To Dissent or not to Dissent? Informative Dissent and Parliamentary Governance

Legislative dissent occurs frequently in most parliamentary systems despite its potential detrimental effects on both party and legislator, i.e., legislators depend on their party label for re-election, whose value in turn depends in part on the party’s ability to appear cohesive. Commonly dissent has been attributed to ‘extreme’ preferences. I provide an informational rationale for dissent. Dissent allows the legislator to credibly, albeit not without cost, signal information about his constituency’s preferences to the Cabinet. As a result the Cabinet can better calibrate its policies with the electorate’s preferences. The model provides richer predictions about the frequency and level of dissent than existing theories have. Dissent is shown not to depend on policy preferences alone but also on the legislators’ electoral strength, electoral volatility, and the cost of dissent. Finally, the model suggests that parties may benefit from tolerating some level of dissent. ∗Département de science politique, Université de Montréal, CP 6128, Succ. Centre-Ville, Montréal, H3C 3J7, Canada. Email: [email protected]. I am grateful to Christopher Kam, Randall Calvert, John Duggan, Mark Fey, and Sona Nadenichek Golder for their helpful comments. I also wish to thank the Political Institutions and Public Choice Program at Michigan State University for its support. All remaining errors are mine.