working papers
Expression at the Edge: Free Speech Boundaries Amidst the Gaza Crisis (with Yphtach Lelkes, Ran Abramitzky, Tamar Mitts, and Hani Mansour).
Abstract
This study examines how college students navigate the tension between free speech and harm prevention, highlighted by recent campus protests around the war in Gaza. Using online survey experiments with 3,065 college students nationwide, we find that the severity of speech and the target’s identity strongly influence support for disciplinary actions in response to objectionable speech. Students generally oppose punishing objectionable speech unless it is deemed highly harmful. Hateful rhetoric targeting minority groups, such as Black, Jewish, Muslim, and transgender individuals, elicits stronger punitive responses than identical statements directed at White students. While students, on average, afford greater protections to minority groups, there is notable variation. Exploratory analysis reveals that students’ responses are shaped by normative principles: about two-thirds believe minority groups should receive greater protection from harmful speech, while one-third advocate universal, equal treatment regardless of the target’s identity. These principles predict responses to speech scenarios, beyond ideology, stance on the Israeli-Palestinian conflict, and other personal characteristics. However, commitment to these principles weakens when individuals have a strong stance on the topic. These findings shed light on how college students balance competing principles of fairness and harm prevention in polarized contexts, offering insights into contemporary campus debates about free speech and inclusion.Informal Connections Outweigh Co-authorship Ties in Academic Impact. (with Danús Lluís, William Dinneen, Carolina Torreblanca & Sandra González-Bailón).
Abstract
Research has documented the importance of teamwork in the form of co-authorship for research productivity and innovation, but we know much less about how informal collaborations relate to academic success. Informal ties allow intangible exchanges like mentoring, guidance, and feedback to flow among scholars: these interactions weave a support structure that improves ideas and encourages project growth. However, these informal exchanges are more difficult to measure because they do not leave as clear a trail as co-authorship ties. Here, we uncover this layer of informal communication around scholarly outputs by parsing the information contained in the acknowledgment sections of published articles. Our data include 130,000 articles authored by 86,000 scholars from the period 2003-2023. We analyze scholars’ embeddedness in this informal structure of collaboration and reveal that (1) informal ties create a larger and denser network of support than co-authorship ties; (2) disconnection from informal networks is associated with gaps in productivity and impact; and (3) informal ties are a more relevant predictor of academic success than formal collaborations, even after matching for gender, seniority, methodology, and geographical location. Using coarsened exact matching and random forest regressions we show that informal structures of support are significantly associated with academic impact, creating gaps in who benefits from those connections.The Evolving Landscape of Political Science: Two Decades of Scholarship in a Growing Discipline. (with William Dinneen and Carolina Torreblanca).
Abstract
This study analyzes the evolution of political science (PS) scholarship using 140,000+ articles from 174 journals (2003–2023). We examine how structural changes—shrinking job markets and increased reliance on publication metrics—affect what gets studied and how. Growing publication pressures push younger scholars to publish more, yet the tripling of PS publication volume stems from a larger contributor base, not individual output. On a positive note, structural shifts have made PS more collaborative, with efficiency gains from team research outweighing credit diffusion. Contrary to fears of topical narrowing, our text-as-data analysis shows consistent topical diversity, though higher-ranked journals form a distinct ecosystem with unique methodological references. We also identify a key tradeoff: topically novel work tends to earn more citations over time but faces hurdles in prestigious journal publication. These findings enhance our understanding of how academic production systems shape the nature of intellectual progress in the discipline.Ethical Oversight in Impact Evaluations: External Advisory Committees to Assess Programming Risks. (with Darin Christensen, Allison N. Grossman, Jon Kurtz, Jeremey Weinstein, and Jessica Wolff).
Abstract
Social scientists not only conduct impact evaluations but also participate in the design and implementation of the programs being evaluated. While Institutional Review Boards (IRBs) oversee research activities, they do not assess risks posed by the interventions themselves. We propose establishing External Advisory Committees (EACs) to provide independent, expert oversight of programming risks. EACs complement IRBs by focusing on potential harms to participants and communities, offering dynamic risk assessments, and advising on program adaptations or termination. By providing impartial expertise, EACs help address potential conflicts of interest that may arise when researchers and implementers are invested in a program’s continuation. We illustrate the value of EACs through our experience implementing a cross-border labor migration program in Niger. Our EAC provided crucial guidance on scaling up the intervention after a pilot study and adapting the program following an unexpected military coup. While EACs introduce additional costs and may limit researcher autonomy, they generate accountability and are particularly valuable for novel and politically sensitive interventions in fragile environments.Liberalizing Refugee Hosting Policies without Losing the Vote. (with Zhou, Yang-Yang, Naijia Liu, & Shuning Ge).
Abstract
Inclusive refugee policies -- granting refugees the right to work, use public services, and move freely -- benefit both refugees and host countries' economies. Yet many governments hesitate to liberalize such policies, fearing electoral backlash. Can governments avoid backlash by pairing refugee rights expansions with policies that reduce burdens on host communities? We examine this question in Uganda, Africa's largest refugee-hosting country. Alongside refugee policy liberalization, Uganda mandated reallocating a share of refugee aid to communities near refugee centers. Combining refugee settlement data with election returns (2001-2021), we show that the incumbent government faced no backlash, but only after implementing these compensation schemes. Using public goods data, public opinion surveys, newspapers data, and parliamentary speech records, we find that investments in hosting communities and the reluctance of opposition parties to rally against popular policies account for our 'no backlash' findings.Calculation and Conscience: Motivations for the Substantive Representation of Ethnic Minorities. (with Apurav Bhatiya, William Dinneen, Stephanie Zonszein).
Abstract
A vast body of work shows that minority legislators are more likely to represent their group's interests compared to dominant group legislators. However, it is unclear whether this is due to intrinsic motivations or electoral incentives. We use a regression discontinuity design (RDD) to analyze ethnic minority representation in the UK Parliament. By comparing white MPs who narrowly beat minority candidates to minority MPs who narrowly beat white candidates, the RDD controls for electoral incentives since it holds constant constituency factors correlated with a minority parliamentary win. Analyzing over 1 million parliamentary questions and speeches, we find that minority MPs are more likely than white MPs to discuss issues important to ethnic minorities. Additional evidence supports that narrowly elected MPs face similar electoral incentives, and that minority MPs representing minorities face reelection penalties. Our findings suggest that minority substantive representation is driven at least in part by intrinsic motivations.The Effects of Cellphone Coverage Expansion on Wealth and Political Behavior. (with Katrina Kosec, Shuning Ge, Apoorva Lal, and Benjamin Laughlin).
Abstract
Taking advantage of Ghana’s gradual extension of cellphone towers in the early 2000s, we analyze the wealth effects of cellphone coverage expansion in a developing country setting using a difference-in-differences (event study)research design. We proxy local wealth using night-time light density over 1996–2016 and an asset ownership-based index from the 2000 and 2010 censuses. We find that cellphone coverage expansion significantly raised wealth in Ghana. We then explore possible downstream effects of cellphone coverage expansion on electoral outcomes. We find no evidence that better-off citizens reward incumbents, either in presidential or parliamentary elections. Using Afrobarometer survey data, this null finding appears to be be-cause citizens do not give the government credit for economic improvements that are due to decisions made by private telecommunications companies. Further, increases in cellphone coverage significantly decrease vote-buying, which may be due to voters being harder to buy off when they are better off.Can Community Policing Improve Police - Community Relations in an Authoritarian Regime? (with Rob Blair, Anna Wilke).