Published online by Cambridge University Press: 13 June 2011
Analysts have long suspected that politics affects the lending patterns of the International Monetary Fund (IMF), but none have adequately specified or systematically tested competing explanations. This paper develops a political explanation of IMF lending and tests it statistically on the developing countries between 1985 and 1994. It finds that political realignment toward the United States, the largest power in the IMF, increases a country's probability of receiving an IMF loan. A country's static political alignment position has no significant impact during this period, suggesting that these processes are best modeled dynamically. An analysis of two subsamples rejects the hypothesis that the IMF has become less politicized since the end of the cold war and suggests that the influence of politics has actually increased since 1990. The behavior of multilateral organizations is still driven by the political interests of their more powerful member states.
1 The literature on the IMF is extensive. For useful surveys, see Bird, Graham, “The International Monetary Fund and the Developing Countries: A Review of the Evidence and Policy Options,” International Organization 50, no. 3 (1996)CrossRefGoogle Scholar; idem, IMF Lending to Developing Countries: Issues and Evidence (London: Routledge, 1995)Google Scholar; Edwards, Sebastian, “The International Monetary Fund and the Developing Countries: A Critical Evaluation,” NBER Working Paper, no. 2909 (1989)Google Scholar; Killick, Tony, IMF Programs in Developing Countries: Design and Impact (London: Routledge, 1995)CrossRefGoogle Scholar; Williamson, John, ed., The Lending Practices of the International Monetary Fund (Washington, D.C: Institute for International Economics, 1982)Google Scholar; and idem, IMF Conditionality (Washington, D.C: Institute for International Economics, 1983).Google Scholar
2 For effective treatments of these and related issues, see Keohane, Robert O., After Hegemony: Cooperation and Discord in the World Political Economy (Princeton: Princeton University Press, 1984)Google Scholar; idem, International Institutions and State Power: Essays in International Relations Theory (Boulder, Colo.: Westview, 1989)Google Scholar; idem, “Multilateralism: An Agenda for Research,” International Journal 45, no. 1 (1990)Google Scholar; Ruggie, John Gerard, ed., Multilateralism Matters: The Theory and Praxis of an Institutional Form (New York: Columbia University Press, 1993)Google Scholar; Krasner, Stephen D., ed., International Regimes (Ithaca, N.Y.: Cornell University Press, 1983)Google Scholar; and Oye, Kenneth A., ed., Cooperation under Anarchy (Princeton: Princeton University Press, 1986).Google Scholar
International institutions and multilateralism are not necessarily equivalent. The IMF fits Ruggie's definition in Multilateralism Matters of multilateral organization as “defined by such generalized decision-making rules as voting or consensus procedures” (p. 14). On IMF decision-making procedures, see Southard, Frank, “The Evolution of the International Monetary Fund,” Princeton Essays in International Finance, no. 135 (1979)Google Scholar; and Lister, Frederick K., Decision-Making Strategies for International Organizations, vol. 20, World Affairs (Denver, Colo.: Graduate School of International Studies, University of Denver, 1984).Google Scholar
4 Bird (fn. 1,1995), 124.
5 Rowlands, Dane, “Political and Economic Determinants of IMF Conditional Credit Arrangements: 1973–1989” (Manuscript, Norman Paterson School of International Affairs, Carleton University, Ottawa, Ont., 1995).Google Scholar
6 Polak, Jacques J., “Monetary Analysis of Income Formation and Payments Problems,” IMF Staff Papers, no. 6 (1957)Google Scholar, cited in Edwards (fn. 1); and idem, “The Changing Nature of IMF Conditionality,” Princeton Essays in International Finance, no. 184 (1991)Google Scholar; and Swedberg, Richard, “The Doctrine of Economic Neutrality of the IMF and the World Bank,” Journal of Peace Research 23 no. 4 (1986).CrossRefGoogle Scholar
8 Southard (fn. 2), 13; Edwards (fn. 1); Finch, C. David, “The IMF: The Record and the Prospects,” Princeton Essays in International Finance, no. 175 (1989)Google Scholar; and Spraos, John, “IMF Conditionality: Ineffectual, Inefficient, Mistargeted,” Princeton Essays in International Finance, no. 166 (1986).Google Scholar
9 Finch (fn. 8), 2.
10 Schoultz, Lars, “Politics, Economics, and U.S. Participation in Multilateral Development Banks,” International Organization 36, no. 3 (1982)CrossRefGoogle Scholar; Cohen, Benjamin J., “International Debt and Linkage Strategies: Some Foreign Policy Implications for the United States,” in Kahler, Miles, ed., The Politics of International Debt (Ithaca, N.Y.: Cornell University Press, 1986).Google Scholar
11 Smith, Fred L., “The Politics of IMF Lending,” Cato Journal 4 (Spring/Summer 1984).Google Scholar The U.S. representative is “ordered by law to clear his or her decisions with the Secretary of the Treasury.” Swedberg (fn. 6), 379.
