3
This paper will draw use literature from symbolic politics and elite framing to develop a
theoretical explanation for the preponderance of conflicted conservatives. I argue that because of the
way political choices are portrayed by both politicians and the mass media, the dominant messages
that individuals receive from elites are themselves conflicted in a way that would lead certain
members of the mass public to hold “conflicted” preferences. In addition, the mass public receives
cues as to the non political-meanings of liberal and conservative (from, for example, church and
religious leaders), and these cues can also color the ways in which individuals perceive how these
ideological terms relate to political stimuli. I find evidence that conflicted conservatives are people
most likely to internalize both “conservative” messages with respect to general ideology and liberal
messages with respect to specific policy areas. In addition, even though conflicted conservatives do
not necessarily know what “conservative” means in the context of politics, “conservatism” as a
general principle is still relevant to political judgment, as the perceived conservatism of candidates
and parties is an important component of political decision-making for conflicted conservatives.
These findings have implications for both the substantive study of ideological self-identification and
American electoral dynamics and the empirical use of ideological self-placement scales in models of
political decision-making.
What is “conflicted conservatism?”
I begin by formally defining the term “conflicted conservative.” Most national surveys ask
people to place themselves on an abstract ideological scale. The scale often has multiple (usually 7)
“points to it, but the essential categories are “liberal,” “moderate,” and “conservative.” In addition,
individuals can be classified as having predominantly liberal or conservative policy preferences. I
operationalize individual-level “policy preferences” using an additive scale of responses to 13 issue-
related questions asked in the General Social Survey, 1974-2002: 11 issues related to government
intervention, one related to the death penalty, and a scale of preferences for the legality of abortion in
various circumstances. All responses are coded as either “liberal,” conservative” or “moderate”
Convention