PEER-REVIEWED PUBLICATIONS
Studies have found that pandemics can heighten xenophobia among host citizens, often explained by the behavioral immune system theory or elite-driven scapegoating. However, most research has overlooked the role of pandemic-related economic restrictions and job loss on sentiment toward immigrants. To isolate this economic mechanism, we examine the case of Venezuelan migrants in Colombia before and during COVID-19. Despite the Colombian government’s severe economic lockdown, few politicians blamed Venezuelans for the pandemic. Thus, any economic impact on xenophobia should be evident. Using a panel experimental survey of 374 Colombians, supplemented with 550 new respondents at endline, we find no evidence that exposure to COVID-19 changed attitudes towards Venezuelans, even for those directly affected by the pandemic. Yet, those who did not lose their jobs viewed Venezuelan migration more positively at endline, providing support for the economic effects of pandemics.

Survey respondents' experiences with COVID-19 (top), and the proportion of respondents who agree with the migration statements at endline in February 2021 (bottom).
Presented at the UBC Comparative-Canadian Workshop and APSA 2022.
Links: Paper | Supplementary Materials | Replication Materials | Pre-Analysis Plan | Slides
Do perceived political views of migrants affect their treatment? For existing studies of migrant reception largely conducted in the Global North, the overlap between ethnicity and partisanship has made it difficult to disentangle political fears from other status and identity concerns. We leverage a case in which migrants come from a similar ethno-linguistic background to explore the role of political fears. Drawing on an original face-to-face survey with over 1,000 Colombians and 1,600 Venezuelans in Colombia, we find that Colombians view Venezuelan migrants as left-wing even though actual Venezuelan migrants are more right-wing than their Colombian hosts. These political misperceptions are consequential: we find that Colombians strongly oppose the settlement of left-wing migrants in their communities. Our research implies that societies can construct out-groups along political lines when the ethnic and cultural bases for migrant exclusion are weaker.

Conjoint results for Colombian and Venezuelan survey respondents highlighting partisanship preferences for migrant profiles. Locations of respondents in Cali and Cúcuta.
Presented at APSA 2020, Global Research in International Political Economy (GRIPE), UCLA Center for the Study of International Migration.
Links: Paper | Supplementary Materials | Replication Materials | Pre-Analysis Plan | Slides
Despite record numbers of internally displaced persons (IDPs) globally, we know little about the causal effects of intergroup contact between migrants and locals in active conflict settings. We conduct a randomized controlled trial of a vocational skills-training program implemented by Mercy Corps that enrolled 2,597 locals and migrants in near equal numbers in Kandahar, Afghanistan, where IDPs face daily prejudice and discrimination. Courses lasting up to six months emphasized collaborative learning and soft skills development. We surveyed participants at endline and followed up eight months later. While the program provided the most sustained duration of intergroup contact (360–720 hours) experimentally evaluated to date, we find no evidence of reported behavioral or attitudinal change by locals (N=1,276) toward migrants generally, regardless of classroom demographics or course duration. Our findings suggest that prolonged contact through vocational training programs is insufficient to improve relations in conflict settings.

Null effects of social contact through vocational training classes on locals' relations with migrants. Subgroup effects by course duration (blue) and composition (red).
Presented at APSA 2020, UBC Centre for Migration Studies, UPenn Perry World House Seminar, Texas A&M Bush School of Government and Public Service, PRIO Peace Research Speakers Series, World Bank-UNHCR JDC Research Conference on Forced Displacement, University of Washington International Security Colloquium, EGAP Priority Themes Seminar Series, Center for Conflict and Violence Prevention at USAID.
Links: Paper | Supplementary Materials | Replication Materials |
Pre-Analysis Plan | Slides | Media: EGAP Priority Theme Brief, Dartmouth News
Inclusive Refugee-Hosting can Improve Local Development and Prevent Public Backlash (with Guy Grossman and Shuning Ge). World Development, Volume: 166, 106203 (June 2023).
Large arrivals of refugees raise concerns about potential tensions with host communities, particularly if refugees are viewed as an out-group competing for limited material resources and crowding out public services. To address these concerns, calls have increased to allocate humanitarian aid in ways that also benefit host communities. This study tests whether the increased presence of refugees, when coupled with humanitarian aid, improves public service delivery for host communities and dampens potential social conflict. We study this question in Uganda, one of the largest and most inclusive refugee-hosting countries. The data combines geospatial information on refugee settlements with original longitudinal data on primary and secondary schools, road density, health clinics, and health utilization. We report two key findings. First, even after the 2014 arrival of over 1 million South Sudanese refugees, host communities with greater refugee presence experienced substantial improvements in local development. Second, using public opinion data, we find no evidence that refugee presence has been associated with more negative (or positive) attitudes towards migrants or migration policy.

