Full text of The Balance of Payments: Free Versus Fixed Exchange Rates

View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

TH E BALANCE OF PAYM ENTS:
FREE VERSUS FIXED EXCH A N G E RATES
Milton Friedman
and
Robert V. Roosa
Published by
American Enterprise Institute
for Public Policy Research

Troubled conversations among monetary authorities
about the United States’ balance-of-payments problems
have given proposals for free exchange rates scant, if
any, attention. Yet many campus economists, removed
from day-to-day problems in banking or government,
contend that the United States’ balance-of-payments
deficit would disappear if Washington were to allow its
own exchange rate among currencies of the world to
seek its level— or to “ float.”
The American Enterprise Institute for Public Policy
Research sought to provide a full discussion of inter
national exchange rates in a debate between Professor
Milton Friedman of the University of Chicago and
former Under Secretary of Treasury for Monetary A f
fairs, Robert V. Roosa, now a partner in Brown Brothers
Harriman and Company in New York. Their papers,
their rebuttals, and the questioning of both by a semi
nar of experts is contained in this new book.
Professor Friedman, president of the American Eco
nomic Association, is a brilliant theorist who character
izes himself as an empirical scientist. His opponent, the
president of the American Finance Association, matured
in the Federal Reserve System as an economist and in the
Treasury; more recently at Brown Brothers Harriman,
he has worked with practical day-to-day problems of
world money exchange.
Results of this debate have been both provocative
and informative. John Davenport, on the Board of Edi
tors of Fortune magazine, called the program "the most
important money debate I have ever heard.”
Dr. Friedman, who delivered the first lecture in early
May, 1967, saw the present fixed United States exchange

rate as another government control, similar to rent con
trol, farm pricing, or the minimum wage. He said that
floating exchange rates would lead to freer world trade,
would promote a dismantling of exchange controls and
import quotas, and would encourage a reduction of
tariffs.
Dr. Roosa, on the other hand, predicted chaos in in
ternational commerce if traders could not have a rate
of exchange stable enough to depend upon through the
term required for transactions in international trade.
Professor Friedman, in his rebuttal, capsuled the key
question:
How can two knowledgeable men reach such diametri
cally opposed conclusions on this, as on other aspects, of
free versus fixed exchange rates?

The excellent papers by both men, their rebuttals,
and the questioning by the seminar participants, may or
may not have answered Professor Friedman’s rhetorical
question. But they did demonstrate the need for debate
on this important public issue. The Roosa-Friedman
encounter was the last of four debates sponsored by the
American Enterprise Institute during the 1966-67
school year at The George Washington University.
Professor Friedman, one of the leaders in the informal
"Chicago school” of economists who stress the so-called
“ free market,” has been a student of floating exchange
rates for the past 20 years. His paper is regarded as the
authentic case for floating exchange rates. Dr. Roosa,
in turn, presented the classic case against free rates and
supporting fixed rates. But Dr. Roosa did not simply
defend the status quo. He argued for changes, including
an international unit of currency. Many experts believe
that under a fixed-rate system more reserves than those
presently provided will be needed to handle the growing
volume of international trade. Dr. Roosa, in 1965, wrote
a book, Monetary Reform for the World Economy, pro
posing a new unit of international currency to help ex
pand international liquidity.
Professor Friedman found the United States’ balanceof-payments deficits directly attributable to fixed ex
change rates, which he called government price fixing.

The problem of the balance of payments is simply another
example of the far-reaching effects of government price
fixing, complicated only by two facts: First, that two
sets of prices are involved— the price of gold in terms of
various national currencies, and the price of national
currencies in terms of one another; second, more than one
country is involved.

He was unimpressed by proposals to secure an inter
national agreement to create what he called "paper
gold,” or new international reserves, as Dr. Roosa sug
gested.
"There is one and only one satisfactory solution:
Abolish governmental price fixing,” Professor Friedman
said. "Let exchange rates become free market prices de
termined primarily by private dealing. Let the govern
ment simply stay out of the picture.”
Friedman argued that balance-of-payments problems
would be eliminated by floating exchange rates because
there could not be a surplus or a shortage in the sense
of eager sellers unable to find buyers or eager buyers un
able to find sellers.
"The price may fluctuate but there cannot be a
deficit or a surplus threatening an exchange crisis,” he
declared.
"Floating exchange rates would put an end to the
grave problems requiring repeated meetings of secre
taries of the Treasury and governors of central banks
to try to draw up sweeping reforms. It would put an
end to the occasional crisis of producing frantic scurry
ing of high governmental officials from capital to capital,
midnight phone calls among the great central banks
lining up emergency loans to support one another’s cur
rency.
"Indeed this is, I believe, one of the major sources
of the opposition to floating exchange rates. The people
engaged in these activities are important people and
they are all persuaded that they are engaged in impor
tant activities. It cannot be, they say to themselves, that
these important activities arise simply from pegging ex
change rates.”

This position distressed Dr. Roosa. Although acknowl
edging Professor Friedman as one of the world’s most
distinguished exponents of market economics, Dr. Roosa
countered that fixed rates of currency exchange pro
vided the most hospitable environment for encouraging
commercial and investment transactions between na-.
tions.
"The same downside rigidities and upward price drift
in our postwar economies that make adjustment more
difficult under fixed exchange rates would, in my view,
make for progressive inflation, and successive waves of
exchange-rate depreciation from one country to the
next, if countries were trying to follow a flexible rate
system,” Dr. Roosa said.
"The one telling influence for relative price stability
that is universally recognized, if not respected, in to
day’s world is that exerted by a country’s balance-ofpayments position. A flexible rate system permits a
country to cut itself off from the international force of
market competition—the greatest defender the world
now has for protecting the stability of domestic mone
tary values.”
Dr. Roosa believes that the economic traffic among
nations has become too vast and complex for anyone to
work out a satisfactory system of fluctuating rates for
day-to-day operations.
"Individual foreign exchange traders and bankers
would have an almost impossible task in groping for a
going rate that could take all these conflicting influences
into account,” he said.
“ Their task would be similar to, though larger than,
that of various individuals attempting to come up with
a firm figure for the wholesale or retail index of the
country, for example, and then being prepared to write
contracts on the basis of an unofficial pooling of each
other’s estimates.”
"I am very much afraid that the rate for any cur
rency against all others would have to fluctuate so widely
that the country’s own trade would be throttled and its
capital misdirected.”

TH E BALANCE OF PAYM ENTS:
FREE VERSUS FIXED
EXCH A N GE RATES
Fourth in the series of Rational Debate Seminars
sponsored by the American Enterprise Institute
held at
The George Washington University
'Washington, D. C.

THE
BALANCE OF PAYMENTS:
FREE VERSUS FIXED
EXCHANGE RATES
Milton Friedman
Robert V. Roosa
RATIO N AL DEBATE SEMINARS

American Enterprise Institute
for Public Policy Research
Washington, D. C.

© Copyright 1967 by the American Enterprise Institute for Public
Policy Research, 1200-17th Street, N.W ., Washington, D. C. 20036.
Ail rights reserved under International and Pan-American Copy
right Conventions.

Second Printing, June 1973
Library of Congress Catalog Card Number 672307

FOREW ORD
Rational debate, with the emphasis on "rational,” is the
keystone of a free society. This is the concept on which
the American Enterprise Institute was founded in 1943
and on which it continues to operate today. This book
records the fourth in a new series of Rational Debates
sponsored by the Institute to explore major public issues.
The format was devised to avoid what happens too often
in the course of debating vital public issues, a degenera
tion into repetitious absolutes which do not present
rational choices. The choice, of course, is seldom be
tween the wholly good or the wholly bad. Far from
being simple, most public issues evoke a wide spectrum
of arguments requiring careful consideration. We are
confident that Dr. Roosa and Professor Friedman have
illuminated the grays as well as the blacks and whites
in the issue of free versus fixed exchange rates.
The purpose of AEI from its inception has been to
help legislators, policymakers, educators, the press, and
the general public to reach informed judgments on
major issues of public policy. The Institute conducts
research, publishes studies, and sponsors seminars and
symposia on major questions of the day. Statements by
lecturers and other participants in AEI projects are
their own. The Institute itself takes no position on any
public policy issue.
In the 1966-67 academic year, the American Enter-

T h e B a l a n c e o f Pa y m e n t s

prise Institute presented four Rational Debates on major
public policy issues. The first, Congress and the Presi
dency: Their Role in Modern Times, featured Arthur
Schlesinger, Jr., and Alfred de Grazia. The second pitted
Charles E. Whittaker against "William Sloane Coffin, Jr.,
on Law, Order and Civil Disobedience. The third R a
tional Debate, on “ The New Economics,” brought
together Arthur F. Burns and Paul A. Samuelson.
This is the fourth and final debate of the 1966-67 series.
It is the hope of the American Enterprise Institute
that these seminars will contribute to wise policy deci
sions at all levels of the governments of the United
States, federal, state, and local.
July 10, 1967

William J. Baroody
President
American Enterprise Institute
for Public Policy Research

PREFACE
Dr. Roosa and Professor Friedman have given us, in
this fourth of the Rational Debate Series, an unusual
opportunity to compare two well-reasoned, brilliantly
argued views of the United States’ balance-of-payments
problems. In their frank exchanges there is less of the
gloss and more of the fundamentals than one finds in
most lengthier discourses on this difficult subject.
The Friedman-Roosa debate’s three sessions, spanning
two weeks last May, were attended by a small, select
group of government officials, academicians, and news
men. Now the public at large can study the speakers’
lectures and rebuttals, as well as their responses to ques
tions from the other participants.
This event brought to a close the first academic season
of Rational Debates sponsored by the American Enter
prise Institute, whose diligence in informing public
opinion on the pros and cons of major policy issues has
helped to reinvigorate the nation’s intellect in a time
of expanding need.
July 9, 1967

G. Warren Nutter
Coordinator
Rational Debate Series

C O N TEN TS
I

FIRST LECTU RE—Milton Friedman

SECOND LECTU RE—Robert V. Roosa

III REBUTTALS
Milton Friedman
Robert V. Roosa

71
81

IV DISCUSSION
First Session 111
Second Session 145
Third Session 169
FOOTNOTES

193

FIRST LECTURE

M ILTO N FRIEDM AN
Economists may not know much. But we do know one
thing very well: how to produce shortages and surpluses.
Do you want to produce a shortage of any product? Sim
ply have government fix and enforce a legal maximum
price on the product which is less than the price that
would otherwise prevail. That is how the great housing
"shortage” of postwar years was produced—by legal
fixing of maximum rents. That is why New York City
which is the only city in the country that still has legal
rent control is also the only city that still has a housing
shortage of the wartime type.
Do you want to produce a surplus of any product?
Simply have government fix and enforce a legal mini
mum price above the price that would otherwise pre
vail, either by making it illegal to pay less or by offering
to buy all that is offered at that price. That is why there
is a surplus of unskilled youths seeking jobs—because the
government makes it illegal for enterprises to pay less
than the legal minimum wage. That is why we were
plagued for so many years by agricultural surpluses—
because the government pegged farm prices at levels

T h e B a l a n c e o f Pa y m e n t s

above those that would have cleared the market.
The same fixed price may at one time produce a sur
plus and at another a shortage. An excellent example is
the price of silver. When,-at the end of 1933, the U.S.
government first offered to buy all newly produced
domestic silver at 64-64/99 cents an ounce, this price was
well above the price that would clear the market—at the
end of 1932, silver had been selling on the open market
for as low as 25 cents an ounce. The result of this action
plus the subsequent Silver Purchase Act of 1934 which
authorized purchases abroad as well, plus subsequent rises
in the fixed price, was a veritable flood of silver. We
drained China, Mexico, and the rest of the world, more
than tripling our stocks of silver. Since 1955, however,
the price has been below the price that would clear the
market— thanks to price inflation at home and abroad
and despite further rises in the pegged price to $1.29.
As a result, there is now a shortage instead of a surplus.
We are keeping the price down only by rapidly depleting
our reserves. We shall be forced to let it rise sometime
in the next few years.*
Wheat may be or may become another example. F o r
many years, the great problem was the surplus generated
by our pegged price. We were forced to build mammoth
storage facilities, to impose extensive restrictions and
controls on farmers to keep down their output, to tol
erate a different price at home and abroad, controlling
foreign trade in wheat in order to do so. Now, as world
* As occurred not long after the lecture was delivered.

F ir st L e c t u r e

population and food needs are booming and inflation
proceeds on its merry way, the pegged wheat price may
be or may become too low. If so, our stocks will be
rapidly drained.
As these examples suggest, the technique of fixing
prices is an extremely powerful tool. The result will
often appear far out of proportion to the cause. Fix the
price only a little too high and there will appear to be
a tremendous surplus because the price will simultane
ously discourage buyers and encourage sellers. In addi
tion, it will cause the disappointed sellers to make multi
ple offers which will make the supply look larger than it
is. Every attempt to curtail supply by government regu
lation will be met by the ingenuity of the myriad of pri
vate suppliers trying to find some way around the regu
lations, so that there will be a continual tug-of-war,
with the regulations piling ever higher. Fix the price only
a little too low and there will appear to be a tremendous
shortage, because the price will simultaneously encourage
buyers and discourage sellers. In addition, disappointed
buyers will stand, or have stand-ins, in more than one
queue.
The situation is reminiscent of Micawber’s law, as
reported by Charles Dickens, "Annual income twenty
pounds, annual expenditure nineteen six, result happi
ness. Annual income twenty pounds, annual expenditure
twenty pounds ought and six, result misery.”
The apparent disproportion between cause and effect
is the major hindrance— as I have discovered again and

T h e B a l a n c e o f Pa y m e n t s

again—to public understanding of the phenomenon.
How can it be, the ordinary man is likely to say, that
prohibiting landlords from raising the rent—surely no
more than a simple act of justice—can have such farreaching effects as long lists of people seeking apartments
relative to apartments available, widespread complaints
of a shortage of housing space— even though the number
of dwelling units per person may be at its all-time maxi
mum—the development of black markets, the deteri
oration of rental housing, and so on and on? Can it
be, the same intelligent layman is likely to say, that
the entire complicated farm surplus problem, with its
panoply of regulations, elections among farmers, plowing
under of hogs, taking land out of cultivation—that this
whole problem simply reflects government’s attempt to
assure parity prices for farmers? Surely something more
basic and fundamental must be involved.
Yet the truth is, nothing more is involved. Fix prices—
and the problems will multiply; let prices find their own
level in free markets—and the problems will disappear.
The abolition of rent control everywhere in the United
States except New York City shortly after the war is
one dramatic example. The “ shortages” disappeared al
most overnight. The real problems of high cost of build
ing and of urban blight of course remained—but the
false problems disappeared. And in this case, New York
City remained as a control to illuminate the source of the
problems. The abolition of price control in Germany by
Ludwig Erhard one Sunday afternoon in 1948 is another

F ir st L e c t u r e

and even more dramatic example. That was all it took
to release Germany from the chains that were producing
stagnation at a level of output half the prewar level and
to permit the German miracle to occur.
All of this may seem far afield from my announced
topic but it is not. The problem of the balance of pay
ments is simply another example of the far-reaching ef
fects of government price fixing, complicated only by
two facts: first, that two sets of prices are involved—the
price of gold in terms of various national currencies, and
the price of national currencies in terms of one another;
second, that more than one country is involved.
The existence of two sets of prices is a relic of an
earlier day, when there was a real gold standard, and
“ dollar,” “ pound,” and "franc” were simply names for
different amounts of gold. Under such a gold standard,
government’s role is primarily simply as a mint, to
certify the weight and fineness of the gold, coin it on de
mand, issue warehouse certificates for gold, and redeem
the certificates—though in practice governments also
issued promises to pay gold not fully backed by gold.
Under such a system, exchange rates were kept in nar
row bounds—within the "gold points”— not by govern
ment price fixing but by the private shipment of gold
whenever the market price varied by enough to make
it worthwhile to acquirc a foreign currency by shipping
gold rather than by. an exchange transaction. Exchange
rates stayed within narrow limits for the same reason
and in the same way that the price of sugar in New York

T h e B a la n c e o f P ay m e n ts

never deviates much from the price of sugar in Chicago
—because if it did deviate, it would pay private traders
to ship sugar.
The movements of gold that kept exchange rates in
line also served to produce adjustments that made the
gold flows self-limiting. The country shipping gold ex
perienced a decline in the quantity of money; the coun
try receiving gold, a rise. The monetary changes in turn
affected incomes and prices, and therewith the demand
for foreign exchange, lowering the demand in the coun
try shipping gold, and raising it in the country receiving
gold. The key feature of this process was that it was
completely automatic and gradual. There was no way the
gold movements could be prevented from affecting the
money stock. A small discrepancy called forth a small
adjustment. There was a unified currency system, not a
collection of national currencies linked by fixed rates.
Such a unified currency exists today among the different
states of the U.S., between Britain and some of its co
lonial territories, like Hong Kong, and in many similar
cases, but not among such areas.
The situation today is clearly very different. Gold is a
commodity whose price is supported by the government
—like wheat or butter. The major difference is that we
support the price only for foreigners not for U.S. citi
zens, since it is illegal for U.S. citizens to hold gold except
for numismatic or industrial purposes. In addition, gold
has the special property that at the moment there is a
highly elastic foreign demand for it, so we can always sell

F ir s t L e c t u r e

it to acquire foreign exchange. Clearly, we could peg
the price of gold even though exchange rates were not
fixed. For example, Canada’s having a floating exchange
rate, as it did from 1950 to 1962, did not prevent us from
continuing to peg the price of gold even though Canada
is a large producer of gold. There would have been a con
flict only if Canada had also tried to peg the price in
terms of Canadian currency.
The levels at which exchange rates are now fixed are
calculated from the official price of gold each nation lists
with the International Monetary Fund. But it is clear
that exchange rates are not kept within narrow bounds
by the movement of gold. Most countries that have fixed
exchange rates with one another do not freely buy and
sell gold. The U.S. does indirectly on the London gold
market with the cooperation of the Bank of England, but
it does so in order to peg the gold price, just as we sell
silver to peg the silver price, not as the primary means
of fixing exchange rates. We could abandon the pegging
of the price of gold and yet continue to peg exchange
rates, just as the pegging of exchange rates does not re
quire the pegging of the price of lead, or copper, or steel.
Gold is now at most window dressing, not the king pin of
the monetary system that determines the quantity of
money. Hence, I propose in this paper to concentrate on
exchange rates, leaving mostly to one side, as a subsidiary
issue, the price of gold.
The second complication is that more than one gov
ernment is involved. Consider, for specificity, the case

T h e B a l a n c e o f Pa y m e n t s

of Britain and the United States. The official price of
the pound sterling in terms of the dollar is $2.80, but
our agreement with the IMF permits the price to fluctu
ate a bit on each side of.that, roughly between $2.82 and
$2.78. The U.S. is committed to keeping the price from
rising above $2.82—since that would constitute a de
preciation of the U.S. currency; the British are com
mitted to keeping the price from falling below $2.78—
since that would constitute a depreciation of the British
currency. O f course, there is nothing to prevent either
country from engaging in transactions that help the
other keep its commitment, but that is the formal divi
sion of responsibility.
The U.S. can keep the price from rising above $2.82
only by offering to sell all the pounds demanded at that
price—i.e., to buy all the dollars offered; the British can
keep the price from falling below $2.78 only by offering
to buy all the pounds offered at that price—i.e., to sell
all the dollars demanded at that price. How can the two
countries succeed?
Suppose, that, at a price of $2.82 per pound, the num
ber of dollars that people or governments wish to use to
buy pounds in order to spend, lend, or give away is
greater than the number of dollars that other people or
governments wish to acquire with pounds. Suppose, that
is, that the U.S. has a potential balance-of-payments
deficit. How can the U.S. keep the price at $2.82?
Clearly, there are basically only two ways: by providing
the additional pounds, either out of its own reserves of

F ir st L e c t u r e

foreign exchange or by borrowing them from someone
else; or by inducing or forcing people to change the
number of pounds they seek to buy. And the converse
statements hold for the British in the contrary case.
To use the language that has become common, there
are two problems: the liquidity problem—having enough
reserves to be able to meet demands; the adjustment
problem—keeping demand in line with supply. This is
the precise counterpart of the problem for wheat: the
liquidity problem—accumulating or decumulating
wheat stocks; the adjustment problem—keeping down
the production of wheat or stimulating its consumption.
Superficially, it looks as if the liquidity problem could
be easily solved simply by reversing the tasks assigned the
United States and Britain. Let Britain keep the price of
the pound sterling from rising above $2.82 by offering
to sell an unlimited number of pounds at that price and
let the United States keep the price of the pound from
falling below $2.78 by offering to buy an unlimited
number of pounds at that price. Each can always do so.
Britain manufactures pounds and the United States
manufactures dollars, so each can always meet its com
mitments. However, in doing so, each is in effect giving
the other country a blank check on its own goods and
serices. If the price of the pound were tending to rise,
Britain would be accumulating dollars. The counterpart
would be a flow of goods from Britain to the United
States. Britain would in effect be giving the United States
an unintended loan at a zero interest rate. This is pre-

T h e Ba la n ce o f Pa y m en ts

cisely what happened to Germany for many years: it
accepted a large inflow of dollars, which meant that
it was selling a larger dollar volume of goods than it was
buying; it was implicitly exporting goods on credit.
Clearly no country will be willing to do this indefinitely.
Yet this approach is worth mentioning, because it is
the lure that underlies all the talk of an international
agreement to create "paper gold,” new international re
serves. At bottom, what is involved is an agreement by
countries to make automatic loans to one another. Every
country will be in favor of such an agreement, in princi
ple. But each will want a different agreement—one that
enables it to borrow much and commits it to lending
little. Thus I predict, without fear of successful contra
diction, that despite all the appearance of agreement in
principle, no effective agreement will in fact be reached.
To return to the United States’ liquidity problem. The
alternative to Britain’s providing an unlimited line of
credit at zero interest is for the United States to build
up reserves in advance from which it can meet excess
demands for pounds—this is indeed the important role
played by our gold stocks—or to arrange to borrow as
the occasion demands.
Clearly, potential deficits cannot be met indefinitely
out of reserves. Reserves are necessarily limited. Clearly,
also, to meet the deficits indefinitely by governmental
borrowing abroad would be costly and undesirable. And,
on the other side, no country will be willing to accumu
late another country’s currency indefinitely. Reserves

F ir s t L e c t u r e

alone cannot do the job. There must be some adjustment
mechanism.
What possible adjustment mechanisms are there? One
is the standard gold-standard mechanism—changes in
the quantity of money, income, and prices internally.
After all, the only reason a problem arises is because the
existence of central banks interferes with this mecha
nism. With central banks, a payment deficit need not
mean a reduction in the quantity of money, because the
central bank can offset it, and a surplus need not mean
an increase. Indeed, central banks are a necessary—and
today almost a sufficient—condition for a balance-ofpayments problem.
A central bank could do deliberately what the real
gold standard did automatically. To correct the United
States balance-of-payments deficit, it could reduce the
quantity of money (or reduce the rate of growth),
lowering incomes and prices—or letting them rise less
rapidly than in other countries. This would reduce the
demand for foreign exchange and increase its supply.
The United States has done this to some extent. It is
clear that monetary policy was tighter than it otherwise
would have been from 1956-61 because of the balanceof-payments problem. But it is also clear that it is both
unlikely that the United States would put major reliance
on this adjustment mechanism and undesirable that it
should do so. It is unlikely because of the government’s
commitment to full employment. It is almost inconceiv
able that any administration, of either party, would be

T h e B a l a n c e o f Pa y m e n t s

willing to force a significant domestic recession or de
pression to resolve a balance-of-payments problem. It is
undesirable that the United States put major reliance on
this adjustment mechanism partly because foreign trade
is so small a part of our economy—it is absurd to force
95 percent of the economy to adjust to 5 percent rather
than the other way around. More basically, it is undesir
able because many of the adjustments forced on us are
likely to be the product not of changes in the real forces
of demand or supply but of monetary manipulations of
other countries.
This adjustment mechanism is the one which the pro
ponents of fixed rates regard as the "discipline” imposed
by the fixed-rate system. But it is a peculiar discipline.
The discipline of fixed rates forced inflation on Germany
in the past decade at least as effectively as it forced de
flation on us. The only discipline is to keep in step with
the rest of the world, not to march in the right direction.
In any event, it is clearly a discipline that we are not
willing to accept.
The only other adjustment mechanism—while peg
ging exchange rates—is to control by direct or indirect
means the amount of foreign exchange people try to
buy— the counterpart to restrictions on the production
of wheat. Britain and other countries have, of course,
extensive exchange control. A resident of Britain may
not exchange pounds for dollars without the explicit
permission of a government official. This has involved
extraordinarily detailed control of the day-to-day life

F ir st L e c t u r e

of the British citizen—where he may go on a vacation,
what books he may read, and so on, ad infinitum.
We have so far avoided explicit exchange control, but
we have interfered in many ways with private trade
—some serious, some niggling, some demeaning. Oil
import quotas, meat quotas, and quotas and tariffs on
many other products have been justified as means of
"saving” foreign exchange. The niggling reduction of
the duty-free tourist’s allowance has the same origin. So
has the demeaning spectacle of our negotiating "volun
tary” quotas on exports from Hong Kong and Singapore
and Japan. Our high officials have gone hat-in-hand to
France and Germany and other countries to plead for
earlier repayments of loans and special purchases of
American goods. We have required recipients of foreign
aid to buy American goods—giving with one hand and
taking away with the other. We have preached free trade
and practiced restriction. And most recently, we have
gone in for "voluntary” controls on foreign lending by
banks and foreign investment by enterprises. And the
end is not yet.
With all this we have not succeeded. The experience of
countless price-fixing schemes has been repeated. Let the
fixed price differ from the price that would clear the
market, and it will take herculean efforts to hold it.
Consequently, we have also been driven to the final
adjustment mechanism—changes in the exchange rate.
We profess to have kept the exchange rate rigid. Yet we
have in effect devalued it selectively. That is what the

T h e B a l a n c e o f Pa y m e n t s

interest equalization tax amounts to. For purposes of
buying foreign securities, the dollar has been devalued
by 15 percent, and a further devaluation is proposed.
That is also what our program of reducing the foreign
exchange component of military expenditures amounts
to. Our military authorities are instructed to compare
the cost in dollars at the official exchange rate of pur
chasing an item abroad with the cost in dollars of buying
it in the United States. If the cost in the United States
exceeds the foreign cost by less than x percent, they are
instructed to buy it at home—paradoxically to save dol
lars. I do not know what x is but I understand that it is
sizable, something over 50 percent. The tying of foreign
aid is another example.
We sneer at South American countries that adopt
* multiple exchange rate systems. Yet that is what we have
adopted—only in concealed form.
There is one and only one satisfactory solution: abolish
governmental price fixing. Let exchange rates become
free market prices determined primarily by private deal
ings. Let the government simply stay out of the picture.
Suppose, under such a system, that, at a price of $2.80
to the pound, the number of dollars that people want to
use to buy pounds to spend, lend, or give away is greater
than the number of dollars holders of pounds want to
acquire. The eager buyers will offer to pay more. The
price of the pound will be bid up. As it rises, buyers of
pounds will be discouraged—because a higher price of
the pound means a higher price in dollars for goods and

