18
matters to any great degree for conflicted conservatives, and supports the commonly-asserted notion
that the use of ideology in structuring political choices is limited to certain (sophisticated) subsets of
the electorate.
But there is more to the story. Precisely one of the reasons that “conflicted conservatives” are
conflicted is because they cannot necessarily understand how ideological labels map onto political
life. They identify as conservative for a multitude of (political or non-political) reasons, but not, it
seems, because of their policy preferences. They perceive themselves as conservative when they are
actually not. In much the same way, they may also inaccurately perceive the ideological leanings of
political parties and candidates—viewing certain candidates, issue positions, and the like as
“conservative” when the opposite is true. But if conservatism as a label is truly important to this
group and not simply an empty label devoid of any meaning, then these perceptions of conservatism,
even if they are inaccurate, should affect political choices. I thus address the impact of
“conservatism” for conflicted conservatives not just by understanding the impact of ideological self-
identification, but also as whether the perceived conservatism of political candidates (even if
incorrect) affects vote choice.
[Table 6 about here]
I analyze this question by examining the effects of perceived conservatism of the two major-
party candidates on the 2000 vote. The 2000 NES asked respondents to place both George Bush and
Al Gore on a seven-point ideological scale. I code responses as “correct” if respondents place Bush
to the right of Gore, and “incorrect” if Gore is placed to the right of Bush. Table 6 gives descriptive
statistics of this measure for conflicted conservatives, consistent liberals, and conflicted
conservatives. First, the distribution of “correct” responses is striking: conflicted conservatives are
considerably more likely than “consistent” ideologues to get candidate placements wrong. Only 52