itself explains part of the lack of importance
of the interest rate, but only part) cannot be of such crucial significance. As for expectations, the General Theory is brilliant in
calling attention to their importance and in suggesting many of the
central features of uncertainty and speculation. It paves the way
for a theory of expectations, but it hardly provides one. 5
4

In the present writer's opinion this "neo-Austrian" demand analysis


of inflation has, if anything, been overdone; there is reason to suspect
that the relaxations of price controls during a period of insufficient
general demand might still be followed by a considerable, self-sustaining rise in prices.
See Chapter XXXI, [of Seymour Harris, ed., The New Economics.
Knopf, New York, 1947].

The General Theory'

321

I myself believe the broad significance of the General Theory


to be in the fact that it provides a relatively realistic, complete system for analyzing the level of effective demand and its Huctuations.
More narrowly, I conceive the heart of its contribution to be in that
subset of its equations which relate to the propensity to consume
and to saving in relation to offsets-to-saving. In addition to linking
saving explicitly to income, there is an equally important denial
of the implicit "classical" axiom that motivated investment is indefinitely expansible or contractable, so that whatever people try to
save will always be fully invested. It is not important whether we
deny this by reason of expectations, interest rate rigidity, investment
inelasticity with respect to overall price changes and the interest
rate, capital or investment satiation, secular factors of a technological
and political nature, or what have you. But it is vital for businesscycle analysis that we do assume definite amounts of investment
which are highly variable over time in response to a myriad of
exogenous and endogenous factors, and which are not automatically
equilibrated to full-employment saving levels by any internal efficacious economic process.
With respect to the level of total purchasing power and employment, Keynes denies that there is an invisible hand channeling the
self-centered action of each individual to the social optimum. This
is the sum and substance of his heresy. Again and again through his
writings there is to be found the figure of speech that what is needed
are certain "rules of the road" and governmental actions, which will
benefit everybody, but which nobody by himself is motivated to
establish or follow. Left to themselves during depression, people
will try to save and only end up lowering society's level of capital
formation and saving; during an inHation, apparent self-interest
leads everyone to action which only aggravates the malignant upward spiral.
Such a philosophy is profoundly capitalistic in its nature. Its
policies are offered "as the only practicable means of avoiding the
destruction of existing economic forms in their entirety and as the
condition of the successful functioning of individual initiative."6
From a perusal of Keynes' writing, I can find no evidence that
G

General Theory, p. 3 80.

322

Paul A. Samuelson

words like these resemble the opportunistic lip service paid in much
recent social legislation to individual freedom and private enterprise.
The following quotations show how far from a radical was this
urbane and cosmopolitan provincial English liberal:
How can I accept [the communistic] doctrine which sets up
as its bible, above and beyond criticism, an obsolete economic
textbook which I know to be not only scientifically erroneous
but without interest or application for the modern world? How
can I adopt a creed which, preferring the mud to the fish, exalts
the boorish proletariat above the bourgeois and intelligentsia
who, with all their faults, are the quality of life and surely carry
the seeds of all human advancement. Even if we need a religion, how can we find it in the turbid rubbish of the Red
bookshops? It is hard for an educated, decent, intelligent son of
Western Europe to find his ideals here, unless he has first
suffered some strange and horrid process of conversion which
has changed all his values ....
So, now that the deeds are done and there is no going back,
I should like to give Russia her chance; to help and not to
hinder. For how much rather, even after allowing for everything, if I were a Russian, would I contribute my quota of
activity to Soviet Russia than to Tsarist Russia. 7
Nothing that I can find in Keynes' later writings shows any significant changes in his underlying philosophy. As a result of the
Great Depression, he becomes increasingly impatient with what he
regards as the stupidity of businessmen who do not realize how
much their views toward reform harm their own true long-run interests. But that is all.
With respect to international cooperation and autonomy of national policies, Keynes did undergo some changes in belief. The
depression accentuated his post-World War I pessimism concerning
the advisability of England or any other country's leaving itself to
the mercy of the international gold standard. But in the last half
dozen years, he began to pin his hopes on intelligent, concerted,
multilateral cooperation, with, however, the important proviso that
each nation should rarely be forced to adjust her economy by deflationary means.

