2019-2020 Spring Colloquia

ABSTRACT: There are many reasons to keep climate change in our attention right now, even while we are in the midst of the COVID-19 crisis. Both are fundamentally political problems, reflecting successes and failures of collective decision-making and cooperation. I’ll describe several new research projects in this talk, each aimed at exploring climate politics in a period of unfolding crisis. The first project uses a “contest methodology” to evaluate interventions to move people to take action about climate change. The second project questions the limits of partisanship to affect costly personal decisions. The third calls philosophers to arms to do more research on climate change, and specifically on climate politics. The fourth explores elements of the moral psychology of living in a heated world.

ABSTRACT: Despite the rapid growth of online political advertising, the vast majority of scholarship on political advertising relies exclusively on evidence from candidates’ television advertisements. The relatively low cost of creating and deploying online advertisements and the ability to target online advertisements more precisely may broaden the set of candidates who advertise and allow candidates to craft messages to more narrow audiences than on television. Drawing on data from the newly-released Facebook Ad Library API and television data from the Wesleyan Media Project, we find that a much broader set of candidates advertise on Facebook than television, particularly in down-ballot races. We then examine within-candidate variation in the strategic use and content of advertising on television relative to Facebook for all federal, gubernatorial, and state legislative candidates in the 2018 election. Among candidates who use both advertising media, Facebook advertising occurs earlier in the campaign, is less negative, less issue focused, and more partisan than television advertising.

ABSTRACT: Severe weather events are becoming more frequent and/or severe due to climate change. Repeated exposure to climate-related events and concurrent individual-level adversity are stressors that may affect subsequent cardiovascular health. The objective of this study was to examine whether location-level severe weather-related deaths (proxy for storm severity and exposure) and recent adversity are associated with new onset cardiovascular health ailments over time. A community-based sample of N=1,963 residents from Texas (n=711), Florida (n=943), and the New York Metropolitan region (n=309) completed surveys between September 2017 and October 2018 after Hurricanes Harvey, Irma, and Michael, and provided reports of doctor-diagnosed cardiovascular ailments before Hurricane Harvey and after Hurricane Michael. Post-stratification weights were used for population-based inferences. The main outcomes and measures included: Self-reported doctor-diagnosed cardiovascular ailments (e.g., high blood pressure, heart problems, and stroke); National Centers for Environmental Information’s (NCEI) Storm Events Data on severe weather events (e.g., number of severe events, number of severe weather event days, weather-related injuries, deaths, and estimated damage costs); Lifetime and recent adversity. In generalized estimating equations analysis, increased incidence of cardiovascular ailments over time was associated with greater exposure to severe weather-related deaths over time (IRR=1.07; 95%CI, 1.01–1.12; P<.05) and recent adversity (IRR=1.08; 95%CI, 1.04 –1.12; P<.001), controlling for covariates. Other severe weather-related exposures were not associated with cardiovascular ailments over time. Therefore, among our community-based sample we found that repeated exposure to severe weather-related deaths and recent adversity were associated with increased incidence of new onset cardiovascular ailments. Our results highlight a link between severe weather and a key metric of human health, and is a potential public health concern given increasing climate-related threats.

ABSTRACT: With increasing use of the internet for social research, Web surveys have become one of the most important and efficient tools for collecting survey data. One of the biggest threats to data quality in Web surveys is breakoff, which we see in this mode of data collection much more prominently than in any other mode. Given the (already) lower response rates in web surveys compared to more traditional data collection modes, it is crucial to keep as many diverse respondents in a given web survey as possible and prevent breakoff bias, maintaining high data quality and producing accurate survey estimates.

We fitted a dynamic survival model to data from a real web survey to predict the likelihood of breaking off at both the respondent and page levels. This model makes use of the survey data, along with rich paradata and accessible administrative information from the sampling frame.

After we evaluated the quality of predictions based on the model, we applied the model as part of a randomized experiment designed to reduce breakoff in the same on-going online survey on sustainability conducted by the Institute for Social Research at the University of Michigan. We used the model to predict page-level breakoff risks in a live fashion while respondents were taking the Web survey. Respondents in the treatment group saw an intervention message once their risk of breaking off passed a certain threshold, while respondents in the control group had the standard collection procedure.

Our analyses show that female respondents and students reacted positively on intervention messages and broke off at lower rates when assigned to the treatment group. Additionally, breakoff respondents within the treatment group answered more survey questions than untreated breakoff respondents.

