Chapter IV

International Financial and Trade Arrangements

The problem of international monetary arrangements is the relation among different national currencies: the terms and conditions under which individuals are able to convert U.S. dollars to pounds sterling, Canadian dollars to U.S. dollars, and so on. This problem is closely connected with the control of money discussed in the preceding chapter. It is connected also with governmental policies about international trade, since control over international trade is one technique for affecting international payments.

Despite its technical character and forbidding complexity, the subject of international monetary arrangements is one that a liberal cannot afford to neglect. It is not too much to say that the most serious short-run threat to economic freedom in the United States today -- aside, of course, from the outbreak of World War III -- is that we shall be led to adopt far-reaching economic controls in order to "solve" balance of payments problems. Interferences with international trade appear innocuous; they can get the support of people who are otherwise apprehensive of interference by government into economic affairs; many a business man even regards them as part of the "American Way of Life"; yet there are few interferences which are capable of spreading so far and ultimately being so destructive of free enterprise. There is much experience to suggest that the most effective way to convert a market economy into an authoritarian economic society is to start by imposing direct controls on foreign exchange. This one step leads inevitably to the rationing of imports, to control over domestic production that uses imported products or that produces substitutes for imports, and so on in a never-ending spiral. Yet even so generally staunch a champion of free enterprise as Senator Barry Goldwater has at times been led, when discussing the so-called "gold flow," to suggest that restrictions on transactions in foreign exchange may be necessary as a "cure." This "cure" would be vastly worse than the disease.

There is seldom anything truly new under the sun in economic policy, where the allegedly new generally turns out to be the discard of a prior century in flimsy disguise. Unless I am mistaken, however, full-fledged exchange controls and so-called "inconvertibility of currencies" are an exception and their origin reveals their authoritarian promise. To the best of my knowledge they were invented by Hjalmar Schacht in the early years of the Nazi regime. On many occasions in the past, of course, currencies have been described as inconvertible. But what the word then meant was that the government of the day was unwilling or unable to convert paper currency into gold or silver, or whatever the monetary commodity was, at the legally stipulated rate. It seldom meant that a country prohibited its citizens or residents from trading pieces of paper promising to pay specified sums in the monetary unit of that country for corresponding pieces of paper expressed in the monetary unit of another country -- or for that matter for coin or bullion. During the Civil War in the United States and for a decade and a half thereafter, for example, U.S. currency was inconvertible in the sense that the holder of a greenback could not turn it in to the Treasury and get a fixed amount of gold for it. But throughout the period he was free to buy gold at the market price or to buy and sell British pounds for U.S. greenbacks at any price mutually agreeable to the two parties. In the United States, the dollar has been inconvertible in the older sense ever since 1933. It has been illegal for American citizens to hold gold or to buy and sell gold. The dollar has not been inconvertible in the newer sense. But unfortunately we seem to be adopting policies that are highly likely, sooner or later, to drive us in that direction.

Only a cultural lag leads us still to think of gold as the central element in our monetary system. A more accurate description of the role of gold in U.S. policy is that it is primarily a commodity whose price is supported, like wheat or other agricultural products. Our price-support program for gold differs in three important respects from our price-support program for wheat: first, we pay the support price to foreign as well as domestic producers; second, we sell freely at the support price only to foreign purchasers and not to domestic; third, and this is the one important relic of the monetary role of gold, the Treasury is authorized to create money to pay for gold it buys -- to print paper money as it were -- so that expenditures for the purchase of gold do not appear in the budget and the sums required need not be explicitly appropriated by Congress; similarly, when the Treasury sells gold, the books show simply a reduction in gold certificates, and not a receipt that enters into the budget.

When the price of gold was first set at its present level of $35 an ounce in 1934, this price was well above the free market price of gold. In consequence, gold flooded the United States, our gold stock tripled in six years, and we came to hold well over half the world's gold stock. We accumulated a "surplus" of gold for the same reason we accumulated a "surplus" of wheat -- because the government offered to pay a higher price than the market price. More recently, the situation has changed. While the legally fixed price of gold has remained $35, prices of other goods have doubled or tripled. Hence $35 is now less than what die free market price would be.¹ As a result, we now face a "shortage" rather than a "surplus" for precisely the same reason that rent ceilings inevitably produce a "shortage" of housing -- because the government is trying to hold the price of gold below the market price.

