Using movers to identify close election effects

Abstract

Many theories of political participation imply that close elections increase voter turnout, but empirical support for this is mixed. One challenge is that close elections occur in unrepresentative places, making it difficult to extend counterfactual inferences across the wider electorate. In this note, I study closeness in an alternative way by leveraging those who move home between elections. With a large-scale longitudinal survey in Great Britain, comparing individuals who move between safe and competitive parliamentary constituencies, I provide evidence that closeness increases campaign contact but generally fails to affect turnout. British movers are politically comparable to the wider electorate, so the results can be cautiously generalised. This contributes to substantive literature on voter and party-led theories of participation, while adopting an empirical strategy seldom used in the study of political behaviour.