اقتصاد سیاسی مشروط: تجزیه و تحلیل تجربی پرداختی های وام بانک جهانی

Traditional aid conditionality has been attacked as ineffective in part because aid agencies – notably the World Bank – often fail to enforce conditions. This pattern undermines the credibility of conditionality, weakening incentives to implement policy reforms. The standard critique attributes this time inconsistency to bureaucratic factors within the aid agency such as pressure to lend, defensive lending, or short-sighted altruism. Pressure from powerful donors provides another potential explanation for lax enforcement. This paper presents an empirical analysis of the political economy of conditionality in international organizations using the case of the World Bank and the United States. The analysis examines panel data on World Bank disbursements to 97 countries receiving structural adjustment loans between 1984 and 2005. Using UN voting as an indicator of alignment with the U.S., the paper presents evidence that World Bank structural adjustment loan disbursements are less dependent on macroeconomic performance in countries aligned with the United States.

Traditional aid conditionality has been attacked as ineffective in part because aid agencies – notably the World Bank – often fail to enforce conditions (Mosley et al., 1995, Collier, 1997 and Dreher, 2004). This pattern undermines the credibility of conditionality, weakening reform incentives for governments in developing countries. The standard critique attributes this time inconsistency to bureaucratic factors within the aid agency such as pressure to meet lending targets, defensive lending to promote repayment of past loans, or short-sighted altruism (Svensson, 2003). Yet the cost for the aid agency of lax enforcement is high since it fuels expectations that other conditions – across the agency's portfolio – also will not be enforced and hence broadly undermines compliance. Pressure from powerful donors with geopolitical or commercial interests in the recipient country provides another potential explanation for lax enforcement of aid conditions (as suggested in Kanbur, 2000). This paper presents an empirical analysis of the political economy of conditionality in international organizations using the case of the World Bank and the United States. Because project-level disbursement data are not publicly available for World Bank lending, the analysis examines overall World Bank disbursements when structural adjustment loans (SALs) are active using a panel of 97 countries from 1984 to 2005. I use macroeconomic variables to measure the imposition of structural adjustment. Ceteris paribus, if disbursements are less sensitive to macroeconomic performance for countries friendly with the U.S., this suggests that either ex ante SAL conditions are less stringent or enforcement is lax for U.S. friends. Separately identifying these effects is difficult since data on ex ante World Bank SAL conditions are not systematically available. However, for reasons discussed below, the empirical specification used here is better suited to detecting lax enforcement. Data are widely available only for two relevant macroeconomic indicators, inflation and the percentage change in the official exchange rate. The U.S. interest variable is a measure of UN voting alignment similar to Andersen et al. (2006b). It reflects countries making concessions to the U.S., i.e., deviating from their normal voting position toward the U.S. position on votes that the U.S. considers important. The estimation finds a significant link between macroeconomic performance and disbursements when countries are not aligned with the U.S. but no substantial effect when countries are aligned with the U.S. This is consistent with the donor pressure explanation of lax enforcement. Understanding the reasons for weak conditionality is important as they may influence the success of efforts to reform international organizations. Many reform proposals focus on changing bureaucratic incentives (e.g., linking pay and promotion to outcomes) or on reducing institutional information and commitment problems (e.g., aid tournaments as suggested by Pietrobelli and Scarpa, 1992 and Svensson, 2003). Such reforms may have significant merit but do not address the issue of donor pressure directly and hence could yield smaller gains and be more difficult to implement than expected. Other reforms that restrict direct donor influence in international financial institutions (IFIs) – changes in governance, donor financing, and perhaps headquarters location – also need to be considered.

This paper presents indirect evidence that pressure from the U.S. has undermined World Bank imposition of structural adjustment conditionality. For countries not friendly with the U.S. (countries that do not make concessions to the U.S. position in important UN votes), there does appear to be a significant degree of enforcement. When these countries have active World Bank structural adjustment loans, poor macroeconomic policy is associated with lower disbursements and the effect can be substantial. For countries that are friendly with the U.S., there is little evidence of enforced conditionality. For this second group, there is no substantial link between macroeconomic policy and disbursements. This pattern reoccurs in a range of specifications, across geographic regions, and over different time periods and is robust to a number of estimation methods. In contrast, no similar pattern is found when SALs are not active, again indicating that the pattern is driven by selective imposition of structural adjustment conditionality. These results highlight donor pressure as an important alternate explanation for the failure of conditionality, one that merits more attention from researchers and reformers. This issue has been explored empirically in the context of the IMF (Stone, 2002, Stone, 2004 and Vreeland, 2005) but not previously for the World Bank. Why does it matter what is the cause of conditionality slippage? Efforts to reform structural adjustment have focused increasingly on selectivity to change bureaucratic incentives, reduce problems of information and commitment, and promote ownership of programs (largely through the PRSP process). These reforms may have significant merit but do not address the issue of donor pressure that can, as before, undermine borrower incentives and World Bank credibility. Other more fundamental reforms that aim to reduce donor influence – changes in World Bank governance, ending the tradition of allowing the U.S. to select the World Bank president, developing alternative sources or methods of funding – also need to be explored. Garnering sufficient donor support for such fundamental reforms is not straightforward but may be aided by more research to better understand the impact of international politics on World Bank programs and the costs associated with resulting distortions. Do case studies and other direct evidence (e.g., new data sources that give World Bank disbursements and tranche release conditions by loan) support the indirect evidence presented here? Does international politics also influence which countries get World Bank SALs and the tightness of the conditions spelled out in loan agreements? Are eventual outcomes worse in cases where conditionality was not enforced? The literature on the IMF has explored many of these questions and may provide important guidance. Ultimately, donors like the U.S. have many bilateral instruments they can use to pursue foreign policy objectives. These include bilateral economic aid, military aid and trade policy. Part of the calculus they engage in when deciding how to reward friends (or punish enemies) is to compare the costs of delivering rewards directly via bilateral instruments with the costs of exerting pressure on IFIs like the World Bank. If donors can be convinced that the cost of using IFIs is too high because of deleterious effects on the unique functions of those institutions, then fundamental reforms may be possible.