Search Results

This introductory chapter traces the development of population-based experiments and highlights some of their advantages over traditional experiments and surveys. There is a tendency to think about population-based survey experiments as simply a hybrid methodology that melds certain characteristics of surveys and experiments. But to say this tells nothing about which advantages and disadvantages of each methodology are inherited. The chapter argues that population-based survey experiments are instead more akin to an agricultural hybrid that produces something that was not present in either of the two original plants. To the extent that population-based survey experiments can be implemented with effective treatments and with the same degree of control over random assignment as in the lab, it is the only kind of research design capable of straightforwardly estimating population average treatment effects without complex statistical machinations.

This chapter discusses how population-based survey experiments can be especially valuable to more particularistic research because the target of applicability is known and specified. With particularistic research, population-based survey experiments are important not so much to explore boundaries of generalizability as to establish that the empirical cause and effect relationship works as predicted on the targeted population. By increasing the variety of experimental subjects, the settings in which research is done, and the kinds of treatments and measures that are utilized, population-based survey experiments may produce greater awareness of the boundaries of various social science theories. Whether this kind of activity is welcome or not, it remains an important contribution to social science knowledge.

This chapter attempts to begin connecting the dots between the results discussed in Chapter 4 and the ultimate argument that representation itself comes in culturally “Red” or “Blue” shades. The first step is the ballot box, and that is the focus of this chapter. First, the chapter shows that in both 2006 and 2008, those who prefer trustee-style representation tended to vote overwhelmingly for Republican candidates at all levels of government. Next, the chapter reports evidence from a controlled survey experiment that asked respondents to evaluate hypothetical candidates based on representation style that is evidenced rhetorically. This experiment reveals that traditionalistic Christians (and Republicans) are more likely to think that candidates sending “trustee”-style messages will be better representatives if elected.

This chapter presents findings from two survey experiments that explore popular support for individual rights: one focused on free speech in Turkey and one focused on torture in the United States. For both countries, we designed a survey experiment to gauge whether people’s support for a policy changes when they are informed that this policy violates the constitution, and whether people are willing to mobilize to protect their constitutional rights in the face of violations. The goal of the survey experiments was to explore the mechanisms through which rights mobilization can occur and, specifically, whether constitutional violations change people’s positions or motivate them to take action. Both experiments found that being informed that a certain policy violates the constitution does not change popular opinion, which might be one of the reasons why constitutional rights are hard to enforce in the absence of formal organizations.

Chapter 5 explains why partisanship can matter for individual religious behaviors and outlooks. The elite-level link between the Republican Party and organized religion allows voters to associate the parties with different levels of religiosity. Evidence from two survey experiments demonstrates that the close relationship between Republicans and organized religion has pushed Democrats out of the organized religious sphere, while bringing Republicans deeper into the religious fold. Panel data provide additional verification of this finding by allowing me to see how partisans respond when the linkages between religion and the Republican Party become more salient. The data show that, thanks to the increasing importance of gay marriage as a political issue in 2004, party identi?cation in?uenced subsequent religious practices: Democrats (Republicans) reported lower (higher) rates of religiosity in 2004 than they did in 2000 or 2002. And, consistent with the life-cycle theory, this relationship is strongest among respondents who were of an age when they were likely considering re-engaging with or increasing participation in religious life.

Chapter 5 tests the main mechanism that central bank communication changes households’ inflation expectations. The chapter reports evidence from a survey experiment conducted on a sample of German households. It shows that households that receive more precise central bank information react more to central bank news than those who receive more vague information. Estimating the average treatment effect across the two experimental groups, the result is that more precise information attenuates both short-term (one-year-ahead) and medium-term (five-years-ahead) inflation forecasts but not longer-term expectations (ten years ahead). Other findings are that while younger people, women, and those less educated and financially literate have higher prior inflation expectations than their older, male, and financially literate counterparts, the evidence suggests that information rather than socioeconomic or demographic features matters most of all.

This chapter purports to give empirical feedback on impartial reasoning to justice by using online survey experiments. More precisely, the study focuses on whether and how the different conceptions of the veil of ignorance and John Rawls’s method of reflective equilibrium affect real people’s impartial reasoning to justice. The findings show that, while ordinary people support impartial reasoning to the difference principle (maximin), their endorsement of it echoes neither John Harsanyi’s nor Rawls’s reasoning. The results illuminate that findings in human psychology, such as the principles of loss aversion and “perceived luckiness,” cannot be dismissed for denoting the relevant impartial reasoning to justice.

