Are there conditioning commonalities between Brexit, the decision of the United States to pull out of the Transpacific Partnership Agreement (TPP), or the Japanese decision to withdraw from the International Whaling Commission (IWC)?
Case-study research has provided us with in-depth knowledge of these withdrawals, whereas emerging large-N time-series cross-sectional studies have unveiled that diverging preferences between individual states and the other members can be a driving force of withdrawal. While both approaches have clear advantages, they also deal with severe drawbacks. The former struggling with issues of external validity, the latter dealing with a large amount of ‘excess zeros’ - cases in which the preference divergence of states would predict a withdrawal, but a withdrawal does not occur. There must exist a conditioning mechanism across cases that determines whether governments that diverge from the preferences of the other members of an intergovernmental organization decide to withdraw. I argue that this conditioning mechanism is the governmental expectation of domestic opposition to the decision to withdraw. Scholarship of coalition behavior in foreign policy-making has developed two conflicting hypotheses. One strand of literature suggests that coalition governments act more extreme, due to the ability of fringe parties in the coalition to hijack the government or the diffusion of blame across coalition partners. An opposing strand of the literature proposes that coalition governments are more restrained in their foreign policy behavior due to the veto capabilities of coalition partners. I analyze the conditioning effects of withdrawal from intergovernmental organizations. I compare first, the difference between coalition governments and their single party counterparts and subsequently analyze the role of ideological fractionalization of coalition parties on the decision to withdraw from an intergovernmental organization. I further highlight the role of key ministers in the decision to exit an intergovernmental organization and assess their influence on exit in a cross-sectional time-series analysis.