Although International Organizations (IOs) have the ability to promote economic and political development throughout the world, political imperatives ensure that they will fail to meet their potential. This essay is a response to the editor's kind request to speculate as to the role of IOs over the next decade. Although my arguments apply broadly, here I consider development, and so I focus on organizations such as the World Bank and IMF. Over the coming decades these organizations will fail in their attempts to alleviate poverty. On a more positive note, political science is rapidly advancing our understanding of the pathologies of IOs. Unfortunately, the failure of IOs to alleviate poverty is, I believe, a politically stable circumstance: intellectual advances will not translate into better-performing IOs. I focus on a specific set of theoretical arguments as to why IOs fail to promote development, concluding that we should anticipate more of the same rather than the radical reforms that are necessary if IOs are to achieve their stated goals.

Over the last few years political scientists have made great strides in explaining international relations by opening up the internal politics of nations and organizations to focus on the incentives of decision makers and how institutional settings shape their choices. As Kenneth Arrow's impossibility theorem guarantees, what constitutes national interest depends critically upon the institutional context in which the question is asked.¹

It is not semantics to distinguish a leader's interests from those of her nation.² Leaders have their own objectives; the extent to which their actions are aligned with the interests of the members of society depends upon the context in which they hold office. Like nations, IOs are complex organizations. How, and even whether, they pursue their stated goals depends upon the individuals who run them, the demands of those who fund the organization, and the institutions of the IO.

Institutions shape the policy choices of leaders. Bruce Bueno de Mesquita and colleagues characterize institutions by the number of people whose support the incumbent requires to retain power—the winning coalition (W)—and the size of the pool from which these supporters can be drawn—the “selectorate” (S).³

By way of illustration, democracies are typically large W and large S systems; while military juntas and monarchies have small W and small S. Selectorate institutions shape which policies best enable leaders to survive in office. When coalition size is small, leaders can lavishly enrich their small number of necessary supporters. However, as coalition size increases, for a leader to reward each supporter with individual private benefits becomes inefficient relative to providing effective public policy and other forms of public goods.

Coalition size also shapes political survival. When rewards to supporters are predominately private in nature, as is the case in small W systems, then supporters of the incumbent jeopardize these benefits if they defect to a political rival. This is particularly the case when the selectorate is large, as the political rival will have a large pool from which to draw supporters and is therefore less likely to include members of the former coalition in his coalition. In contrast, when supporters defect from the incumbent in a large coalition system they risk less since rewards in such systems are predominantly public in nature. As a consequence, as coalition size increases leaders must work harder to maintain the loyalty of their supporters and political survival becomes more difficult.

Leaders in large coalition systems commit to policies of low levels of expropriation and taxes and high levels of public goods. These are the policies most likely to help them retain their hold on office given the institutional context in which they serve. These policies also promote economic growth. Leaders in small coalition systems cannot commit to such growth-enhancing policies, as they are incompatible with the institutional context in which the leader holds office. For small coalition leaders, bad public policy is often good politics. IOs operate against this political backdrop. National leaders follow the policies that keep them in office rather than the policies that IOs recommend to promote growth.

IOs provide expertise and resources ostensibly for economic development and poverty relief. Unfortunately, all too often these resources are misappropriated, as illustrated by recent events in Kenya. In December 2002 Mwai Kibaki was elected president following the retirement of the long-term incumbent Daniel Arap Moi. Billions of dollars were stolen under Moi's rule. Given worsening economic conditions, aid agencies such as the IMF agreed to the resumption of aid to Kenya. The Kenyan government promised to reduce corruption. Unfortunately, rather than being used to root out corruption, these funds have been largely stolen. The BBC reports that graft has cost Kenya $1 billion under Kibaki's increasingly autocratic regime. A majority of Kenyans believe they are worse off under Kibaki than Moi.⁴

While the occasional small-coalition leader might be civic-minded enough to carry through reforms—Lee Kwan Yew of Singapore springs to mind—for the most part venality and the imperatives of political survival ensure that economic development fails in small-coalition systems. Political institutions drive policy choices. Without political reforms, development programs fail.

IOs are not blind to the need for political reform. Unfortunately, political leaders in recipient countries are extremely reluctant to engage in genuine political reforms since such reforms jeopardize their tenure in office; under most circumstance they would prefer to forgo development assistance. The personnel of IOs face a terrible dilemma. Although they probably recognize the unlikelihood of subsequent political reforms, they are confronted by terrible need.

