Cabinet Decision Rules and Political Uncertainty in Parliamentary Bargaining

game theoretic models in political science frequently suggest that the outcomes of bargaining processes in democratic systems should be significantly influenced by the precise character of proposal and veto opportunities, on the one hand, and by informational asymmetries that exist among participants, on the other. In parliamentary systems, although the cabinet is widely recognized as the central arena for strategic bargaining, scholars have examined neither variation in cabinet decision rules nor the effect of private information on bargaining processes in the cabinet. Our analysis attempts to fill this gap by arguing for the central importance of distinguishing between collective cabinet decision institutions and institutions that permit unilateral action by the prime minister on confidence motions. Our models suggest that this institutional distinction profoundly affects the distribution of bargaining power within the government. Under collective rules, pivotal parties within the prime minister’s majority have the advantage of being able to make the final policy proposal; under unilateral rules, the prime minister has this advantage. The distribution of proposal power, we argue, affects not only the nature of policy outcomes but also the types of political considerations that parties face when choosing a prime minister. These institutions also influence the ability of parties to exploit political uncertainty. They affect, for example, the capacity of the privileged actor within a coalition to extract policy concessions, the propensity for inefficient bargaining failures to occur, and the circumstances surrounding such failures. 14 We used Keesings Contemporary Archive to code government terminations (from the Woldendorp data) according to whether (1) they occurred because of policy conflict, and (2) they ended with resignations before floor debate as opposed to defeats on the floor. We excluded governments that ended for reasons other than policy conflict. For the remaining cases, our dependent variable takes the value 1 if resignation occurred, 0 otherwise. There are 130 observations, 81 of which take the value 1 (for resignation). Cabinet Decision Rules and Political Uncertainty in Parliamentary Bargaining June 2001