This dissertation examines the involvement of political elites in direct legislation contests in the United States. In order to address this objective, this analysis proceeded by asking three questions. The first question addresses the level of political elites’ involvement. Using social media data, Chapter 2 and Chapter 3 of the dissertation focus on the involvement - defined as the endorsement of ballot propositions - of governors and state legislators. Using data extracted from the Twitter or Facebook profiles of 60 governors between 2012 and 2016 revealed that less than 10 percent of all measures were endorsed by state's chief executives. In addition, considering the official voter information guide printed by the Secretary of State, ballot campaign websites, coverage in two newspapers, and social media data shows that about 20 percent of all observations were endorsed by state legislators in at least one of the four platforms. With the aim of explaining the behavior of political elites, the second question asked in this research concerns the conditions that shape the behavior of political elites, and in particular, the involvement of governors and state legislators in direct legislation contests put on statewide ballots in recent years. With that goal in mind, I have built on previous work in direct legislation but also on executive and legislative politics to provide a theoretical framework. This framework emphasizes two general approaches. Policy and strategy are the two answers to the question of “Why do political elites get involved in direct legislation contests?” The analyses provide much support for my proposed general framework that political elites are strategic players, though not without important qualifications and reservations. Lastly, and addressing the consequences of their involvement, the third question asks: Does political elites’ involvement affect candidate elections? I assessed the impact of candidate positioning from two different perspectives. First, I examined aggregate-level voting behavior in California using vote returns from the 2014 elections. I found that Proposition 1, the salient water bond, did not play any role in moving counties to choose the Democratic gubernatorial candidate. Despite the intense campaigning by Governor Jerry Brown in favor of the water bond, the aggregate-level results failed to find any priming effect of the measure on candidate choice. Secondly, using survey data, I examined the effect of two statewide ballot measures on the 2016 ballot in the Golden State and the endorsements of these propositions by two political elites on the vote for president and state senate. / Doctorat en Sciences politiques et sociales / info:eu-repo/semantics/nonPublished

The citizen legislator is both a controversial and recurring phenomenon of interest in political science research. A longstanding concern for the discipline has been whether or not involvement of the public in the lawmaking process is an asset or a liability to quality governance. This study explores the desirability of citizen legislating in the American states. A four dimensional index is created that includes empirical indicators of "substantive" and "procedural" governance. These indicators include voter turnout, fiscal health, the ideological distance between government and the citizenry, and the diversity of a state's interest group system. The total number of initiatives and popular referendums that appear biennially within each of the fifty states is employed as the key explanatory variable to capture the degree of citizen legislating that is occurring in the states between 1980 and 2000. A random-effects generalized least squares regression reveals that higher ballot measure counts are statistically and substantively associated with better quality governance, indicating that citizen legislation is a quality input into the political system. Key control variables such as divided government, interparty competition, citizen ideological extremism, state legislative term limits, and legislative professionalism also tell particularly poignant stories about the road to good governance.