12 Stiles, Kendall W., Negotiating Debt: The IMF Lending Process (Boulder, Colo.: Westview, 1991).Google Scholar
14 See Lister (fn. 2).
15 Lichtensztejn, Samuel and Baer, Mónica, Fondo Monetario Internacionaly Banco Mundial: Estrategiasy Politicas delPoder Financiero (Mexico City: Ediciones de Cultura Popular, 1987), 60–61.Google Scholar
16 Miles Kahler notes that the U.S. has in the past refused to support a renewal of the managing director's tenure when his “accomplishments did not meet American expectations.” Kahler, , “The United States and the International Monetary Fund,” in Karns, Margaret P. and Mingst, Karen A., eds., The United States and Multilateral Institutions (Boston: Unwin Hyman, 1990), 94.Google Scholar
17 The origins of this procedure date back to the Fund's early years, when the U.S. executive director went to great lengths to muffle the strong voice of U.S. power, which nevertheless was decisive. See Southard (fn. 2), 5–6,19–20.
18 Eckaus (fn. 7), 237; Stiles (fn. 12), 37.
19 Ruggie (fn. 2), chap. 1; James A. Caporaso and Miles Kahler attribute part of the postwar economic cooperation to this type of “minilateralism.” The creation of the Bretton Woods monetary order through U.S. and British coordination and the subsequent adjustments made by the G-7 after its breakdown (for example, the Plaza and Louvre accords) can be profitably understood in these terms. Caporaso, “International Relations Theory and Multilateralism: The Search for Foundations,” and Kahler, “Multilateralism with Small and Large Numbers,” in Ruggie (fn. 2).
20 Payer, Cheryl, The Debt Trap: The International Monetary Fund and the Third World (New York: Monthly Review Press, 1974)Google Scholar; and Swedberg (fn. 6).
21 Kahler (fn. 16), 110.
23 Killick (fn. 1), 118–19.
24 Stiles (fn. 12), 196–97.
25 Conway, Patrick, “IMF Lending Programs: Participation and Impact,” Journal ofDevelopment Economics 45, no. 2 (1994).Google Scholar
26 He finds statistically significant positive results for prior participation and the percentage of available funds drawn down. A series of dummy variables for each year had generally significant results.
27 Lindert, Peter H., “Response to Debt Crisis: What Is Different about the 1980s?” in Eichengreen, Barry J. and Lindert, , eds., The International Debt Crisis in Historical Perspective (Cambridge: MIT Press, 1989).Google Scholar
28 Bird (fn. 1,1995).
29 This result supports the catalytic impact of IMF lending as providing a “seal of approval” that encourages private banks to resume lending to a country that has negotiated an agreement with the Fund. Bird (fn. 1,1995), 122. A negative result would suggest a substitution effect between IMF and private lending.
30 Spraos (fn. 8); and Finch (fn. 8).
31 Bird (fn. 1, 1995), 109.
34 See Bello, Walden and Kinley, David, “The IMF: An Analysis of the International Monetary Fund's Role in the Third World Debt Crisis, Its Relation to Big Banks, and the Forces Influencing Its Decisions,” Multinational Monitor 4 (1983).Google Scholar
35 Lindert (fn. 27), 245.
36 Killick (fn. 1).
37 Lindert (fn. 27), 243; Bird (fn. 1, 1995), 112.
38 Barry Eichengreen has questioned the impact of the “default penalty” on future credit access. Eichengreen, , “The U.S. Capital Market and Foreign Lending, 1920–1955,” in Sachs, Jeffrey D., ed., Developing Country Debt and the World Economy (Chicago: University of Chicago Press, 1989), 247.Google Scholar Cf. Eaton, Jonathan and Gersovitz, Mark, “Debt with Potential Repudiation: Theoretical and Empirical Analysis,” Review of Economic Studies 48 (April 1981).CrossRefGoogle Scholar
39 See Brett, E. A., “The World's View of the IMF,” in Bureau, Latin America, ed., The Poverty Brokers: The IMF and Latin America (London: Latin America Bureau, 1983)Google Scholar; Pastor, Manuel, “The Effects of IMF Programs in the Third World: Debate and Evidence from Latin America,” World Development 15 (Fall 1987)CrossRefGoogle Scholar; and Swedberg (fn. 6).