ABSTRACT: Scholars have argued that leaders pay domestic audience costs for backing down from a prior position. We challenge this argument theoretically and methodologically. We argue that scholars have erred by measuring “costs” exclusively through disapproval of a leader’s handling of the situation when general job approval more accurately reflects audience cost theory. This distinction matters because Americans often have strong existing opinions of the president, such that situational disapproval does not damage general approval. We also argue that the use of hypothetical leaders compounds this problem. We test these assertions using two experiments. Our primary design examines approval for both a hypothetical president, as is common in the literature, and the then sitting president, President Obama. Our secondary design allows us to alter the president’s partisanship. The results strongly support our theory, suggesting scholars have missed an important piece of the puzzle by focusing on situational approval for hypothetical leaders.

ABSTRACT (2 Papers):

1. By many measures, Americans have become increasingly polarized in recent decades. We study the role of the Internet and social media in explaining this trend. We find that polarization has increased the most among the demographic groups least likely to use the Internet and social media, suggesting that the role of these factors is limited.
2. We measure trends in affective polarization in nine OECD countries over the past four decades. The US experienced the largest increase in polarization over this period. Three countries experienced a smaller increase in polarization. Five countries experienced a decrease in polarization. These findings are most consistent with explanations of polarization based on changes (e.g., changing party composition, growing racial divisions, the emergence of partisan cable news) that are more distinctive to the US, and less consistent with explanations based on changes (e.g., the emergence of the internet, rising economic inequality) that are more universal.

ABSTRACT: Social media and online portals are becoming the primary source of information for many voters across the globe. To organize the vast, growing, amount of information online, internet portals employ algorithmic news feeds that relay a combination of news stories selected with a very high degree of personalization. Personalized political information may affect democratic outcomes in many ways; for instance, it bolsters false consensus beliefs by selecting stories consistent with browsing history, and, in contexts with high levels of polarization, may exacerbate policy differences by amplifying partisan views. This paper looks at the impact of personalization on political beliefs and preferences, by implementing a pre-registered, globally replicable, lab-in-the-field experiment with a custom-developed news app, that delivers either a personalized or a chronological news feed. Subjects are monitored continuously and evaluated with surveys on changes to their political and democratic beliefs, including partisan preferences, issue valence, exposure to disinformation, and, crucially, second-order beliefs, that are beliefs about voters of the other parties. The experiment is also a patient-preferred trial to explicitly account for the impact of self-selection. Results from the pilot show that algorithmic news feed are quite powerful in determining media diets, but ineffective at changing preferences.

ABSTRACT: Political factors can shape social determinants of health or can be proximate determinants of health themselves. In the United States, an important political factor is the political party of the executive. The two main parties differ in their ideologies and policy agendas, and these differences have sharpened since the 1960s. We examine the effect of prenatal exposure to the political party in office at the national level (the president’s party) and the state level (the governor’s party) on infant health between 1971 and 2018. Fixed effects models show a beneficial effect of a Democratic president, but no effect of a Democratic governor, on the probability of preterm and low-weight birth. The positive effect of a Democratic president varies dramatically by race: It is weak for white mothers and large for black mothers. The effect of a president’s party does not materialize immediately after the president’s inauguration. Rather, it takes approximately two years after the transition to a Democratic administration to fully emerge, and remains elevated until the end of the party’s spell in office. The effect is only partially mediated by a battery of measurable social policies. The findings suggest that the party in power is an important distal determinant of population health even when experienced before birth, particularly among vulnerable populations.

ABSTRACT: Over the past decades, trust in public opinion polls has declined from historically high levels. Recently, prominent examples of polls that have made inaccurate predictions (e.g., the 2015 Israeli presidential election, the 2016 US presidential election, the Brexit referendum) might seem suggest that this mistrust is warranted. Headlines suggesting that polling is “broken” abound. Yet research suggests that polls that use sound methodology, such as employing random sampling, continue to show a high level of accuracy. The current studies assess how polls that employ probability sampling are evaluated by the public. They test how information about a poll’s methodology affects trust in these polls while considering other influential factors that shape trust in polls, such as whether a poll’s results are congruent with or conflict with a person’s prior views on a topic. Results suggest that providing the public with information stating that a poll employed random sampling methods does not increase trust in this poll on its own, but that educating the public about the need for random sampling in poll methodology can bolster trust in polls that employ probability samples. Helping the public to appreciate the importance of random sampling could encourage public trust in well-conducted polls.