Back Matter
  • 1, International Monetary Fund


  • Angell, James W. (1926), The Theory of International Prices.

  • Bergsten, C. Fred (1975), The Dilemmas of the Dollar. New York: New York University Press.

  • Bickerdike, C. F (1906), Theory of Incipient Taxes, Economic Journal (December).

  • Bickerdike, C. F (1907), Protective and Preferential Import Duties, Economic Journal (March).

  • Bickerdike, C. F (1920), The Instability of Foreign Exchange, Economic Journal (March), pp. 118122.

  • Bordo, Michael D. and Finn E. Kydland (1990), “The Gold Standard as a Rule,” NBER Working Paper 3367 (May).

  • Bordo, Michael D. and Eugene N. White, A Tale of Two Currencies: British and French Finance during the Napoleonic Wars (August). Manuscript.

    • Search Google Scholar
    • Export Citation
  • Brock, P., M. Connolly and C. Gonzalez-Vega (1989), Latin American Debt and Adjustment: External Shocks and Macroeconomic Policies. Praeger: New York.

    • Search Google Scholar
    • Export Citation
  • Cassel, Gustav (1925), The Restoration of Gold as a Universal Monetary Standard. In Young (1925), pp. 2056.

  • Clement, Simon (1695), A Discourse of the General Notions of Money. Trade. and Exchanges

  • Crowther, G. (1957), Balances and Imbalances of Payments. Boston.

  • Dimand, Robert W. (1988), The Origins of the Keynesian Revolution. Stanford: Stanford University Press.

  • Edgeworth, F. Y. (1908), Appreciations of Mathematical Theories, Economic Journal (December), pp. 54156.

  • Fischer, Stanley, and Jacob A. Frenkel (1974a), Economic Growth and Stages of the Balance of Payments. In Horwich and Samuelson, pp. 50321.

    • Search Google Scholar
    • Export Citation
  • Fischer, Stanley, and Jacob A. Frenkel (1974b) Interest Rate Equalization and Patterns of Production, Trade and Consumption in a Two-Country Growth Model. The Economic Record (December), pp. 55579.

    • Search Google Scholar
    • Export Citation
  • Frenkel, Jacob A., Morris Goldstein, and Paul A. Masson (1989), International Dimensions of Monetary Policy: Coordination versus Autonomy. In Monetary Policy Issues in the 1990s. Federal Reserve Bank of Kansas, pp. 183233. A Symposium Sponsored by Fed. Reserve Bank of Kansas City.

    • Search Google Scholar
    • Export Citation
  • Halevi, Nadav (1971), An Empirical Test of the “Balance of Payments Stages Hypothesis”, Journal of International Economics 1: pp. 103117.

    • Search Google Scholar
    • Export Citation
  • Horwich, George and Paul A. Samuelson (1974), Trade. Stability and Macroeconomics: Essays in Honor of Llovd A. Metzler. New York and London: Academic Press.

    • Search Google Scholar
    • Export Citation
  • Host-Madsen, Paol (1962), “Assymmetries Between Balance of Payments Surpluses and Deficits,” IMF Staff Papers. Vol. 9, pp. 18299.

  • Kindleberger, Charles P. (1985), Keynesianism vs. Monetarism and Other Essays in Financial History. London: Allen & Unwin.

  • Kojima, Kiyoshi (1990), “Proposal for a Multiple Key Currency Gold-Exchange Standard,” Journal of Asian Economics. Vol. 1, No. 1, pp. 1933.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Marshall, Alfred (1925), Memorials of Alfred Marshall.

  • Mundell, Robert A. (1968a), Man and Economics. New York: McGraw-Hill.

  • Mundell, Robert A. (1968b), International Economics. New York: MacMillan.

  • Mundell, Robert A. (1968c), “A Plan for a World Currency,” In Next Steps in International Monetary Reform. Reuss Subcommittee (September 9), pp. 2528.

    • Search Google Scholar
    • Export Citation
  • Mundell, Robert A. (1974), “The Optimum Balance of Payments Deficit and the Monetary Theory of Empires,” In Salin and Claassen, pp. 6986.