The organizational arrangements within IOs reinforce the incentives to do something rather than do nothing, even if the effects are marginal. William Easterly categorizes many institutional deficiencies at the World Bank. For instance, it is a standard practice to measure success by inputs, such as the amount of money loaned, rather than outputs, such as the amount of poverty relieved. IOs generate detailed plans for policy reforms; however, few resources are applied to evaluating why similar previous proposals remain unimplemented. Under such incentives, an ambitious bureaucrat should chase bad loans with more money, rather than find out why the prior attempts increased debt rather than alleviated poverty.⁵

IOs typically fail to promote sustained growth in small-coalition systems because they lack the leverage to induce leaders in these systems to make consequential political reforms. As in the case of Kenya, all too often the resources destined for poverty relief end up in the pockets of politicians and their cronies. My recent work with James Vreeland shows that IMF agreements helps small-coalition leaders survive longer when they are experiencing economic difficulties.⁶

Thus an unfortunate consequence of IOs' programs is that they help maintain the political regimes whose policies produce poverty.⁷

The kleptocratic policies endemic in small-coalition systems risk running down the economy, such that a minor shock can cause an economic crisis. This creates a political crisis for the leader since, as the economy contracts, it becomes increasingly difficult for the incumbent to raise the revenues required to provide the private goods required by supporters. Since further expropriation only makes the economic problem worse, political leaders can no longer resist calls for political reforms. Only through such reforms can leaders credibly commit to the types of policies that encourage economic activity. Of course, such political concessions endanger a leader's long-term political survival. They are therefore only willing to assent to political change when economic deterioration becomes a political crisis.

IOs' assistance during economic crises often lets political leaders off the hook. If leaders are offered an alternative source of revenue from which to reward supporters, they can avoid the need for political reform. While an IO's help often entails a commitment to political reform, once aid is delivered and the leaders' political problems are solved, they renege on promises of political reform. By offering aid in advance of genuine political change, IOs undermine endogenous demands for political reforms.⁸

In small-coalition systems, IO assistance provides little genuine alleviation of poverty and perpetuates political systems that discourage economic development.

Given the failure of IOs to alleviate poverty, it is reasonable to ask why donor nations like the United States, which provide the fuel for IOs, perpetuate their existence when either scraping or radically reforming them would offer greater prospects for promoting long-term growth in recipient nations. IOs have not changed much—nor, I predict, are they likely to—for one simple reason. While they are not effective at alleviating poverty, IOs are an attractive tool of leaders in donor states to reward political leaders in recipient states, that is, to buy policy concessions. Vreeland has recently demonstrated this, showing that nations that align their UN votes with those of the United States get greater access to IMF funds. Vreeland also suggests that there are fewer conditions attached to loans for nations friendly toward the United States.⁹

Donor nations, such as the United States, are typically large-coalition systems. Leaders in such systems survive by providing public goods and effective policies. Buying policy concessions from abroad provides an effective means to deliver rewards to their supporters at home. Democratic leaders promote growth at home because these policies help them survive in office. They do not promote growth abroad because citizens abroad do not vote for them. Rewarding foreigners at the expense of their own supporters jeopardizes large-coalition leaders' tenure in office. While they might wish to promote growth abroad, doing so is not conducive to their institutionally induced survival interests. Leaders in donor nations are interested in obtaining international concessions that reward their voters. Seen from the perspective of donor state leaders, the pathologies of IOs do not appear as such a disadvantage. Perversely, it might actually be disadvantageous to promote political development abroad, since an enlargement of a recipient state's winning coalition makes buying policy concessions more expensive and difficult.¹⁰

My analysis paints a depressing picture for the role of IOs in promoting development over the coming decades. I predict little will change. Although our understanding of IOs and their interaction with donors and recipients will improve, the modus operandi of IOs will persist. The policy prescriptions I propose for effective poverty reduction are that IOs should do less rather than more, at least until consequential political reforms are enacted in recipient nations. This advice runs counter to the prescriptions of others, such as Jeffrey Sachs, who advocate massive increases in development assistance.¹¹

However, IOs will follow neither tack as long as they continue to play a useful role for leaders in donor nations by providing a means through which to reward foreign leaders for policy concessions.