Discriminating between the directional and proximity spatial models of electoral competition

Abstract

The directional spatial model of electoral competition – recently introduced by Rabinowitz and Macdonald – specifies utility as keyed to intensity of preference and communality of direction between voters and candidates on dichotomous issues. This models stands in contrast to the traditional proximity spatial model which represents utility as a declining function of distance between voter and candidate ideal positions. This paper tests the contrasting predictions of the proximity and directional spatial models, employing data from American and Norwegian Election Studies and controlling for the effects of voter sophistication and issue-placement of candidates by voters. In contrast to previous studies which heavily favour the directional model, the current analysis does not support either pure model over the other but rather a mixed model, which reflects both similarity in issue positions between voters and candidates and the intensities of the voters' and candidates' positions on these issues.