The legal price of gold would long since have been raised -- as wheat prices have been raised from time to time -- except for the accident that the major producers of gold, and hence the major beneficiaries from a rise in its price, are Soviet Russia and South Africa, the two countries with whom the United States has least political sympathy.

Governmental control of the price of gold, no less than the control of any other price, is inconsistent with a free economy. Such a pseudo gold standard must be distinguished sharply from the use of gold as money under a real gold standard which is entirely consistent with a free economy though it may not be feasible. Even more than the price fixing itself, the associated measures taken in 1933 and 1934 by the Roosevelt administration when it raised the price of gold represented a fundamental departure from liberal principles and established precedents that have returned to plague the free world. I refer to the nationalization of the gold stock, the prohibition of private possession of gold for monetary purposes, and the abrogation of gold clauses in public and private contracts.

In 1933 and early 1934, private holders of gold were required by law to turn over their gold to the federal government. They were compensated at a price equal to the prior legal price, which was at the time decidedly below the market price. To make this requirement effective, private ownership of gold within the U.S. was made illegal except for use in the arts. One can hardly imagine a measure more destructive of the principles of private property on which a free enterprise society rests. There is no difference in principle between this nationalization of gold at an artificially low price and Fidel Castro's nationalization of land and factories at an artificially low price. On what grounds of principle can the U.S. object to the one after having itself engaged in the other? Yet so great is the blindness of some supporters of free enterprise with respect to anything touching gold that in i960 Henry Alexander, head of the Morgan Guaranty Trust Company, successor to J. P. Morgan and Company, proposed that the prohibition against the private ownership of gold by U.S. citizens be extended to cover gold held abroad! And his proposal was adopted by President Eisenhower with hardly a protest from the banking community.

Though rationalized in terms of "conserving" gold for monetary use, prohibition of private ownership of gold was not enacted for any such monetary purpose, whether itself good or bad. The nationalization of gold was enacted to enable the government to reap the whole of the "paper" profit from the rise in the price of gold -- or perhaps, to prevent private individuals from benefiting.

The abrogation of the gold clauses had a similar purpose. And this too was a measure destructive of the basic principles of free enterprise. Contracts entered into in good faith and with full knowledge on the part of both parties to them were declared invalid for the benefit of one of the parties!

In discussing international monetary relations on a more general level, it is necessary to distinguish two rather different problems: the balance of payments, and the danger of a run on gold. The difference between the problems can be illustrated most simply by considering the analogy of an ordinary commercial bank. The bank must so arrange its affairs that it takes in as service charges, interest on loans, and so on a large enough sum to enable it to pay its expenses -- wages and salaries, interest on borrowed funds, cost of supplies, returns to stockholders, and so on. It must strive, that is, for a healthy income account. But a bank which is in good shape on its income account may nonetheless experience serious trouble if for any reason its depositors should lose confidence in it and suddenly demand their deposits en masse. Many a sound bank was forced to close its doors because of such a run on it during the liquidity crises described in the preceding chapter.

These two problems are not of course unrelated. One important reason why a bank's depositors may lose confidence in it is because the bank is experiencing losses on income account. Yet the two problems are also very different. For one thing, problems on income account are generally slow to arise and considerable time is available to solve them. They seldom come as sudden surprises. A run, on the other hand, may arise suddenly and unpredictably out of thin air.

The situation of the U.S. is precisely parallel. Residents of the United States and the U.S. government itself are seeking to buy foreign currencies with dollars in order to purchase goods and services in other countries, to invest in foreign enterprises, to pay interest on debts, to repay loans, or to give gifts to others, whether private or public. At the same time, foreigners are seeking to acquire dollars with foreign currencies for corresponding purposes. After the event, the number of dollars spent for foreign currency will precisely equal the number of dollars purchased with foreign currency -- just as the number of pairs of shoes sold is precisely equal to the number bought. Arithmetic is arithmetic and one man's purchase is another man's sale. But there is nothing to assure that, at any given price of foreign currency in terms of dollars, the number of dollars that some want to spend will equal the number others want to buy -- just as there is nothing to assure that at any given price of shoes the number of pairs of shoes people want to buy is exactly equal to the number of pairs other people want to sell. The ex post equality reflects some mechanism that eliminates any ex ante discrepancy. The problem of achieving an appropriate mechanism for this purpose is the counterpart of the bank's problem on income account.

