The New Monetarism

by Nicholas Kaldor

T HE Keynesian Revolution of the late 1930s has completely

displaced earlier ways of thinking and provided an entirely
new conceptual framework for economic management. As a
result, we think of day-to-day problems—of inflationary or
deflationary tendencies, unemployment, the balance of payments or
growth—on different lines from those of economists of earlier
generations. We think of the pressure of demand as determined by
autonomous and induced expenditures, and we seek to regulate the
economy by interfering at various points with the process of income
generation: by offsetting net inflationary or deflationary trends
emanating from the private sector or the overseas sector by opposite
changes in the net income generating effect of the public sector.
Previously, economists had thought of the level of demand—the
volume of spending—as being directly determined by the supply of
money and the velocity of circulation; and thought of regulating the
level of expenditure mainly by monetary controls.
The Bank is not necessarily in agreement with the For the last twenty or thirty years we have felt we have much
views expressed in articles appearing in this Review. better insight into the workings of the market mechanism than our
They are published in order to stimulate free discussion predecessors, and felt much superior to them. However, we now
and full enquiry. have a "monetary" counter-revolution whose message is that
during this time we have been wrong and our forbears largely, if not
perhaps entirely, right; anyhow, on the right track, whereas we
have been shunted on to the wrong track. This new doctrine is
assiduously propagated from across the Atlantic by a growing band
of enthusiasts, combining the fervour of early Christians with the
suavity and selling power of a Madison Avenue executive. And it
is very largely the product of one economist with exceptional
powers of persuasion and propagation: Professor Milton Friedman
of Chicago. The "new monetarism" is a "Friedman Revolution"
more truly than Keynes was the sole fount of the "Keynesian
Revolution". Keynes's General Theory was the culmination of a
great deal of earlier work by large numbers of people: chiefly
Wicksell and his followers, Myrdal and Lindahl in Sweden, Kalecki
in Poland, not to speak of Keynes's colleagues in Cambridge and of
many others.
The author is Professor of Economics in the University of Cambridge and
Fellow of King's College. This article is the text of a public lecture given at University
College, London, on Thursday, March 12.
 The new school, the Friedmanites (1 do not use this term in any (2) Money cannot change "real" things, except temporarily,
pejorative sense, the more respectful expression "Friedmanians"
sounds worse) can record very considerable success, both in terms
of the numbers of distinguished converts and of some rather
L and in the manner of throwing a spanner into the works—a
" monkey-wrench into the machine ", to use Friedman's more
homely expression1—at the cost of painful adjustments afterwards.
glittering evidence in terms of "scientific proofs", obtained through There is a unique real equilibrium rate of real interest, a unique
empirical investigations summarized in time-series regression real equilibrium real wage, an equilibrium level of real unemploy-
equations. Indeed, the characteristic feature of the new school is ment. By monkeying around with money, these things can tem-
"positivism" and "scientism"; some would say "pseudo-scientism", porarily be made to change—interest reduced, unemployment
using science as a selling appeal. They certainly use time-series cut, the real wage cut (or raised, I am not sure which)—only by
regressions as if they provided the same kind of "proofs" as making, in each case, reverse changes (abnormally high interest
controlled experiments in the natural sciences. And one hears of rates, abnormal unemployment, etc.) the inevitable sequel.
new stories of conversions almost every day, one old bastion of old- All this part of the Friedman doctrine is closely reminiscent
fashioned Keynesian orthodoxy being captured after another: first, of the Austrian school of the 'twenties and the early 'thirties—the
the Federal Reserve Bank of St. Louis, then another Federal theories of von Mises and von Hayek—a fact which so far (to my
Reserve Bank, then the research staff of the IMF, or at least the knowledge) has received no acknowledgment in Friedmanite
majority of them, are "secret", if not open, Friedmanites. Even the literature. (Very few people these days know the works of the
"Fed" in Washington is said to be tottering, not to speak of the Mises-Hayek school; unfortunately, I am old enough to have been
spread of the new doctrines in many universities in the United an early follower of Professor Hayek, and even translated one of
States. In this country, also, there are some distinguished and lively his books, and there is nothing like having to translate a book,
protagonists, like Professor Harry Johnson and Professor Walters, particularly from the German language, to force you to come to
though, in comparison to America, they write in muted tones and grips with an argument.) Friedman differs from Mises and Hayek
make more modest claims; which makes it more difficult to discover in being more liberally spiced with 'the new empiricism. On the
just what it is they believe in, just where the new doctrine ceases to other hand, he misses some of the subtleties of the Hayekian
be a matter of semantics and becomes a revelation with operational transmission mechanism, and of the money-induced distortions in
the "structure of production."
significance. (3) While the money supply alone determines money expendi-
ELEMENTS OF NEW DOCTRINE tures, incomes and prices, it does so with a time lag which is,
What are the essential propositions of the new doctrine ? For unhappily, not a stable one. It can vary, for reasons yet unknown,
this, it is no good turning to the "moderates", who do not really say between two quarters and eight quarters. This is what the regression
equations show.
