The ministerial condition: Political survival and cabinet change

This dissertation explores how leaders extend their time in office by manipulating the tenure of their cabinet members. Leaders need ministers to help them rule and so conventional wisdom suggests that leaders appoint competent ministers to their cabinet. This dissertation shows that this is not necessarily the case. It demonstrates that in small coalitions systems, such as autocracies, leaders maintain incompetent ministers in the cabinet in order to clear the political landscape from potential internal rivals this results in governments with poor performance. In large coalition systems, such as democratic parliamentary regimes, prime ministers maintain moderately competent ministers in the cabinet, while in democratic presidential systems, leaders appoint comparatively more capable cabinet members. In large coalition systems, these strategies maximize a leaders probability of reelection and produce more effective governments. An original database of more than 7,300 ministers of foreign affairs was constructed to test this theory. In addition, the dissertation develops a measure of ministerial competence related to the occurrence of interstate war. This measure of competence builds on a well known theoretical result that demonstrates that governments can reach pre-war bargains that leave them as well off as if they had fought a war while avoiding the costs of fighting, hence effectively preventing armed conflict. The dissertation argues that competent governments, and particularly capable ministers of foreign affairs, are simply more able to reach these ex ante bargains. Consequently, governments run by competent cabinets should be comparatively more associated with long-lasting periods of peace than governments run by mediocre cabinet members. The dissertation tests the effect of this measure of ministers competence on their likelihood of deposition, its impact on the tenure of leaders, and how this depends on political institutions. The empirical tests provide evidence in favor of the theory, especially in small coalition systems such as autocracies, where coalition dynamics incentivize ministers---and leaders---to perform poorly in order to extend their tenure in office.