Constituent Interest and Congressional Voting

TLDR: The authors showed that congressional voting behavior can be analyzed using a simple principal-agent model, in which political competition constrains legislative agents to serve the interests of those who "pay" for their services with votes and other forms of political currency (for example, campaign funds).

Abstract: THIS article shows that congressional voting behavior can be analyzed usefully with a simple principal-agent model. This model, in which political competition constrains legislative agents to serve the interests of those who "pay" for their services-with votes and other forms of political currency (for example, campaign funds)-is often the starting point for economic analysis of legislation. But a frequent conclusion has been that political ornithology is at least as important as the interests of constituents in explaining legislative voting behavior. Thus, "liberal democrats," for example, tend to vote alike on many specific issues where the diversity of their constituencies would seem to suggest otherwise. This result has emerged in a number of empirical studies of voting that share a common methodology.' This typically starts with a statistical model such as