    • Search Google Scholar
    • Export Citation
  • Mundell, Robert A. (1988), “Latin American Debt and the Transfer Problem,” In Brock, Connolly and Gonzagez-Vega.

  • Mundell, Robert A. (1989a), “Trade Balance Patterns as Global General Equilibrium: The Seventeenth Approach to the Balance of Payments,” Rivista di Politica Economica (June).

    • Search Google Scholar
    • Export Citation
  • Mundell, Robert A. (1989b), “The Global Adjustment System”, Rivista di Politica Economica (December), pp. 351466.

  • Mundell, Robert A. (1990a), “The International Distribution of Saving,” Rivista di Politica Economica. Paper presented at the second Mondragon Conference in July 1990 (forthcoming).

    • Search Google Scholar
    • Export Citation
  • Mundell, Robert A. (1990b), “Fiscal Policy and the Theory of International Trade,” forthcoming in Giersch (ed.), Proceedings of the Sohmen Memorial Conference: Tegernsee, Bavaria.

    • Search Google Scholar
    • Export Citation
  • Mundell, Robert A. and Alexander Swoboda (1969), Monetary Problems of the International Economy. Chicago: University of Chicago Press.

  • Mundell, Robert A. and John McCallum (eds.) (1990), Global Disequilibrium in the World Economy. Oxford University Press (forthcoming).

  • Neher, Philip A. (1970), International Capital Movements and Balanced Growth Paths. Economic Record 46, pp. 393401.

  • Onitsuka, Yusuke (1970), International Capital Movements and Patterns of Economic Growth, unpublished Doctoral Dissertation, University of Chicago.

    • Search Google Scholar
    • Export Citation
  • Onitsuka, Yusuke (1974), International Capital Movements and the Patterns of Economic Growth, American Economic Review 64 (1) (March), pp. 2436.

    • Search Google Scholar
    • Export Citation
  • Patinkin, Don and J. Clark Leith (eds.) (1978), Keynes. Cambridge and the General Theory. Toronto: University of Toronto Press.

  • Poehl, Karl Otto (1987), Cooperation—A Keystone for the Stability of the International Monetary System. First Arthur Burns Memorial Lecture at the American Council on Germany, New York (November 2).

    • Search Google Scholar
    • Export Citation
  • Robertson, Dennis H. (1938), Professor Viner Again, Economic Journal.

  • Salin, P. and E. Claassen (1974), Stabilization Policies in Independent Economies. Amsterdam: North Holland.

  • Schumpeter, Joseph A. (1954), History of Economic Analysis. New York: Oxford University Press.

  • Swoboda, Alexander (1989), Global Disequilibrium and International Monetary Reform. Rivista di Politica Economica (December). Reprinted in Mundell and McCallum (1990).

    • Search Google Scholar
    • Export Citation
  • Taussig, Frank W. (1928), International Trade. New York.

  • George S. Tavlas (1990), “On the International Use of Currencies: The Case of the Deutsche Mark,” IMF Working Paper, 90/3. (Washington, D.C., International Monetary Fund (January)). Forthcoming revised version in Essays in International Finance. Princeton University (January 1991). See also an unpublished paper by Tavlas and Yusuru Ozeki (December 1990). The Japanese Yen as an International Currency.

    • Search Google Scholar
    • Export Citation
  • Jacob Viner (1937), Studies in the Theory of International Trade. London.

  • Jacob Viner (1938), Reply to Robertson: Rejoinder, Economic Journal.

  • Young, John P. (1925), European Currency and Finance. Government Printing Office, Washington, D.C.


Professor of Economics at Columbia University and Visiting Scholar in the IMF Research Department. This paper was prepared for presentation at the Conference on International Adjustment at the Institute for International Economics, Washington, D.C., October 27-28, 1990.


