A Theoretical Model of Roll Call Vote Requests

Parliamentary voting data are used increasingly to analyze the behavior of members of parliament (MPs). These data are one of the very few systematic records of legislative behavior. Yet in many parliaments, recorded and published votes might be selected strategically and thus form a non-random and potentially biased subset of all votes taken. Consequently, in the extreme, inferences based on such parliamentary voting records may be meaningless. I thus address the following research question: under what circumstances are roll call votes requested in parliaments? In order to answer this question, I develop a game-theoretic model of roll call vote requests. The model shows that -- depending on the legislative context -- roll call votes might not be a representative sample of the universe of votes cast. Thus, specific knowledge of the legislative setting is necessary in order to understand the sample properties of roll call vote data