Linking Party Platforms to Perceptions of Presidential Candidates' Policy Positions, 1972-2000
Simas, Elizabeth N., Evans, Kevin A., Political Research Quarterly
If candidates sometimes seek to distinguish themselves from their parties and are ambiguous about their policy positions, to what extent do the policy platforms of parties affect individuals' perceptions of presidential candidate positions? Using data from the American National Election Study and the Comparative Manifesto Project from 1972 to 2000, we show that citizens are able to use party platforms in their assessments of presidential candidates. We also demonstrate that an individual's level of education is important in the process of linking Republican Party platforms to Republican presidential candidates. Our findings have important implications for the role of parties in presidential elections.
party platforms, candidate positioning, voter perceptions, elections
According to the proximity model of voting proposed by Downs (1957), voters choose the candidate who locates closest to them in an ideological space. If this model is to accurately capture voting behavior in modern American elections, one major assumption must be true: voters must be able to identify the ideological positions of candidates. Given the generally low levels of knowledge among American citizens (Delli Carpini and Keeter 1996), this assumption appears to be problematic for Downsian theory. Further compounding the problem is the fact that many scholars, including Downs himself, have shown that it is actually in a candidate's best interest not to make his or her policy positions clearly identifiable to the public (Downs 1957; Shepsle 1972; Page 1976, 1978). Thus, if candidates sometimes seek to distinguish themselves from their parties and do their best to conceal issue positions, and voters have little interest or motivation to actively seek those positions out (Graber 1984; Keeter and Zuckin 1983), to what extent do the policy platforms of parties affect individuals' perceptions of presidential candidate positions?
Although partisanship is the major predictor of vote choice in American elections (Campbell et al. 1960; Bartels 2000), the strong influence of parties' election manifestos that has been found in other democracies (Meguid 2005; Tavits 2007; van der Brug 1999) is absent in the study of American presidential politics. As such, this article aims to explicitly demonstrate that a link between perceptions of presidential candidates and party platforms exists in the United States. In doing so, we build upon the previous literature (i.e., Koch 2002; Franklin 1991) by (1) going beyond the House and Senate to find a connection between parties and candidates at the presidential level and (2) demonstrating that the actual positions of parties, not only peoples' subjective perceptions of them, influence impression formation.1
Using the logit-scale corrected Comparative Manifesto Project (CMP) codings of the Democratic and Republican Party platforms (Lowe et al. 2009)2 as unbiased measures of party positions, we analyze how party positions affect citizen placements of presidential candidates from 1972 to 2000. The results show that voters do have a factual basis for their perceptions of the ideological positioning of presidential candidates. We also find that education plays an important role in the process of linking Republican candidates and Republican Party platforms. Overall, our research suggests that the Downsian assumption that voters can assess their ideological proximity to candidates is not as overly optimistic as previous literature implies.
Linking Parties to Their Candidates
Our expectation that party platforms will have a significant effect on citizen perceptions of candidate positions is motivated by two major findings in previous studies. First, citizen perceptions of candidates have been shown to have a factual basis. Koch (2002) demonstrates that House candidates' true3 positions on abortion significantly affect where citizens place them on the issue. Likewise, Franklin (1991) finds that a Senator's roll call record, as measured by his American Conservative Union rating, impacts where citizens place him on the seven-point liberal-conservative scale. …