Examples of this stabilizing element in expectations is afforded by the currency history of Britain and other countries. Instead of devaluation in the 1690s to finance the great recoinage, the government harked back to Elizabeth I’s recoinage over 130 years earlier that paid for the coinage out of taxes rather than depreciation. Britain was able to compound its public debt to finance the several wars of the 18th century without devaluation; and when the Bank of England succumbed to inconvertibility in 1797, interest rates for two decades reflected the strong probability that Britain would restore the sacred parity after the war (which happened in 1819). Again, the uninterrupted history of the gold standard in Britain in the nineteenth century, despite periodic crises, contributed to the belief, when payments were suspended during World War I, interest rates again reflected the expectation that, after the war, convertibility would be resumed at the old parity. This occurred in 1925, but it proved abortive when, for systemic reasons, the undervaluation of gold that had been a consequence of world war inflation led to a liquidity crisis, tight money and the great deflation. In the face of this external stability Britain correctly opted out of convertibility. In a recent paper, Bordo and Kydland (1990) analyze the gold standard as a contingent rule, meaning that the authorities could “temporarily abandon the fixed price of gold during a wartime emergency on the understanding that convertibility at the original price of gold would be restored when the emergency passed.” See also Bordo and White (1990) for a recent discussion of British and French finance during the Napoleonic Wars.


If fiduciary monetary issues were permitted, they had to be kept within the limits of safeguarding the nation’s central gold reserves. Gold also ensured that the domestic price level would remain, in the long run, basically constant relative to the world price level. The stability of the world price level would itself depend on the relation between the world demand and supply of gold. When gold supply increased more rapidly than gold demand, the value of gold would fall and the world price level would rise; when gold supply increased less rapidly than gold demand, gold would appreciate and the world price level would fall. Although there were disturbing trends in the value of gold over the long period due to changes in supply, new gold production in any given year was a small proportion of the outstanding stock of gold so that annual changes in the price level were comparatively small. Over the long run, relative changes in the real value of gold would affect supply and demand in a stabilizing way, with the result that periods of rising prices tended to be offset by falling prices.


I have argued elsewhere (Mundell 1968a) that John Stuart Mill at least, among the classical economists, did not confuse Walras’ Law and Say’s Law.


See Mundell (1989a and especially 1989b) for a recent review of the literature on this subject and for detailed references. An excellent review of the literature up to 1937 is offered by Viner (1937).


Keynes did not admit Ohlin’s point in his reply to Ohlin’s article in the Economic Journal but later in correspondence with Ohlin he wrote: “As to your point that reparations cause a shift in the demand curve of the receiving country irrespective of any rise in the price level of that country, I do not think I disagree with you.” This is quoted in Dimand (1988, p. 124) from Patinkin and Leith (1978 p.162).


This assumes trade is initially balanced; a slight adjustment is required where this is not the case.


Constant costs and incomplete specialization imply that the elasticity of demand for imports is infinite. Meade (1950) produced the first algebraic formula for the change in the terms of trade explicitly integrating expenditure and price effects. Note that because expenditure propensities and price elasticities are related by the Johnson-Slutzky condition, import propensities high enough (greater than one-half on the average) to shift demand onto the goods of the paying country (the anti-Mill direction) are sufficient to ensure the elasticities of demand are large enough (greater than one-half on the average) to ensure stability of exchange equilibrium. See Mundell (1960 and 1968).


Chipman’s analysis was presented at the 1988 meeting of the Eastern Economic Association in Baltimore.


This paragraph was part of my comment on Chipman’s Baltimore paper.


Jacob Viner addressed this question in his Studies (1937), analyzing the issue on the assumption that what he called the “final-purchases velocity” of money was constant. Viner’s analysis resulted in an exchange with Dennis Robertson (1938) in the Economic Journal, Robertson arguing that it was more natural to assume that the income velocity of money was constant. In the ensuing debate a compromise position was arrived at, with Viner conceding that the money balances required to service the increase in expenditure in the receiving country would be, while still positive, less than that required to service expenditure from income produced at home, with the result that gold would still flow in the direction of the receiving country—changes in prices being abstracted from—but by a smaller amount than he had initially concluded. Robertson, however, held to his view that the receiving country would not require more money to service the additional expenditure mad possible by tax reductions than was previously required to service the payment of taxes that the inward transfer made unnecessary.

