Scholars of international relations have asked similar questions about the use of US power in security affairs: Why does the United States sometimes choose to use military force unilaterally as opposed to relying on multilateral action? A common answer in the security studies literature is that "powerful states go it alone because they can.”15 Unipolarity, it is argued, breeds unilateralism.  Yet this answer is not very satisfying. Although the ability to act alone is without question a prerequisite for US actions as an ILLR, it does not explain the decision to actually employ this capability. In fact, I contend the United States has never really wanted to be in the international bailout business. For all of the reasons cited above, its general preference is for the IMF to assume this role. However, despite the fact that acting through an established multilateral process provides numerous benefits to the United States, situations can arise where it has incentives to "go it alone” outside of existing institutions. Stated broadly, my argument is that the United States will act independently as an ILLR when it believes a multilateral response via the IMF is either too slow or too small to protect vital US economic and financial interests.
Multilateralism, as defined by John Ruggie, demands that states surrender decision-making flexibility and resist short-term gains in exchange for benefits over the long run. During times of "normal politics” states operate with lengthy time horizons and should be more willing to think long-term and give up some flexibility. However, when faced with extraordinary and unforeseen circumstances, multilateralism "will entail higher transactions costs than centralized mechanisms” and can "create problems for an organization attempting ... to respond quickly to some exogenous crisis.” Similarly, Stewart Patrick notes that multilateral action
“can slow decisions, dilute objectives, constrain instruments, and culminate in policies of the lowest common denominator.” Elsewhere, Robert Keohane admits that even staunch advocates of multilateral institutions have trouble arguing they are “more efficient than states” and notes they tend to “respond slowly and often partially to rapidly changing events.” Echoing Keohane, Sarah Kreps adds that despite the benefit of burden sharing, multilateral actions are “more time-consuming, less reliable, and more limiting” than going it alone. In sum, despite the very real benefits of multilateralism, scholars of international relations have long recognized that it also brings with it some risks. This is particularly the case during unforeseen crises when policy responses need to be fast, flexible, and forceful. Thus, under certain conditions, states may prefer the freedom of action that is associated with policymaking outside of existing multilateral forums. If a state cares enough about a particular policy outcome and believes that multilateralism will fail to adequately protect its interests, then the expected value of unilateral action can be greater than the expected value of relying on an existing multilateral solution—even if it must bear all the costs of action.
I argue that US international bailouts are a direct response to two chronic weaknesses of the IMF that limit its effectiveness as an ILLR. Both of these shortcomings derive directly from Bagehot’s classic conception of the lender of last resort that lends automatically and freely. The first of these I call the problem of unresponsiveness. Because of the bureaucratic and multilateral process by which IMF loans are negotiated and approved, the institution has rarely lived up to Bagehot’s classical conception of a crisis lender that provides credit automatically to stem panics. What good is a fire truck if it arrives after the house has already burned to the ground? The second weakness plaguing the Fund I refer to as the problem of resource insufficiency. Because IMF resources are finite (limited by the amount its members pay in) and because member countries are limited in the amount of funds they can draw at a given time (called an “access limit,” based on their quota) situations can emerge where the institution simply does not have the resources necessary to stem a crisis on its own. Because the IMF cannot create money like a central bank, the institution does not live up to Bagehot’s classical conception of an LLR that lends without limit during panics. What good is a fire truck if it runs out of water? I contend that when vital US economic interests are threatened by foreign financial conflagrations and US officials fear that the IMF is incapable of effectively protecting these interests on its own, they will pursue actions outside of the Fund that will. To paraphrase a William J. Clinton administration axiom, the United States opts for multilateralism when it can but acts unilaterally when it must.
