Current Research
Books in progress and active projects, organized by research domain. For published work, see Publications.
Political Violence
Aid in Conflict In Progress Book
Aid in Conflict tells the story of the international aid industry in conflict-affected settings. Over the past three decades, conflict-affected countries have become the largest recipients of foreign aid. Donors have pledged not only to give large amounts of aid to these areas but to do so in a way that builds peace and reduces violence. Despite this commitment, pathologies within the aid industry make it unlikely that aid will solve the challenge of violent conflict. Based on 165 elite interviews and subnational statistical analysis of aid allocation across more than seventy donors operating in the DRC, Nepal, South Sudan, and Sudan (1990–2015), the book shows that it is the structure of donors, not violence itself, that determines how donors respond to conflict. Within the same country context, different donors pursue radically different and often contradictory strategies, jockeying for influence often to the detriment of recipient countries.
Rebel Sophistication in Aid Targeting: Insights from LRA Inner Circles In Progress
Rebel groups can display remarkable sophistication during armed conflict that goes unrecognized because such tactics are difficult to observe and rarely documented at the operational level. We tackle a puzzle of variation in rebel targeting of foreign aid, illuminating the formulation and adaptation of associated tactics. We triangulate original primary data from interviews and focus groups with ex-commanders of the Lord’s Resistance Army (LRA), high-level leaders of the Ugandan Peoples’ Defence Forces (UPDF), elected government officials, and a survey of over 1,200 residents of areas affected by LRA activities. The LRA’s tactics tended to be most sophisticated when it had high strategic aspirations alongside facing high material demands. The LRA exploited violence against civilians to cause humanitarian needs, precipitating aid flows targetable consistent with its requirements. Our findings provide insight into the development of tactics in seemingly ragtag rebel groups.
Leaders, Factions, and Political Violence: Evidence from China's Cultural Revolution In Progress
China’s Cultural Revolution (1966–1976) ranks among the deadliest events of the twentieth century, yet it remains understudied in conflict research. A central puzzle is the widespread fragmentation of rebel groups at the local level. This paper conceptualizes rebel fragmentation as a bargaining failure driven by two forces. First, secret alliances with different factions in Beijing fueled mutual optimism, consistent with local power seizures that often drew counterclaims from rival patrons. Second, the military, tasked with “supporting the left” under vague and shifting directives, struggled to identify who qualified as leftists, leaving its commitment to enforce agreements in doubt. Using a novel monthly dataset of violence across 2,267 counties, we show that both power seizures and military intervention significantly increased the likelihood of rebel fragmentation. Moreover, fragmentation is associated with higher subsequent violence, consistent with the interpretation of fragmentation as bargaining failure.
Experimental Evidence on Ethnic Identification and Deception In Progress
Micro-level theories of ethnic politics commonly assume that individuals can identify the ethnic backgrounds of the people with whom they interact. While this assumption underlies a vast number of theories, we lack systematic evidence regarding under what conditions it is more or less justifiable. This study fills a gap by proposing a theoretical structure of social classification that enables us to examine ethnic classification and understand the variability in ethnic identifiability. Drawing on lab experiments conducted in ethnically diverse Uganda, South Africa, and the United States, ethnic identification errors provide the basis for constructing a new measure of group distinctness. We use this measure to examine the potential for passing through the strategic use of signals and find that successful passing is more likely when individuals attempt to pass into groups that are similar to their own but distinct from the group of the person trying to place them.
Illicit Finance
Bank of Guineastan: Research Synthesis In Progress
External validity is a central challenge in evaluating financial-compliance interventions: policymakers must know not only whether mechanisms work, but whether their effects generalize or transport across countries, financial institutions, legal standards, compliance outcomes, and time. We address this through a coordinated global audit/field experiment research program (2011–2022) comprising two overarching waves, four major studies, 11 study modules, and more than 50,000 email audit approaches to financial institutions in over 200 countries. We deliberately investigated core mechanisms across diverse settings, treatments, outcomes, units, and time. We introduce a new preregistered external-validity comparison: a structured evaluation of individual-study methods versus multi-study synthesis methods. Substantively, the design enables rare characterization of the external validity of core financial-compliance mechanisms; methodologically, it provides guidance for how to approach external validity assessment.
Using Field Experiments in International Business Research: Pricing Risk in Shell Corporations In Progress
Field experiments allow international business researchers to address critical questions that cannot be effectively studied by employing other methodologies such as surveys, non-experimental databases, or small sample cross-country/firm studies. We highlight the value of field experiments and conduct a study testing the price-sensitivity of firms offering incorporation services for creating shell companies. Drawing upon data collected from 782 incorporation firms in 148 countries, resulting in 1,046 responses, we test whether incorporation firms charge a price premium when asked to create an anonymous shell corporation, in violation of international law, by customers with varying risk profiles. Confounding bedrock conventional assumptions about businesses responding to higher risks by requiring higher rewards, we find that businesses supplying shell companies are remarkably price insensitive to even the most obvious risks of serious criminal intent by their customers.
