3. The Rational State Choice of Multilateralism + Lisa L. Martin ‘TATRS can choose from a wide array of organizing forms, including multilateralism, as patterns by which to organize interactions among. themselves. The papers in this voltime explore situations in which states have used varying degrees of multilateralism to steucture their relations. Some of these seudies argue thae functional theories cannot explain when governments adopt multilateral norms and organizations. At some level of empirical detail this is surely costece, if for no ocher reason than the high levels of uncertainty that surround the construction of new institu- tions. However, the argument of this paper is that studies of state choice can achieve high payoffs by giving serious consideration co functional- contractual arguments hat see insticutions as a solucion to dilemmas of seraregic interaction.” , ‘Wichoue such consideration, the inadequacy of functional arguments remains an assumption rather than an empirical conclusion. Explanations based on the cole of ideas, domestic politics, or any other set of factors cannot establish their analytic superiority over structural explanations until che conseraints and incentives afising from strategic interaction are taken into account. I argue here that consideration of che power and interests of state actors in different sicuations leads co hypotheses about the “form of successful cooperation," or modal tendencies im che types of norms and organizations thae states create to facilicate pursuit of their I refer to the “institution of multilateralism” as defined by John [Ruggie in the introduction to chis volume. According to this definition, the instieution of malilateralism consists of the three principles of indi visibility, generalized organizing principles (nondiscrimination), and dit- fase reciprocity. In this paper, I ask abour the instrumental value of My shanks to Rober: Kestae wll he pripaes in his poke othe alae commen on hi tec, 992 * Lisa L. Martin multilateral norms under different configurations of state interests, i.c., in different types of “cooperation problems.” The cases examined in this volume demonstrate that state officials believed multilateralism was a valuable means of reaching goals such as liberalization ofthe international trading system, the security of Western Europe, and economic growth. By creating multilateralism as a means rather than a goal, we open the possibility that alternative organizing devices will be equal or superior in their utility for reaching higher level ends such as liberalization. The choice of rools depends, at least in pert, on the configuration of state [power and interests in particular issue areas. Thus, T assume that states ace self-interested and turn to multilateralism only if ie serves theie purposes, whatever these may be. A belief in the utility of multilateralism was expressed after World ‘War It ina drive co create issue-specific organizations as well asa general- purpose organization, the United Nations. In general, multilateral orga- nizations make decisions on a multilateral basis; no small group of _members is privileged co make decisions forthe others. However, we see 1 great deal of variation in the degree co which actual organizations conform to the norms of multilateralism. This paper attempcs to explain vaciation in theit organizing principles and strength on the basis of the strategic problems facing states. In addition, ie suggests hypotheses about felationships among norms, formal orgenizations, and behavioral out- The first section of this paper discusses a typology of cooperation problems and the potential role of the institution of multilateralism (IM) ‘and multilateral organizations (MO) in helping states overcome these problems. Each of four ideal types of cooperation problems—colabora- tion, coordination, suasion, and assurance—presents states with unique cullenges. In some, the functions performed by formal orgenizations, such as monitoring and enforcement, will be essential to the achievement of cooperation. [a ochers, multilateral norms such as nondiscrimination will be more efficient. ‘This consideration illuminaces che functional considerations behind alternative institucional solutions fr different types of games. However, at this abstract level of analysis che outcomes remain inde- terminate. Multiple feasible sotutions exist for each problem. For ex- ample, states can achieve coordination in a game of common aversions through either multilateral discussion and notification of intentions or ‘he establishunene of a focal point by a dominane player. Therefore, the second section of this paper atcemprs co decrease the level of uncertainty "The Rational State Chie * 93 by taking into consideration two elements of international structure in the postwar era: U.S. hegemony within the Western subsystem and the bipolar distribution of power in the international system as a whole. ‘These factors lead us to consider the strengths and weaknesses of multi- lateralism from the hegemon's point of view and che impact of a bipolar secutity structure on relations embedded within this steuctue. ‘The fitst two sections of this paper adope the approach of comparative statics to develop expectations about the role of norms and organizations. ‘The third introduces a dynamic element to the analysis by asking about changes in instiutions in the face of changing distributions of power and other exogenous changes. Some solutions to cooperation problems are preferable co others because certain structures can adapt t0 changes in relative power; this Feature is one ofthe Key advantages of a multilateral architecture. Other forms of cooperation, such as those based on tactical issue linkages forged by « hegemon, will be britele and susceptible o challenge as the hegemon declines relative to other members, Changes in the