anything, or to the "extremists"—like Messrs. Anderson, Jordan
and Keran of St. Louis—who both vulgarize and discredit the new (4) Hence, while control of the money supply is the only
creed by the blatant simplicity of their beliefs and the extravagance powerful instrument of control, it is hopeless for central banks to
of their claims. One must turn to the archpriest, Friedman himself, pursue a positive stabilization policy by varying the money supply
and such of his close disciples, like Meiselman, Anna Schwartz and in a contra-cyclical manner. Indeed, their attempts to do so may
Philip Cagan, who can be relied on to follow the master closely and have been the very cause of the cyclical instabilities in the economy
which they aimed to prevent. Hence, the best thing for stability is
interpret him correctly. to maintain a steady expansion of the money supply of 4-5 per
The essential elements of the creed can, I think, be summarized
cent, (in the latest version, the ideal has come down to 2 per cent.)
in the following four propositions: and, sooner or later, everything will fall into line. There will be
(1) Money alone matters in determining "money things", such
steady growth without inflation.
as the money GNP, the level and the rate of changes of money
All this is argued not, like the Keynesian theory, in terms of a
prices, and the level and the rate of change of money wages. Per
contra, other things—such as fiscal policies, taxation, trade union 1
p. 12.
"The Role of Monetary Policy", American Economic Review, March, 1968,
behaviour, etc.—do not (or do not really) matter.
 sensitive to interest rates. But who are "the people" in this
structural model which specifies the manner of operation of various
connection? Are they the wage- and salary-earners, who, between
factors. The results are based on direct and conclusive historical
them, account for 70 per cent, of the national income, but hold, at
evidence; on statistical associations which appear—to the authors
any one time, a much lesser proportion, perhaps 10 to 20 per cent.,
—so strong and clear as to rule out other interpretations. The
of the total money supply? Or are they the "rentiers", whose
actual mechanism by which exogenous changes in the supply of
"portfolio selection" and "portfolio shifts" are much influenced at
money influence the level of spending—how the money gets into
any time by short-term expectations, as well as by the relative
circulation, who it is received by, whether the recipients treat it as
yields of various types of financial assets ? Or are they businesses,
an addition to their spendable income or to their wealth, or whether
for which holding money is just one of a number of ways of securing
it comes into existence in exchange for other assets without
liquidity—unexploited borrowing power, unused overdraft limits
augmenting either wealth or income—is hardly considered by the
and so on being other ways—and for which the state of liquidity is
orthodox Friedman school. It is significant perhaps that when
only one of a number of factors that influence current expenditure
Friedman in his latest essay does attempt a graphic description decisions ?
of how an increase in the money supply leads to a rise in prices
and incomes, the money is scattered to the population from the RELATION OF MONEY TO GNP
air by a helicopter.1 * * * Before we consider these contentions further, one might pause
to ask whether there is anything surprising in a "stable money"
The basis of all this is the "stable demand function for money", function.
derived from empirical observations over longer and shorter Clearly, in a broad sense the "money supply", however
periods; with varying definitions of money, and varying time lags defined, correlates with the money GNP—so does everything else:
between changes of money and income, where the choice of the consumption, investment, wealth, the wage-bill, etc. All these
time lag, and the choice of the definition of what is "money", are things move over time, normally upwards, and in any time series
both determined by the criterion of the best statistical "fit" (in the movement of any one item is bound to be highly correlated with
terms of R2 and "t" values) for the regression equation. It is the others. Thus Richard Stone demonstrated years ago that for
sometimes expressed in terms of a "money supply multiplier" the U.S. economy in the inter-war years all principal items of
which is clearly implied by the "stable demand function", though income and expenditure (eighteen of them) were closely correlated
the empirical values of the "multiplier" are not consistent with a with three independent factors, which he identified as the GNP,
unity elasticity in the demand for money (i.e., an equi-propor- the change in the GNP and a time trend.1
tional relationship between the change in money and that of money The important questions to ask are:
income) which the quantity theory postulates; sometimes in terms First, does a high correlation indicate a causal relationship
of a relationship between changes in the money supply and changes either way? Does it imply that the supply of money determines
in consumption expenditure, together with the demonstration that the level of income, or the other way round? Or are both determined
the money multiplier invariably "outperforms" the Keynesian simultaneously by a third factor (or factors) ?
multiplier. (This latter contention, for what it is worth, has been Second, does the existence of a strong statistical association
shown to be dependent on arbitrary and inappropriate definitions imply that by controlling one of the variables, say the money
of "autonomous" expenditures in a Keynesian model.)2 supply, one can induce a predictable variation in the other? In
Friedman interprets his empirical findings in a strict Walrasian other words, would the "money multiplier" survive if it were
(or Marshallian) manner, as an indication that "people" wish to subjected to serious pressure?
keep a constant proportion of their real income (or their permanent In the U.K., the best correlation is undoubtedly found, not
real income) in the form of money, a proportion which is not (very) between the so-called "money supply" and the GNP, or that and
'ntn ,|ff consumers' expenditure, but between the quarterly variation in the
1 R. Stone. "On the Interdependence of Blocks of Transactions", Journal of the
Royal Statistical Society, Supplement, vol. 8 (1947).
