Inequality and Violence
State Capacity and Gender Inequality: Experimental evidence from Papua New Guinea.
Job market paper.
How does expanding state capacity change power inequalities between men and women? I design a field experiment in a remote area of post-conflict Papua New Guinea that randomly assigns permanent police officer presence to 17 villages and not to 22 others. After the police have been working in treatment villages for over eight months, I measure outcomes in interviews with 1,383 respondents, and broaden the findings through analysis of over 30,000 historical case records. Whereas many expect the state to supplant non-state institutions, I find the opposite can occur. As police enforce a more equal rule of law, and empower women, men seek to preserve their advantage by increasing their reliance on local chiefs. Previous accounts overlook such strategic complementarities because they assume use of alternative resolution mechanisms is zero-sum. The findings show how forum-shopping can limit the ability of the state to reduce power inequalities between men and women.
Countering violence against women by encouraging disclosure: A mass media experiment in rural Uganda.
With Donald P. Green and Anna Wilke.
Conditionally accepted at Comparative Political Studies.
Violence against women (VAW) is widespread in East Africa, with almost half of married women experiencing physical abuse. Those seeking to address this issue confront two challenges: some forms of domestic violence are widely condoned, and it is the norm for witnesses to not report incidents. Building on a growing literature showing that education-entertainment can change norms and behaviors, we present experimental evidence from a media campaign attended by over 10,000 Ugandans in 112 rural villages. In randomly assigned villages, video dramatizations discouraged VAW and encouraged reporting. Results from interviews conducted several months after the intervention show no change in attitudes condoning VAW yet a substantial increase in willingness to report to authorities, especially among women, and a decline in the share of women who experienced violence. The theoretical implication is that interventions that affect disclosure norms may reduce socially harmful behavior even if they do not reduce its acceptability.
Police and Corruption
Political Corruption Cycles in Democracies and Autocracies: Evidence from micro-data on extortion in West Africa.
Revise and resubmit, QJPS.
Using two large cross-national micro datasets on extortion and commodity flows, I provide evidence of corruption cycles around elections in five West African states. In democracies but not in autocracies, police and other officials extort bribes that are 27% higher in the buildup to elections. These cycles occur on the intensive margin—the price at which bribes are set—rather than on the extensive margin—the total number of agents extorting. When incumbents lose, prices remain abnormally high. They only return to normal levels if incumbents win. I find no evidence of political cycles in the composition, quantity, or direction of commodity flows. This pattern of results lends little support to the claim that new democracies have higher corruption than autocracies because politicians use extortion for illicit campaign fundraising. Traditional political business cycles also do not appear to explain corruption cycles in this context. Rather, I argue that corruption cycles may result from independent decision-making by bureaucrats who need to insure against the uncertainty of future leadership.
How robust is institutionalized corruption? A field experiment on extortion along West African highways.
Are dramatic policy interventions necessary to disrupt forms of corruption sustained through mutually reinforcing expectations, or can small shocks to uncertainty destabilize corrupt equilibria? I attempt to answer this question by randomly introducing the presence of a highly unexpected foreigner into interactions in which officials typically extort citizens for bribes according to a stable going rate. Specifically, I rode with truck drivers along 1,500km of highways in Burkina Faso, Ghana, Togo and Benin. At some of the 123 checkpoints manned by customs, police and gendarmerie, I openly observed attempts to extort bribes from drivers, at others I hid from sight. In contradiction to the expectations of well-informed experts, I find statistically significant evidence against large treatment effects. The findings suggest established patterns of corrupt behavior can be resilient to small shocks to uncertainty, and that experts may overestimate the effectiveness of novel anti-corruption measures.
Declaring and Diagnosing Research Designs.
With Graeme Blair, Alex Coppock, and Macartan Humphreys.
American Political Science Review, 113(3): 838—859, 2019.
Researchers need to select high-quality research designs and communicate those designs clearly to readers. Both tasks are difficult. We provide a framework for formally “declaring” the analytically relevant features of a research design in a demonstrably complete manner, with applications to qualitative, quantitative, and mixed methods research. The approach to design declaration we describe requires defining a model of the world (M), an inquiry (I), a data strategy (D), and an answer strategy (A). Declaration of these features in code provides sufficient information for researchers and readers to use Monte Carlo techniques to diagnose properties such as power, bias, accuracy of qualitative causal inferences, and other “diagnosands.” Ex ante declarations can be used to improve designs and facilitate preregistration, analysis, and reconciliation of intended and actual analyses. Ex post declarations are useful for describing, sharing, reanalyzing, and critiquing existing designs. We provide open-source software, DeclareDesign, to implement the proposed approach.
The Broad Impact of a Narrow Conflict: How Natural Resource Windfalls Shape Policy and Politics.
With Sung Eun Kim and Johannes Urpelainen.
Published in JOP.
Access to natural resources creates a political conflict between the expected economic winners and their environmental opponents, but the effects of such conflict on policy and politics remain unclear. To examine the scope of such effects, we exploit the rapid and unanticipated technological breakthroughs in the “fracking” of shale gas. During the past decade, shale gas production around the Marcellus Shale formation in the Northeastern United States expanded rapidly. Using a quasi-experimental design, we examine how access to shale gas in electoral districts changed the voting record of House Representatives on environmental policy relative to neighboring districts without access. Votes become 15–20 percentage points less likely to be in favor of the environment. The best explanation for this effect is the strong electoral performance of (anti-environmental) Republicans in shale-affected districts. The narrow conflict has a broad impact: access to natural resources puts downward pressure on environmental policy across the board.
A Placebo Design to Detect Spillovers from an Education-Entertainment Experiment in Uganda
With Anna Wilke and Donald P. Green.
Revise and resubmit at Journal of the Royal Statistical Society: Series A.
Education-entertainment refers to dramatizations designed to convey information and change attitudes. Buoyed by observational studies suggesting that education-entertainment strongly influences beliefs, attitudes, and behaviors, scholars have recently assessed education-entertainment using rigorous experimental designs in field settings. Studies conducted in developing countries have repeatedly shown the effectiveness of radio and film dramatizations on outcomes ranging from health to voting. One important gap in the literature is estimation of social spillover effects from those exposed to the dramatizations to others in the audience members' social network. In theory, the social diffusion of media effects could greatly amplify their policy impact. The current study uses a novel placebo-controlled design that gauges both the direct effects of the treatment on audience members as well as the indirect effects of the treatment on others in their family and in the community. We implement this design in two large cluster-randomized experiments set in rural Uganda using video dramatizations on the topics of violence against women, teacher absenteeism, and abortion stigma. We find several instances of sizable and highly significant direct effects on the attitudes of audience members, but we find little evidence that these effects diffused to others in the villages where the videos were aired.