Development aid and international politics: Does membership on the UN Security Council influence World Bank decisions?

Journal of Development Economics 88 (2009) 1 – 18 www.elsevier.com/locate/econbase

Development aid and international politics: Does membership on the UN Security Council influence World Bank decisions? Axel Dreher a,b,⁎, Jan-Egbert Sturm a,b , James Raymond Vreeland c a

ETH Zurich, KOF Swiss Economic Institute, Weinbergstrasse 35, CH-8092 Zürich, Switzerland b CESifo, Germany c Yale University, Department of Political Science, USA Received 13 July 2007; received in revised form 2 February 2008; accepted 9 February 2008

Abstract We investigate whether elected members of the UN Security Council receive favorable treatment from the World Bank, using panel data for 157 countries over the period 1970–2004. Our results indicate a robust positive relationship between temporary UN Security Council membership and the number of World Bank projects a country receives, even after accounting for economic and political factors, as well as regional, country and year effects. The size of World Bank loans, however, is not affected by UN Security Council membership. © 2008 Elsevier B.V. All rights reserved. JEL classification: O19; O11; F35 Keywords: World Bank; UN Security Council; Voting; Aid

1. Introduction Founded in the aftermath of the Great Depression and World War II, the World Bank was created to help rebuild war torn Europe. In the years following its inception, the focus of the World Bank gradually shifted to developing countries. Today it is the primary international institution responsible for promoting economic development in the world. Sponsoring projects of various scopes in both emerging market countries as well as the world's poorest countries, in 2006 alone the World Bank provided $23.6 billion in loans and grants through 279 projects around the globe. Critics (e.g. Easterly, 2005) allege that the World Bank has fallen far short of its goals of improving living standards and reducing poverty. Many argue that failure is due to the imposition of misguided policy conditions through the development projects. One possible reason for this is that instead of enforcing sound development policies, the World Bank has been used as a tool of foreign policy to funnel money to corrupt governments in strategic positions who ally themselves with the major share⁎ Corresponding author. E-mail addresses: [email protected] (A. Dreher), [email protected] (J.-E. Sturm), [email protected] (J.R. Vreeland). 0304-3878/$ - see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jdeveco.2008.02.003

holders of the World Bank — the United States, Japan, Germany, France, and the United Kingdom. The World Bank itself freely admits that during the Cold War its lending was politically driven.1 With only a few notable exceptions, however, researchers have not systematically investigated this claim. Do international political imperatives guide the so-called development lending of the World Bank? Anecdotes are plentiful, but only Schneider et al. (1985), Frey and Schneider (1986) and, more recently, Andersen et al. (2006) and Dreher and Sturm (2006) put this question to a large-n test. They use clever proxies to capture the importance of developing countries to the major shareholders of the World Bank, such as the quantity of exports from major shareholders to developing countries, their former colonial status, and voting patterns at the United Nations General Assembly.2 These studies present some 1

http://web.worldbank.org/WBSITE/EXTERNAL/EXTSITETOOLS/0,, contentMDK:20264002~menuPK:534379~pagePK:98400~piPK:98424~ theSitePK:95474,00.html (accessed June 14, 2007). 2 Voting patterns in the UN General Assembly are also used to investigate the politics involved in bilateral aid (Ball and Johnson, 1996; Boschini and Olofsgard, in press; Alesina and Weder, 2002; Fleck and Kilby, 2006). Kilby (2006) employs UN voting patterns in his analysis of donor influence on the Asian Development Bank.

2

A. Dreher et al. / Journal of Development Economics 88 (2009) 1–18

evidence that political imperatives do influence the Bank. In this paper, we offer the first systematic study analyzing a new measure of the international importance of a country: temporary membership on the United Nations Security Council (UNSC). Recent evidence indicates the importance of temporary UNSC members to powerful countries. Kuziemko and Werker (2006) find that US foreign aid increases when countries serve on the UNSC, as does UN Development Program aid, and Dreher et al. (2006) find that the ten temporary members of the UNSC are more likely to receive IMF assistance than other countries. They attribute these increases in various forms of foreign aid to vote-trading activities: temporary members can trade their Security Council votes for cash. There is reason to believe that foreign aid is not only provided to help countries in economic distress but also to achieve the donor's political objectives. In fact, since the late 1940s, every US administration considered foreign aid to be important in achieving foreign policy goals (Ruttan, 1996). It has even been claimed that the primary purpose of US economic assistance is in promoting overall US policy objectives (Zimmermann, 1993). According to Morgenthau (1962: 302), “the transfer of money and services from one government to another performs here the function of a price paid for political services rendered or to be rendered.” Our question is whether World Bank lending is used by the institution's major shareholders for a similar purpose. While nearly all countries in the world are members of the World Bank and all have votes, these votes are pegged to economic size, and the G7, for example, has an inordinate amount of power at the Bank as a result. They control well over 40% of the votes. When they coordinate, they have veto power over certain important decisions that require supermajorities. Alone they constitute a near majority, and need the support of only a handful of allies to guarantee control of the Bank's loans and grants to developing countries. Clearly they control the World Bank. Do they use this control purely to help developing countries in need or do international politics also play a role in how they choose to guide the development institution? To anticipate our results, we find that temporary Security Council membership does increase the number of World Bank projects a country receives. The qualitative results are robust to the inclusion of economic and political factors, as well as regional, country, and year specific effects. We proceed as follows. The next section provides some background on the UNSC and the World Bank and develops our hypothesis. Section 3 presents anecdotal evidence, while Section 4 presents rigorous analysis of large-n data where we test alternative hypotheses. We discuss extensions in Section 5. Section 6 concludes. 2. The argument The Security Council is the principal organ of the United Nations with responsibility for the maintenance of international peace and security. Its duties include taking military action against aggressors and investigating disputes or situations likely to lead to international frictions. The Security Council has the

power to make binding resolutions and may adopt legally binding measures in order to maintain or restore international peace — including the use of armed forces. Ten of the fifteen seats on the UN Security Council are held by temporary members, while five members – China, France, Russia, the United Kingdom, and the United States – serve on a permanent basis. Each year, five temporary members are elected for a two-year term. They are nominated by their regional caucus and have to be approved by at least two thirds of the votes in the General Assembly. Since 1966, the composition of the 10 temporary members has been: 3 from Africa, 2 from Asia, 2 from Latin America and the Caribbean, 1 from Eastern Europe, and 2 advanced industrial countries.3 The elections are held about three months before the term starts on January 1. Each member of the UNSC has one vote. Decisions on substantive matters require a majority of nine votes. The permanent members have veto powers. Thus, for any given matter, the votes of four out of the ten temporary members are required, in addition to all five votes of the permanent members. All votes are by open ballot. According to Malone (2000), there is extensive competition for the temporary seats on the Security Council. Countries might expect to receive net rewards during their tenure. Kuziemko and Werker (2006) find, for example, that average US aid increases by 54% and average UN development aid by 7% when a country is elected to the Security Council. Another benefit of serving on the Council is the likelihood of receiving greater support from the IMF (Dreher et al., 2006). Why would the US seek to influence the temporary members of the UNSC? As discussed above, the body has jurisdiction over such important UN matters as military action and sanctions. When acting unilaterally to pursue its global interests, the US may face substantial costs. When acting through the UNSC, the US bears a smaller share of military burdens and needs fewer soldiers (Sandler and Harley, 1999). So, perhaps the US and other powerful countries attempt to pressure members of the UNSC to vote their way on important issues.4 Yet, less than half of the votes of the ten elected members are required for UNSC measures to pass. Of the nine out of fifteen votes required, five must come from the permanent members with veto power and only four from the elected members. Elected members are rarely pivotal. O'Neill (1996) shows, for example, that the cumulative voting power (using the Shapley– Shubik index) of all ten elected members of the UNSC is less than 2%. Nevertheless, the US may seek out the support of UNSC members as insurance votes. It is well established in the

3 For the five combined seats for Africa and Asia, one country always comes from an Arab country. Before 1966, there were only six elected members of the UNSC. Composition was typically: two from Latin America, one from the Middle East, one from Eastern Europe, and two advanced industrial countries. For more on the UNSC, see Russett (1997), Dreher and Vreeland (2008), Hurd (2007, 2008), Hurd and Cronin (2008), and Voeten (2008). 4 Bennis (1997) claims that “US influence in (and often control of) the UN comes in the form of coercing the organization to take one or another position, or to reject some other position, or pressuring a country or countries to vote a certain way in the General Assembly.”

A. Dreher et al. / Journal of Development Economics 88 (2009) 1–18

vote-buying literature that over-sized coalitions tend to be established (see, e.g., Volden and Carrubba, 2004). Besides seeking a coalition of countries to ensure the passage of resolutions, the US and other important countries may seek the support of the UNSC for reasons of legitimacy. In the absence of UNSC legitimacy, US domestic public support might be more difficult to achieve, and Congress might also be recalcitrant (Voeten, 2001; Hurd, 2007).5 Chapman and Reiter (2004) indeed find that “UN Security Council support significantly increases the rally behind the president (by as many as 9 points in presidential approval)… This [robust] effect is unique among international institutions because other actions by the UN or regional security organizations do not significantly affect rallies.”6 The view that the UNSC matters mostly for reasons of legitimacy is consistent with the observation that there is a premium for getting (near) unanimous votes (see, e.g., Doyle, 2001: 223). For example, it is well known that the US threatened Yemen with dire consequences (which materialized) if it did not vote in favor for resolution 678, on the use of armed forces against Iraq in 1990, even though Yemen's vote was in no way pivotal (e.g. Pilger, 2002). As another example, consider the run up to the recent war in Iraq where President Bush was going to take the vote to the UNSC even though he knew the French were going to veto. The US was lobbying some of the non-permanent members with aid packages in an attempt to win a simple majority in the Council (Eldar, in press). This makes no sense in terms of the UNSC's institutional rules, but it might have provided some legitimacy to the war for the US and international audiences. The legitimacy view of the UNSC is also consistent with Kuziemko and Werker's (2006) claim that the impact of UNSC membership on foreign aid holds even when the outcome of the vote is obvious a priori.7 To the extent that the US and other powerful countries care about how elected members of the UNSC vote, they may be willing to use other means beyond direct foreign aid to influence them. As the World Bank annually lends between US$15 and $25 billion for projects in over 100 countries,8 the institution may be well suited for this purpose. Its most powerful members

5

Voeten (2001) provides examples. He cites the memoirs of James Baker (1995: 278), emphasizing domestic support to be the main reason for the US government to seek a multilateral solution to the Gulf War. He also cites Malone (1998: ix), arguing that it was easier for the Clinton administration to secure support of the UNSC as compared to those of the US Congress. 6 For a general argument and a case study of the Gulf War, see Thompson (2006). 7 In contrast to arguments about legitimacy, Fang (2006) argues that leaders seek approval by the UNSC to (falsely) signal their type to voters. She argues that under certain conditions, sincere and insincere leaders regarding their foreign policy intentions may both bring actions before the UNSC. An alternative explanation comes from the literature on the informational role of committees. Elected representatives may have influence in their regions because of the access to information of various sorts they have through the UNSC. Thus, the support of regional representatives on the UNSC may be sought because of the influence they have with countries of their region. We are grateful to Joanne Gowa for this suggestion. 8 http://web.worldbank.org/WBSITE/EXTERNAL/EXTABOUTUS/0,, contentMDK:20103853~menuPK:1697035~pagePK:51123644~piPK: 329829~theSitePK:29708,00.html and http://web.worldbank.org/WBSITE/ EXTERNAL/PROJECTS/0,,contentMDK:20120731~menuPK:41390~pagePK: 41367~piPK:51533~theSitePK:40941,00.html. Accessed October 17, 2007.