amount of cash (that is, notes and coins) in the hands of the public, the people who can be trusted not to spend in excess of what they
and corresponding variations in personal consumption at market can afford to spend—would thus live on credit cards. The rest of
prices.1 This, of course, was broadly known long before multiple the population—the mass of weekly wage-earners, for example,
regressions were invented (or computers to calculate them with who have no "credit", not being men of substance—would get
ease). Every schoolboy knows that cash in the hands of the paid in chits which would be issued in lieu of cash by, say, the top
public regularly shoots up at Christmas, goes down in January and five hundred businesses in the country (who would also, for a
shoots up again around the summer bank holiday. consideration, provide such chits to other employers). And these
Nobody would suggest (not even Professor Friedman, I five hundred firms would soon find it convenient to set up a clearing
believe) that the increase in note circulation in December is the system of their own, by investing in some giant computer which
cause of the Christmas buying spree. But there is the question that would at regular intervals net out all mutual claims and liabilities.
is more relevant to the Friedman thesis: Could the "authorities" It would also be necessary for the member firms of this clearing
prevent the buying spree by refusing to supply additional notes system to accord mutual "swops" or credit facilities to each other,
to take care of net credit or debit balances after each clearing.
and coins in the Christmas season?
Of course, most people would say that it would be quite When this is also agreed on, a complete surrogate money-system
impossible to prevent the rise in the note circulation without and payments-system would be established, which would exist
disastrous consequences: widespread bank failures, or a general side by side with "official money".
closure of the banks as a precautionary measure. If I were asked to
advise, I would say that it could be done by less dramatic means:
CHARACTERISTICS OF MONEY
by instructing the banks, for example, not to cash more than £5
at any one time for each customer; by keeping down the number of What, at any time, is regarded as "money" are those forms of
cashiers, so as to maintain reasonably long queues in front of each financial claims which are commonly used as means of clearing
bank window. If a man needed to queue up ten times a day, half debts. But any shortage of commonly-used types is bound to lead
an hour a time, to get £50 in notes, this would impose a pretty to the emergence of new types; indeed, this is how, historically,
first bank notes and then chequing accounts emerged. To the
effective constraint on the cash supply.
But would it stop Christmas buying? There would be chaos extent that no such new forms have emerged recently—in fact,
for a few days, but soon all kinds of money substitutes would they are emerging, though not as yet in a spectacular way—this
spring up: credit cards, promissory notes, etc., issued by firms or is only because the existing system is so managed as to make it
financial institutions which would circulate in the same way as unnecessary—with the "authorities" providing enough money of
bank notes. Any business with a high reputation—a well-known the accustomed kind to discourage the growth of new kinds. They
firm which is universally trusted—could issue such paper, and any thereby also condition our minds into thinking that money is some
one who could individually be "trusted" would get things on"credit". distinct substance, a real entity, whose "quantity" is managed and
People who can be "trusted" are, of course, the same as those who controlled quite independently by the monetary authority.
have "credit"—the original meaning of "credit" was simply Of course, within limits, the ultimate monetary authority can
"trust". There would be a rush to join the Diners Club, and every- and does exercise control over the volume of borrowing, because
one who could be "trusted" to be given a card would still be able it can control interest rates, particularly at the short end, through
open market operations, far more powerfully than other operators;
to buy as much as he desired. and because, within limits, it can control the volume and direction
The trust-worthy or credit-worthy part of the population—
of lending by the clearing banks, which have such a powerful role
in the system as suppliers of credit. But, as the Radcliffe Committee
1 Thus, for 83 observations in the period 1948-69, the R 2 is -884, the "cash
multiplier" 2-3, the "t" value 3 • 7, after allowing for seasonally. The "cash multiplier" has shown, when credit control is operated as an independent
is 6 • 1, the "t" value 9, the R 2 is -494, without correction for seasonally. Even better instrument—as a substitute for fiscal policy, and not as a comple-
sounding results can be attained by relating the change in expenditure to both current ment to it—any forceful initiative by the monetary authorities
and lagged changes in the cash supply, lagged for each of the four quarters, which
yield positive and negative multipliers in regular sequence—which only goes to show weakens their hold over the market by diverting business from the
what "t" values and R2s are worth. (For equations, see Appendix on page 18.)
 More fundamentally (and semi-consciously rather than in full
clearing banks to other financial institutions. The post-war experi- awareness) it may have sprung from the realization of the monetary
ments in monetary policy caused a lot of disorganization—"a authorities, be it the Federal Reserve or the Bank of England, that
diffused difficulty of borrowing", in the words of Radcliffe, with they are in the position of a constitutional monarch: with very
firms having to borrow money from unaccustomed sources, or wide reserve powers on paper, the maintenance and continuance of
else to delay paying bills so as to achieve a better synchronization which are greatly dependent on the degree of restraint and
between receipts and outlays—but with little discernible effect on moderation shown in their exercise. The Bank of England, by virtue
spending. When the central bank succeeds in controlling the of successive Acts of Parliament, has a monopoly of the note
quantity of "conventional money", lending and borrowing is issue, at least in England and Wales. But the real power conferred
diverted to other sources, and the "velocity of circulation", in by these Acts depended, and still depends, on maintaining the
terms of conventional money, is automatically speeded up. central role of the note issue in the general monetary and credit
system; and this, in turn, was not a matter of legal powers, but of
VELOCITY OF CIRCULATION the avoidance of policies which would have lead to the erosion of
Friedman's main contention is that the velocity of circulation, this role.