F ir st L e c t u r e

services bought abroad—and the sellers of pounds will be
encouraged—because a higher price of the pound means
that they can buy more United States goods and services
with a given number of pounds. A t some price, say $3.08,
the number of dollars offered will be equal to the num
ber of dollars demanded.
This rise in the price of the pound by 10 percent will
have had precisely the same effect on the relative costs
to Americans and Britons of American and British goods
as a decline of 10 percent in United States prices with
no change in British prices, or a rise of 10 percent in
British prices with no change in United States prices. But
how much easier it is to have the exchange rate change
by 10 percent than to get a general decline in all United
States prices by 10 percent. Why not have one price—
and that a potentially highly flexible one— do the adjust
ing rather than require the myriads of domestic prices
to vary, with all their stickiness and all the side effects?
Why not have the dog wag the tail, instead of the tail
wag the dog?
As this example suggests, a system of floating exchange
rates completely eliminates the balance-of-payments
problem—just as in a free market there cannot be a
surplus or a shortage in the sense of eager sellers unable
to find buyers or eager buyers unable to find sellers. The
price may fluctuate but there cannot be a deficit or a
surplus threatening an exchange crisis. Floating exchange
rates would put an end to the grave problems requiring
repeated meetings of secretaries of the Treasury and

T h e B a la n c e o f Pa y m e n t s

governors of central banks to try to draw up sweeping
reforms. It would put an end to the occasional crisis pro
ducing frantic scurrying of high governmental officials
from capital to capital, midnight phone calls among the
great central banks lining up emergency loans to support
one or another currency.
Indeed this is, I believe, one of the major sources of the
opposition to floating exchange rates. The people en
gaged in these activities are important people and they
are all persuaded that they are engaged in important
activities. It cannot be, they say to themselves, that these
important activities arise simply from pegging exchange
rates. They must have more basic roots. Hence, they say,
it is simpleminded to believe that freeing exchange rates
would eliminate the problem. That is what the allied
advisers engaged in price control, rationing, and the like
told Erhard that summer in 1948. That is why he re
moved price controls on a Sunday, when they were not
in their offices to countermand his edicts.
Under a system of floating exchange rates, the liquid
ity problem disappears. There is no need for official
foreign exchange reserves. Private individuals will pro
vide the reserves needed—just as they do in commodities
that trade in a free market. If a given movement in ex
change rates seems temporary, it will be in the selfinterest of private holders of exchange to dampen the
move by speculation and they can be counted on to do so.
With floating rates, we could therefore terminate at
once the frustrating and ineffective negotiations for a

F ir st L e c t u r e

new international liquidity arrangement— negotiations
that are in any event bound to fail. More important, we
could abolish at once the interest-equalization tax and
informal exchange controls.
Most important of all, floating rates would enable us
to separate issues and determine our national policies
on the right grounds. Monetary and fiscal policy could
be directed toward pursuing internal stability without
being hamstrung by the. balance of payments. We could
decide how much foreign aid to give in terms of our re
sources and our values, not by the irrelevant considera
tion of the currency in which it is expressed. We could
instruct the military to buy in the cheapest market and
keep the real costs to a minimum—not turn them into
a foreign exchange authority. We could conduct foreign
policy in terms of our true national interests—not in
terms of the effect on gold flows. We could behave in
foreign trade like a great nation, not like a mendicant,
by unilaterally moving toward freer trade without hav
ing to be concerned about balance-of-payments prob
lems.
This last point perhaps deserves a slight digression. Not
the least of the advantages of floating rates, in my opin
ion, is that it makes it so much easier for the layman to
understand the merits of free trade. With rigid rates,
the first effect of a reduction in tariffs is an increase in
imports without any immediate effect on exports. It
looks as if imports have simply displaced domestic prod
ucts and so produced unemployment. It takes a subtle

T h e B a l a n c e o f Pa y m e n t s

chain of reasoning to show that this is only part of the
story, that the increase in imports will have indirect
effects that will ultimately lead to an expansion of ex
ports so that the final result is an increase in foreign
trade not an increase in unemployment. And, indeed,
with our present nearly paralyzed adjustment mecha
nism, the indirect effects may be long delayed and highly
unreliable.
With floating rates, a reduction in tariffs will also
produce an attempted increase in imports. But how can
this be realized? Only if the importers can get some for
eign exchange. To do so they will bid up its price which
immediately makes exports more attractive to foreigners.
The first effect of a reduction in tariffs is thus a rise in the
price of foreign exchange and a simultaneous increase in
imports and exports. There is not even a temporary im
portation of unemployment.
The floating rate provides the protection to the balance
of payments that is essential if we are to move signifi
cantly to ease barriers to trade. In the absence of such
protection, it appears as if we can afford to reduce bar
riers only in return for a reciprocal reduction of barriers
by others. The result is the kind of drawn-out and in
effective negotiations that are currently nearing their
appointed end in connection with the Kennedy round.
What objections have been raised against floating
rates?
One is the allegation that we cannot move to floating
rates on our own, that just as two governments are now

F ir s t L e c t u r e

involved in pegging each rate, so it will take two to
unpeg. This is in one sense correct, yet it is irrelevant.
The United States can announce that it will no longer
try to keep the dollar from depreciating—i.e., in the
case of the pound, no longer try to prevent the price of
the pound from rising above $2.82. If Britain wants to
take on the task of keeping the price of the pound from
rising, fine. It can do so only by either being willing to
accumulate dollars indefinitely—which is to say, by ex
tending us an unlimited line of credit— or by adapting
its internal policy to ours, so that the free market rate
stays below $2.82. In either case, we can only gain not
lose. Similarly, if it chooses to continue to keep the price
of the pound from falling, that again is no cause for con
cern on our part. It can only do so by using dollar re
serves, which we must be ready to permit, or again by
aligning its internal policy with ours.
I think it highly likely that if we announced that our
government will no longer intervene in the exchange
market, a fair number of other countries would peg their
currencies to ours. I see no harm in that and much good.
Perhaps we could begin to build up a truly unified cur
rency area—not a collection of national currencies linked
by pegged rates. A system of floating exchange rates has
basically much more in common with a real gold stand
ard—in that both leave private individuals free to buy
and sell currencies as they wish and both are free of
government intervention— than either has with our pres
ent system.

T h e B a l a n c e o f Pa y m e n t s

A second objection that is raised is that floating ex
change rates would be highly unstable and that unstable
rates would add to the uncertainty and difficulty of con
ducting foreign trade. However, floating rates need not
be highly unstable. Canada had floating rates from 1950
to 1962 and they were highly stable. If floating rates are
highly unstable, it will be because the internal monetary
policies of the countries or some other aspects of their
economy are highly unstable. But in that case, the un
certainty is there and the only question is what form it
takes. Under a real gold standard, the uncertainty would
be about internal price levels, because they would reflect
the instability. Under pegged exchange rates, the uncer
tainty would be about whether exchange would be
available, that is, what the exchange controls would be
like. If anything, the uncertainty about the price of for
eign exchange under a floating rate system is the easiest
for a trader to protect himself against by hedging in a
futures market.
A related argument is that the uncertainty under
floating rates would be greater than under other systems
because floating rates would give rise to destabilizing
speculation. When I first began writing on this subject
nearly two decades ago, I took this objection seriously. I
no longer do. In the interim, there have been a consider
able number of careful empirical studies of speculation
under floating rates. None has produced a clear example
of destabilizing speculation on any significant scale. And
the bulk of the evidence strongly supports the view that

F ir st L e c t u r e

speculation has generally been clearly stabilizing. I think
it is time therefore that this bug-a-boo is given a decent
burial —at least until somebody can come up with some
real evidence that it is more than a bug-a-boo.
Another objection to floating rates is that it reduces
the attractiveness of a country as a financial center. This
can be correct. It may well be that Britain was at one
time well advised on this score to maintain rigid rates
with other countries or that Switzerland is now. But this
seems to me not a relevant objection for the United
States. First, our international financial activity is not a
major industry. Second, its development is interfered
with at least as much by the measures—like the interest
equalization tax and "voluntary” controls on foreign
lending—that we take to peg the rates as it would be by
floating rates. Third, the formation of a "dollar” bloc,
suggested as a possibility above, might be a favorable
development. Fourth, without the interest equalization
tax, informal exchange controls, and extensive trade bar
riers, the dollar would very likely be used even more ex
tensively than it is as an international currency. Para
doxical though it seems, letting rates float, and removing
controls, may be the most effective way to strengthen
New York’s role as a financial center.
The major objection raised against floating rates is one
already mentioned—that it would remove the "disci
pline” which fixed rates are said to impose on domestic
economic policy, that it would open the door to irre
sponsible inflationary monetary policy. This objection

T h e B a l a n c e o f Pa y m e n t s

has merit if the alternative were a real gold standard. It
has some merit for countries like Italy and Japan that
have been susceptible to highly inflationary policies, that
have been willing to submit to the discipline of the bal
ance of payments, and for which foreign trade is a
substantial part of total trade. It has negligible merit
for the United States. Foreign trade is so small a part
of total trade, and our reserves are so large, that we
can neglect the balance of payments for long stretches
of time, letting small disturbances build up into
big ones. And even then, we are not willing to
submit to the discipline. Instead, we resort to import
quotas, tariffs, multiple exchange rates, and infor
mal exchange controls. The same discipline which
produced these, incidentally also produced inflationary
pressure from 1945 to 1956 when we were accumulating
gold and foreign exchange. The discipline is asym
metrical: we yield to it when it imposes inflation on us;
we resist it when it calls for deflation. That is a kind o f
discipline that I think we can do without.
These are the objections to floating rates. But they are
not the reasons why we do not—and very likely shall not
—adopt floating rates.
The most important reason we stick to pegged rates is
the tyranny of the status quo. The United States has taken
the public position that the dollar will be defended. The
President and other high officials have committed them
selves over and over again to the proposition that the
dollar will not be devalued, that the present system o f

F ir st L e c t u r e

pegged rates is one of the great postwar achievements,
which the United States will support with might and
main. Once such a position is taken, it takes a major
crisis to produce a change.
A second reason is the confusion between a real gold
standard and the pseudo-gold standard we now have. The
public at large and in particular the financial community
hankers after the freedom from government interven
tion of a real gold standard. It confuses the pegged rates
of our present system with the rigid rates of a real gold
standard.
A third reason is the confusion between devaluation
and a system of floating rates. A particular exchange
problem can be met by changing the level at which the
exchange rate is pegged. Such a system, under which the
level at which the exchange rate is pegged is changed at
substantial intervals of time, is the worst of both worlds.
An adjustable peg provides neither the certainty of a
truly fixed rate nor the flexibility of a floating rate. It is
certain to be subject to destabilizing speculation. Such a
system must be sharply distinguished from a system of
floating rates. Devaluation of the dollar to a new pegged
level would, in my opinion, be most unwise; whereas
establishment of a system of floating rates is eminently
to be desired.
A final reason is what may be called the Arizona effect.
As you may know, the death rate from tuberculosis is
higher in Arizona than in any other state in the country.
Clearly, Arizona must be a most unhealthy place to live.

T h e B a l a n c e o f Pa y m e n t s

Similarly, floating exchange rates have often been
adopted as a last resort by countries experiencing grave
financial crises when all other devices have failed. That
is a major reason why they have such a bad reputation.

SECOND LECTURE

RO BERT V. ROOSA
Any debate on a subject as vast as the balance of pay
ments needs narrowing down to particular issues. Fortu
nately this one, as I understand the intention of our
sponsors, is to be centered on the special significance of
the exchange rate in balance-of-payments adjustment.
That relieves me of any need for introductory remarks
on the current plight of the United States, and projects
me right into a few generalizations about the relations
between external and internal adjustments in a modern
economy. After that brief preface, I can take up my
main theme—that the best way to understand the value
of the present system of fixed exchange rates is to see
what would be wrong with a system of "free” or "flex
ible” rates.
A country’s external accounts, summarized through
the flows recorded in its balance of payments, are under
standably important for what they show, looking out
ward, about the country’s economic relations with the
rest of the world. These accounts may, when in deficit,
or in surplus, or even at times when in equilibrium, point
to distortions that need correction in the country’s own

T h e B a la n c e o f Pa y m e n t s

behavior toward foreign trade or aid or capital move
ments. Or they may simply show that disturbances or
deficiencies elsewhere are creating difficulties for a par
ticular country. But a country’s balance-of-payments
accounts can also be important for what they frequently
reveal, looking inward, about the country’s own do
mestic economy—whether it is achieving the mix o f
saving, investment, and consumption, or of prices and
wages and profits, or of commodities, industries, and
services, for example, that would be most likely to assure
sustained advances in employment, incomes, and general
welfare at home.
Not only outwardly but also inwardly, then, the bal
ance of payments can have a pervasive significance for
the economic policy of any nation. Over and above all
this, for us, of course, there are the special balance-ofpayments concerns of the United States as the supplier
of much of the world’s internationally usable money,
and the special implications of this country’s dominant
size as both exporter and importer, and lender and bor
rower—serving in effect as a pivot for the movements
of money, trade, and capital throughout the world. B ut
what I want to stress, from the very outset, is my own
conviction that the balance of payments of this country
or of any country often mirrors, in outline if not in
detail, the mistakes and the achievements of its own do
mestic economy.
In general, a country whose external accounts are
seriously and continually out of balance often has some-

Second L ec tu re

thing going seriously awry within its own economy, and
a look through the window provided by the balance-ofpayments accounts will usually help to locate the cause.
This proposition is not particularly novel when it is
expressed concerning countries in deficit. But countries
with sustained large surpluses, though less likely to be
urgently concerned, will also usually find that their swol
len earnings are partly the result of conditions at home
that will need correction or adjustment sooner or later in
order to avoid potential (or possibly even present) dif
ficulties in the domestic economy.
To many of my economist friends the comments I
have just made are not merely polite introductory cliches,
but challenging, fighting words. For I am really saying
that balance-of-payments viability is, or should be,
among the principal tests and guides for the functioning
of domestic economic policy. And in the context of the
topic for tonight’s discussion, I am going to go further
and assert that this essential, indeed inescapable, testing,
guiding, or even "disciplining” that the outside world
provides for each individual national economy is what,
in the end, impels all countries toward some sort of ap
proximation of a market economy. It seems to me that
national economies simply project on a wider canvas,
and in their own more complicated way, images of many
of the same problems and the same needs that we iden
tify much more familiarly with individual firms, and the
relations among firms. I even suspect, or at any rate can
hope, that Professor Friedman will agree that movement

T h e B a l a n c e o f Pa y m e n t s

toward the characteristic relationships of the market
economy is as relevant for the flourishing of nations as
for firms.
I must concede, of course, that the path toward the
market economy is very long and roundabout for many
countries—measured in generations rather than in years.
I must also concede that the working out of balance-ofpayments disciplines often occurs in strange ways—
harsh, crude, and extreme. But what I want to argue,
while enjoying the privilege of joining issues with one of
the world’s most distinguished exponents of market
economics, is that a system of fixed rates of currency ex
change provides the most hospitable environment for en
couraging market-oriented adjustments within and
among nations.
To do that I will first have to state a little more con
cisely what it is that I mean by a fixed exchange rate
system, and contrast that with several variants of flexible
exchange rates. After identifying my side of this argu
ment, I can return for a lengthier look at the nature of
the need that I see, in principle, for a fixed-rate system,
and also at the reasons why, conversely, it seems to me
persuasive even on purely theoretical grounds that a
flexible-rate system would undermine rather than en
courage marketplace adjustments.
In much of the conventional discussion of these issues,
there has too often been a tendency to concede a theo
retical case for flexible rates as a counterpart of the
flexibility that characterizes a market economy. I am

Seco n d L e c t u r e

not that congenial. Moreover, after I have outlined my
reasons for challenging the case for flexible rates on
broadly theoretical grounds, I intend to go further to
describe some of the practical considerations that would
render a system of flexible rates unworkable, even if
there were good theoretical grounds for making the at
tempt. And beyond the generally applicable constraints
of realism, there are also the direct and special needs of
the United States. I make no apologies for regarding
these, too, as crucial, and so will add to my general list
of operational obstacles some mention of the unique dis
advantages that a flexible rate system would impose on
the United States.
There is still a further and final set of considerations,
however, which must also be faced frankly. For even
though we might agree that a system of fixed exchange
rates is clearly preferable on theoretical grounds, and
even though it is clearly a much more workable system
for the world as a whole, and for the United States, we
do have to ask whether some other kind of fatal flaw
might yet develop in the structure of the fixed exchange
rate system—a crack in the structure that could require
its abandonment. Specifically, a fixed-rate system needs
a gradually increasing supply of internationally usable
monetary reserves. Is there a risk that the world will run
out of acceptable monetary reserves; that the countries
of the world, taken together, cannot add to their hold
ings of reserves fast enough, as world trade grows, to
support their own exchange rates on a fixed standard

T h e B a l a n c e o f Pa y m e n t s

through alternating periods of relative strength or weak
ness for one country after another? If the fixed exchange
rate system is inexorably headed for this kind of collapse,
because there are not enough reserves to go around for an
expanding world economy, then we might by default
have to fall back on a system of flexible exchange rates,
despite the forbidding shortcomings of that system.
My concluding comments will suggest that a satisfac
tory answer can and will be found. Indeed, perhaps
paradoxically, I suspect it will be an answer reminiscent
of some of Professor Friedman’s other writings. N ot the
answer of flexible rates, but instead I expect that the
countries of the world will initiate for their global reserve
needs some version of that proposal for a regular and con
tinuing creation of money that has so long been identi
fied, for the needs of domestic economies, with Professor
Friedman’s name. But before getting to that, there is
much other intervening ground that I must cover. First,
a brief sketch of what I think is meant conceptually by
a fixed-rate system, in contrast with one consisting o f
"free” or flexible exchange rates.
WHAT ARE FIXED EXCH AN GE RATES?
Under a system of fixed exchange rates, each country
defines a parity for its own currency that is, if I may be
unambiguously clear, neither rigidly fixed nor freely
flexible. Moreover, a number of countries—ordinarily
including some that are less developed, and others that

Seco n d L e c t u r e

are going through a radical transition of some kind—will
set no parity at all.
What all of this does mean, though, is a system which
presumes that a central core of leading countries, through
whom most of the trade and payments move, will each
set a parity for its own currency; that these parities will
be defined in terms of some commonly accepted stand
ard or norm, such as the dollar; and that cross rates can
consequently be calculated among these various cur
rencies for determining with fair accuracy how many
Japanese yen, for example, would be equivalent to one
Swedish kroner, or one Mexican peso. There can, around
these parities, be some moderate fluctuation, such as the
plus-or-minus 1 percent permitted by the International
Monetary Fund for countries that have declared a parity
to the Fund, have met certain other criteria, and are
prepared freely to buy or sell their own currency at
prices within that band.
Any of these countries, too, on proper notice to the
Fund can actually alter their own declared parity (mak
ing up to the IMF for any losses that it might otherwise
suffer on its own holdings of that currency, in the event
the change in parity is a devaluation). Those who use
the currency of such a country—individuals, firms, or
nations—may possibly take out protection against a
devaluation, for example, by making certain they owe
some debts in a vulnerable currency to offset their claims
denominated in that currency. Or they may avoid com
mitments in suspect currencies by trying to get much

T h e B a l a n c e o f Pa y m e n t s

of their business with such countries denominated an d
payable in a universally usable currency, such as U nited
States dollars. Or, in the case of about one-half dozen
of the leading countries, they may be able actually t o
"hedge” by selling short in the forward markets that are
maintained in these few currencies by specialized foreign
exchange traders.
There will also be other countries which have n o t
declared a parity. Their monetary authorities make their
own rules for buying and selling their own currency in
terms of some universal standard, such as the U nited
States dollar. But there is a general presumption, in a
world that is "on” a fixed exchange rate system, th a t
these countries will, when their own conditions show
some reasonable stability and strength, establish a parity
of their own. Meanwhile, these non-parity countries, a s
well as the traders, investors, and bankers of every
country, can conduct their daily affairs with the con
vertible countries in the comfortable assurance th a t
changes of any seripus magnitude in those currencies w ill
be quite unusual, and ordinarily preceded and high
lighted by special circumstances which prompt users o f
any such currency to take special precautions. That is,
most of the time, though with a wary eye for the o c
casional exception, anyone can assume that the yardstick
for measuring values among the leading countries, a n d
calculating in different currencies the prices of the prod
ucts which they buy or sell, has on it a fixed scale.
The undergirding of the fixed exchange rate system

Second L ec tu r e

is its common reference point, into which all parities can
be readily translated. Though one might perhaps imagine
others, the one we now have, which has the sanction of
time, usage, and universal recognition, is the dollar price
of gold. This is not the place, at least not yet, to debate
the gold price issue or the gold policy of the United
States. But I should make clear that the fixed-rate sys
tem, as presently conceived, does presume a fixed price
of gold. The parities of other currencies may be expressed
in terms of dollars, but that of the dollar is in terms of
gold.
The alternative system, that of free or flexible rates,
has presumably already been thoroughly described here
by Professor Friedman. Subject to correction after I learn
what he said, I need only make a few definitional com
ments. I do not want to quibble over some of the socalled middle-ground proposals. The notion of a slightly
wider band for spot-rate fluctuations around parity, say
to 2 or even 3 percent, I am going to leave aside. Simi
larly, at least in these prepared remarks, I will not discuss
the suggestions for slow but frequent and regular in
creases in the gold price. Both are essentially, I think,
rooted in acceptance of what I consider the fundamental
advantages of a fixed-rate system. My principal question
about them is whether, in trying to help strengthen the
functioning of that system, they introduce too many
new hazards of the kind that I see in a fully flexible
system. To be brief and blunt about it, in my view, the
wider band idea might someday be of some use; the

T h e B a l a n c e o f Pa y m e n t s

incremental gold price change would be an unmitigated
disaster.
The crucial issues concerning the nature of the world’s
monetary system really come into focus in a contrast
between the kind of fixed-rate world I have just sketched
and one in which no country attempts to set a parity for
its currency. In the purest formulation of a "free-rate”
world, a market springs up for the exchange of cur
rencies, and changing quotations for those currencies
are produced by supply and demand forces from hour
to hour. The market may have many locations, but pre
sumably active arbitrage and rapid communication
among centers will produce a consistent set of quotations
almost anywhere at any time. Even in this system, con
venience would dictate use of one or a very few cur
rencies as the common denominator, or unit of account.
And the rates of most smaller countries would be ex
pressed almost exclusively in terms of only one or two o f
the leading currencies.
In this rather rarified kind of truly "free” system,
there would be no intervention in the foreign exchange
market, and no action directly impinging on that mar
ket, by the central bank of any country. I leave it for
Professor Friedman to say whether this means that there
could not, in turn, be any central bank anywhere that
could take discretionary action to affect the money sup
ply of its own domestic economy; but I suspect that in
a rigorous analysis this would have to be the case. A t any
rate, because I feel happier when the dancing angels are

Second L e c tu re

brought down from the head of a pin on to solid ground,
I do not propose to spend very much time with the
"pure” version.
For most of what I shall try to say during these
debates, I will be talking not about "free” exchange
rates—the romantically appealing term used in the title
set for this program—but about "fluctuating” or "flex
ible” exchange rates. That is a system in which, broadly
speaking, there are no parities for any currency and the
exchange rate of any country is expected to decline
whenever its aggregate outpayments exceed its inpay
ments, to rise when its inpayments are the larger, and to
steady out when external payments are back in balance.
Short-lived erratic influences, it is assumed (and I stress,
assumed) , would be smoothed out by active spot and for
ward markets for all, or virtually all, currencies. Central
banks would be presumed to exist, but any intervention
by them in the exchange markets should be quite limited
—mainly to offset any misleading effects of their own
domestic actions that might otherwise obscure the under
lying supply and demand situation.
Having worked the idealized versions of both the fixed
and the flexible systems down somewhat closer to earthy
reality, I now feel more comfortable about asking how
they compare in terms of principle. After a look at that
question, I will move on to more operational considera
tions.