A correct analysis of the issue requires splitting the demand for money balances into sectoral demands by consumers, producers and governments. Because production in both countries is assumed to be constant, there would not ordinarily be a change in producer’s money balances. But the money requirements of consumers would rise in the receiving country and fall in the paying country on account of the expenditure shifts. Government spending, however, would rise in both the receiving and the paying countries in view of the transactions associated with the disbursement of the proceeds through (say) an income subsidy in the payee and the raising of the proceeds by (say) taxation in the payer.


I have noted this transactions effect in Mundell (1988). As noted in the preceding footnote, a precise treatment of the increased transactions demands would require splitting national demands into household, producer and government sectors.


An alternative formulation would make the demand for money a function of wealth. Insofar as we are restricting our analysis to unilateral transfers, wealth, and therefore the demand for money, is lower in the paying country and higher in the receiving country, resulting in a balance-of-payments surplus in, and a flow of gold to the receiving country.

This formulation would have to be modified in the case of capital movements. Unlike unilateral transfers, capital movements do not imply a change in wealth. They represent a geographical redistribution of wealth. The purchasing power (absorption) of the borrowing country is increased, and of the paying country is reduced, by the capital transfer, but the GNPs, which take into account international interest payments, are only slightly affected (by the extra rent associated with differences in rates of return). On these grounds, therefore, the inter-temporal pattern of consumption would not theoretically be affected by borrowing; most of borrowed money should therefore be devoted to the formation of physical or human capital. This implies that the product mix of marginal expenditure effects will be influenced by the proportion of the transfer that is unilateral and the proportion that is merely a loan.


Among a number of other important contributions to theory and policy, Bickerdike (1907) originated the theory of the optimum tariff. Edgeworth (1908), following along the lines of Bickerdike’s optimum tariff work, had explicitly developed the criterion for stability of the real exchange equilibrium. Bickerdike was the first, however, to relate the elasticity formula to changes in exchange rates, using the device of two inconvertible currencies.


Bickerdike uses the inverse of elasticities, what he calls “inelasticities.”


Bickerdike’s “inelasticities” are zero.


The next few paragraphs draw on and summarize the analysis of a paper delivered first at the Caracas Conference on Financial Markets in January 1979 and published in Spanish in the Conference volume. See also Mundell (1989, 1990).


This abstracts from asymmetries in the treatment of the reserve assets that result in a recorded non-zero balance of trade for the world as a whole.


Even where the inter-temporal borrowing is legitimate, however, there might be externalities associated with changes in the terms of trade that may create a divergence between private and social benefits from external borrowing.


Some attention was, it is true, paid to the impact of the trade balance on the level of employment, theorizing that was at least partly based on errors in economic theory. Superficial mercantilist reasoning, arguing from the national income accounts, Keynesian equations, and the fact that exports are source of demand while imports are not (except for repercussion effects), seems to imply that an improvement in the balance of trade increases employment. This conclusion may be correct if the disturbance to the balance of trade originates from an increase in export demand on the part of the rest of the world; but it is false if the improvement arises from a reduction in domestic expenditure. It is a fallacy to argue that measures that improve the balance of trade ipso facto increase employment.

In late 1971, after the breakdown of the Bretton Woods system, an improvement in the trade balance became, for a brief period, an explicit target of government policy. Secretary-of-the-Treasury John Connally made his famous assertion that the United States wanted a turnaround in its trade balance of $13 billion. Treasury officials naively hoped that the devaluation of the dollar negotiated at the Smithsonian Institution would bring about that turn-around in the trade balance.

The devaluation failed to make a dint in the trade balance. The trade balance deficit actually severely worsened in 1972 to over $6 billion, the largest deficit up to that time in US history. Although the balance became positive in 1973, it worsened again in 1974, only to become positive again in the recession year 1975.