distribution of power can lead to shifts in the kind of game being played, in addition to affecting the outcome within specified games. For example, suasion’ games will tend to give way to collaboration problems a the distribution of power becomes less asymmetric, with predictable changes in the form of successal cooperation. In this analysis, 1 find it useful to differentiate among the role of ‘multilateralism at three separate points in the cooperation process. First is the stage of arviving at decisions. Seates can reach decisions through genuinely mulilaceral discussions, a series of bilateral agreements, or the imposition of decisions on a unilateral basis. Second, we need to specify the scope of state decisions. Decisions may apply only to those directly involved in their negotiation oF may be extended co a broader range of actors, Finally, norms of multilateralism may apply at the stage of implemencation. Centeal problems at thi stage involve the monicoring and enforcement of agreements, and states may utilize mechanisms rang ing from highly centealized co completely decentralized to solve them. Multilaceral norms may, for example, apply to the scope of ageeements bbue not co their negotiation or enforcement. The questions I address in this paper are about the utility of multilateral norms and organizations at cach stage and thus about expectations for che muleilateral characee of the forums within which specific instances of cooperation are embedded. ‘The conclusions I draw should: be seen as hypotheses Yor purposes of future empicicel examination.94 + Lisa L. Martin Strategic Interaction and Multilateralism ‘As Rugaie notes in the introduction co this volume, multilateralism is 8 “highly demanding form.” Ie requires that states sacrifice substantial levels of flexibility in decisionmaking and resist the temptations of short- erm gain in the interest of long-term benefits. Therefore, ic is unrealistic to expect state behavior co conform co pure multilaceralism. Lastead, we need to ask about the role IM and MO can play under specified condi- tions. Here, I suggest chae focusing on the fundamental problem of strategic inceraccion within an issue area provides some answers t0 ques- tions about the likelihood of successful use of IM and MO. Drawing on ehe work of Duncan Snidal, Arthur Stein, and others, I present a simple four-category typology of cooperation problems. Each of these problems—collaboration, coordinstion, suasion, and sssurance— presents unique challenges to states considering cooperation. Thus, they lead to different expectations about the role of norms and organizations. Consideration of the strategic dilemmas underlying paeticula issue areas in theie simplest form suggests particular relationships between IM and ‘MO. In addition, each situation leads co different relations becween IM for MO and behavioral outcomes. To some extent, these games represent aleernative paradigms in international relations theory. Liberals have often implicitly assumed coordination or assurance games between states. Realism, with its emphasis on power asymmetries, has often assumed something close to a suasion game, while ncoliberals have explicitly focused on collaboration problems. Snidal and Seein differentiate becween ewo prototypical cooperation problems. Snidal refers co chese as coordination and Prisoners’ Dilemma, while Scein discusses coordination and collaboration. In this paper, I use Scein's terminology, for itis more general than Snidal's. I also argue that ‘wo additional game rypes require consideration. Coordination and col- laboration both assume symmetrical incerests. In a suesion game, dehned below, states have asymmetrical interests, 0 that the equilibrium out- come leaves one of them dissatisfied. Although analysis have downplayed the significance of regimes or institutions in such symmetrical siua- tions, I argue that the postwar distribution of power often created pre- cisely this kind of problem and that insticutions could, nevertheless, facilitace cooperation. In assurance games, institutions have little to concribute t0 cooperation under conditions of complete information However, given the structural unceresinty of international relations, The Rational State Choice + 95 states may find some modest level of instieutionalization in these situa- tions conducive to their ability co achieve mutual gains COLLABORATION PROBLEMS Collaboration games are characterized by situations where equilibrium outcomes are suboptimal. These games, and the nature of potential solutions co them, have been che subject of extensive discussion among ‘economists and political scientists, who often consider them “the” collec ‘ive action problem.’ Prisoners’ Dilemma, es shown in figure 3.1, is the most choroughly studied collaboration game in the international relations Hiceratuee.® Figur 3.1. A Colisboration Game (Prisoners Dilemma) Resolving the dilemma of collaboration games is a matter of mutual policy adjustment, for both players must agree to move away from the subopeimal equilibrium, rejecting their dominane strategy. Many authors have identified factors chat allow states to overcome collaboration prob- lems, which focus on using che proper strategy, extending the shadaw af the furure, and relying on centralized mechanisms such as formal inter= national organizations. Multilaceral organizations can play a role in facil- iating cooperation in these types of problems. However, the norms of ‘muleilateralism do not meet the demands of collaboration games, leading. us c0 expect divergence berween the institution of multilateralism and. ‘multilateral organizations in these cases. In the rest of this section, 1 identify ehese demands and the ways in which multilateral norms fail eo resolve, or may even exacerbare, collaboration dilemmas, Collaboration problems contain strong incentives to defect from estab-