40 The net effect of DFI exposure may depend on the sectoral location of the investment. If it serves primarily the domestic market, a negative result might be expected. If it serves mostly export markets, a positive result would be more likely. The impact of export exposure may depend on whether the product exported is a final consumption good (negative) or an input into the export sector (positive).
41 Bello and Kinley (fn. 34), 14.
42 Awanohara, Susumu, “Fiscal Interdiction: U.S., Japan Block IMF Effort to Support Vietnam,” Far Eastern Economic Review, September 28, 1989.Google Scholar
43 Amuzegar (fn. 33); Assetto, Valerie J., The Soviet Bloc in the IMF and the IBRD (Boulder, Colo.: Westview, 1988).Google Scholar
45 Horowitz, Irving Louis, “The ‘Rashomon Effect’: Ideological Proclivities and Political Dilemmas of the IMF,” in Myers, Robert J., ed., The Political Morality of the International Monetary Fund (New Brunswick, N.J.: Transaction Books, 1987), 96.Google Scholar
46 Assetto (fn. 43), 184.
48 Timothy J. McKeown, “Resolving the ‘Conditionality Paradox’ in U.S. Bilateral Foreign Aid” (Manuscript, University of North Carolina, Chapel Hill, n.d.).
49 Bird (fn. 1,1995), 149–50.
50 Regressions were run on the amount of the loan divided by GNP, and the general results were similar to those reported here, particularly for the political variables.
51 This figure represents all of the developing countries, as defined by the IMF, for which data were available. See IMF, AnnualReport (Washington, D.C.: IMF, 1986), 162.Google Scholar Data for the indicator of political alignment used here are not available before 1983. For some countries, data are available only for certain years. See Appendix B for a list of countries used in the data analysis.
53 Two other IMF lending programs, the Structural Adjustment Facility (SAF) and the Enhanced Structural Adjustment Facility (ESAF), are not included in this analysis for a number of reasons. First, only low-income developing countries qualify for SAF and ESAF loans. A large number of countries in the sample would therefore not qualify for these programs, while all are eligible for SBA and EFF packages. Second, the SAF and ESAF are structural adjustment rather than economic stabilization programs. To include them in the analysis would require a different underlying macroeconomic model than that specified for SABs and EFFs. Third, 1987 was the first full year of operation for the SAF and 1988 for the ESAF. Only SBA and EFF programs were operational throughout the entire time period examined here. See Polak (fn. 6, 1991); and Schadler, SusanBennett, AdamCarkovic, MariaDicks-Mireaux, LouisMecagni, MaruoMorisink, James H. J., and Savastano, Miguel A, “IMF Conditionality: Experience under Stand-By and Extended Arrangements. Part I: Key Issues and Findings,” IMF Occasional Paper, no. 128 (1995).Google Scholar Compared to the number of SBAs and EEFs, there have been few SAF and ESAF loans made. Regressions run on a variable including all of these programs together yielded results generally consistent with those reported in the following section.
54 Logit transforms this variable, which has a nonlinear relationship to the independent variables, into the log-odds of receiving a loan, which has a linear relationship to the independent variables. The new dependent variable, or logit, is then regressed on the independent variables using maximum likelihood estimation (MLE). Data for this variable were gathered from IMF, Annual Report (Washington, D.C.: IMF, various issues).
55 All economic variables except ratios are expressed in millions of 1990 U.S. dollars, using the 1990 U.S. GDP deflator reported in IMF, International Financial Statistics Yearbook. (Washington, D.C.: IMF, various issues).
56 These variables make the figures for large and small countries more comparable. I also tested the ratio of balance of payments to GNP and the change in that ratio with the same substantive results. Data are from IMF, International Financial Statistics Yearbook (Washington, D.C.: IMF, various issues).
57 The World Bank's debt ratios (DEBT/GNP, INT/GNP, and RES/DEBT) appear to have been multiplied by 100. To make comparisons across units consistent, I multiplied the CACCT/GNP ratios calculated from (but not listed in) World Bank data by 100. World Bank, World Debt Tables (Washington, D.C.: World Bank, various issues); and idem, Global Development Finance (Washington, D.C.: World Bank, various issues).
58 These figures are from the World Bank, World Debt Tables (Washington, D.C.: World Bank, various issues); and idem, Global Development Finance (Washington, D.C.: World Bank, various issues); with population data taken from IMF, International Financial Statistics Yearbook (Washington, D.C.: IMF, various issues).
