A History of the Federal Reserve, Volume 2

by Allan H. Meltzer

  46. Switzerland agreed to this arrangement but said it was obligated under Swiss law to buy from and sell gold to Swiss citizens.

  47. The two-tier gold market was discussed and discarded earlier. Arthur Okun explained that much of the contingency planning after 1965 was done by the Deming group, chaired by Treasury Undersecretary Frederick Deming, a former Federal Reserve Bank president. The other members were Dewey Daane of the Board of Governors, Okun while he was a member of the Council of Economic Advisers, Francis Bator of the National Security Council, and Anthony Solomon of the State Department. This group both preferred to pay out gold and doubted that other countries would agree to close the gold window. Okun claimed that in March “the outflow got up to hemorrhage proportions. It was at this point, I think, that the negotiability of the two-tier system first became clear. . . . They wanted to stop their losses of gold. They did not want to keep paying it into the gold pool” (Okun, Oral History, Interview II, tape I). Okun credits Guido Carli, governor of the Banca d’Italia, with the idea of the two-tier system.

  The U.S. gold stock rose slightly in 1969–70. One reason was that Germany and Japan, the countries with large payments surpluses at the time, limited gold purchases. France had regularly demanded gold, but it experienced a payments deficit in 1968 and 1969, so it had to sell gold.

  The German decision was formally stated in a March 1967 letter from President Karl Blessing of the Bundesbank to Chairman Martin supported by a letter from the chancellor of the West German government. Germany committed to refrain from purchasing gold from the U.S. Treasury (Solomon, 1982, 111). The decision came “following threats of a partial troop withdrawal” (Holtfrerich, 1999, 384). Japan did not make a public pledge or require direct pressure. It simply accumulated dollars and did not demand much gold.

  Table 5.3 shows the change in short-term dollar assets and gold between 1966 and 1970 for France, Germany, and Japan. After it recovered from the 1968 strikes, disruptions, and wage increases, France accumulated dollar assets and did not replace the gold sold before the August 1969 devaluation of the franc. But it must have been clear to everyone that the agreement eroded the gold base of the fixed exchange rate system. De facto, the dollar was no longer convertible for most purposes.

  The March 16–17 meeting took a long step toward ending the Bretton Woods system. It solved the short-term problem but left the long-term problem untouched. Vietnam spending had increased. The projected United States budget deficit reached a postwar record ($25 billion), and the dollar remained overvalued. The Johnson administration did not devalue the dollar against gold and offered no other adjustment toward long-term equilibrium. The Federal Reserve did not commit to a less inflationary policy. Until 1970–71 there was little pressure from Congress or others to develop a long-term policy.

  48. The commitment not to sell official gold (no longer to supply) was much stronger than the decision not to buy (no longer necessary because the existing stock is sufficient). The claim of sufficiency reflected the belief that the SDR would soon be approved as a source of international reserve growth. The Board’s staff pointed out that IMF members had a right to obtain foreign currencies for gold and to repay the IMF in gold. The gold could be obtained from (or sold to) the market (memo, Holland to FOMC, Board Records, March 29, 1968, 9). Among large gold holders, France, Australia, and South Africa did not subscribe to the agreement. The last two were important gold producers.

  The meeting solved the immediate problem by introducing another exchange control; this one went to the basis of the system. It could only work as long as countries feared the consequences of collapse and did not ask to exchange dollars for gold. Closing the gold market, passage of the surtax in June, the small budget surplus in fiscal 1969,a new administration in 1969, higher interest rates, and a capital inflow from the euro-dollar market postponed the long-term problem. Firm resistance at the Washington meeting to the British-Dutch proposal to devalue the dollar showed that United States officials had not yet concluded that they had to adjust. The most likely explanation is that they hoped that what they said was true; creation of SDRs and the surtax would solve their problem.

  In the squabble over the discount rate with New York, the majority and minority set out their positions. The difference in the two positions was not in their analysis but in the goals each sought to achieve. New York recognized that a deflationary or disinflationary policy was the only way to keep the dollar price of gold fixed. A 6 percent discount rate would show “determination . . . to defend the dollar at all costs” (Maisel diary, March 18, 1968, 5). Maisel then wrote that the Board “should reject the concept that the defense of the dollar is worth ‘all costs’” (ibid.) He was willing to “hold the growth rate of the economy to 4 percent and perhaps even to 3.5 percent, but anything beneath this is not logical . . . [F]or equilibrium to come about at existing exchange rates there would have to be a much faster rise in European prices and it does not seem likely that this can be brought about” (ibid., 6). The choice, he said, was additional controls or eventual devaluation of the dollar.

  Martin did not share Maisel’s view. He told the American Society of Newspaper Editors that the United States was in a crisis—“the worst financial crisis that we have had since 1931.” But he added: “This is not a disaster story. The world would not come to an end if we did that [devalued]” (extemporaneous remarks, Martin speeches, April 19, 1968, 4, 5).49 Press reaction to the speech called Martin an alarmist. Like many of his contemporaries, Martin believed that devaluation of the dollar or floating rates would end the postwar system and bring back trade restrictions.

