Leader or Party? Evaluating the Personalization Thesis and Vote Choice 1996-2016

• In Event: Democratic Satisfaction, Leaders, and Mobilization: Fresh Insights from CSES

Abstract

It has become almost a political science truism that elections have become more personalized. With the expansion of the visual media, leaders’ are considered to have become essential to voters and parties alike. While a recent profusion of studies has highlighted that leaders can and do have an influential impact on the vote, the more dominant view in the literature is that leaders matter much less than conventional wisdom suggests. In testing this personalization politics thesis empirically, existing studies have been stymied by frequent changes of leader, measurement challenges including separating leader and party effects from one another, and a lack of over-time cross-national data. In this paper, we break new ground by overcoming many of these shortcomings using 20-years of data from Comparative Study of Electoral Systems (CSES) Integrated Module Dataset (IMD) covering 130 elections in 35 states. Using consistent measures of leader and party popularity, we provide the first longitudinal test of the personalization thesis cross-nationally and one which also takes account of leader changes and distinguishes between voters whose choice is motivated solely by the leader and voters driven more by the party. Our results have implications for how we understand the electoral significance of leadership in national elections.

Authors

• Ian McAllister, Australian National University

• Stephen Patrick Quinlan, GESIS Leibniz Institute for the Social Sciences