Significant alarm has been sounded in recent years about state exits from international organizations (IOs). Some recent exits have been high-profile, for example, the United States’ withdrawal from the World Health Organization (WHO) or Britain’s farewell to the EU. Others have gone relatively unnoticed, such as Honduras’ withdrawal from the International Centre for the Settlement of Investment Disputes. Asserting an increase in exits, a burgeoning literature on the “crisis of multilateralism” has identified alternative drivers of IO exit—from growing populism and nationalism to institutional overreach, regime complexity, and global power shifts. Others have questioned whether exit is really on the rise, suggesting that recent cases are an insignificant ripple on a long trajectory of states joining and leaving multilateral organizations.
Exit from International Organizations makes an incisive intervention into such debates by offering a “unified theory” of state withdrawals and suspensions from IOs and providing an abundance of statistical data and analysis. Examining all instances of exit from IO between 1914 and 2022 (488 in total), the book offers a much-warranted reality check on general patterns, drivers, and consequences of states taking leave from multilateral organizations.
What do we learn? A key finding of Exit is that, contrary to present anxieties, IO exits remain infrequent. Historically, the maximum frequency of exits per year is 14; not insignificant, but also not indicating a collapse of international cooperation. Crucially, the normalized rate of exit has not risen lately. Some readers will undoubtedly push back on this finding, pointing to a slate of recent IO exits that are not captured in the extensive dataset on which Exit builds, including the recent US departure from the WHO; the UN Human Rights Council; the UN Educational, Scientific and Cultural Organization (UNESCO); and the UN Relief and Works Agency for Palestine Refugees—all in 2025. Sceptics might also cite Argentina’s departure from the WHO; Burkina Faso, Mali, Niger, and Hungary’s exit from the ICC; and the three West African countries’ simultaneous withdrawal from ECOWAS and the International Organization of La Francophonie (OIF) (also in 2025), along with a slate of suspensions from the African Union, ECOWAS, and OIF. Others may object that recent IO exits differ from the past in that they include powerful countries turning their backs on high-profile multilateral institutions, which have been integral to the so-called Liberal International Order since 1945.
Neither objection sticks when it comes to evaluating the core theoretical and empirical insights of Exit. As Vabulas and von Borzyskowski meticulously document, not only are IO exits a long-standing feature of global multilateralism, but we have seen “high-profile” exits before. Think only of Germany, Japan, and Italy’s departure from the League of Nations, which was followed by a cascade of further withdrawals and suspensions. Indeed, when weighting IO exits by the leaving country’s relative importance in terms of national capabilities, the authors find no increase in “significant” exits between 2000 and 2020 (given data limitations, the book does not provide such a measure beyond 2020).
A second important finding of Exit is that most exits are neither severely penalizing for the departing states nor particularly damaging to remaining IO participants. The reason is simple; whenever exit involves high costs, rational states do not entertain the idea. Dissatisfied states expecting high exit costs are deterred from exiting (opting instead to use “voice” to push for institutional change), whereas remaining participants foreseeing negative consequences from exit prefer to concede institutional change. What we observe instead are “low-cost exits” in which either states with limited bargaining power leave organizations in which they cannot effect desired change, or powerful states withdraw from IOs because they are, all things considered, better off outside. Even low-cost exits are often temporary, as states often continue bargaining over institutional change, paving the way for re-assession (in 45% of cases).
This brings us to the core theoretical and empirical claim of Exit: Goodbyes to IOs are generally not intended to end cooperation; rather, exit is a structural feature of an ongoing negotiation process through which dissatisfied states seek institutional change. We also learn, however, that bidding farewell is an imperfect tool for negotiating institutional change. Given institutional inertia, most IO withdrawals result only in shallow changes to organizational structures and policies. As for suspensions (which typically aim at pressuring states to reform domestic policies to abide by IO rules), the cost of being excluded is generally “priced in” by rule violators, meaning that suspensions mostly fail to deliver reform. This does not imply that IO exits have no effect on international cooperation; rather, the strongest effects may be the hardest to observe. Much like economic sanctions, exit is most effective when credibly threatened but not implemented (via deterrent effects on states for whom exit would indeed incur high costs). Indeed, as we learn from the extensive data, 91% of exit threats are never executed, presumably because adjustments are made by either side to preempt them. The exits we do observe tend, therefore, to be “outlier” cases in which withdrawal/expulsion is preferred to compromise.
Running 325 pages, Exit offers an abundance of fascinating case-based and statistical evidence on exit threats, processes, and consequences. The authors deal deftly with inferential problems and potential confounders, delivering clear and credible findings. The concluding chapter is cautiously optimistic in suggesting that IO exits can help to defend international norms and strengthen international order. That may be the case for some types and instances of exit. However, the finding that many IO exits have led to renegotiation of international commitments in the past does not per se imply this dynamic will continue to hold amid growing geopolitical tensions and conflict over resources and ideology. Indeed, one might argue that lessons from the past century offer somewhat cold comfort when assessing the future of multilateralism: The last period of hyper-nationalism, deglobalization, and democratic backsliding began in the 1920s and lasted until the end of World War I. Like today, economic protectionism skyrocketed, authoritarian populism spread in Europe and beyond, and both great powers and smaller states left the League of Nations and other IOs, which were mired in conflict. From the vantage point of the early 1930s, it would have felt like international cooperation was grinding to a halt—and indeed it was. While IO exits remained low at first, they began to climb rapidly from 1935 (as I show in Eilstrup-Sangiovanni, “Death of IOs. The organizational ecology of intergovernmental organizations 1815–2015.” Review of Intl. Orgs. 15, 2020). A decade later, nearly a quarter of all existing IOs had terminated.
One of the strongest findings of Exit is that states are significantly more likely to exit IOs as their preferences diverge from the mean. This finding seems to bode ill for the (near) future. Indeed, while exits from formal IOs were not much above the norm before 2022, unilateral treaty withdrawals (another indicator of diverging preferences) increased sharply in the past decade. So while the extensive analysis in Exit offers some welcome reassurances regarding IO withdrawals not necessarily leading to a decline of international cooperation, there are still reasons to worry. Clearly, no book can capture every aspect of multilateral cooperation, and while Exit (by the author’s own admission) does not provide a full picture of the current state of multilateralism, it offers an unusually rich and nuanced account of exit from formal IOs and sets out a promising agenda for further research.