In addition, the United States has a problem like the bank's of avoiding a run. The U.S. is committed to sell gold to foreign central banks and governments at $35 an ounce. Foreign central banks, governments, and residents hold large funds in the United States in the form of deposit accounts or U.S. securities that can be readily sold for dollars. At any time, the holders of these balances can start a run on the U.S. Treasury by trying to convert their dollar balances into gold. This is precisely what happened in the fall of i960, and what is very likely to happen again at some unpredictable date in the future (perhaps before this is printed).

The two problems are related in two ways. In the first place, as for a bank, income account difficulties are a major source of loss of confidence in the ability of die U.S. to honor its promise to sell gold at $35 an ounce. The fact that the U.S. has in effect been having to borrow abroad in order to achieve balance on current account is a major reason why holders of dollars are interested in converting them into gold or other currencies. In the second place, the fixed price of gold is the device we have adopted for pegging another set of prices -- the price of the dollar in terms of foreign currencies -- and flows of gold are the device we have adopted for resolving ex ante discrepancies in the balance of payments.

We can get more light on both of these relations by considering what alternative mechanisms are available for achieving balance in payments -- the first and in many ways the more fundamental of the two problems.

Suppose that the U.S. is roughly in balance in its international payments and that something comes along which alters the situation by, let us say, reducing the number of dollars that foreigners want to buy compared to the number that U.S. residents want to sell; or, looking at it from the other side, increasing the amount of foreign currency that holders of dollars want to buy compared to the amount that holders of foreign currency want to sell for dollars. That is, something threatens to produce a "deficit" in U.S. payments. This might result from increased efficiency in production abroad or decreased efficiency at home, increased foreign aid expenditures by the U.S. or reduced ones by other countries, or a million and one other changes of the kind that are always occurring. There are four, and only four ways, in which a country can adjust to such a disturbance and some combination of these ways must be used.

1. U.S. reserves of foreign currencies can be drawn down or foreign reserves of U.S. currency built up. In practice, this means that the U.S. government can let its stock of gold go down, since gold is exchangeable for foreign currencies, or it can borrow foreign currencies and make them available for dollars at official exchange rates; or foreign governments can accumulate dollars by selling U.S. residents foreign currencies at official rates. Reliance on reserves is obviously at best a temporary expedient. Indeed, it is precisely the extensive use by the U.S. of this expedient that accounts for the great concern with the balance of payments.

2. Domestic prices within the U.S. can be forced down relative to foreign prices. This is the main adjustment mechanism under a full-fledged gold standard. An initial deficit would produce an outflow of gold (mechanism 1, above); the outflow of gold would produce a decline in the stock of money; the decline in the stock of money would produce a fall in prices and incomes at home. At the same time, the reverse effects would occur abroad: the inflow of gold would expand the stock of money and thereby raise prices and income. Lowered U.S. prices and increased foreign prices would make U.S. goods more attractive to foreigners and thereby raise the number of dollars they wanted to buy; at the same time, the price changes would make foreign goods less attractive to U.S. residents and thereby lower the number of dollars they wanted to sell. Both effects would operate to reduce the deficit and restore balance without the necessity for further gold flows.

Under the modern managed standard, these effects are not automatic. Gold flows may still occur as the first step, but they will not affect the stock of money in either the country that loses, or the country that gains gold, unless the monetary authorities in the separate countries decide that they should. In every country today, the central bank or the Treasury has the power to offset die influence of gold flows, or to change the stock of money without gold flows. Hence this mechanism will be used only if the authorities in the country experiencing the deficit are willing to produce a deflation, thereby creating unemployment, in order to resolve its payments problem, or the authorities in the country experiencing the surplus are willing to produce an inflation.

3. Exactly die same effects can be achieved by a change in exchange rates as by a change in domestic prices. For example, suppose that under mechanism 2 the price of a particular car in the United States fell by 10 per cent from $2,800 to $2,520. If the price of the pound is throughout $2.80, this means that the price in Britain (neglecting freight and other charges) would fall from £1,000 to £900. Exactly the same decline in the British price will occur without any change in the United States price if the price of a pound rises from $2.80 to $3.11. Formerly, the Englishman had to spend £1,000 to get $2,800. Now he can get $2,800 for only £900. He would not know the difference between this reduction in cost and the corresponding reduction through a fall in the U.S. price without a change in exchange rate.

