UC San Diego

Why do political parties nominate candidates who have the potential to damage parties' reputation among voters? Specifically, why do parties in India nominate candidates who have been charged with committing violent crimes? In this dissertation, I examine the relationship between parties' goals, candidates' attributes, and electoral competition in order to understand why parties nominate candidates with normatively bad reputations. Chapter 1 provides the motivation for the project, highlighting the importance of understanding parties' candidate nomination logic when trying to understand why candidates with criminal charges are electorally successful in India. Chapter 2 develops a theory of candidate selection that serves as a framework for understanding why parties nominate certain types of candidates. In Chapter 2, I contend that parties nominate candidates who have a comparative advantage in helping parties increase their chances of winning a given seat (vote-generating candidates) into competitive electoral constituencies; parties nominate candidates who have a comparative advantage in helping parties increase their organizational capacity (party-capacity building candidates) into safer electoral constituencies. In Chapter 3, I apply this theory to India, arguing that parties' demand for vote- generating candidates has been increasing over time and that criminal candidates are primarily skilled at increasing parties' vote-share in the constituencies they are nominated into. In Chapter 4, I empirically test the relationship between constituency level electoral competition and the probability parties nominate candidates with violent criminal charges. I analyze party- constituency level data from 22 state level elections between 2003 and 2007. I find that parties that were the closest to winning a seat in their prior election --- parties that faced the highest level of electoral competition --- are the most likely to nominate a candidate with violent criminal charges into that constituency in the current election. I also find that electorally competitive parties that nominate these candidates are more likely to win their constituencies, compared to parties that nominated non-charged candidates. In Chapter 5, I examine whether, consistent with existing scholarship, parties nominate candidates with criminal records because they are wealthier than non-charged candidates. If this conventional wisdom were true, then I would expect parties to nominate their wealthiest candidates into their most competitive electoral constituencies. In contrast to this expectation, I find that parties nominate their wealthiest candidates into their safest electoral constituencies, thus implying that wealthy candidates are party-capacity building candidates and that the electoral success of candidates with violent criminal charges is not due solely to wealth. In Chapter 6, I examine if parties face any downstream consequences related to governance because they nominated candidates with violent criminal charges. Using a regression discontinuity design, I compare the attendance rates of Members of Parliament (MPs) with violent criminal charges who barely defeated a non-charged challenger with the attendance rates of MPs without criminal charges who barely defeated a challenger with a violent criminal charge. I restricted this analysis to MPs who were elected in the 2004 or 2009 national elections and who served out their entire term. I find that MPs with violent criminal charges attend office less often than their non-charged peers