Electoral context

Abstract

This paper explores the importance of context in election studies. A distinction is made between global and compositional effects, and those resulting from measurement error. While compositional effects are essentially spatial, global effects may also result from temporal variation. The mechanisms through which such effects come about are outlined and then some different designs for exploring and assessing such effects are considered. Particular attention is paid to work on contextual effects using British election studies, the work of Huckfeldt and Sprague on one US locality, and comparative studies such as the European election studies and the Comparative Study of Electoral Systems project. It is argued that global contextual effects merit more attention than compositional ones.