UCLA

In this dissertation, I write three papers on American politics. The first essay, ``Do Americans Respond to Presidential Pork?,'' examines the impact of presidential pork, party identification, and citizen ideology on presidential vote choice and voter turnout. It could be possible that increased government spending in a person's local area will increase his income and thus induce him to evaluate the incumbent presidential candidate through retrospective economic pocketbook voting. I also give a counter reasons of why voters would not respond to presidential pork: Americans do not pay much attention to how much the government is spending in their local area especially when people weigh heavily their ideology and party ID when making vote and turnout decisions. I then find, contrary to previous studies, strong evidence that voters do not respond to pork by changing their who they will vote for but find some evidence that people do respond by turning out to vote. My statistical analysis reveals that pork barrel spending did not switch people's vote choice in favor of the incumbent candidate in the 1988, 2008, and 2012 presidential elections. In fact, I find that the absolute level of spending in per capita terms for a county actually hurt McCain 2008. I do find that pork slightly increases the probability that a person turns out to vote for McCain 2008. But even though the results are statistically significant, they are not substantively significant enough to believe that presidential pork has a huge mobilizing effect for the candidate of the incumbent party. Furthermore, when I analyze the 1988 election between Bush and Dukakis with ANES data, I find that pork has no statistically significant effect on turnout. This leads me to conclude that presidential pork has no significant influence over vote choice or voter turnout.

In the second paper, ``A Factional Theory of Parties,'' I investigate the conditions under which groups in society coalesce to form large-tent parties previous to any election. I construct a formal model where groups seek to win office to implement their ideal points. These groups can either run on their own in the election or form coalitional parties. Each group has activists who are able to influence independent voters by contributing campaign resources. Coalitional parties can form in two cases. First, if there are no intense policy demanders (groups who care mostly about one policy dimension), then groups will coalesce and combine activist resources only if their influence in campaigning is significant. Second, if there are many policy dimensions and intense policy demanders then the influence of activists can be very weak and coalitional parties will still form. The reason why they form is very similar to vote-trading (logrolling). And coalitions need not be minimal winning in my model.

In the third paper, ``Who Are the Moderates?,'' I investigate which Americans are extreme in their political opinions. I focus on two schools of thought: the political economy school and the public opinion school. The political economy tradition argues that people are mainly interested in their material interest: rich people want less income redistribution and poor people want more. Thus, rich people should be conservative and poor people should be liberal. The public opinion tradition argues that the more politically informed a person is the more likely she is to agree with her partisan affiliation. I extend this to imply that more informed people are more likely to be extreme in their political beliefs. Since Liberal and Conservative ideologies hold policy positions on many different issues (the economy, abortion, immigration, gay marriage, the environment), I do a study of how income and political knowledge effects a person's opinion on specific issues. I find support for both of these schools of thought on public opinion.