7J.

M. Keynes, Essays in Persuasion (1932), pp. 300 and 31I.

The General Theory 323


PORTRAIT OF THE SCIENTIST
There is no danger that historians of thought will fail to devote
attention to all the matters already discussed. Science, like capital,
grows by accretion, and each scientist's offering at the altar blooms
forever. The personal characteristics of the scientist can only be
captured while memories are still fresh; and only then, in all honesty,
are they of maximum interest and relevance.
In my opinion, nothing in Keynes' previous life or work really
quite prepares us for the General Theory. In many ways his career
may serve as a model and prescription for a youth who aspires to
be an economist. First, he was born into an able academic family
which breathed in an atmosphere of economics; his father was a
distinguished scholar, but not so brilliant as to overshadow and stunt
his son's growth.
He early became interested in the philosophical basis of probability theory, thus establishing his reputation young in the technical
fields of mathematics and logic. The Indian Currency and Finance
book and assiduous service as assistant editor and editor of the
Economic Journal certified to his "solidity" and scholarly craftsmanship. His early reviews, in the Economic Journal, of Fisher,
Hobson, Mises, and of Bagehot's collected works, gave hints of the
brilliance of his later literary style. The hiatus of the next few years
in his scientific output is adequately explained by his service in the
Treasury during the first World War.
The first extreme departure from an academic career comes, of
course, with the Byronic success of the Economic Consequences of
the Peace, which made him a world celebrity whose very visits to
the Continent did not go unnoticed on the foreign exchange markets.
As successful head of an insurance company and bursar of King's
College, he met the practical men of affairs on their own ground
and won the reputation of being an economist who knew how to
make money. All this was capped by a solid two-volume Treatise on
Money, replete with historical accounts of the Mycenean monetary
system and the rest. Being a patron of the ballet and theater, a
member of the "Bloomsbury set" of Virginia Woolf and Lytton
Strachey, a governor of the Bank of England, and peer of the realm
simply put the finishing gilt on his portrait.

324 . Paul A. Samuelson


Why then do I say that the General Theory still comes as a surprise? Because in all of these there is a sequence and pattern, and
no one step occasions real astonishment. The General Theory, however, is a mutant, notwithstanding Keynes' own expressed belief that
it represents a "natural evolution" in his own line of thought. Let
me turn, therefore, to his intellectual development.
As far back as his 1911 review of Irving Fisher's Purchasing
Power of Money,S Keynes expressed dissatisfaction with a mechanical quantity theory of money, but we have no evidence that he
would have replaced it with anything more novel than a Cambridge
cash balance approach, amplified by a more detailed treatment of
the discount rate. All this, as he would be the first to insist, was
very much in the Marshallian oral tradition, and represents a view
not very different from that of, say, Hawtrey.
Early in life he keenly realized the obstacles to deRation in a
modern capitalistic country and the grief which this process entailed.
In consequence of this intuition, he came out roundly against going
back to the prewar gold parity. Others held the same view: Rist
in France, Cassel in Sweden, et al. He was not alone in his insistence, from the present fashionable point of view vastly exaggerated, that central bank discount policy might stabilize business
activity; again, compare the position of Gustav Cassel. Despite the
auspicious sentence concerning savings and investment in its preface,
the Tract on Monetary Reform on its analytical side goes little beyond a quantity theory explanation of inRation; while its policy
proposals for a nationally managed currency and Ructuating exchange are only distinguished for their political novelty and persuasiveness.
In all of these, there is a consistency of pattern. And in retrospect it is only fair to say that he was on the whole right. Yet this
brief account does not present the whole story. In many places, he
was wrong. Perhaps a pamphleteer should be judged shotgun rather
than riRe fashion, by his absolute hits regardless of misses; still one
8

This is a characteristically "unfair" and unfavorable review, to be


compared with Marshall's review of Jevons, which Keynes' biography
of Marshall tries weakly to justify. That Keynes' first publication of
a few years earlier was a criticism does not astonish us in view of his
later writings.

