I’m going to talk today about a path that the IMF, the International Monetary Fund, might have taken at its founding but did not, and how different the organization would have been had it adopted that alternative path. When I finish what I plan to say about the IMF, I will conclude with some observations about the design of the World Bank, the second post-World War II international financial agency.

To develop the theme of this talk, I need to discuss the influences on Harry Dexter White, the main designer of the IMF at its origin, that led him to the choices he made rather than other choices that were available to him. So I shall begin by telling you who Harry Dexter White was and how he came to occupy a pivotal position in the creation of the IMF.

Although the talk will begin with a focus on the 1930s and 1940s, by the end I shall focus on the IMF as it currently functions.

Harry Dexter White got a Ph.D. in economics from Harvard in 1930, and had a brief academic career at Lawrence College in Appleton, Wisconsin. His dissertation, The French International Accounts, 1880-1913, published in 1933, won the Welles Prize at Harvard, and apparently brought him to the attention of Jacob Viner of the U of Chicago, who was then conducting a series of studies for the U.S. Treasury Department. Viner invited White to come to Washington in the summer of 1934 to work with him as a summer research appointee. Neither White’s nor Viner’s name appears in the 1934 Treasury Annual Report. In 1935 White is listed as one of five assistants to the director of the division of research and statistics, and one of six in 1936-37. In 1938, when the division of monetary research was established, White became its director, with one assistant director, a position he held thereafter, with two or more assistant directors. In December 1941, after Pearl Harbor Day, White was appointed assistant to Henry Morgenthau, the Treasury secretary, to act as liaison between the Treasury and the State Department on all matters having a bearing on foreign relations and "responsibility for the management and operation of the stabilization fund without a change in its procedures." In a minute I’ll tell you about the U.S. Exchange Stabilization Fund, one of the influences on Harry White’s thinking.

In January 1945, White was appointed Assistant Secretary of the Treasury. He resigned from the Treasury on 1 May 1946 to become the first U.S. Executive Director of the IMF. During the IMF’s first year, whenever the Managing Director was not at headquarters, White served as Acting Managing Director.

He became embroiled in a political assault on his loyalty and resigned his post in March 1947. He died the following year a few days after testifying before the House Committee on Un-American Activities. One other point about White’s career at the Treasury is that early on he became indispensable to Morgenthau in preparing memoranda for him and drafts of letters for Morgenthau to send to Roosevelt on international policy. The memoranda and letter drafts are included among White’s unpublished papers of 1935-46 that are available in a Princeton Manuscript Library.

In 1938-40, White worked on a proposal for loans to Latin America and participated in plans for an Inter-American Bank, which did not materialize. The plan for an Inter-American Bank, however, inspired White’s first draft of the subsequent plans for the IMF and the World Bank that White prepared in 1941 at Morgenthau’s direction.

Having sketched Harry White’s background, I now want to discuss his experience at the Treasury that influenced his conception of the IMF. To describe his experience, I shall first talk about the U.S. Exchange Stabilization Fund that was established at the Treasury Department by a provision in the Gold Reserve Act of 31 January 1934. The fund began operations in April 1934, financed by $2 billion of the $2.8 billion paper profit the government realized from raising the price of gold to $35 an ounce from $20.67 The Act authorized the fund to use its capital to deal in gold and foreign exchange in order to stabilize the exchange value of the dollar. The ESF as originally designed was a creature of the Executive Branch not subject to legislative oversight, but I doubt that this feature has played any role in more recent criticism of the IMF for lack of transparency.

The Gold Reserve Act authorized the Fund to use such assets as were not needed for exchange market stabilization to deal in government securities. It had no statutory authority, however, to engage in other activities that it began to undertake. The principal such extraneous activity it devoted itself to was lending dollars to politically favored governments.

The selection of countries to which the ESF extended loans was obviously a political decision, made by the Treasury. Harry White’s memoranda supporting loans to various countries, attest to his influence on which ones the ESF selected between 1936 and 1944 . The countries included China, Mexico, other Latin American countries, and Russia.