3

can use the Bank's resources to influence recipient countries' behavior in line with their interests. The World Bank consists of two institutions, the International Bank for Reconstruction and Development (IBRD), which lends to middle income countries, and the International Development Association (IDA), which lends to the world's poorest countries.9 A World Bank “project” – be it small or large in scope with either long-run or short-run goals – is supposed to provide a loan to developing countries to improve standards of living and reduce poverty. The Bank does not provide the entire loan upfront, however, but rather practices “conditionality.” To avoid subsidizing bad policies or corrupt governments, the Bank typically conditions its loans on policy changes the government must undertake. Continued disbursements of the loan are supposed to depend on compliance with the agreed upon policy conditions. One common condition for the Bank's structural adjustment lending is that a loan with its sister institution – the IMF – be in good standing, which means that the government may be practicing fiscal austerity and tight monetary policy, undertaking structural reforms, and perhaps devaluing the national currency. In addition, the World Bank has also stressed good governance and anti-corruption in recent years. Whether governments actually comply with conditions is hotly debated.10 Despite the guidelines on how World Bank lending is supposed to work, there is some evidence that the Bank has been used for political purposes. The World Bank candidly acknowledges on its website, “It is true that during the Cold War years aid was politically motivated.”11 The Bank officially denies that lending continues to be politically motivated, and there is evidence in favor of this claim. Dreher and Sturm (2006), for example, show that World Bank (and IMF) lending influenced voting in the UN General Assembly during the Cold War but not thereafter. While countries voting in line with the G7 received larger loans from the Bank until 1990, this pattern does not hold for the years 1991–2002. Nevertheless, there are post-Cold War anecdotes. In the early 1990s, the US facilitated World Bank meetings with Egypt in return for support in the Gulf (Momani, 2004). The US, Japan, Germany, France, and the UK clearly have much control over the World Bank.12 Why would these powerful countries employ the World Bank to favor key 9 The World Bank Group refers to the IBRD, the IDA, as well as three other organizations: the International Finance Corporation (IFC), which encourages private sector investment in coordination with the IBRD; the Multilateral Investment Guarantee Agency (MIGA), which facilitates foreign direct investment in developing countries by providing insurance, mediating disputes, advising governments, and providing information; and the International Centre for Settlement of Investment Disputes (ICSID), which facilitates the conciliation and arbitration of investment disputes between members and nationals of other countries. 10 See Vreeland (2006, 2007). 11 http://web.worldbank.org/WBSITE/EXTERNAL/EXTSITETOOLS/0,, contentMDK:20264002~menuPK:534379~pagePK:98400~piPK: 98424~theSitePK:95474,00.html (accessed June 14, 2007). 12 Decisions by the Executive Board are usually taken by consensus. The power of the World Bank's major shareholders might thus be larger than the sum of their votes as it is unlikely that a decision on a project will be adopted when representatives of the major powers object.

4

A. Dreher et al. / Journal of Development Economics 88 (2009) 1–18

countries?13 The benefits for powerful countries of bribing or rewarding other governments indirectly via the Bank instead of directly with their own aid programs might be substantial. We identify political benefits, leverage benefits, and cost benefits. As for political benefits, national governments may delegate unpleasant tasks or “dirty work” to international organizations. This might allow governments to escape the nationalist resentment such actions would create if imposed through more obvious and direct means (Vaubel, 1986, 1991, 1996). According to Eldar (in press), Security Council votes are almost always traded behind the scenes because most countries prefer to conceal vote-trading arrangements to escape public condemnation. This holds for both the pressuring and the pressured countries. The conditionality attached to World Bank loans may also prove useful because it provides a leverage mechanism. International Financial Institutions can impose punishments and rewards much wider in scope than any single donor can (Harrigan et al., 2006). World Bank conditionality is frequently used in concert with the conditionality of its sister institution, the IMF. In terms of the cost benefits, Eldar (in press) argues that loans extended through international organizations are usually much more valuable to potential borrowers than they are costly to the US. This holds for the other major shareholders of the World Bank as well. Not only are burdens shared, but the lending facility for emerging market countries, the IBRD, raises most of its lending funds by issuing bonds on financial markets.14 IDA, which lends to poor countries, does get most of its resources from the major shareholders, but some of its budget is financed by income generated by the IBRD and part of the budget comes from the repayments of borrowers from IDA.15 So what is the precise mechanism by which the World Bank is used to deliver favors to temporary members of the UNSC? We suggest that both the major shareholders and the temporary UNSC members understand the importance of votes on the UNSC. For their part, most developing countries are in routine negotiations with the World Bank to receive new loans or continued disbursements from existing projects. Decisions over such requests first go through the staff and then are decided by 13

What happens when the G7 do not coordinate? No one has yet looked at this question for the World Bank, but for the IMF, Copelovitch (2007) suggests that when the five most powerful members coordinate (the G5, as he calls them), they pursue their common political objectives. When they do not, they leave a power vacuum filled by IMF staff who pursue technocratic but also bureaucratic goals. For more on such a Public Choice view of the IMF and the World Bank, see Vaubel (1996). An extension of this paper would be to check whether there is a correlation between voting coincidence with the US, the UK and France in the UNSC and the number of World Bank projects initiated. This exercise would be hampered by the fact that most UNSC resolutions are adopted by consensus, but some split votes have occurred. We intend to study this in future research. We are grateful to the anonymous reviewers for raising these points. 14 http://web.worldbank.org/WBSITE/EXTERNAL/EXTABOUTUS/EXTIBRD/0,, menuPK:3046081~pagePK:64168427~piPK:64168435~theSitePK:3046012,00. html. Accessed October 17, 2007. 15 http://web.worldbank.org/WBSITE/EXTERNAL/EXTABOUTUS/IDA/0,, contentMDK:21206704~pagePK:51236175~piPK:437394~theSitePK:73154,00. html. Accessed October 17, 2007.

the Executive Board. The staff may recognize the importance of a temporary UNSC member and expedite the request so that it reaches the Executive Board faster. Or requests may simply come up routinely at the weekly Board meetings, where country representatives push them forward. Ultimately, the World Bank Executive Board makes the key decisions. We suspect that the major shareholder representatives on the Executive Board of the World Bank are made well aware of the potential importance of the ten current elected members of the UNSC by their home governments. The home governments instruct them to facilitate projects for such countries. All of the major shareholders can easily agree that temporary members of the UNSC are potentially important. Should a significant issue come up during the tenure of a temporary UNSC member, it behooves the major shareholders to have the country in their debt.16 Providing them World Bank projects is a relatively low cost way to provide them a favor. Hence, if temporary members request, they facilitate getting the projects through the board, with the implicit promise that the loans will be cut off if they misbehave on the UNSC.17 In sum, it is best for all major shareholders to put the elected UNSC members in a position of owing them favors.18 We thus hypothesize: UN Security Council membership increases the probability of receiving World Bank projects. 3. Anecdotal evidence Examples of G7 countries manipulating International Financial Institutions to achieve foreign policy goals abound. Consider the Gulf War period. In 1991, the US supported a World Bank loan for China in exchange for China's support of the Security Council resolution to deploy armed forces in Iraq (Eldar, in press). A similar exchange transpired in 1994, when China agreed to abstain on the Security Council resolution to restore democracy in Haiti. The US again facilitated a World Bank loan for China (Eldar, in press).19 Turning to elected members of the UNSC during the Gulf War, Eldar (in press) reports that the US promised “financial help to Columbia, Côte d'Ivoire, Ethiopia and Zaire.” In the run up to the more recent invasion of Iraq, the US (unsuccessfully) tried to buy the votes of temporary UNSC members, principally targeting elected members who “had not decided how to vote, namely: Angola, Guinea, Cameroon, Pakistan, Chile and Mexico” (Eldar, in press). 16 Again, the World Bank may be a less effective tool when the major shareholders disagree on a specific resolution, but we suspect that they can agree on the potential importance of a temporary UNSC member a priori, even if they subsequently disagree over a specific issue before the Security Council. 17 It is not just the US who cares, of course. Consider the second largest shareholder at the World Bank, Japan. Japan cares a great deal about the UNSC, as evidenced by its attempts to get a permanent seat, by its numerous successful attempts to get elected to the UNSC (more than any other country), and by the direct foreign aid it provides to countries that get elected to the UNSC. See Dreher and Vreeland (2008). 18 The logic of the incentive approach to gaining policy favors is explored more recently in Dorussen (2001), Drezner (1999), Bernauer and Ruloff (1999), Cortright (1997), Long (1996), and Wagner (1988). For a more skeptical view, see Moon (1985). 19 The US also granted China security guarantees regarding Taiwan.

A. Dreher et al. / Journal of Development Economics 88 (2009) 1–18

Beyond the prominent Gulf War examples, consider the following examples from around the world. In Latin America, Argentina had no new World Bank projects in 1970, but two new projects were initiated in 1971, when Argentina joined the UNSC. Later, Argentina had two new World Bank projects in 1985, and it did not belong to the UNSC. It was elected to the UNSC in the fall of 1986, however, and the number of new World Bank projects doubled that year. Four projects started during the first year of its term on the UNSC in 1987, and the number of new projects jumped to seven in 1988 while Argentina continued to serve. The number of projects then dropped down to one in 1989, but peaked again in 1995 — Argentina had again been elected to the UNSC and was serving the second year of its term. It would be difficult to establish on a case by case basis whether Argentina was not “deserving” of these World Bank projects on purely economic grounds. After all, according to its level of GDP per capita, Argentina is exactly the kind of emerging market country that the IBRD is supposed to help. In terms of economic variables that robustly predict World Bank projects, there is no consensus in the literature as there is for IMF programs. Nevertheless, it is instructive to consider some economic indicators. Consider the peak in new World Bank projects in 1995.20 Debt service as a percentage of gross national income came down from nearly 4.6% in 1990 to 2.3% in 1994. Foreign reserves held steady throughout the early 1990s, between 5 and 7 times monthly imports. Trade was steadily rising: 15% of GDP in 1990, 18% in 1994. GDP per capita growth did slow from an amazing 15% in 1993 to just 4% in 1994. Certainly, one could find some economic indicator that makes Argentina worthy of World Bank assistance for any year, but there was nothing outstanding economically in the run up to 1995 that made this the year for the number of new World Bank projects to peak. Turning to Africa, consider Ghana. Its first ever World Bank project was coincident with its first term on the UNSC in 1962– 3. From 1964 to 1985, the country received an average of about 1.5 new World Bank projects per year, with the highest number of new projects being four in some years. Then Ghana was elected to serve again on the UNSC in 1986–7, and the number of new projects jumped to its highest level ever: eight in 1987. As for an economic justification for this jump, again, one can always find reasons for the World Bank to loan to a poor African country like Ghana. Growth of GDP per capita slowed from 4.7% in 1984 to 1.5% in 1985 and 0.7% in 1986. Yet, this was nothing compared to the contraction in 1983 when growth was − 8.7%. Foreign reserves were steady throughout. Debt service showed no obvious trajectory, and trade was growing. In terms of economic indicators, there was nothing outstanding in the run up to 1986. As another Cold War example, consider Zaire. From 1960 to 1981, Zaire received between zero and three new projects each year, with an average of 1.2 per year. In 1982 and 1983 the country served on the UNSC, and the number of new projects 20 In this section, economic data are taken from World Bank (2006a), except growth data which come from Heston et al. (2002).

