Download in PDF format

August 1999 JEC Study

Research Findings Regarding the Costs of U.S. Participation in the IMF


------------------------------------------------------------

Executive Summary

    The finances of the IMF are relevant to U.S. taxpayers because of the prominent role of the U.S. in funding the IMF and guiding its decisions. Consequently, IMF quota increases and most significant IMF gold sales are considered as important policy issues before the Congress. However, informed policy decisions by Congress require adequate IMF transparency to ensure the availability of necessary information. Only with adequate transparency can Congress, in consultation with academic and other experts, develop a thorough understanding of the IMF's financial structure and the costs of U.S. participation in the IMF.

    Over the last two years, the Joint Economic Committee (JEC) has promoted essential transparency through hearings, research papers and press statements.  This paper reviews some of the key conclusions of JEC research concerning IMF financial structure and costs of U.S. IMF participation.

    The lack of IMF transparency makes many dimensions of the costs of U.S. participation in the IMF unclear to policymakers and the taxpaying public.  These costs include a disproportionate U.S. burden in financing the IMF,  subsidized interest rates, absorption of risk, and cost-shifting and other aspects of non-restituted IMF gold sales.  Conservative estimates of the costs of U.S. participation in the IMF indicate that these costs are substantial.  The evidence shows that the U.S. is shouldering a significantly greater proportion of the IMF's financial burden than the oft-cited 17.7 percent quota share suggests.  Instead, the U.S. contributes about 26 percent of usable IMF resources.

    In addition to these costs, it is important to highlight the changing nature of the IMF financial structure.  The IMF's portfolio has become riskier in a number of ways over the last two decades.  Further, the IMF has evolved into an organization that relies on a narrow base of donors to provide the funds borrowed by a separate group of borrowers.  More specifically, financial support is increasingly supplied by a small  number of industrialized countries while borrowers are for the most part developing countries facing long-term structural problems.  Furthermore, IMF lending is highly concentrated.  The five largest borrowers from the IMF account for 70 percent of outstanding loans.

    This information, while essential for informed Congressional decisionmaking, has not been readily available to Congressional policymakers or the taxpaying public.  Much of the reason policymakers and the public are not fully informed on these matters is a documented lack of transparency on the part of the IMF.
 
 

------------------------------------------------------------

Research Findings Regarding the Costs of U.S. Participation in the IMF


Introduction

    The finances of the IMF are relevant to U.S. taxpayers because of the prominent role of the U.S. in funding the IMF and guiding its decisions. Consequently, IMF quota increases and most significant IMF gold sales are considered as important policy issues before the Congress. However, informed policy decisions by Congress require adequate IMF transparency to ensure the availability of necessary information. Only with adequate transparency can Congress, in consultation with academic and other experts, develop a thorough understanding of the IMF's financial structure and the costs of U.S. participation in the IMF.

    Over the last two years, the Joint Economic Committee (JEC) has promoted essential transparency through hearings, research papers and press statements.  This paper reviews some of the key conclusions of JEC research concerning IMF financial structure and costs of U.S. IMF participation.1

    This paper reviews several findings derived from an examination of the IMF financial structure and costs of U.S. IMF participation.  While there are obviously many other important policy issues involved in examining IMF policy (e.g. moral hazard, conditionality), this paper focuses on summarizing the costs of U.S. participation in the IMF.  The paper also discusses the changed IMF's financial structure and comments on IMF transparency.

Costs of U.S. Participation in the IMF

    The U.S. contributes about 17.7 percent of total IMF quota subscriptions.  It is this percentage of quota resources that determines the voting rights of IMF member countries and is often equated to the member country's official financial contribution.

    However, the actual economic costs of U.S. participation in the IMF differ significantly from this widely cited 17.7 percent figure.  In particular, the evidence demonstrates that in a number of ways the U.S. is shouldering a significantly larger burden than is suggested by the Treasury or official IMF statements.  These additional costs are often obscured by accounting practices and procedures as well as by related difficulties in calculating various hidden costs, opportunity costs, subsidies, risk factors, or their accruals over time.  Some observations relating to this issue include the following:

    In its quarterly operating budget, the IMF separates usable contributions from the unusable contributions from countries in weak economic condition.  (There is little, if any, demand for the currencies of these later countries for international transactions.)  With these unusable quota contributions set aside, the U.S. share of usable IMF quota contributions rises to about 26 percent.  Since this figure represents the proportion of those contributions that actually can be used for lending, it is an economically more meaningful figure than the overall quota percentage.  It demonstrates that the U.S. is contributing a significantly larger share of financial resources than suggested by the quota percentage often cited by the Treasury and the IMF.     Under IMF procedures, 25 percent of quotas are initially paid in international reserve assets, primarily foreign exchange.  The other 75 percent ordinarily takes the form of local currency or promissory notes.  But some countries, such as the U.S., persistently contribute more than initial 25 percent of quota in reserve assets, including hard currency.  In fact, the U.S. reserve tranche position as a percentage of quota currently is about 42.3 percent.  The reasons for this are that (1) the U.S. -- unlike many other countries -- does not normally run down its reserve tranche position and (2) the IMF "encashes" U.S. promissory notes to raise additional funds for IMF lending.  In so doing, the IMF enlarges the U.S. reserve tranche position.  Accordingly, countries like the U.S. provide the extra marginal resources used for additional IMF lending programs.  At the same time, many other countries -- nearly one half of IMF membership -- maintain virtually no reserve positions.2   Furthermore, under the recent quota increase, over half of the IMF members immediately withdrew the economically meaningful portion of their contribution.  According to the GAO:

                Between January 1999, when the Eleventh General Review quota became effective,
                and April 30, 1999, …. 92 members withdrew the entire $3.6 billion of usable currencies
                or SDR, replacing them with their national currencies.3

    In sum, the additional reserve positions financing the IMF are largely from the U.S. and a relatively small group of other advanced countries, whereas a large share of IMF members contribute little or no usable monies.  Indeed, of the total amount contributed in excess of 25 percent of member quota in hard currency, about 30.8 percent is contributed by the United States.

    In short, the U.S. is contributing significantly more than 17.7 percent of both usable and hard currency contributions to the IMF.  Thus, U.S. taxpayers are contributing proportionately more meaningful financial resources than the 17.7 percent implied by their oft-cited quota-determined voting rights percentage.

    The U.S. share of total credit arrangements of the GAB amounts to 25%.  Again, this amounts to a commitment whose share exceeds that of the oft-cited quota-based voting rights share.  Accordingly, it indicates the U.S. is shouldering a larger share of the credit line commitment than suggested by its quota contribution share.  This is case even when the New Arrangements to Borrow (NAB) are considered, although the combined share does decline to 19.7 percent.     The current rate of interest that is remunerated on U.S. funds is about 3.4 percent, a rate below that on U.S. Treasury securities.  In other words, the U.S. government is lending at more favorable interest rates than the cost of money to the government.  The government, therefore, is lending at "below-cost" rates involving a subsidy that should be recognized.4

    From another perspective, the U.S. is essentially financing and providing the IMF with predominately long-term money (its reserve position at the IMF) while receiving a short-term rate of remuneration.  This involves a subsidy to the IMF and a corresponding cost to U.S. taxpayers.5

    In addition to the portion of the U.S. reserve position that receives interest payments (or remuneration), there is an unremunerated portion as well that does not generate interest payments.6   Accordingly, this portion involves an even larger (percentage wise) subsidy than the remunerated portion.  In other words, there is a portion of our reserve position that is, in effect, providing "free money" to the IMF.  According to the new report by the General Accounting Office, the cumulative value of lost interest payments amounts to $2.7 billion dollars.7   This unremunerated portion of the U.S. reserve position constitutes another sizable subsidy to the IMF and a significant cost to the U.S. taxpayer.     Another issue relating to IMF operations is the increased riskiness of IMF lending.  In particular, over the years IMF lending has changed from making a large share of relatively safe, short-term (low risk) loans to high-grade industrial countries such as the U.S., U.K., France, and Italy to making significantly higher risk, longer-term loans to lower-rated countries such as Russia, Indonesia, Brazil, Mexico, and Korea.  At times, the IMF loan portfolio has been highly concentrated with the latter type of loans since IMF lending limits have been raised substantially.  In short, the IMF loan portfolio has become considerably riskier over time.

    Since these higher risks expose the lender, the IMF's donor countries should be compensated for these higher risks via higher interest rates; the interest rate charged the borrower and compensating the lender should reflect this higher risk.  Yet this additional risk factor is generally not recognized or reflected in the interest rates paid by borrowers from the IMF and received by lenders to the IMF.  This uncompensated risk factor is another subsidy and cost borne by lenders such as the U.S. and its taxpayers.  But this risk factor is exceedingly difficult to accurately calculate or quantify for a number of reasons.  One recent study recognizes and discusses the problems of calculating such risk factors.  In particular, Lerrick recommends that "a risk premium or allowance for credit losses should be included in the cost of providing resources" to the IMF.8