in terms of conventional money1, has been relatively stable. That * * *
may well be, but only because, in the historical periods observed, The explanation, in other words, for all the empirical findings
the supply of money was unstable. In other words, in one way or on the "stable money function" is that the "money supply" is
another, an increased demand for money evoked an increase in "endogenous", not "exogenous".
supply. The money supply "accommodated itself" to the needs of This, of course, is the crux of the issue, and it is vehemently
trade: rising in response to an expansion, and vice versa. In denied by the monetarist school. They base their case on two kinds
technical terms, this may have been the result of the objective of of evidence:
"financial stabilization", of maintaining the structure of interest (1) The first is the time lag. Peaks and troughs in the money
rates at some desired level, or the so-called "even keel policy", of supply (in the U.S., at any rate) have regularly preceded peaks and
ensuring an orderly market for government debt.2 troughs in GNP, though with a variable lag of two to six quarters,
and one that tended to shorten in the post-war era to one quarter
1 The precise meaning of "conventional money" differs from author to author
(and from country to country); in the U.K. context it is usually defined as cash plus circumstances would directly increase spending forget that, barring helicopters, etc.,
clearing bank deposits (both current and deposit accounts) in the hands of the public. the "excess supply" could never materialize.
2 A great deal of the current discussion on the importance of "money" is devoted One of the main contentions of the Friedman school is that, whenever the central
to the issue of the "interest elasticity" of money balances—i.e. to the question of how bank changes the money supply by open market operations, say, by selling bonds in
the ratio between the "money supply" (as conventionally denned) and the national exchange for cash, it does not follow that the individuals who buy the bonds which
income can be expected to vary with changes in interest rates. Evidence of a low- the central bank sells will reduce their holding of money correspondingly—they may
interest elasticity is supposed to support the "monetarist school", while a high- continue to hold the same amount of money, and economize instead on the buying of
interest elasticity is supposed to lend support to the "Keynesian" view. In fact, it "goods". In this way, it is contended, a reduction in the money supply will have a
does neither the one nor the other. The interest-elasticity of the demand for money "direct effect" on the demand for goods, and not only an "indirect effect", via the
really concerns a different issue: the power of the monetary authorities to vary the rate of interest. But there is a confusion here between "stocks" and "flows". The
money supply in an exogenous manner. The less prepared the public is to absorb amount of money held by an individual is part of his stock of wealth; if he buys
more cash in response to a reduction in interest rates, or to release cash in response additional bonds, and this purchase represents an addition to his total stock of
to a rise, the less is it possible for the monetary authorities to expand the "money wealth, and not merely a substitution between one form of holding wealth and another
supply" relative to demand, or to prevent it from rising in response to a rise in the (i.e. he continues to hold the same amount of money, plus a larger amount of bonds)
public's demand. This is because the authorities' sole policy instrument for changing this is only another way of saying that the individual bought the additional bonds out
the "money supply" is the buying and selling of financial assets in exchange for of income (i.e. out of forgone consumption), which in plain language means that he
money; this presupposes that such sales or purchases can be effected in reasonable was induced to save more as a result of the opportunity of buying bonds on more
amounts without creating violent instabilities in the financial markets. Hence, the attractive terms. No one has ever denied that monetary policy operating through
more Friedman and his followers succeed in demonstrating the insensitiveness of the changes in interest rates (or through direct controls over the volume of bank lending)
demand for money to interest rates, the more they denigrate the role of money as an could have an effect on the propensity to save as well as on the inducement to invest.
autonomous influence on the economy. The "stable money function" is evidence, But, unless the monetarists assume a high-interest elasticity in the propensity to save,
not of the "importance of money", but only of the impotence of the authorities in and attribute the major influence of monetary action to this factor (in which case this
controlling it. If it required a 50 per cent, fall in Consols to effect a 5 per cent, should be made explicit), they cannot be saying anything different from Keynes—i.e.
reduction (or to prevent a 5 per cent, rise) in the amount of money held by the public that the effects of "monetary action" on the level of demand depend on the effects of
(i.e., assuming an interest elasticity of 0-1), any autonomous regulation of the "money the consequential changes of interest rates (or, what comes to the same thing, of
supply" would in practice be rendered impossible by the exigencies of the financial credit rationing by the banks) on the level of investment.
and banking system. Those who hold that an "excess supply" of money under these
 11
10
feature of the scene, especially in the post-war years. I am referring
or less.1 If the money supply changes first, and the level of income to the so-called "built-in fiscal stabilizer", which means that the
(or business activity) afterwards, it is contended that the one that fiscal deficit automatically rises in times of declining activity and
came first must have been the cause of the other. automatically falls in times of rising activity. Owing to lags in tax
(2) The second is the contention that in the U.S., at any rate,
collection, particularly in taxes on corporate profits, this operates
banks are always "loaned up", more or less. Hence, the "money so that the maximum swings occur sometimes after the turning
supply" (which includes bank deposits, as well as notes and coins point in economic activity.
