Less than the Sum of Its Parts: Monitoring Legislative Teamwork in Multi-Member Districts

I present a formal model of the comparative effects of multi-member and single-member districts on legislative and electoral behavior and outcomes. In my model voters choose legislators to maximize pork production. In both systems voters face a moral hazard problem and a learning problem as they attempt to asses how hard the incumbents have worked and how skillful the incumbents are. However, in multi-member districts (MMDs) there is an additional team production problem: voters do not know which of their multiple representatives to credit for pork, which creates incentives for legislators in multi-member districts to free-ride. The model yields several results that are consistent with empirical regularities identified in the extant literature. Legislators from MMDs invest less time and effort in work related to the specific interests of their district constituents than do legislators from SMDs. The incumbency advantage, while present in both systems, is weaker in MMDs. The incumbency advantage and constituency service co-vary. And, the incumbency advantage within MMDs is weaker in districts with unified delegations than those with divided delegations. The model also generates several new hypotheses that could be subject to empirical testing. Decreasing effort with tenure in office will be present in both SMDs and MMDs, however the decline is expected to be smaller in MMDs. Districts with more partisan voters will be more likely to select unified delegations than districts that are more centrist which will choose divided delegations. And legislators from divided delegations within MMDs are expected to invest more energy in constituent relevant work than are legislators from unified delegations. Finally the model sheds light on institutional variations across different MMD systems (e.g., Japan vs. the US) that affect legislative outcomes. 1I am indebted to Scott Ashwoth, Bruce Bueno de Mesquita, Chuck Cohen, Dan Gingerich, Ken Kollman, Jonathan Nagler, Ken Shepsle, Alan Wiseman, the participants in the Harvard Political Economy Workshop and the NYU Political Science Department Seminar for valuable comments. All errors are my own. An earlier draft of this paper was presented at the 2002 American Political Science Association meetings in Boston, MA. Please send comments to [email protected]