In most democracies, the public record of legislative votes in national and local parliaments is an important basis for holding elected officials accountable. In political science, that record is also an important source of data on legislator and party behavior. In practice, many legislatures create a public record of the votes cast by individual legislators for only a fraction of the issues on which votes occur. These recorded votes often are not a representative sample of all votes cast and may exhibit systematic biases that have implications for political accountability and for the science of political behavior. Therefore, understanding the characteristics of the issues that receive a publicly recorded vote (a roll-call vote) is essential to our understanding of democratic processes and evaluating the limits of scientific inferences that can be drawn from roll-call data. This study will advance our understanding of the voting record through examination of national parliamentary bodies around the world.
Intellectual merit. Currently in political science, two accounts, derived from alternative political motivations, serve as the basis of contending explanations of the decision to conduct a recorded vote: disciplining arguments and position-taking arguments. This study (a) develops theory about the implications of these accounts and specifies testable propositions about both what issues come to a vote and what are the consequences for legislators' observable voting patterns; (b) examines the policy domains and legislative sponsors in the full population of votes to the policy domains and legislative sponsors subject to recorded votes; (c) examines the formal rules and political context that condition the issues that receive a recorded or unrecorded vote; and (d) generates and estimates statistical models that account for the effect of expected outcomes on the decision to request a recorded vote.
In doing so, the study provides a deeper understanding of the roll-call vote data generating process than political science has had to date, which, in turn, will allow stronger inferences to be drawn from observed voting behavior.
Broader impact. This study is the first systematic effort to understand the roll-call vote data generating process for multiple national parliaments. The study will advance our understanding of democratic representation, accountability, and responsiveness. In addition, it will provide a large corpus of data on parliamentary voting that will be important to scholars in the field of comparative legislative politics. Finally, the project will provide graduate students with invaluable research experience.