held by the public) is fairly closely related to "high-powered Now, it is well known that changes in the government's net
money"—to the so-called "monetary base", which is under the borrowing requirement are the most important cause of changes in
sole control of the Federal Reserve, and who exercise their power, the money supply. This is only partly due to the fact that the
wisely or foolishly, but quite autonomously. government's own balances are excluded from the "money supply",
In my opinion, neither of these arguments proves that money so that any depletion of such balances automatically augments the
plays—in the U.S., let alone in the U.K.—the causal role: that the money supply. Partly it is due to the fact that the government is
"money supply" governs the level and the rate of growth of money the one borrower with unlimited borrowing power: an increase in
incomes or expenditures. government borrowing, whether due to a decline in tax receipts, a
rise in expenditure, or both, involves an increase in the money
THE TIME LAG supply as an automatic result of a "passive" monetary policy,
With regard to the time lag, it is now fairly generally admitted which supplies reserves as part of a policy of stabilizing interest rates
that it does not prove anything about the nature of the causal or simply to ensure orderly conditions in the bond market.l More-
relationships. If one assumed a purely Keynesian model where over, since, in the U.S. at any rate, the government's borrowing
expenditure decisions govern incomes, and if one assumed a purely requirement is largest when the economy is depressed, it occurs
passive monetary system—with reserves being supplied freely, at at a time when the Federal Reserve system is least inclined to
constant interest rates—it would still be true that the turn-round follow a "tough" credit policy; whilst in times when it wishes to
in the money supply would precede the turn-round in the GNP, restrain the expansion of credit, the government itself is likely to
for much the same reasons for which the Keynesian multiplier be in surplus. Hence, the large observed fluctuations in the money
invariably involves a time lag. supply, preceding in time the business cycle, may merely be a
Suppose the initiating change is a decision of some firms to reflection of the operation of the built-in fiscal stabilizer.
increase their inventories, financed by borrowing. The first impact An interesting bit of evidence for this view is the abnormal
is to cause some other firms whose sales have increased un- behaviour of the money supply following the Korean War, when
expectedly to incur some involuntary disinvestment. It is only the money supply peaked about a year after, not in the year pre-
when that is made good by increased orders that productive ceding, the peak of the post-Korean boom. A possible explanation
activity is expanded; any such expansion will cause higher wage is that the rise in government expenditure (and the deficit) followed
outlays, which in turn may involve further borrowing. The ultimate on this occasion the sharp rise in activity, which was induced, no
effects on income involve further increases in productive activity doubt, by the large rise in military procurement but which had been
arising from the expenditure generated by additional incomes. reflected in a sharp increase in federal expenditure only some time
There is every reason for supposing, therefore, that the rise in the later on, when the bills came to be paid.
"money supply" should precede the rise in income—irrespective of
whether the money-increase was a cause or an effect. 1 As Hawtrey has repeatedly emphasized, in the case of private borrowing the
There may be other explanations which would need to be maintenance of orderly conditions in the bond market invariably involved some
investigated, such as the contra-cyclical behaviour in the fiscal policy of "credit-rationing" or rather "issue rationing" by the issuing houses, who
made sure that the volume of issues for public subscription at any one time was no
balance which, particularly in the U.S., has been a very important greater than what the market could absorb. This is his explanation for the long-term
rate of interest being largely a "conventional phenomenon". (Cf. e.g. A Century of
Bank Rate, London 1938, pp. 177ff.) But there is nothing equivalent to this in the
1 Richard G. Davis, "The Role of the Money Supply in Business Cycles", case of government borrowing.
Monthly Review of the Federal Reserve Bank of New York, April, 1968, p. 71.
12 13

CHANGES IN MONEY SUPPLY: U.S. EXPERIENCE studies have demonstrated that the facts are precisely the reverse: the U.S.
monetary authorities followed highly deflationary policies. The quantity of money
This brings me to Friedman's second contention and the one in the United States fell by one-third in the course of the contraction. And it
fell not because there were no willing borrowers—not because the horse would
on which he would himself lay the most emphasis: that in the not drink. It fell because the Federal Reserve System forced or permitted a sharp
United States, at any rate, changes in the money supply have been reduction in the monetary base, because it failed to exercise the responsibilities
assigned to it in the Federal Reserve Act to provide liquidity to the banking
"exogenous" and were largely determined by autonomous policy system. The Great Contraction is tragic testimony to the power of monetary
decisions of the Federal Reserve Board. Since Friedman and Anna policy-—not, as Keynes and so many of his contemporaries believed, evidence
of its impotence.
