Designing multilateral environmental agreements

Multilateral environmental agreements have exploded in number in the thirty-five years since the 1972 United Nations Conference on the Human Environment (UNCHE) in Stockholm. Over the same period, there has been a considerable resurgence in the popularity of institutional approaches to the study of international relations (IR). This dissertation evaluates the different explanations for institutional design that are provided by three leading theoretical approaches to IR: realism, rational institutionalism, and constructivism.^ This dissertation argues that there are three critical elements of institutional form that any theory of institutional design must explain: membership, delegation, and flexibility. Membership encompasses two related concerns: first, who may participate in a given agreement, and, second, who must participate for an agreement to succeed. Delegation is understood in terms of its structure and substance: what resources and authority are delegated to third parties, and to what specific ends? Institutional flexibility can take three forms. Adaptive flexibility allows members temporarily to suspend participation in specific circumstances, transformative flexibility allows members to alter the terms of cooperation over time, and interpretive flexibility provides discretion to members in implementing agreement-related obligations.^ After reviewing recent literature on institutional design, the dissertation derives hypotheses from realist, rational institutionalist, and constructivist theory concerning each of these three elements of institutional form. These hypotheses point to the importance of five explanatory variables, distribution problems, enforcement problems, hegemony, the number of relevant states, and scientific uncertainty/knowledge.^ These hypotheses are tested against a database of international environmental agreements compiled by the author and based in part on the recently published International Regimes Database (IRD). Membership rules are found generally to reflect a norm of non-exclusion. Delegation is highly circumscribed among surveyed agreements, and is predicted primarily by the distribution of power among negotiating states. Finally, modest exceptions to reduce transaction costs notwithstanding, institutional flexibility is dramatically undersupplied compared to the expectation of rational institutionalist theory. The dissertation concludes by suggesting how these findings are relevant to ongoing theoretical debates, as well as policy debates concerning the reform of specific international environmental agreements and institutions.