5

went up to six and four, respectively. The deplorable economic conditions of Zaire are well known, and again the country certainly qualified as needy of World Bank support. But still, there was nothing economically remarkable about 1982 or 1983. The economy contracted almost every year throughout the 1980s. Trade dominated the small economy throughout. Debt service was unremarkable. As for foreign reserves, the World Bank does not even have any reliable data available for Zaire. Yet, for some reason, Zaire received more World Bank projects when it served on the UNSC. Algeria received an average of about one new World Bank project per year from 1962 to 1987, with some years receiving none and other years as much as four new projects. In 1988, however, Algeria served on the UNSC, and the number of new projects went up to their highest number: five. Maybe this was just due to economic circumstances. The economy contracted in 1987 by 0.03%, and then more severely in 1988, by nearly 8%. There were four new projects in 1989, during the second year of the term. Yet, there was no similar bump up for Algeria when it served again after the Cold War had ended. When Algeria served on the UNSC in 2004, not one new project started. We see the pattern exists throughout Asia as well. A large country, Indonesia always has a lot of Bank projects in place. For example, from 1968 to 1972, Indonesia had from two to eight new projects per year with eight new projects in 1970, four in 1971, and seven in 1972. When it served on the UNSC in 1973, it nevertheless received a bump up in the number of new projects: eleven. Following the UNSC term, the number of new projects remained high, averaging about nine new projects per year from 1975 to 1994. However, the record number of new projects in Indonesia is seventeen, which was reached while it was serving the second year of a term on the UNSC in 1996. Bangladesh is a very poor country that typically has multiple World Bank projects in place regardless of international politics. Consider that from independence in 1972 until 1978, the average number of new World Bank projects was almost five per year, ranging from one to seven. Yet, when Bangladesh served on the UNSC in 1979–1980, the number of new World Bank projects jumped up to nine and eleven new projects, respectively. This was unprecedented in Bangladesh and unmatched since — except in 1999 when the country again received eleven new projects. Bangladesh had announced its candidacy and been elected again to the UNSC that year. For all of these countries, the economic need for World Bank assistance is constant. Appropriate candidates for development assistance to be sure, there is no obvious economic reason that they received increased World Bank attention during the years they served on the Security Council. We suspect the reasons were political. The above anecdotal evidence from Latin America, Africa and Asia, while intriguing, is nonetheless circumstantial. The World Bank can defend just about any loan to a single developing country as part of its overarching goal to promote development or reduce poverty. We have no doubt that staff

6

A. Dreher et al. / Journal of Development Economics 88 (2009) 1–18

economists would be able to justify every project mentioned above as worthy of World Bank assistance. Surely every one of them was a worthy project, but certainly there were also many other worthwhile projects that never received Board approval. Governments of developing countries constantly negotiate and work with the World Bank to develop new projects. The question for us is, why is it that more projects get approved for some countries in some years? When we consider one country's historical experience and then compare it to the experience of many other countries, we begin to see a pattern. UNSC participation seems to be associated with more World Bank projects. Yet, one could dismiss our anecdotes. We have seen such skepticism before. Similar research on the lending activities of the International Monetary Fund was covered by a reporter for the Washington Post, who confronted the IMF with some of our anecdotes. A spokesman for the IMF contended, “the evidence is anecdotal and circumstantial” (Washington Post Wednesday, November 1, 2006; Page A19). With such skepticism in mind, we turn to more rigorous analysis of large-n data. 4. Data, Method, and Results 4.1. Data Consider what we observe21 : Our dataset includes a maximum of 5690 country-year observations of 157 countries22 from 1970 to 2004. The panel is unbalanced because countries enter and leave the sample in different years — this includes all independent countries for the time period (according to Cheibub and Gandhi, 2004), minus the observations for the five permanent members of the UNSC and industrial countries that never received Bank loans over the period under study.23 Our dependent variable is the number of World Bank projects starting in a particular year. Our principal explanatory variable of interest is UNSC membership, a dummy variable coded 1 if a country is temporarily serving on the UNSC, and 0 otherwise. Membership begins on January 1 of the year and lasts two years, although there are some rare cases where service lasted only one year. The number of World Bank projects starting in a particular year is taken from the World Bank's projects website.24 World Bank projects were initiated during 2749 of the 5690 country-year observations (48%), while governments served as temporary members of the UN Security Council in only 357 observations (6.3%). Serving on the UN Security Council is thus a relatively rare event, whereas 21 Appendix B summarizes the sources and definitions of the variables we use, while descriptive statistics are reported in Appendix C. 22 Some countries, like the Yemen Arab Republic, no longer exist. 23 Following Cheibub and Gandhi (2004), we do not include Ukraine or Belarus before 1991, as they are not generally considered to have been independent countries. However, they were separate members of the UN and even served as UNSC members in 1948–9, 1984–5 (Ukraine), and 1974–5 (Belarus). 24 http://web.worldbank.org/WBSITE/EXTERNAL/PROJECTS/0,,menuPK: 115635~pagePK:64020917~piPK:64021009~theSitePK:40941,00.html (accessed October 5, 2007).

participation in World Bank projects is quite common throughout the developing world. In our sample, the average number of new Bank projects is 1.34, with a maximum of 18 (China in 1999). Brazil and Indonesia had 17 new projects in 1997 and 1996, respectively. India had 16 in 1977. Note that all of these outlying observations represent years when the countries in question were either elected to the UNSC (Brazil) or were already serving (China, India, Indonesia). One possibility that comes immediately to mind, of course, is the selection problem: if selection onto the UNSC is nonrandom, some outside factor could cause both UNSC membership and participation in World Bank projects. As shown in Dreher and Vreeland (2008), however, election to the UNSC is idiosyncratic. They establish that election to the UNSC is clearly not related to variables that also affect the World Bank's decision to lend. The only factor that seems somewhat to matter is population size in Latin America and Asia, although even this factor does not predict the timing of a country's election to the UNSC, only the frequency. Thus, we rule out the possibility of reverse causation. We do recognize that bigger countries are likely to have more World Bank projects, however. In our empirical analysis below, we capture this by the inclusion of dummies for each country. In addition, we control for population size. As Fig. 1 illustrates, governments not serving on the UN Security Council receive fewer World Bank projects.25 On average, non-members receive about 1.3 new projects per year, while temporary members of the UNSC get 2.1. Restricting our attention to only those observations of countries that actually serve on the UNSC, we see that the average number of new World Bank projects is high during both years a country serves on the Council. Comparing the Cold War to the post-Cold War period reveals that the difference between members and nonmembers prevails for both periods. What about patterns within the regional caucuses that nominate candidates for UNSC membership? Fig. 2 shows that the pattern holds at varying levels of strength for every developing region in the world — Africa, Asia and the South Pacific, Latin America and the Caribbean. The pattern is particularly strong for South Asia. Part of this is driven by the strategic importance of Pakistan, which has served six terms on the UNSC. When not serving, the average number of World Bank projects initiated annually is 3.02; when serving in the UNSC, however, the average number of World Bank projects initiated annually is 4.58. We suspect that the geopolitical significance of Pakistan gives this country a particularly important voice when serving on the UNSC. Apparently the voice is worth about 1.56 additional projects per year. The basic pattern holds, but appears weak in Africa south of the Sahara.26 In fact, however, the regional differences turn out not to be statistically significant when subjected to rigorous analysis (as we show below). 25 The horizontal line in Fig. 1 and the following figure shows the average number of new projects across our entire sample. 26 Interestingly, we have found in other work that the descriptive pattern of UNSC membership and IMF lending is quite pronounced for Africa (Dreher et al., 2006).

A. Dreher et al. / Journal of Development Economics 88 (2009) 1–18

7

Fig. 1. Number of new World Bank projects by non-permanent UN Security Council membership. Note: The horizontal line shows the average number of new projects across the entire sample.

4.2. Method Does this pattern hold when put to more rigorous tests? To explore this, we analyze various statistical models with the number of new World Bank projects as our dependent variable and UN Security Council membership as principal explanatory variable of interest. Following a bivariate analysis of the relationship between UNSC membership and the number of new World Bank projects, we introduce different control variables to ensure that our results are not spurious and present for each specification four statistical models — (1) Poisson with fixed country and year effects, (2) Poisson with fixed country effects, (3) Negative Binomial with fixed country and year effects, and (4) Negative Binomial with fixed country effects.27 As the data on new Bank projects are strongly skewed to the right, estimation with OLS would be inappropriate. Furthermore, significant overdispersion leads us to prefer the Negative Binomial estimator (with fixed effects).28 In choosing our covariates, we mainly follow Dreher et al. (2006), including the most robust predictors of IMF participation following the Extreme Bounds Analysis (EBA) of Sturm et al. (2005) in their analysis of UNSC membership on IMF program participation. Arguably, the determinants of new World Bank projects may be related to those of IMF programs. We thus include debt service, measured as public and publicly 27 We also experimented with models using random instead of fixed effects. None of the conclusions is affected by this. 28 Overall, likelihood ratio tests comparing the Negative Binomial results with those of the Poisson model, as depicted in Tables 1 and 2 below, confirm this.

guaranteed debt service as a percentage of gross national income, taken from the World Development Indicators (2006). Investment, measured as private and public gross national investment as a share of gross domestic product (GDP) comes from Penn World Tables 6.1 (Heston et al., 2002). We include lagged elections, which is a dummy variable coded 1 if elections were held the previous year and 0 otherwise. It has been argued that World Bank involvement is more likely after national elections, as Bank conditionality is used to fight the effects of expansionary policies in the run up to the election (Dreher and Vaubel, 2004). We do not follow Dreher et al. (2006) in using the level of international reserves which is important for IMF programs but not likely to be related to new World Bank projects.29 However, we include (log) per capita GDP. This is because development is an important element in official World Bank policies.30 As mentioned above, we also include (log) population to the regressions. Finally, we employ a dummy variable that is one, when a country is under an IMF program. For some World Bank (structural adjustment) projects, IMF programs are a pre-condition. Not controlling for IMF participation might thus simply capture the effect of UNSC membership on the IMF. The next section presents our results. 29 When included to the regression, the coefficient is completely insignificant, while the main results are unchanged. The robustness analysis reported below includes this variable. For a study of why foreign reserves may be a problematic variable, see Trudel (2005). 30 In addition, per capita GDP proves to be a significant predictor of IMF programs, as the robustness section in Dreher et al. (2006) shows.