Schwartz have written a book of eight hundred pages to prove this
point,1 it is not easy to deal with their massive evidence in a few I cannot understand the reference to the "sharp reduction in
sentences at the tail end of a lecture. Nonetheless, I shall try, but the monetary base" in the above passage, which is absolutely
will confine myself to some key issues and to some general critical to the argument. According to Friedman's own figures,1 the
observations. amount of "high-powered money", which is Friedman's own synonym
In the first place, while the correlation between the "monetary forthe "monetary base" (i.e. currency held by the public plus member
base" (defined above) and the "money supply" was good in general, bank reserves with the Federal Reserve) in the U.S. increased,
it was not all that good to be able to regard changes in the one as not decreased, throughout the Great Contraction: in July, 1932, it
being the equivalent of changes in the other. In particular, it was more than 10 per cent, higher than in July, 1929, whereas it was
appears that on occasions when the Federal Reserve went out of its held constant in the three previous years (1926-29). The Great
way to increase reserves (as in the 1929-39 period), the reaction Contraction of the money supply (by one-third) occurred despite
on the total money supply was small. Moreover, the effects of this rise in the monetary base. This was partly because the ratio of
changes in the "monetary base" on the "money multiplier" were currency held by the public to bank deposits rose substantially.
consistently negative in all periods.2 This is attributed by Friedman to a confidence crisis: the public's
More important than this, the variations in the "monetary diminished confidence in the banks. But it is important to observe
base" are themselves explained by factors—such as the desire to that this dramatic rise in the ratio of currency held by the public
stabilize interest rates, or to ensure government debt financing (the to bank deposits was never reversed subsequently. In July, 1960, it
so-called "even keel" objective3)—which makes the "monetary was still at approximately the same level as in July, 1932, which in
base" automatically responsive to changes in the demand for turn was nearly twice as high as in July, 1929. If it was a matter of
money. In other words, if variations in the money supply were confidence in the banks, why was it not reversed in the subsequent
closely related to changes in the "monetary base", this is mainly thirty years? The fact that the currency-deposit ratio was at its
because the latter has also been "endogenous", as well as the highest during the war years, 1944-45 (when it stood 45 per cent.
former. above the July, 1932, level) suggests rather that the main explanation
Friedman himself regards the monetary history of the Great may lie elsewhere—in the change in the pattern of expenditure
Contraction, 1929-33, as the ultimate test of his basic contention. between goods (or assets) normally paid for in cash, and those
It is worth quoting the critical passage in his Presidential Address normally paid for by cheque; which was due partly to the fall in
to the American Economic Association4 at some length:— the volume of financial transactions in relation to income trans-
The revival of belief in the potency of monetary policy was fostered also by a actions (this would explain why the deposit-currency ratio rose so
re-evaluation of the role money played from 1929 to 1933. Keynes and most much during the years of the Wall Street boom2); and partly also
other economists of the time believed that the Great Contraction in the United
States occurred despite aggressive expansionary policies by the monetary to the rise in the share of wages, and the fall in the share of property
authorities—that they did their best, but their best was not good enough. Recent incomes, during the slump.
1 A Monetary History of the United States, 1867-1960, National Bureau of 1 Friedman and Schwartz, op. cit. Table B-3, pp. 803-804.
Economic Research, Princeton University Press, 1963. * The demand for money is usually considered as a function of income and
2 Cf. Keran, "Monetary and Fiscal Influences on Economic Activity—The wealth; this is legitimate on the assumption that the volume of money transactions is
Historical Evidence", Review of the Federal Reserve Bank of St. Louis, November, itself uniquely related to income and wealth. However, in times when people make
frequent "switches" in their portfolios, and the volume of financial transactions is
1969, Tables VII and VIII. large relatively to the total value of assets, it is inevitable that the amount of money
3 Keran, op. cit. Table VI. held by speculators as a group should also relatively be large, even if no one individual
4 "The Role of Monetary Policy", American Economic Review, March, 1968, p.3.
intends to hold such balances for more than a short period.
(My italics.)
14 15

The other reason was the fall in the ratio of bank deposits to provided by Friedman himself, in comparing U.S. and Canadian
bank reserves—in other words, a rise in commercial bank liquidity experience during the Great Contraction.1 In Canada, there were
by some 27 per cent, between July, 1929, and July, 1932—which no bank failures at all; the contraction in the money supply was
may have reflected prudential motives by the banks, but may also much smaller than in the U.S.—only two-fifths of that in the U.S.,
have been the consequence of an insufficient demand for loans—- or 13 against 33 per cent.—yet the proportionate contraction in
of the horse refusing to drink (particularly the fall in the demand money GNP was nearly the same. The difference in the propor-
for loans for speculative purposes). There is nothing in these tional change in the money supply was largely offset by differences
figures, in my view, to support the far-reaching contentions which in the decline in the velocity of circulation: in the U.S. it fell by
I have just quoted; and, in a complex issue of this kind, I would put 29 per cent., in Canada by 41 per cent. This clearly suggests that
far more trust in the "feel" and judgement of contemporary the relative stability in the demand for money is a reflection of the
observers, like Keynes or Henry Simons, than in some dubious instability in its supply; if the supply of money had been kept more
(and tendentious) statistics produced thirty years later. stable, the velocity of circulation would have been more wwstable.