The first stabilization loan that the ESF extended was to Mexico in January 1936. We do not know who in the Division of Research and Statistics at the Treasury devised the terms of the loan. White’s unpublished papers give no indication that he was a principal in the design of the loan, but he unquestionably was aware of its unique features. In any event, it became the model for subsequent loans to other countries in addition to Mexico. When he became the director of the division of monetary research in 1938, White administered stabilization loans.

The terms of the Mexican loan were enclosed in a letter to the Federal Reserve Bank of New York as the Treasury’s fiscal agent, which was instructed to send them to the Banco de Mexico.

Let me note the unique features of the stabilization agreement with Mexico.

1. For the first time, instead of a loan of dollars by one country to another, the transaction is described as an exchange of currencies of the two contracting parties. The U.S. advanced $5 billion to Mexico for pesos, which the Banco de Mexico agreed to repurchase on demand by the U.S. Treasury at any time at the exchange rate at which they had been acquired.

2. Mexico agreed to pay a 3% interest rate on the peso account credited to the New York Fed, when the discount rate was then 1½% at the Federal Reserve Bank of New York, and the prevailing rate on 4- to 6-months commercial paper was ¾ of a percent.

3. Mexico also was required to provide silver collateral equal to the dollar amount of the exchange of currencies.

Let me briefly describe the mission of the IMF as set forth in the Articles of Agreement (dated 22 July 1944, as the Final Act of the United Nations Monetary and Financial Conference held at Bretton Woods, New Hampshire). The Articles bear the imprint of White’s formulation, and I shall next check for evidence of the ESF’s influence.

The IMF embodied the idea that a properly working international monetary system was based on currencies convertible into gold or U.S. dollars for current account transactions, but that allowed capital controls. The aim was to avoid competitive devaluations and assure exchange rate stability in the service of expanding international trade. The plan for the functioning of the IMF was outlined in the Articles of Agreement. To qualify for membership, each country would provide gold, its currency, and its government securities payable on demand in that currency to the Fund. Each country was assigned a quota, which was determined by its economic size and was to declare the par value of its currency in terms of gold or the U.S. dollar. The fund was authorized to provide financial assistance to individual countries to eliminate balance of payments deficits. Operations of the fund were to be directed by executive directors appointed by the members, and a managing director selected by the executive directors.

I now turn to a comparison of the ESF and the IMF noting the respects in which they were similar or differed. I begin with similarities.

1. The ESF was established to stabilize the exchange value of the dollar, the IMF "to provide an instrument with the means and the procedure to stabilize foreign exchange rates."
2. The composition of the assets of both the ESF and the IMF with one exception was similar. Each set consisted of gold, foreign currencies, and securities; in the case of the ESF, U.S. government securities; in the case of the IMF, securities of member governments. The ESF also held silver assets.
3. For both the ESF and the IMF rates of exchange were based upon the value of a currency in terms of U.S. dollars which had prevailed since the Gold Reserve Act of 1934. For the IMF foreign exchange dealings were based on par values, which the ESF did not specify.
4. Both the ESF and the IMF had authority to buy, sell, and hold gold, currencies and government securities, to earmark and transfer gold, and to buy and sell government securities. The IMF in addition in White’s draft was authorized to issue its own obligations, and to offer them for discount or sale in member countries. This provision did not appear in the IMF’s Articles of Agreement but the IMF was authorized in the Articles to borrow. The IMF has borrowed from a broad range of member countries and central banks to supplement the amounts it can lend.
5. The ESF provided dollars in exchange for the currencies of countries with which it established exchange stabilization agreements. The IMF provided foreign exchange to a member country in exchange for its own currency in order for it to meet an adverse balance of payments predominantly on current account
6. Both the ESF and the IMF designated the agencies with which they would deal: a country’s Treasury, central bank, stabilization fund (in the case of the ESF, beginning with the Tripartite Agreement).

The foregoing are common features with divergences between a domestic and an international institution. There are also differences unrelated to common features.