Effects of refugee presence in Uganda on public school access, health clinics, health utilization, road density, and a public goods index from 2001 to 2020.
Links: Paper | Supplementary Materials | Replication Materials | Slides | Media: World Bank Blog Post
From textbooks and articles to seminars and online resources, advice on how to successfully design and conduct randomized controlled trials (RCTs) abounds. We agonize over the research design, practitioner partnerships, and participant recruitment to name just a few concerns. But we rarely talk about those who conduct the field experiments—us. Even rarer is a discussion on how the researcher’s identity have methodological consequences, particularly when a researcher is from a background traditionally underrepresented
in academia. We the co-authors, all researchers of color, have found that our identity has posed additional challenges—with our expertise, objectivity, status doubted, occasionally followed by muted enthusiasm for participating in experiments. When researcher
identity defies the expectations of a typical profile of an academic affiliated with North American or European-based institutions (white and male, in particular), it has important implications for
the inferences we draw from field experiments. In this article, we reflect on these challenges and the potential biases that can arise from a researcher’s identity, highlighting our own experiences in the field.
Links: Paper
A large literature suggests that the presence of refugees is associated with greater risk of conflict. We argue that the positive effects of hosting refugees on local conditions have been overlooked. Using global data from 1990 to 2018 on locations of refugee communities and civil conflict at the subnational level, we find no evidence that hosting refugees increases the likelihood of new conflict, prolongs existing conflict, or raises the number of violent events or casualties. Furthermore, we explore conditions where provinces are likely to experience substantively large decreases in conflict risk due to increased development. Analysis examining nighttime lights as a measure of development, coupled with expert interviews, support our claim. To address the possibility of selection bias, we use placebo tests and matching. Our research challenges assertions that refugees are security risks. Instead, we show that in many cases, hosting refugees can encourage local development and even conflict reduction.

Map of refugee sites and average number of violent events during our study period 1990 - 2018. Negative effects of concentrated refugee presence on conflict outcomes.
Presented at ISA 2014, APSA 2015, Empirical Studies of Conflict (ESOC) 2019 Annual Meeting at UCSD.
Links: Paper | Supplementary Materials | Replication Materials | Slides | Media: Political Science Now, The Economist
How do major national events influence attitudes toward non-nationals? Recent research suggests that national sports team wins help foster national pride, weaken ethnic attachments, and build trust among conational out-group members. This paper asks a related question: By heightening nationalism, do these victories also affect attitudes towards foreign out-groups, specifically refugees? We examine this question using the 2019 Africa Cup football match between Kenya and Tanzania, which Kenya narrowly won, coupled with an online survey experiment conducted with a panel of 2,647 respondents recruited through Facebook. We find that winning increases national pride and preferences for resource allocation toward conationals, but it also leads to negative views of refugees’ contribution to the country’s diversity. However, we present experimental evidence that reframing national sports victories as a product of cooperation among diverse players and highlighting shared superordinate identities can offset these views and help foster positive attitudes toward refugees.

