Electoral Fragmentation: Changing Voter Criteria in U.S. Elections, 1984-2020

• In Event: Evaluations of Politicians and Voting Behavior

Abstract

This study illuminates a previously underexplored aspect of American presidential elections: the evolving dynamics of voter evaluation criteria for presidential candidates. Through the analysis of open-ended responses regarding likes and dislikes about presidential candidates in U.S. presidential elections from 1984 to 2020, drawn from the American National Election Studies, and utilizing a specially developed BERT model, we uncover a range of evaluative criteria. These criteria cover candidate experience, leadership and personal qualities, party and government connections, management, philosophy, and domestic and foreign policy stances.

Our research is centered on longitudinal trends in three primary areas: the diversity of evaluative criteria among citizens, the alignment of criteria preferences across different voter groups, and the uniformity in applying these criteria to candidates from various political affiliations. The investigation into the diversity of criteria seeks to discern whether voters are moving towards a more streamlined or comprehensive approach in their decision-making. Additionally, the analysis of criteria congruence among voters aims to reveal shifts towards individualistic or collective evaluative strategies, offering insights into the commonality of political perspectives. Lastly, by studying the consistency in criteria application, we gauge the extent of comparative analysis voters undertake for candidates from different parties, which might suggest an ability for cross-party understanding and dialogue.

Our findings reveal a significant shift in the criteria voters use to judge Republican and Democratic presidential candidates. There is a noticeable downward trend in the diversity of criteria employed by voters over time for evaluating candidates. Additionally, we observe an increasing divergence in chosen criteria among individual voters, indicative of a growing fragmentation in the evaluation process. We also find a diminishing tendency for voters to use shared evaluation standards when assessing candidates from two different major parties.

This research contributes a systematic, comparative, and longitudinal perspective to the understanding of voter behavior in assessing political candidates. It not only enriches the existing body of literature but also provides critical insights into the shifting dynamics of voter decision-making. By focusing on the temporal evolution of voter assessment criteria, our study offers a novel understanding of the interaction between voters and candidates, highlighting the progressive nature of voter behavior and its implications in today's political landscape.

Authors

• WEI ZHONG, New York University

• Mengqing Zhang

• Zeyu Peng, Massachusetts Institute of Technology