Interdisciplinary Workshops on Politics and Policy

Joint work with Ben Weidmann: Attrition has been described as “the Achilles Heel of the randomized experiment” (Shadish et al., 1998). The main virtue of randomization is its ability to create groups whose only systematic difference is treatment status. Attrition looms as a threat because it can undermine group equivalence, eroding the methodological strength at the heart of a randomized evaluation, a cornerstone of education research. But how large are these biases in practice?  To investigate this question, we estimate the magnitude of attrition bias for 10 randomized controlled trials (RCTs) in education. We make use of a unique feature of administrative school data in England that allows us to analyze post-test academic outcomes for nearly all students, including those who originally dropped out of the RCTs we analyze. We find that the typical magnitude of attrition bias is 0.015 effect size units (ES), with no estimate greater than 0.034 ES. This suggests that, in practice, the risk of attrition bias is limited. However, this risk should not be ignored as we find some evidence against the common ‘Missing At Random’ assumption. We recommend that researchers incorporate uncertainty due to attrition bias, as well as performing sensitivity analyses based on the types of attrition mechanisms that are observed in practice. We provide an approach for both of these things.

Explicit racism in political campaigns is on the rise. While some work has examined how minoritized groups mobilize into politics in the face of such political discrimination, little direct causal evidence exists on the effects of inflammatory campaign rhetoric on members of marginalized groups being directly attacked, and also on those who hold neighboring identities to the direct targets of political attacks. Identity and membership in social groups are powerful forces in politics, producing effects such as partisan bias, racial affinity and discrimination, welfare chauvinism, and many others. And, existing research suggests that while members of minoritized groups do punish candidate derogation aimed at other minoritized groups, responses to attacks on their own ingroup might be stronger. In the face of hostile political rhetoric targeting one outgroup, do individuals with neighboring identities also respond as strongly? We focus on how anti-immigrant attacks affect both immigrant and non immigrants in Canada and the United States. Using Latino-American and South-Asian Canadian samples, we conduct two experiments and randomize respondents into one of three fictional social-media style campaign videos on immigrants, a specific panethnic group (Latinos / South Asians) or a control condition. Results show strong treatment effects on both emotional responses and candidate evaluations. There is some evidence of a difference between targeted and neighboring identities, with immigrants reacting slightly stronger to the anti immigrant ad. However, the difference is quite small, and swamped by the much larger general treatment effects on all respondents. Together, these results point to the importance of neighboring identities, and how social and psychological connections can produce effects as large as actual membership.

Seeking international support to counter Russia’s February 2022 invasion, Ukrainian President Volodymyr Zelensky has publicly addressed politicians from several democratic nations. Media coverage paid special attention to his explicit comparison of Ukraine’s current situation to salient historical events in the audience countries. Given that public opinion can influence foreign policy decisions in democracies, we investigate whether evoking the audience country’s past effectively increases popular support for Ukraine. We conducted preregistered survey experiments simultaneously in four countries where Zelensky delivered speeches rich in historical parallels – Germany, Israel, United Kingdom, and United States. Exposure to excerpts from his speeches triggered distinctive emotional reactions in all countries, reactions consistent with the substantive content tailored for each country. Only in Israel, however, did this rhetoric increase public support for measures intended to bolster Ukraine’s war efforts. Thus, while rhetoric emphasizing past-present commonalities might pull at the heart strings, its persuasive potential appears limited.

Politics is a social enterprise, with most of its activities taking place in public, sometimes in open confrontation with others. As such, personality traits that define sociability govern levels of political involvement that involve social aspects – e.g., attending a rally and canvassing. We add to this literature in two ways. First, we show that extraversion benefits men’s political activities more than women’s using panel data gathered around the 2016 and 2020 U.S. elections and a module of the American National Election Study. That is, the gender gap opens among extraverts. Second, we argue that this is due to the perceived pleasantness of political activities, a mechanism which has been overlooked by previous work studying the link between personality and political participation. Extraverts are only more likely to participate in certain actions if activities are perceived as pleasant. With pre-registered hypotheses, we directly test the connection between gender, extraversion, and pleasantness of political activities in our 2020 data. Our findings suggest women are less likely to experience political activities as pleasant, which serves to curb the positive effect of extraversion. Finally, I will detail a current study in the field, using round-robin, dyadic, in-person conversations to uncover how gender and extraversion shape political and nonpolitical discussions and perceptions of conversation quality.

A key challenge for countries with large refugee populations is to establish policies that help refugees while ensuring that host country nationals do not develop negative attitudes towards them. In principle, international aid programs can help to address this dilemma by generating additional resource flows into the country. Can international aid induce more altruistic attitudes and behaviors towards refugees in the local population and, if so, which types of aid programs are more or less likely to do so? Based on a two-month randomized control trial in Turkey with multiple survey waves conducted using an online platform, we explore whether distinct types of aid, notably cash transfers, vocational training, and social cohesion programs, have different effects on local population attitudes and altruistic behavior vis-a-vis refugees, our main outcomes of interest. We further assess whether these effects vary depending on whether the funder is the host government or international donor and if international aid is framed in terms of helping the local economy versus the refugee population. Understanding the conditions under which international aid policies successfully integrate refugees in host countries may improve the effectiveness of programs aimed at supporting refugees and nationals in receiving countries alike.