In practice, there are several ways in which the change in exchange rates can occur. With the kinds of pegged exchange rates many countries now have, it can occur through devaluation or appreciation, which is to say, a governmental declaration that it is changing the price at which it proposes to peg its currency. Alternatively, the exchange rate does not need to be pegged at all. It can be a market rate changing from day to day, as was the case with the Canadian dollar from 1950 to 1962. If a market rate, it can be a truly free market rate determined primarily by private transactions as the Canadian rate apparently was from 1952 to 1961, or it can be manipulated by government speculation as was the situation in Britain from 1931 to 1939, and in Canada from 1950 to 1952 and again from 1961 to 1962.

Of these various techniques, only the freely floating exchange rate is fully automatic and free from governmental control.

4. The adjustments produced by mechanisms 2 and 3 consist of changes in flows of commodities and services induced by changes either in internal prices or exchange rates. Instead, direct governmental controls or interferences with trade could be used to reduce attempted U.S. expenditures of dollars and expand U.S. receipts. Tariffs could be raised to choke off imports, subsidies could be given to stimulate exports, import quotas could be imposed on a variety of goods, capital investment abroad by U.S. citizens or firms could be controlled, and so on and on up to the whole paraphernalia of exchange controls. In this category must be included not only controls over private activities but also changes in governmental programs for balance of payments purposes. Recipients of foreign aid may be required to spend the aid in the U.S.; the military may procure goods in the United States at greater expense instead of abroad in order to save "dollars" -- in the self-contradictory terminology used -- and so on in bewildering array.

The important thing to note is that one or another of these four ways will and must be used. Double entry books must balance. Payments must equal receipts. The only question is how.

Our announced national policy has been and continues to be that we shall do none of these things. In a speech in December 1961, to the National Association of Manufacturers, President Kennedy stated "This administration, therefore, during its term of office -- and I repeat this and make it as a flat statement -- has no intention of imposing exchange controls, devaluing the dollar, raising trade barriers or choking off our economic recovery." As a matter of logic, this leaves only two possibilities: getting other countries to take the relevant measures, which is hardly a recourse that we can be sure of, or drawing down reserves, which the President and other officials repeatedly stated must not be permitted to continue. Yet Time magazine reports that the President's "promise drew a burst of applause" from the assembled businessmen. So far as our announced policy is concerned, we are in the position of a man living beyond his income who insists that he cannot possibly earn more, or spend less, or borrow, or finance the excess out of his assets!

Because we have been unwilling to adopt any one coherent policy, we and our trading partners -- who make the same ostrich-like pronouncements as we do -- have perforce been led to resort to all four mechanisms. In the early postwar years, U.S. reserves rose; more recently they have been declining. We welcomed inflation more readily than we otherwise would have when reserves were rising, and we have been more deflationary since 1958 than we would otherwise have been because of the drain of gold. Though we have not changed our official price of gold, our trading partners have changed theirs, and thereby the exchange rate between their currency and the dollar, and U.S. pressure has not been absent in producing these adjustments. Finally, our trading partners used direct controls extensively and, since we instead of they have been faced with deficits, we too have resorted to a wide range of direct interferences with payments, from reducing die amount of foreign goods that tourists can bring in free of duty -- a trivial yet highly symptomatic step -- to requiring foreign aid expenditures to be spent in the U.S., to keeping families from joining servicemen overseas, to more stringent import quotas on oil. We have been led also to engage in the demeaning step of asking foreign governments to take special measures to strengthen the U.S. balance of payments.

Of the four mechanisms, die use of direct controls is clearly die worst from almost any point of view and certainly the most destructive of a free society. Yet in lieu of any clear policy, we have been led increasingly to rely on such controls in one form or another. We preach publicly the virtues of free trade; yet we have been forced by the inexorable pressure of the balance of payments to move in the opposite direction and there is great danger that we shall move still farther. We can pass all the laws imaginable to reduce tariffs; the Administration may negotiate any number of tariff reductions; yet unless we adopt an alternative mechanism for resolving balance of payments deficits, we shall be led to substitute one set of trade impediments for another -- indeed, to substitute a worse set for a better. While tariffs are bad, quotas and other direct interferences are even worse. A tariff, like a market price, is impersonal and does not involve direct interference by government in business affairs; a quota is likely to involve allocation and other administrative interferences, besides giving administrators valuable plums to pass out to private interests. Perhaps worse than either tariffs or quotas are extra-legal arrangements, such as the "voluntary" agreement by Japan to restrict textile exports.