8

A. Dreher et al. / Journal of Development Economics 88 (2009) 1–18

Fig. 2. Number of new World Bank projects by UN Security Council membership across regions.

4.3. Results Table 1 presents the results of the bivariate analysis under various statistical models. In all cases, the effect of UN Security Council membership on new World Bank projects is positive and significant at the 1% level. Columns 1 and 5 do not include additional control variables and therefore represent the situation as depicted in Fig. 1. These estimates suggest that countries which are temporary members of the UNSC do on average receive 0.8 World Bank projects more per year than non-member countries. The estimated coefficients in columns 1 and 5 imply a 73% increase, i.e., the associated incidence-rate ratio equals (exp(0.55)=) 1.73. Fig. 2 highlighted that regional differences may be substantial. For that reason columns 2 and 6 include, besides fixed effects for each year, dummies for the following regions: South Asia, East Asia, Latin America/Caribbean, Middle East/ North Africa, Sub-Saharan Africa, Eastern Europe, and Western Europe. As with any complete set of dummy variables, one of them must be left out of the analysis. In Table 1, we omit Western Europe. So the significantly positive coefficients for the other regions indicate that these regions are more likely to get World Bank projects than Western European countries — not controlling for other determinants of participation like income and investment (as already suggested by Fig. 2). Again, the significantly positive relationship between UN Security Council membership and participation in new World Bank projects persists and remains high. We further explore regional influences in Section 5 below (specifically, see Table 4).

In a next step, we replace the regional-fixed effects by conditional country-fixed effects. Although this has a sizeable impact on our parameter of interest, it remains significant at the 1% level. Once country-specific elements which might drive both the number of new World Bank projects and UNSC membership are included (Columns 3 and 7), the incidencerate ratios drop to 1.14. Hence, UNSC membership increases the number of new World Bank projects on average by approximately 14%. Finally, this result is not affected by removing the year dummies from the specification. Note, though, that likelihood-ratio tests (as shown at the bottom of the table) indicate that the results presented in Column 7 are preferred.31 To summarize Table 1, the statistically significant relationship between UN Security Council membership and new World Bank projects holds when we introduce regional and/or conditional country- and year-fixed effects and is independent of whether we use Poisson or Negative Binomial regression methods. Table 2 introduces the control variables discussed above to the fixed effects models. Subsequently, we test whether the fixed time effects can be removed from the specification. In both the Poisson and the Negative Binomial models likelihoodratio tests show that the null hypothesis that the two models are drawn from the same distribution cannot be rejected. Hence, once control variables are added, it is no longer necessary to

31 Specifically, restrictions imposed on the model by setting country and year fixed effects to zero are rejected at the 1% level of significance. Also at the 1% level, restrictions imposed by the Poisson model are rejected.

A. Dreher et al. / Journal of Development Economics 88 (2009) 1–18

9

Table 1 The impact of UN Security Council membership on new World Bank projects under different statistical models Poisson

Temporary UNSC member South Asia dummy East Asia dummy Latin America/Caribbean dummy Middle East/North Africa dummy Sub-Saharan Africa dummy Eastern Europe dummy

Negative Binomial

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

Pooled

Regional effects

Fixed country and year effects

Only country effects

Pooled

Regional effects

Fixed country and year effects

Only country effects

0.55 0.55 (14.43)⁎⁎⁎ (14.00)⁎⁎⁎ 1.83 (23.93)⁎⁎⁎ 1.27 (16.59)⁎⁎⁎ 1.23 (16.95)⁎⁎⁎ 1.15 (14.47)⁎⁎⁎ 1.14 (15.96)⁎⁎⁎ 1.22 (15.83)⁎⁎⁎

Country effects Country effects — p-value Year effects Year effects — p-value Log likelihood −9606.23 LR-test w.r.t. Poisson model — Chi2 LR-test w.r.t. Poisson model — p-value

0.13 (3.21)⁎⁎⁎

Fixed 0.00 Fixed Fixed 0.00 0.00 − 9155.17 − 6320.69

0.15 (3.67)⁎⁎⁎

0.55 (7.19)⁎⁎⁎

Fixed 0.00

− 6403.51

−8176.67 2859.12 0.00

0.48 (6.42)⁎⁎⁎ 1.78 (16.35)⁎⁎⁎ 1.23 (12.17)⁎⁎⁎ 1.21 (12.90)⁎⁎⁎ 1.13 (10.63)⁎⁎⁎ 1.12 (12.39)⁎⁎⁎ 1.20 (11.65)⁎⁎⁎

0.13 (2.68)⁎⁎⁎

Fixed 0.00 Fixed Fixed 0.00 0.00 −8012.35 − 6271.73 2285.64 97.91 0.00 0.00

0.15 (2.95)⁎⁎⁎

Fixed 0.00

− 6325.57 155.88 0.00

Absolute value of z statistics in parentheses. ⁎Significant at 10%; ⁎⁎significant at 5%; ⁎⁎⁎significant at 1%. All regressions contain 4711 observations covering 149 countries and 38 years and include a constant term.

also include time fixed effects. We furthermore report likelihood-ratio tests comparing the Poisson to the Negative Binomial models. As in Table 1, we reject the restrictions imposed by the Poisson model. According to these results, countries currently under an IMF program also have roughly (exp(0.25)− 1=) 28% more new World Bank projects (evaluated at the mean of the explanatory variables and assuming the fixed effects to be zero). Debt service has a strongly positive significant effect in all of our models. More debt leads to more need for World Bank projects. The effect is not surprising.32 Investment, however, has a surprisingly robust positive and significant effect on the number of new World Bank projects. Perhaps the Bank acts like a bank when it comes to investment levels – it prefers country episodes with higher levels. As expected, the number of new World Bank projects decreases with GDP per capita – countries that are poorer turn to the Bank more frequently. Population has a strangely negative effect when country- and year-fixed effects are included in the Poisson model (Column 1). Note that country size, as measured by population, is already captured by the country-specific effects. In case we estimate without country-specific effects, the population coefficient has the expected positive sign and is highly significant. When we 32 Marchesi (2003) tests whether countries having arrangements with the IMF are more likely to obtain a rescheduling of their external debt than others. She concludes that the adoption of an IMF program works as signal of a country's “good intent” which is rewarded with debt relief. Her results confirm the existence of a significant effect of the adoption of an IMF program on the subsequent concession of a debt rescheduling by private creditors.

estimate conditional on just country-specific effects, the significant result also disappears. Lagged elections do significantly reduce the number of new World Bank projects according to all specifications. Potentially, incoming governments need time to establish relations with the Bank. Turning to our principal independent variable of interest, UNSC membership has a positive effect on participation in World Bank projects in all of our models. The finding is significant at the 10% level in all models.33 Note that this is true even with the loss of observations due to missing data in our control variables; the sample size is reduced to 2135 observations from the original 4711 observations in Table 1. Even after introducing control variables, the impact of UNSC membership remains substantial. Evaluated at the sample means of the covariates, the number of new World Bank projects starting in a particular year increases by 10% when a country participates on the UNSC. According to column 4 of Table 2, a 1% increase in debt service increases the number of new World Bank projects by 10%. The corresponding (absolute) elasticities for investment, GDP per capita, and population are 26%, 58% and 4%, respectively. Lagged elections reduce the number of Bank projects by 8%. To summarize this section, using a battery of different statistical models and control variables used in the previous 33 We also separately analyzed the number of new IDA and IBRD projects. UNSC membership is marginally insignificant in both specifications, with a positive coefficient. The coefficient of the UNSC dummy is not significantly different in the two specifications.

10

A. Dreher et al. / Journal of Development Economics 88 (2009) 1–18

Table 2 The impact of UN Security Council membership on new World Bank projects under different statistical models, extended model Poisson

Negative Binomial

(1)

(2)

Fixed effects Temporary UNSC member IMF program Debt service (in percent of GDP) Investment (in percent of GDP) GDP per capita (log) Population (log) Lagged election Country effects Country effects — p-value Year effects Year effects — p-value Log likelihood LR-test w.r.t. Poisson model — Chi2 LR-test w.r.t. Poisson model — p-value

0.09 (1.74)* 0.25 (6.66)*** 0.01 (3.73)*** 0.02 (4.18)*** −0.59 (5.05)*** −0.75 (2.48)** −0.09 (2.31)** Fixed 0.00 Fixed 0.20 −3086.85

Only country effects 0.10 (1.95)* 0.24 (6.64)*** 0.01 (3.81)*** 0.02 (4.56)*** − 0.46 (4.40)*** − 0.11 (1.03) − 0.08 (2.05)** Fixed 0.00

−3105.04

(3)

(4)

Fixed effects

Only country effects

0.09 (1.69)* 0.25 (6.30)*** 0.01 (3.70)*** 0.02 (4.25)*** −0.62 (5.46)*** −0.10 (0.53) −0.09 (2.31)** Fixed 0.00 Fixed 0.58 −3083.31 7.08 0.01

0.09 (1.74)* 0.25 (6.31)*** 0.01 (3.67)*** 0.02 (4.83)*** − 0.58 (5.50)*** − 0.04 (0.39) − 0.08 (2.07)** Fixed 0.00

− 3097.25 15.57 0.00

Absolute value of z statistics in parentheses. *Significant at 10%; **significant at 5%; ***significant at 1%. All regressions contain 2135 observations covering 113 countries and 31 years.