The General Theory . 325


must note that, even when most wrong, he is often most confident
and sure of himself.
The Economic Consequences of the Peace proceeds from beginning to end on a single premise which history has proved to be
false or debatable. Again, he unleashed with a flourish the Malthusian
bogey of overpopulation at a time when England and the Western
European world were undergoing a population revolution in the
opposite direction. In his controversy with Sir William Beveridge
on the terms-of-trade between industry and agriculture, besides being
wrong in principle and interpretation, he revealed his characteristic
weakness for presenting a few hasty, but suggestive, statistics. If it
can be said that he was right in his reparations transfer controversy
with Ohlin, it is in part for the wrong reasons-reasons which in
terms of his later system are seen to be classical as compared to the
arguments of Ohlin. Again, at different times he has presented
arguments to demonstrate that foreign investment is (I) deflationary,
and (2) stimulating to the home economy, without appearing on
either occasion to be aware of the opposing arguments.
None of these are of vital importance, but they help to give the
flavor of the man. He has been at once soundboard, amplifier, and
initiator of contemporary viewpoints, whose strength and weakness
lay in his intuition, audaciousness, and changeability. Current quips
concerning the latter trait are rather exaggerated, but they are not
without provocation. It is quite in keeping with this portrait to be
reminded that in the early twenties, before he had an inkling of
the General Theory, or even the Treatise, he scolded Edwin Cannan in no uncertain terms for not recognizing the importance and
novelty of modem beliefs as compared to old-fashioned-I might
almost have said "classical"-theories.
Where a scientist is concerned, it is not inappropriate, even in
a eulogy, to replace the ordinary dictum nihil nisi bonum by the
criterion nihil nisi verum. In all candor, therefore, it is necessary
to point out certain limitations-one might almost say weaknesses,
were they not so intrinsically linked with his genius-in Keynes'
thought.
Perhaps because he was exposed to economics too young, or perhaps because he arrived at maturity in the stultifying backwash
of Marshall's influence upon economic theory-for whatever reason,

326 . Paul A. Samuelson


Keynes seems never to have had any genuine interest in the theory
of value and distribution. It is remarkable that so active a brain
would have failed to make any contribution to economic theory; and
yet except for his discussion of index numbers in Volume I of the
Treatise and for a few remarks concerning "user cost," which are
novel at best only in terminology and emphasis, he seems to have
left no mark on pure theory.9
Just as there is internal evidence in the Treatise on Probability
that he early tired of somewhat frustrating basic philosophic speculation, so he seems to have early tired of theory. He gladly "exchanged the tormenting exercises of the foundations of thought and
of psychology, where the mind tries to catch its own tail, for the
delightful paths of our own most agreeable branch of the moral
sciences, in which theory and fact, intuitive imagination and practical judgment, are blended in a manner comfortable to the human
intellect" (Essays in Biography, pp. 249-50).
In view of his basic antipathy to economic theory, it is all the
more wonder, therefore, that he was able to write a biography of
Alfred Marshall which Professor Schumpeter has termed one of the
best treatments of a master by a pupiPo Never were two temperaments more different than those of the two men, and we can be
sure that the repressed Victorianism and "popish" personal mannerisms which Keynes found so worthy of reverence in a master and
father would have been hardly tolerable in a contemporary.
From Marshall's early influence, no doubt, stems Keynes' antipathy toward the use of mathematical symbols, an antipathy which
already appears, surprisingly considering its technical subject, in the
early pages of the Treatise on Probability. In view of the fact that
Indeed, only in connection with Frank Ramsey's article on "A Mathematical Theory of Saving" eEJ, 1928) does he show interest in an
esoteric theoretical problem; there he gave a rather intricate interpretation in words of a calculus-of-variations differential equation condition of equilibrium. His reasoning is all the more brilliant-and I
say this seriously!-because it is mathematically unrigorous, if not
wrong. The importance which Keynes attached to this article is
actually exaggerated and can be accounted for only in terms of his
paternal feeling toward Ramsey and his own participation in the
solution of the problem.
10 Keynes' discussion of Marshall's monetary theory is much better than
his treatment of Marshall's contribution to theory.
9