T h e B a la n c e o f P a y m e n ts

THEO RETICAL ASPECTS OF BO TH SYSTEMS
Both systems rest on the premise that money should
serve trade, and that the best money system is one which
serves not just passively but also constructively—that
is, it not merely assures the availability of an adequate
means of payment, but also helps to provide a climate
for confident expansion of the economy and for the most
productive allocation of resources. The differences be
tween the two systems show up in the way they might
be expected to work for the ordinary businessman and
banker; in the way in which they might affect the pat
tern of payments flows among countries; in their differ
ing impacts on the mechanism for balance-of-payments
adjustment among countries; and more broadly in their
implications for the flourishing expansion of trade in the
world economy over the years.
Under a fixed-rate system, there is an established scale
of measurement, easily translatable from one country to
another, which enables the merchants, investors, and
bankers of any one country to do business with others
on known terms. The flexible-rate system cannot offer
that kind of assurance. No single trader can know enough
about all the developments likely to affect the rate o f
exchange between his own country and that of his client
to make a firm contract without including a substantial
allowance for the risk that the rate of exchange between
the currencies involved will change while the transaction
is underway. Prices in international trade and the costs

Second L ectu re

of doing international business of any kind would thus
almost inevitably become higher under a flexible-rate
system—higher because the businessman must include a
charge for the added element of risk.
The customary reply of the "flexible rate school” is
not, in fact, an answer. Members of this school do cor
rectly point out that an exchange rate fixed at a dis
equilibrium level is also bad for trade and adjustment;
but as I have already stressed, individual rates (with the
exception of the United States, as further described
later on) can and should be changed when there is a
persisting disequilibrium under a fixed-rate system. But
there is also, under a fixed system, a two-way function of
established rates— that is, oftentimes the stability of a
rate provides the framework for a useful shifting of
supply and demand relationships into a settled equilib
rium, just as, at much less frequent times, imbalances
between the supply and demand for a currency may
require a change in the fixed rate itself. Moreover, the
obligation to maintain a fixed rate can often provide
a country with the needed incentive for carrying
through internal adjustments that are vitally needed
on purely domestic grounds.
But the further reply of the “ flexible raters,” with
respect to the potential burden of rate instability or
fluctuations upon the costs of doing international busi
ness, is that the banks or other foreign exchange traders
can easily generalize the cost of any uncertainty, and
in effect "average it out,” by maintaining an active for

T h e B a l a n c e o f Pa y m e n t s

ward market in all of the currencies. Then the trader
can pay a known and minimal cost for obtaining for
ward "cover” for his risk of exchange-rate fluctuation.
To be sure, a full-scale organization of foreign exchange
markets may help minimize the overall cost of this kind
of hedging. But the task of "making” a market in cur
rency futures when there are no known par values is
much more complicated than the operations of the
futures trader in a single commodity. When expecta
tions about the future performance of a given country,
and in turn about the strength of its currency, all begin
to run in one direction, where—in the absence of some
clear signal from monetary authorities who might be
prepared to intervene—do the proponents of flexible
rates expect to find private underwriters to make for
ward markets, except perhaps at costs which any foreign
trader would have to consider excessive?
Under a flexible-rate system, the trader must try to
take into account the whole range of prospective fu
ture payments from and to the particular country for
goods and services of every kind, as well as all manner o f
capital movements, and various states of mind. N or can
the relatively small per annum charges made by the
futures traders of today’s fixed-rate system be projected
into the conditions of a fluctuating rate system. For to
day’s futures traders work against known parities, not
an unknown array of unknowables, and they often op
erate within spread ranges that are protected by cen
tral bank intervention. Moreover, the very limits set by

Second L ectu re

the margins for spot-rate fluctuation do, as I can testify
from considerable experience, tend by themselves to
limit (though admittedly in a somewhat elastic way)
the range of forward-rate movements—even without
central bank intervention directly in the forward mar
ket itself. Yet even with all this shelter, the very aggres
sive and competitive community of foreign exchange
dealers in the world today only maintains active and re
liable forward markets in about a half dozen of the
world’s strongest and most widely used currencies.
Moreover, the burden of the risk premium on trade
and other payments is only the first of several compara
tive disadvantages of the flexible-rate system. Another
comes in the distorting influence exerted upon the com
position and pattern of payments flows among countries.
For as mentioned earlier, one other form of “ hedging”
against the exchange risk on any given set of payments
to or from a country is to have a debt transaction of
about the same magnitude running in the opposite direc
tion. So with or without active forward markets, business
firms of all kinds will have a compelling inducement to
buy from the countries in which they sell, and financial
institutions to incur debts to match their claims in each
country. To be sure, that kind of bilateral pairing, by
firms and by countries, will not always be manageable,
but what I am stressing is the ominous significance of a
system which inherently generates strong inducements of
this kind. They work exactly opposite from the presumed
objectives of a market economy, which should be to en-

T h e B a l a n c e o f Pa y m e n t s

courage an environment in which everyone may be able
to sell or buy, or borrow or lend, wherever the gain from
each transaction can be maximized.
Problems of this kind, added to the sheer wear and
tear of trying to do business with a rubber yardstick for
a measure, would not only be deterrents to the spread o f
diversification across national frontiers but actually
would, I am convinced, contribute to a greater economic
isolationism. A wall of currency uncertainty would be
built around every country. What this means is that the
inducement to bilateralize foreign trade would be ac
companied by a further inducement to trade at home,
within the area over which one’s own currency can be
freely used (that is, where exchange-rate fluctuations
are automatically avoided).
Granting all these dampening influences on the longrun expansion of multilateralized world trade that would
come with a flexible-rate system, might it still be possible
that gains of another kind could more than offset these
losses? What about the widely discussed need for im
provement in the balance-of-payments adjustment mech
anism? If a replacing of fixed rates with flexible rates
could assure a more effective and less disruptive pro
cedure for bringing the balance of payments of various
countries into viable relationships with each other, then
perhaps the price of some dampening in the growth o f
total trade might be well worth paying. And indeed, a
rather impressive case has at times been made that the
post-World War II premises of economic life now leave

Seco n d L ec tu r e

exchange rates as the only usable variable for achieving
adjustment flexibility. For wages, and prices, and em
ployment within a country can no longer, it is plausibly
argued, be reduced in order to improve the country’s
trading position abroad— that is, deliberately induced
slack is simply not acceptable as the way for a country
to “ adjust” itself into a position of overall balance or
surplus in its payments to and from other countries.
I certainly recognize that this is, when put starkly in
these terms, the most troublesome dilemma faced by the
fixed-rate system. It ignores, however, the role of capital
flows in balance-of-payments disequilibrium and in the
restoration of equilibrium. And so far as the trade ac
counts themselves are concerned, my own answer—with
out questioning the downside rigidity of prices, the
persistent upward pressure of wages, or the priority
appropriate for sustained employment—is that suit
able adjustment can still be accomplished under a
fixed-rate system through variations in the rate and
pattern of advance. For another premise of postwar
political economy is that each country’s economy will be
dynamic—continually growing, adapting, and changing
within a world of rising expectations. The engineering
of economic policy may be a little more difficult in such
a dynamic world; the analysis must be a little more
sophisticated and the policy action a little more delicate;
and we do not yet have accumulated patterns of experi
ence to make the judgments easier; but I see no reason
conceptually why the same kinds of adjustments cannot

T h e B a la n c e o f P ay m e n ts

now be achieved through variations which merely affect
the speed of an economy’s forward motion, and the com
position of its output, that could earlier have been at
tempted through deliberate deflation in the days when
the world lived on the premises of a relatively static
economy. And, of course, with capital markets grow
ing, the opportunities for aggravating, and for offset
ting, effects upon the balance of payments from capital
flows become increasingly important in qualifying any
generalizations made with respect primarily to trade.
The same can be said, moreover, concerning transactions
on government account.
One difference, at least for some years until the policy
makers of the world have more experience to build upon,
and a readier acuity for the fine tuning that will occa
sionally be needed, is that countries facing a need for
major adjustment in the structure or pattern of trade
will sometimes have to have a longer period to com
plete necessary adaptations than would have been con
sidered appropriate in earlier years. This means that in
dividual countries in deficit will in some circumstances
need access to more reserves—reserves that they own, or
reserves that can be borrowed, or both. And countries
in surplus will need to accept the implications of sur
pluses—both in the holding and the using of reserves.
But the essentials of the adjustment process under the
fixed-rate system, and now familiar to all countries, can
surely remain the same. Loss of reserves, perhaps accom
panied by a need to resort to borrowing, will impel a

Seco n d L e c tu r e

country to look inward upon the performance of its
own economy to discover the cause. And if prices are
rising rapidly, or wage increases are exceeding productiv
ity, or investment demands are outrunning resources, or
the production of salable export crops is being restricted,
or government requirements (including overseas spend
ing) are adding undue strain upon total capacity, for
example, then appropriate action can be taken, with
results that may alter the economy’s forward speed for
a time, but need not mean general recession and unem
ployment. Concern over depleting reserves, as well as
pressure to repay reserves that have been temporarily
borrowed, will under a fixed-rate system provide the
balance-of-payments "prod” to carry through such
action—even though the need for that action, if not the
recognition of the need, could be as great or greater for
purely domestic reasons. Reserve gains, in the opposite
direction, will be the stimulus to further reduction of
trade restrictions, the opening of capital markets to
foreign borrowers, the extending of foreign aid or
investment, and the expansion of the home market.
This is not the place, and I certainly lack the compe
tence, to digress further into these familiar and tantaliz
ing subjects. I do strongly urge, however, that it is not
only possible, but also productive, to resist the easy
tendency to see a hopeless conflict between balance-ofpayments viability and progressive gain in the domestic
economy. Conflict can exist, do not misunderstand me,
and not every proposed prescription for balance-of-pay-

Bala n ce

Paym en ts

ments correction under a fixed-rate system would fit
the premises of the modern economy. But I am sure
there can also be ways of using balance-of-payments
signals constructively to meet, in combination, both ex
ternal and internal needs. The strength and significance
of any such signals under a system of flexible rates would,
I am afraid, be much weaker.
For the intention, under a flexible-rate system, would
be for any country to let its exchange rate against other
currencies fall to a new equilibrium position whenever
aggregate outpayments exceeded inpayments. In effect,
whenever the balance of payments might point to some
thing wrong in the home economy—when the product
mix or the savings mix or the investment mix or some
other aspects of the domestic economy are out of joint
and need correction—the risk is that the exchange rate
would simply move to compensate for things the way
they are and the correction would be avoided. That
might theoretically be all right if balance-of-payments
problems were always merely warts on the economic
body, quite unrelated to its own functioning, arid an
annoying nuisance best removed through simple surgery.
But, as I would hope you might agree, that is rarely the
case.
To be sure, a declining exchange rate itself might be
considered sufficient cause for alarm to set corrective
domestic economic reactions in motion. But with no
norms to defend, and with no pressure from reserve
losses or from needed borrowing to reinforce the policy

Seco n d L ec tu r e

maker’s resolve, the likelihood of any corrective reaction
pattern is a little hard to predict. Indeed, perhaps the
most likely reaction, to take the case of a country experi
encing rapid inflation, would be toward accentuation,
not containment, of the exchange-rate decline, for the
domestic prices of exports could then increase behind
the screen of lower exchange costs to foreigners, import
costs would rise, wages would no doubt be raised even
further, and a new wage floor would have, in effect, been
built under the inflation already realized. And given a
commitment to flexible rates, it is hard to see where this
process would end, except in a sequence of competitive
devaluations. That is the kind of chaos experienced in
the thirties that the Bretton Woods system was specifi
cally designed to avoid.
The same downside rigidities and upward price drift
in our postwar economies that make adjustment more
difficult under fixed-exchange rates would, in my view,
make for progressive inflation, and successive waves of
exchange-rate depreciation from one country to the next,
if countries were trying to follow a flexible-rate system.
The one telling influence for relative price stability that
is universally recognized, if not respected, in today’s
world is that exerted by a country’s balance-of-payments
position. A flexible-rate system permits a country to cut
itself off from the international force of market compe
tition—the greatest defender the world now has for
protecting the stability of domestic monetary values.
Moreover, I am really puzzled over what could hap-

T h e B a l a n c e o f Pa y m e n t s

pen to the allocation of capital among and within
countries, over time, with exchange rates fluctuating
frequently and at times widely. Often, indeed, those dif
ferentials in interest rates and profits that help to bring
capital to its most productive uses would simply be off
set. The mere beginning of the capital flows themselves
might set off exchange-rate adjustments that would
bottle up any further flows, or if not, the exchange risks
might well exceed the potential interest or dividend
gain. And as I mentioned earlier, this could mean that
capital flows to any particular country would be roughly
limited to the amount of corresponding obligations that
one could obtain as a hedge in that same country or cur
rency. By contrast, a fixed-rate system, for all its imper
fections, does provide a reasonably stable set of bench
marks within which long-range capital commitments
can be planned and worked out in terms of calculated
profitability.
But I must not go on any longer with this list of the
shortcomings of a flexible-rate system as seen from
the theoretical side. What it all comes down to is that the
economic traffic among nations has become too vast and
too complex—including raw materials and processed
goods, services, and all forms and maturities of capital—
for me, at least, to be able to conceive of any satisfactory
way in which a system of fluctuating rates could really
determine the rates that people need to use from day to
day. Individual foreign exchange traders and bankers
would have an almost impossible task in groping for a

Sec o n d L e c t u r e

going rate that could take all of these conflicting influ
ences into account. Their task would be similar to,
though larger than, that of various individuals attempt
ing to come up with a firm figure for the wholesale or
retail price index of a country, for example, and then
being prepared to write contracts on the basis of an un
official pooling of each other’s estimates. I am very
much afraid that the rate for any currency against all
others would have to fluctuate so widely that the coun
try’s own trade would be throttled and its capital mis
directed. But that leads directly into the next area of
difficulty. Thus far we have been talking in terms of
the theory of what either system might attempt to do.
Now I want to turn for a little while to actual operating
considerations.
OPERATIONAL ISSUES
The hard fact is, I am convinced, that no country
able to control its own exchange rate will in practice
allow it to float. Even if flexible in form, the exchange
rate of any such country is going to be watched over
by the financial authorities of that country. That is to
say, even if a compelling theoretical case could be made
for "flexible” rates among countries—and you have seen
that I do not think it can— the same forces that have
given us the downward rigidities of prices, wages, and
employment in this postwar era would impel the govern
ment and central bank of any reasonably developed

T h e B a la n c e o f P a y m e n ts

country to try to control its own exchange rate. For the
exchange rate can be a powerful weapon. When settled
at a particular level for any country, it determines the
comparative costs of all the country’s imports and all of
its exports. If the entire system were to become un
hinged, few countries could as a matter of practical
politics stand by and let their rate against other cur
rencies be influenced by the intervention of the author
ities of the other countries without at least taking some
defensive action.
The practical answer to the natural wish of each
country to gain the apparent advantage of a slightly un
dervalued currency has had to be the kind of armed
truce provided by a fixed-rate system, which allows only
narrow margins for market-rate fluctuation in response
to shifting supply and demand conditions. All of that
is now given status through the par-value procedures of
the International Monetary Fund. And whenever the
price and wage structure of a country is persistently out
of line, the country may, of course, change its parity,
though even then only under the watchful and apprais
ing eyes of all other members of the Fund.
The alternative opportunity that would be opened by
a worldwide system of flexible rates would, I very much
fear, be a continuous invitation to economic warfare as
countries maneuvered their rates against each other—or
more charitably, influenced their own rates to reflect in
each case the immediate interest of the country con
cerned. There then would be no widely recognized,

Seco n d L ec tu r e

established rate levels, and no presumption of any obliga
tion to maintain rate stability. The advantage of being
able to sell abroad a little cheaper, without necessarily
lowering domestic prices, would always make “ just a
little more” depreciation attractive. Countries would
simply be unable to leave their rates alone. And without
such abstinence, whatever may have been claimed as the
theoretical advantages of a flexible-rate system would in
practice surely be dissipated, if not lost.
Every foreign exchange trader whom I know, includ
ing to be sure the traders of most of the central banks
of the leading countries, has at one time or another told
me in puzzlement that he has never been able to visualize
the operational arrangements that could prevail under
a flexible-rate system. Nor can I. None of us is able to
imagine that any private traders, or groups of traders,
could have the courage or capacity to make markets in
all currencies, or even only in the major currencies, with
out some benchmarks to guide them. That is one impel
ling reason why I think it inevitable that every central
bank will always have to be a factor of some importance
in the market trading of its own currency against others
—if not through active intervention, then at least
through the setting of a par value and buying or selling
at the outer limits of the agreed margin for variation
around that par value.
Moreover, without the underlying steadiness afforded
by an official presence (or some official participation) in
the spot market, I doubt that forward markets could

T h e B a l a n c e o f Pa y m e n t s

ever as a practical matter get started in any currencies—
except perhaps at discounts so large as to make the nom
inal markets meaningless, if not ridiculous. And yet the
existence of forward markets for virtually all currencies
is usually given as a prerequisite for the functioning of
a flexible-rate system in all of the theoretical discussions.
Nor can one forget the need for people to perform
the functions of dealers and traders. While the little
coterie of foreign exchange practitioners certainly does
not include all who might have talent for this occult
art, it is an open society which anyone with a trading
knack is welcome to enter. Talent in this field seems
always, in fact, to enjoy a sellers’ market. Yet it must be
significant that I have never met anyone who has
attained the competence of a seasoned trader who would
be prepared to continue in the business if, by some sleight
of hand, all parities were to be abandoned and the central
banks were barred from entering the markets in their
own currencies. Many, and I include myself, would
probably want to withdraw from trading activities even
under the sort of flexible-rate system in which the central
banks were allowed a role, so long as there were no parity
guidelines to get us into the right ball park. A t any rate,
so far as forward trading is concerned, if I had no parity
guidelines in the spot market, I would certainly not
want to be crunched between the pressures generated by
central banking giants in a free-for-all. They would be
pursuing aims and using tactics, in jockeying rates
against each other, that I simply could not interpret in

Sec o n d L e c t u r e

the minute-by-minute environment that makes a trad
ing market.
Mind you, I am not trying to confront Professor
Friedman with an organized strike of my fellow traders
in the foreign exchange markets of the world, but I
do submit that there surely would, even if all of my
other objections could be overcome, be no little recruit
ing problem in getting the trading desks capably manned
for a launching of his system.
SPECIAL SITU ATIO N OF U N ITED STATES
In addition to the considerations that make flexible
rates undesirable and unworkable for most of the world’s
more advanced countries, there are further problems
that would center on the United States. To be sure,
because of its dominating size in world trade, and be
cause of the widespread use of dollars as a “ currency of
convenience” for the international transactions of other
countries, the United States will inevitably have unique
problems as well as unique opportunities under any
system of international monetary arrangements. Its cen
tral role in the system of fixed-exchange rates has super
imposed upon the conventional balance-of-payments
problems ordinarily encountered by any industrialized
country a vulnerability to monetary disturbances or
pressures of many kinds from many sources outside the
United States. Some of the hazards and burdens related
to these pressures will be described shortly, when I try
to review with you some of the additional reforms that

T h e B a la n c e o f P ay m e n ts

are going to be needed in order to keep the present fixedrate system functioning effectively over the years and
decades ahead. At this point, though, I want to focus
on a number of added risks or strains that would be
imposed on the United States if the world were to switch
over to a flexible-rate system.
The most significant overall point to be recognized is
that the United States can never expect, as a practical
matter, to bring about at its own initiative any effective
change in its exchange rate vis-a-vis the other countries
of the world. This is a consequence of our size and the
world’s need for use of the dollar as a “ vehicle” currency.
For us to attempt to initiate a change in our rate against
other countries, under a fixed-rate system, or to expect
a change to develop to our advantage under a flexiblerate system, is to reverse the natural laws of gravity and
magnetism in the monetary system. Any change orig
inating in the United States will in turn be evaluated by
all other countries who will then, under either a fixedor flexible-rate system, take offsetting action to defend
themselves.
There is always the possibility, of course, that a
system-wide change in parities could be made simul
taneously against .gold under the present fixed-rate sys
tem, or that under a flexible-rate system the gold price
might fluctuate frequently—with uniform consequences
for all currencies in either case. Yet to change the gold
price for the system as a whole, under the fixed-rate
system, would undermine confidence for the future in

Seco n d L ec tu r e

the stabilizing central point of reference to which all
other elements of the system are hinged. It cannot be
considered by the responsible governments of the leading
countries. Under a flexible system, frequent changes in
the price of gold would only be another manifestation
of the built-in uncertainties which would weaken the
reliance that businessmen and bankers could then place
on the continuity of any monetary values.
Aside from a change in the price of gold—which is the
one way that might be theoretically considered for ex
erting a uniform change in the value of all currencies—
the interesting and relevant questions concern instead the
kinds of changes that might in fact occur in the ex
change rates among countries. Having already noted that
the United States cannot effectively maintain rates de
signed for its own advantage, vis-a-vis the outside world
as a whole, I should also stress that it nonetheless can be
the source of changes which will then subsequently be
more precisely tuned by the individual countries on the
other end of the pairing (i.e., between the dollar and
each of the other currencies).
The additional hazards for the United States under a
flexible-rate system come mainly from the fact that dol
lars are widely used a vehicle currency, and consequently
may be held in transactions balances in larger or smaller
amounts by the businessmen, investors, and bankers of
other countries. This inevitably means a special kind of
exposure for the United States, as well as giving us the
opportunity to serve as bankers for much of the world.

B a la n ce

Pa y m e n t s

We have certainly not yet worked out an easy accom
modation to the exposure aspect of this unavoidable
opportunity. But we have, under the fixed-rate system,
been able to minimize the purely speculative capital
flows, and to neutralize the impact of all such flows on
the reserve position of the United States—so long as the
causes are temporary. Where the causes are lasting, there
are consequences in long-term capital flows and adjust
ments that we are learning to recognize. But the over
riding point is this: whatever has been accomplished has
come about only because the central banks and the pri
vate markets have created new facilities within the
framework provided by a set of fixed parities among
the leading countries. And changes in those parities, as
in the cases of France, Germany, the Netherlands, and
Canada during the past decade, for example, have been
made as discrete steps, within a structure of assurances
that rates were going to be kept in place, once declared.
But the exposure that the United States must face
would be magnified under a flexible-rate system. As if
earlier conditions would not have made those risks grave
enough, the development in recent years of the so-called
Euro-dollar market, in which deposits of many billions
of dollars are held on the books of foreign banks, has
greatly enlarged the scale of such problems. I am not
even talking here about the ways in which shifts of these
Euro-dollar funds may affect the different methods
of accounting for our balance of payments. What m at
ters most for this discussion is that there now can be,

Second L ec tu re

in magnitudes much larger than anything experienced
in the past, massive movements into dollars from other
currencies, or out of dollars back into other currencies—
shifts that can amount to several billions of dollars
within a few days, or even hours. Under a system of
flexible rates, shifts of this kind (which have now be
come commonplace, and for which we now do have a
variety of effective cushioning devices to minimize or
neutralize possible disturbing side effects) would be ex
pected to work themselves out entirely in changes in the
exchange rate.
The consequences of these movements of volatile
funds could not, under a flexible-rate system, be cush
ioned by a network of forward-market transactions;
they are much too large and capricious for that. Instead,
they would have to be reflected in changes in the spot
rates themselves—changes that would inevitably have
the same effect as changes in the terms of trade between
the United States and all other countries. Thus, when
the net flow might be inward into additional dollars,
the exchange rate for the dollar would rapidly strengthen
against other currencies. American importers would find
that the prices of foreign goods had suddenly become
much cheaper for them, provided they could quickly
complete transactions through the immediate purchase
of the needed foreign currencies. And conversely, Amer
ican exporters would find a sudden decline in their sales
contracts. As soon as the flow into dollars ceased, quite
apart from the further effect of a reverse flow, the rate

Ba la n ce

Pa ym en ts

would necessarily decline, the importing fall back, the
exporting become more competitive, and if this same
whip-sawing had not occurred too often in the recent
past, the merchants engaged in foreign trade might try
to resume more normal business. But the costs in physical
adjustment—or in forward cover where that could pos
sibly be arranged—would place such a strain on the mer
chants engaged in foreign trade that only the hardiest
could survive at all.
To be sure, the enlarged potential for shifts into and
out of dollars that has been generated by the develop
ment of the Euro-dollar market only enlarges the scale
of problems that have often been recognized, in discus
sions of a flexible-rate system, when consideration was
given to the possible impact of short-term capital move
ments on changes in exchange rates. There is no ques
tion, in my mind, that the unusual exposure of the
United States to this type of hazard would have results
that could be completely disruptive to the orderly con
duct of our commercial trade.
The argument has been made at times that flexible
rates could work well for one or two industrialized
countries, so long as all of the other countries maintained
a fixed set of exchange-rate relationships. The most not
able case, of course, was that of Canada until 1962.
There it was often argued that movements of short-term
capital produced a change in the exchange rate against
the United States dollar that was equilibrating, helping
in turn to achieve a balance between inpayments and

S econd L ec tu re

outpayments without the need for extensive use of
Canada’s monetary reserves. While that history will long
be studied and re-studied, and argued and re-argued, my
own feeling is that even in that very special case, the
avoidance of reserve strains, as changes occurred in the
structural relations between Canada and the United
States, lulled Canada into a false sense of security. In the
end, it became necessary for Canada to establish a fixed
parity for its dollar. But by that time, the problems of
internal structural realignment then confronting Can
ada were much more difficult, at least as I see the situa
tion, than might have been the case if Canada had been
maintaining the trade and investment patterns con
sistent with a fixed exchange rate and intermittent
swings in foreign exchange reserves over the years. Even
so, the test is a poor one because, either through influ
encing the interest rates that affected capital flows be
tween our two countries, or by affecting the exchange
rate itself, the Canadian authorities were certainly not
standing aside, to permit the full range of possible rate
fluctuation.
It is this inevitability of intervention by the monetary
authorities of other countries, whenever they see capital
flows or trade moving significantly against them, that
makes the position of the United States so peculiarly
vulnerable under a flexible-rate system. And in turn,
the exchange-rate fluctuations that will be generated, not
only by market forces but by official intervention, can
certainly be expected, at least some of the time, to gen-

T h e B a l a n c e o f Pa y m e n t s

erate destabilizing additional movements of short-term
capital.
I am quite aware that this description of the extra
strains that a flexible-rate system would impose on the
United States runs counter to the picture described by
Professor Friedman in his testimony before the Joint
Economic Committee in November, 1963, when he im
plied that the United States could have been insulated
from what he called "enormous power [given] to offi
cials of foreign governments” by our balance-of-pay
ments deficits. I can certainly overlook his characteriza
tion of my own efforts at that time as a "frantic search
for expedients.” But I have never been able to under
stand how he could argue that we would have been
spared any balance-of-payments strain, and our domestic
economy kept completely insulated, if only our exchange
rate against all other countries had been free in those
circumstances to decline. For that was the period, you
will remember, when the first of a series of measures had
to be introduced to help check the tumbling outflow of
American capital, both long term and short term. The
momentum of those outflows, in a flexible-rate environ
ment, would undoubtedly have produced so sharp a
decline in our exchange rate against other countries that
American exporters would have had a field day. And I
have little doubt that the monetary authorities of other
countries, seeing an impending avalanche of American
goods would promptly have taken offsetting action to
adjust their own exchange rates and return the relative

S eco n d L e c t u r e

relations closer to the level that might have originally
prevailed. But even if I should be wrong in that pre
sumption, the result then actually would have been a
tremendous stimulus for American exports, a sharp rise
in the price of our imports, an all-around sequence of
other internal cost and wage increases, and the initiation
of internal inflationary pressures that would have quite
upset the relative price stability which was doing so
much, at that time, to help create the conditions for an
improvement of productivity and the orderly growth of
the American economy.
I realize, of course, that meditations of this kind on
the history of "what might have been,” are no more than
personal daydreams. But I do submit that they are at
least as credible, as descriptions of the potentials of a
flexible-rate system in 1963, as those which Professor
Friedman adduced in his testimony.
Moreover, having reflected on his characterization of
the impact that a flexible-rate system might have had in
these circumstances, I find it hard to reconcile the kinds
of advantages he claims in these circumstances with the
description that he gives of a flexible-rate system at other
times. For he has also said something else. In attempting
to defend a flexible system against the charge (which I
have certainly been making here) that exchange rates
would often fluctuate widely, with harmful conse
quences for the volume and the composition of trade,
he has asserted that the expected result should ordinarily
be a relatively stable pattern of rates. He wrote in his

B ala n ce

Paym ents

classic essay on "The Case for Flexible Exchange Rates”
that “ advocacy of flexible exchange rates is not equiv
alent to advocacy of unstable exchange rates. The ulti
mate objective is a world in which exchange rates, while
free to vary, are in fact highly stable.” It is that stability,
and its advantages as well as its shortcomings, which I
think best serves the interests of the United States—but
I do not see how Professor Friedman can have it both
ways.
A FATAL FLAW?
One advantage that I have in this debate is that I am
defending the essentials of the system that is already in
place. But, I can already hear the murmur, "What a
system!” And I must agree that it is far from perfect.
Nonetheless, I do have the comfort of knowing that it
does actually work. And I have the conviction, born of
some bias as well as some experience, that most of the
harsh and unsettling aspects of the working of this
system, as they have been most conspicuous most re
cently in the United Kingdom, are very largely manmade and can be man-corrected within the premises of
a fixed-rate system. All of us are learning through the
years what can be helpful in reducing the anguish while
enjoying the achievements which this system generates.
There is, however, one emerging problem in the func
tioning of a fixed-rate system that is so fundamental that
I must frankly concede that the system could disinte
grate if that problem is not carefully diagnosed and

Seco n d L e c t u r e

resolved within the next few years. I am not going to
wander off into the many other kinds of improvements
that could be visualized to make the system work more
effectively, nor will I try here to describe the further
implications for our own domestic economic policy in
the United States that are, at least to my eye, being so
clearly signaled by our present balance-of-payments
deficits. I do want to conclude with a few comments
and suggestions on this crucial fault which, if it widens
much further before it is corrected, could bring the
system down. Should that happen, I suspect the result
would be a world divided into smaller trading blocs,
enjoying the advantages of fixed exchange rates within
each bloc and the hardships of barter between blocs—
rather than any turn toward Professor Friedman’s flexible-rate system. But perhaps we can argue about that
next week.
What threatens to undermine the fixed-rate system
now is the fact that the supply of gold, the ultimate
reserve on which the system depends, cannot be expected
to grow rapidly enough in world monetary reserves to
provide the primary liquidity that the countries will
need for making settlements among themselves at fixed
rates of exchange. Up until now, it has been possible to
build upon gold another kind of primary reserve, first
in the form of British pounds sterling, then for more
limited areas the French franc, the Dutch guilder, and
others, and then on a worldwide scale, the United States
dollar. In principle, this is the way— the supplying of

B a la n ce

Pa ym en ts

usable national currencies—in which an adequate answer
can be found for meeting the world’s growing reserve
needs as the world’s scale of economic activities expands.
The reason that this need cannot simply be solved by
a change in the price of gold is very similar to the reason
why the world cannot effectively function on a flexiblerate system. In my view, the premise for an effective
functioning of a market economy, guided by changes in
relative prices, is that the numeraire must remain rela
tively constant. That is why, as I see it, we need a fixedrate system. For in the developed world, at any rate, we
simply face an unresolvable problem in which there are
more unknowns than there are variables, once we intro
duce fluctuations in the unit of measurement itself
which may be so large or so capricious that a stable reso
lution of all the forces cannot be found. The unit of
measurement must remain reasonably constant (in the
sense of predictability) in order that all other elements
can move up and down in a measurable, and thus mean
ingful, way. The alternatives, for the world as a whole—
though individual countries may go otherwise so long
as the leading countries do not—is surely either barter
or exchange control, introducing another kind of cer
tainty into the measurement of relationships. Without
the temerity to attempt a mathematical demonstration
of a point that most eminent mathematical economists ap
parently do not accept, I can at least suggest the need for
further scientific testing of a hypothesis that the whole
world of finance and trade now lives by.