literature, we find that UNSC membership robustly increases the number of new World Bank projects — with an estimated increase of about 10%. We take this as evidence in favor of our argument that the major shareholders systematically use the World Bank to pursue international political goals. 5. Extensions and tests for robustness In this section, we present several extensions to our basic analysis. First, we investigate the timing of UNSC membership influence on World Bank projects in greater detail. Recall Fig. 1, where we display the development of new Bank projects during the two years of UNSC membership. The number of new projects appears to slightly increase in the second year. How significant is this increase? Following Kuziemko and Werker (2006), we employ a series of dummy variables to examine the trend over time. Importantly, this also allows us to test an alternative hypothesis. Perhaps election to the UNSC simply raises the visibility of a country on the international stage and helps it capture the attention of an international organization like the World Bank. If this is true, World Bank projects should not drop off following the end of a term. Once a country has the attention of an international organization like the World Bank, there is no reason to think it should then lose the attention, unless the reason for the attention is directly related to the importance of the country's behavior while on the UNSC. Arguably, an increase in the global orientation of a developing country's government may prompt it both to run for

election to the UNSC and become more confident in dealings with international organizations like the World Bank.34 This is certainly a plausible explanation, and one can think of cases that fit.35 While such a global orientation is unlikely to disappear over just a two-year stint on the UNSC, we extend the regression and include an index of political globalization. The index is provided by the KOF Swiss Economic Institute for 122 countries on a yearly basis and is frequently used in the recent empirical literature.36 Basically, it combines membership in international organizations, participation in UN peace keeping missions, and the number of embassies in a country. As an additional test, we split the sample according to the Cold and post-Cold War era. Fig. 1 suggests that the membership effect on new World Bank projects is similar during and after the end of the Cold War. Does this hold once controlled for other relevant determinants of World Bank projects? Vote buying may have been more important during the Cold War. However, there has been some discussion about whether the end of the Cold War actually introduced a structural shift in countries' positions in the UN, as measured by General Assembly voting (see, e.g., Voeten, 2000). Arguably, the determinants of voting in the UN need not be constant over the 30 years under study. Yet, with the end of the Cold War,

34

We are grateful to an anonymous reviewer for raising this point. Nyerere's Tanzania comes to mind. 36 For a detailed description and a long list of articles using the index see http://www.kof.ethz.ch/globalization. 35

A. Dreher et al. / Journal of Development Economics 88 (2009) 1–18 Table 3 The impact of UN Security Council membership on new World Bank projects over time Negative Binomial (1)

(2)

(3)

(4)

Temporary UNSC member two years before Temporary UNSC member year before Temporary UNSC member year 1 Temporary UNSC member year 2 Temporary UNSC member year after Temporary UNSC member two years after IMF program

0.03 (0.43) 0.06 0.06 (0.75) (0.72) 0.02 0.02 0.02 (0.27) (0.23) (0.30) 0.16 0.16 0.16 0.16 (2.15)** (2.14)** (2.24)** (2.22)** −0.13 − 0.13 (1.51) (1.56) 0.00 (0.04) 0.25 0.25 0.26 0.26 (6.29)*** (6.31)*** (6.34)*** (6.36)*** Debt service (in percent 0.01 0.01 0.01 0.01 of GDP) (3.75)*** (3.74)*** (3.67)*** (3.69)*** Investment (in percent 0.02 0.02 0.02 0.02 of GDP) (4.22)*** (4.22)*** (4.27)*** (4.27)*** GDP per capita (log) −0.62 − 0.61 − 0.62 −0.62 (5.44)*** (5.43)*** (5.46)*** (5.46)*** Population (log) −0.09 − 0.09 − 0.10 −0.10 (0.47) (0.46) (0.55) (0.56) Lagged election −0.10 − 0.10 − 0.10 −0.10 (2.45)** (2.44)** (2.45)** (2.48)** Log likelihood − 3080.59 − 3080.69 − 3082.32 −3082.36 Equal UNSC coeff. — Chi2 7.78 7.54 1.98 Equal UNSC coeff. — p-value 0.17 0.06 0.16 Absolute value of z statistics in parentheses. *Significant at 10%; **significant at 5%; ***significant at 1%. All regressions include conditional country- and year-fixed effects. All regressions contain 2135 observations covering 113 countries and 31 years.

countries are less constrained by alignments and might thus be more likely to vote according to their preferences when not being bribed. Economically weak countries no longer need protection by “their” bloc, and now need to be bribed to achieve alignment. Vote buying might thus prevail in the post-Cold War era, as our Gulf War anecdotes above indicate. Fig. 2 suggests substantial differences in the impact of UNSC membership on new World Bank projects across geographic regions, however, these differences might be due to differences in other important factors like income or population. We therefore test whether these regional differences hold when we control for such determinants. Specifically, we include interactions between the UNSC membership dummy and the regional dummies.37 We also investigate whether UNSC membership influences the size of World Bank loans. Different World Bank projects are of substantially different economic magnitude, and the amount of money associated with these projects may matter more than the number of projects approved. It has been shown, for example, that countries voting with the US in the UN General Assembly receive larger IMF loans (Oatley and Yackee, 2004).

37

We are grateful to an anonymous reviewer for raising this point.

11

When the World Bank's major shareholders use the Bank to bribe or reward temporary UNSC members, we expect them to receive larger loans, all else equal. Yet, the results in Dreher et al. (2006) show that – while UNSC membership does significantly affect the probability of participating in an IMF program – there is no statistically significant impact on the size of the IMF loans. We test whether the same holds for the World Bank. As our final extension, we examine the robustness of the relationship between UNSC membership and participation in World Bank projects with Extreme Bounds Analysis. Almost 300 specifications with different combinations of control variables are analyzed. The EBA approach is described in detail in Appendix A. Following Sala-i-Martin (1997), we consider the impact of UNSC membership on IMF programs to be robust if the fraction of the cumulative distribution function lying on one side of zero (CDF(0)) exceeds 0.90. The variables our EBA includes in all specifications are: Participation in IMF programs, debt service, investment, GDP per capita, population, and lagged election. Each regression also includes up to three variables out of: GDP growth, change in international reserves, level of international reserves, economic globalization, social globalization, political globalization, regime type, voting inline with the average of France, the UK, and the US in the UN General Assembly,38 government budget balance (in percent of GDP), rate of inflation, current account balance (in percent of GDP), and (the natural logarithm of) the number of checks and balances in the political system. These variables have all been proposed in the previous literature as potential determinants of World Bank involvement, and are described in more detail in Appendix B. The statistical model we employ from here onwards is the most rigorous one — Negative Binomial regression with fixed country and year effects.39 We use the same set of control variables included in Table 2 above. Table 3 presents the results of the first extension, regarding the effect of the timing of UNSC membership on World Bank projects. We include a series of dummy variables for two years and one year before UNSC membership, the first year and second year of membership, and one year and two years after membership ends. Subsequently, we remove the insignificant dummies in three steps. As indicated by the statistically significant coefficients for the UNSC Year 2 dummy shown in all columns, the impact of UNSC membership is by far the strongest during the second year of the two-year term. During the second year of UNSC membership, a country received on average about 17% more World Bank projects. This is contrary to the results in Dreher et al. (2006), showing that the effect of UNSC membership on IMF program participation is strongest

38 These countries have permanent seats in the UNSC and appoint an Executive Board member at the World Bank. Replicating the analysis using the G7 instead shows similar results. The same is true for the IMF's five biggest shareholders and the five permanent UNSC members. Following Dreher and Sturm (2006), we weigh the voting behavior of each country with its share in World Bank votes, to reflect their power in the institution. 39 Note that the year fixed effects are jointly insignificant in most – but not all – models. Omitting them from the specifications does not affect the results.

12

A. Dreher et al. / Journal of Development Economics 88 (2009) 1–18

Table 4 Extensions, Negative Binomial regression (1) During Cold War Temporary UNSC member

0.14 (1.97)**

(2) Δ after Cold War

0.21 (4.16)*** Debt service (in percent of GDP) 0.004 (1.77)* Investment (in percent of GDP) 0.02 (4.76)*** GDP per capita (log) − 0.45 (3.43)*** Population (log) − 0.16 (0.71) Lagged election − 0.03 (0.58) South Asia * UNSC

Δ after Cold War

−0.14 (1.23)

Temporary UNSC member year 2 IMF program

During Cold War

0.10 (1.33) 0.004 (1.76)* −0.02 (2.43)** 0.08 (2.00)** −0.03 (1.21) −0.12 (1.51)

(3)

(4)

Regional differences

Political globalization

− 0.18 (0.58) 0.20 (2.16)** 0.22 (4.21)*** 0.004 (1.79)* 0.02 (4.79)*** − 0.44 (3.38)*** − 0.18 (0.77) − 0.04 (0.73)

East Asia * UNSC Latin America/Caribbean * UNSC Middle East/North Africa * UNSC Sub-Saharan Africa * UNSC

− 0.13 (0.85) 0.10 (1.33) 0.004 (1.69)* − 0.02 (2.46)** 0.08 (1.96)** − 0.03 (1.30) − 0.12 (1.44)

0.25 (6.28)*** 0.01 (3.66)*** 0.02 (4.15)*** − 0.61 (5.32)*** − 0.10 (0.54) − 0.10 (2.44)** 0.45 (1.39) 0.22 (0.63) 0.18 (0.56) 0.54 (1.62) 0.12 (0.36)

Political globalization, index Number of countries 113 Number of observations 2135 Log likelihood − 3073.56 No Cold War effect — Chi2 19.53 No Cold War effect — p-value 0.00

113 2135 − 3073.04 18.69 0.00

113 2135 − 3532.84

0.10 (1.77)*

0.23 (5.47)*** 0.01 (4.07)*** 0.02 (4.46)*** − 0.60 (5.02)*** 0.005 (0.03) − 0.09 (2.05)**

− 0.0083 (3.75)*** 81 1736 − 2671.30

Absolute value of z statistics in parentheses. *Significant at 10%; **significant at 5%; ***significant at 1%. All regressions include conditional country- and year-fixed effects.

in the first year of membership. Most likely, the difference is due to the typically longer preparation phase of World Bank projects. While IMF loans are usually negotiated relatively quickly, the World Bank explains that their project preparation phase can range from a few months to three years, depending on the complexity of the project proposed.40 Also, for some World Bank (structural adjustment) programs, an IMF program is required to be in place. Table 4 contains the results separated along the Cold War/ post-Cold War dimension, the regional differences, and the impact of political globalization. Whereas the left parts of columns 1 and 2 report the coefficient of the variable over the full sample, their right parts show the interaction of the respective variable with a dummy for the post-Cold War period. In other words, it documents the change in the respective variable's influence on new World Bank projects after the end

40 http://web.worldbank.org/WBSITE/EXTERNAL/PROJECTS/0,,contentMDK: 20120731~menuPK:41390~pagePK:41367~piPK:51533~theSitePK:40941,00. html (accessed June 15, 2007).

of the Cold War. As can be seen, there is no significant change of the impact of UNSC membership on participation in World Bank projects after the end of the Cold War period. Note that the coefficient of the UNSC membership dummy after the Cold War is only marginally insignificant and quantitatively matches the dummy for the whole period. To some extent, this might be taken in favor of the World Bank's claim that politics became less important after the end of the Cold War. Focusing on the second year of membership, however, column 2 shows a completely insignificant (and quantitatively smaller) coefficient on the interaction of the UNSC dummy with the post-Cold War dummy. There is thus some evidence that the Bank is still being used as political tool of their major shareholders. Column 3 examines potential regional differences. Recall from Fig. 2 that some regions appear to have stronger effects of UNSC membership than others. Of course, these could be driven by other factors we control for, like income or population, which are also correlated with region. The regional differences also might disappear when we control for countryspecific effects. This turns out to be the case. We do not see significant differences in the effects of UNSC membership by