Chapter 8 asks whether changing the types of information provided to voters would sufficiently move public opinion to make such a strategy viable for political actors. Three original survey experiments explore the role of positive factual information, partisan factual information, and simple altruistic framing in shaping opinions. In the first case, a randomly selected half of respondents watched a trade supportive political campaign ad narrated by John McCain. In the second case, respondents received positive messages from experts about the benefits for the United States of the World Trade Organization and the costs to the United States of responding to Chinese currency manipulation, but the partisan attribution of the expert cited in these messages varied. In the final case, respondents identified in random order their preference for U.S. trade policy and their preference for Chinese trade policy. Although all three affected individuals’ beliefs, those effects were not strong enough to overcome most participants’ support for trade protection. Positive messages also increased, rather than decreased, gender and race gaps in preferences for trade protection. The chapter concludes by arguing that these findings support the decision of most individuals seeking reelection not to embrace pro-trade messages.

This chapter describes the methodological challenges involved in answering the question whether constitutional rights matter. While there is no easy way to overcome these challenges, the approach of this book is to use a range of methods to study this topic. This book’s primary method is large-N statistical analysis. But because no single quantitative research design is likely to isolate the causal effects of constitutional rights, it supplements large-N analysis with case studies and survey experiments. By using a range of methods, we can triangulate onto a more accurate understanding of the effect of constitutional rights than any single method could provide. This chapter also reflects upon the challenges involved in measuring human rights violations and describes the data we use to capture rights violations.

This chapter explores the psychology of assessing uncertainty in international politics. It presents a series of survey experiments that test important concerns about the extent to which transparent probabilistic reasoning could potentially warp the quality of foreign policy analysis and decision making. The most important of these concerns is the idea that clear assessments of uncertainty create illusions of rigor that make leaders insensitive to risk. Contrary to this assertion, the chapter’s experimental evidence suggests that decision makers’ choices are sensitive to subtle variations in probabilistic reasoning, and that making this reasoning more transparent encourages decision makers to be more cautious when placing lives and resources at risk. These experiments involve hundreds of real national security professionals and thousands of non-elite respondents.

Chapter 4 provides an original explanation both for why women and minorities are more likely to express protectionist sentiments and for why those protectionist sentiments are not reflected in their voting. The chapter provides an extension of standard models of individual economic well-being to consider trade’s effect not only on wages but also on employment volatility, which is increased by openness to foreign trade. The chapter offers analysis of original survey data from 2006 and 2010 and three decades of American National Election Studies to confirm the previously observed gender gap and newly identified racial gap in trade preferences. The chapter then presents two experimental surveys testing alternative causal mechanisms for the divides. Both experiments vary the type of information provided to respondents about trade partners and potential benefits of trade. In both cases, experiments show stability in women and non-whites preferences for trade and variability in white men’s preferences. Next, the chapter reinvestigates the salience of trade by gender and racial groupings and shows low salience among women and non-whites. The chapter concludes with a description of who might benefit from women and minorities stable preferences and why so few organizations seek to do so.

Chapter 9 assesses the effects of exposure to different author bylines and frames in news coverage of birth control policy. Using a series of survey experiments, this chapter identifies how exposure to the kinds of articles typically produced by male and female reporters influences media trust, political cynicism, and opinions about contraception. The results show that women asked to read articles on the contraceptive mandate written by other women viewed the media as more credible than women asked to read articles written by men. In addition, the results demonstrate that exposure to strategic game coverage significantly decreased perceptions of media credibility, enhanced feelings of political cynicism, reduced issue-specific information retention, and encouraged more frequent expressions of negativity among experimental participants. In line with previous research, the experimental results also show that support for the contraceptive mandate increased after reading an article framed around women’s health, reproductive rights, and sexual morality.

In this chapter, Miquel Pellicer and Eva Wegner make the case for employing quantitative analysis in the study of the Middle East and North Africa. The chapter recognizes the difficulties in obtaining good and reliable data, but employs examples of recent methodological innovations to discuss how such difficulties can be reduced. The authors also point out that quantitative methods in isolation cannot do justice to the political and social complexities of the region and in-depth qualitative knowledge is required to make sense of raw data.

