"Overlooked Realignment?" Cohort differences in Consideration Set Models

A vast amount of social science literature has repeatedly demonstrated that generational replacement is a steady and fundamental driver of political change. In that way, we have seen resulting generational differences in values, attitudes, turnout, voting behavior, and electoral volatility. In particular, it is established that generational replacement is one of the key reasons for electoral realignment in Western democracies. Those processes have fostered a general fragmentation of Western European party systems with increasing numbers of parties on the macro-level and larger electoral volatility at the micro-level. In that respect, the literature has demonstrated cohort differences at the early phase of the electoral process (understanding of left-right, party identification) as well as at the final phase of it (the party a voter votes for at the ballot box). Little is known, however, about cohort differences in the stage between those two processes - the consideration stage. This lack of research stands in contrast to the potential relevance of this stage in the voting process: A voter’s consideration set size and composition is even more likely to reflect generational shifts than just the act of voting: It might more precisely reflect cohort-based differences in attitudes, salience, party preferences, or the perception of the political space per se. Hence, studying consideration sets from a cohort perspective links the more fundamental understanding of the political space with the increasingly volatile vote decision. In that matter, it serves as an important factor in better understanding cohort-driven electoral realignment. In this matter, this paper tries to study this "overlooked stage" of electoral realignment. I will analyze differences in consideration sets across cohorts over time: Related to this, I study both the size of different cohorts’ consideration sets, as well as their composition. I argue that both are diverging from cohort to cohort: More recent cohorts, socialized in a fragmented party space and less bound to structural determinants of voting, are expected to be more likely to include more parties in their consideration set than less recent cohorts. Furthermore, they have different reasoning for which parties to include in their consideration set: Parties’ positions on the socioeconomic dimension matter less than spatial closeness on the sociocultural dimension. In this regard, increasing volatility among more recent cohorts might be an effect of those cohorts’ increased consideration set size when compared with older voters, while electoral shifts might then be a result of divergent preferences on which parties land up in a consideration set. I test these assumptions by using cross-sectional longitudinal election data from the European Election Studies (EES; 1999-2019) covering eight Western European countries. I apply APC models (including Age, Period, and Cohort) to disentangle the different dimensions of age.