Cognitive dissonance, political participation, and changes in policy preferences

We investigate how participation in the electoral process can causally change policy preferences drawing on the framework of cognitive dissonance theory. We present an innovative experimental design, which allows us to isolate the net effect of cognitive dissonance on preference changes. Our results suggest that cognitive dissonance created by expressing support for a losing candidate causally led participants to align their policy preferences with that of the supported candidate more closely. Our results, however, also uncovered a strong dependency of such preference changes on the outcome of the election. When supported candidates won the election, no preference change was observed. Our results may be an indication that previous studies overestimated the cognitive dissonance effect on preference changes.