Parliaments, parties, and MPs: A comparative perspective on backbench dissent, party discipline, and intra-party politics (Canada, Great Britain, Australia)

Parliamentary systems are typically studied under the assumption that parties are unitary actors, but in fact maintaining party unity is the central strategic problem of modern parliamentary government. My dissertation delves beneath the assumption that parties are unitary actors and asks what factors make it easier or harder for parties to operate cohesively or—phrased at the level of the individual MP—why do some MPs dissent while others toe the party line?; I consider this question from three theoretical perspectives, economic, institutional, and sociological. In an economic (or purely preference-driven) model of parliamentary behavior the MP's decision to dissent from or remain loyal to her party follows directly from her unconstrained policy preferences. In an institutional and sociological framework, however, the MP's behavior is constrained either by the formal rules or social norms of party and parliament.; The dissertation demonstrates that a purely preference-driven model of parliamentary behavior is simply not empirically supportable. In its place I propose an institutional model in which the party's control of the MP's career prospects is the most important element. This career advancement-dissent model starts from the assumption that MPs desire electoral security, office perks, and policy influence. Institutional rules, specifically the Cabinet's monopoly on power and the doctrine of collective responsibility fuse promotion and policy influence. As a consequence, MPs are institutionally limited to choosing between electoral security on the one hand, and promotion and policy influence, on the other. Dissent comes about when the promise of career advancement is not sufficient to offset the electoral costs of unpopular policies. My career advancement-dissent model finds a good deal of empirical support, across countries and across time. Sociological effects are harder to pin down, however, and they appear to act in a secondary or reinforcing capacity, shoring up party cohesion when instrumental rationality is no longer sufficient to elicit loyalty.