Replacement of a prime minister by another from the same party is a relatively unexplored subject in the study of cabinet termination. This article examines the effect of leadership selection mechanism on prime-ministerial turnover.
We develop the two simple game-theoretic models; (1) a “closed” leadership selection model in which parliamentary party group selects the new leader and (2) an “open” leadership selection model in which extra-parliamentary party selects him/her. We argue that the policy preference, competence, and popularity of the newly elected leader depend on which mechanism is used. We show that under closed selection, prime ministers are more likely to be replaced than under open selection.
We evaluate our theory empirically using the data of Australia, Canada, Japan and the UK for the period after 1975. Using survival analysis, we find that prime ministers are more likely to be replaced when his/her party adopts the closed leadership selection rule under which only parliamentarian can participate in the selection process.