59 Adequate data on the exposure of U.S. banks in particular countries are unavailable. In any event, the largest creditor banks are likely to be based in the U.S. and the IMF's other principal shareholder countries.
60 A variable measuring the total number of cancellations that a country experienced from 1975 through t-1 did not yield statistically significant results. Data were gathered from IMF, Annual Report (Washington, D.C.: IMF, various issues).
61 Data are from IMF, Direction of Trade Statistics Yearbook (Washington, D.C.: IMF, various issues).
62 Data have been taken from the U.S. Department of Commerce, Survey of Current Business, various issues.
63 Using these annual reports, I coded votes in agreement with the U.S. 1.0, votes in disagreement with the U.S. 0.0, and abstentions or absences by the sample country 0.5.1 then added and divided these numbers by the total number of key votes each year to come up with the annual decimal measure for each country. This method differs slightly from the technique of discarding absences and abstentions from the total count of UNGA votes used in Kegley, Charles W. Jr. and Hook, Steven W., “U.S. Foreign Aid and U.N. Voting: Did Reagan's Linkage Strategy Buy Deference or Defiance?” International Studies Quarterly 35 (September 1991).CrossRefGoogle Scholar Rather than not count those nonvotes on “key” issues, I interpret them as neutral.
64 The transmission of United States foreign policy preferences from the State Department is not necessarily direct in the case of the multilateral development banks and the IMF, where Treasury plays a critical role. See Schoultz (fn. 10). The (American) deputy managing director has typically been “a ‘Treasury man,’ reinforcing the close ties between that agency of the U.S. government and the IMF.” Kahler (fn. 16), 94. Furthermore, Kahler argues that Treasury maintains tight control over U.S.-Fund relations and that “other agencies that might attempt to politicize the IMF for broader foreign policy goals tended to be excluded from direct access to it.” Kahler (fn. 16) 94, 97. On the other hand, Joanne Gowa notes that Treasury has adopted an ordering of priorities that “subordinates the demands of the international monetary order to the imperatives of domestic economic policy and foreign security policy,” suggesting some coordination—or at least compatibility—between different agencies within the executive branch. Gowa, , Closing the Gold Window: Domestic Politics and the End ofBretton Woods (Ithaca, N.Y.: Cornell University Press, 1983).Google Scholar The present analysis of policy output (as opposed to interagency input) is an indirect test of these two competing hypotheses. Future work should address the interagency dynamics more directly.
65 U.S. Department of State, Report to Congress on Voting Practices in the United Nations (1985), 2.Google Scholar
67 See Soo Kim, Yeon and Russett, Bruce, “The New Politics of Voting Alignments in the United Nations General Assembly,” International Organization 50, no. 4 (1996)CrossRefGoogle Scholar; Holloway, Steven K. and Tomlinson, Rodney, “The New World Order and the General Assembly: Bloc Realignment at the UN in the Post-Cold War World,” Canadian Journal of Political Science 28, no. 2 (1995)CrossRefGoogle Scholar; Pallansch, Leona and Zinni, Frank Jr., “Demise of Voting Blocs in the General Assembly of the UN? A Multidimensional Scaling Analysis” (Paper presented at the annual meeting of the Southern Political Science Association, Atlanta, 1996)Google Scholar; Tomlin, Brian W., “Measurement Validation: Lessons from the Use and Misuse of UN General Assembly Roll-Call Votes,” International Organization 39, no. 1 (1985)CrossRefGoogle Scholar; and Menkhaus, Kenneth J. and Kegley, Charles W. Jr., “The Compliant Foreign Policy of the Dependent State Revisited: Empirical Linkages and Lessons from the Case of Somalia,” Comparative Political Studies 21, no. 3 (1988).CrossRefGoogle Scholar
68 Ed Lansdale, “Memo Re: Long Range Impact FPF-II,” April 24, 1964, National Archives, Record \ Group 59, Lot file 67D554, Under Secretary for Political Affairs, Records of the Special Assistant ] 1963–65, Box 2.1 thank Tim McKeown for providing me with a transcription of this document.
69 For example, Argentina sent troops to the 1991 Persian Gulf conflict. Escudé, Carlos, “Entrevista i a Escude realizada por Lorena Kniaz” (1997)Google Scholar, cited May 19, 1999, http://www.geocities.com/CapitolHill/Congress/4359/reporta.html.