The General Theory . 327


mathematical economists were later to make some of the most important contributions to Keynesian economics, his comments on
them in the General Theory and in the Marshall and Edgeworth
biographies merit rereading. l l
Moreover, there is reason to believe that Keynes' thinking remained fuzzy on one important analytical matter throughout all
his days: the relationship between "identity" and functional (or
equilibrium-schedule) equality; between "virtual" and observable
movements; between causality and concomitance; between tautology
and hypothesis. Somewhere, I believe in his early writings, he already
falls into the same analytic confusion with respect to the identity of
supply and demand for foreign exchange which was later to be his
stumbling block with respect to the identity of saving and investment.
Perhaps he was always too busy with the affairs of the world to
be able to devote sufficient time for repeated thinking through of
certain basic problems. Certainly he was too busy to verify references
("a vain pursuit")' His famous remark that he never learned anything from reading German which he didn't already know would
be greeted with incredulity in almost any other science than economics. 12 What he really meant was that his was one of those original
minds which never accepts a thing as true and important unless he
has already thought it through for himself. Despite his very considerable erudition in certain aspects of the history of thought, there
was probably never a more ahistorical scholar than Keynes.
Finally, to fill in the last little touch in this incomplete portrait
of an engaging spirit, I should like to present a characteristic quotation from Keynes:
Keynes' critical review of Tinbergen's econometric business cycle
study for the League of Nations reveals that Keynes did not really
have the necessary technical knowledge to understand what he was
criticizing. How else are we to interpret such remarks as his assertion
that a linear system can never develop oscillations?
12 Around 191 I - I 5, he was the principal reviewer of German books for
E]; also he must have read-at least he claimed to have-innumerable
German works on probability. That he could not speak German with
any fluency is well attested by those who heard him once open an
English lecture to a German audience with a brief apology in German.
11

328 . Paul A. Samuelson


In writing a book of this kind, the author must, if he is to put
his point of view clearly, pretend sometimes to a little more
conviction than he feels. He must give his own argument a
chance, so to speak, nor be too ready to depress its vitality with
a wet cloud of doubt.
Is this from the General Theory? No. From the Treatise on
Money or the Tract? No and no. Even when writing on so technical
a subject as probability, the essential make-up of the man comes
through, so that no literary detective can fail to spot his spoor.

THE ROAD TO THE


GENERAL THEORY

It was not unnatural for such a man as I have described to wish


as he approached fifty to bring together, perhaps as a crowning life
work, his intuitions concerning money. Thus the Treatise was born.
Much of the first volume is substantial and creditable, though hardly
exciting. But the fundamental equations which he and the world
considered the really novel contribution of the Treatise are nothing
but a detour and blind alley.
The second volume is the more valuable, but it is so because of
the intuitions there expressed concerning bullishness, bearishness,
etc. And even these might have been prevented from coming into
being by too literal an attempt to squeeze them into the mold of the
fundamental equations. Fortunately, Keynes was not sufficiently
systematic to carry out such a program.
Before the Treatise was completed, its author had already tired
of it. Sir Isaac Newton is alleged to have held up publication of his
theory for twenty years because of a small discrepancy in numerical
calculation. Darwin hoarded his theories for decades in order to
collect ever more facts. Not so with our hero: let the presses roll
and throw off the grievous weight of a book unborn! Especially
since a world falling to pieces is ripe to drop Pollyanna and take up
with Cassandra on the rebound.
Perhaps not being systematic proved his salvation. A long line
of heretics testifies that he is not the first to have tried to weld intuition into a satisfactory unified theory; not the first to have shot his
bolt and failed. But few have escaped from the attempt with their
intuition intact and unmarred. In an inexact subject like economics,