Secon d Lec tu r e

It is just as misleading, in my view, to think that the
external value of a currency can be determined in some
detached way by fluctuating supply and demand as it is
to think that the price of gold can be varied frequently
without weakening its usefulness as a standard of value.
Instead, just as the supply of money for individual na
tions must be man-made, within a framework of ar
rangements that aims at maintaining stable values, so the
supply of gold at the base of the world’s monetary sys
tem must be augmented by new arrangements. In effect,
while preserving the fixed price, those arrangements
would add a common currency to the gold supply,
through the combined action and commitment of a
number of countries, acting together. The world would
not then, henceforth, have to rely primarily on the
issuance of one currency to provide the acceptable sup
plement to gold.
The need, in effect, in order to preserve the $35
price as the kingpin of the whole structure of fixed ex
change rates, while providing for regular and controlled
increases in the supply of gold for reserve purposes, is to
find an effective and convincing method for creating the
equivalent of additional gold. That is the object of the
international monetary discussions which have now been
under way for more than three years. This is no simple
matter. Countries will have to make lasting commit
ments to contribute to the creation of, and to accept and
hold and use, a new kind of international money that
rests upon the joint and several obligations of all of them.

T h e B a la n c e o f P a y m e n ts

Such obligations cannot be undertaken lightly. They de
mand from each country a careful search of the risk
that its own direct interests may at some time be jeopard
ized by these arrangements. It is small wonder, then, that
much time has been consumed in working toward agreed
principles of operation, long after the stage has been
reached of preliminary agreement on the need for plan
ning of this kind.1
I feel certain that the members of the International
Monetary Fund will, in what still may be many months
but will be less than many years, find a way to agree on
prudent methods for creating what Secretary [Fred
erick L.] Deming, [Under Secretary of the Treasury
for Monetary Affairs] has called, in a spirit of whole
some compromise, a "Drawing Unit Reserve Asset.” I
would suggest that his "D U R A ” be re-christened
"DURA-GOLD.” It will be an asset that central banks
can readily accept and contentedly hold— transferable,
usable for obtaining the dollars needed for trading in the
foreign exchange markets, and counted as part of each
country’s primary monetary reserves.
My certainty that a way will be found springs from
my conviction that most of the countries of the world
want to keep and improve the present environment for
a multilateral expansion of trade and investment—the
environment among the countries on fixed exchange
rates which has made possible in this past decade the
greatest expansion of capacity and output that has been
achieved by any group of countries at any time. They

Second L ec tu re

will be impelled to find an answer because they cannot
accept either of the alternatives— the anarchy of an en
tire world on flexible exchange rates, or (and this would
be the more probable) the protectionism and economic
autarchy of the sort of currency blocs that prevailed
in the 1930s.

REBUTTALS

M ILTON FRIEDM AN
Dr. Roosa regards as a key disadvantage of free ex
change rates the likelihood that they will exert a damp
ening influence on world trade and will promote bi
lateralism and autarchy.
I regard as a key advantage of free exchange rates
the likelihood that they will lead to freer world trade,
will promote a dismantling of exchange controls and
import quotas and a reduction of tariffs.
How can two knowledgeable men reach such dia
metrically opposed conclusions on this, as on other as
pects, of free versus fixed exchange rates?
Needless to say, I believe that my conclusions are
correct. So I shall look for defects in Dr. Roosa’s argu
ments, not my own, to explain the conflict. After all,
we both believe in division of labor according to com
parative advantage, so I can surely leave the defects
in my argument to his capable hands.
The explanation, I believe, is that Dr. Roosa applies
a double standard. The advantages that he cites for
fixed rates are valid and important advantages of a
real gold standard, of a unified currency system in

T h e B a la n c e o f P ay m e n ts

which there are no central banks to interfere with the
effects of balance-of-payments flows on the quantity of
money. Such a standard does weld the world together,
does promote international trade, does impose an effec
tive and reasonably mild discipline on internal policies.
If I had to choose between such a standard and a
system of national currencies linked by flexible exchange
rates, I might very well choose the gold standard, de
spite the waste of resources involved in digging out gold
in one part of the world to bury it in another.
But a real gold standard bears as little relation to
the existing system of pegged exchange rates, if I may
quote my ex-chancellor, as football does to a college
education. Yet, Dr. Roosa discusses the existing system
as if it were a real gold standard.
Do pegged rates really provide international traders
with certainty when they do not know whether they
will be able to convert their exchange a year hence? O r
whether they will be able to get import permits? O r
whether they will be faced by different pegged rates?
Dr. Roosa recognizes the problem but then resolves
it by bland faith. "What about the . . . balance-ofpayments adjustment mechanism?” he asks. " I see no
reason conceptually,” he answers, "why . . . adjust
ments cannot. . . be achieved through variations which
merely affect the speed of an economy’s forward
motion.”
"I am sure,” he goes on, "there can also be ways o f
using balance-of-payment signals constructively to

R ebuttals

meet, in combination, both external and internal needs.”
Perhaps so, but do we know these ways? There is
not the slightest sign in his paper that we do. I rubbed
my eyes as I read all of this. Do we really have an
interest-equalization tax? And is it really being pro
posed to double it? Is there a "voluntary” credit re
straint program? Is there a "voluntary” program to
restrict foreign investment? Are there oil import
quotas? Has the tourist duty-free allowance been
reduced?
Have Japan and Hong Kong been pressured to im
pose voluntary export quotas? Does Britain have ex
change control? Has Germany taken measures to re
strict the import of capital?
And so on and on, or are all of these only figments
of my overheated imagination induced by my inability
to rejoice in the expansion of multilateral world trade
with ever-fewer restrictions and impediments?
In Dr. Roosa’s story, the discipline of fixed rates is
always in the direction of desirable internal policies.
Has Germany not existed these past years? Is it my
imagination that fixed rates forced Germany to inflate?
In short, when Dr. Roosa considers fixed exchange
rates, he implicitly considers not our actual system but
a glittering gold man with only an occasional side
glance at the reality it conceals.
When he comes to free exchange rates, he also con
siders not the system as it would actually operate, com
pared with the alternatives to it actually available, but

T h e B alance o r Paym en ts

a hypothetical system. This time, however, the hypo
thetical system is not a gold man but a straw man, a
scarecrow of shreds and patches to frighten children
with. There is none of that bland faith that somehow
methods will be found to cope with possible weaknesses.
No, this time everything is for the worst in the worst
of all possible worlds.
Free rates, Roosa apparently believes, are subject to
Murphy’s law, which I am sure you all know. It goes,
“ If anything can go wrong, it will.” Let me illustrate,
and I quote. "Prices in international trade and the costs
of doing international business of any kind would . . .
almost inevitably become higher under a flexible-rate
system” because of "the added element of risk.”
What added element of risk? Compared with a real
gold standard, yes. Compared with a system of pegged
rates, held together by exchange controls, import
quotas, capital restrictions, and occasional changes in
pegged rates? No. Or, at the very least, unproved.
Even more important, as I emphasized in my paper,
under a free rate system some currency or currencies
would undoubtedly become widely used for denominat
ing international transactions. There need not be even
a presumption of greater risk in such a case.
Or, to look another of the monsters created by D r.
Roosa’s fertile imagination in the face, he conjures up
the possibility that free exchange rates would work
towards “ accentuation, not containment” of rapid

R ebuttals

inflation. Dr. Roosa envisions that a decline in the ex
change rate would mean that the "domestic price of
exports could then increase behind the screen of lower
exchange costs to foreigners, import costs would rise,
wages would no doubt be raised even further,” and
so on.
This is simply wrong, if free exchange rates are
compared with the correct alternative. Suppose the
exchange rate is kept pegged despite the inflationary
pressure. How can it be kept pegged? Only by con
trolling imports and exports. But this will make domes
tic prices of the controlled items differ from their
foreign prices. Given the same volume of trade, the
domestic prices will be precisely the same under either
system.
How does a fixed-rate system provide the country
with more resources with which to get more goods than
it otherwise could get? Implicitly, again, what Dr.
Roosa is doing is to compare the flexible-rate system,
not with the actual alternative, but with some hypo
thetical ideal world in which there is neither exchange
rate change nor inflationary pressure.
Dr. Roosa reserves his finest flights of the imagination
for the horrors facing exchange traders in spot and for
ward markets under free exchange rates. I have heard
this story before. It is exactly what people said would
happen, as I pointed out in my paper, in the housing
market before rent control was removed, what people
said would happen in Germany in 1948 before price

T h e B a la n c e o f P a y m e n ts

controls were removed. I have little doubt that if we
could disinter the files of the Office of Price Admin
istration for 1944 and 1945 we would find a memo
explaining the horrible chaos that would follow if price
controls and rationing were suddenly removed, pro
ducing large and unpredictable fluctuations in prices.
To go farther afield, I am sure that the Gosplan files
would reveal dozens of similar memos explaining why
a free market could not possibly work under the spec
ial conditions of the Soviet Union.
Though Dr. Roosa apparently has unlimited faith
in the ability of government bureaucrats to devise sen
sitive and effective substitutes for market adjustments,
he seems to have very little faith in the ability of the
market itself to adjust. Like every good administrator,
he knows the ins and outs of the present cumbrous
system. He knows it works after a fashion and he can
not conceive that there is any other way to run it.
Let me assure him that he will develop a similar ex
pertise in a free exchange rate world and, if only he
will let himself go, he would find himself speculating
less and enjoying it more. I might add that it’s rather
curious to find a partner in a major financial house
standing here before us and extolling the virtues
of government bureaucracy, explaining to us how
civil servants can control things fine, while an acad
emician argues that you can’t trust those civil servants,
you’d better leave it to the private market. But all sorts
of strange things happen in this world.

R ebu ttals

In any event, I must repeat, the world that Dr. Roosa
conjures up, the world of bilateral balancing with er
ratic movements of volatile funds and "fluctuations in
the unit of measurement', which may be . . . large . . .
or capricious,” is a figment of his imagination.
We have had extensive experience with free rates in
many countries, including the United States. Let Dr.
Roosa show us any empirical case where his fears have
been realized, except where there have been initially
unstable internal policies. He is a victim of what, in my
earlier paper, I called the Arizona effect.
If countries separately follow stable internal policies,
exchange rates, while free to move, will be highly
stable. Stability is not rigidity.
Finally, I come to Dr. Roosa’s comment about the
special situation in the United States: "The most sig
nificant overall point to be recognized,” he writes, “ is
that the United States can never expect, as a practical
matter, to bring about on its own initiative any effec
tive change in its exchange rate vis-a-vis the other
countries of the world.”
Who ever said it could or should? Certainly not I. I
have apparently failed completely to convey to Dr.
Roosa my position. So perhaps there is some hope we
can still get together.
Would he agree that the United States, could, by it
self, do the following four things?
One— Get rid of all elements of exchange control,

T h e B a la n ce o f Pa y m en ts

direct and indirect.
Two—Announce to the world that it will not try to
peg the rate of exchange between the dollar and cur
rencies of other countries.
Three—Announce that it will sell gold to all and
sundry at $35 an ounce until its' supply runs out or,
alternatively, that it will simply stop selling or buying
gold.
Four—Proceed to follow a stable internal monetary
and fiscal policy designed to foster a reasonably con
stant price level. (I have my own pet scheme for doing
this but, for the present purposes, the broader state
ment will suffice.)
I find it hard to believe there can be any doubt that
the United States by itself can do those four things.
Now if it did those four things and other countries
chose to tie their currencies to the dollar, fine and
dandy. We would then be on the way to a unified cur
rency and Dr. Roosa and I would both be happy.
If, under those circumstances, other countries chose
to let their currencies float relative to the dollar, that’s
also fine and dandy.
If we really followed my Points One and Four, that
is, if we really eliminated exchange controls and fol
lowed a stable internal policy, we would be on the way
to the widespread use of the dollar as an international
vehicle currency with other countries able to pursue
their own internal policies.
If we did not follow my Point Four, that is if we

R ebuttals

did not follow a stable internal policy, some other cur
rency would develop as an international vehicle cur
rency and the rest of the world would at least be insu
lated against our monetary mistakes and we against
theirs.
Under any of these circumstances, those able men
who man the emergency telephones in the great central
banks and who spend endless hours trying to devise in
genious means whereby everybody can borrow from
everybody else without anybody being committed to
lend to anyone, those able men could be released to do
some truly productive work.

RO BERT V. ROOSA
We both have the same problem. We both think that
the other has so glamorized his own world that it has
lost contact with the kinds of reality that each of us
believes he represents. And I am afraid that unless we
can bring these contrasts, as Milton has presented them,
down to some closer approximation of an earthy reality
we are in danger of spending the evening on opposite
sides of a revolving door walking through in opposite
directions.
I think we do have to find whether there isn’t some
way of bringing both of our presentations back to the
mainland, back from what he regards, and I regard, as
islands of abstraction so remote as to be unrelated to
the real potential that the countries of the world now
confront.
I think perhaps we will find the beginnings of an
answer in the questions that he asks toward the close.
In themselves, they help to illustrate the ambitious
striving for the ideal which, I believe, characterizes his
efforts and colors his judgment as to the potential for
achievement.

T h e B a l a n c e o f Pa y m e n t s

He asks, just for example, "can the United States
get rid of exchange controls?” I’ll put it the other way
around: Will the United States ever be able to live with
out its own government having a sense of involvement
and responsibility for the rate of exchange between its
currency and that of other currencies in the world?
The United States, as I have already explained, isn’t
going to be able to do very much about its own rate
against other currencies, not on its own initiative. But
the other countries of the world, and Professor Fried
man has indicated he would welcome the kind of de
velopment in which this might occur, the other coun
tries of the world are going to have to react continually
to the pattern of influences that is generated by condi
tions in the United States, in setting their exchange rates
against the dollar.
And we, here, are certainly never going to be in a
position to proceed on the premise that he offers by
implication, the premise that we will simply take the
exchange rate offered in an environment in which there
are no ground rules, in which every other country is
free by definition to proceed to the advantages" of com
petitive depreciation, to beggar each other’s neighbor,
as was done through the thirties, an experience that is
all too searing still in our memories to forget and which
I do commend to him when he asks whether there is any
experience with the sort of approach that he is recom
mending.
The end result then was a deterioration into cur-

R ebu ttals

rency blocs which became competitive and abrasive
to the point of producing bilateralism and the shrinking
of trade patterns. I am sure these are inherent when
there are no clear ground rules by which countries are
guided, and when changes are made in exchange rates in
response not only to the trade flow but also to the capi
tal flows that are of overriding importance (and will
be) in the determination of the exchange rate of this
or any other country under his system.
Or, look at the other question. Can I give him any
assurance that the United States will follow a consistent
policy aimed at internal stability? I can’t.
I can assure him, as I am sure all of you could, that
the United States is committed and every free country in
the world is committed to a set of objectives—and that
their objectives limit Professor Friedman’s conditions of
pure and idealistic freedom with, a number of premises.
Neither the United States government, nor any gov
ernment, is going to withdraw from responsibility for
the maintenance of employment, for providing mini
mum levels of wages and the acceptance of a downward
rigidity both in prices and in wages. These are parts of
the real world that I had in mind when I said there is a
risk that in trying both to formulate his questions and to
imply his answers he is forgetting where it is we live and
what it is we are going to have to find a solution for.
None of the solutions are going to be simple and none
are going to be in any sense capable of providing the

T h e B a l a n c e o f Pa y m e n t s

smooth perfection that either of us would like to ascribe
to our own systems.
Throughout my own effort I indicated step-by-step
that I am supporting the existing system, an admittedly
imperfect system, one whose imperfections are clear fo r
all to see because it is meeting the test of practical and
continuous current application and subject to the con
straints and the premises of the kind I have just men
tioned and with which we are bound to live. Whatever
any of us would like to do about recreating the image
of mankind for the sake of providing a simpler setting
for our own formulation of economic arrangements, we
don’t really have that choice.
As a matter of fact, I would like to turn in the other
direction and ask him a number of questions that were
raised by his paper, in the hope that as we work these
over back and forth we may be able to get a little
closer, at least, to a clearer understanding of our differ
ences. But I hope we can see them as they become im
portant in the real world, and not this time through a
fanciful creation of his imagination.
I really wonder, first of all, Milton, why you used
the description you did throughout your paper of the
patterns of central bank action that occur in foreign
exchange markets. I am sure you know that operations
in foreign exchange markets have never been conducted
in that way. I realize that economists don’t have to know
ail of the mechanics of every process with which they
are dealing. We can generalize about big issues in the

R ebuttals

same way that we can drive a car without having to
know how the motor works. But we do have to know,
and be able to tell others, which end to put in water and
which end to put in fuel.
Yet the whole analysis of the way in which foreign
exchange markets function as set forth throughout your
paper, Milton, is simply completely out of context with
the way in which these markets do operate. To have a
starting point, we ought to begin at least with recogni
tion of the fact that foreign exchange rates are main
tained by the various central banks of the world by trad
ing in their own currency against dollars. Each main
tains its own upper and its own lower limit. Each,
within those limits, most of the time maintains a rate
which is close to parity and never even reaches either
outer limit. The central banks are operating under the
discipline of a set of guidelines which they realize, in
self-defense, they must respect. Otherwise, the whole
system, if there weren't these guidelines, would break
down into a sequence of competitive depreciations which
would create the conditions of bilateralism. There
would be an insulation of the frontiers of each country
from the trading potentials of others. That seems to
me inherent in the nature of an unstable exchange-rate
system. (By the way, if I were, in my language, to trans
cribe the terminology that describes this debate, it
would not be the romantically appealing "free versus
fixed” exchange rates. Quite the contrary, what we are
talking about is "unstable” as opposed to "stable” ex-

T h e B a la n c e o f P a y m e n ts

change rates and the conditions which either of these
make possible for the maintenance and expansion of
trade and capital movements in the world.)
Now, my second question is: Why is it that he as
sumes at one part of his analysis that there can be com
plete elasticity in domestic supply and demand to make
adjustment to rate changes instantaneous within an econ
omy? Then elsewhere he assumes that a rate change will,
in fact, completely insulate the economy as it exists from
the need for any internal change?
In 'Newsweek one week ago, Professor Friedman wrote
that the way in which to solve the whole balance-ofpayments problem is just to let the rate move, nothing
else will have to happen. That would, he said, free the
authorities within every country to maintain whatever
domestic policy they wish to pursue.
What sort of an adjustment process is this, other
than an assumption that the rate would so change as to
preserve at all times the status quo?
And if that is the way the rate is going to function,
what it really means is the building of a wall around
every country; this is going to prevent those healthful
corrective adjustments that he has already identified in
beginning the discussion tonight.
And then, a third question. Why does he say at one
point in the discussion that rate fluctuations alone can
remove adjustment problems, then at another point
claim that the ideal result, the world we want to have,
is one in which, although rates might be free to move,

R ebuttals

they are in fact stable? How can he have it in this re
spect both ways?
Or a fourth question. If stable rates are the ideal,
what is introduced by having the potential for flexi
bility? What do you get except more uncertainty, es
pecially among those countries in a real world which
will in each successive episode of internal domestic diffi
culty find a way, if they aren’t limited by guidelines of
norms and behavior, to depreciate their rates, one after
another, to gain a momentary trade advantage? Ours is
a world of red-blooded competition. In that kind of
world, the existence of a fluctuating rate system is in
evitably going to mean a wider band of uncertainty,
with all the complications from it that I have already
mentioned.
And then from this how can he go on to say that he
is going to bury decently the view that speculation un
der floating rates can be destabilizing? He says actually
there has never been a study that showed this destabi
lizing effect. Now I don’t pretend to read all that I
should, but the Aliber work in the Yale Economic
Essays, about five years ago certainly shows that in the
case of all the European countries after World War I,
that is, all those which Aliber analyzed, under the con
ditions then prevailing, the effect of fluctuating rates
was to create a speculative aggravation, a sequence of de
stabilizing influences.
And then my fifth question. Why did he leave out of
his paper, and also leave out of what he has just been

T h e B a l a n c e o f Pa y m e n t s

saying, all analysis of capital movements? Any part of
his exposition, if it holds at all, can be turned upside
down if there were to be the kind of capital movement
which would weaken or strengthen an exchange rate
and in turn create opportunities for exporters or im
porters in a given country. This would lead, in turn, just
because the capital movement had occurred, to a decisive
change in the attractiveness within a given country o f
one kind of export, one kind of production, one kind of
import or another.
How can you say, if the rate is free to move and if
you expect it to move, that you are going to keep the
domestic economy completely insulated and free to
follow its own program regardless of what goes on out
side? Particularly if you take into account the fact that
trade and capital movements both have to be recon
ciled in the given exchange-rate system. There has to
be a compromise. There’s no perfect solution.
The nearest approximation or series of successive ap
proximations we have been able to devise is that coun
tries will pursue domestic policies under the armed truce
that prevails with a fixed-rate system. We can’t deny
them the right to pursue domestic policies. They are just
determined to do it, whether we legislate it in or out.
They will pursue domestic policies that in their own
way are bound to be affected by the reactions from any
change in exchange rates produced either by capital
movements or the resulting changes in trade flow.
Another question, my sixth, getting back to the disci

R ebuttals

pline that he talked about. How can he assert that, as
concerns the system I’ve been trying to defend, all ad
justment under this system occurs only with controls?
Certainly there have bqen a lot of controls. This system
is imperfect. At the present stage it is certainly doing
much less well than at many other times in the past.
But a good many effective adjustments have been ac
complished too.
What is going to change the nature of American
readiness to take the kinds of steps necessary to ac
complish an effective adjustment if the exchange rate is
free, as against the present discipline of a fixed-rate
system? To be sure, the discipline has led us into some
strange and wonderful things, but it has also persistently
been a dominating influence in guiding us toward the
same decisions that we should seek for domestic reasons,
the same emphasis. We fail, as human beings do, to do
the perfect job, but the same emphasis remains on the
need for price stability, for productivity improvement,
and so on.
The present system is geared to the assumption that
there will be imperfections all the way around. But we
will use reserves to provide time for correction. We
won’t have to have the sort of instantaneous shake of
a system from one side to the other, as a rate change
occurs, with one industry flooded with demand one
day and drought the next, because capital movements
have turned the exchanges and altered the exchange rate.
Finally, my seventh question, and this only repeats

T h e B a l a n c e o f Pa y m e n t s

a point that I have woven through many of these other
questions, how can he under a system of flexible rates,
if he wants to live in a real world, avoid the risk, the
exposure to competitive exchange depreciation, the re
lated reinforcing of potential inflationary developments,
and the serious risk of misallocation of capital around
the world?
PROFESSOR G. W ARREN N U T T E R , University
of Virginia, coordinator of the Rational Debate:
It is the desire of both participants to have a free
exchange insofar as we have time left within our first
hour, so I am going to bow to this and allow the de
baters to talk to each other for the next 15 minutes.
PROFESSOR FRIEDM AN: Let me take up first
the problem of competitive depreciation. This is really
a false issue. I don’t know what color herring but it’s a
herring.
The experience to which Dr. Roosa refers is the ex
perience of the 1930s. Now, we need to have historical
perspective. First of all, it’s important to know that the
1930s were a product of fixed rates, not of floating
rates. The United States and the other countries of the
world had a fixed rate in 1929. When the United States
embarked on deflation and proceeded to reduce its
money stock, the rest of the world was forced into a
major catastrophe. So the origin of that move was fixed
rates, not flexible rates.
It was only when countries got off the fixed rates
that improvements were made. When Britain went off

R ebuttals

gold in 1931, she was able to break the link with our
contraction.
It is worth noting that the countries that stayed on
the fixed-rate system were the ones that saw the depres
sion come to an end latest, the United States in 1933,
France in 1934, and so on. On the other side, the coun
tries that broke the link with fixed rates and went on
to floating or flexible rates were the ones that got out
of the depression first.
In the second place, a system under which you have
adjustable pegs is a very different system from a sys
tem under which you have floating rates.
In the third place, the 1930s saw widespread un
employment in many countries. Countries were seek
ing to control their exchange rates by governmental
means and to depreciate against one another. In the
present situation you have relatively full employmfent,
and I trust you will continue to. What does a coun
try have to do to depreciate competitively against
the U.S. dollar? It has to provide us with interest-free
loans.
We are utter fools if we don’t say to any country
in the world, if you want to depreciate competitively,
fine and dandy. What that means is that you are pre
pared to sell us your goods at a lower price than that at
which we could otherwise buy them. There’s nothing
wrong with that. There is no mileage to be gained from
competitive depreciation by anybody in a world of
reasonably full employment.

T h e B a la n c e o f P a y m e n ts

This is, again, I think, something that Dr. Roosa has
dreamed up to scare children with.
But let me go to the more important analytical issue
that Bob raises. Throughout his discussion, he confuses
different kinds of adjustments. I ask him what his ad
justment mechanism is and he says to me, “ Well, how
can you say that under a floating rate you have instan
taneous adjustment? Aren’t you going to have a coun
try with goods flooding in one day and flooding out the
next day? Aren’t capital movements going to cause all
of the farmers of the country to plant twice as many
acres this month and next month to go out and dig
them up?”
The answer is, I believe, that you must distinguish
three kinds of adjustments which are quite different:
adjustments to monetary disturbances; adjustments in
the international sector itself; adjustments in the rest
of the economy.
In the first place, there are those cases in which ex
ternal disturbances represent monetary changes abroad
or at home, in which there is no need for “ real” ad
justments, for adjustments in the flows of goods and
services. If a foreign country, to take an extreme case,
doubles its price level, then a halving of the exchange rate
means that no adjustment is required by either country
because relative prices are left unchanged.
In such a case, rate fluctuations can remove adjust
ment problems because they make it unnecessary to make
real adjustments in response to monetary disturbances.