1. The ESF was a U.S. nationalist response to the British Exchange Equalisation Account, established in July 1932, of which the U.S. was suspicious. The U.S. believed that Britain planned to depreciate the pound in order to obtain a competitive advantage over the dollar. White deplored nationalism, writing, "A breach must be made and widened in the outmoded and disastrous economic policy of each-country-for-itself- and-the-devil-take-the-weakest."
2. The ESF extended very short-term currency swaps. Mexico borrowed in January 1936 and repaid in May and June 1936. White proposed a term of one year or less for a member country borrowing another currency from the IMF.
3. Investment by the ESF and the IMF was limited to government securities. White proposed to extend the right to include commercial drafts.
4. White would have denied currency to a borrowing country when the currency was to be used to make possible adjustment of a government foreign debt which had been defaulted. Had Mexico defaulted in 1995, White’s limitation would have prevented the ESF and IMF rescue, when dollars were used to pay domestic and foreign holders of tesobonos.
5. The ESF did not fix the exchange rates at which it transacted. In White’s proposal the IMF fixed the rates at which it would exchange a member’s currency for another.
6. The ESF charged an interest rate above the discount rate at the Federal Reserve Bank of New York. White proposed that the IMF be a source to which member countries "could turn in time of need to borrow at very low rates of interest."

The transformation of the ESF into the IMF involved many features that uniquely characterized the international institution, such as the quota system, subscriptions by the members, how it was organized and managed, the emphasis on scarce currencies. The aspects of the ESF that carried over to the IMF, however, shaped the institution in ways that lead to speculation on whether the performance of the IMF would have benefited had it operated under a different set of ESF guidelines. I next turn to this speculation

White’s proposals for the IMF were based on the ESF principle that all currencies are equivalent and that the goal of each fund was to stabilize exchange rates. There were other features of the ESF, however, that he might have deemed of sufficient importance for the IMF to replicate. The terms of the stabilization loans that the ESF extended during White’s tenure were commercial: short-term, extended at a market rate of interest, and collateralized with silver or gold. Commercial terms did not fit with White’s vision for the IMF. He was imbued with idealism in formulating his utopian plan for the future international financial institutions. They were to "pave the way and make easy a high degree of cooperation and collaboration among the United Nations in economic fields hitherto held too sacrosanct for international action or multilateral sovereignty." Thus it was that White opted not for commercial terms for IMF loans but "very low interest rates" in a setting of pervading neighborliness and good will.

Once the IMF began operations, its practices in providing credit to member countries evolved in ways that White had not envisioned. Countries draw on the fund in tranches that are based on the size of their quotas, the first tranche equal to 25% of a member’s quota that was to be automatically available with no conditions attached. By 1952, both within the IMF and outside it, "the death of automaticity" was recognized. Higher tranches were to be negotiated and the credits would come with conditions the member country was expected to fulfill. The method of operations henceforth also provided for approving "stand-by facilities" – approving drawings in advance, should they be needed – which made balance-of-payments assistance available on condition that the member agreed to implement detailed changes in macroeconomic performance.

The path the IMF has followed in providing financial assistance to member countries has relied on conditionality as the implicit price borrowers pay rather than the explicit price of market interest rates and collateral that the ESF adopted and that White did not lay down as the rule of action by the IMF. From the point of view of the borrower, conditionality is an implicit interest rate over and above the explicit rate charged for a loan. From the IMF viewpoint, conditionality should not be treated as a cost by the borrower. Fulfillment of conditions would improve economic performance.

One objection to conditionality is that the IMF does not uniformly and predictably enforce its conditions. Some of the reforms the conditions exact are not in any event achievable over the duration of the loan. Another objection is that the number of conditions the borrowing country is asked to accept is so multifarious that, with the best will, making the required changes is beyond its ability to achieve. Indonesia, for example, was given a list in excess of 100 conditions when the IMF negotiated the loan it extended in 1997. Compliance was poor.

Some critics of IMF conditionality terms believe that conditionality should "require a country to do things that it does not want to do." On this view, conditionality is a device to ration access to cheap IMF credit. If all the IMF required a borrower to do was the continuation of past sensible policies, it would be overwhelmed by loan requests.

One way for the IMF to avoid this dilemma would have been for it to lend on the terms that the ESF observed. The amount that a country borrowed would be available at a market rate of interest backed by collateral. No intrusion into the country’s institutional and legal arrangements would be part of the IMF’s mandate. The borrowing countries would seek assistance only if they were willing to meet a market test of their creditworthiness.