There are only two mechanisms that are consistent with a free market and free trade. One is a fully automatic international gold standard. This, as we saw in the preceding chapter, is neither feasible nor desirable. In any event, we cannot adopt it by ourselves. The other is a system of freely floating exchange rates determined in the market by private transactions without governmental intervention. This is the proper free-market counterpart to the monetary rule advocated in the preceding chapter. If we do not adopt it, we shall inevitably fail to expand the area of free trade and shall sooner or later be induced to impose widespread direct controls over trade. In this area, as in others, conditions can and do change unexpectedly. It may well be that we shall muddle through the difficulties that are facing us as this is written (April, 1962) and indeed that we may find ourselves in a surplus rather than deficit position, accumulating reserves rather than losing them. If so, this will only mean that other countries will be faced with the necessity of imposing controls. When, in 1950,1 wrote an article proposing a system of floating exchange rates, it was in the context of European payments difficulties accompanying the then alleged "dollar shortage." Such a turnabout is always possible. Indeed, it is the very difficulty of predicting when and how such changes occur that is the basic argument for a free market. Our problem is not to "solve" a balance of payments problem. It is to solve the balance of payments problem by adopting a mechanism that will enable free market forces to provide a prompt, effective, and automatic response to changes in conditions affecting international trade.

Though freely floating exchange rates seem so clearly to be die appropriate free-market mechanism, they are strongly supported only by a fairly small number of liberals, mostly professional economists, and are opposed by many liberals who reject governmental intervention and governmental price-fixing in almost every other area. Why is this so? One reason is simply the tyranny of the status quo. A second reason is the confusion between a real gold standard and a pseudo gold standard. Under a real gold standard, the prices of different national currencies in terms of one another would be very nearly rigid since the different currencies would simply be different names for different amounts of gold. It is easy to make the mistake of supposing that we can get the substance of the real gold standard by the mere adoption of the form of a nominal obeisance to gold -- the adoption of a pseudo gold standard under which the prices of different national currencies in terms of one another are rigid only because they are pegged prices in rigged markets. A third reason is the inevitable tendency for everyone to be in favor of a free market for everyone else, while regarding himself as deserving of special treatment. This particularly affects bankers in respect of exchange rates. They like to have a guaranteed price. Moreover, they are not familiar with the market devices that would arise to cope with fluctuations in exchange rates. The firms that would specialize in speculation and arbitrage in a free market for exchange do not exist. This is one way the tyranny of the status quo is enforced. In Canada, for example, some bankers, after a decade of a free rate which gave them a different status quo, were in the forefront of those favoring its continuation and objecting to either a pegged rate or government manipulation of the rate.

More important than any of these reasons, I believe, is a mistaken interpretation of experience with floating rates, arising out of a statistical fallacy that can be seen easily in a standard example. Arizona is clearly the worst place in the U.S. for a person with tuberculosis to go because the death rate from tuberculosis is higher in Arizona than in any other state. The fallacy is in this case obvious. It is less obvious in connection with exchange rates. When countries have gotten into severe financial difficulties through internal monetary mismanagement or for any other reason, they have had ultimately to resort to flexible exchange rates. No amount of exchange control or direct restrictions on trade enabled them to peg an exchange rate that was far out of line with economic realities. In consequence, it is unquestionably true that floating ex- change rates have frequently been associated with financial and economic instability -- as, for example, in hyperinflations, or severe but not hyperinflations such as have occurred in many South American countries. It is easy to conclude, as many have, that floating exchange rates produce such instability.

Being in favor of floating exchange rates does not mean being in favor of unstable exchange rates. When we support a free price system at home, this does not imply that we favor a system in which prices fluctuate wildly up and down. What we want is a system in which prices are free to fluctuate but in which the forces determining them are sufficiently stable so that in fact prices move within moderate ranges. This is equally true of a system of floating exchange rates. The ultimate objective is a world in which exchange rates, while free to vary, are, in fact, highly stable because basic economic policies and conditions are stable. Instability of exchange rates is a symptom of instability in the underlying economic structure. Elimination of this symptom by administrative freezing of exchange rates cures none of the underlying difficulties and only makes adjustments to them more painful.

