On the psychological reality of party identification: Evidence from the accessibility of voting intentions and of partisan feelings

Results from a CATI survey of the 1993 Canadian federal election are presented. Response latencies to a voting intention question and a party identification question were obtained to measure the accessibility of these constructs among three groups of respondents: unconflicted partisans, who identified with a party and intended to vote for that party; conflicted partisans, who identified with a party but intended to vote for a different party; and nonpartisans. The voting intentions of unconflicted partisans and nonpartisans were expressed faster after the call of the election than before the election was called whereas those of conflicted partisans were expressed more slowly. Similarly, the party identification of unconflicted partisans was expressed faster after the call of the election whereas that of conflicted partisans was expressed more slowly. The implications of these and related results for the debate between proponents of the classical and revisionist views of party identification are discussed.