Chapter 6 executes the empirical strategy from Chapter 5 to test the theory presented in Chapter 2 on the South African case. It uses multivariate regression analysis to estimate the effect of ethnic distance on the propensity to vote with one's group and to change one’s vote due to the 2014 election campaigns. To estimate the latter, it uses ethnic distance to predict changes in vote choice across the two surveys. The chapter also uses data from the survey experiment to investigate the effects of campaign appeals more precisely. The chapter then maps survey estimates of racial distance onto real electoral outcomes from the 2014 national elections to test the effect using actual vote returns. And finally, it provides tests of the core social treatment mechanism. The chapter finds support for the argument that differential social treatment weakens identification with one’s group and reduces the likelihood of voting with the group.

Chapter 6 argues that the redistributive nature of trade policy also affects individuals’ trade preferences. Trade protectionism differs from other redistributive policies both in its mechanism for redistribution and the most common portrayal of its beneficiaries. As shown by analysis of trade-related ads from multiple election cycles, images in political ads overwhelmingly present white workers as the beneficiaries of trade protectionism. The chapter describes an original survey experiment that found whites’ support for trade protection depended on the depicted race of trade protection beneficiaries in a newspaper article provided to survey respondents. Analysis of three decades of US public opinion data provides evidence that white support of redistribution via trade protection is higher and support for redistribution via welfare is lower in communities where high levels of racial diversity heighten in- and out-group dynamics. The chapter concludes with a discussion the mobilization of race-based protectionist sentiment in the 2016 election cycle.

Chapter 7 tests the life-cycle theory on African Americans and explores the political consequences of their unique constellation of identities—as both ardent Democrats and frequent churchgoers. Because African-American religious traditions are frequently consistent with Democratic economic messages, African Americans may not feel pressure to update their religious or political attachments in the same way that white Americans might. We should therefore expect that the life-cycle theory will have different implications for African Americans. Whereas African Americans should undergo similar religious life-cycle experiences—decreasing religiosity in young adulthood and then increasing religious involvement in adulthood—they should not necessarily feel that their religious and political leanings are in conflict. Empirical tests bear out these expectations. The chapter then considers the political consequences of African Americans' religious and political attachments. The results from two experiments show that although African Americans are comfortable with religion and politics mixing, as one might expect from a highly religious group, they are not open to this mix when Republicans are doing the mixing. This occurs because the political values valorized in black Protestantism differ from those emphasized by religious denominations and groups associated with the Republican Party.

This chapter studies the relationship between consumer fairness, political preferences, and policy uptake. Americans who support Donald Trump are especially likely to believe the government should be judged by the standards of private companies. New experimental evidence documents that, when politicians of both parties use consumer rhetoric, co-partisans of those leaders subsequently come to view politics in strikingly consumerist terms. In another experiment, results show that voters with low levels of political knowledge look most positively upon a hypothetical political candidate who promises cost-benefit alignability, compared to a candidate who promises more benefits than costs. The chapter then describes a field experiment administered in cooperation with a health insurance cooperative funded under the Affordable Care Act (ACA). A message that framed the cooperative as meeting the standards of cost-benefit alignability caused people to enroll in the cooperative.

The study of public opinion towards Muslim minorities in contemporary liberal democracies where majorities are not themselves Muslims needs to start by highlighting the persistence of hatred and intolerance. There is ample evidence from politics and society in Western Europe and North America that right-wing extremism is on the rise and that this ideology of hate increasingly has vilified Muslims. A close inspection of the free text associations with the word “Muslim” of more than 1,000 randomly recruited Norwegian citizens confirms the existence and vicious nature of both demonization and Islamophobia. Still, the largest share of free associations hint that differentiation may be more common than demonization and Islamophobia. A differentiation experiment embedded in a simultaneous survey conducted in France, Germany, Norway, Sweden, and the Netherlands confirms both that discrimination against Muslims persists and that differentiation is far more common than demonization.

In this chapter, I argue that the salience of particular groups is affected by the rhetoric of institutions and political leaders. To test this argument, I draw on an observational difference in survey of Canada, Germany, and the United Kingdom and experiments in that survey and two others, one of which includes Italy. I show that ethnic and regional arguments shape the relationship between group incomes and attitudes toward redistribution, but that these cues do not affect everyone equally.