Political advertising in presidential elections and law making in the U.S. Congress

This thesis contains two research areas including political advertising in presidential elections and legislating in the Congress. In the first chapter, we focus on political advertising, and the following two chapters investigate legislating output and success, as well as one special kind of legislation: earmark. Specifically, chapter 1 investigates the effects of political advertising in the 2000, 2004 and 2008 U.S. presidential elections. We integrate the decision of voter turnout and candidate choice in a discrete choice model, with candidate shares derived from the number of registered voters at county level. To address the endogeneity of advertising, we instrument for advertising exposure and tone with information about the advertising price and its interaction with candidate, polling statistics and voter turnout in prior elections. In concert, our model and empirical strategy allow us to discern the role of qualitative and quantitative aspects of advertising on voter choice and election results. This paper contributes to the literature by finding robust effects of political advertising on candidate choice, that advertising exposure benefits candidates who advertise extensively. Advertising negativity are found to be positively associated with higher vote shares in two out of three elections considered. Our results are also robust to different definitions of voter market. After the investigation of pre-election political advertising, we turn to study post-election legislative activities. In Chapter 2, we focus on the output and outcome of legislating in the US. Congress, which have received increasing attention in the legislative literature. This paper mainly employs the data from 2007 to 2010 for the 110th and 111th Congress, explores the factors influencing bill sponsorship and cosponsor- ship, and presents evidence from institutional position, member characteristics and electoral concerns. We show that members holding leadership or committee chairs sponsor more bills but cosponsor less. Female, racial minority and ideologically more extreme members are more active in cosponsoring legislation. Rich committee experience encourages a high level of bill sponsorship and cosponsorship, while freshman class members are generally disadvantaged. Finally, chamber and party difference exist. Further, this paper investigates factors affecting legislating success from the 93rd to 113th Congress (1973-2014), and finds that the number of cosponsors, referrals, majority status and committee chair sponsors significantly help legislation to be enacted into law. Chapter 3 is a natural extension of chapter 2, which investigates into Congressional earmarks in legislations from 2008 to 2010, and finds that retiring members request significantly fewer earmarks, after factors such as seniority, committee membership, general efforts level and local demand controlled for. The result suggests the suboptimal use of earmarks that could be contributed to favor exchange or re-election purposes of the Congressional members