Paul Quirk

What is the nature of citizen, policymaker, and scientific expert deliberation on U.S. nuclear weapons policies? Are citizens and policymakers appealing to the best available information in nuclear weapons policy debates? Are citizens prone to emotion-driven responses that impede effective deliberation? This dissertation is motivated by a normative concern for broader public participation in national security policy, and it evaluates some of the claims made by rational public optimists in a hard case for citizen competence. It undertakes original research on three case studies related to U.S. nuclear weapons policies from 1985 to 2020: nuclear arms reduction treaties with Russia, ballistic missile defense of the U.S. homeland, and nuclear crisis management in Iraq, Iran, and North Korea. The dissertation examines over 232 public opinion poll items and performs content analysis on an original dataset of 1,000 letters to the editor published in 42 U.S. newspapers to assess the content of public opinion and citizen deliberation on these three nuclear weapons policy issues. The letters dataset provides evidence of reasoning and argumentation on nuclear weapons policy issues, and these expressions of opinion are indicative of what citizens might say in other deliberative forums, such as emails to elected officials, social media posts, and at town hall meetings. Overall, this dissertation finds only partial validation for a competent atomic public. It finds that public opinion on nuclear weapons policy issues is stable and more consistent with the policymaker debates than the conventional wisdom would suggest, though it lags behind the technical-scientific debates. The content analysis finds little evidence of emotion-driven responses that impede effective deliberation. At the same time, the research also finds that citizen and policymaker learning is constrained by a failure to consistently absorb, accept, and appeal to the best available information provided by technical experts on nuclear weapons policy issues.

View record

There are large gaps in opinion between policy experts and the public on a wide variety of issues. Scholarly explanations for these observations largely focus on the tendency of citizens to selectively process information from experts in line with their ideology and values. These accounts are likely incomplete. This dissertation is comprised of three papers that examine other important limitations of expert influence on public opinion on topics featuring widespread expert agreement. The first paper looks at the degree to which information on expert agreement is available in the information environment of the average citizen – the news media – and whether or not such information is clouded by media bias towards balance and conflict. An automated and manual content analysis was conducted on over 280,000 news stories on 10 issues featuring widespread expert agreement. The results show that discussion of expert agreement is extremely rare in news content. On occasions when such discussion is featured, it is typically found in the midst of claims and counter claims by polarizing political actors. The second paper seeks to explain rising climate skepticism in the American public and related polarization of Democratic and Republican Party supporters on climate science. An automated content analysis was conducted on over 26,000 news stories to measure over time dynamics in polarizing information, such as party elite and ideological identity cues, messages from organized climate skeptics, and economic cost frames. Results show that the prevalence of party elite cues is strongly associated with aggregate levels of climate skepticism and polarization even after controlling for other possible factors. The third paper explores the role of anti-intellectualism as a predisposition that governs persuasion by expert agreement and the possibility that anti-elite rhetoric may prime this predisposition in information processing. Findings from the General Social Survey and an original survey of over 3,600 American citizens show that anti-intellectualism is strongly associated with opposition to a variety of positions of expert agreement. Results of an embedded survey experiment demonstrate that anti-intellectuals are less persuaded by messages of expert agreement and that this is particularly true when primed with anti-elite rhetoric.

View record

This dissertation is composed of three papers broadly examining the relationship between the mass media and political polarization in the United States. The first paper examines whether the media might have played a role in the polarization of the American public. Using an automated content analysis of almost 600,000 news articles and transcripts from a variety of prominent news media sources over the past four decades, the paper analyzes whether coverage of ten issues has changed over time along several dimensions of tone (affect, incivility, conflict) and source cues (in particular, whether the media cover increasingly more extreme politicians). The results indicate that the media likely contributed to the process of partisan sorting by increasingly providing the public with partisan signals in the news coverage. There is also some evidence that the media contributed to the affective polarization of the public. The second paper focuses on the nature of media coverage of climate change and its effect on public opinion polarization of climate change attitudes, finding that despite the common perception, the media, including conservative media, did not overwhelmingly promote climate change skeptics, industry groups, or denialist organizations. Instead, the coverage featured an increasing number of partisan cues as the issue rose in salience, which polarized the public. In the third paper, I examine the relationship between climate change attitudes and news media diets. Previous work has focused extensively on Fox News and posits that Fox has been a dominant player in turning the Americans, and especially Republicans, into climate skeptics. Utilizing a large national survey, I find that the relationship is more nuanced than previously argued. Fox News does seem to have a negative effect on supporting governmental action in reducing greenhouse gas emissions, though that effect is limited to a small group of purists stuck in the conservative echo chamber. Most people, and importantly, most Republicans, are not very likely to be members of that group.

View record

This dissertation considers two aspects of legislative representation: (1) how citizens use information about legislative activities and outcomes to assess the performance of the US president and the congressional majority party, and (2) why Canadian MPs debate government bills—even when the government controls the outcome. An investigation of these questions is divided into three principal chapters. First, I examine the effects of legislative outcomes on citizens’ assessment of the president and the majority party in Congress. Prominent theories of legislative behavior argue—and media pundits often assert—that Americans reward these actors if they succeed in passing their bills. But what if the bill is divisive, as is likely the case with well-publicized legislation? Using survey experiments, I show that, on average, citizens still express greater approval for the president and the majority party if Congress passes their ideologically contentious bills—compared with if Congress does not pass them. However, I also find that this reward is typically concentrated among those who already favor the underlying policy change; among policy opponents, the effect is often statistically indistinguishable from zero. Second, I investigate the sophistication of citizens’ judgments of legislative performance. Specifically, do inferential biases—common in other domains—interfere with how citizens evaluate the president and the congressional majority party in light of bill failure? Again using survey experiments, I find that citizens avoid the serious inferential mistake of treating these actors as if they had performed poorly. Instead, I show that their assessments—even in the absence of diagnostic information about those involved—are broadly consistent with realistic beliefs about legislative performance and the obstacles to success in Congress. Third, I explore why Canadian MPs debate government bills. Whereas recent research tends to emphasize legislative speech as a means of communicating with the electorate, the particular rules of government bill debate—coupled with the relatively low visibility of such deliberations—suggest alternative motivations. Using an original dataset of 53 debates, I find no evidence of personal vote seeking; instead, I find patterns of debate participation consistent with attempted obstruction by bill opponents and attempted persuasion by bill proponents.

View record

Over the last two decades, the majority party in the U.S. House of Representatives has increasingly bypassed committee deliberations and restricted floor debate to prevent the minority party from shaping provisions in bills. In response to this rising partisanship, scholars have attempted to measure the direct influence that parties have on legislative outcomes, often by examining roll-call votes. However, they have done relatively little work on the decline of the committee system and the increasing control that majority-party leaders exert on legislative deliberations. In this project, I examine the way in which increasingly cohesive parties draft bills at the prefloor and floor stages of the legislative process. I ask two questions. First, when does the House majority party seize full control of bill development and exclude the minority party from decision-making? Second, what are the policy consequences of the House majority party barring the minority party from formal legislative deliberations? To answer the first question, I construct multiple measures of legislative actions and procedures for major bills developed between 1983 and 2008 in the House. I find that the majority party controls deliberations when its majority status is threatened or its policy goals face strong opposition. I also find that the majority limits discussion on bills designed to promote its electoral brand. These party-brand bills involve tax and welfare policies as well as moral issues. To answer the second question, I create new measures of bill defects and bill extremity to assess how legislative processes affect bill quality. I determine that truncated deliberation tends to produce substantively problematic legislation. Thus, the findings in this project reinforce concerns over and provide evidence of the negative effects of one-party lawmaking.

View record