The General Theory' 329


concepts are not (psychologically) neutral. Decisions based upon
ignorance or the equi-probability of the unknown are not invariant
under transformation of coordinates or translation of concepts. Simply to define a concept is to reify it, to breathe life in it, to create
a predisposition in favor of its constancy; viz., the falling rate of
profit and the organic composition of capital, the velocity of circulation of money, the propensity to consume, and the discrepancy
between saving and investment.
The danger may be illustrated by a particular instance. Shrewd
Edwin Cannan, in characteristic salty prose, throughout the first
World War "protested."13 At first his insights were sharp and incisive, his judgments on the whole correct. But in the summer of
1917, to "escape from an almost unbearable personal sorrow," he
undertook to set forth a systematic exposition of the theory of money.
The transformation of Cinderella's coach at the stroke of twelve is
not more sudden than the change in the quality of his thought.
Here, I am not so much interested in the fact that his voice becomes
shrill, his policies on the whole in retrospect bacl-as in the fact
that his intuitions were perverted and blunted by his analysis, almost in an irrecoverable way! Not so with Keynes. His constitution
was able to throw off the Treatise and its fundamental equations.
While Keynes did much for the Great Depression, it is no less
true that the Great Depression did much for him. It provided challenge, drama, experimental confirmation. He entered it the sort of
man who might be expected to embrace the General Theory if it
were explained to him. From the previous record, one cannot say
more. Before it was over, he had emerged with the prize in hand,
the system of thought for which he will be remembered.
Right now I do not intend to speculate in detail on the thought
process leading up to this work, but only to throw out a few hints. In
the 1929 pamphlet, Can Lloyd George do it? written with H. D.
Henderson, Keynes set up important hypotheses concerning the
effects of public works and investment. It remained for R. F. Kahn,
that elusive figure who hides in the prefaces of Cambridge books,
to provide the substantiation in his justly famous 1931 Economic
Journal article, "The Relation of Home Investment to Unemploy-

18E. Cannan, An Economist's Protest (1927).

330 . Paul A. Samuelson


ment." Quite naturally the "multiplier" comes in for most attention,
which is in a way too bad, since the concept often seems like nothing
but a cheap-jack way of getting something for nothing and appears
to carry with it a spurious numerical accuracy.
But behind lies the vitally important consumption function:
giving the propensity to consume in terms of income; or looked
at from the opposite side, specifying the propensity to save. With
investment given, as a constant or in the schedule sense, we are in
a position to set up the simplest determinate system of underemployment equilibrium-by a "Keynesian savings-investment-income cross"
not formally different from the "Marshallian supply-demand-price
cross."
Immediately everything falls into place: the recognition that the
attempt to save may lower income and actually realized saving; the
fact that a net autonomous increase in investment, foreign balance,
government expenditure, consumption will result in increased income
greater than itself, etc., etc.
Other milestones on the road to Damascus, in addition to the
Lloyd George pamphlet and the Kahn article, were Keynes' contributions to a report of the Macmillan Committee14 and his University of Chicago Harris Foundation lectures on unemployment in
the summer of 193 I. In these lectures, Keynes has not quite liberated
himself from the terminology of the Treatise (vide his emphasis on
"profits"); but the notion of the level of income as being in equilibrium at a low level because of the necessity for savings to be equated
to a depressed level of investment is worked out in detail.
From here to the Means to Prosperity (1933) is but a step; and
from the latter to the General Theory but another step. From hindsight and from the standpoint of policy recommendations, each such
step is small and in a sense inevitable; but from the standpoint of
14

