Who Needs the IMF?
Federal Reserve Swap Network, 1963
loan without the drawbacks. As Figure 3.1 depicts, by the end of 1963 Chairman Martin had in place a system that gave the Fed access to $2.3 billion in foreign exchange. More importantly, these swap lines had a number of advantages over alternative means of external financing. Central bank swaps were incredibly flexible and could be expanded at a moment’s notice over the phone. This contrasted greatly with the months of difficult negotiation that were necessary to implement the GAB. Additionally, swap credits were available on demand without negotiation. Thus, swap resources could be deployed far more swiftly during a quickly unfolding crisis. Charles Coombs summed up just how fast and flexible the swap lines were:
It is quite true . . . that many of these defenses were quickly improvised, sometimes within a matter of hours, to deal with sudden emergencies. In most cases, they were negotiated on a bilateral basis and may give the impression of being no more than an unrelated patchwork. But these bilateral defenses have the most important advantage of being solidly based on market and institutional realities in each country and are capable of being flexibly adapted to new and unforeseeable needs. One cannot overemphasize the importance of being able to move quickly—on the basis of telephone consultations if necessary—against speculative pressures before they gain momentum.
Speed and flexibility were not the only advantages either. Swap drawings also came without conditions, meaning the United States could borrow and still maintain its domestic policy autonomy. Finally, swaps also met an occasional "desire for secrecy which an IMF drawing could not provide.”
One need not look further than the actual use of the central bank swap lines by the United States relative to its use of the GAB to see which facility was preferred. Despite the fact that the central purpose of the GAB was to make financing in foreign currency available to the United States, not once during the 1960s did the United States activate the arrangement. Borrowing from the IMF and GAB was a cumbersome and slow process and was not well suited to defend the dollar in the face of speculative pressure. Furthermore, as one scholar explains, "Successive U.S. administrations were unwilling, for domestic political reasons, to accept the conditions attached by the Fund to drawings.” The swap network meant that the United States and its partners no longer needed the IMF for assistance. They had "managed to develop sources of liquidity [they] deemed preferable.”  Or, in Richard Cooper’s words,
The apparent need of the United States for swap facilities suggest certain deficiencies in the International Monetary Fund as a source of and supplement to international liquidity. IMF lending was evidently felt to be too costly, too clumsy to arrange, too small in amount, or too visible to the public to satisfy the requirements of the countries in
Aggregate Swap Credits by Quarter, 1962-1969
In contrast, the United States consistently drew on its swap arrangements throughout the decade and beyond. They emerged as a clear alternative to the IMF that was far closer to Walter Bagehot’s ideal ILLR. Figure 3.2 plots aggregate quarterly swap drawings from 1962 through the end of 1969. Foreign central bank drawings from the Fed are indicated by the solid line. Fed drawings from foreign central banks are indicated by the dotted line.
Perhaps the only downside to the swap arrangements, as compared with borrowing from the IMF, was the fact that the former were quite short-term in nature. Typically, a swap drawing formally expired in three months and therefore in principle had to be reversed in that same time frame. This meant that while they were great at addressing short-term dollar outflows and pressures at the gold window, they were not designed well for financing medium- to long-term imbalances. However, in practice, it turned out that the swap lines effectively provided longer-term assistance as well. Robert Roosa made this point to President Kennedy in a meeting with his economic advisors in 1962. Roosa explained to the president that, practically speaking, swaps are not short term: "They are, in general, renewable. The usual practice in these is you simply roll them over until they are reversed. They can go on for years if they have to.” In practice, Roosa was correct. All the swap lines opened in 1962 were consistently renewed, without exception, for decades to come. In addition, Roosa pointed out that the "technically” short-term nature of the swaps was desirable for domestic political reasons: "It keeps us in control ... . We don’t want the lights to [be] put out, you see. We’ve had our problems with the Congress on this. We don’t like to put out a swap arrangement that they say, well, you’re just really getting into this for good.” These seemingly contradictory statements highlight another reason swaps were so desirable: they had the effect of looking temporary but, in practice, could be quite durable and long-standing.
-  FOMC 1961b, p. 85.
-  Data are from relevant historic Federal Reserve Monthly Review publications, availableat http://www.ny.frb.org/research/ monthly_review/ 1963.html.
-  Coombs 1946, p. 91. A great example of this flexibility came in the immediate aftermathof the assassination of President Kennedy when, within minutes, the Fed had increased itsswap lines with the Swiss National Bank and the BIS by 50 percent each "in a move to headoff any panic or speculative sale of dollars for other currencies” (Cowan 1963).
-  Creditors did retain the discretion to activate the swap, and so they could ask for certain assurances during the period of consultation. At times, such assurances included apromise to turn to the IMF, if needed, rather than request a swap renewal (Henning 1999,pp. 51-52). In practice, however, this was uncommon.
-  Cooper 1968, p. 214.
-  Ainley 1984, p. 26.
-  Bird and Rajan 2001, p. 10.
-  44. Cooper 1968, p. 213. Emphasis added.
-  Data were collected by the author from relevant historic Federal Reserve MonthlyReview publications.
-  Naftali 2001, pp. 499, 509.