International Development
Aid in Conflict In Progress Book
Aid in Conflict tells the story of the international aid industry in conflict-affected settings. Over the past three decades, conflict-affected countries have become the largest recipients of foreign aid. Donors have pledged not only to give large amounts of aid to these areas but to do so in a way that builds peace and reduces violence. Despite this commitment, pathologies within the aid industry make it unlikely that aid will solve the challenge of violent conflict. Based on 165 elite interviews and subnational statistical analysis of aid allocation across more than seventy donors operating in the DRC, Nepal, South Sudan, and Sudan (1990–2015), the book shows that it is the structure of donors, not violence itself, that determines how donors respond to conflict. Within the same country context, different donors pursue radically different and often contradictory strategies, jockeying for influence often to the detriment of recipient countries.
Do Aid Shocks Cause Conflict? Evidence from the Dismantlement of USAID In Progress
Summary TBA.
Short-Term Gains and Long-Term Losses in International Peacebuilding In Progress
International peacebuilding represents a sustained commitment of billions of dollars annually, yet surprisingly little systematic evidence exists on whether these investments reduce violence at scale. We introduce PeaDa (Peacebuilding Data), a novel cross-national dataset coding 167 peacebuilding projects across 56 African countries from 1995 to 2018. Merging PeaDa with battle fatality data from ACLED and UCDP, we employ difference-in-differences and generalized synthetic control methods. We find a central paradox: peacebuilding reduces battle fatalities by approximately 100–259 deaths annually in the short run (1–3 years), yet this effect attenuates substantially by year four and may reverse by year five or six. Strikingly, while peacebuilding dampens fatality counts, it is associated with a 30–60 event increase in violent incidents. This fatality-incident disconnect suggests a compositional shift toward more frequent but less lethal violence. Our findings challenge the presumption that peacebuilding effects are durable.
Employing Ex-Combatants: A Resume Experiment in Colombia Data Collection
Economic reintegration is a critical part of peacebuilding, helping former combatants transition into civilian life. Yet stigma from employers may block access to formal jobs, threatening the success of reintegration programs. We implement a résumé field experiment to test whether ex-combatants in Colombia face labor market discrimination, and whether signals of rehabilitation reduce bias. Partnering with the government’s Agency for Reincorporation and Normalization, we use real résumés from eight job-seeking ex-combatants. We randomly vary whether they disclose ex-combatant status, highlight education or reconciliation experience, signal work experience, or mention tax incentives. We track employer responses to identify which signals increase interest and under what conditions discrimination is most severe. This design provides unique causal evidence on employer behavior in post-conflict societies and will inform policies for how best to structure reintegration support.
When Technology Meets Necessity: Digital vs. Traditional Humanitarian Aid in Pakistan In Progress
Ex-Combatant Demobilization and Support for Peace: A Quasi-Experiment in Colombia In Progress
Methods, Ethics & Scientific Standards
External Validity for Social Inquiry Under Review Book
External validity is both everywhere and nowhere. It is everywhere because very few studies aim only to make inferences for one specific case or sample. External validity is also nowhere because the number of studies that make credible inferences beyond the cases or samples that they analyze is so small. Building on insights from the credibility revolution associated with an increased focus on causal inference, we demonstrate how researchers can generalize inferences from samples to populations with similar characteristics and make transportable inferences to new populations that often have different characteristics. Achieving these end goals necessitates better conceptualization, including more attention to the distinction between populations and samples, generalizability and transportability, and the dimensions of external validity: Mechanisms, Settings, Treatments, Outcomes, Units, and Time (M-STOUT). We develop three key criteria—Model Utility, Scope Plausibility, and Specification Credibility—each with concrete step-by-step components that deliver independent analytical value.
Wrong Is Right: How Experts Can Learn Better from Error In Progress Book
Most people, and especially most experts, feel they are right most of the time. But are we really that smart? Confirmation bias, motivated reasoning, hindsight bias, and other well-known cognitive distortions may be lulling people – including experts on the knowledge frontier – into a false sense of correctness. When generating new knowledge especially, experts likely err much more often than they know. The irony: we learn best and most deeply when we make prediction errors. This book advocates changing experts’ mindsets toward that of students curious about the social world and promotes methods of learning that enable more frequent and more transparent error detection and reporting. It is no shame to prove yourself wrong; it is an opportunity to learn. Based on the history of science and insights from psychology, it introduces a theory of knowledge generation founded on updating after error, reports evidence on what causes greater openness to error, and ends with encouragement toward deep curiosity about the limits of our current knowledge.
Research Map
An interactive map of my full research portfolio — publications, teaching, and policy work — organized by substantive domain. Click any sub-topic cluster to expand. Hover for details. Curved arcs show cross-domain connections.