UC San Diego

This dissertation broadly explores the increasingly prevalent but understudied phenomenon of famous political amateurs, or celebrities, running for and serving in political office. As there is yet no comparative empirical study of this phenomenon, chapter 1 takes the preliminary step of tracking the rates of celebrity candidacy in democratic polities around the world, wherever data is available. In the rest of the dissertation I take advantage of a wealth of voter survey data, election records, and legislative behavior data from Japan to test hypotheses about the causes and consequences of this phenomenon. In chapters 2 and 3 I provide both experimental and observational evidence that electoral systems play a large role in the electability of these candidates, and that contrary to popular belief their supporters are not disproportionately young, less educated, nor apolitical. In chapter 4 I find experimental evidence that such support may also be linked to an individual's trust in government, however not through the commonly assumed mechanism of a protest vote. Rather, the results suggest that information about political scandal induces a sincere preference for celebrities as viable political outsiders. Finally chapter 5 takes the first steps towards empirically testing the popular notion that celebrities make for less reliable and less competent legislators. Tracking individual legislative performance in Japan's upper house from 1968 to 2015 using 11 different metrics, I find that by many measures celebrity legislators are indistinguishable from their less famous counterparts, while on some key indicators - such as bill sponsorship and parliamentary questioning - they are actually significantly more active