  49. He noted that some economists preferred floating rates. That would be “the greatest setback, financially, that this country has faced, certainly in my lifetime, and I think it will take us a long time to recover from it” (Martin Speeches, April 19, 1968, 5). But his speech showed willingness to consider devaluation despite his reluctance.

  When the London market reopened on April 1, gold sold for $38 an ounce. The price remained between $38 and $43 for the next year, then fell to $35 following sales by South Africa (Solomon, 1982, 124). The price remained close to $35 until new disruptions culminated in the formal end of U.S. gold sales in 1971. The two-tier agreement remained in effect until November 1973, when the market price was almost $100 and the official price had increased to $42.22 (ibid., 127).

  Between November 1967, when Britain devalued, and March the federal funds rate increased nearly a percentage point, from 4.12 to 5.05. Growth of the monetary base remained at a 6 percent annual rate. In response to a question, Martin reported that the European central bankers held mixed views about the desirability of additional rate increases. They preferred a reduction in the budget deficit by fiscal contraction. Robert Solomon added that the Europeans would accept a rise in interest rates that shifted about $200 million in euro-dollars a month to the United States, reducing the capital outflow. A more aggressive policy would likely bring matching increases abroad.

  Despite a decision at the April 2 meeting to make only a modest change toward firmer policy, short-term rates rose following the meeting. By midmonth, the Board agreed to increase the discount rate to 5.5 percent and the regulation Q ceiling rate to 6.25 percent. Much of the argument for the changes referred to the balance of payments. Before making the changes, the Board notified the Secretary of the Treasury and the Council of Economic Advisers. The Council informed the president. No one objected, but the White House staff asked the Board to delay the announcement until the following Monday, “pending an opportunity for the President to discuss the matter with the Chairman” (Board Minutes, April 18, 1968, 13). The Board agreed to delay the announcement (ibid., 14). This was the first time that the White House interceded to delay announcement of an action already voted. The Board Minutes do not report on the discussion with President Johnson, but the Board announced the rate increases after the meeting.

  MINTING PAPER GOLD50

  For the years 1960–67 as
a whole, the non-U.S. members of the G-10 (including Switzerland) acquired 150 million ounces of gold, an increase of one-third over their holdings at the end of 1960 (IMF, 1990, 65). Every country except Britain and Canada added to its holdings. Britain sold 38 million ounces, the United States 164 million ounces. France acquired two-thirds of the G-10 increase, 100 million ounces. The following year, it sold 40 million ounces to delay devaluation.

  50. Except as noted, this section follows Solomon (1982, chapter 8) and Meltzer (1991). Solomon was an active participant in most of the meetings. For details of the negotiations, see Solomon (1982).

  The steady decline in the United States’ gold stock pushed the liquidity issue to the forefront. From 1965 to 1968, the major countries discussed changes in the international monetary system to increase the stock of settlement balances (liquidity). The outcome was the special drawing right (SDR) widely described as “paper gold,” a new reserve asset to be used for settlement between countries. The intention was to free the provision of international means of payment between central banks or monetary authorities from dependence on the supply of gold and U.S. dollars. Although most of the emphasis was on finding a solution to the “Triffin problem,”51 France often tried to get attention to the adjustment problem. The United States usually assumed that its payments deficit would end.

  The French argued, correctly, that there was no reason for the liquidity discussion as long as the United States and Britain had large payments deficits.52 France opposed any planning for a new reserve asset as long as world dollar balances continued to increase. Its spokesmen wanted to end the special role of the dollar as a reserve currency, what President de Gaulle called its “exorbitant privilege,” but they opposed putting control of a new money at the IMF. They preferred a new form of credit controlled by the G-10.

  On important votes, the other European countries did not support France. This was particularly true of French proposals to increase world reserves by raising the gold price and reestablishing gold as a “neutral currency” (Solomon, 1982, 136). But the principal European countries agreed with France, for a time, that the new source of reserves should be a repayable form of credit, not an addition to international money as the U.S. proposed.53

  51. Haberler (1965, 46) wrote, “Some of the ingenuity now so lavishly spent on how to guard against the possibility that international liquidity may become scarce could be more profitably applied to the more basic and neglected problem of how to improve the adjustment mechanism.” Policy officials ignored all such comments.

  52. The French position was that the special position of the dollar gave the United States the great advantage that it could settle its payments deficit by printing more of its own currency. The United States’ position accepted the claim that it had a unique position, but recognized that it had committed to maintain convertibility of dollar claims into gold. After March 1968, this response was less prominent (Report of the President’s Task Force on Foreign Economic Policy, Department of State, S/P Files, Lot 70D194, November 25, 1964, 23). The United States also paid interest on the foreign claims.

  53. One early French proposal known as the Collective Reserve Unit (CRU) would have tied growth of international money to the stock of monetary gold, acting as a multiplier of the existing gold stock. The United States wanted increases in reserves to remain independent of the gold stock. The CRU proposal called for action by the Group of Ten. The United States favored action under the IMF, where it could veto proposals.