The year 1975 marks a watershed in the history of the international accounts. It was the last year the United States had a trade surplus. Deficits rose after 1976 to over $30 billion and after 1983 to over $100 billion.


At the 1966 Chicago Conference on International Monetary Problems. See Mundell and Swoboda (1969).


See Mundell and Swoboda (1969, p. 38).


There was a short period in which the decline in the US creditor position made Saudi Arabia the largest creditor country, before Saudi Arabia was overtaken by Japan and Germany.


See Cairnes (1874) for the development of an pioneering version of the stages approach; see also Taussig (1927). Crowther (1957) elaborates six stages: immature debtor-borrowers; mature debtor-borrowers; debtor-repayers and debtor-lenders; immature creditor-lenders; mature creditor-lenders; and creditor-drawers and borrowers. Halevi (1971) develops a twelve-fold classification (allowing for borderline cases). Other early work in the modern analysis of stages includes Onitsuka (1970, 1974), Neher (1970), Fischer and Frenkel (1974a, 1974b).


Note, however, that the model would predict a turnaround again after the first decade of the 21st century as the baby-boomers retire and are replaced as senior workers by another small cohort.


It is, of course, possible that inflation can shift the distribution of wealth, resulting in a rearrangement of world demand and a change in relative prices. Inflation also has fiscal effects insofar as it reduces the real fiscal burden of the public debt. If saving increases (because of the wealth-saving relation) and the budget deficit is reduced (because of reduced real interest payments), the trade balance will improve unless these changes are offset by increases in investment financed by increased capital inflows.


In offer curve analysis of the Marshall-Meade type there is no place for a balance-of-payments deficit. A price vector that differs from the equilibrium produces an excess demand for the foreign country’s good, which can only constitute a transactions configuration if there is intervention: Domestic commodity authorities must be selling foreign goods out of their own stockpiles, and simultaneously buying stocks of the domestic good. When official transactions are added to the private transactions, there is no deficit or surplus.


We might note that it is also necessary to determine the effect on the net capital outflow of the policy that creates the change in the exchange rate. If, for example, the central bank buys foreign exchange with newly-printed money and invests the foreign exchange in a foreign security, the trade balance must improve by the amount of the capital outflow.


As noted earlier, this was pointed out by Bickerdike (1920, p. 118-119). After discussing the fall in stock prices in terms of dollars as one of the mitigating factors that would encourage greater American demand for pound assets, he writes: “ has to be noted that if the stocks and shares pay interests in pounds, a fall in value arising in this way must exceed the fall in the exchange rate before purchases by Americans becomes attractive, unless the lower rate is expected to be temporary.”


A stronger case can be made for changes in relative prices in the manufacturing sectors of large countries than in small open economies producing primary products for which there is a world market. The relative price of basic commodities are determined by real variables and it is unlikely that relative price changes play an important role in the process of adjustment.


Commodity standards were subject to instability arising from two main sources: fluctuations in the supply of the commodity; and shifts of demand. Although changes in supplies of the commodity were disturbing enough, it was possible to anticipate them, in view of long production lags. More disturbing were sudden and dramatic shifts of demand due to countries going from one standard to another.

From the 1820s to the late 1840s, when bi-metallism in France and the United States gave the world a monetary unity, expanding silver supplies from Mexico were insufficient to make up for the shortfall in gold production. This position was reversed in mid-century when the gold discoveries in Australia and California were sufficient to double in a decade world monetary gold stocks; gold drove silver out of circulation in France and world prices began to rise. The United States currency became inconvertible during the Civil War, leaving France alone to bear the brunt of bi-metallism; but France suspended convertibility during the Franco-Prussian War. By that time silver production had rapidly increased, forcing France to safeguard her gold currency by abandoning bi-metallism for a limping gold standard; even earlier the new German Empire had dumped silver for gold, further lowering the price of silver.