It may help bring out in concrete terms the implications of this discussion if I specify in detail the measures that I believe the U.S. should take to promote a truly free market in both gold and foreign exchange.

1. The U.S. should announce that it no longer commits itself to buy or sell gold at any fixed price.

2. Present laws making it illegal for individuals to own gold or to buy and sell gold should be repealed, so that there are no restrictions on the price at which gold can be bought or sold in terms of any other commodity or financial instrument, including national currencies.

3. The present law specifying that the Reserve System must hold gold certificates equal to 25 per cent of its liabilities should be repealed.

4. A major problem in getting rid completely of the gold price-support program, as of the wheat price-support program, is the transitional one of what to do with accumulated government stocks. In both cases, my own view is that the government should immediately restore a free market by instituting steps 1 and 2, and should ultimately dispose of all of its stocks. However, it would probably be desirable for the government to dispose of its stocks only gradually. For wheat, five years has always seemed to me a long enough period, so I have favored the government committing itself to dispose of one-fifth of its stocks in each of five years. This period seems reasonably satisfactory for gold as well. Hence, I propose that the government auction off its gold stocks on the free market over a five-year period. With a free gold market, individuals may well find warehouse certificates for gold more useful than actual gold. But if so, private enterprise can certainly provide the service of storing the gold and issuing certificates. Why should gold storage and the issuance of warehouse certificates be a nationalized industry ?

5. The U.S. should announce also that it will not proclaim any official exchange rates between the dollar and other currencies and in addition that it will not engage in any speculative or other activities aimed at influencing exchange rates. These would then be determined in free markets.

6. These measures would conflict with our formal obligation as a member of the International Monetary Fund to specify an official parity for the dollar. However, the Fund found it possible to reconcile Canada's failure to specify a parity with its Articles and to give its approval to a floating rate for Canada. There is no reason why it cannot do the same for the U.S.

7. Other nations might choose to peg their currencies to the dollar. That is their business and there is no reason for us to object so long as we undertake no obligations to buy or sell their currency at a fixed price. They will be able to succeed in pegging their currency to ours only by one or more of the measures listed earlier -- drawing on or accumulating reserves, co-ordinating their internal policy with U.S. policy, tightening or loosening direct controls on trade.

A system such as that just outlined would solve the balance of payments problem once and for all. No deficit could possibly arise to require high government officials to plead with foreign countries and central banks for assistance, or to require an American President to behave like a harried country banker trying to restore confidence in his bank, or to force an administration preaching free trade to impose import restrictions, or to sacrifice important national and personal interests to the trivial question of the name of the currency in which payments are made. Payments would always balance because a price -- the foreign exchange rate -- would be free to produce a balance. No one could sell dollars unless he could find someone to buy them and conversely.

A system of floating exchange rates would therefore enable us to proceed effectively and directly toward complete free trade in goods and services -- barring only such deliberate interference as may be justified on strictly political and military grounds; for example, banning the sale of strategic goods to communist countries. So long as we are firmly committed to the strait jacket of fixed exchange rates, we cannot move definitively to free trade. The possibility of tariffs or direct controls must be retained as an escape valve in case of necessity.

A system of floating exchange rates has the side advantage that it makes almost transparently obvious the fallacy in the most popular argument against free trade, the argument that "low" wages elsewhere make tariffs somehow necessary to protect "high" wages here. Is 100 yen an hour to a Japanese worker high or low compared with $4 an hour to an American worker? That all depends on the exchange rate. What determines the exchange rate ? The necessity of making payments balance; i.e., of making the amount we can sell to the Japanese roughly equal to the amount they can sell to us.

Suppose for simplicity that Japan and the U.S. are the only two countries involved in trade and that at some exchange rate, say 1,000 yen to the dollar, Japanese could produce every single item capable of entering into foreign trade more cheaply than the U.S. At that exchange rate the Japanese could sell much to us, we, nothing to them. Suppose we pay them in paper dollars. What would the Japanese exporters do with the dollars ? They cannot eat them, wear them, or live in them. If they were willing simply to hold them, then die printing industry -- printing the dollar bills -- would be a magnificent export industry. Its output would enable us all to have the good things of life provided nearly free by the Japanese.