R ebu ttals

The simplest example in modern times is Germany.
Germany in the past ten years had to make real ad
justments. She had to export capital she didn’t intend
to export. She had to change interest rates internally
she otherwise wouldn’t have had to change because, with
a fixed-rate system, Germany was being affected by the
monetary inflation in the rest of the world.
Point number one therefore is: Insofar as exchangerate adjustments offset monetary disturbances in differ
ent countries, the exchange-rate movements eliminate
the necessity for making physical adjustments. In those
cases they do make adjustment painless.
Second, there are external disturbances which re
flect changes in real conditions of trade— the fact that
some country has developed a new way of producing
automobiles which makes them more competitive with
respect to us; or that there is an increase in the world
demand for copper. These real changes do require real
adjustment in an economy.
But they require it only in certain sectors. They re
quire it in the sectors which are affected by these changes.
Under a fixed-rate system the problem is that, if such
a change occurs, it requires the adjustment in one sec
tor to be made by an adjustment in the rest of the
economy. To illustrate, let us suppose that in the U.S.
there is, suddenly, a great increase in the demand for
copper. Under a fixed-rate system, this disturbs our
balance of payments. We tend to have a deficit. To
adjust to this, we would have to lower our whole price

T h e B a la n c e o f Pa y m en t s

level just to permit a rise in the relative price of copper.
Under a floating-rate system, the exchange rate makes
that part of the adjustment which would otherwise have
to be made by the economy as a whole and the relative
prices of the particular commodities that are involved
make that part of the adjustment which corresponds to
a real adjustment.
The virtue of a floating-exchange-rate system is not
that it eliminates the need for all adjustments. The virtue
is rather that it minimizes the problem of adjustment
because it makes it unnecessary to make real adjust
ments to monetary changes and to change relative prices
through changing the total price level. O f the three types
of adjustments I listed initially, it eliminates the neces
sity for two.
Now, I may say that covers quite a number of Bob’s
comments. I think most of his comments and half of his
questions derive from not distinguishing among these
three kinds of adjustments.
With respect to capital movements, I don’t intend to
leave out capital movements at all. We want a world in
which you have free capital movements. We want to
welcome and encourage capital movements. Capital
movements do require real adjustments. Dr. Roosa some
times talks as if under a fixed-rate system you could have
a capital movement without an adjustment. That isn’t so.
If Canada is going to import capital, in the real sense,
then there has to be a difference between its exports and
its imports.

R ebuttals

If the U.S. is going to export capital, there has to be
an excess of exports over imports to finance it.
O f course, capital movements have to affect the real
balance of payments. The question is: What is the mech
anism of adjustment?
Under a fixed-rate system the mechanism of adjust
ment must be differential price movements. The country
that seeks to export capital must have its prices decline
relative to countries that want to import capital.
Under a floating-rate system the adjustments can take
place much more easily because the exchange rate will
then move and this will tend to create the balance-oftrade difference required to accommodate the capital
movement.
Let me just make one more comment—about Dr.
Roosa’s differentiation between my description of ex
change markets and the way they really work. He says
that as a day-to-day matter foreign exchange rates are
maintained by central banks with upper and lower
limits, that my description of foreign banks having a
requirement to maintain an upper limit and the U.S.
having the requirement to maintain a lower limit is not
valid. In an operational way, of course, he’s right. He
knows much more about the day-to-day operations than
I do. But on the basic level of principle I do not believe
that my description is invalid.
The point is that foreign countries are willing to sell
and buy within these ranges and do what I described as

T h e B a la n c e o f P ay m e n ts

our job because we have been willing to trade their extra
dollars for gold. At that point, they are, in effect, acting
as our agents and it has been our willingness to provide
them with gold when they felt they had acquired too
many dollars that has made them willing to appear to
act in the market to support our rate at the bottom.
That,really is the mechanics of the way we have operated
our side of the deal.
DR. ROOSA: Thanks, Milton. I do think that it is
important to repeat again that our aims, as they persist
through all of this, are clearly the same. Despite the
variations that are introduced for the purpose of making
one point or another, we both see that the world is going
to function more effectively on an exchange-rate system
which provides, most of the time, stable relations among
countries so that there is a basis for measuring what the
impact is of a change within any country.
In technology, we both hope to produce a result indi
cating whether or not the automobile now producible
at cheaper prices in the United States is going to be
exported or not.
If, whenever there were a technological improvement
on the part of this or any other country, all that hap
pened was that exchange rates instantly moved against
the country then in a position to move into another
foreign market, what would have been accomplished, of
course, through a flexible-rate system would be the kind
of isolation that would prevent —
PROFESSOR FRIEDM AN: No, Bob. No, no, Bob.

R ebuttals

What would move the rate? Only that they were selling
the goods.
DR. ROOSA: Well, if they find that they have a
competitive advantage and if the surplus that they are
acquiring is not providing them with a stronger rate,
then their competitive advantage would go on providing
a continued opportunity for them to export for a while.
If, finally, the rate moves, the net result of their having
a stronger rate is—
PROFESSOR FRIEDM AN: No, no, no, no.
DR. ROOSA: — for them eventually to import. Well,
all right.
PROFESSOR FRIEDM AN: Your mechanics are
wrong.
DR. ROOSA: Let me skip over that one and we’ll
come back, because I shouldn’t have digressed too far
into this one because this is only the starter and I do
want to also get back to your point about the nightmare
of competitive depreciation, to which this leads and we
can come back to the technical point about the produc
tivity change in a minute.
First of all, the study I referred to covered the 1920s,
after World War I, and not the 1930s, although I did, in
other connections, refer to the 1930s as well. Anything is
going to be a result of the system that has been prevail
ing so we can’t say that just because flexible rates during
their period didn’t work that only shows that fixed rates
are a failure. The net result is that after flexible rates
were allowed to function for a while, were supposed to

T h e B a l a n c e o f Pa y m e n t s

be taking over, every country in its own wisdom and
judgment returned to a fixed-rate system because they
needed the protection against mutual combativeness that
is embodied in the drive produced for competitive exchange-rate depreciation.
So I can’t agree, Milton, that this is something to be
dismissed only as my private nightmare. I have a lot of
them. But not this one.
And on another point—I know that this isn’t exactly
what you mean, but it is an illusion for you to say that
the end result of all this is a satisfactory solution because
we just get a lot of interest-free loans from other coun
tries. They obviously aren’t interest-free in any case.
Foreigners do hold interest-earning money market assets
here. But what does actually produce the difference
leads me into my comment on capital movements. It is
not merely that we were prepared to pay out gold. It is
the fact that the dollar is the transactions currency, the
balance-holding currency and, indeed, the interest-earning currency which provides another part of this total
system.
What is part of the real world in which we live today
and is a byproduct of this interest-earning dollar is that
we also have a Euro-dollar market of a $15 billion
dimension. In this Euro-dollar market there can be
changes that will, under a free-rate system, affect the
exchange rate of the dollar without making one iota of
difference in the movement of goods in or out of the
United States in relation to those transactions. Now,

R ebuttals

this is a fact of life. This is no one’s creation of imagina
tion, Milton, and we have to live with some of the con
ditions of the world as they have evolved.
If we were to inject a free-rate system amid this
existing set of arrangements, arrangements that have
evolved out of the convenience, necessity, and require
ments of responses to free impulses, if not fully free
markets, we would find that shif ts in this kind of capital
movement would be producing changes in the exchange
rate for the dollar that would have repercussions on the
whole economy. These repercussions you can’t escape
if the terms of trade for American products are changed
by movement into or out of Euro-dollars on the part of
the holders of other currencies.
Now, to get to your point on adjustments, Milton.
I have to confess, I don’t know whether I was daydream
ing or what went wrong, but I only heard two of the
three kinds you promised to describe. I do think what
you have said is important and it has been a real contri
bution to get the clarification of these distinctions, be
cause I agree that it does make a difference as to which
of these we are talking about.
If, in the first case, we talk about differences that
originate from monetary changes at home or abroad,
and if in those circumstances you rely upon a rate change
to neutralize whatever the superficial influences in the
monetary system are inserting into an otherwise balanced
relationship, then that conceptually is possible. Certainly
it is conceptually possible. But this does mean that in

100

T h e B a l a n c e o f Pa y m e n t s

order to take care of every change of any significance in
the monetary performance of any other country, the
exchange rate for the world against the United States
is going to have to change.
Now, Milton, you can’t have it both ways. You can’t
say that you want to live in a world of stable rates and
get the advantage of it and, at the same time, say that
as a part of the adjustment process that the United States
must, in this way, be subjected to the influence of wrong
or right monetary policies on the part of every other
country in the world. You have to choose.
Under the present system we have those hard choices
to make. I am not trying to say that we can escape them.
I am saying that these are a part of the inherent issues
that have to be worked through under either system and
you just can’t escape them by a flip definition.
In the same way, when there are real changes in the
conditions of trade, we get down to the problem that
you and I were trying to talk about a minute ago. We
don’t want to run too far beyond our time limit here
but I think we could come back to that for a minute.
Suppose that the exchange rate of the United States
is improved because, everything else having remained
the same, there has been an improvement in our exports
because we produce something more cheaply and those
goods move abroad. Starting here, the first effect should
be an increase, a strengthening of our own exchange
rate. We are earning more.

R ebuttals

101

PROFESSOR FRIEDM AN: Because we are selling
more.
DR. ROOSA: Because we are selling more. And
now —
PROFESSOR FRIEDM AN: We’re buying more.
DR. ROOSA: Yes. Now, what happens —
PROFESSOR FRIEDM AN: You’re buying more.
DR. ROOSA: What I want to know, because this is
where we get the answer, is: are you saying that the
minute that this rise of exports happens we, in fact, get
an instantaneous adjustment in rate and thereby get an
instantaneous adjustment in terms of importing an equal
additional amount?
PROFESSOR FRIEDM AN : By selling more, the ex
porters acquire additional foreign exchange. Who pro
vides it to them?
DR. ROOSA: The person who buys the exports,
whoever he is.
PROFESSOR FRIEDM AN: All right. Now what do
the exporters do with the additional foreign exchange?
If you’ve got a market that has to clear and there
is no central bank for filling in this gap, they sell the
foreign exchange to people who want to use it to buy
foreign goods. That is what makes our exchange rate
strengthen—that is, makes the price of the dollar in
terms of foreign exchange go up.
DR. ROOSA: Oh, now, that’s another assumption.
That’s really important to track back. You are now

102

T h e B a lan ce of Paym en ts

saying that this version of the flexible-rate system is one
in which central banks do nothing.
PROFESSOR FRIEDM AN: They do something, they
follow their internal policy. They do not intervene in
the foreign exchange market.
DR. ROOSA: All right. Then we just have an abso
lute dichotomy because my premise is that you can’t
keep them out.
PROFESSOR FRIEDM AN: Just for the moment,
assume that they are out and that they’re not at liberty
themselves to take any action affecting the rate itself.
DR. ROOSA: Yes.
PROFESSOR FRIEDM AN: Now, we sell abroad,
and we sell for foreign exchange. What do the sellers
do with the foreign exchange they get? Do they hold it?
If they hold it, the rate won’t change.
DR. ROOSA: O f course, it’s possible that this will
then take the form of this kind of capital movement.
Capital has moved and it is undergirded with a move
ment of exports.
PROFESSOR FRIEDM AN: It could be that the
people who sell the additional goods abroad finance it
by making a loan essentially to the foreign country to
buy it.
DR. ROOSA: Yes, that is one possibility.
PROFESSOR FRIEDM AN: Then you are in identi
cally the same position as the position you approve o f
under a fixed rate.
DR. ROOSA: Yes, quite.

R ebu tta ls

103

PROFESSOR FRIEDM AN: And the rate won’t
change.
DR. ROOSA: And the rate doesn’t change, that is
right.
PROFESSOR FRIEDM AN: That won’t happen.
DR. ROOSA: Why?
PROFESSOR FRIEDM AN: If these people didn’t
want to make a loan before, why should they want to
make a loan now? What happens is that the people who
now acquire the extra foreign exchange try to sell it for
dollars. How can they sell it for dollars? Only by offer
ing more foreign exchange per dollar—
DR. ROOSA: Yes.
PROFESSOR FRIEDM AN: — and why do other
people then want to buy the foreign exchange?
DR. ROOSA: Yes.
PROFESSOR FRIEDM AN: Because if they can get
more foreign exchange for a dollar, this makes foreign
goods cheaper to them. So they buy the foreign exchange
to spend it on foreign goods, which means that our im
ports go up. So that the exchange rate has the effect of
adding to exports and adding to imports, except for
transitional capital movements.
DR. ROOSA: All right. Now, you see this is again
an illustration of what I mean.when I say you can’t
have it both ways. I’ll take that version because that
could be one version.
PROFESSOR FRIEDM AN: All right.
DR. ROOSA: You can’t at the same time say that,

104

T h e B a l a n c e o f Pa y m e n t s

this having occurred, that we are able with a flexible
rate to insulate the American economy from the impact
of a change.
PROFESSOR FRIEDM AN: We don’t insulate the
American economy from the change in our real com
petitiveness in exports or our increased imports. What
we do is to insulate the American economy from any
secondary effect of monetary changes abroad; and we
make this adjustment to the change in real factors with
out requiring all prices in the United States to go up
and down.
DR. ROOSA: Yes, I thought that’s what I was saying.
PROFESSOR FRIEDMAN: I quite agree with you
that it is wrong to say that we insulate the American
economy from everything. We insulate it from those
external events that do not require changes in the pat
tern of production and consumption. We do have to
adjust to external events that do require changes in the
pattern of production and consumption. It is desirable
that we should.
DR. ROOSA: Yes.
PROFESSOR FRIEDMAN: You and I both want us
to adjust to those.
DR. ROOSA: That’s right, that’s right. I have no
problem with that.
PROFESSOR FRIEDMAN: And it is precisely be
cause we eliminate the unnecessary adjustments that we
can have a larger tolerance for those adjustments that
both you and I want us to make.

R ebuttals

105

DR. ROOSA: The other condition is that at the same
time that this isolated sequence is occurring that the
same freedom of rates which has been operating here in
this delightful way is not also being subjected to other
kinds of influences. For example, in the area of short
term capital movements.
PROFESSOR FRIEDM AN: Take the case of short
term capital movements. With fixed rates, you undoubt
edly have destabilizing movements of short-term capital
because whenever there is a possibility that the rate will
be devalued, people have nothing to lose by getting out
of the currency. If they are wrong, they can go back in.
But consider the case of floating rates in the Euro
dollar market. How can people get out of dollars? Only
by persuading somebody else to buy the dollars. In order
to persuade somebody else, they have to offer dollars at
a lower price.
DR. ROOSA: Sure.
PROFESSOR FRIEDM AN : They pay something for
the capital movement. Actual and potential exchangerate fluctuations inhibit capital movement, so that with
flexible rates, you are far less likely to have volatile
capital movements than with fixed rates.
I have looked at Aliber’s study and, as you know,
there are also a series of studies that have appeared in
International Monetary Papers by Tsiang on postwar
European experience.
And in addition, there have been a series of studies on
South American countries, and more recently the Cana

106

T h e B a l a n c e o f Pa y m e n t s

dian experience. All of these suggested the absence of de
stabilizing capital movements, but this isn’t the place
where we can thrash it out.
DR. ROOSA: No, to be sure. But all I was saying on
this one was that there is evidence both ways, not only
one way. Your original statement was that you were
going to bury the thing because no one else had any
contrary experience.
PROFESSOR FRIEDMAN: As I interpret Aliber’s
study — well, again, I don’t think we ought to be going
into that. Maybe we can put it into the record.
[Note subsequently added by Professor Friedman:
The study referred to is Robert Z. Aliber, “ Specula
tion in the Foreign Exchanges: The European Experi
ence, 1919-1926,” Yale Economics Essays, Spring, 1962,
pp. 171-245.
Aliber studied experience in five countries: Britain,
France, Belgium, the Netherlands, and Switzerland. On
the issue under discussion, he concludes that speculation
was destabilizing in France in the sense that speculatively
induced changes in exchange rates produced cost-push
inflation internally, which was subsequently validated by
governmental policy. Thus while speculators were ex post
proved correct—hence in one sense stabilizing—he argues
that they themselves produced the internal price move
ments that proved them correct. This is certainly a possi
bility and perhaps it is correct, but I find Aliber’s
evidence far from convincing. It consists mainly of the

R ebuttals

107

inability to find evidence supporting one alternative
interpretation of the inflation, namely, that it reflected
government deficits, rather than of affirmative evidence
of the influence of speculatively produced changes in
exchange rates on internal policy.
For Belgium, Aliber also finds evidence of destabilizaing speculation, arising, he argues, from "the strong
speculative belief that the Belgium franc and the French
franc should exchange on a one-for-one basis.”
For Britain, he concludes that there was no destabi
lizing speculation but that "speculators forced the
United Kingdom authorities to honor their commitment
to return to the gold standard at the prewar parity when
this was not the path of economic wisdom.”
For Netherlands and Switzerland, he also finds that
speculation was not destabilizing but argues that "both
countries . . . easily could have become subject to a
speculative attack which proved self-justifying.”
At bottom, therefore, Aliber’s negative conclusions
about flexible rates rest primarily on the experience of
France, and even for France, on a possible but not dem
onstrated link between speculation and internal policy.]

DISCUSSION

FIRST SESSION
LOUIS DOMBROWSKI, Chicago Tribune: Professor
Friedman, what effect would a serious economic crisis
in the United States have on floating exchange rates
throughout the world?
PROFESSOR FRIEDM AN: You mean a serious in
ternal crsis?
MR. DOMBRO WSKI: Yes.
PROFESSOR FRIEDM AN: If you had a worldwide
system of floating exchange rates and the U.S. had a
serious internal crisis, the effect would largely be re
stricted to the United States.
In most circumstances, the effect would be that the
dollar would appreciate in terms of other currencies.
Because if the U.S. had a serious economic crsis, it would
have unemployment and declining prices. Therefore
American goods would tend to become cheap relative to
world goods. This would, in the first instance, make for
a balance-of-payments surplus, which would be offset by
an appreciation of the exchange rate.
The situation would be very different from what
happened from 1929 to 1931. A t that time we had rigid

112

T h e B a l a n c e o f Pa y m e n t s

exchange rates, so the U.S. crisis pulled down the rest
of the world. The key reason why there was a world
wide economic crisis from 1929 to 1931 was because the
United States deflated and because most of the world
was on a real gold standard linked to the United States.
Evidence that that was the case for 1929 to 1931
turns out to be readily available in the case of China.
China was on a silver standard when the rest of the
world was on a gold standard. As a result China had
the equivalent of floating exchange rates, because the
price of silver in terms of gold could change. China did
not feel the worldwide depression from 1929 to 1931.
The total exports of China stayed up, income within
China stayed up. China was affected by the world de
pression for the first time in September, 1931, when
Britain went off gold, and the pound sterling depreciated
relative to the Chinese currency.
That’s a very dramatic example of the effectiveness
of a floating exchange rate in insulating a country from
disturbances within other countries.
Obviously, I don’t mean to say that other countries
wouldn’t be affected at all. If the United States had a
great depression, our demand for their goods would go
down, there would be a decline in real demand. This
would produce adverse effects on other countries. But
these adverse effects would not be compounded by our
forcing monetary deflation on them.
They might have some decline in exports, but they
would not be required in addition to adjust their whole

D is c u s s io n

113

price level, which is what we forced the rest of the world
to do in 1929.
HARVEY SEGAL, Washington Post: Professor Fried
man, two questions. First, are you surprised at the
durability of this adjustable peg system that was set
up after Bretton Woods? And the second question: Can
we get a change in the present system of movement,
let’s say toward flexible exchange rates, without having
a crisis as a practical matter?
PROFESSOR FRIED M AN : On the first one, the du
rability of the adjustable peg system, one has to ask what
you mean by being durable. Let’s look at the period
since Bretton Woods. There have been quite a substan
tial number of major adjustments. Britain devalued
sharply, twice I guess it was, wasn’t it? Germany appre
ciated once.
You had a switch from a dollar shortage, when, most
of the European countries had extensive exchange con
trols, import quotas, restrictions on trade to a dollar
surplus, which meant that restrictions on trade were
reduced in Europe and increased in the United States.
So I’m not sure it has really been so terribly durable.
And, of course, if you go outside the range of the
European countries, why then, in the rest of the world
it clearly has not been. But I grant you that the major
issue is about Europe and the U.S. and not the rest of
the world.
Now, your second question. As a practical matter
can you get a change away from the present system?

114

T h e B a l a n c e o f Pa y m e n t s

As a practical political matter, I think there are only
two sets of circumstances under which you are likely
to get floating rates. One is if you have a major crisis.
The second is if you have an effective change in gov
ernment from one party to another, and the floating
rates are established within the first two weeks. Let me
illustrate the second possibility with a concrete case.
Harold Wilson made a terrible mistake. Suppose within
the first week of his coming to office the first time, not
the second time, he had gone on the BBC (British
Broadcasting Corporation) and had said, “ On coming
into office and examining the figures on the balance of
payments and our foreign balance position, I was shocked
at the state in which I found them. I hadn’t realized
that the Conservative Government had done such a
terrible job and left us in such an awful position. Under
these circumstances I can see only two things to do.
Either we can go in for a severe austerity program, for
extensive controls and tightening of the belt in order
to redeem the mistakes of our predecessors, or else we
can devalue. I don’t know quite where to devalue to,
so we ought to float for a while until we find out. I have
chosen to cut the Gordian Knot and get us out from
under the mess we were left in by floating.”
If Wilson had done that, in my opinion he would
have done for the Labor Party by that one act what
Erhard did in 1948 for the Christian Democratic Party
in Germany. Erhard’s one act on that Sunday in 1948
put the Christian Democratic Party in power for over

D isc u ssio n

115

20 years. And I believe that if Wilson had had the fore
sight to do what I have just suggested, and if he had
accompanied floating the exchange rate with the elimina
tion of other restrictions, you might have had the same
kind of economic miracle in Britain that you had in
Germany after 1948.
But once two weeks or so had passed, it was too late,
because by then Wilson had said, “ We shall defend the
pound, we shall not devalue.” He was committed and
it’s very difficult to back out of a commitment.
President Kennedy could have done it in the United
States within his first two weeks. President Johnson
could not do it because he was the same party. If the
Conservatives come into power in Britain, they could
do it in Britain. If the Republicans come into power in
the United States, they could do it in the United States.
Speaking politically and realistically, it seems to me
that those are the only possible circumstances: a crisis,
or during the first two weeks of a new government of
a new party.
What are the chances? With respect to a crisis, any
body who tries to predict the occurrence of a crisis is,
I think, very foolish. You are sitting on a powder keg,
but if nobody lights the fuse, nobody will know you’re
sitting on a powder keg. A major international crisis
is the kind of thing that happens every 20 or 30 years.
The odds against it happening in any particular sixmonths period are very great.
Therefore I would not want to predict that at any

116

T h e B a la n ce of Pa y m en ts

immediately foreseeable date you are going to have a
crisis. Yet I don’t want to rule out the possibility that
one of these days you may very well have an inter
national crisis which blows up to such dimensions that
the only way you can get out of it is by letting the ex
change rates go.
HOWARD S. PIQUET, Library of Congress: Pro
fessor Friedman, I have admired always the manner
in which you have testified before the committees. I
think that you have not been too effective in getting
anything done.
PROFESSOR FRIEDM AN: I hate to have you
denigrate the role of truth in that way.
DR. PIQUET: I’m coming to the question. You have
pretty much taken for granted that adoption of a
floating exchange rate such as in answer to your last
question would also be accompanied by a change of
heart on the part of ourselves and other countries with
respect to trade controls, quotas, and so on.
If we were to adopt the one without the other, that
is, the floating exchange rate without a real determined
movement toward giving up the idea of quotas, wouldn’t
we have the danger of simply doing what we did be
tween the two world wars, of having competitive ex
change bargaining—not bargaining, but—
PROFESSOR FRIEDM AN: Depreciation?
DR. PIQUET: —exchange warfare—
PROFESSOR FRIEDM AN: No.
DR. PIQUET: —exchange depreciation.

D isc u ssio n

117

PROFESSOR FRIED M AN : No. No, we would not.
Let me separate my answer to that.
Even if we didn’t give up any quotas, it would be
better to have floating exchange rates than what we
now have, because what we now have keeps forcing
more and more quotas, more and more restrictions, on us.
Moreover, if we once had floating rates, I think it would
be very much more difficult to maintain the system
of quotas and restrictions.
Maybe I am wrong, maybe we would maintain them.
However, the fear of exchange depreciation is a particu
lar example of confusing a system of adjustable pegs with
a floating rate.
The competitive exchange depreciation that we had
during the 1930s was competitive depreciation with ad
justable pegs, not with a floating rate.
DR. PIQUET: That was gold.
PROFESSOR FRIEDM AN: What’s that?
DR. PIQUET: That was gold still, in some cases.
PROFESSOR FRIEDM AN: Well, yes. But it in
volved changing the price of gold.
DR. PIQUET: That’s right.
PROFESSOR FRIEDM AN: It involved an adjust
able peg, moving the exchange rate from one level to
another. There is another and even more fundamental
difference between the situation in the thirties, and what
the situation would be now with floating rates. In the
thirties you had worldwide unemployment. Countries
were anxious to give goods away if it provided employ-

118

T h e B a l a n c e o f Pa y m e n t s

ment at home. And other countries, foolishly enough,
were not willing to take goods for nothing if it created
unemployment.
At the moment we are not in that situation, and
we’re not likely to be. If other countries want to en
gage in competitive exchange depreciation, we are crazy
if we don’t welcome it. What are they doing when they
competitively depreciate? They are saying to us "Look,
if you’ll take some of our goods for cheap, we’ll give
them to you.” Well, let’s not be fools, let’s take it.
So I think the answer to your question is that in a
world where you have reasonably full employment,
something like a reasonably operating system, there is no
incentive to countries to engage in competitive depre
ciation, and there is no incentive for other countries to
try to avoid it.
DR. PIQUET: Would you be equally as firm in say
ing there would be no incentive to impose import quotas?
PROFESSOR FRIEDM AN: There surely would be
no incentive to impose import quotas.
DR. PIQUET: Why not?
PROFESSOR FRIEDM AN: Why should we impose
import quotas?
DR. PIQUET: I don’t say we. They.
PROFESSOR FRIEDM AN: Why should they? Why
should they impose import quotas? They can’t—under
floating exchange rates they can only import more from
us if they export more to us.

