Value Differentiation, Policy Change, and Cooperation in International Regime Complexes Tyler Pratt∗ December 3, 2020 Abstract In many important policy areas, interstate cooperation is governed by a dense net- work of separate but overlapping international institutions. Whether this environment of \regime complexity" strengthens or undermines cooperation is a subject of debate. While some argue that overlapping institutions enhance legitimacy and flexibility, oth- ers claim that opportunistic forum shopping will enable states to escape compliance with rigorous rules. This paper reconciles this debate, demonstrating that regime com- plexity has contrasting effects depending on the degree of value differentiation among institutions. In issue areas where undifferentiated institutions function as substitutes, forum shopping will reduce the regime's ability to discipline state behavior. However, in issue areas where institutions are differentiated by value|i.e., institutions with deeper rules provide greater benefits—institutional overlap can increase policy change among states. I demonstrate these dynamics formally and provide empirical evidence in a comparative analysis of the development finance and election-monitoring regime complexes. ∗Assistant Professor of Political Science at Yale University ([email protected]). I am grateful to Sarah Bush, Christina Davis, Alex Debs, Randall Henning, Kosuke Imai, Amanda Kennard, Robert Keohane, Nico- las Lampach, Melissa Lee, Julia Morse, Abraham Newman, Kelsey Pratt, Charles Roger, Duncan Snidal, Diana Stanescu, and participants at the July 2020 Workshop on International Regime Complexity for com- ments on earlier drafts. 1 1 Introduction Rapid growth in the number and scope of multilateral institutions since World War II has transformed the structure of global governance in many issue areas. Instead of a single unified regime, states frequently confront a regime complex: a set of international institutions that operate in a common issue area and the (formal and informal) mechanisms that coordinate them (Henning and Pratt, 2020). Regime complexes feature a dense network of institutions that compete for authority over the same issue area. This environment gives rise to strategic behavior by states, who can select among multiple institutions when crafting new rules or seeking judgments about compliance. Recent scholarship has improved our understanding of how regime complexity shapes interstate bargaining (Alter and Meunier, 2009; Jupille, Mattli, and Snidal, 2013; Morse and Keohane, 2014) and influences power relations among states and institutional actors (Drezner, 2009; Lipscy, 2015; Pratt, 2020; Henning, 2017). However, existing work provides inconsistent answers to the most fundamental question raised by the increasing density of institutions: how does institutional overlap affect international cooperation? Many scholars argue that the fragmentation of governance across multiple institutions harms cooperation by fomenting ambiguity, encouraging rule conflict, and undermining compliance (Raustiala and Victor, 2004; Alter and Meunier, 2009; Struett, Nance, and Armstrong, 2013). Others contend that regime complexes facilitate more effective cooperation: they can increase flex- ibility (Keohane and Victor, 2011), boost legitimacy (Kelley, 2009), and engender greater expertise (Lesage and Van de Graaf, 2013) compared to unified regimes. These contradictory findings have driven calls for conditional theories of regime complex- ity that explain variation across issue areas (Henning and Pratt, 2020). This paper con- tributes to this agenda by introducing a key omitted variable|value differentiation|that moderates the effect of regime complexes on state behavior. Put simply, regime complexes that are value-differentiated can impose discipline and extract meaningful policy adjustment from member states. Undifferentiated regime complexes are significantly weaker in generat- ing policy change. 2 The paper seeks to accomplish three tasks. First, I propose a useful criterion for assessing cooperation in regime complexes: depth of policy adjustment by states. Policy adjustment is a practical, clear, and widely applicable measure of regime complex effectiveness. It is consistent with the fundamental goal of international regimes, as articulated by Keohane (1984). It is also applicable across a wide range of issue areas, facilitating comparisons across distinct regime complexes. Second, I develop a theory of overlapping institutions that explains why some regime complexes outperform others in generating depth of policy adjustment. The theory explains how expected policy adjustment shifts when new institutions are layered onto a particular issue area. In some issue areas|particularly those where international institutions offer near-identical benefits to member states|the emergence of a regime complex will decrease depth of policy adjustment. As states gain the ability to forum shop (Busch, 2007; Alter and Meunier, 2009), they will opportunistically empower institutions with weaker standards and thereby reduce the need for policy change. This dynamic has generated concern, for example, about the proliferation of development aid institutions.1 Because states can obtain similar benefits (e.g., grants or loans) from an array of sources, they select into institutions where policy conditions are more lenient. The ability to forum shop provides states with leverage vis-a-vis multilateral institutions, weakening the ability of the latter to demand policy reforms. In other issue areas, however, regime complexity can have the opposite effect. The presence of overlapping institutions will deepen policy adjustment if institutions are value- differentiated: i.e., the benefits they provide increase as institutional rules become more rigorous. One example is election monitoring bodies, where institutions with strict standards send a stronger signal about the quality of an election. In this issue area, states face more complicated tradeoffs when substituting one institution for another. Selecting into a more lenient election observer lowers the conditions a government must meet to have its election 1While this paper examines overlap among multilateral institutions, studies of aid competition and frag- mentation typically analyze both bilateral and multilateral aid flows (Annen and Kosempel, 2009; Steinwand, 2015; Bueno de Mesquita and Smith, 2016; Brazys, Elkink, and Kelly, 2017; Gehring, Michaelowa, Dreher et al., 2017). 3 certified as compliant. But it simultaneously reduces the benefit of compliance: clearing a lowered bar sends a less favorable message to domestic and international audiences. In these value-differentiated regime complexes, some states will forum shop to more rigorous institutions in order to extract a greater benefit. If the degree of differentiation is sufficiently large, a regime complex can actually generate more policy adjustment than a unified regime with a single institution. The paper first describes the sources of value differentiation and then demonstrates its effect on policy adjustment. The degree of value differentiation in a regime complex is determined by the type of incentives international institutions offer states to encourage com- pliance. Institutions that induce compliance via certification (sending public signals about a government's behavior) and gatekeeping (granting access to reciprocal policy adjustment by other states) tend to feature high levels of value differentiation. In conducting these ac- tivities, more rigorous institutional standards generate greater rewards. On the other hand, institutions that provide members with private benefits (e.g., aid or technical assistance) tend to have lower levels of value differentiation. Most international institutions perform a mix of these activities. The degree of value differentiation in an issue area is determined by the relative importance of each in motivating compliance with institutional rules. To illustrate the effect of value differentiation on state behavior, I develop a simple model of state compliance with international institutions. The model envisions states as consumers in a \market" for institutional rewards. States decide whether to comply with institutional rules by weighing the costs and benefits. Like consumers, a state's decision depends on the structure of the market. The creation of a regime complex is equivalent to transitioning from a monopolistic market (one international institution) to an oligopoly (multiple institutions). As states gain additional institutional options, their willingness to comply with each body's rules depends on the relative costs and benefits of compliance. The model predicts significant differences in state behavior depending on the differentiation of international institutions. If institutions in the issue area are undifferentiated, the emergence of a regime complex de- creases policy adjustment by states. If institutions are value-differentiated, however, regime 4 complexes can actually increase policy adjustment. Finally, I test these expectations by comparing the effect of overlapping institutions in the domains of development finance and election monitoring. These regime complexes are similar in many of the dimensions identified by (Henning and Pratt, 2020). Both display moderate levels of informal hierarchy and feature institutions differentiated by regional focus. They differ, however, in the degree of value differentiation. Development finance institutions induce compliance with policy conditions by offering private benefits