But, of course, Japanese exporters would not want to hold die dollars. They would want to sell them for yen. By assumption, there is nothing they can buy for a dollar that they cannot buy for less than the 1,000 yen that a dollar will by assumption exchange for. This is equally true for other Japanese. Why then would any holder of yen give up 1,000 yen for a dollar that will buy less in goods than the 1,000 yen will? No one would. In order for die Japanese exporter to exchange his dollars for yen, he would have to offer to take fewer yen -- the price of the dollar in terms of the yen would have to be less than 1,000, or of the yen in terms of the dollar more than i mill. But at 500 yen to the dollar Japanese goods are twice as expensive to Americans as before; American goods half as expensive to the Japanese. The Japanese will no longer be able to undersell American producers on all items.

Where will the price of the yen in terms of dollars settle? At whatever level is necessary to assure that all exporters who desire to do so can sell the dollars they get for the goods they export to America to importers who use them to buy goods in America. To speak loosely, at whatever level is necessary to assure that the value of U.S. exports (in dollars) is equal to the value of U.S. imports (again in dollars). Loosely, because a precise statement would have to take into account capital transactions, gifts, and so on. But these do not alter the central principle.

It will be noted that this discussion says nothing about the level of living of the Japanese worker or the American worker. These are irrelevant. If the Japanese worker has a lower standard of living than the American, it is because he is less productive on the average than the American, given the training he has, the amount of capital and land and so on that he has to work with. If the American worker is, let us say, on the aver- age four times as productive as the Japanese worker, it is wasteful to use him to produce any goods in the production of which he is less than four times as productive. It is better to produce those goods at which he is more productive and trade them for the goods at which he is less productive. Tariffs do not assist the Japanese worker to raise his standard of living or protect the high standard of the American worker. On the contrary, they lower the Japanese standard and keep the American standard from being as high as it could be.

Given that we should move to free trade, how should we do so? The method that we have tried to adopt is reciprocal negotiation of tariff reductions with other countries. This seems to me a wrong procedure. In the first place, it ensures a slow pace. He moves fastest who moves alone. In the second place, it fosters an erroneous view of the basic problem. It makes it appear as if tariffs help the country imposing them but hurt other countries, as if when we reduce a tariff we give up something good and should get something in return in the form of a reduction in the tariffs imposed by other countries. In truth, the situation is quite different. Our tariffs hurt us as well as other countries. We would be benefited by dispensing with our tariffs even if other countries did not.² We would of course be benefited even more if they reduced theirs but our benefiting does not require that they reduce theirs. Self interests coincide and do not conflict.

I believe that it would be far better for us to move to free trade unilaterally, as Britain did in the nineteenth century when it repealed the corn laws. We, as they did, would experience an enormous accession of political and economic power. We are a great nation and it ill behooves us to require reciprocal benefits from Luxembourg before we reduce a tariff on Luxembourg products, or to throw thousands of Chinese refugees suddenly out of work by imposing import quotas on textiles from Hong Kong. Let us live up to our destiny and set the pace not be reluctant followers.

1 have spoken in terms of tariffs for simplicity but, as already noted, non-tariff restrictions may now be more serious impediments to trade than tariffs. We should remove both. A prompt yet gradual program would be to legislate that all import quotas or other quantitative restrictions, whether imposed by us or "voluntarily" accepted by other countries, be raised 20 per cent a year until they are so high that they become irrelevant and can be abandoned, and that all tariffs be reduced by one-tenth of the present level in each of the next ten years.

There are few measures we could take that would do more to promote the cause of freedom at home and abroad. Instead of making grants to foreign governments in the name of economic aid -- and thereby promoting socialism -- while at the same time imposing restrictions on the products they succeed in producing -- and thereby hindering free enterprise -- we could assume a consistent and principled stance. We could say to the rest of the world: We believe in freedom and intend to practice it. No one can force you to be free. That is your business. But we can offer you full co-operation on equal terms to all. Our market is open to you. Sell here what you can and wish to. Use the proceeds to buy what you wish. In this way co-operation among individuals can be world wide yet free.

1 A warning is in order that this is a subtle point that depends on what is held constant in estimating the free market price, particularly with respect to gold's monetary role.

2 There are conceivable exceptions to these statements but, so far as I can see, they are theoretical curiosities, not relevant practical possibilities.