D is c u s s io n

119

DR. PIQ UET: The logic is impeccable. That I under
stand.
PROFESSOR FRIEDM AN: Well.
DR. PIQUET: But I’m talking about the political
reality of how politicians react.
PROFESSOR FRIEDM AN: Right. But how would
politicians react under those circumstances? "Why would
they have an incentive to impose import quotas that
they otherwise would not impose? As we all know, the
forces of protection are always very strong. Particular
industries have special vested interests. They are always
going to try to get governmental measures that they
expect to operate in their favor.
The argument I am making is that the existence of
rigid rates pegged by government strengthens the spe
cial interests who are trying to get measures in their
favor. Maybe they would succeed anyway, but their
road would be a little harder with floating rates.
DR. PIQ UET: I follow your logic completely. How
do I tell Senator Dirksen though?
PROFESSOR FRIEDM AN: I’m not sure I know
what it is that you have difficulty telling Senator Dirk
sen. What is it that you can’t tell him?
DR. PIQUET: Well, I’m giving him simply as an
example.
PROFESSOR FRIEDM AN: I know. Oh, I under
stand that. The thing you tell him is very simple. You
tell him that the floating rate provides general pro
tection for all export industries at one fell swoop and

120

T h e B a lan ce of Paym en ts

you don’t have to have it for one industry after another
separately. The point I’ve tried to emphasize is that
under a fixed rate, the argument against protection is
subtle; sophisticated, and difficult to get across. Under
a floating rate it is simplicity itself.
What are you afraid of? Are you afraid that we are
going to import a lot from abroad? We can’t. The at
tempt to do so would drive the price of foreign cur
rencies up which will stimulate our exports.
The movable exchange rate provides automatic pro
tection. You therefore can dispense with this wasteful,
silly system of protecting industry A separately and B
separately and C separately. Let’s do it in one fell swoop.
And it seems to me that is the argument you make.
DR. PIQUET: I’ll change jobs with you.
PROFESSOR FRIEDM AN: At what price?
HERBERT BRATTER, Banking magazine: If I un
derstand you correctly, these floating rates would soon
result in a congregation of countries attached to the dol
lar and perhaps other countries attached to the pound
and others to the yen and so on. So in effect you would
have floating rates not between a mass of individual cur
rencies, but between a few blocs of currencies. Is that
right?
'PROFESSOR FRIEDM AN: As a practical matter
that is probably the way it would work. You would
have also many separate individual countries, probably
Brazil, Argentina, Chile, Colombia, countries like that,

D isc u ssio n

121

which would not attach themselves. They might, but
very likely they would not attach themselves.
MR. BRA TTER: Well, wouldn’t the countries at
taching themselves to the dollar be opting for fixed
rates rather than floating rates by that very action?
PROFESSOR FRIEDM AN: They would be opting
for fixed rates. And it is not necessarily wrong for them
to do so.
Don’t misunderstand me. I’m not saying that there
never is a case for fixed rates. I’m saying there is never
a case for pegged rates, which is quite a different thing.
The point is that it makes a great deal of sense for
these other countries to tie themselves to the dollar in
the sense of unifying their currency with the dollar—
provided we adopt a reasonably stable internal policy.
O f course if we are foolish and stupid, if we let our
selves in for another crisis such as was suggested before,
well then, they would be very smart to break the tie
with the dollar.
But so long as we maintain a reasonably stable internal
policy, it makes a great deal of sense for smaller countries
for whom foreign trade is a large part of their total
trade to tie themselves to the dollar.
MR. BRA TTER: How would those smaller countries
which are working for or asking for commodity agree
ments to peg the price of coffee or cocoa or something
else fare under such a system?
PROFESSOR FRIED M AN : The commodity stabili-

122

T h e B a la n ce o f Pa y m en ts

zation programs, that is a different question. That s a
question—
MR. BRATTER: They want fixed prices.
PROFESSOR FRIEDM AN: What’s that?
MR. BRATTER: Those who want stabilization want
fixed prices, fixed relationships.
PROFESSOR FRIEDM AN: Yes. I think that the
United States makes an enormous mistake by partici
pating in any such stabilization agreements. They are
cartel agreements that we ought to oppose. We ought
to be in favor of free market prices for commodities
as well as free markets for currencies. If those coun
tries still want to engage in those stabilization agree
ments, we ought not to support them or cooperate with
them.
MR. BRATTER: Why did Canada abandon floating
rates after 12 years?
PROFESSOR FRIEDMAN: That is a very good
question, and I’ll be glad to indicate why. The first
point that has to be made is that floating rates are not
a guarantee of sensible internal monetary policy. You
can have silly internal monetary policy with fixed rates,
you can have silly internal monetary policy with float
ing rates. All floating rates do is to make it possible for
you to have a sensible internal monetary policy without
considering the rest of the world.
What happened in Canada? It’s a very interesting
and instructive story, because the reason—I’ll come to
the end first and then I’ll go back and trace it out. The

D isc u ssio n

123

reason Canada went off floating rates was because they
were working so well, and their internal monetary policy
was so bad. Let me illustrate that a little bit. From 1950
to 1952 the Bank of Canada was interfering in the mar
ket for the Canadian dollar. You had a truly floating rate
from about 1952 on.
From 1952 to 1962, so far as exchange rates were
concerned, they worked very well. The rate floated,
but it didn’t move around a great deal. Speculation was
clearly stabilizing. Short-term movements were mild.
The Bank of Canada largely stayed out of it.
But side by side with that, the Canadian internal
monetary policy was a very bad monetary policy. It
was an extremely erratic monetary policy, particularly
when J. E. Coyne was Governor of the Bank of Canada.
He tried to lengthen the maturity distribution of the
Canadian debt, and did it in a very clumsy way. In
general, he followed an erratic and generally disturbing
monetary policy that left Canada with relatively high
levels of unemployment.
He also, as you know, was very much in favor of
Canadianization of industry, opposed to the import of
capital from the United States. So he was in favor of
trying to interfere with the free flow of capital from
the United States. That was one of the reasons why he
followed the kinds of monetary policy he did.
When Coyne finally left, there was an attempt to
do something effective about the unemployment prob
lem in Canada. But instead of correcting their bad mone

124

T h e B a l a n c e o f Pa y m e n t s

tary policy internally, which would have been a sensible
thing to do, they said, “Ah, here is a floating rate. We
will force the rate down and stimulate employment in
side Canada by discouraging imports and encouraging
exports. We will try to engage in the kind of thing Mr.
Piquet was talking about, competitive currency de
valuation.”
And so what happened? The Bank of Canada an
nounced that it was going to try to force the price of
the Canadian dollar down by exchange speculation. The
speculators didn’t believe it. The Bank of Canada spec
ulated against the Canadian dollar, and the speculators
absorbed the funds, and nothing happened. The Cana
dian rates stayed fairly fixed. So the Bank of Canada,
said, "We haven’t been doing this on a big enough scale;
we’re going to do it yet.” Then the Bank made bigger
and bigger announcements and engaged in larger and
larger speculative actions.
Finally it started the rate moving down. Once the
rate started moving down, the speculators said to one
another, "The government really is going to do it.” So
what do the speculators do in that case? They jump on
top of the government speculation and all of a sudden,
the rate started to go down much faster than the Gov
ernment had intended it to go down.
Now the government was stuck. "W e’ve got a tiger
by the tail,” they said. "How do we stop this downward
slide of the rate?” When they initially started this op

D isc u ssio n

125

eration, they had no intention of pegging the rate. What
they were trying to do was drive it down.
When under their stimulus there was a rapid move
ment down, obviously the sensible thing for them to
have done would have been to announce that they had
made a mistake, that they were getting out of the mar
ket and that they were going to let the dollar resume its
former behavior.
But no government in the world has ever done a
thing like that. The key principle of a government is
that you make a different mistake each time, not the
same one. And you never admit that what you did before
was wrong.
To go back, the only reason Canada got on a floating
rate in the first place in 1950 was because of a similar
earlier sequence in which they had made a mistake,
when they had appreciated the Canadian pound earlier.
And instead of undoing that mistake, they went on and
floated. Well, similarly, this time, instead of simply un
doing it, they said, "Well, the way we’ll stop the rate
from going down is by pegging it.” So they announced
that they were going to peg it at 92/z cents. As it hap
pened, it took very large operations at that stage to break
the speculative movement and to peg it. But they finally
succeeded in pegging it at 92 /l z cents. Then they were
stuck with it there.
It was a very unwise thing for them to do in terms
of their own internal policy. It is not evidence in the
slightest that the floating exchange rate wasn’t working.

126

T h e B a l a n c e o f Pa y m e n t s

The Royal Bank of Canada has been a strong advocate
of a floating rate in Canada and has published a con
siderable number of very interesting empirical analyses
of the operations of it before and of the effects of freez
ing the rate afterwards.
Thanks to the exemption under the interest equali
zation tax which requires that Canada hold her foreign
exchange holdings relatively constant, Canada is now
pegged to the dollar in thet literal sense. She is having
a real gold standard in relation to the dollar. She is uni
fying her currency with the dollar.
MR. BRATTER: And at the annual meetings of the
International Monetary Fund there is no country strong
er than Canada for the system of fixed parities.
PROFESSOR FRIEDMAN: Of course! Why should
this surprise you? They have to defend the action they
took. When was the last time you heard an administrator
get up and say, "Well now, the action I took was the
wrong action for me to take” ?
Do you hear that very often? The tyranny of the
status quo is enormous.
When Canada first went on a floating rate in 1950,
almost all of the banks of Canada were opposed to going
on the floating rate. When Canada went off the floating
rate in 1962, a considerable fraction of the banks were
opposed to going off the floating rate. The basic prin
ciple of administration is that every administrator knows
that the way he is conducting the particular program
he is conducting is the only way it can be done. I was

D isc u ssio n

127

first taught this lesson in a very striking way back in
1941 to 1943 when I was working at the Treasury. I
was involved in the development of the withholding sys
tem for personal income taxes.
If you asked Internal Revenue today whether they
could collect the present income taxes without with
holding at the source, there is no doubt that they would
say it would be utterly impossible. But who was our
biggest opponent when we tried to get withholding
taxes in 1941 to 1943? Internal Revenue. They said it
was unworkable, it was unfeasible, it was administra
tively not possible. So I think the fact that people who
three years ago decided to shift to a fixed rate are de
fending the action they took is hardly evidence that the
action they took was a wise action. I think the Canadian
case is a very strong case on the other side, myself.
MR. SEGAL: As an interim step in reform, what
would you think about widening the dollar gold point,
let’s say between $30 and $40 an ounce?
PROFESSOR FRIEDM AN: I think that would be
a serious mistake. But I think widening the exchange
rate limits would be very desirable. I would urge on you
to separate the gold problem from the exchange rate
problem. What we ought to do about gold, it seems to
me, is a separate question from what we ought to do
about exchange rates.
We could stop buying and selling gold and sit on
it, let the gold price be a free market price like the price
of lead or copper or anything else. And we could, as an

128

T h e B a la n c e o f Pa y m e n ts

interim matter, instead of having the pound pegged at
$2.78 to $2.82, peg it between say $2.70 and $2.90.
That would certainly be an improvement for exchange
rates. But I think it is not a desirable thing to do it with
gold.
MR. BRATTER: Prominent bankers recently sug
gested that the United States discontinue selling gold
and there has been some speculation as to whether they
were launching a trial balloon for the Secretary of the
Treasury. If that were done, would we then be on float
ing rates?
PROFESSOR FRIEDM AN: No. As I tried to em
phasize in my paper, there are two problems. If we tinpegged silver, would that put us on floating rates? The
problem of fixing, the price of a commodity like gold or
silver and the problem of pegging currencies are two
separate problems.
Many countries — Germany and France — peg ex
change rates. But they don’t necessarily, as an official
matter, peg the price of gold. In fact, in France you can
buy and sell gold freely, but the price varies. The price
of a French napoleon is not pegged.
So we could unpeg gold, stop buying and selling it,
and yet continue to peg exchange rates for a consider
able period.
MR. BRATTER: That is by official intervention, you
mean.
PROFESSOR FRIEDM AN: That’s what we do now.

D is c u s sio n

129

How do we peg exchange rates now? We don’t really
do it by shipping gold.
MR. BR ATTER: If we stopped buying gold and we
did not peg—
PROFESSOR FRIED M AN : Peg gold or peg exchange
rates?
MR. BR A TTER: Well, peg exchange rates.
PROFESSOR FRIEDM AN: Yes?
MR. BRATTER: Then what would be the initial ef
fect of that step on the $2.82 rate for the pound sterling?
PROFESSOR FRIEDM AN: Well, in the case of the
pound sterling, because Britain is in balance-of-payment
difficulties, the pound sterling price would probably fall,
not rise. What would happen would be that the dollar
would very likely appreciate relative to the pound sterl
ing, and depreciate relative to the franc and the mark.
So the different prices would behave in different ways.
Trying to predict these things is much more compli
cated than it appears.
Let’s suppose for a moment that tomorrow we elim
inated the interest equalization tax and all our voluntary
exchange controls. Let’s suppose we eliminated oil im
port quotas and a bunch of the other quotas, and let’s
suppose we set the dollar free.
I am not sure that the dollar would fall. It might tem
porarily. But I’m talking about what would happen
over six months or a year. Do you suppose that when a
bank has lots of outstanding liabilities and that banker
gets on the stairs of his bank and says, "You know, we’re

130

B ala n ce

Pa y m e n t s

a very sound bank, but, gee, I wish you would hold off
coming in and trying to ask to get your deposits for six
months or a year.” Do you think that’s the way to
strengthen confidence in that bank? I don’t think so.
Similarly, we have been doing everything we possibly
could to weaken confidence in the dollar and to tell
everybody in the world, "You’re a fool if you hold
onto any more dollars than you have to.”
If we acted from strength, which we have—and sim
ply said, “ We’re going to stop all this nonsense, we’re
going to remove the interest equalization tax, we’re not
going to force our military to buy on the most expen
sive market and so on,” I would not be a bit surprised
if the market price of the dollar strengthened rather
than weakened.
You know, it’s an interesting thing that economic
events often work very much differently than you ex
pect.
What killed silver as a monetary metal? The fact that
we raised its price.
That seems kind of paradoxical, doesn’t it?
MR. BRATTER: No. Silver — the silver standard
started to disappear in the last century. We just—
PROFESSOR FRIEDM AN: Yes, to some extent it
did.
MR. BRATTER: India went off the silver standard
in 1893.
PROFESSOR FRIEDM AN: Yes. But when did China
go off it?

is c u s s i o n

131

MR. BRATTER: China went off when we started
buying silver heavily in 1934.
PROFESSOR FRIEDM AN: Right.
MR. BRATTER: That’s when they officially went off.
PROFESSOR FRIEDM AN: Right. "What happened
to Mexico?
MR. BR A TTER: Mexico went off much earlier.
PROFESSOR FRIEDM AN: But Mexico still had an
extensive full-bodied silver coinage. What happened to
it in 1934?
MR. BR A TTER: It went to the melting pot after
we started buying.
PROFESSOR FRIEDM AN: Right. If we had not
engaged in a silver-purchase program in 1934, silver
would today be a monetary metal.
What killed gold as a monetary metal?
MR. BR A TTER: I wouldn’t follow you on that.
PROFESSOR FRIEDM AN: What killed gold as a
monetary metal is that we raised the price to $35 an
ounce in 1934 and accompanied that by measures mak
ing it illegal for individuals to hold gold and in effect
declaring gold clauses unenforceable. Paradoxically, that
is what killed gold as a monetary metal. And that is
why the gold standard at the moment can only be re
established, if ever, by first getting rid of the vestiges
of it that we now have and letting it re-emerge as a
real honest-to-God gold standard.
MR. SEGAL: To come back to the point about
widening the gold points, would you agree with this

132

Bala n ce

Pa y m e n t s

chain of events which I deduce logically? That if we
were to widen the gold points and therefore made it risky
for a country like France to hold gold, then France
would have the choice of either continuing to peg on
the dollar—which she now does—or pegging on gold.
If she chose to peg on gold, would she have flexible
exchange rates?
PROFESSOR FRIEDM AN: That’s right.
MR. SEGAL: I mean would you accept this as a
strategy that might help to put us on the right road?
PROFESSOR FRIEDM AN: Well, I don’t know.
Maybe.
I was directing my attention to the question of
whether it was desirable to keep the gold as a means
through which you adjusted exchange rates.
Your idea would be to let the gold price be determined
on the London gold market?
MR. SEGAL: Yes.
PROFESSOR FRIEDM AN: And to peg it between
$30 and $40 an ounce?
MR SEGAL: Yes. With plenty of play.
PROFESSOR FRIEDM AN: Well, maybe if it went
up to $40 an ounce and we could get rid of most of our
gold at that price, it might be worth doing. It’s better
than getting rid of it at $35 an ounce.
You see, I must say that my own favorite scheme is
completely impossible politically, and therefore I’m not
talking about it. But if you were being rational, the
rational thing—and this is supposed to be a rational

D isc u ssio n

133

debate—in my opinion for the United States to do
would be first:
Repeal all prohibitions on private individuals owning
gold, trading in gold, or exchanging gold.
Second, announce that on one Monday we are going
to have an auction and we’re going to get rid of the gold
stock to the highest bidder and go out of the business.
I see no reason why the storage of gold should be
a nationalized industry, any more than I see why the
delivery of mail should be a nationalized industry. They
are equally illogical, and we see the inefficient results in
both cases.
But, as I say, I realize that this is highly unfeasible
politically. So I have been inclined to content myself
with saying, “ Okay, let’s stop buying and selling gold
and just sit on our gold stock.”
As a political matter maybe your device would be a
better one for getting off. I really haven’t thought about
that particular device.
MR. BRA TTER: Are there many Milton Friedmans
abroad advocating this sort of thing?
PROFESSOR FRIED M AN : It depends on what you
mean by “ this sort of thing.” If you mean by “ this sort
of thing” more flexibility in exchange rates, the situa
tion today is that if you were to poll the professional
people in money and international trade, the academic
people, you will find that at least three-quarters of them
are in favor of a greater degree of flexibility of exchange
rates.

134

T h e Ba la n ce

Paym ents

This is a tremendous change in opinion. Fifteen to 20
years ago you would not have found 5 percent.
But today I don’t think there is the slightest doubt
that an overwhelming majority of people in this area
would favor a greater degree of flexibility of exchange
rates.
Not all of them would go to completely floating rates.
Most probably would not.
Most people would favor widening the range of fluc
tuation and seeing how that worked. Some people would
go to a more complicated system of widening the range
and having a movable peg, saying that in any year in
which the exchange rate is toward the lower end of the
band the parity rates would be lowered. That would be
a possible in-between case. That is, so far as the aca
demic world is concerned.
If you go to the financial community, the number
of people who are in favor of flexible exchange rates
or floating exchange rates is many fewer. That’s under
standable, I think. This is their business. And everybody
is always in favor of free prices for everybody else, but
not for himself.
MR. BRATTER: That is especially true if they get
a phone call from Washington.
PROFESSOR FRIEDM AN: Yes, I’m sure that’s
right.
That is part of the problem in this area. It is impossible
for any high official to hint at the possibility that ex
change rates might be changed. And understandably.

D is c u s s io n

135

If I were a high official, I wouldn’t hint at it so long as
the government is committed to holding it. Therefore
it is very difficult to get an intelligent, open public de
bate on this issue. It has to come entirely from the out
side.
For example, a couple of years ago the Council of
Economic Advisers professed in one section of its report
to discuss all the alternative ways in which you might ad
just the balance of payments. They said, ''We’re not
talking about policy, we’re just going to give the eco
nomics of it.” Yet that section did not contain the
words “ exchange rate.” It obviously was not a scientific
discussion. It was understandably, but nonetheless actu
ally, a political discussion.
If you will pardon me for saying so, I think the
banking community is being extremely shortsighted be
cause it seems to me that it would be in their interest
to move toward floating rates. Here they are, accepting
ever-increasing controls over their business, controls
over whom they may lend to abroad and at what terms.
Why? For what gain to themselves? They are the people
who could specialize in this business of foreign exchange
transactions. Who would it be who would run the
futures markets?
MR. BRA TTER: Did you notice that the Bankers
Association of Foreign Trade just this week protested
against this?
PROFESSOR FRIEDM AN: Maybe, but that doesn’t
make them any the less shortsighted. I am an empirical

136

Bala nce

Pa y m e n t s

scientist. In many other areas I have observed time and
again that the business community, in these issues of
public policy, is very shortsighted. Time and again, they
oppose measures which, after they are adopted, they
welcome.
We had a very simple recent case of this, the invest
ment credit. As it happens, I have always been opposed
to the investment credit. I think it’s a bad tax measure.
It should never have been on the books. When it was
first proposed the business community at large was vio
lently opposed to it.
Last year, when it was proposed to take it off, the
business community said this was the greatest thing that
had ever happened and that it was disgraceful to take
it off. This is one example. But there are numerous other
examples of the fact that the business community in this
respect tends to be very shortsighted. If it is true that
banks engaging in foreign trade came out against floating
exchange rates, it’s another—
MR. BRATTER: I’m not saying that. They came out
against all of these controls that you have mentioned.
PROFESSOR FRIEDM AN: I beg your pardon. I
thought you said against floating rates.
MR. BRATTER: Oh, no, no, no.
PROFESSOR FRIEDM AN: But what is their alter
native? There is no point in coming out against these
controls unless you have an alternative.
MR. BRATTER: I couldn’t tell you. I didn’t see
the full statement.

D isc u ssio n

137

PROFESSOR FRIEDM AN: Otherwise they are just
spitting into the wind. Unless you have some alternative
adjustment mechanism, you must have direct or indirect
exchange controls. There is no alternative. There’s no
use kidding yourself. You cannot have your cake and
eat it too.
Under present circumstances in the world, you can
not have fixed rates and freedom from control indefi
nitely. You can have it for a time, but you can’t indefi
nitely.
PROFESSOR PAUL M cCRACKEN, University of
Michigan: I have just a comment on this. I think the
Bankers Association, or this statement, probably was
opting for what really would be the kind of austerity
program that the British have— in other words, to try
to take the adjustment in the form of a lower level of
business activity.
PROFESSOR FRIEDM AN: Again, if that is so, it’s
another evidence of shortsightedness. What do they gain
from that? Is it really in the self-interest of the bank
ing community to force a recession on this country in
order to enable the price of the mark, let’s say, in terms
of the dollar to be kept at 25 cents?
What is there about this particular price that makes
it a be-all and end-all of policy? I think the explanation
is that the alternative has not been made clear to them.
If they examined the alternatives and if they had the
choice of either a slight fluctuation in the price of the
dollar in terms of foreign exchange or a substantial re-

138

B ala n ce

Pa y m e n t s

cession or depression in this country, it’s very hard for
me to see how anybody could opt for the second under
present circumstances.
MR. BRATTER: I think you have to distinguish be
tween two kinds of bankers. We have a certain number
of large banks which are engaged in international ac
tivity and the rest of the 14,000 banks in this country
are domestically oriented and know very little about
this excepting what they hear from the leaders.
PROFESSOR FRIEDM AN : Yes.
MR. BRATTER: They have no original ideas on it.
PROFESSOR FRIEDM AN: I’m sure you’re right.
PROFESSOR M cCRACKEN: Both groups tend to
agree on this.
PROFESSOR FRIEDM AN: As Paul says, both the
large banks and the small banks tend to agree?
MR. BRATTER: That’s what I say, they get their
leaders, who lead the way.
PROFESSOR FRIEDM AN: Then it’s the large banks
who need to be educated on this issue.
MR. SEGAL: Just one footnote. I think that in the
case of these large banks really what they were for,
and I can document this in the case of the interest equali
zation tax, was a corporative system of controls. They
were opposed to the interest equalization tax in 1963.
What they really wanted was a capital issues committee.
They are all for that.
PROFESSOR FRIEDM AN: Sure, I don’t blame them.
MR. SEGAL: This could be a cozy cartel.

is c u s s i o n

139

PROFESSOR FRIED M A N : Let me go back. One of
the reasons why the bigger banks have ambiguous atti
tudes is because, of course, the so-called voluntary agree
ment to restrict foreign lending is simply a cartel agree
ment.
MR. SEGAL: O f course.
PROFESSOR FRIEDM AN: Under which they can
charge higher prices to foreign borrowers than they
otherwise could and under which they are protected
from competition of people who are not in the business.
MR. SEGAL: Exactly.
PROFESSOR FRIEDM AN: As you all realize, the
President asked and received a legislative exemption
from the Antitrust Act in order to make this voluntary
restriction on foreign lending effective. So they had
mixed motives.
On the one hand, their cartel interests led them to
favor it but again I think this is shortsighted. They may,
in the short run, get something out of this cartel but, as
exchange controls tighten up, they are going to find
that the cartel is not operated in the way in which their
own interests would dictate. But maybe I’m wrong.
Maybe the cartel is in their own interest. Well, then it
is up to the rest of us to whose interests it is adverse to
try to do something about it.
MR. BR A TTER: Well, the additional charge which
is represented by the tax does not go to the banker.
PROFESSOR FRIEDM AN: No, no. N ot the Interest
Equalization Tax. I am talking about the voluntary

140

T h e B a la n c e o f P a y m e n ts

agreement on bank lending. There is no charge on that.
JAROSLAV HABR, Academy of Sciences, Prague,
Czechoslovakia: Is it not an advantage for people who
are engaged in long-term planning to have fixed prices?
Otherwise, in addition to all the other uncertainties,
they are also subject to Uncertainty about prices.
PROFESSOR FRIEDM AN: These people are faced
with a risky situation with fixed prices. If the prices are
fixed wrong, then something else is unstable. Again, it
seems to me what is involved in that argument is the
belief that you can have your cake and eat it too.
O f course, it would be very nice if you could have
stable prices and also the equality of supply and demand
in all markets, but you cannot, if conditions are chang
ing. You can fix the price, but then you have to do
something else about the quantity. You may have greater
stability in prices, perhaps under an arrangement with
fixed prices but you have greater uncertainty about
everything else. And, in fact, the long-term planners are
in a far better position, if they have the flexible and
prompt signal of .prices to give them information about
the state of demand and supply in various markets, than
if they have to infer themselves that state of demand
and supply from the length of queues of various kinds.
That’s very complicated to do.
Just consider in this country for a moment, leave
aside Czechoslovakia and Poland, consider in this coun
try the situation in agriculture. Look at the problem

is c u s s io n

141

our agricultural planners have been up against. Why?
Because they don’t have a price system to do the job for
them and they are trying to do crudely what this much
more effective instrument would do in a sensitive way.
If, in fact, internal monetary policy in the United
States is stable and in other countries is stable, then,
under a system of floating exchange rates, exchange
rates will be free to move but they will, in fact, be
highly stable.
The price of sugar is free to move in our markets but
it’s highly stable. The prices of other products which
are free to move are highly stable. There is a difference
between being free to move and actually moving around
a lot. And, if the price does actually move around a lot,
it’s because something is happening to move it. That
something would still be there if you froze the price.
DR. H A BR: May I comment?
PROFESSOR FRIEDM AN: Sure, I would like to
have you.
DR. H A BR: Because it seems to me that there is a
difference, if you can make your decisions on the ground
of the development of prices on a free market. So it’s
not so difficult to take into consideration the various
factors which involve your field of activity. But if be
sides this you must take into consideration, let’s say, this
movement of exchange rates in a country, then you
don’t know what are the real reasons for these changes.
They are connected with various factors which are not
necessarily within your own sphere of activity.

142

h e

B a la n ce

Pa y m

ents

PROFESSOR FRIEDM AN: Right, right, right.
DR. HABR: So it’s much more risky a situation —
PROFESSOR FRIEDM AN: No, no, it’s the other
way around. If I don’t have the exchange rates moving,
if they are pegged, then I really have to go and find out
about those real things. But, if the exchange rates are
moving, then they give me the information without my
knowing the real things. I don’t have to know why the
exchange rate is rising. You say to me, "Well, I have to
know whether it’s going to be higher a year from now.”
Fine, I have a futures market on which specialists in
exchange rates speculate. They provide me, free, for no
cost, with information that I could never in a hundred
years get myself by trying to become an expert in all
these different areas.
The great virtue of a price system, and it’s just as
much a virtue in exchange as it is in other areas, is that
it’s an extremely effective way of giving each man the
information he needs.
How am I going to get that information some other
way? Suppose I freeze the exchange rate? I still have
to know whether it’s going to be possible ten years from
now, if I invest now in Brazil, to get my dollars out.
If I have frozen exchange rates, I have to ask myself,
ten years from now will Brazil be in a situation where
exchange control will be loose enough so they will per
mit me to take my dollars out? I submit that’s a much
harder thing to forecast than it is simply to sell some
dollars forward on a forward exchange market.

isc u ss io n

143

DR. HABR: It is the sort o£ argumentation that this
would be better than the present type, which is worse,
but it is not an argumentation that either is right.
PROFESSOR FRIEDM AN: What’s a better one?
Tell me a better one?
DR. HABR: Oh, yes, with that I quite agree. We are
in a tight situation.
PROFESSOR FRIEDM AN: The truth is that the
future is uncertain. That’s the truth. Now, the question
is, among alternative mechanisms for dealing with un
certainty, which one comes closest to giving the right
answer? None can give the right answer because the
future is unknowable. It really is uncertain. Among the
various mechanisms that imperfect men have invented
or that have grown up, the price system seems to be
about the most efficient as a transmitter of information
and as a means whereby people can deal with true
uncertainty.
And I think the problem is the one you have suggested.
Everybody wants to have his cake and eat it too. He
wants to have a fixed exchange rate and also always
have the freedom to buy and sell an unlimited amount
of exchange and also have stable prices internally and
externally.
He can’t. Something has to give and, therefore, you
have to choose which is the thing that it is best to have
give.

