In this dissertation, I study campaign effects on turnout and vote choice. I analyze different campaign messages and the way they affect voters across various situations. First, through an online survey experiment, I study the impact of campaign messages and ideological cues on voters as they make inferences on candidates. Next, through a field experiment, I test whether microtargeted messages or general messages on the economy have any effect on turnout. Lastly, using online survey data, I examine how cross-pressured voters behave electorally when holding an opposing party's position on social issues. These three studies indicate that different messages do, in fact, matter. Furthermore, voters are not fools--they are reasoning and rational. While partisanship does continue to heavily impact voting decisions, voters do consider issue positions and different voting dimensions (i.e., social, economic, moral). In light of this, campaigns should continue their efforts to persuade and inform the electorate.

My work is on the importance of elite rhetoric in shaping public opinion despite the fact that party identification operates as an efficient cue for most voters. With a representative sample of respondents, nine experiments, five different elite cues, and four different issues, I show that the role of elite rhetoric on public opinion is more than just simple cueing. Elites matter to public opinion, but in a more dramatic way than we recognize. I show consistently across all experiments that elites must justify their issue positions with coherent content in order to generate support, and in this way, elites can amplify the role that party identification has on shaping attitudes about issues. Lastly, when elites use emotional stories to personalize an issue, they are able to reach across party lines and generate support among out-party members. These justifications, if delivered as a story, can also lead to more favorable evaluations of the elites delivering the justification.

But it’s not that simple. While elites can lead opinion in more ways than we previously realized, they are also subject to more constraints than we understood. To demonstrate this, I test the alternative hypothesis that elites can say anything and voters will blindly follow. I show that elites cannot give any justification and expect voters to support their issue positions; they must provide coherent reasons to generate support. Elites also cannot gain popular support for wholly undesirable policies. Because elites are constrained by justifications and issues, their ability to shape public opinion is more limited than previously characterized.

The three papers in this dissertation examine public policy, representation, accountability, and elections. The first paper studies the effects of two New Deal-era economic relief programs on voting behavior. I find that voters rewarded the Democratic Party for providing direct economic benefits, but not for indirect beliefs delivered through the private sector. The paper suggests that policy, and policy design, influences voting behavior. The second paper tests whether city governments are more responsive to certain demographic or economic groups at the expense of others. To study this, I collect 25 million citizen-initiated 311 calls for city goods and services, and find that cities respond marginally faster to calls from high-income and white neighborhoods. The third paper tests the effects of a popular electoral policy reform --- independent redistricting commissions --- on electoral competition in U.S. House and state legislative elections. I compare competition under the enacted districts to competition under simulated districts as well as competition under a set of of district maps that were considered but not enacted. I find that districts drawn by partisan legislators are just as competitive as those drawn by nonpartisan commissions.

In this paper, I use survey experiments to try to better understand attitudes around political correctness, shifting cultural power, and competing standards for what constitutes racism/discrimination (all highly polarized issues at present). I find no evidence that cultural threats nudge attitudes in a conservative direction the way racial, demographic threats have been shown to do. However, in a separate experimentally manipulated scenario, I find that Democrats are more sympathetic to racial minorities, while Republicans are roughly equally sympathetic to whites and racial minorities complaining about a racially offensive event. This suggests that Democrats tend to consider the historical plight of groups when making judgements about speech norms and what is considered legitimately offensive while Republicans do not. Additional basic survey items support the notion that Democrats and Republicans conceive of racism and discrimination in highly divergent ways, making productive dialogue across partisan camps difficult on these issues, even assuming good faith on all sides.

Background

Objective

Design

Participants

Measures

Key results

Conclusions

Objectives

Methods

Results

Conclusions