The 1870s witnessed an explosion of silver production, leading to the abandonment of bi-metallism and the march to the gold standard. This shift of standards created an excess demand for gold and an excess supply of silver which led to deflation in the gold bloc and inflation in the remaining silver countries. Criticism at this time was directed at the shift away from bi-metallism, which aggravated deflation in the gold brigade and inflation in the remaining silver countries. Objections to deflation led to agitation to restore bi-metallism, which, however, failed when South African gold arrived to prevent further deflation, and, indeed, induce a mild inflation for almost two decades before the outbreak of World War I.


Gold had become unstable after the outbreak of World War I when the belligerent countries engaged in inflationary finance, exporting gold to the few countries (including the United States and Japan) remaining on the gold standard. The commodity value of gold fell in half as US prices doubled. The immediate post-war deflation in the United States was insufficient to restore the pre-war gold prices. The stock of gold was sufficient to maintain the US on gold, but inadequate for an international gold standard of the pre-war type. Nevertheless, the international gold standard was restored in the midst of a state of monetary uncertainty occasioned by wide fluctuations in exchange rates. Despite feeble attempts at international monetary coordination, the gold scarcity led to deflationary policies that inaugurated the great deflation, made worse by mass unemployment. Belatedly, some countries left the gold standard or devalued, leading to a revised system.


The new system relied heavily on the U.S. dollar after its devaluation in 1934. The devaluation more than corrected the gold shortage (given the prohibition on privately-owned gold in the United States), leading to an initially under-valued dollar (relative to gold). The dollar shortage, however, lasted only from 1934 to 1950. By 1950 wartime and post-war inflation had raised prices, creating a gold scarcity, concealed somewhat by the prohibition of gold for US citizens and the disproportionately-large U.S. gold holdings. Attracted by interest returns and confidence in the dollar, the rest of the world was initially content to use dollars in lieu of gold. But gold losses to European central banks, revealed the true nature of the global excess demand for gold.


I have discussed this effect—what I have called the “Thornton effect”—in Mundell (1989, 1990). Hume and Smith had recognized that an increase in credit or paper notes would not cause prices to rise provided convertibility of the currency was maintained; gold or silver money would be exported to the same extent that it was replaced by soft money. Thornton, however, took account of the effects of the export of specie on the price level in the rest of the world, which would have a corresponding, if small, effect in raising the domestic price level.


One of the problems associated with the confusion between systemic and national issues is the concept of the balance of payments deficit, conventionally defined as a loss of gold or increase in liquid liabilities. This definition is inappropriate if deficits and surpluses are thought of as error signals because it makes no distinction between desired and undesired changes in reserves. In Mundell (1965) I advanced a definition of the balance of payments restricted to undesired changes in reserve assets or foreign liquid liabilities; I still believe this is the appropriate concept, despite its operational difficulties.


A dollar shortage is not, however, incompatible with a gold shortage. In the I.M.F., the dollar was the currency that was “needed to be drawn” even while countries with balance of payments surpluses were exchanging dollars for gold.


Keynes had earlier suggested what he called an “international clearing union” in 1943. I presented a plan for a world currency in testimony to the Joint Economic Committee in Mundell (1968).


Certainly the Viet-Nam War did much to sour the trans-Atlantic climate. More important, however, the dominant power—always the country with, in the short run, the most to lose by powerful supranationalism—was not yet willing to demote itself.


There was a provision in the Articles of Agreement of the International Monetary Fund for a change in the par values (specified in gold) of all currencies, so the founding fathers of the IMF anticipated the possibility. Note that this proposal would not necessarily involve any changes in exchange rates.


Other (often mutually contradictory) objections were that an increase in the price of gold would (1) be inflationary; (2) create expectations of future increases and lead to a gold shortage in the future; (3) tend to reinstate the gold standard; (4) unfairly redistribute wealth toward gold-holding countries; (5) penalize countries that had accumulated dollars rather than gold; (6) lower the gold value of existing contracts; (7) help South Africa and/or the Soviet Union.


The difficulties attendant upon official devaluation of the dollar became clear in the debates at the Smithsonian meeting when a political compromise was reached at which the United States devalued the dollar against gold and some other countries revalued their currencies against gold.