I have also perused the one hundred and twenty pages devoted This last statement may appear to be in contradiction to
to the Great Contraction in the book on the monetary history of Friedman's empirical generalization according to which the move-
the U.S.; and, while I would agree that he makes out a good case ment in the velocity of circulation in the U.S. has historically been
for saying that the policy of the Federal Reserve, particularly after positively correlated with movements in the money supply—the
Britain's departure from the gold standard, was foolish and velocity of circulation was at its most stable when the money supply
unimaginative, and that the succession of bank failures in the was most stable. But the two propositions are not inconsistent,
course of 1932 might have been avoided if the Federal Reserve had which shows how easy it is to draw misleading conclusions from
followed more closely the classic prescription for a financial panic statistical associations. If one postulates that it is the fluctuation in
of Mr. Harman of the Bank of England in 1825 (quoted by the economy that causes the fluctuations in the money supply (and
Bagehot)1—of lending like mad on the security of every scrap of not the other way round), but that the elasticity in the supply of
respectable looking paper—I do not believe that it would have money (in response to changes in demand) is less than infinite,
made all that difference. In particular, I do not believe that the then, the greater the change in demand, the more both the supply
Great Depression (with all its tragic consequences, Hitler and the of money and the "velocity" will rise in consequence. If the supply
second world war) would not have occurred but for Governor of money had responded less, the change in velocity would have
Benjamin Strong's untimely retirement and death in 1928. Indeed, been greater; if the supply of money had responded fully, no.
I am not sure whether Governor Strong's policies in the years prior change in velocity would have occurred (under this hypothesis).
to 1928 might not have contributed to the financial crisis following
the crash in 1929. For he kept the volume of reserves—the supply WHAT ABOUT BRITAIN?
of "high-powered money"—rigidly stable in the years 1925-29. In this country, at least since the second world war, it is even
This occurred at a time when the U.S. economy and the national less plausible to argue that the "money supply" is under the direct
income was expanding, with the result that the banking system control of the monetary authorities, regulated through the rate of
became increasingly precarious: the ratio of bank deposits to bank creation of bank reserves. Clearly, it is not controlled through the
reserves, and the ratio of deposits to currency in the hands of the 8 per cent, minimum cash ratio, for there is an agreement between
public, rose well above the customary levels established prior to the the Bank and the clearing banks to supply sufficient reserves to
first world war, and to very much higher levels than these ratios validate this ratio week by week without any window-dressing.
have ever attained subsequently.2 Nor can it be said that the "money supply" is controlled by the
* * * agreement of the clearing banks to observe the 28 per cent.
prudential liquidity ratio, since there are numerous ways open to
Indeed, the best answer to Friedman's main contention is the banks to maintain this latter ratio which do not involve recourse
1 Bagehot, Lombard Street,London, 1873, pp. 51-52; quoted in Friedman and to central bank credit.
Schwartz, op. cit. p. 395. 1 Ibid, p. 352.
2 Friedman and Schwartz, op. cit. Table B-3, pp. 800-808.
 17
16
What, then, governs, at least in the U.K., the changes in relationship of the business sector which, in turn, was a reflection
"money supply" ? In my view, it is largely a reflection of the rate of of the big improvement in the receipt-expenditure relationship of
change in money incomes and, therefore, is dependent on, and the public sector, only partially offset by the (more recent) improve-
varies with, all the forces, or factors, which determine this ment in the receipt-outlay relationship of the overseas sector.
magnitude: the change in the pressure of demand, domestic invest-
ment, exports and fiscal policy, on the one hand, and the rate of
wage-inflation (which may also be partly influenced by the pressure What, if anything, follows from all this ? I have certainly no
of demand), on the other hand. This basic relationship between the objection to Friedman's prescription that the best thing to do is to
money supply and GNP is modified, however, in the short period secure a steady expansion of x per cent, a year in the money supply.
by the behaviour of the income-expenditure relation (or, as I But I doubt if this objective is attainable by the instruments of
would prefer to call it, the receipt-outlay relation) of those particular monetary policy in the U.S., let alone in the U.K. If it is ever
sectors whose receipt-outlay relation is particularly unstable—in attained, it will be because, contrary to past experience, we shall
other words, whose net dependence on "outside finance" is both succeed in avoiding stop-go cycles emanating from abroad, or from
large and liable to large variations, for reasons which are the private business sector, or, what is more likely, from the very
endogenous, not exogenous, to the sector. This is true, of course, to changes in fiscal policy which aim to compensate for other instabili-
a certain extent of the business sector, though business investment ties; and if, by some combination of incomes policy and magic
in fixed capital and stocks has not been nearly as unstable in the (but more by magic), we shall also succeed in keeping the rate of
last twenty years as it was expected to be in pre-war days. But it is increase in money wages in both a stable and a reasonable relation-
chiefly true of the public sector, whose "net borrowing require- ship to the rate of growth of productivity.
ment" has been subject to very large fluctuations year by year. I
am convinced that the short-run variations in the "money supply" March, 1970 Nicholas Kaldor
—in other words, the variation relative to trend—are very largely
explained by the variation in the public sector's borrowing
requirement.1
Over the last five years we have witnessed a dramatic change
in the rate of increase in the money supply: it fell from 9 • 8 per cent,
in 1967 to 6f per cent, in 1968 and to only 2-9 per cent, in 1969.