SECOND SESSION
PROFESSOR H E N R Y BRIEFS, Georgetown Uni
versity: I want to get to the question of adjustment in
the balance of payments under conditions of pegged
exchange rates. This view of the way in which balance
can be restored has a lot of appeal. I think it has been
part of the central doctrine about the adjustment mech
anism under pegged rates.
One of the difficulties is that you have to accept vari
ations in the unemployment rate. That is, if you vary
the growth rate, you are going to get fluctuations in
the unemployment rate. Unfortunately, because of do
mestic considerations, countries have tied themselves
down to rather rigid criteria of what an acceptable un
employment rate is. If you are willing to let the unem
ployment rate fluctuate, let’s say, in our case, between
3.7 and 4.5 percent, this might make some sense.
But, if you get the view, which you are beginning to
get, that any time the unemployment rate moves to
wards 4 percent or possibly beyond that we really have
a recession underway, balance-of-payments adjustments
in terms of variations in the growth rate seems to go

146

T h e B a la n c e o f P a y m e n ts

down the drain. In other words, if you are rigidly tied
to maintaining a very low level of unemployment, you
can’t have such flexibility in the growth rate and you
just paralyze the adjustment mechanism.
DR. ROOSA: The best answer is to say that there
will always have to be some range of variation in the
unemployment rate, as a byproduct of the change that
is always going to be underway within a dynamic domes
tic economy. Whatever is found by experience to be
the acceptable range of unemployment required by
purely domestic adjustments should set the outer limits
on what is produced by any action considered necessary
for balance-of-payments adjustments.
That isn’t a clear and resounding answer in terms of
percentages. There are people who will still raise ques
tions as to whether for domestic reasons at times the
rate may have to go as high as 4.5 percent. We will be
refining these criteria over the years ahead as we already
have in the recent past.
I can remember vividly a time when some of my
associates and I thought we were talking about pie in
the sky when we tried to visualize a set of arrangements
that would lead to 4 percent unemployment and, at the
same time, bring this about in an environment of rela
tive price stability and thereby also be helpful to the
balance of payments.
I think, if we keep this same objective, we will find
as the unemployment target is further sharpened over
the years, as we learn more about the intractable or

D is c u s s io n

147

malleable nature of the composition of the remaining
unemployment, that we’ll be able to make the necessary
balance-of-payments adjustments without necessarily
increasing unemployment beyond that range of fluctua
tion which would be implied and inevitable for the via
bility of a dynamic economy.
I think that can even be said about the thinking of
some of the people in the United Kingdom and the
change that they’ve induced to increase the unemploy
ment rate to 2 percent. I don’t myself agree that such
an objective was inevitable. I think the British govern
ment was drawn to this (relatively mild) extremity
because it didn’t take other kinds of appropriate balanceof-payments 'action earlier. This is only to highlight,
though, the need all of us face to make the judgments
in advance that we could have made better in retrospect.
But I don’t think there has to be anything inherently
contradictory here, between my suggestion for varying
the rate of economic advance and the widespread com
mitments to hold unemployment at low levels. What
I’m suggesting is going to be harder, because it’s a very
simple and easy prescription to be brutal and to say,
we’ll take the unemployment, we’ll rattle things around
and we’ll bust through in a hurry. My way is going to
have to be longer and slower moving, but I don’t see
any fundamental reason why, if we are convinced of the
objectives, the engineering of it can’t be brought about.
HARVEY SEGAL, Washington Post: Two questions,
really.

148

B a la n ce

Pa y m

ents

The first one is: What are the grounds for your con
fidence that the adjustments under a system of fixed
exchange rates can be confined to, first differences, all of
them positive in the rate of output? I think I could
argue that the experience from 1958 to 1961 doesn’t
support your case.
Secondly, what about the very serious problems of
forecasting, if we are going to continue along the lines
that you are suggesting? And I am thinking of the situ
ation that confronts the country at this moment. In
view of your appraisal of the balance-of-payments out
look, what sort of domestic policy would you advocate?
I am curious to know because I am wondering if
they are not in conflict.
DR. ROOSA: As to the first, I can’t introduce any
proof. There is no necessary reason, of course, why all
of the changes are going to work out in such a way that
there will only be variations in the quarter-by-quarter
pluses, as between larger or smaller figures in the process
of achieving adjustment.
I do say that this is a reasonable aim of economic
policy and that I think most of the time it should be
workable. The demonstration of it is going to be limited
or the evidence is going to be limited by our success in
reasonable fulfillment of the challenge posed by your
second question and there will always be room for human
error there too.
The forecasting problem for any purpose, whether
that of a completely insular economy or that of one

is c u s s io n

149

exposed to the outside world, is always difficult. Even so,
the spillage that we have to contemplate for the balance
of payments of the United States, however important
that may be now because we have run deficits for so
long, is, in our case, relatively much smaller than for
any other country.
But I do feel that the need—in accepting what I re
gard as the premises of modern economics or as Arthur
Okun says, "good economics”— the need is to make the
forecasting art one that is continually being improved.
This need exists and persists for the domestic economy.
Folding into any forecast the foreign economic policy
implications—choosing from among a variety of choices,
those choices that will also be most useful in balance-ofpayments terms—seems to me relatively the easier part
of the process.
Now, to take the immediate situation and the balanceof-payments difficulty that we face, I have no better
figures than those of the first quarter, but I am assuming
that we are now running, on a conventional basis, a
balance-of-payments deficit of over $2 billion at an
annual rate and that this is going to require some repair
action and some corrective action.
The corrective action broadly is the same that I think
is needed for the continued strength and advance of the
home economy and for absorbing the continued increase
in the labor force. I think that does mean that we first
will benefit from the restoration of the investment credit
and the stimulus to productivity-raising investment that

150

B a la n ce

Pa y m

ents

this provides. Second, that, and here I am being very
brash and without any analysis am just trying to give
illustrations which any of you can improve upon but,
second, I think we are going to have to rely, if the ex
pansion becomes a problem on the inflationary side (with
the continued growth of Vietnam expenditure) we are
going to have to rely more on overall tax increases.
That means that, as of now, my guess would be that
a recommendation should be made for a tax surcharge,
that the timing will be a little later than originally con
templated, probably by the 1st of October but no later
as an effective date than the 1st of January. Based on
information neither you nor I have, and I don’t suppose
Mr. McNamara has yet but will soon have to have, on
the scale of the further increments on the expenditure
side, I would feel that there is a risk that this will have
to be a 10 percent, not a 6 percent surcharge, and I would
think that accompanying that there will be some slowing
down in the pace of monetary expansion that has pro
ceeded so rapidly in the early months of this year; that
will still mean an expansion in the magnitudes of these
days that is within the bounds of a normal seasonal
growth.
I doubt that it’s possible to drive the whole structure
of interest rates very much lower, because the economy
itself will be so strong that you will be fighting an irre
sistible force, if you try to contrive rates which don’t
equate supply and demand in the market. But the same
balancing of the components of fiscal and monetary

is c u s s io n

151

measures, and the same need for restraint in the series
of large wage negotiations ahead of us, that will be nec
essary for domestic balance and growth will also be
appropriate for improving our own competitive position
balance-of-payments wise.
I suppose it’s unlikely that anything we do can ac
complish a reduction in the rate of price increase below
2.5 percent and we may get something more this year;
but I think, even at that, we may gain ground in the
competitive race, balance-of-payments wise, around the
world. I also think the fact that we have passed the
period of peak inventory accumulations and should now,
certainly by the next quarter, be proceeding at a rate
which is more nearly normal in comparison with sales
advances, perhaps still a little on the low side, will in
turn mean that that volume of marginal demand which
reaches outside the country and produces an increase in
imports will not be present. The economy will still ex
pand but the high ratio of import increases that often
accompanies an extraordinary rise in inventories will not
be present.
And I would then think that we should see for this
year an improvement on the trade account in the mag
nitude of one-half to one billion dollars, all of which
will be lost on the other side through Vietnam expendi
ture and through some further slippage on the capital
outflow with the result that the total ordinary liquidity
deficit will run somewhat higher than last year. And
just because the Euro-dollar market plays a role here too,

152

h e

B a la n ce

Pa y m

ents

the deficit on the official settlements basis this year will
probably be higher than the liquidity deficit. That
doesn’t worry me a bit.
I do think, insofar as this is a problem, that the im
portant thing is that the holding of the additional dollars
flowing from this operation will end up either in the
IMF through other transactions or in Germany where,
at least, there is a kind of contented holding, or in the
countries on the periphery of Vietnam who are interested
in rebuilding their reserves.
So that I think, with the present combination of de
velopments and Treasury tactics, we can now avoid the
kind of balance-of-payments problem that has any ex
plosive connotations this year. But next year will be
another story.
NORMAN TURE, National Bureau of Economic
Research: I would like to pursue the same line of ques
tioning that Henry Briefs and Harvey Segal have initi
ated but, Dr. Roosa, I first want to offer a comment, if
I may, on your tax prescriptions.
The restoration of the investment credit ought to re
duce the tax on the returns to corporate capital by some
thing of the order of $2 billion. The proposed increase
in corporate taxes under a 6 percent surcharge proposal
would be something of the order of magnitude of $3
billion. It is hard for me to see why this represents a
combination of measures that will, in fact, increase the
incentive for investment in productivity-enhancing cap
ital in the corporate sector.

D is c u s s io n

153

But may I now ask my question?
DR. ROOSA: Certainly. I am going to treat that as
a question too, if you don’t mind.
MR. TU R E: All right.
It seems to me that when you make the observation
that one virtue of a fixed-exchange system is that it
sends up clues when a payments deficit occurs and indi
cates what should be done to domestic policy, under
lying that there has got to be some sort of assumption
that the change in the deficit in either direction is a
function of the rate of resource utilization in the do
mestic economy.
But it seems to me that the experience of the late
1950s and the early 1960s does not confirm that. What
was happening then, if memory serves me correctly, was
that the unemployment rate was high, well above 5 per
cent, that prices were quite stable, and that our balanceof-payments deficits were intractable.
Furthermore, is it not so that the reliance on a fixedexchange system had the effect of making us minimize
domestic expansionary policies and when at last we saw
fit to pursue somewhat more expansionary policies than
we otherwise would have, this forced us to such exotic
manipulations in the money markets as Operation Twist
and exhortations to the private business community to
inhibit their investments abroad? This is hardly one of
the virtues of a fixed-exchange system; that is, it hardly
tends toward improvement in capital allocation inter
nationally.

154

T h e B a la n c e o f P a y m e n ts

DR. ROOSA: The answer to all of your questions is
that everything is relative and nothing is perfect.
But, first, on your point about the tax impact, actu
ally we don’t have just one monolithic corporation
which pays all the taxes. If, in an environment in which
taxes are increasing, there are some who have an oppor
tunity of getting a remission of tax by taking certain
action, they have every incentive to take the action and
the incentive that is indicated by the investment credit
is that of, hopefully, adding to the stock of productive,
creative capital in the economy.
So that I don’t see any problem of inconsistency there.
On the one side, given a composition of output which
is underway, there is a need for an incentive which will
exert some influence on the composition of that output
in the direction of increasing the productive capability
of the economy—a little more going toward investment,
within the total of all resources being used for all
purposes.
Given the total that is available for all purposes, con
siderably more also has to be devoted to meeting govern
ment expenses in a period when otherwise the aggregate
of demand would itself add to additional inflationary
pressures which, I assume, you wouldn’t want to see
for domestic reasons.
When translating any of this into balance-of-pay
ments terms, I think there is a built-in tendency to want
to "think simple.” I would enjoy doing so. But in this
area, perhaps because I’m defending what we have, I

is c u s s io n

155

can’t. There simply cannot be the direct and one-way
sort of a tradeoff that is implied in your formulation—
at least not often.
Now, the United States’ situation in the late fifties
was, of course, a special one. It didn’t fit any of the
books because we had a balance-of-payments deficit at
the same time that we had a very large current account
surplus. And still do. We, therefore, had to look for the
cause in the behavior of capital movements and trans
fers. The nature of the forces at work was such that if
there had been further reductions of interest rates, as
such, at that time, more money would have flowed out.
Actually, without reductions in short rates, indeed while
they were rising, there were continuing increases in the
availability and directed used of borrowed funds and it
proved possible to influence their allocation construc
tively, without worsening the deficit in our balance of
payments. If there had simply been the technique of
flooding out liquidity and then letting it spill wherever
it would go, the consequences would have been greater
outflows balance-of-payments wise. We might have been
able to sponge some of them up, as we did sponge up
much of what did flow anyway, but probably with
more difficulty.
We didn’t have to make that choice. There were other
ways of accomplishing what we wanted for the domestic
economy in an orderly manner and in a lasting way that
also held great promise for the balance of payments.
And here is an alibi if you have ever heard one, but

156

Ba la n ce

Pa y m

ents

in 1964, if it had not been for the capital outflow prob
lem which we had still not fully resolved (and which
had taken on new characteristics of its own) the re
sponse to the whole program initiated in 1961 and
worked out over the next three years, the response to
that whole program had by 1964, along with a lot of
good luck, produced a balance-of-payments surplus on
current account that could have covered all of our
outflows. As you remember, the current account surplus,
conservatively measured, was well over $7 billion. This
was achieved when the economy had, at the same time,
been rapidly expanding, the most rapid expansion we
had had in ages, and when unemployment was being
absorbed.
We devised a program that had some chance of work
ing in that environment toward the resolution of both
problems—unemployment at home and deficits abroad.
There were slippages along the way to be sure, but as a
conception of a program I still don’t have any regrets
about the way in which that was designed. I am in no
position, however, to be an apologist for the period that
just preceded it.
I would say that out of the experience of that pre
ceding period a good many lessons were learned and we
are still, I hope, all of us learning them. I doubt that it
will ever be possible to say that balance-of-payments
problems exist for the United States just because there
is something wrong with the rate of resource utilization
in the United States, as you put it. That may or may not

is c u s s io n

157

be what a balance-of-payments problem at any given
moment points to.
But certainly in the early sixties we did have a prob
lem of utilizing resources and we attempted to develop
a combination of measures that would not only increase
the resource utilization but would reduce the deficits.
While the deficits certainly have continued each year
due to some other proximate cause, the nature of the
problem as it existed in 1961 was not intractable and
the problem as it existed in 1961 was solved in 1964.
We live in a dynamic world and you can’t always
foresee all of the next year’s problems. But the new
problems then were essentially those on the capital side.
To say that a free-rate system, as of that time, when
capital was flowing out at a rapid rate, could have been
helpful is to me inconceivable.
Now, I know what Professor Friedman says. He said
it in his testimony before the Joint Economic Committee
in 1963. I was still down here, then, and I had to try to
imagine what his prescription would mean.
The outflow of capital, if it had been able to reflect
itself in a flexible rate, would have cut the United States
dollar exchange rate against other countries very sharply.
Exports from the United States would have become
much less costly to others, not just by pennies but (in
the magnitudes of that time) by dimes and quarters.
Other countries could not have tolerated massive in
creases in their imports and would have taken retaliatory
defensive measures. Meanwhile, with supplies of our

158

T h e B a la n c e o f P a y m e n ts

goods for export not wholly elastic, their domestic prices
in the United States would have begun running up—
disrupting our own domestic price stability.
How was that going to solve the problem that we
were confronting at that time? It seems to me it’s abso
lutely upside down. I still can’t see what it was he had
in mind but perhaps next week he’ll explain.
DANIEL EDWARDS, Joint Economic Committee:
I would like to follow up the ideas of Norm and Harvey.
As a side observation, the stock of money right now
is less than 1 percent above where it was last year and
although most of the econometric models that are being
used indicate the economy might roar in the last half of
this year, most of these models are not very sensitive to
money and credit conditions.
You hinted that a more elastic view of gold wouldn’t
lead to the end of the world, as many people have sug
gested. I am wondering what the pattern of events
would have been after 1958, if the United States had
gone off of gold internationally as well as domestically
in the thirties?
DR. ROOSA: There are a lot of different meanings
about going off gold. I’ll try to take that one quickly
first.
Bernard Baruch used to call me up every few weeks
in the days when we first had some thought of taking
off the gold cover back early—I’ve forgotten now, 1961
sometime — and he always referred to that as going off
gold and he was terribly worried about it.

isc u ss io n

159

The assumption that I am making about your statement is that you mean by “ going off gold” that we
simply ceased either to buy or sell.
MR. EDW ARDS: That is correct.
DR. ROOSA: And maintained no price.
MR. EDW ARDS: That’s correct.
DR. ROOSA: We would use it if we found a chance
to auction it off to meet a balance-of-payments deficit
sometime but it would have served no role in the world
monetary system.
MR. EDW ARDS: That’s right.
DR. ROOSA: In that case, we would now, in my
view, be living in a world in which Hjalmar Schacht
would be very much at home.
MR. EDW ARDS: But what do you think the se
quence of events would have been internationally after
1958?
DR. ROOSA: We just wouldn’t have had the con
vertibility change that occurred in 1958, there couldn’t
have been the move. The result then would have been
that much of what provided the currency environment
for the really fabulous expansion of internal and external
trade over the past decade— the fixed certainty of being
able to do business at known prices and reasonably lasting
values—would have been gone.
There would have been nothing that the world mone
tary system could hold to as an anchor for the fixed-rate
system. In time there would have been the evolution of
a dollar bloc and a good many countries would have

160

B a la n ce

Pa ym

ents

joined it. Within the dollar bloc we would be permitting
relatively free capital movement to those who were
members of it, and none, except perhaps through cen
tralized licensing, to those who were outside. As to what
would have happened to those outside, anybody could
conjure up a dozen schemes.
But the evolution of a dollar bloc itself, once we went
off gold, would just consist of those countries which,
in order to have some element of certainty, said they’d
be willing to set a' fixed exchange rate with the dollar,
that they would invoice their transactions with others
in the group in dollars, and stay with us, provided that
they would have access to our capital market.
This dollar bloc, I suppose, would have included all
of Latin America very quickly, and much of the less
developed part of the world, where the capacity for
developing capital needs is enormous. There would have
to be some scheme I think of priorities and queuing, as
has developed, as far as I know, in every other bloc
system that has been developed.
I am not just trying to create a horrendous picture
here. I am trying to give you a quick answer as to what
I think would have happened. That is why I was so
concerned to see that it didn’t happen when I had a
little chance to begin dealing with the question at the
end of I960’.
As you may remember, there was widespread expec
tation that it might happen as soon as President Kennedy
came in. This is why I didn’t go to anything more than

isc u s s io n

161

the swearing in on Inauguration Day and was on the
phone in touch with the gold market all that day. Then
subsequently, of course, over the weekend there was no
problem, and things settled down thereafter. But some
of the traders in critically sensitive parts of the world
were all ready for a real run on gold. They expected, in
that event, that we would have to split off into some
kind of a defensive dollar bloc system. The more gold
they had, the better their bargaining position would then
be because they could always sell it for something in the
trade they had with the United States. And it would give
them time to decide whether or not, at least, they wanted
to join such a bloc.
As to the other question that you asked, the more
immediate one, that the money supply is only up 1 per
cent above last year, the question of the choice of dates
is always a problem there. I am not close enough to the
details. I have the luxury of knowing so little that I can
have strong views.
I do feel that it is not likely that the rate of increase
that has prevailed since November can continue. That
rate, I think, and I may be wrong here, but I think at
an annual rate the increase in the money supply over
that period has been somewhere in the area approaching
5 percent,1 and I would think we have made enough
recovery now from that gushing so that the continuing
increase will be at a somewhat slower rate.
I would be surprised, though, if we ended up with
an increase in the money supply, strictly defined, of less

162

T h e B a la n c e o f P a y m e n ts

than somewhere in the 3 percent area on an average for
this year as a whole over last year.
From what I know, and it isn’t much, I think that
would be about right.
MR. EDWARDS: I agree with you. The problem is
that, if you follow it, the money supply has decreased
the last four weeks and, although that is too short a
period to set a trend, you can get concerned when the
administration is known to be looking at the econometric
models and believing that the explosion is going to take
place sometime in the third or fourth quarter.
You can interpret this recent contraction as evidence
that the Fed has already reversed policy in anticipation
of an inflation. It could be a case of again over-reacting.
DR. ROOSA: No, I doubt it. My hunch would be
that, even though this is a time in which the more rapid
rate of increase in money supply can appropriately slow
down, at the same time, with the prospect of what I
would regard as appropriate fiscal action, at least in the
wings, I would not think that this is a situation in which
the Fed needs to be thinking for balance-of-payments
reasons or any other reasons in restrictive terms.
Getting back to a position of nourishing a good
average growth rate, I would think, if it becomes a
problem of inflationary pressure, at this stage (and I
certainly can be wrong, these are things that need the
combined judgment of dozens of people all debating all
the time and on the basis of some knowledge, which I
don’t have) I would think that this will be the time

D is c u s s io n

163

when the emphasis ought to be on the general tax side.
I did promise a reply to this chap here—I know we’re
getting up to the time limit but I made a promise I had
better go through with.
PROFESSOR EDW ARD MURPHY, Georgetown
University: Thank you. I had the feeling that part of
your defense of fixed rates was based on some of the
problems associated with flexible rates. One of these had
to do with capital movements. I believe you suggested
that under the system of fixed rates capital tended to be
internationally distributed in a more economic way than
it would under flexible rates.
I am not so sure that this is immediately obvious but
I believe that the history of fixed rates, at least since
World War I, has been that governments have not really
permitted capital flows to be as large as they probably
would have in the absence of controls.
We see today a large number of controls, formal and
informal, being instituted by the United States govern
ment. The controls which exist in Europe, even today,
on capital flows are known. While we don’t restrict
private capital flow too much to underdeveloped coun
tries, I would assume we might if they got very large.
We certainly do restrict government capital flows to
underdeveloped countries through AID.
I really wonder the extent to which this is a viable
criticism of flexible rates. Perhaps in the private capital
flows which do take place under fixed rates, there is
some assurance of a future return, but isn’t the relevant

164

T h e B a la n c e o f P a y m e n ts

comparison really the capital flows that don’t take place
because of the controls?
DR. ROOSA: You are going to think that I can only
play one broken record, but the same answer is that we
have to make comparisons between degrees of inade
quacy rather than assume that either approach could
assume the total fulfillment of some of these objectives. I
think the key is in what you said about the fact that
with fixed rates there is a readier opportunity to make
the calculations on which a rational allocation of capital
around the world can be accomplished.
Now, the fact that governments interfere with this
is a fact of life that we have to deal with on its own.
This only illustrates that that system is better which is
at least geared to the assumption that governments will
stay out and not to the assumption that governments
must always be in the market, in a significant way. I
take a “flexible,” though not a “ free,” system to mean
that governments will always be jiggling or at least in
fluencing the current movement of exchange rates.,
My feeling is that we would have found the same
governmental aim of restricting capital movements
under any system. Governments were restricting capital
flows before they nominally went convertible. They de
fined convertibility in the Monetary Fund way back in
the war period as meaning only current account trans
actions. They never wanted to have capital move freely.
This is a big question. It’s built into a great proportion
of the history of investment and development around

isc u ss io n

165

the world.
And, in my view, the framework in which you have
a greater opportunity, at least, to urge, press, and plead
for some freeing of the governmental restrictions and
thereby allow the private allocation process to work, is
the fixed-rate system.
The fact that the United States has fallen from grace
in this respect after years of pleading with others to look
toward virtue is, without alibiing too much, the result
of the fact that even we weren’t big enough to go it
alone, completely alone, on the road of freedom for capi
tal movements, and what we have done has still very
carefully—with the exception of certain provisions to
favor the less developed countries, as you mentioned—
all been designed to retain the freedom of judgment of
the source of capital here as to where it should go, where
it can be most productive, and where the earnings can
be greatest.
You may not believe it, but that’s the same theory
on which the interest equalization tax was devised: that
government action would just raise the cost of putting
money in a foreign country and then, with the cost
generalized, it would be up to the investor to decide
where the investment goes, if it goes.
So that the whole notion of our restrictions, undesir
able and unfortunate as they are, has still been consistent
with the allocative principle that is possible under a fixedrate system. That is the allocation according to where
on the basis of a reasonable calculation, you can earn

166

B a la n ce

Pa y m

ents

the most over time. I don’t see anything wrong with
that if you believe in a market economy.
JAROSLAV HABR, Academy of Sciences, Prague,
Czechoslovakia: After hearing you and Professor Murphy
and Professor Milton Friedman and the gentleman who
raised the last question, I am rather puzzled.
DR. ROOSA: No wonder.
DR. HABR: It seems to me that the difference is
more in the meaning of the words fixed and flexible than
in reality. For your fixed rates are practically flexible
and Professor Friedman’s flexible rates are in reality fixed.
But I would like to know if the real point of difference
is that according to your view you would like to adjust
the internal situation between the development of econ
omies to the rate of exchange; whereas, according to the
view of Professor Friedman, the rates ought to be ad
justed to the development of individual economies.
DR. ROOSA: This is a very profound way of putting
a difference which is rather confused, as you first noted.
There are formulations that I have seen of Professor
Friedman’s that would correspond to the point you have
made, that he thinks in terms of letting the rate, the
external rate, adjust so that you can devote yourself
fully to the domestic economy.
In that sense I, perhaps unfairly, extend this in an
exaggerated way to say, if you accomplish that (and I
don’t think the United States can, but perhaps other
countries can) then you really are successfully using the
rate to build a wall of isolation around your own country.

isc u ss io n

167

It’s understandable that that could be chosen as an
aim of policy. But I am a multilateralist. I feel that the
distribution of resources around the world— always
somewhat haphazard and always lagging and subject to
the constraints you mention—but the distribution of
productive capacity and the selling of the goods pro
duced within any given country is likely to be greater,
that is, the flourishing of trade as a whole is likely to
be greater if you can, in effect, through the fixed ex
change rates at least make a first approximation toward
establishing all the world as your market.
And that will mean that there will be times when,
in order to adapt successfully in a lasting way to the
needs of making your own way in the other markets,
you have to make some changes at home, perhaps more
changes than would be necessary if you were isolated
and had a defensive wall of a flexible exchange rate
around your island.
But what I have tried to argue is that there will often
be value in making the adjustment in the home economy
that increases the capability to perform as a part of the
world economy. With the use of an ample supply of
reserves to tide over the period that the adjustments are
occurring, it becomes possible to make these adjustments
simply by changing the rate of increase in the domestic
economy. Such adjustments never have to imply en
forced unemployment, but they do require time. The
end result, of that kind of environment, as I see it, is a
wider area for the opportunities of freer trade.

168

T h e B a la n c e o f P a y m e n ts

Now, there are a good many other obstacles. Trying
to mesh the trade with the capital movements becomes
terribly complicated. But the objective that would be
set by a fixed-rate system is that each country removes
as it can, piece by piece, the other restrictions. Then it
can enjoy, the approximation that all are reaching for,
in a sense, that is, a free trade area where differences
among currencies do not matter.