Young economists who disbelieve in the novelty of the Keynesian


analysis, on the ground that no sensible person could ever have
thought differently, might with profit read Hawtrey's testimony before the Macmillan Committee, contrasting it with the Kahn article
and comparing it with Tooke's famous demonstration in his History
of Prices, Volume I, that government war expenditures as such cannot l'ossibly cause inflation-because what the government spends

would have been spent anyway, except to the extent


created.

of "new

money"

A Brief Post-Keynesian Survey . 331


having stumbled upon and fonnulated a new system of analysis,
each represents a tremendous stride.
But now I shall have to desist. My panegyric must come to an
end with two conHicting quotations from the protean Lord Keynes
between which the jury must decide:
In the long run we are all dead.
. . . The ideas of economists and political philosophers, both
when they are right and when they are wrong, are more powerful than is commonly understood. Indeed, the world is ruled by
little else. Practical men, who believe themselves to be quite
exempt from any intellectual inHuences, are usually the slaves
of some defunct economist. Madmen in authority, who hear
voices in the air, are distilling their frenzy from some academic
scribbler of a few years back. I am sure that the power of vested
interests is vastly exaggerated compared with the gradual encroachment of ideas ... Soon or late, it is ideas, not vested interests, which are dangerous for good or evil,15

A Brief Survey of Post. .


Keynesian Developments
[1963J
MUCH HAS HAPPENED in the quarter century since
Keynes published his General Theory. What are some of the changes
that a second edition might call for? Had he lived Keynes would be
eighty today, hard as that is to imagine. And having worked all his
life (overworked, according to his friend Bertrand Russell), Keynes
might be excused from the chore of preparing a revision-particularly
since, for better or worse, the General Theory is a classic and even

332 . Paul A. Samuelson


its creator must not tamper with a classic. The pyramids may not be
perfect, but they are the most perfect pyramids we shall ever know.
Still Keynes at eighty would be younger than most of us at thirty.
Toward the end of his life he declared that, if he had it to do all over
again, he would drink more champagne. Heaven protect us from
that! As William James said, there are some men who are born with
a bottle of champagne to their credit; in the case of Keynes, it must
have been a jeroboam. Believing in creative obsolesence, Keynes
would not be content to change a few commas and improve a few
derivatives. Instead he would write a sequel, which might be entirely
new-as different from the General Theory as that was from the
Treatise on Money. The body of economics could not take another
such shock in the same century.
In the long run we are each dead. And now posterity must take
care of itself without the promptings from J. M. Keynes. Very
briefly, therefore, I shall give a personal interpretation of some major
trends in macroeconomic analysis that have grown out of the General
Theory.

THE PIGOU EFFECT


The most shocking view in the General Theory was the allegation
that economic equilibrium need not produce full employment.
Economists like Schumpeter found this to be simply incredible.
Smith's Invisible Hand was brought under direct attack. This was
revolution, not evolution.
What is most shocking in a book is not necessarily most important and lasting. Had Keynes begun his first few chapters with the
simple statement that he found it realistic to assume that modern
capitalistic societies had money wage rates that were sticky and resistant to downward movements, most of his insights would have
remained just as valid-particularly in view of his later discussions
that showed how money wage cuts were, apart from interest rate
changes, by no means followed by commensurate real wage cuts.
We usually associate with the wartime writings of A. C. Pigou the
demonstration that within a Keynesian system full employment can
in all probability be fully restored by the following mechanism:
Let unemployment drive down money wage rates and price
levels to, say, one-hundredth of their previous levels; the real