The devaluation of the dollar against other currencies could have been effected in either of two ways: (1) An increase in the dollar price of gold, the par values expressed in gold of other currencies remaining constant; (2) The raising of the gold values of other currencies, the par value of the dollar remaining constant at .888671 grams = 1/35 of an ounce. Either method would have required other countries to raise the price at which they bought dollars in the exchange markets. These difficulties were by no means insurmountable—they were resolved when exchange rates were changed at the 1971 Smithsonian meeting—but they indicated the great complications associated with the asymmetry of the dollar’s position in the Bretton Woods system.


As always, systems have to be compared against the alternatives. Despite the arguments against such an arrangement, it might have been better than the alternatives, including that actually adopted. There are worse systems than one which required a change in the price of the international monetary asset every generation or so.


See Machlup and Malkiel (1964). The other three proposals coming in for detailed discussion were the gold standard, a world central bank and system of currency areas or large floating blocs.


The facts proved otherwise. National demands for international reserves would not abate with flexible exchange rates. On the contrary, as Harrod pointed out even in the 1960s, countries would need more rather than less liquidity under flexible exchange rates because of the increase in uncertainty. Moreover, as I argued, a flexible-exchange-rate world without an official standard would naturally use the dollar as the most important reserve asset, creating a normal current account deficit equal to the secular demand for dollar reserve assets.


Penalties in the form of asset conversions would have to be negotiated in advance, rather than imposed. At the Copenhagen meetings of the IMF in 1970, IMF Managing Director Pierre-Paul Schweitzer proposed that the major reserve country accept some gold losses in view of its balance-of-payments deficit worsened, a proposal which, at the time, was resented by the U.S. Treasury.


The collapse of Bretton Woods was stretched out over three episodes: 1968, when the market price of gold was severed from the official price and the members of the I.M.F. withdrew from the private gold market; 1971, when the dollar became inconvertible; and 1973, when flexible exchange rates was adopted.


The gold pool, organized at the initiative of the United States, with the central banks of Belgium, France, Italy, the Netherlands, Switzerland, West Germany, and the United Kingdom. The pool was actually a gentleman’s agreement to divide the burden of stabilizing the price of gold in the London gold market, with the United States having a 50 per cent share. The Bank of England acted as agent for the pool in its market operations.


The communique of March 17, 1968, stated that the seven members of the gold pool (France had dropped out in 1967) “decided no longer to supply gold to the London gold market or any other gold market” and further asserted that in view of the prospective establishment of the Special Drawing Rights, the existing stock of monetary gold is sufficient.


The accumulation of dollars by the rest of the world was partly desired to compensate for the immobility of gold; but it is probable that US monetary policy was more expansive over the period 1965-73 than the countries on the European continent desired. The other countries temporarily had an excess supply of dollars which, however, disappeared with the rising prices level in the early 1970s.


Gold was still, however, legally used as a numeraire. A few days before the Smithsonian agreement of December 17-18, 1971, Presidents Nixon and Pompidou met at the Azores and agreed that the United States would devalue in terms of gold. The dollar was devalued by 7.89 per cent, raising the price of gold to $38 an ounce. The DM was appreciated by 4.61 per cent against gold (13.58 per cent of the old dollar parity); the yen was raised 7.66 per cent against gold (16.88 per cent against the old dollar parity). The other Group of Ten countries, with the exception of Canada, whose currency was left floating, revalued against the dollar, but some devalued against gold.


Exchange rate changes could not, as we have argued, remedy the systemic problem of the undervaluation of gold against all currencies. The Smithsonian agreement failed to correct the basic problem of the undervaluation of gold and in 1973 a second dollar crisis emerged with another equally futile, devaluation of the dollar. Only a few months after this devaluation (June 1973), the system broke down into flexible exchange rates, removing what little monetary discipline remained. The subsequent increase in oil prices was quickly ratified by inflationary finance and the explosion of credit in the Eurodollar market.