The last of these years has also witnessed a dramatic turn-round
in the balance of payments. This is regarded as a "feather in the
cap" for the monetarists, who point with pride to the effectiveness
of monetary policy—not in stopping wage and price inflation, for
this unfortunately has not happened—but at least in restoring a
healthy balance of payments. They forget that the same period
witnessed an even more dramatic turn-round in the net borrowing
requirement of the public sector—from over £2,000 millions in
1967-68 to minus £600 millions in 1969-70. The recent "credit
squeeze" is not really a "credit squeeze" but a "liquidity squeeze".
It is a direct consequence of a big fall in the receipt-expenditure
1 In fact, a simple regression equation of the annual change of the money supply
on the public sector borrowing requirement for the years 1954^1968 shows that the
money supply increased almost exactly £ for £ with every £1 increase in the public
sector deficit, with t = 6-1, R 2 = -740, or, in fashionable language, 74 per cent, of the
variation in the money supply is explained by the deficit of the public sector alone.
(See Appendix.)
18
TECHNICAL APPENDIX
Regression Equations relating Changes in Consumers' Expenditure in the
U.K. to Changes in Currency in Circulation held by the Public.
Data: Quarterly changes in £ millions; 1948 II—III to 1969 H—III
Notation: A C = Change in Consumers' Expenditure
The Multinational Enterprise
A N = Change in average currency in circulation with the By John H. Dunning
public;
r ~ ~]
i ii
J ii—ml NDIVIDUALS, firms and businesses have long traded with
Dummy variables = 1 for quarter to quarter changes ] III—IV f, 0 otherwise
UV- ij
Standard deviation in brackets; R 2 unadjusted; s = standard error
(adjusted for degrees of freedom).
Lags in quarters denoted by negative subscripts.
I each other across national boundaries; to this extent, the
internationally-oriented enterprise is no new phenomenon.
Similarly, the economic prosperity of nations has always been
influenced by the terms on which they have exchanged goods and
Results: services. Since the early 19th century, an active international
R2 0-494
AC = -65-99 6-127 AN
s 183-8 capital market has existed, while the international flow of know-
(24-23) (0-681)
166-77d a 3-71d 3 476 -48 d4 ledge has an even longer pedigree, dating dack to the exodus of
AC = 170-35 (40-57) (49-83) the Huguenots in the 17th century and the smuggling of drawings,
(38-78) (29-74)
R2 0-884 designs and machinery out of Britain to the American colonies
2-350 AN s 89-5
(0-636) more than one hundred years later. But, until fairly recently,
AC = 5-77 + 3-855 AN - 2-565 AN., + 2'725AN- 2 most international transactions had two things in common. First,
(23-52) (0-639) (0-411) (0-417)
5-640AN.3 + 4-220AN.4
R2 0-878 each was generally undertaken independently of the other and by
s 94-62
(0-417) (0-644) different economic agents. Second, most transactions were between
AC = 96-84 - 31 -51dz + 37-54d3 - 310-88d4 + 2 - 6 2 4 AN - unassociated buyers and sellers, and were concluded at market or
(49-17) (63-52) (69-18) (68-15) (0-672)
2-062 AN.! + 1 -528 AN_ 2 - 2-205i AN_3 + R2 0-920 "arm's length" prices.
(0-706) (0-699) (0-702) (0- s 78-33 During the last half century, and particularly in the last
twenty years, a new and separately identifiable vehicle of in-
II Regression Equations showing the Relationship of Changes in the Money ternational economic activity has emerged: production by the
Supply in the U.K. to the Public Sector Borrowing Requirement. rapid expansion of foreign direct investment. The distinctive
Data: Annual figures in £ millions, relating to calendar years. features of foreign direct investment are two-fold. First, it embraces,
Notation: AM = increase in money supply.
P = Net acquisition of financial assets by the public sector. usually under the control of a single institution, the international
Standard deviation in brackets. transfer of separate, but complementary, "factor inputs"—
notably equity capital, knowledge and entrepreneurship—and
Results: sometimes of goods as well. Nowadays, direct investment accounts
(1) Period 1954-68 R2 0-740
A M = —299-1 — 1-035P s 210-2 for 75 per cent, of the private capital outflows of the leading in-
(0-170)
(2) Period 1960-68
dustrial nations, compared with less than 10 per cent, in 1914.
R2 0-714
AM = —246-3 — 0-979P s 212-1 Payments for proprietary knowledge, e.g. royalties, technical
(0-231)
service fees etc. between related institutions accounted for over
half of all such payments made across national boundaries by
British enterprises in 1968 and, in the same year, about a quarter
of their manufactured exports were sent directly to their foreign
subsidiaries.
The second unique quality of direct investment is that the
The author is Professor of Economics in the University of Reading. This article is
an abridged version of a paper presented by Professor Dunning at a conference on
the multinational enterprise held at the University of Reading, May 28-30, 1970. The
full proceedings of the conference will be published by Allen & Unwin in due course.