A. Dreher et al. / Journal of Development Economics 88 (2009) 1–18

13

Table 5 The impact of UN Security Council membership on World Bank disbursements and commitments (1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

Disbursements Disbursements Log disbursements Disbursements Commitments Commitments Log commitments Commitments (as % of GDP) (as % of GDP) (as % of GDP) (per capita) (as % of GDP) (as % of GDP) (as % of GDP) (per capita) OLS Temporary −0.0011 UNSC member (1.46) IMF program 0.0020 (5.83)*** Debt service (in 0.0002 percent of GDP) (11.01)*** Investment (in 0.0001 percent of GDP) (2.03)** GDP per capita −0.0085 (log) (5.99)*** Population (log) −0.0002 (0.06) Lagged election −0.0003 (0.54)

Tobit − 0.0011 (1.42) 0.0020 (6.15)*** 0.0003 (12.91)*** 0.0002 (3.53)*** − 0.0084 (12.88)*** − 0.0023 (6.51)*** −0.0003 (0.56)

Tobit − 0.0011 (1.42) 0.0020 (6.30)*** 0.0002 (12.99)*** 0.0002 (3.54)*** − 0.0082 (13.12)*** − 0.0022 (6.53)*** − 0.0003 (0.56)

Tobit

OLS

− 0.0073 (1.57) 0.0130 (6.60)*** 0.0015 (12.97)*** 0.0009 (3.12)*** − 0.0428 (11.40)*** − 0.0141 (6.88)*** − 0.0016 (0.53)

− 0.0006 (0.38) 0.0048 (6.61)*** 0.0002 (5.31)*** 0.0002 (1.80)* − 0.0105 (3.44)*** − 0.0117 (1.55) − 0.0012 (1.16)

Tobit −0.0006 (0.34) 0.0049 (7.05)*** 0.0003 (8.12)*** 0.0004 (4.06)*** −0.0109 (12.63)*** −0.0028 (6.51)*** −0.0012 (1.15)

Tobit − 0.0005 (0.32) 0.0047 (7.12)*** 0.0003 (8.25)*** 0.0003 (3.98)*** − 0.0105 (12.94)*** − 0.0027 (6.52)*** − 0.0011 (1.13)

Tobit −0.0032 (0.33) 0.0299 (7.39)*** 0.0019 (8.34)*** 0.0019 (3.76)*** −0.0556 (11.38)*** −0.0173 (7.02)*** − 0.0057 (0.92)

Absolute value of t statistics in parentheses. *Significant at 10%; **significant at 5%; ***significant at 1%. The OLS models include conditional country- and year-fixed effects. All regressions contain 2121 observations covering 113 countries and 31 years. The Tobit models are estimated with country random effects and regional- and year-fixed effects.

region once we account for these other factors. The regional differences apparent in Fig. 2 disappear when put to more rigorous analysis. Indeed, due to problems of multicollinearity and the inclusion of irrelevant variables, the standard errors on most of our variables of interest increase so much that we find little significance at all. Column 4 adds the index of political globalization. Surprisingly, the results show that political globalization reduces the number of new World Bank projects rather than increasing them.41 It seems that politically more globalized countries have other sources of aid, so they are less needy of the Bank. Most importantly, however, the impact of UNSC membership on new World Bank projects remains significant at the 10% level. In Table 5 we present results for disbursed and committed World Bank loans, respectively. Columns 1 and 5 show the results when we use Ordinary Least Squares corrected for both country and time fixed effects. Columns 2 and 6 report results using a Tobit model including random country effects and fixed regional and year effects. In columns 3 and 7 we present Tobit specifications in logs as in Neumayer (2003) and Andersen et al. (2006). Columns 4 and 8 report results for aid in per capita terms (suggested in Frey and Schneider, 1986). As can be seen, neither disbursements nor commitments are significantly affected by temporary UNSC membership. This is in line with

41 Arguably, there might be some correlation between the measure of political globalization and better economic fundamentals, so that the result would reflect economic fundamentals rather than globalization. We therefore replicated the analysis focusing on low income countries only (according to the World Bank (2006a), low income countries are those with 2004 GNI per capita less than US $825). The results with respect to globalization (and temporary UNSC membership) remain.

the results for the IMF reported in Dreher et al. (2006). Why might loan availability be affected even though disbursements are not? The result is actually not surprising. Today's disbursements to a large extent depend on projects arranged years before. The increased amount associated with the larger number of new projects seems to be dominated by these amounts. However, commitments directly refer to new projects. Does the insignificant coefficient on commitments together with the increase in the number of new projects imply that the average loan is smaller? We tested for this possibility and did not get a statistically significant result. So, it seems that the additional amount of money is not large enough to be detected by our statistical model, while the number of additional projects is. Our model has been chosen to explain the determinants of projects, not the amount of loans. Potentially, the determinants of projects do not fully correspond to those of loans, so our loan equation might be misspecified.42 We leave this for future research. Finally, results of the Extreme Bounds Analysis are summarized in Table 6. Again, we report results for our preferred specification which includes both country- and year-fixed effects. We also report the results including only conditional fixed country effects, or when replacing the UNSC dummy by the dummy for the second year of membership. According to the results, the impact of UNSC membership on the number of new World Bank projects is indeed robust to the choice of control variables. The coefficient is significant in

42 Clearly, the loan equation would have to account for being selected as recipient of World Bank projects in the first place. Estimating a Heckman selection model, however, requires demanding exclusion restrictions on the allocation equation.

14

A. Dreher et al. / Journal of Development Economics 88 (2009) 1–18

Table 6 The impact of UN Security Council membership on new World Bank projects, Extreme Bounds Analysis

Average beta coefficient Average robust std. error % Sign. at 10% level CDF(0) Lower bound Upper bound # of combinations Average # of observations

Negative Binomial fixed effects

Negative Binomial only country effects

Negative Binomial fixed effects, 2nd UNSC year only

0.105 0.057 71.1% 0.96 − 0.028 0.277 298 1680

0.106 0.057 72.5% 0.96 − 0.030 0.279 298 1680

0.172 0.075 99.0% 0.99 − 0.004 0.400 297 1681

Variables included in all specifications are: IMF program, debt service, investment, GDP per capita (log), population (log), and lagged election. The vector of variables from which up to three are included in each of the 298 possible combinations consists of: GDP growth, change in international reserves, level of international reserves, economic globalization, social globalization, political globalization, regime type, voting inline with the average of France, the UK, and the US in the UN General Assembly, government budget balance (in percent of GDP), rate of inflation, current account deficit (in percent of GDP), (log) checks and balances.

more than two thirds of the specifications. With a CDF(0) of 0.96 the threshold of 0.90 suggested by Sala-i-Martin (1997) is easily passed. Focusing on the second year of membership only, the results again become stronger; the CDF(0) rises to 0.99, with a coefficient significant at least at the 10% level in 99% of the regressions. On average, a country receives close to 19% more new World Bank projects during the second year than otherwise would be the case. We, thus, conclude the following from our analysis: elected members of the UNSC are more likely than other countries to receive projects from the World Bank. The projects are likely to start during the second year of the two-year UNSC term. The actual loan size is not affected by UNSC membership. 6. Conclusions Our results contribute to the growing literature showing that International Financial Institutions have been employed as a tool of foreign policy by their major shareholders. Whether used to bribe or reward, the World Bank's projects have been funneled to politically important developing countries, such as those serving a term on the UN Security Council. Given the nature of our large-n study, we do not know who took the initiative – whether the temporary UNSC members increased their requests for projects, having increased confidence the World Bank would accept – or, alternatively, whether the Bank's major shareholders took initiative. In any case, the projects of the World Bank are one mechanism by which the major stakeholders of the Bank – mainly the US, but also Japan, Germany, France and the United Kingdom – can win the favor of voting members of the UN Security Council. This is not what the institution was intended to be used for. Originally, the World Bank was set up to promote development. Many blame the failure of World Bank projects, which involve both loans and policy conditions designed to promote development and reduce poverty, on the design of programs and on the failure of countries to adopt World Bank reforms.43 Additionally, previous research suggests the pernicious effects of politically motivated foreign aid. The World Bank often subsidizes inefficient policies and corrupt governments, and long-run 43

Governments may have their own agenda. See Vreeland (2003).

economic prosperity suffers (e.g., Easterly, 2005). Thus, part of the failure of the World Bank may be that its loans have not been exclusively used to promote development in the first place. Nevertheless, we hesitate to jump to the conclusion that the two institutions should be reformed. Regarding the UNSC, perhaps we can view temporary membership on the UNSC as a mechanism of international redistribution. Here is one way of viewing the institution in light of our analysis: almost all countries get their turn to participate in the important deliberations of the UN Security Council, and, when they do, they receive perks in the form of World Bank loans so long as they play ball with the truly powerful countries in the world and do not rock the boat. Their role in most deliberations may be practically inconsequential, but they rise to prominence on the international stage and are duly rewarded for their service. As for reforming the World Bank, we should be cautious. It may be that the very prospect of manipulating the World Bank for political goals is a necessary condition for its major shareholders to lend their financial support to the international institution. If the World Bank can use some of its resources to promote development – perhaps more often in politically important countries – this may be a fine arrangement.44 Moreover, if the international political goals of the countries controlling the Bank are laudable (e.g., to pass Security Council decisions promoting human rights), it is arguable that the overall costs of the votetrading are not substantial.45 So, perhaps the political manipulation of international institutions is a necessary evil to engage the participation of powerful countries in international cooperation. Even if nonpolitically motivated aid might be preferable to politically motivated aid, this may not be a realistic alternative. The alternative to a world with politically manipulated international institutions may simply be a world without international institutions. One must therefore weigh the costs of political manipulation against the benefits of having various institutions that facilitate multilateral deliberations. 44 Other studies confirm that economic need also certainly helps guide World Bank lending decisions. See, e.g., Frey and Schneider (1986), Neumayer (2003), Dollar and Levin (2004). 45 We are grateful to an anonymous reviewer for this suggestion. See also Eldar (in press).