  The major issues that had to be resolved included whether the new units would be credit or money, who would get the new units, whether the new unit would be tied to gold, and who would decide on the size and frequency of additions to the stock—how many votes would be required and who would be able to vote. The United States and France generally differed on these issues. The U.S. task was to move ahead without attacking the French position in a way that would end the meetings.

  By June 1967, the group had resolved several issues. However, France continued to insist that reserves supplied by the new mechanism had to be repaid. Differences about who would have power to veto additions to reserves also remained. One reason for the deadlock was that France and others believed that the United States intended to pay off its liabilities by creating new pieces of paper unrelated to gold or other assets. The French did not want a system that would permit the United States to pay its debts with a new paper asset.

  France gained a concession when the Group of Ten agreed to name the new unit a special drawing right (SDR). Unlike other IMF drawings, however, the new unit was transferable and remained available for future transfers. France accepted the decision to make the new units permanent. This left only decisions about voting as a major issue. At the IMF annual meeting in Rio de Janeiro in September 1967, the members voted to approve the new asset as a supplement to existing reserve assets.54 Finally at Stockholm on March 29 and 30, 1968, ministers agreed on the SDR and once again reaffirmed their commitment to the $35 per ounce gold price. To obtain agreement by the Europeans (other than France), the United States accepted that the European Union, like the United States, could veto any future increases in IMF quotas.

  The IMF could issue SDRs only if an 85 percent majority approved. Under the 1944 Bretton Woods enabling legislation, Congress had to approve any increase. When approved, IMF members could create an asset to serve as a substitute for gold in settlement between central banks and governments but not as payment for quota increases at the IMF. France reiterated the need for adjustment and did not sign the agreement until autumn 1969. Congress quickly approved the amendments to the IMF agreement, and President Johnson signed the agreement on June 19, 1968. The agreement received enough support to enter into effect on July 28, 1969. The first issue of SDRs came on January 1, 1970.

  54. France tried to the end to get agreement on an increase in the gold price. It proposed making the U.S. use profits from depreciation to redeem part of its outstanding liabilities.

  Robert Solomon, who did a great deal to achieve agreement on the SDR and the broader issue of providing liquidity, summed up the experience. At the time he thought that the decisions in March to close the gold market and to adopt the SDR “may turn out to have marked a turning point in world monetary history” (FOMC Minutes, April 2, 1968, 13). In his book, he revised this conclusion. “Even if establishment of a reserve-creating mechanism was a necessary condition for international payments balance . . . earlier attention should have been given to introducing greater flexibility of exchange rates” (Solomon, 1982, 167). Much earlier, Haberler and Willett (1971) criticized the decision to devote much time to the liquidity problem and not to the adjustment problem.

  The case for SDRs was far less than compelling. As the French never tired of pointing out, the system had excess liquidity. The United States argued that this would not always be true, that ending the U.S. payments deficit would, sooner or later, lead to a liquidity shortage.

  This argument suffered from two weaknesses. First, U.S. representatives had no plan for achieving balance and spent little effort on that problem. Second, the argument was false or, at best, incomplete. The United States could add to reserves by buying other stable currencies, just as those countries bought dollars. The Council of Economic Advisers (1964, 145) recognized that this argument was correct.

  Principal world monetary officials had made enormous effort to solve the lesser of two problems while ignoring the more serious problem. French efforts to shift the emphasis met with hostility. To the last, at Stockholm, the French Finance Minister, Michel Debré, reminded his colleagues that they had neglected the adjustment problem. Perhaps because of the history of French policy, the others rejected his plea. Within three years, the Bretton Woods system broke down, exchange rates changed, the gold price increased, and the international monetary system later acquired an adjustment mechanism. SDRs remained in the system.

  It seems unlikely that the SDR would have become a dominant medium of exchange or s
tore of international reserves if the fixed exchange rate system had survived. Developing countries treated SDR allocations as a wealth transfer; most quickly exchanged them for hard currencies. Furthermore, the SDR did not dominate alternatives. Gold is an established store of value with a long history. SDRs had to compete with currencies that had superior properties—the dollar and later the mark and the yen. Balances held in each of these assets paid interest. At first SDR balances did not earn interest and could not pay interest since there was no source of revenue or earnings.

  Revaluing gold would have solved the liquidity problem, as the French insisted. French proposals did not limit future changes in the gold price, so they were open to the charge that expectations of future devaluation would lead to a run on the dollar. Much earlier, Keynes (1924) had proposed a commodity standard in which gold served as a medium of exchange. He tied the gold price to the price of a commodity basket. In 1922, Irving Fisher persuaded a congressional committee to hold hearings on a similar proposal for a compensated dollar (Meltzer, 2003, 182–83).

  When governments agreed to the SDR in 1968, the U.S. price level was approximately 2.5 times its 1929 level. If the gold price had increased in the same proportion, the 1968 gold price would have been $52, making the U.S. gold reserve $17.6 billion, $1.6 billion more than its total liabilities to central banks and governments. The Bretton Woods system could have continued. Of course, it could not have continued indefinitely or even very long without some restriction on U.S. monetary policy, a restriction that U.S. authorities were unlikely to accept. That was true of any solution that tried to maintain fixed exchange rates.55