Dynamics of Third Party Support: The New Democratic Party in Canada.

Most analyses of electoral support of third parties consider the correlates of this support at election time. Existing explanations of support for smaller third parties focus on voter distrust of/alienation from major parties, lack of/declining identification with a major party and the strategic calculations voters who fail to vote sincerely. The focus on election time support of third parties can mask theoretically interesting and important dynamics that may characterize trends in support for such parties between elections. This paper develops a dynamic model of electoral support for the New Democratic Party (NDP): a small, ideological third party in Canada. Espousing a social democratic ethos, the NDP serves as a classic example of a small, ideologically-committed third party.The analysis considers the extent to which two types of factors explain trends in electoral support for the NDP: economic conditions and strategic calculations. The economic voting literature finds that incumbent support is influenced by economic conditions in a reward-punishment fashion. While deteriorating economic conditions may undermine incumbent support, it may also be the case that these same conditions strengthen support for small, non-governing parties such as the NDP. In particular, because the NDP is a left-wing party, it is expected that support will be prominently influenced by changes in the unemployment rate. Secondly, a variation of the strategic vote theorem is considered. Specifically, the hypothesis is that NDP support is a function of time elapsed from the previous federal election and expected time to the next federal election. As a result, it is expected that NDP support will be greatest at the projected midpoint between federal elections because the costs of expressing a sincere electoral preference are lowest at this point.The proposed paper develops a time series analysis using quarterly economic and vote intention data from 1977-2001. Results indicate support for both of the central propositions of the paper. Findings will be discussed in light of contributions to comparatively understanding electoral support for third parties. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]