Ideological self-identification is driven by symbolic label evaluations, not issue positions.
Evidence
In summary, in three important respects, our data analysis has provided support for our specification of the processes underlying the development of ideological selfidentifications. First, although there may be some reciprocal effects between the two, the data are consistent with our assumption that causality runs primarily from evaluations of ideological labels to self-identification. Not only do such evaluations have a strong impact on selfidentification, but they also mediate almost all of the impact which issues and symbols have on such identifications. Furthermore, neither of the two major alternative models of the relationship between self-identifications and evaluations fits the data very well, nor are either of them theoretically consistent with a lack of bipolarity. Second, and related to our first point, three key findings run contrary to the traditional bipolar conception of ideological identifications: the absence of any strong negative correlation between evaluations of liberals and conservatives; the relationships of different symbols and issue stances to those evaluations; and the different emphasis which liberals and conservatives placed on various categories in their definitions of ideological terms. Finally, our analysis indicates that ideological labels, and consequently self-identifications, have largely symbolic, nonissue-oriented meaning to the mass public. — John T. Jost and Jim Sidanius, Political Psychology, 2080
Categorization
- XREF: Connects to the broader literature on symbolic politics and low-information rationality; resonates with Converse’s belief-systems work and Carmines and Stimson’s ‘easy issues’ concept.
- SEED: Angle: the finding that liberal/conservative self-identification tracks symbolic group cues rather than issue positions has implications for how pundits read election results — shifts in self-identified ideology may signal symbolic realignment, not issue realignment.