A. Dreher et al. / Journal of Development Economics 88 (2009) 1–18

With the above caveats in mind, our suggestions for reform are conservative. We do suspect there is room for improvement, to the extent that long-run economic gains matter more than short-run political gains. Perhaps much like central banks are isolated from the vagaries of day to day politics, the World Bank can be reformed to provide Executive Directors with a little more independence from political pressures and still provide the major shareholders with enough voice to maintain their support of the institution. With this in mind, consider the governance of the World Bank. Some reformers call for a redistribution of vote shares so that recipient countries have a greater voice.46 Adjusting vote shares, however, would not change the fact that the World Bank can still be used to achieve foreign policy objectives; it would just change the players who get to use the World Bank to pursue foreign policy objectives. Others argue that the World Bank should shift from reliance on the IBRD to the IDA (e.g. the Meltzer Commission, 2000; Rogoff, 2004; Lerrick, 2007). This would focus the World Bank's attention on low-income countries instead of emerging market countries. It would also change where it gets its resources from.47 As mentioned above, the IBRD raises most of its funds on international financial markets, while the IDA relies much more heavily on the resources of major shareholders. Increased reliance on shareholder resources could make the Bank even more beholden to their short-term political goals, unless other reforms are also taken. Thus we suggest the World Bank's main governing body – the Executive Board – might be more effective if it had some increased degree of insulation from short-term political goals. The Directors who sit on the World Bank Executive Board could be appointed for long, non-renewable terms. Certainly, they would and should still represent the interests of their specific countries, but they might resist pressure to pursue some projects that are short-sighted and motivated purely by politics. Acknowledgements We thank David Bearce, Terrence Chapman, Pamela Chasek, José Cheibub, Donald Daniel, Han Dorussen, Michael Doyle, Ofer Eldar, Songying Fang, Christina Fattore, John Freeman, Martin Gassebner, Joanne Gowa, Birger Heldt, Nathan Jensen, Christopher Kilby, Ilyana Kuziemko, Thomas Markussen, Katja Michaelowa, Bessma Momani, Pablo Pinto, Bruce Russett, Holger Schmidt, Friedrich Schneider, Randall Stone, Shawn Treier, Burcu Uçaray, Erik Voeten, and seminar participants at Bond University, the Free University of Brussels, Columbia University Political Science Department, Duke University School of Public Policy, the Penn State Political Science Department, the UMass Amherst Political Science Department, the Georgetown University School of Foreign Service, the University of Pittsburgh Political Science Department, the Princeton University International Relations Colloquium, the University of Illinois 46 See Bird and Rowlands (2006) for a recent discussion. Also see Buira (2005) and Woods (2005). 47 We are grateful to an anonymous reviewer for this suggestion.

15

Political Science Department, the Vassar College seminar, the UCLA Political Science Department, the Manhattan College International Studies Program, the Minnesota University Political Economy Colloquium, the Folke Bernadotte Academy and Georgetown University Working-Group on Peacekeeping Operations, RWI Essen, ETH Zurich, the University of Fribourg, Lund University, Essex University, WIFO Vienna, the Silvaplana Workshop on Political Economy 2007, the Annual Conference of Verein für Socialpolitik: Development Economics (Göttingen 2007), the Canadian Economic Association (Halifax 2007), the Royal Economic Society Annual Conference (Warwick 2007), the International Society for New Institutional Economics (Reykjavik 2007), the European Public Choice Society Meeting (Amsterdam 2007), the International Studies Association (Chicago 2007), the Christmas Meeting of the German-speaking Economists Abroad (Mannheim 2007), the European Economic Association (Budapest 2007), the NPSA (Boston 2006) and two anonymous referees for helpful comments on earlier drafts. Appendix A: Extreme Bounds Analysis (EBA) To examine both the sensitivity of our baseline model and the coefficients of our explanatory variables of interest to changes in model specification we apply (variants) of the so-called Extreme Bounds Analysis (EBA) as suggested by Leamer (1983) and Levine and Renelt (1992). EBA has been widely used in the economic growth literature.48 The central difficulty in this research – which also applies to the research topic of the present paper – is that several different models may all seem reasonable given the data, but yield different conclusions about the parameters of interest. The EBA can be exemplified as follows. Equations of the following general form are estimated: Y ¼ aM þ bF þ gZ þ u;

ð1Þ

where Y is the dependent variable; M is a vector of ‘standard’ explanatory variables; F is the variable of interest; Z is a vector of up to three possible additional explanatory variables, which according to the literature may be related to the dependent variable; and u is an error term. The extreme bounds test for variable F states that if the lower extreme bound for β – i.e. the lowest value for β minus two standard deviations – is negative, while the upper extreme bound for β – i.e. the highest value for β plus two standard deviations – is positive, the variable F is not robustly related to Y. As argued by Temple (2000), it is rare in empirical research that we can say with certainty that one model dominates all other possibilities in all dimensions. In these circumstances, it makes sense to provide information about how sensitive the findings are to alternative modeling choices. The EBA provides a relatively simple means of doing exactly this. Still, the EBA has been criticized in the literature. Sala-i-Martin (1997) argues that the test applied in the Extreme Bounds Analysis poses too rigid a threshold in most cases. If the distribution of β has some positive and some negative support, then one is bound to find at 48

See, e.g. Levine and Renelt (1992), Sala-i-Martin (1997).

16

A. Dreher et al. / Journal of Development Economics 88 (2009) 1–18

least one regression for which the estimated coefficient changes sign if enough regressions are run. We therefore not only report the extreme bounds, but also the percentage of the regressions in which the coefficient of the variable F is significantly different from zero at the 10% level. Moreover, instead of analyzing just the extreme bounds of the estimates of the coefficient of a particular variable, we follow Sala-i-Martin's

(1997) suggestion to analyze the entire distribution. Following this suggestion, we not only report the unweighted parameter estimate of β and its standard deviation but also the unweighted cumulative distribution function (CDF(0)), i.e. the fraction of the cumulative distribution function lying on one side of zero. We base our conclusions on the Sala-i-Martin variant of the EBA.49

Appendix B: Sources and definitions Variable

Description

Source

World Bank projects World Bank projects, IDA World Bank projects, IBRD World Bank disbursements World Bank commitments Temp. UNSC member IMF program

Number of new World Bank projects starting in a particular year. Number of new IDA projects starting in a particular year.

www.worldbank.org www.worldbank.org

Number of new IBRD projects starting in a particular year.

www.worldbank.org

IBRD and IDA net disbursements as a share of GDP.

World Bank (2006a)

IBRD and IDA commitments as a share of GDP.

World Bank (2006a)

Dummy coded 1 if a country is a non-permanent member of the United Nations Security Council, and 0 otherwise. Dummy coded 1 if a country participates in an IMF program during part of the year under Stand-by, Extended Fund Facility, Structural Adjustment Facility, and Enhanced Structural Adjustment Facility/Poverty Reduction and Growth Facility, and 0 otherwise. Total debt service outstanding in percent of GNI.

www.un.org

Debt service (% of GNI) Investment (% of GDP) GDP per capita (log) Population (log) Lagged election Growth Foreign reserves Regime

Voting inline with G3 Budget balance (% of GDP) Inflation

World Bank (2006b)

Private and public gross national investment as a share of gross domestic product (GDP).

Przeworski et al. (2000), extended Gross Domestic Product measured in constant 2000 PPP US dollars. Przeworski et al. (2000), extended A country's (log) population. World Bank (2006b) Dummy variable coded 1 if elections were held the previous year and 0 otherwise. Beck et al. (2001) Growth rate of GDP per capita. Przeworski et al. (2000), extended Gross international reserves in terms of the number of months of imports of goods and services which could be World Bank (2006b) paid for. Dummy variable coded 1 for dictatorships and 0 for democracy. Przeworski et al. (2000), extended by Cheibub and Gandhi (2004) Percent of times a country votes in line with the average of France, UK and US in the UN General Assembly. own calculations

Overall budget balance is current and capital revenue and official grants received, less total expenditure and lending minus repayments for central government in percent of GDP. Inflation as measured by the consumer price index reflects the annual percentage change in the cost to the average consumer of acquiring a fixed basket of goods and services that may be fixed or changed at specified intervals, such as yearly. The Laspeyres formula is generally used. Current account balance Sum of net exports of goods, services, net income, and net current transfers (in percent of GDP). Changes in Changes in net reserves is the net change in a country's holdings of international reserves resulting from international reserves transactions on the current, capital, and financial accounts (in percent of GDP). log(checks) Log of the number of checks and balances. Economic globalization KOF index of economic globalization, measuring long distance flows of goods, capital and services as well as information and perceptions that accompany market exchanges. Social globalization KOF index of social globalization, measuring the spread of ideas, information, images, and people. Political globalization

IMF annual report (various years)

KOF index of political globalization, measuring diffusion of government policies.

World Bank (2006b) World Bank (2006b)

World Bank (2006b) World Bank (2006b) Beck et al. (2001) Dreher (2006), updated in Dreher et al. (2008) Dreher (2006), updated in Dreher et al. (2008) Dreher (2006), updated in Dreher et al. (2008)

49 Sala-i-Martin (1997) proposes using the (integrated) likelihood to construct a weighted CDF(0). However, the varying number of observations in the regressions due to missing observations in some of the variables poses a problem. Sturm and de Haan (2001) show that as a result this goodness of fit measure may not be a good indicator of the probability that a model is the true model and the weights constructed in this way are not equivariant for linear transformations in the dependent variable. Hence, changing scales will result in rather different outcomes and conclusions. We therefore restrict our attention to the unweighted version. Furthermore, for technical reasons – in particular our unbalanced panel setup – we are unable to use the extension of this approach called Bayesian Averaging of Classical Estimates (BACE) as introduced by Sala-i-Martin et al. (2004).

A. Dreher et al. / Journal of Development Economics 88 (2009) 1–18

17

Appendix C: Descriptive statistics Variable

Obs.

Mean

Std. Dev.

Min

Max

World Bank projects World Bank projects, IDA World Bank projects, IBRD World Bank disbursements World Bank commitments Temporary UNSC member IMF program Debt service (%GNI) Investment (%GDP) GDP per capita (log) Population (log) Lagged election Growth Foreign reserves Regime Voting inline with G3 Budget balance Inflation Current account balance Changes in international reserves log (checks) Economic globalization Social globalization Political globalization

5690 5690 5690 3656 3620 5690 5484 2740 3668 4791 5114 5690 4887 3611 4770 4999 3097 3972 3683 3673 3681 3377 3827 3827

1.345 0.725 0.710 0.010 0.013 0.063 0.305 17.198 15.410 7.394 15.307 0.149 3.431 3.431 0.591 0.303 −3.271 50.854 −4.623 −0.009 0.620 49.173 36.420 45.239

2.032 1.557 1.659 0.015 0.025 0.243 0.460 14.229 9.083 1.520 1.906 0.356 6.562 2.997 0.492 0.148 5.880 530.931 20.931 0.044 0.661 20.090 20.584 24.395

0.000 0.000 0.000 − 0.001 0.000 0.000 0.000 0.000 − 3.460 4.035 9.741 0.000 − 51.031 − 0.092 0.000 0.000 − 64.493 − 100.000 − 894.543 − 0.483 0.000 7.839 1.927 1.362

17.000 16.000 17.000 0.178 0.598 1.000 1.000 152.270 72.410 10.877 20.771 1.000 106.280 25.360 1.000 1.000 58.713 23,773.130 56.698 0.489 2.890 98.487 93.097 96.276