Effects of winning the Africa Cup match and receiving inclusion primes for Kenyans (vs. Tanzanians) on National Pride and Attitudes on Refugees bringing positive diversity.
Presented at MIT GOV/LAB and APSA 2020.
Links: Paper | Supplementary Materials |
Replication Materials | Pre-Analysis Plan
| Slides | Media: MIT News, UBC Q&A, UW PE Forum Podcast, The Conversation, Digital Fieldwork
Recent studies of efforts to increase citizen engagement in local governance through information campaigns report mixed results. We consider whether low levels of self-efficacy beliefs limit engagement, especially among poor citizens in poor countries. Citizens may be caught in an "efficacy trap" which limits their realization of better public goods provision. We describe results from a series of experimental studies conducted with over 2,200 citizens in rural Tanzania, in which we compare the effects of standard information campaigns with Validated Participation (VP), an intervention designed to socially validate citizens' participation. We implement a staged approach to experimental research, seeking to balance ethical and cost concerns about field experimentation. In our main analyses, we find that VP did not lead to increased levels of self-efficacy or more active citizen behaviors relative to standard informational treatments. Nonetheless, we find some promising evidence for VP in a follow-up qualitative study with teachers. We conclude by discussing lessons from this research and directions for future investigation of the possible role of self-efficacy traps in development.

Effects of Validated Participation vs. Information Only on efficacy (Kilosa study, N = 1,633 parents) and school outcomes 2 years later (Bukoba study, N = 24 schools).
Presented at WGAPE NYU Abu Dhabi 2016, APSA 2017, Yale ISPS Experiments Workshop 2018, Ideas & Evidence at Twaweza East Africa 2018, Harvard Experimental Political Science Graduate Student Conference 2018, African Studies Association 2019.
Links: Paper | Supplementary Materials | Replication Materials | Pre-Analysis Plan | Slides | Media: JEPS Blog
Governments, militaries, and aid agencies use economic interventions to influence wartime support for combatants. Yet credible evidence of whether these programs can shift support for governments and insurgents remains scarce. We experimentally evaluate a program of livelihood training and one-time unconditional cash transfers on combatant support among 2,597 at-risk youths in Kandahar, Afghanistan. Employing survey methodologies for sensitive questions, we find that training alone has little effect on combatant support. Cash has a boom-and-bust dynamic, initially increasing pro-government sentiment before reversing itself months later, leading to higher Taliban support. When combined with livelihood training, cash increased support for the Afghan government in the medium term. We interpret these results as consistent with a credit capture argument. While each intervention alone was a weak signal of government competency, the combination of training and cash provided participants with sufficient information to revise their beliefs about government performance and responsiveness.

Effects of training (TVET) and cash (UCT) on combatant support measured with endorsement and randomized response at endline and 7 months later.
Presented at the 11th Annual NYU CESS Conference on Experimental Political Science and MPSA 2018.
Links:
Paper | Supplementary Materials | Replication Materials | Pre-Analysis Plan | Slides | Media: Washington Post, Nature Human Behaviour
The randomized response method is a survey technique that seeks to reduce potential bias due to non-response and social desirability when asking questions about sensitive behaviors and beliefs. This survey methodology asks respondents to use a randomization device, such as a coin flip, whose outcome is unobserved by the enumerator. By introducing random noise, the method conceals individual responses and consequently protects respondent privacy. While numerous methodological advances have been made, we find surprisingly few applications of this promising methodology. In this paper, we address this gap by (1) reviewing standard designs available to applied researchers, (2) developing various multivariate regression techniques for substantive analyses, (3) proposing power analyses to help improve research designs, (4) presenting new robust designs that are based on less stringent assumptions than those of the standard designs, and (5) making all described methods available through open-source software. We illustrate some of these methods with an original survey about militant groups in Nigeria.

Power of four standard designs across probabilities of asking for the truth (x-axis) for N=500 and proportions of .1, .2, and .3 with the sensitive trait.
Presented at MPSA 2015 and the UNFPA Expert Group Meeting on Population Data for the 21st century.
Links: Paper | Replication Materials | Slides