THIRD SESSION
PROFESSOR JOSEPH ASCHEIM, George Washington University: I wonder what both speakers would
think of the idea of conducting an experiment with
respect to exchange rates for a duration to be specified
by both of them as an appropriate one to disclose sig
nificant results of this experiment?
PROFESSOR FRIEDM AN: I don’t really know how
to answer that. I am willing to take any period of time
that Dr. Roosa will give me.
DR* ROOSA: My answer is that I would just put it
on a computer and run it through. That’s the only one
I’d be willing to take.
HOWARD PIQUET, Library of Congress: I would
like to ask both gentlemen to comment on the band
proposal, the proposal for widening the bands on for
eign exchange rates as discussed primarily by Professor
George N. Halm. I address the question to both gentle
men with a thought that maybe this is a compromise
between the two positions.
PROFESSOR FRIED M A N : My position is simple.
I prefer a wide band to a narrow band and a wider band

170

h e

B a la n ce

Pa y m

ents

yet to a wide band. So that’s a movement in the correct
direction, although I am concerned that if you institute
it, there is a possible tendency to drift up to one side
of the band and then stick there.
DR. ROOSA: I haven’t any hard and fast view about
some widening of the band. I have thought that we
ought to do some experimenting in that direction. So
that would be part of the answer to your question,
[Professor Ascheim]. We have begun experimentation
with the few forward markets that exist by having the
forward rates get to a range well outside the band of
spot-rate fluctuation. As I recall, fluctuation has been
by an amount at the most equal to about 4 percent per
annum, in one case.
This is an elastic kind of relationship because as soon
as the spot rate is pulled down toward the lower limit
there has to be greater official support for the spot.
Then the forward rate can’t go any further. But this
is the kind of experiment that I think deserves further
testing, even though I am not conceding the principles
of flexible rates. This is like deciding how much room
to allow for expansion and contraction in a basically
rigid steel structure. You want to discover how much
you can safely allow to avoid the risk of cracking, while
still keeping the structure itself fixed.
Incidentally, I quite agree with Milton that if the
band were made very wide, the risk would be that rates
would move to one end or the other and there still would
not be room for variation. What you do have in such

is c u s s io n

171

circumstances is a clear signal that there should have
been a change in the parity. I think there may at some
time be a place for a wider band. But this would be
mainly as a means of getting a clearer indication of the
time when a change in the fixed rate should be made.
HERBERT STEIN , Committee for Economic De
velopment: Following up on that comment, under a
fixed-rate system would controls over capital movements,
exchange equalization tax, et cetera, be indications that
lead to a change in the rate?
DR. ROOSA: No, the problem that these present
restraints symbolize is the one that I have been hammer
ing at here so hard ad nauseum tonight, that it is ex
tremely difficult to get an equilibrium rate that takes
into account at the same time both the major forces at
work affecting capital movements and those affecting
goods in trade.
It’s particularly difficult because I suppose I was a
little too easy in granting too much too quickly to
Milton in this last exchange with respect even to the
trade adjustment. This is because, alongside the influ
ences of large and sustained capital movements, there are
actually inelasticities in either supply or demand for the
trade flows that are presumed to parallel the capital
movements. And these inelasticities make it impossible
to get the quick related adjustments in real terms, even
if everything else were done properly.
So that there will often be times, in a world of sub
stantial capital movement (and there will be more and

172

T h e B a la n c e o f P a y m e n ts

more capital flows as we become more multilateralized,
I hope) when divergence appears between the kind of
equilibrium that is indicated by the pattern of capital
flow over a given brief period—brief meaning two,
three, four years— and the pattern that would prevail
in purely trade terms.
I think we do have to live with some impediments of
this kind. Now, Milton’s answer is that a free rate has
a radar eye and can pick out which of the things to
adjust for and which not to adjust for. My feeling is
that nobody can do that. Under any system we are
going to have to have some selective arrangements when
the imbalances, particularly of this kind between trade
and capital flows, are not readily realizable, or readily
reconcilable in a single rate.
PROFESSOR FRIEDMAN: But it is precisely, of
course, the fact that it takes time to make the adjust
ments, that supply and demand are not perfectly elastic,
which is why you want to have a flexible rate rather
than a fixed rate. Let us suppose that there is a move
ment of capital that tends to come into a country and
the country cannot adjust its production and consump
tion pattern instantaneously. To begin with, the rate
will take on most of the burden. Then, as the adjust
ment is made, the rate will come back to that long-term
rate which is appropriate to a full adjustment.
The great virtue of a market is precisely that short
term movements in prices can fill in the gap between
the prior position and the new long-run position. The

is c u s s io n

173

notion that somehow or other the people administering
capital controls or voluntary exchange-rate restraints
are going to be able to do their job in any kind of a
delicate way is not a notion that you would find it very
easy to support.
This is sort of off the track a little, but in your last
interchange I realized all of a sudden that earlier I
didn’t give the right example in my remark about what
your comments reminded me of. I suddenly realized
that I was reminded of 1951 and the bond support peg
ging program.
Before the Federal Reserve gave up the pegging of
the bond price, we heard all over the lot that a free
market in bonds was going to be chaotic, that the in
terest rate might go heaven-high or down, there might
be capital losses, savings institutions might well be wiped
out by their capital losses, and that we needed some basic
peg price on which the market could form its anticipa
tion.
We abandoned the pegged price. None of these things
happened in the bond market. And it seems to me you
are doing the same thing over again. In each case—
DR. ROOSA: And, of course, the bond market has
been surprisingly stable without government interven
tion ever since!
PROFESSOR FRIEDM AN: Well, you have had a
good deal of Federal Reserve purchases and sales in the
bond market for monetary purposes. That has made the
bond market somewhat more unstable than it otherwise

174

h e

Ba la n ce

Pa y m

ents

would have been without the government intervention.
I’m glad to agree with you on that.
DR. ROOSA: You see, I didn’t say it.
PROFESSOR FRIEDM AN: No, I know, but the im
plication is there.
PROFESSOR GOTTFRIED HABERLER, Harvard
University: I should like to ask Dr. Roosa and Professor
Friedman a question.
My question for Dr. Roosa is this: Are you not worried
by the fact that we are getting into more and more
controls of international payments, especially controls
over capital movements? Only a few years ago you your
self used pretty strong language rejecting controls of
capital flows.
[Professor Haberler later supplied a paragraph from
a speech by Dr. Roosa in 1962 before the American
Bankers Association to support his recollection:
This country rejects direct controls on the flow
of capital not only because they would be incon
sistent with our traditional and fundamental ob
jectives of freeing trade and payments between
countries, but for immediate dollars-and-cents
reason—they would cost us more than they could
possibly save. Our own money and capital markets
are the most highly organized, most efficiently di
versified, of any in the world. To try to impose
controls over outward capital movements in any
one sector of these markets—say bank loans—
would only invite capital flight through many

D is c u s s io n

175

others, and to try instead a comprehensive ap
proach—clamping the cold hand of capital issues
controls, or credit rationing, over the entire sweep
of the markets—would literally congeal the blood
stream of American capitalism.]
Your prediction that controls of one segment of the
international capital market would "invite capital flight
through other channels, has been proved correct by
later events. The controls had to be extended to more
and more areas.
How do you feel about that today? Don’t you think
the maintenance of fixed exchanges in the face of a
continuing deficit has something to do with these dis
turbing developments?
DR. ROOSA: I am worried about a drift toward more
controls, and just to show you how foolhardy I am, I
am republishing several essays in which I have said that—
in a book that will be out in another few months. So
that I don’t shrink from the embarrassment or from
the implication.
I think what has happened has been the result of the
interaction both of events from outside, which have been
greater in impact than could have been visualized, and,
let’s face it, a comparative failure in the execution of
domestic economic policy over the past year or year and
a half.
To say that there have been these human failures is
not to say that the objectives should not remain the
same and that the system isn’t going to, all things wash

176

B a la n ce

Pa y m

ents

ing out, still be superior, because the same human beings
are going to be at work, whether Milton likes it or not,
manipulating the exchange rates of individual countries’
under a system in which the superficial conditions of
freedom or flexibility were accepted.
I grant that this is a regrettable trend, that it should
be reversed, that we have not yet accomplished the
reversal and though I don’t want to drag in a real red
herring, of course, there is something in the fact that we
are fighting a war and we don’t have any other form of
exchange control or internal restraint.
I don’t want to lean on that unfairly hard, but it does
have something to do with the continuation of this
whole array of unfortunate impediments to fully free
capital movements now.
I would just say that as of the year 1964 the results
of the other aspects of the domestic program, which
had been undertaken beginning in 1961, were good
enough that they produced a current account surplus
of nearly $8 billion. That should have been enough to
get away from the interest equalization tax and to make
unnecessary the voluntary restraints program, but it
wasn’t.
PROFESSOR HABERLER: I cannot resist asking
Dr. Roosa a second question. You speak of "manipula
tion” of the exchange rate under the system of flexible
exchange rates. You said earlier that there would be a
danger of "competitive devaluation.”
Would such policies not be the very opposite of the

is c u s s io n

177

freely floating-rate system? In other words, you are not
denying that the system may work, you are saying that
countries will not be prepared to let the system of freely
floating exchange rates operate.
DR. ROOSA: My argument is on two levels and I
just argue that both are important. On the first level,
as a matter of principle, I don’t accept what 90 percent,
at least, of the academic community in this country
apparently accepts, that the theoretical case for fluctuat
ing rates is a good case. I don’t accept it.
But, going beyond that, I say, even if the case were
good, even if I were going to grant you that this is
the way in which a free-rate system should function,
I go on to say that as a matter of hard fact indi
vidual monetary authorities, country by country, are
not going to stand by and take the impact any more
than they will take the impact of a deflationary policy
that would provide more effective and quicker correc
tion in the American economy.
They are going instead to temporize and find ways of
accomplishing this result, as I think they should, more
slowly, gradually, less precipitately. This pattern of ad
justment to new norms in the internal economy is some
thing, I think, that has to be taken as given, just as I
think it must be taken as given that no nation is going
to stand by and let its own exchange rate, in fact, be
moved by other countries.
PROFESSOR FRIEDM AN: But you say take the
impact. What impact? There isn’t any impact. The float

178

T h e B a la n c e o f P a y m e n ts

ing exchange rate prevents the impact of those things
that the country should not adapt to. You implicitly
assume each time that all of the factors affecting ex
change rates somehow or other hurt the individual coun
try or require some real adjustment in the individual
country. If what the exchange rate does is to make it
unnecessary for them to adjust their own economy to
external monetary changes, then what impact?
DR. ROOSA: This gets back to my monotonous point
that you can’t have it both ways. If all that happens is
that the rate adjusts to keep the status quo, that’s one
thing. If, on the other hand, a rate change means that
there is a change in the allocation of resources in both
countries on both sides, then there are internal conse
quences.
PROFESSOR FRIEDM AN: Right, right.
DR. ROOSA: And if these internal consequences are
to be allowed to work their way through, they will,
from time to time, intrude upon the objectives currently
being maintained for general economic policy. They will
cause unemployment in some areas and an increase in
employment somewhere else because the elasticities aren’t
perfect. You won’t get the labor moving. You will get
strain where the new demand is. You will get slack
where there isn’t demand.
You will then have domestic economic policy trying
to cushion the impact and, before you know it, what you
thought was being screened off by the change in rate has
in fact been conditioned, altered, and given a new dimen

isc u s s io n

179

sion by the fact that government, because of its other
responsibilities, has moved in. And in the same way,
before this even happens, the central bank, recognizing
what is coming with an exchange-rate change, sees that
in order to moderate the other kinds of policy action it
may be impelled to take internally later, it should first
move on the exchange rate.
This is exactly what happened in Canada during the
period of 1950 to 1962.1 was in Canada. I worked across
the exchanges with the Canadian authorities. I sat at the
trading desk with the officials of the Bank of Canada.
I know how they at times had to jiggle the interest rate
in order to maintain an appearance of relative stability
in the Canadian dollar-U.S. dollar rate.
And this is supposed to be the example of what a
flexible rate system can do. They had the ideal condi
tions: Nobody else in the world moving against them.
One country able to move the rate against the whole
world to its own advantage and they did it.
PROFESSOR FRIEDM AN: They hardly did it to
their own advantage.
DR. ROOSA: In the end, of course, what they had
to do was to recognize that the combination of structural
relations between them and us, and occasional human
failure in the execution of this set of maneuvers—human
failures that correspond to the kind you are criticizing
us for under the present system— the combination of
these two things led them to see that the only workable
arrangement was one under which they accepted some

180

B a la n ce

Pa y m

ents

norms and went to work under them and set a parity.
PROFESSOR FRIEDM AN: I find it hard to accept
the view that what happened had to happen and you
don’t really want to argue that either. You don’t want
to argue that every measure any country ever took was
necessary—
DR. ROOSA: No, I just wanted to learn from that
experience/
PROFESSOR FRIEDM AN: I do too. But let’s go
back for a moment. The existence of floating rates does
not guarantee good internal policy in any country. It
only permits it. It only facilitates it. So a country can
have floating rates and bad internal policy. Canada did.
Second, I am not trying to have it both ways. What
happens, each time you say that, is that you grant in
principle that there may be two kinds of adjustments.
Then you assume that there is only one kind and that
somehow—
DR. ROOSA: There’s only one rate.
PROFESSOR FRIEDM AN: You assume that there
are no adjustments of rates that offset unnecessary ad
justments. You assume that all adjustments of rates re
quire physical adjustments because implicitly what you
do is to assume that rates change only because of real
changes in conditions of international trade.
Now, insofar as they do, under a fixed rate you have
to make exactly the same kinds of adjustments to exactly
those same kinds of changes and you have to do it in a
way which will generate more difficult internal problems

is c u s s io n

181

than under a flexible rate. Because, if there are changes
in real forces of demand and supply, then to adjust to
them with a fixed rate, you have to move your whole
price level up and down.
Over and over again you implicitly assume that there
is an adjustment mechanism under fixed rates which
simply is not there. You state over and over again that
all you need to have is a differential pace of adjustment
and then you grant that we haven’t had it. We’ve had
exchange control and we’ve had one thing and another.
So I am not trying to have it both ways. What I am
trying to say is that precisely because countries have
committed themselves to internal policies of full employ
ment, the amount of adjustment they are willing to
undertake is limited. Let’s not waste that. Let’s reserve
that adjustment capacity for those real changes in inter
national trade which require adjustment capacity. Let’s
not fritter it away on forcing prices up and down,
on introducing exchange controls or capital controls,
because there have been monetary changes externally.
Finally, the argument for free rates is not that a coun
try can’t make a mistake, but that with free rates you
have a chain in which, if one link is weak, the other links
are not affected. If one country makes a mistake and goes
off haywire, it pays the price, it pays it rather quickly,
and it doesn’t force -unnecessary adjustments on other
countries.
A fixed-rate system has the characteristic that, so long
as it is maintained, a country which acts to disturb the

182

T h e B a la n c e o f P a y m e n ts

system can, for a considerable time, be free of the neces
sity of paying for its actions.
DR. ROOSA: Milton, I can take everything critical
that you said and say that the description applies in re
verse, that this is exactly what happens with my defini
tion of a flexible-rate system, which I insist must be one
in which you can’t assume that the central banks are
nonparticipants.
PROFESSOR FRIEDM AN: Let me ask you a differ
ent question, a wholly different question, if I may.
What kind of evidence would it require to persuade you
that the fixed-rate system is not a viable one? What
would have to happen in the world? How far would
exchange rates have to get out of line? How much ex
change control do you have to have? Can you conceive
of a sequence of events over the next ten years so that
ten years from now you would be willing to say, okay,
a fixed-rate system is not going to work?
DR. ROOSA: Well, it’s a pleasant invitation. I invite
you to make the contrary suggestion—what do you
think would make a fixed-rate system work, if it’s not
working already? But I simply cannot see how a flexiblerate system could work. I just am unable to give you
that much satisfaction, much as I would like to try.
PROFESSOR FRIEDM AN: I’m not asking you to say
how it can work. I am pursuing the line that Professor
Haberler raised. Professor Haberler said to you, you
believe in free trade and so do I. You believe in free capi
tal movements. You and I alike agree—

isc u s s io n

183

DR. ROOSA: Yes.
PROFESSOR FRIEDM AN: — that the development
of exchange controls and so on is congealing the blood of
capitalism. All right, now, following up on Professor
Haberler’s question, how far does the blood have to
congeal before you will be willing to say there is some
thing wrong with the system?
DR. ROOSA: I put it the other way around. If the
blood congeals, the same conditions of national commit
ment to national objectives that I have been talking
about as relevant here will lead not to a flexible-rate
system but to a system of organized trading blocs in
which the conditions of a fixed-rate system can be pre
served within a smaller geographical area.
Now, I think this would certainly be unfortunate for
the pattern of evolution in world trade and of economic
and political relations. But I cannot conceive of any
congealing that would reach a stage where the world
and the people that I know in it are going to turn, in
fact, to a flexible system because the effects of a flexible
system are clear enough. They are going to want to pre
serve, within whatever area they can make coherent, the
advantages of the fixed-rate system within that area.
PROFESSOR FRIEDM AN: But then you are really
evading the issue, Bob.
DR. ROOSA: No, I am just giving you the only
alternative.
PROFESSOR FRIEDM AN: No, you’re not, because
whether they turn to the flexible-rate system or not

184

Ba la n ce

Pa y m

ents

depends on whether people like you are willing to look
at it clearly and willing to face up to the issues of a flexible-rate system and answer each case not by the state
ment, "We won’t have it,” but rather ask yourself the
question, "How would it in fact operate?” Whether
they turn toward a flexible rate or toward the narrow
regional blocs will depend in large part on what people
like you urge and recommend them to do.
And to say I am not even going to look and examine
and study in detail how a flexible-rate system will work
because no country will adopt it—
DR. ROOSA: All I am trying to say is that, having at
least looked at a flexible rate long enough to engage in
this little debate, the look that I have taken convinces
me that it is not either theoretically sound or operation
ally practicable and therefore I ask myself, What does
happen if the present system breaks down and it can’t
be continued? What is the alternative?
PROFESSOR FRIEDM AN: What do you mean by
the statement it is not theoretically sound? That seems
to use the word "theory” in a different sense than I un
derstand it.
DR. ROOSA: Conceptually.
PROFESSOR FRIEDM AN: I don’t understand what
you mean. I have read your paper. There is not a word
in there which can be interpreted as saying that a system
of floating exchange rates is not conceptually sound.
DR. ROOSA: There are really two principal points
on which it hinges and then I suppose a number of other

isc u ssio n

185

ancillary ones. The first is this point that we have labored
pretty hard, the assumption that adjustment will occur
instantaneously, that countries can accept it as an in
stantaneous adjustment and that there is an elasticity on
both sides which permits this to occur without strain.
PROFESSOR FRIED M A N : But those are not valid
points of a floating-rate system. The second of them is a
political judgment. So let’s leave it out for a moment.
DR. ROOSA: All right.
PROFESSOR FRIEDM AN: We want to know
whether it’s theoretically sound. Do you deny that the
market will set a price?
DR. ROOSA: I deny that an actual market will exist.
PROFESSOR FRIEDM AN: You deny that a market
will exist in exchange?
DR. ROOSA: I do, yes.
PROFESSOR FRIEDM AN: In foreign exchanges?
DR. ROOSA: You will find, if all countries have no
fixed parity, that instead, because there isn’t a real going
and lasting market, the relationships that will begin to
develop will be the kinds which will lead to the creation
of the bloc system that I have described; fixed rates
within each bloc, and barter among them.
PROFESSOR FRIEDM AN: You don’t believe that
there is a securities market in the world? You don’t be
lieve that there is an international stock market and
you don’t believe there are bond markets and you don’t
believe there are markets for commodities?
DR. ROOSA: Indeed I do. My partners and I trade

186

h e

Ba la n ce

Pa y m

ents

in. most of them every day—at fixed rates of exchange.
But the difference between all of these things and money
is that the money which has to represent the composite
of all these things, the numeraire in which they all have
to be measured, has to have some capability of stability.
That is the capability which you also want. Without it,
the drive of every organized society is going to be to
find that stability.
PROFESSOR HABERLER: My question, addressed
to Professor Friedman, is this: I was surprised when you
said at the beginning of your speech that you would be
in favor of fixed exchange rates, if we have a real gold
standard. By that you meant, I believe, that the quantity
of money was fixed for the whole world.
PROFESSOR FRIEDM AN: No. What I meant was
a unified currency system among the nations of the
world in which there was a single money, gold, which
had different names in different countries. It might be
called the dollar in the United States. But a dollar was
defined as one thirty-fifth of an ounce of gold and a
pound was defined as so much, so that you had in the
world as a whole the equivalent of what we have among
the different states in this country.
PROFESSOR HABERLER: Does that not mean that
the quantity of money, in terms of gold, is fixed for the
whole world?
PROFESSOR FRIEDM AN: Oh, yes, indeed. I mis
understood you. I thought you meant by fixed that it is
constant.

isc u s sio n

187

PROFESSOR HABERLER: Then let me ask this
question: Suppose under such a monetary arrangement
one part of the world expands faster than the other.
Would that not mean that the fast expanding area would
drain money away from the slowly expanding area and
would inflict a painful deflation on the latter? Would it
not be preferable to have flexible rates?
PROFESSOR FRIEDM AN: Yes, it would.
PROFESSOR H ABERLER: So, in that case you
would probably again drop the fixed rate.
PROFESSOR FRIEDM AN: As a purely technical
matter, there is no doubt that floating exchange rates
would be preferable but I do believe that the political
argument for a unified currency is a valid argument. If
you have a real gold standard, I would argue that the
adjustment would be so gradual and slow that you would
not in fact have a major real problem of the kind that
you are suggesting.
PROFESSOR HABERLER: This is a factual assump
tion which does not happen. Half of the world expands
at a very fast rate and the other half does not.
PROFESSOR FRIEDM AN: You see, my view, in
contrast to Bob’s is that what central banks do in prac
tice is not solely to smooth out minor movements. They
do that sometimes but what they often do is to prevent
adjustments which would be minor and which instead
accumulate into major problems. Even if nine times
out of ten when they smooth out a minor movement
they prevent an unnecessary adjustment, the one time

188

T h e B a la n c e o f P a y m e n ts

they make a mistake, they convert what otherwise
would have been a minor problem of adjustment into
a sizable problem.
My hunch is that under a real gold standard, because
adjustments would be set in train very gradually, be
cause a slight discrepancy between two countries would
produce a slight adjustment, the problem would be what
it now is among the different areas of the United States,
where major discrepancies do not accumulate. Maybe
I am wrong on that but that was the assumption under
which I was operating.
JAROSLAV HABR, Academy of Sciences, Prague,
Czechoslovakia: I have listened to both discussions, both
lectures, and I would like to express my thanks for this
unique opportunity for someone who is coming from
Central Europe. But, in one sense, it was not a unique
experience. It seemed to me sometimes that I was listen
ing to our top, top planner and our top, top reformer.
And you know who is who.
It is relatively easy to make a conceptual framework
either for the present existing world or for some future
world, but it is always very difficult to create a concep
tual framework for a transitional period. It appears to
me to be difficult for someone who believes in the present
world not to be rather skeptical about the new construc
tions of the future. And on the other side, those who
are making the constructions are perhaps underrating
the programs of the representatives of the present world.
DR. ROOSA: We do know each who we are.

D is c u s s io n

189

PROFESSOR FRIED M AN : I will get even with him
later for calling me a reformer.
PROFESSOR FR A N K TAM AGNA, American Uni
versity: I have been listening to you with some surprise,
because one of the things you are saying is that once the
adjustments have been made, either of the systems would
work.
PROFESSOR FRIEDM AN: I think the basic differ
ence, Professor Tamagna, is whether, if countries do not
follow the right policy, the burden then falls on the
country which has made a mistake or it is spread
throughout the system and so disturbs the whole system.
In my opinion, the great virtue of a world in which
exchange rates are free to vary but in which countries
are able to follow stable policies so the rates are in fact
relatively stable, is that a country which departs from
the rules itself pays the price and, therefore, the chain
is as strong as its strongest link.
The defect of the fixed exchange-rate system on the
level that you are raising is that if a country does not
follow what we regard as the correct policy it can, for
a considerable time, shift the burden onto the other
countries. This produces difficulties in the whole system
so that the chain is only as strong as its weakest link.
DR. ROOSA: On the same point, we have just
reached the stage where I think Milton has made his
most telling thrust at the system I am defending. I recog
nize that the greatest defect, and I didn’t have the
temerity to advertise it on my own, is the asymmetry

190

Ba la n ce

Pa y m

ents

of this impact on deficit and surplus countries. Surplus
countries can, if they are so inclined, get off easy and
shift most of the burden of correction onto the deficit
countries.
I think this is part of a fundamental defect which does
require considerable additional effort both to devise
norms of behavior and to impose them through agreed
means. I grant that this is the area in which the most
needs to be done and where the present system is, in fact,
weakest.
On the other side, I do not feel that you can say that
a choice between the systems becomes a matter of indif
ference. It is important to see that the pattern of reac
tions that is created, once the system is in being, the pat
tern of reactions that is created on the part of the pri
vate sector that engages in the trade and provides the
investment, will be that which has the greatest assurance
of maximizing the growth and the rational distribution
of resources around the world.
Now, if it were true that Milton’s system would always
produce stable rates, then I think we would have some
thing close to the equivalent of a single currency in the
world and we would get the distribution of resources
that is ideal. My difference hinges at least in part but
importantly on the fact that I do not believe that the
conditions of a flexible-rate system will make it possible
for the ordinary merchant and trader and banker and
investor to have the conditions of reasonable assurance,
the benchmarks for making choices, the reasonable sta-

is c u s s io n

191

bility of conditions for committing resources, that is
necessary to get the greatest result.
With all its imperfections, and granting that, as I say,
I think Milton has put his finger on what is a major
defect of a fixed-rate system, I still feel that in the total
result the gain, because of its influence on the private
commercial sector, is greater and will continue to be
greater under a fixed-rate system. I do think that the
flourishing of trade and capital flows that we have seen,
with all of the troubles, in the last decade since converti
bility was restored, provides some evidence to support
my argument.
PROFESSOR FRIEDM AN: But those conditions,
Bob, under which floating rates will not be stable, are
precisely the conditions under which your fixed rates
can be maintained only by exchange controls, interfer
ences with trade, and so on. And these have the same dis
ruptive effects, I would say much worse ones, because
at least if floating rates aren’t stable the countries that go
off half-cocked are the ones that bear the burden.
DR. ROOSA: Not always. But partly my opposition
comes also because I think the mere existence of the
flexible system creates an environment, as I say, for de
stabilizing speculation at times. But on that we really
can’t agree. We just identify our differences.
PROFESSOR FRIEDM AN: I should say that I once
wrote an essay entitled "In Defense of Destabilizing
Speculation,” because I am willing to go so far as to say
that in point of fact if there is destabilizing speculation

192

T h e B a l a n c e o f Pa y m e n t s

there is nothing wrong with that either. That simply
means that the speculators are making a gift to the
countries involved. If there is destabilizing speculation,
speculators lose. Who gains? Essentially the citizens of
the two countries who provide gambling services to the
speculators.
DR. ROOSA: On that difference I am afraid we will
have to rest. As I said in the beginning, I will leave with
even more respect, if that’s possible, for your capability
of making any case plausible and persuasive and I will
undertake to study a little further the offer that you have
given me to find the conditions under which a flexible
system could possibly be considered. There has to be a
presumption, I confess, that such a brilliant jockey could
not have chosen a horse as poor as the one I think I see.

FOOTNOTES
SECO N D L E C T U R E
Page 66
1 My own appraisal of the issues to be resolved in providing an acceptable supple
ment to gold through multi-national creation of a new reserve asset will appear
in The Dollar and World Liquidity, scheduled for publication by Random House
in September, 1967.

DISCUSSION
SECO N D SESSION
Page 161
1 Including time deposits the rise was about 10 percent, but not all time deposits
can be considered a part of the active money supply.

Full text of The Balance of Payments: Free Versus Fixed Exchange Rates

FRASER