References Alesina, A., Weder, B., 2002. Do corrupt governments receive less foreign aid? American Economic Review 92 (4), 1126–1137. Andersen, T.B., Hansen, H., Markussen, T., 2006. US politics and World Bank IDA-lending. Journal of Development Studies 42 (5), 772–794. Baker III, J.A., 1995. The Politics of Diplomacy: Revolution, War and Peace: 1989–1992. Putnam's, New York. Ball, R., Johnson, C., 1996. Political, economic, and humanitarian motivations for PL 480 food aid: evidence from Africa. Economic Development and Cultural Change 44 (3), 515–537. Beck, T., Clarke, G., Groff, A., Keefer, P., Walsh, P., 2001. New tools and new tests in comparative political economy: the database of political institutions. World Bank Economic Review 15 (1), 165–176 (updated August 24, 2005). Bennis, P., 1997. The United Nations and Palestine: partition and its aftermath — UN stance on Palestine's displacement by creation of Israel. Arab Studies Quarterly 19 (3), 47–77. Bernauer, T., Ruloff, D. (Eds.), 1999. The Politics of Positive Incentives in Arms Control. University of South Carolina Press, Columbia, SC. Bird, G., Rowlands, D., 2006. IMF quotas: constructing an international organization using inferior building blocks. Review of International Organizations 1 (2), 153–171. Boschini, A., Olofsgard, A., in press. Foreign Aid: An Instrument for Fighting Poverty or Communism? Journal of Development Studies. Buira, A., 2005. The Bretton Woods Institutions: governance without legitimacy? In: Buira, A. (Ed.), Reforming the Governance of the IMF and the World Bank. Anthem Press, London, pp. 7–44. Chapman, T.L., Reiter, D., 2004. The United Nations Security Council and the rally 'round the flag effect. Journal of Conflict Resolution 48 (6), 886–909. Cheibub, J.A., Gandhi, J., 2004. Classifying political regimes: a six-fold measure of democracies and dictatorships. Paper prepared for annual meeting of the American Political Science Association in Chicago, September 2004. Copelovitch, M., 2007. Master or servant? Agency Slack and the Politics of IMF Lending. Ms. Department of Political Science, University of Wisconsin, Madison.

Cortright, C.R., 1997. The Price of Peace. Rowman & Littlefield Publishers, Inc. Dollar, D., Levin, V., 2004. The increasing selectivity of foreign aid, 1984–2002. World Bank Policy Research Working Paper 3299. World Bank, Washington DC. Dorussen, H., 2001. Mixing carrots with sticks: evaluating the effectiveness of positive incentives. Journal of Peace Research 38 (2), 251–262. Doyle, M.W., 2001. The new interventionism. Metaphilosophy 32, 212–235. Dreher, A., 2006. Does globalization affect growth? Evidence from a new index of globalization. Applied Economics 38 (10), 1091–1110. Dreher, A., Sturm, J.E., 2006. Do IMF and World Bank influence voting in the UN general assembly? KOF Working Paper 137. ETH, Zurich. Dreher, A., Vaubel, R., 2004. Do IMF and IBRD cause moral hazard and political business cycles? Evidence from panel data. Open Economies Review 15 (1), 5–22. Dreher, A., Vreeland, J.R., 2008. Idiosyncratic Election to the UN Security Council. Ms. Dreher, A., Sturm, J.E., Vreeland, J.R., 2006. Does membership on the UN Security Council influence IMF decisions? Evidence from panel data, KOF Working Paper 151. ETH, Zurich. Dreher, A., Gaston, N., Martens, P., 2008. Measuring Globalization — Gauging its Consequences. Springer, Berlin. Drezner, D.W., 1999. The Sanctions Paradox: Economic Statecraft and International Relations. Cambridge University Press, New York. Easterly, W., 2005. What did structural adjustment adjust? The association of policies and growth with repeated IMF and World Bank adjustment loans. Journal of Development Economics 76, 1–22. Eldar, O., in press. Vote-Trading in International Institutions, European Journal of International Law 19;1. Fang, S., 2006. The informational role of international institutions and domestic politics. Manuscript. University of Minnesota, mimeo. Fleck, R.K., Kilby, C., 2006. How do political changes influence US bilateral aid allocations? Evidence from panel data. Review of Development Economics 10 (2), 210–223. Frey, B.S., Schneider, F., 1986. Competing models of international lending activity. Journal of Development Economics 20 (3), 225–245. Harrigan, J., Wang, C., El-Said, H., 2006. The economic and political determinants of IMF and World Bank lending in the Middle East and North Africa. World Development 34 (2), 247–270.

18

A. Dreher et al. / Journal of Development Economics 88 (2009) 1–18

Heston, A., Summers, R., Aten, B., 2002. Penn World Table Version 6.1. Center for International Comparisons at the University of Pennsylvania (CICUP), Philadelphia, Pa. Hurd, I., 2007. After Anarchy: Legitimacy and Power in the UN Security Council. Princeton University Press, Princeton. Hurd, I., 2008. Myths of membership: the politics of legitimation in UN Security Council Reform. Global Governance 14 (2). Hurd, I., Cronin, B. (Eds.), 2008. The UN Security Council and the Legitimacy of International Authority. Routledge, New York. Kilby, C., 2006. Donor influence in MDBs: the case of the Asian Development Bank. Review of International Organizations 1 (2), 173–195. Kuziemko, I., Werker, E., 2006. How much is a seat on the Security Council worth? Foreign aid and bribery at the United Nations. Journal of Political Economy 114 (5), 905–930. Leamer, E.E., 1983. Let's take the con out of econometrics. American Economic Review 73, 31–43. Lerrick, A., 2007. All eyes on the World Bank. Statement Presented to the Subcommittee on Security and International Trade and Finance of the Committee on Banking, Housing and Urban Affairs of the United States Senate. Levine, R., Renelt, D., 1992. A sensitivity analysis of cross-country growth regressions. American Economic Review 82, 942–963. Long, W.J., 1996. Economic Icentives and Bilateral Cooperation. Michigan University Press, Ann Arbor: MI. Malone, D., 1998. Decision-Making in the UN Security Council: The Case of Haiti, 1990–1997. Oxford University Press, New York. Malone, D.M., 2000. Eyes on the prize: the quest for nonpermanent seats on the U.N. Security Council. Global Governance 6 (1), 3–24. Marchesi, S., 2003. Adoption of an IMF programme and debt rescheduling. an empirical analysis. Journal of Development Economics 70 (2), 403–423. Meltzer Commission, 2000. Report of the International Financial Institution Advisory Commission to the US Congress. Momani, B., 2004. American politicization of the International Monetary Fund. Review of International Political Economy 11 (5), 880–904. Moon, B.E., 1985. Consensus or compliance? Foreign-policy change and external dependence. International Organization 39 (2), 297–329. Morgenthau, H., 1962. A political theory of foreign aid. American Political Science Review 56, 301–309. Neumayer, E., 2003. The Pattern of Aid Giving The Impact of Good Governance on Development Assistance. Routledge, New York. Oatley, T., Yackee, J., 2004. American interests and IMF lending. International Politics 41 (3), 415–429. O'Neill, B., 1996. Power and satisfaction in the United Nations Security Council. Journal of Conflict Resolution 40 (2), 219–237. Pilger, J., 2002. How the Bushes bribe the world. New Statesman, September 23, 2002. http://www.newstatesman.com/200209230006. Przeworski, A., Alvarez, M., Cheibub, J.A., Limongi, F., 2000. Democracy and Development: Political Regimes and Economic Well-being in the World, 1950–1990. Cambridge University Press, New York. Rogoff, K., 2004. The sisters at 60. The Economist July 22. Russett, B., 1997. The Once and Future Security Council. St. Martin's Press, New York. Ruttan, V.W., 1996. United States Development Assistance Policy: The Domestic Politics of Foreign Economic Aid. The Johns Hopkins University Press, Baltimore.

Sala-i-Martin, X., 1997. I just ran two millions regressions. American Economic Review 87 (2), 178–183. Sala-i-Martin, X., Doppelhofer, G., Miller, R.I., 2004. Determinants of longterm growth: a Bayesian averaging of classical estimates (BACE) approach. American Economic Review 94 (4), 813–835. Sandler, T., Harley, K., 1999. The Political Economy of NATO: Past, Present, and into the 21st Century. Cambridge University Press, Cambridge. Schneider, F., Frey, B.S., Horn, H., Persson, T., 1985. A formulation and test of a simple model of World Bank behavior. Weltwirtschaftliches Archiv 121 (3), 438–447. Sturm, J.E., de Haan, J., 2001. How Robust is Sala-i-Martin's Robustness Analysis. University of Groningen, mimeo. Sturm, J.E., Berger, H., Jakob de Haan, J., 2005. Which variables explain decisions on IMF credit? An extreme bounds analysis. Economics and Politics 17 (2), 177–213. Temple, J., 2000. Growth regressions and what the textbooks don't tell you. Bulletin of Economic Research 52 (3), 181–205. Thompson, A., 2006. Coercion through IOs: the Security Council and the logic of information transmission. International Organization 60 (1), 1–34. Trudel, R., 2005. Effects of Exchange Rate Regime on IMF Program Participation. Review of Policy Research 22 (6), 919–936. Vaubel, R., 1986. A public choice approach to international organisation. Public Choice 51, 39–57. Vaubel, R., 1991. The political economy of the international monetary fund: a public choice analysis. In: Vaubel, R., Willett, T.D. (Eds.), The Political Economy of International Organizations: A Public Choice Approach. Westview Press, Boulder, CO, pp. 204–244. Vaubel, R., 1996. Bureaucracy at the IMF and the World Bank: a comparison of the evidence. The World Economy 19, 185–210. Voeten, E., 2000. Clashes in the assembly. International Organization 54 (2), 185–215. Voeten, E., 2001. Outside options and the logic of Security Council action. American Political Science Review 95 (4), 845–858. Voeten, E., 2008. A strategic approach to understanding Security Council authority. In: Hurd, Ian, Cronin, Bruce (Eds.), The UN Security Council and the Legitimacy of International Authority. Routledge, New York. Volden, C., Carrubba, C.J., 2004. The formation of oversized coalitions in parliamentary democracies. American Journal of Political Science 48 (3), 521–537. Vreeland, J.R., 2003. The IMF and Economic Development. Cambridge University Press, New York. Vreeland, J.R., 2006. IMF program compliance: aggregate index versus policy specific research strategies. Review of International Organizations 1 (4), 359–378. Vreeland, J.R., 2007. The International Monetary Fund: Politics of Conditional Lending. Routledge, New York. Wagner, R.H., 1988. Economic interdependence, bargaining power, and political influence. International Organization 42 (3), 461–483. Woods, N., 2005. Making the IMF and the World Bank more accountable. In: Buira, Ariel (Ed.), Reforming the Governance of the IMF and the World Bank. Anthem Press, London, pp. 149–170. World Bank, 2006a. World Development Indicators, CD-ROM, Washington, DC. World Bank, 2006b. Global Development Finance, CD-ROM, Washington, DC. Zimmermann, R., 1993. Dollars, Diplomacy and Dependency — Dilemmas of US Economic Aid. Lynne Reinner Publishers, Colorado.