72 Because of the UNGA's voting calendar, the voting variables have a longer lag structure than the economic variables. The fact that UNGA votes are taken in the last four months of the calendar year means that there is a 67 percent chance that a given loan decision will be made before the UNGA meets in a given year. The chances that such a decision will be made before the session is complete and final votes are tallied approaches 100 percent. Conversely, movement at t-1 occurs immediately before the next calendar year's loan cycle begins.
73 Stimson, James A., “Regression in Space and Time: A Statistical Essay,” American Journal of Political-Science 29, no. 4 (1985)CrossRefGoogle Scholar; Beck, NathanielKatz, Jonathan N., and Tucker, Richard, “Taking Time Seriously: Time-Series-Cross-Section Analysis with a Binary Dependent Variable,” American Journal Political Science 42, no. 4 (1998).CrossRefGoogle Scholar
74 This assumes a Chi Square distribution for the −2 ⋉ LLR figure. While this assumption may not be entirely valid for individual level data, the strong results are still encouraging.
75 Beck, Katz, and Tucker (fn. 73). This approach is designed for longitudinally dominant data with typically twenty or more time periods. The authors have not yet tested this exploratory method on shorter time periods like the one used here (T = 10). Richard Tucker, conversation with the author, August 1998. We may therefore have somewhat less confidence in a negative diagnostic for autocorrelation than in the positive one obtained here.
76 Conway (fn.25).
77 Per capita debt reached the 0.90 level of confidence, and the following variables attained the 0.99 level: the debt to GNP ratio, the interest to GNP ratio, the change in the interest to GNP ratio, and the ratio of reserves to debt. Curiously, the coefficient for debt to GNP is negative (all others are correctly signed). I have no explanation for this anomalous result, except to surmise that the impact of high relative levels of debt may be sensitive to the burden of repayment as captured by the interest to GNP ratio.
78 On the Mexican crisis, see Roett, Riordan, “The Mexican Devaluation and the U.S. Response: Potomac Politics, 1995-Style,” in Roett, , ed., The Mexican Peso Crisis: International Perspectives (Boulder, Colo.: Lynne Rienner, 1996).Google Scholar
79 Bird (fn. 1, 1995).
80 This result may be spurious. Bird suggests that requesting a loan from the IMF has a threshold effect; once a country requests one loan, it is more likely to request additional loans. Since any country that has a loan canceled has already crossed this threshold, it may be more likely to receive loans in the future. Bird (fn. 1, 1995).
81 Regressions were also run using each variable without the other, yielding similar negative results.
82 There was a possible collinearity problem with this variable. Specifically, it correlated strongly with MKVOTEit-t. Because the inclusion of the interactive term is likely to have inflated the standard error of the movement variable and because its inclusion did not significantly improve the model's fit, I did not retain it.
83 Because of the potential for omitted variable bias and the negative diagnostic for multicollinearity in the original specification, I have greater confidence in the results of the full model. I therefore present the refined model results for the interested reader but focus most of the substantive interpretations on the full model.
84 The change in the current account from t—1 to t is significant at the 0.95 level, and the ratio of the current account to GNP is significant at the 0.90 level. The unexpectedly positive coefficient for the change in the current account from t-1 to t could be due to the fact that an IMF loan at time t can itself cause an improvement in the balance of payments at time t.
85 I retained absolute alignment position here to facilitate a clearer comparison of the hypothetical scenarios and to create more difficult conditions for demonstrating the strength of the impact of political realignment. Omitting KVOTEit-2 would lower the probability for the static U.S. ally even more, relative to any country moving toward the U.S.
86 Killick (fn. 1), 128.
87 Killick (fn. 1).
88 See McKeown (fn. 48).
89 Only one developing country (China) is a permanent member of the Security Council, so we cannot use Security Council votes to measure alignment.
90 Data from World Bank, World Development Indicators 1998 (Washington, D.C.: World Bank, various issues); idem, World Debt Tables (Washington, D.C.: World Bank, various issues); and idem, World Bank Global Development Finance (Washington, D.C.; World Bank, various issues).
91 The results for some of the economic variables differ from those in Table 1. Given the smaller number of cases used in Table 4 (a result of data availability), I base my substantive interpretations on the results presented in Tables 1 and 3. Several intermediate and refined specifications yielded similar results.
92 Pallansch and Zinni (fn. 67).
93 Ruggie (fn. 2).
94 See McKeown (fn. 48).
95 See Kahler (fn. 16), 93
96 Cited in Butler, Nick, The IMF, Time fir Reform (London: Fabian Society, 1982), 24.Google Scholar
97 Cf. Ruggie (fn. 2), chap. 1.