A Brief Post-Keynesian Survey . 333


wealth of people embodied in their hard coin, their fiat currency, and [part of] their holdings of government securities
would then be one hundred times as great as previously; at the
same level of real current income, a man possessed of much
real wealth can be expected to want to save less and consume
more; hence, this Pigou effect can shift downward the community's real saving schedule enough to make it intersect even
a low investment schedule at full employment.
While Pigou, rightly in my opinion, did not consider this mechanism a useful one for policy, it did serve to save face and honor for
the believers in the harmony of equilibrium. This was a worthy
achievement, purchased at little cost. And now the air was cleared to
tackle matters of substance and not of ideology.
Later writers, such as Oskar Lange, Don Patinkin, and Franco
Modigliani pursued these issues further. 1 Writers before Pigou, such
as Gottfried Haberler, in his Prosperity and Depression, had of
course anticipated the Pigou effect. Indeed from 1932 to 1936 at
Chicago and Harvard I was generally taught that at low enough
wages all could get employed, although just exactly what was the
general demand curve for labor along which we were to travel was
never made clear to me. Here are two instances: I can remember
Wassily Leontief saying: "If wages are low enough, this dime in my
hand will employ everyone in the nation; and my only requirement
on them is that they not show up at my office for work." That was
around 1935-6, and, although the term had not yet been invented, I
was left with the feeling that some kind of "disguised unemployment" was involved in the Leontief operation. Any student of the
quantity theory who had been assigned the works of such a writer as
Sir Dennis Robertson had before 1936 the feeling that prices would
fall enough so that an irreducible MV could form a full-employment
output ratio Q=MV"-7-P. It is noteworthy that Pigou's Theory of
1

Space does not permit discussion of the revival of the sophisticated


quantity theory of money in recent years by Friedman, Modigliani,
and others. Let me only note that in the most abstract model, instead
of having the absolute price level be proportional to M narrowly defined, it should be made a homogeneous function of the first degree in
the variables (M,GB), where GB is the total of interest-bearing
public debt unbacked by public capital assets. If n is the finite average expectancy of human life, the form of the homogeneous function
varies with n and many other variables.

334 . Paul A. Samuelson


Unemployment, written just before the General Theory and to which
Keynes was partially reacting, did not anticipate the Pigou effect. I
have long argued that the General Theory provided the tools of
analysis for classical writers to understand and defend their own
views. If only Say, James Mill, and Ricardo had lived after 1936,
think of what sense they might have made of Say's Law!
STOCKS AND FLOWS

Much of the General Theory is concerned with flows of income,


with stock effects being implicit rather than explicit. (Liquidity
preference and money provide an obvious exception.) An important
development subsequent to 1936 was to make explicit the role of
stocks and to give them increased emphasis. Indeed, the Pigou effect
just discussed could be placed under this heading.
Sir Roy Harrod, in his 1936 Trade Cycle, had already made explicit the accelerator relationship between capital stock and output
How, foreshadowing his famous 1939 Economic Journal article and
1948 book on Dynamic Economics; Evsey Domar was led to similar
analysis, as were Alvin H. Hansen and myself. Even prior to the
General Theory, Gustav Cassel had worked with what would today
be called Harrod-Domar growth paths, and in the 1935 Econometrica, the Hungarian economist Edward Theiss had published a
similar model. 2
Nicholas Kaldor, in a justly famous 1940 Economic Journal article,
proposed a model of the trade cycle which made investment explicitly a function directly of the level of income and inversely of the
stock of capital. This may be considered either as an alternative to
the acceleration models; or, as Hollis B. Chenery, Richard M. Goodwin, and others have shown, as a generalization which includes the
"tight accelerator" as the special case when investment succeeds in
adjusting so rapidly to the discrepancy between actual and desired
capital as to lead to a reasonably stable capital-output ratio. From the
standpoint of business cycle analysis, the Kaldor model placed
2

Since the General Theory, there has grown up a fascination with exponential rates of growth of the Harrod-Domar and J. Robinson type,
and which Sir Dennis Robertson had experimented with in the
1920'S. Some later remarks touch on these matters but I shall not
here attempt a survey of this area.