    IMF proposals to sell gold from its sizable reserves and use the proceeds for various IMF purposes, instead of restituting the gold to the original contributors, also entail significant costs to U.S. taxpayers.  Costs associated with U.S. gold contributions are often hidden or obscure and usually not adequately taken into account in most discussions of IMF gold sales.  Such hidden costs are another non-transparent element of the costs of U.S. participation in the IMF.9

    Gold is carried on the IMF books at about $48 per ounce, well below current market prices.  Accordingly, this below-market value hides the (higher) economic value of these reserves.  Consequently, any gold sale which occurs at market prices will entail sizable gains to the seller.  Relative to the restitution provisions under the IMF charter, such gains would come at the expense of the original contributors of the IMF gold.  In any event, the potential profits from gold sales were nonexistant when the gold was initially contributed and should not be usurped by the IMF, but returned to the member nations.

    This was forcefully recognized in a bipartisan manner in 1975, when earlier discussions about the proceeds of gold sales occurred.  According to the view of Senator Ribicoff (D-Conn.) and Senator Taft (R-Ohio), expressed in a joint statement from a Joint Economic Committee subcommittee document:

                Either the gold belongs to the IMF, or it belongs to the members states,
                which contributed the gold in proportion to their quotas.  In either case, the profits
               (of sales) should be distributed to the member nations in proportion to their quotas.10

    The IMF currently owns 103 million ounces of gold.  Since the IMF's restitution formula provides for an U.S. share of about 23 percent of this gold, any restitution to the U.S. would entail a sizable sum.  For purposes of illustration, for example, at a market price of $260/ounce, the U.S. taxpayer share of potential restitution of the entire IMF gold stock would be approximately 23 percent of the gain, or about $5.02 billion.11   The U.S. gain amounts to about $190 million for each billion dollars of gold sales.

    On the other hand, U.S. taxpayers would forgo this amount in the case of non-restituted sales.  For example, a 10 million ounce sale without restitution, as recently proposed, would cost the U.S. taxpayer about $488 million.12   These are not trivial sums.  But these costs are seldom recognized in discussions of alternative gold sale proposals.

    There are many dimensions to the costs of U.S. participation in the IMF that policymakers, taxpayers, and the public should understand.  These include a disproportionate shouldering of financial contributions and commitments to the IMF, subsidized interest rates, absorption of risk, and aspects of gold sales.  Conservative estimates of the costs of U.S. participation in the IMF (following the President's Commission on Budget Concepts) suggest that these costs are substantial.13   In summarizing these costs, it is reasonable to conclude the following:     All of this indicates that the U.S. is shouldering a significantly greater proportion of the IMF's financial resources than the oft-cited 17.7 percent quota share would suggest.  Furthermore, these facts have not been transparent to policymakers, the public, or the taxpayer.  These costs to the U.S. taxpayer too often have been understated, hidden, or obscured by IMF accounting practices and procedures.  Indeed, a number of these costs are not adequately accounted for in U.S. budgetary documents as recommended, for example, in the 1967 President's Commission on Budget Concepts.14  In this context, it will be recalled that a year or so ago, IMF and Treasury sources were claiming that U.S. participation in the IMF was costless!  Efforts to make IMF finances more transparent would help to put an end to such misrepresentations.

The changing IMF financial structure.

    Identifying the many dimensions to the costs of U.S. participation in the IMF is one aspect of our examination of IMF financial practices.  Another key point relates to the IMF's changing financial structure.

    The IMF's original procedures, practices, and structure were designed in an era of fixed exchange rates, with gold and the U.S. dollar at the center of the Bretton Woods System.  For the most part, IMF lending largely pertained to (short-term) loans to highly-rated, creditworthy, industrialized countries experiencing temporary balance of payments disturbances.  Being temporary, such lending was seen as an approximation to short-term asset exchanges.  Since interest rates were relatively low, interest charges and related subsidies were deemphasized and viewed as of secondary importance.

    In the years since the demise of the Bretton Woods System, however, IMF practices and its clientele have changed significantly in a number of important ways.  The IMF's portfolio, for example, has become riskier with a longer-term maturity structure.  The IMF has evolved into an organization that is redistributive in nature in that the flow of subsidies is from one group of donors to a different and much larger group of borrowers.

    The IMF's portfolio has become riskier in a number of ways.  For the most part, IMF lending to lower-rated, higher-risk developing economies has replaced the lending to higher-grade, lower-risk industrial countries.

    Further, loans are significantly longer-term and often more structural or developmental in nature than was earlier the case.  Indeed, developing country lending now constitutes virtually all of IMF lending as compared to a significantly lesser share, for example, in 1970 (see graph).