The connection between the two events was brought home to me in January 1972, before the Smithsonian agreement had been fully ratified. Participants at two independent meetings—one of OPEC ministers and the other of the Bellagio group of academics and officials—were lodged at the Intercontinental Hotel in Geneva debating simultaneously the same Smithsonian agreement from entirely different directions. The OPEC group then and there decided to raise the dollar price of oil in reaction to the increase in the dollar price of gold.


It had been, like devaluation, an “unmentionable” at US Treasury Consultants meetings in the 1960s.


In the 1960s, George Schultz, as Dean of the Graduate School of Business overlapped with Milton Friedman as Professor of Economics at the University of Chicago.


The gross US deficit, measured by the increase in liquid liabilities of the United States to foreign countries, has amounted to over a trillion dollars since 1973; total liabilities rose from over the period from $92.5 billion to $1,101.7 billion. Over the same period, liquid external claims rose from $26.6 to $658 billion. On an official settlements basis alone, US liquid liabilities to foreign central banks and governments has increased from $66 billion to over $300 billion.


Although Bickerdike pays close attention to capital movements, his elasticity formula is restricted only to the trade account.


Estimates of foreign exchange market put it over $600 billion per working day, which implies total yearly transactions over $150 trillion.


The comparative monetary properties of a currency are related to the slopes of the liquidity preference schedules in each country, the absolute value of which is normally proportionate to size. See Mundell (1974).


The cumulative improvement in the monetary properties of a currency area as it expands leads to the theoretical proposition that the optimum number of currencies in the world is one. Flexible exchange rates has usually been presented, by its major proponents, as a second best arrangement, as an adaptation necessary in the real world to overcome nominal wage rigidities arising from the existence of labor unions. There is, however, no reason to expect the zones of labor bargaining power to overlap with the domain of currencies.


Quoted in Frenkel, Goldstein and Masson (1989, p. 194).


See Tavlas (1990). The rise of the mark is accounted for by the growth in activity of the EMS and the use of European currencies for intervention purposes. Tavlas supplies, among others, the following figures: Dollars accounted for 71.5 per cent of EMS intervention in 1979-82, but only 26.3 per cent in 1986-87; whereas the mark accounted for only 23.7 per cent in the early period but 59 per cent in the later period. As far as US intervention is concerned, almost 90 per cent was in marks in 1979-82, with virtually the rest in yen; but in 1986-87, the mark share fell to 57.5 per cent and the yen share rose to 42.5 per cent.


There is nothing automatic about the advantages of a large currency area to a single country. It depends, among other things, on whether the prospective partners have a more stable monetary policy than the members; or more exactly whether the new currency area will have a more stable monetary policy than individual country. The ERM system offers a good example where some currencies, previously more inflationary than the DM, have been able to use the system as a political excuse for a more stable monetary policy. Germany may have been induced to follow a slightly more inflationary monetary policy as a result of the ERM, but offsetting this cost is the beneficial effects of European monetary leadership and a more important currency.


The real exchange rate seems to have been an innocent bystander—a victim—of the speculation and erratic monetary policy. After the 1979 oil price increase, an easy monetary policy was adopted to try to prevent the economy from falling into a recession in the 1980 election year; the result was too much inflation and an undervalued dollar. The rise of the nominal and real exchange rate in the first Reagan term was motivated by five factors: (1) a corrective of preceding inflationary policies; (2) reflux into the dollar from foreign currencies as confidence was restored; (3) an increase in capital imports due to the effect of the tax cuts in raising the marginal efficiency of capital in the United States; and (4) overkill by an over-zealous Federal Reserve Chairman (Paul Volcker) as a reaction against the presumed effect of the Kemp-Roth tax cut on the budget deficit; and (5) an underestimation by the influential Friedman wing of the Republican party of the increase in demand for money at home (reduction in velocity) due to a reduction in the rate of expected inflation. There was little or no attention to the role of the real exchange rate in these deliberations and little thought to the harmful consequences of arbitrary fluctuations in the real exchange rate.


The first two of these observations were made to me by Herbert Grubel.

Do Exchange Rates Work? Another View
Author: Mr. Robert A. Mundell