One may speculate on White’s reasons for rejecting the idea of collateral to backup IMF loans. Was he concerned about enforcement of collateral requirements? No problem of this nature arose with ESF stabilization loans to Mexico, Brazil and China in the late 1930s. Did he believe that demanding collateral violated national sovereignty? It may well be that he regarded a demand for collateral as inconsistent with the notion that the postwar world should be built on foundations of trust and cooperation among countries. Collateral smacked of distrust.

Harry Dexter White did not conceive of the IMF as a lender of last resort, although Keynes, its other architect, though of lesser importance, did. However, the idea was revived by Stanley Fischer, until this year the first deputy managing director of the IMF. He has proposed that it could function as a lender of last resort who would require collateral from the borrower.

Those who have discussed the future role of the IMF tend to agree that it is needed to address liquidity crises. What is at issue is on what terms it should provide liquidity to international capital markets that temporarily for whatever reasons are frozen. There is a longstanding prescription for central bank action when confronted by a national liquidity crisis: lend freely at a penalty rate on asset collateral that would be sound under normal conditions.

If the IMF were transformed along the lines of the forgoing prescription, it would cease to function as an institution that offers cheap credit to countries in trouble that White had in mind and instead would become a lender with strict standards in conformity with the demands of market discipline. Both the IMF and the borrowing countries would be constrained to alter their mode of operation. Instead of fifty years and more of IMF solicitude for emerging market countries, their progress to developed country status might have been achieved more consistently had the IMF been a tougher disciplinarian.

White transferred from the ESF to the IMF the form of exchange of currencies of the two contracting parties and not loans by one party to the other, to avoid "the indignity that might seem to attach to borrowing." This rhetoric falsifies the actual transaction of a stabilization loan by the ESF and lending by the IMF to client governments. When a member borrows from the IMF, the rhetoric states that it buys currencies of other members (or Special Drawing Rights), paying the equivalent in its own currency. When the member repays the credit, the rhetoric states that it repurchases the IMF’s holdings of its currency with SDRs or other members’ currencies.

The fiction fosters the idea that the borrower has not assumed a debt. Going into debt becomes behavior that a country need not restrain. What matters in this culture is emphasis on the availability of funds, not on whether it is in the best interest of a country to be a borrower. The culture encouraged the piling up of indebtedness in some countries and the extension of IMF credit year after year in others. This may also reflect the fact that until 2000 the IMF had no formula for forgiving loans.

This obscurantism was reflected in the balance sheet of the IMF, which until this year did not provide usable information on the actual transactions between the IMF and the member country. The assets side, for example, did not include an entry for loans to members. What it showed was the total amount of member country currencies that the IMF owned. The currency veil lacked transparency.

Had the doctrine that all currencies are equal not been transmitted from the ESF to the IMF, the latter’s role could have been defined in a straightforward way as a lender on commercial terms to countries in need of financial assistance. Had White carried forward to the IMF the ESF elements pertaining to interest rates and collateral, its operations would havw evolved in a markedly different direction from the one it has taken. An interesting question is, would the world have been better off

In addition to his proposal for the IMF, White also designed the IBRD, the International Bank for Reconstruction and Development, now known as the World Bank The objective of the Bank was to provide both short-and long-term loans or to participate in loans for relief, reconstruction, and development to any member country, or any political subdivision thereof, and to any business or agricultural enterprise in the territory of a member. The Bank had to be assured that the borrower could not otherwise obtain a loan at a reasonable rate of interest. When the member was not itself the borrower, it or its central bank was to guarantee repayment to the Bank of principal and interest. The Bank was also to guarantee loans of private investors with suitable compensation for risk. The Bank would buy, sell, and hold gold and the obligations and securities of member governments.

White’s experience as a Treasury official made a contribution to several features of his plan for the IBRD. Three events were part of his experience. These events influenced two features of his plan. I first identify the three events.

1. The first event is the ESF stabilization loan to Mexico of January 1936, which I’ve discussed in connection with the IMF. The loan became the prototype for pre-World War II lending. It was notable for requiring collateral and for its limited amount, limited duration, and above market interest rate. These aspects of the loan must have impressed White as unresponsive to Mexico’s needs.