Click here to see the figure.

    An additional element of risk has been added by the liberalization of IMF loan limits that formerly capped borrowing to 100 percent of a member's quota.  Not only have these limits been relaxed, but even the new rules have been grossly exceeded in recent years.  As the JEC has pointed out, this relaxation of IMF lending limits has led to an unhealthy concentration of IMF lending to a small number of high-risk borrowers, thereby subjecting the IMF's lenders to considerable risk exposure.  These JEC findings have also been noted by the GAO:

                In the past 4 years, the IMF has provided financing to five large developing
                countries that have experienced financial crises.  This financing was in amounts
                that were all well in excess of the IMF's limit on cumulative borrowing.15

    In addition to having a riskier, longer-maturity portfolio, the IMF has evolved into an organization that is redistributive in nature.  As mentioned, loans and associated subsidies are increasingly concentrated among lower-rated, lower-income developing countries in contrast to the industrialized country lending of earlier periods.  At the same time, financial support is narrowly-based among the G-10 countries.  In fact, the IMF does not have broad-based financial support.  This is illustrated by the fact that as much as 77 percent of the IMF's usable resources are supplied by the G-10 countries, and nearly half of IMF members maintain little or no reserve positions.  In short, financial support is increasingly supplied by a small number of industrialized countries while borrowers are typically developing economies facing long-term structural problems.  In sum, the IMF portfolio has become more redistributionist over time.

Transparency and the IMF.

    The above-cited costs as well as the IMF's changing financial structure have been obscured from Congressional policymakers, the public, and U.S. taxpayers.  Yet in order for well-informed Congressional decisions pertaining to the IMF to be made, such information is essential.  Part of the reason policymakers have been uninformed on some of these matters is due to a lack of transparency on the part of the IMF.

    Transparency, of course, has many different dimensions, and it means different things to different people.  In brief, however, transparent policy is characterized by a lack of secrecy, obfuscation, or ambiguity and should be clear, simple, and understandable to policymakers as well as to the taxpaying public.  It involves goal clarification as well as clear reporting on a real time, "ex-ante," and "ex-post" basis.16   Unfortunately, while the IMF has made some limited improvements in the dissemination of data and information, the IMF has a long way to go before it can be viewed as a truly transparent institution.

    Numerous examples support this finding of non-transparency.  The language used by the IMF, for example, illustrates this contention.  The IMF refers to its lending from its principle account not as "loans," but rather as "currency purchases."  Furthermore, changed IMF objectives -- as reflected in its dramatically altered financial structure -- have never been spelled out in a meaningful way.  IMF financial documents as well as its accounting practices and procedures are neither clear nor understandable even to some senior IMF officials themselves.

    Many of the costs of participation in the IMF (cited earlier) are difficult to calculate because much of the information needed to make such calculations is unavailable, obscure, hidden, or difficult to understand or to collect.  Accurate information related to usable versus unusable resources, for example, are in operational budgets which are not available to the public.  Similarly, the calculation of remunerated interest subsidies can be difficult.  The costs of non-remuneration of interest are even less well-known since the concept has been hidden and calculation of the unremunerated reserve tranche position is not straightforward.  Similarly, calculation of risk premium and associated costs of risk is a complicated and difficult task that is not undertaken by the IMF.  The cost related to non-restituted gold sales is hidden and obscure as well-documented in a recent JEC study.17   Much of the essential information pertaining to the recent IMF gold sales proposal, for example, was not available to the public.  Accordingly, no rational and informed discussion was able to occur among policymakers in consultation with academic experts and academics.18

    In sum, despite recent improvements in disseminating some data and information, a good deal more reform should be undertaken in order to make the IMF a truly transparent organization.

Summary and Conclusions:

    The U.S. Congress is responsible for decisions related to IMF quota increases as well as to IMF gold sales.  To make informed decisions, policymakers must have a certain amount of essential, yet understandable information.  An understanding of the IMF's financial structure and the costs of U.S. participation in the IMF is especially pertinent.  The JEC has worked to provide and highlight some of this essential information in the form of hearings, research papers, and press releases.

    This paper has reviewed some of the key findings identified in examining IMF financial structure and costs of U.S. IMF participation.  Further areas of investigation remain (e.g., the workings of the SDR department, accurate quantification of risk, etc.).  Nonetheless, it is important that the information conveyed in these lessons be made available and accessible to policymakers and the taxpaying public.