A Brief Post-Keynesian Survey . 335


emphasis on nonlinearity of structure, thereby making possible a
theory of determinate amplitude of fluctuations. (Harrod had elements of this in the Trade Cycle; and my 1939 Journal of Political
Economy synthesis of accelerator and multiplier had shown how
nonlinearity could accomplish this.) Richard M. Goodwin, in
numerous articles, lectures, and in a long-awaited book, had explored
such nonlinear models in the war and postwar years. The Trade
Cycle (1950) of J. R. Hicks built on nonlinearity, combining it with
exogenous exponential trends.
THEORIES OF THE
CONSUMPTION FUNCTION
Most of the stock effects above put emphasis on real capital relationships. At the same time, writers like Franco Modigliani and
James Tobin 3 were emphasizing the role of financial wealth effects
on the propensity to consume. In particular, Modigliani and Brumberg worked out in detail the implications of a model involving lifetime patterns of saving and consuming. And in recent years American economists have been much interested in Milton Friedman's
hypothesis that permanent income (as against transitory income)
gets spent in about the same percentage whether its average level is
$5,000 or $15,000 per year.
A little background to this discussion may be usefully sketched.
Around 1940, Simon Kuznets estimated American consumption and
income for the decades going back to 1900. His finding that the
ratio of consumption to income did not fall with long-run higher income levels was thought by some to be a deathblow to Keynesianism. However, writers like Hansen had already been insisting on the
shift of living standards and consumption propensities through
time, having been led to this position by common-sense observation
and by comparative budget studies-as for example the 1919 and
1935 evidence on Washington civil servants' saving habits. As
Keynes had earlier insisted in controversy with H. Staehle and E. W.
8

I am omitting in this brief survey an important post-I936 development: the analysis of liquidity and general asset preference in tenus
of attitudes toward uncertainty and probability distributions, associated with the writings of J. Marschak, J. R. Hicks, J. Robinson, H.
Markowitz, J. Tobin, F. Modigliani, Musgrave-Domar, A. C. Hart,
L. J. Savage, M. Richter, H. Houthakker, P. Cootner, and others.

336 . Paul A. Samuelson


Gilboy, a straight-line propensity-to-consume schedule was grist for
his mill, the only important requirement being that but a fraction of
extra income go into consumption .spending.
Being a skeptic by nature, I was not willing to go so far as Hansen
in postulating a long-run consumption function showing about the
same percentage of saving at full employment levels. In Seymour
Harris' Postwar Economic Problems (1943), I preferred to state: the
consumption function shifts upward through time as tastes and habits
change and as new consumption items are developed to tempt us; at
the same time our full employment potential shifts upward as
technical invention and capital formation improve productivity; the
fact that these counter trends have about canceled each other out for
half a century can be regarded as in the nature of a coincidence, and
provides no guarantee of repetition in the future. Despite twenty
more years of experience, I incline to pretty much the same opinion,
having learned how treacherous are economic "laws" in economic
life: e.g., Bowley's Law of constant relative wage share; Long's Law
of constant population participation in the labor force; Pareto's Law
of unchangeable inequality of incomes; Denison's Law of constant
private saving ratio; Colin Clark's Law of a 25 per cent ceiling on
government expenditure and taxation; Modigliani's Law of constant
wealth-income ratio; Marx's Law of the falling rate of real wage and/
or the falling rate of profit; Everybody's Law of a constant capitaloutput ratio. If these be laws, Mother Nature is a criminal by nature.
Experience has also taught me not to be necessarily suspicious of
coincidences; in many cases, even if they do not explain the facts,
they do describe the facts, up until they cease to describe the facts.
In the Harris book I pictured the hysteresis loops characteristic of
cyclical moves in the production function. In principle, a quick
transitory increase in incomes during an upswing might get spent
more fully on durable and other goods than might a slow and steady
rise to a new level; in principle, the opposite could happen. Tu