What are these vital interests that US policymakers are protecting? The answer depends on historical context. In some respects, this book is as much a historical narrative as it is an empirical investigation. Although the primary subject—t he role of America as an ILLR—is a constant, the context wherein these actions take place varies greatly over time. Consequently, the specific threats to US interests that motivated US policymakers to act as an ILLR depend on the circumstances unique to the context within which a particular crisis occurred. During the 1960s, policymakers acted to protect the country’s gold reserves and the dollar’s exchange rate—the linchpin of the Bretton Woods monetary system. In the 1980s and 1990s, policymakers acted to protect the US financial system from sovereign debt and currency crises throughout the developing world. In 2008, policymakers acted to protect the US financial system from foreign bank defaults and the domestic housing market from rising interest rates. In every case, what motivated policymakers to act was a desire to protect US economic interests from a gathering threat. Still, even as its actions have been without exception self-interested, indirectly they help produce the global public good of stability in the international financial and monetary systems. Thus, the outcome of American ILLR actions resembles a joint product: when two (or more) outputs are generated by a single production process.
Although the specific interests that prompt policymakers to intervene in order to protect change over time, the process through which these interests are revealed to be threatened is consistent. The process begins with some transformation in the global financial system. Far from being static, over the past half century the global financial system has undergone a series of changes as national economies have become increasingly integrated with each other. Each period of change has brought with it new economic possibilities but also new risks and attendant challenges for managing these risks. Often, policymakers are unaware of the full scale of such risks until the risks reveal themselves in the form of a crisis. When an unforeseen international crisis finally erupts, it poses unique challenges to maintaining global financial stability. Typically, a crisis reveals the Fund to be incapable of effectively acting as the ILLR due to the problem of
Stages of the Argument
unresponsiveness, resource insufficiency, or a combination of the two. In many cases, the IMF has worked to implement reforms aimed at addressing these problems once they have been made apparent. For example, the Fund may increase member quotas to expand its resources. Or it may introduce a new lending mechanism designed to provide financing more swiftly. However, implementation of these reforms tends to be difficult and necessarily takes time. Reforms like these tend to come too late to address the immediate crisis. With the Fund ill equipped to manage the situation multilaterally, states will look for a solution outside of the IMF. If policymakers believe vital US economic interests are threatened by a crisis, the United States will step in and provide international liquidity. Figure 1.4 presents the order of these stages visually.
The argument ultimately rests on two key testable claims: (1) the belief among US economic policymakers that a particular international crisis poses a serious threat to vital US economic and financial interests, and (2) an inability of the IMF to protect those interests on its own due to the problem of resource insufficiency or unresponsiveness (or both). This book assesses the veracity of both assertions by employing a combination of methods. First, and most important, I look at what policymakers actually said when facing a particular crisis. How grave did they perceive the threat to US interests to be? Could the Fund be trusted to protect these interests on its own? I accomplish this through a careful review of primary historical documents, including congressional testimony, Federal Open Market Committee (FOMC) transcripts, Federal Reserve and Treasury quarterly reviews, and IMF executive board transcripts, as well as interview data. These sources are also supported by secondary historical sources to further uncover what policymakers were thinking and to reconstruct the risks facing the US economy when they made their decisions. The core of the book rests on these detailed historical accounts. In addition to the historical case-study analysis, I also develop and test multiple empirical models of the bailout selection criteria employed by the Treasury and Fed. I do this by exploiting geographic and temporal variation in the recipients of US rescues. To accomplish this, I constructed two unique datasets. The first includes all requests for IMF assistance between 1983 and 1999. The empirical results shed light on why only a small fraction of these countries received an additional bailout from the US Treasury. The second dataset, focusing on the 2008 crisis, includes all countries that had signed the IMF’s Article VIII (the Fed’s informal requirement for a swap agreement). Analysis of these data helps to unpack why the Fed selected 14 foreign central banks for liquidity lines but passed over others. These statistical analyses enable me to further uncover the motives of policymakers by determining whether their observed actions are consistent with the public and private justifications I discuss in my case studies.
-  Kreps 2011, p. 4.
-  Boot 2002; Brooks and Wohlforth 2005; Jervis 2009.
-  Ruggie 1992.
-  Martin 1992, p. 772.
-  Patrick 2002, p. 10.
-  Keohane 2006, p. 4. Emphasis added.
-  Kreps 2011, p. 6.
-  Helleiner 1994.
-  To be clear, this is not intended to suggest causality, only the temporal order of thesestages. For example, a financial crisis does not cause the IMF to be inadequate. Rather, itsimply reveals the underlying inadequacies of the institution.