    There are many dimensions to the costs of U.S. participation in the IMF that policymakers and the taxpaying public need to understand.  These include the disproportionate U.S. burden of financial contributions and commitments to the IMF, subsidized interest rates, absorption of risk, and aspects of non-restituted gold sales.  Conservative estimates of the costs of U.S. participation in the IMF suggest that these costs are substantial.  All of this suggests that the U.S. is shouldering a significantly greater proportion of the IMF's financial resources than the oft-cited 17.7 percent quota share would indicate.

    In addition to these U.S. costs, it is important to highlight the changing nature of the IMF financial structure.  The IMF's portfolio has become riskier in a number of ways.  Longer-term loans are increasingly made to lower-grade, higher-risk developing countries.  The IMF's portfolio has become concentrated with a small number of large loans of this type.  Further, the IMF has evolved into an organization that is increasingly redistributionist in nature.  Financial support is increasingly supplied by a small number of industrialized countries while borrowers are for the most part developing countries facing long-term structural problems.

    All of this information, while essential for Congressional decisionmaking, has generally not readily been available to Congressional policymakers or the taxpaying public.  A major reason policymakers and the public are not well informed on these matters is a lack of transparency on the part of the IMF.
 

Robert Keleher                                                                                Christopher Frenze
Chief Macroeconomist                                                                     Chief Economist
to the Vice Chairman                                                                        to the Vice Chairman
 
 



------------------------------------------------------------


 







 
 

Endnotes

1 The General Accounting Office has provided helpful assistance to the JEC in this effort in response to a request by then-Chairman Saxton in 1998.  See, the GAO reports "The Transparency and Financial Structure of the IMF," hearing before the Joint Economic Committee, One Hundred Fifth Congress, Second Session, July 23, 1998; and "Transparency and the Financial Structure of the IMF," hearing before the Joint Economic Committee, One Hundred Sixth Congress, First Session, July 21, 1999.  JEC research papers related to the IMF include: Christopher Frenze, IMF Gold Sales in Perspective, August 1999; Robert Keleher, An International Lender of Last Resort, the IMF, and the Federal Reserve, February 1999; Christopher Frenze and Robert Keleher, IMF Financing: A Review of the Issues, March 1998; and Robert Keleher, Financial Crises in Emerging Markets: Incentives and the IMF, August 1998.  For this and more information on the IMF, please visit our website at www.house.gov/jec.

2 See "Transparency and the Financial Structure of the IMF," hearing before the Joint Economic Committee, One Hundred Sixth Congress, First Session, July 21, 1999.

3 GAO, "Observations on the IMF's Financial Operations," September 1999, p.36.

4 As Lerrick has argued, the 1967 President's Commission on Budget Concepts "defines the budget cost of an 'exchange of assets' program as the difference between the Treasury's cost of funds for the term of the provision of resources and its rate of remuneration."  See Adam Lerrick, Private Sector Financing for the IMF: Now Part of an Optimal Funding Mix, The Bretton Woods Committee, Washington D.C., April 1999, p.11.

5 See Lerrick, op. cit., pp.11-13.

6 This portion originated in the mid-1970s with the demonetization of gold.  When the second amendment to the articles of agreement was passed in 1978, the gold stayed with the Fund.  The U.S. reserve account was credited with an equivalent amount of reserve assets.  This portion was deemed non-interest paying.

7 GAO, "Observations on the IMF's Financial Operations," September 1999, p.56.

8 See Lerrick, op. cit., p.15.

9 These costs are examined and detailed in a recent JEC study.  See Christopher Frenze, IMF Gold Sales in Perspective, Joint Economic Committee, August 1999.

10 Comments of Senators Ribicoff and Taft, The Proposed IMF Agreement on Gold, Report of the Subcommittee on International Economics, Joint Economic Committee, December 17, 1975, p.11 (emphasis and parenthesis added).

11 The U.S. share of a total103 million ounce sale would be 23 percent of the $260-$48 gain, or about $5.02 billion.

12 The U.S. share of the 10 million ounce sale would be 23 percent of the $260-$48 gain, or about $488 million.

13 See Lerrick, op. cit.

14Under the 1967 President's Commission of Budget Concepts, approaches were recommended to costing various programs, including loan programs or asset exchanges.  Subsidization of the cost of funds as well as default risk were to be considered in making cost calculations.  See Lerrick, op. cit.

15 GAO, "Observations on the IMF's Financial Operations," September 1999, p.28.

16 See, for example, Robert Keleher, Transparency and U.S. Dollar Policy, Joint Economic Committee, July 1999, p.2.

17 See Frenze, op. cit.

18 See Frenze, op. cit., p.3.
 

JEC Return Home
Return Home