2. The second event is an undated, possibly 1939 6-page memorandum by White with the title, "Loans to Latin America for the Industrial Development of Latin America." It included a proposal that the U.S. establish a bank with a capital of $300 million (funded by the ESF’s unused gold), authorized to issue up to $1.7 billion in U.S. government-guaranteed bonds, to make long-term loans for industrial enterprises in Latin America under the control of the borrowing country or its nationals. Loans would be offered at no more than 1 percent in excess of the cost of borrowing by the bank. Three-quarters of the dollar loans were to be expended on imports from the U.S.

3. The third event was a proposal for an inter-American bank that originated at a meeting in Panama in 1939. White was appointed by Morgenthau to serve on an interdepartmental committee on plans for this bank, the aim of which was to frustrate Nazi penetration in Latin America. The charter that was drafted in 1940 would have authorized the bank to make loans, or guarantee loans made by others, in any currency or gold to member governments, their nationals, fiscal agencies, and central banks. Governments were to guarantee that loans of more than two-years maturity would be repaid. The bank could hold obligations and securities of member governments, their currencies or foreign exchange, for its own account or the account of member governments. It could hold deposits of member governments, rediscount bills arising from intra-hemisphere trade that commercial banks discounted, and issue its own obligations. Only Mexico ratified the Inter-American Bank Convention. A U.S. Senate subcommittee, which held hearings on the subject in 1941, took no action. White, who championed economic assistance to Latin America, China, and the Soviet Union, extracted some of the elements in the Inter-American Bank proposal in preparing his blueprint for the World Bank.

1. One feature of the World Bank that White espoused was that it should make loans at artificially low interest rates. He insisted that the Bank should limit its loans to cases where otherwise private capital markets would lend at high (relative to those that would obtain in advanced countries) interest rates. In part, his attitude derived from his genuine concern to help the impoverished and war-torn countries of the rest of the world. White, however, in effect substituted the decisions of the Executive Directors of the Bank for the market’s decision on which projects were worth funding to the extent that the lending decisions did not involve projects where the social return exceeded the private return and the private sector would not have funded them. In fact, many of the Bank’s loans in the 1950s and 1960s to finance steel mills and the like were abysmal failures. Because demand at artificially low interest rates exceeded supply, the IBRD had to ration its available funds among borrowers. It was decided in negotiations between the British and the Americans at Bretton Woods that the Bank would not differentiate among borrowers by varying the interest rate that it charged. If there was a difference in the rates that it charged, it reflected the interest rate that the Bank had to pay to obtain funds.

2. A second feature of the blueprint of the World Bank was the emphasis on the need for government guarantees of the repayment of private loans for reconstruction and long-run development. White was responding to the default of international loans in the 1930s and the concurrent shutoff of foreign lending. In order to generate a flow of private capital to foster capital formation and growth in the postwar period. White was convinced that government guarantees of the debts of private borrowers were essential.

Low interest rates was not an ESF characteristic, but White introduced them into the IMF and IBRD. Loan guarantees were not a characteristic of the ESF nor of the IMF, but White introduced them into the IBRD. Keynes at Bretton Woods supported White’s emphasis on Bank guarantees of private overseas investment.

In practice, in the 1950s the World Bank interest rate charge for loans has been described as "near-commercial." The loan period was 10-20 years with an initial grace period after the loan was activated. For very poor countries, near-commercial interest rates were excessive. An International Development Association therefore was added to the IBRD to extend credits to countries with per capita incomes below some cut-off level for a period of around 50 years at a service charge of less than 1 percent annually. Thus White’s vision with respect to interest rates has been realized.

A change in practice from White’s blueprint for the World Bank pertained to the private sectors in countries in which the World Bank was active. Even when a project was to be undertaken in the private sector, World Bank rules required that there be a guarantee of the loan by the government. To enable the World Bank to lend directly to the private sector, the International Finance Corporation was established in 1956. As a result the IBRD has made little use of its guarantee powers. The IBRD, IDA, and IFC, are collectively the World Bank, with regional development banks supplementing World Bank activities. Since 1980 private capital flows to developing countries have been multiples of World Bank flows, something White could not have foreseen.