Research

I work on the political economy of conflict and development in sub-Saharan Africa and South Asia. My research asks whether, in conflict-affected contexts where state capacity is low, changes in incentives and institutions at the micro-level can (i) deliver direct material benefits to individuals and (ii) increase citizens’ ability to hold state and non-state actors accountable. I am especially interested in the relationship between individual behavior and legal institutions, both formal and informal, about which the developing country literature is largely silent. I approach these questions through a range of methods that enable causal inference, including natural and randomized experiments, lab-in-the-field studies, and formal modeling. I take a comparative perspective, drawing on original fieldwork conducted over the past five years in rural Liberia, Sierra Leone and Pakistan.

Forum shopping and legal aid in Liberia
with Justin Sandefur (CGD)

How do people choose between informal and formal institutions for resolving conflicts and legal disputes? We provide formal theoretical analysis and evidence from original survey data collected in Liberia that individuals make rational—albeit, severely constrained—choices between the costs and punitive norms of the dysfunctional formal legal system, and the demographic biases common in customary law. We then ask—can engendering greater competition between formal and informal institutions improve outcomes for those least able to afford formal justice? We show using a unique legal aid experiment that, when formal law is offered in customary, ‘remedial’ trappings, increased access can have significant material benefits for those most excluded—women, minorities, and the poor.

Methods: formal modeling, randomized controlled trial, panel survey [N=2,500]
Partners: Carter Center, Catholic Justice and Peace Commission, United States Institute of Peace
Funding: DFID, UN Peace Building Fund, IGC, OSF, IDLO
Related publications
Delivering justice to the poor: Theory and experimental evidence from Liberia

Reconciliation, conflict, and development in Sierra Leone
with Oeindrila Dube (NYU) and Jacobus Cilliers (Oxford)

The remedial justice offered by informal institutions can only be effective when parties are willing to reconcile—yet in conflict-affected settings, unresolved wartime grievances prevent reconciliation. We ask—can truth-telling and reconciliation improve local governance, social cohesion, and material welfare? Preliminary results from a community-based reconciliation experiment in Sierra Leone suggest that citizens of reconciled communities are happier with traditional dispute resolution mechanisms, express more pro-social attitudes, and participate more in civic life—despite exhibiting signs of being ‘retraumatized’ by the truth-telling process. We are currently extending this research to explore the longer-term impacts on reconciled individuals and the institutions they encounter.

Methods: randomized controlled trial, behavioral games, panel survey [N=2,500]
Partners: Fambul Tok
Funding: 3ie, JPAL
Related publications
“‘White Man’s Burden’? A field experiment on generosity and foreigner presence”
“Reconciliation, conflict, and development: Experimental evidence from Sierra Leone”

Police corruption and excessive detention in Sierra Leone
with Justin Sandefur (CGD) and Alaina Varvaloucas (Yale Law School)

If trust in informal institutions can be restored, what of the ‘informalized’ post-conflict state that is patrimonial and weakened by conflict, with little capacity to deliver essential public services? We ask—can citizen-led monitoring improve the functioning of state institutions?—and offer evidence in support from a detailed study of the criminal justice system in Sierra Leone. The study uses a novel experimental design to test different components of a legal aid intervention, and finds that legal aid reduces excessive detention, while simple observation alone improves police arresting behavior.

Methods: quasi-experiment, panel surveys [N=3,100]
Partners: Timap for Justice, Open Society Justice Initiative
Funding: DFID, IGC, OSF
Related publications:
“The economics of due process: Police monitoring versus legal aid in Sierra Leone”

Incentivizing health service delivery in Sierra Leone
with Oeindrila Dube (NYU), Johannes Haushofer (JPAL) and Vivek Maru (Namati)

If citizen-led monitoring can change the behavior of public officials, are ‘top-down’ monitoring-based incentives sufficient to engender institutional change, or is ‘bottom-up’ citizen involvement necessary? In partnership with the Government of Sierra Leone and the World Bank, we conduct a randomized controlled trial of two social accountability interventions in Sierra Leone’s public health sector that offer non-financial incentives to health workers, using scorecards based on citizen feedback and behavior. In the first, ‘top-down’, intervention, health clinics compete on scorecard criteria and receive non-financial awards for good performance. In the second, ‘bottom-up’ intervention, clinic staff and community members are motivated using scorecards to develop joint action plans to improve community health. Comparison of the two interventions will allow us some understanding of the importance of direct citizen involvement in monitoring. Results from this study are forthcoming.

Methods: randomized controlled trial, panel surveys [N=10,100]
Partners: Ministry of Health, Decentralization Secretariat, World Bank, IRC, Concern, Plan International
Funding: Govt of Sierra Leone, World Bank, USAID-DIV, Namati

Citizenship, security, and access to justice in northwest Pakistan
with Sarah Khan (Columbia)

This project extends my comparative research on legal institutions and state accountability to northwest Pakistan, on the border of Afghanistan, in districts severely affected by conflict, displacement and natural disaster. We ask questions unanswered by research thus far—(i) does citizen engagement with formal institutions increase trust in the state? (ii) does experience of conflict drive citizens away from the state and towards informal institutions? (iii) can citizen-led monitoring, aside from improving service delivery, also increase citizen demand for state services? We have begun to investigate these questions using a combination of ethnographic fieldwork, survey experiments, behavioral experiments, and a randomized controlled trial in partnership with the Sarhad Rural Support Program, the largest NGO in the region.

Methods: randomized controlled trial, behavioral games, survey experiments, panel surveys [N=2,100]
Partners: Sarhad Rural Support Program, Open Society Justice Initiative
Funding: OSF, IGC

Mapping and monitoring the Africa land grab
with Jeremy Weinstein (Stanford) and Darin Christensen (Stanford)

This project explores the impact of citizen monitoring of land concessions agreements in sub-Saharan Africa. We ask—can civic mobilization and citizen monitoring improve the agreements that communities sign with investors and/or improve investor compliance with those agreements and existing national regulations? The project involves, firstly, using high-resolution satellite imagery and remote sensing technology to map investor compliance with national regulations across sub-Saharan Africa. Secondly, we are currently working with Timap for Justice, the top legal aid NGO in Sierra Leone, to experimentally study an intervention intended to raise the bargaining power of dispossessed communities, through third-party mediation and training in non-violent civic activism.

Methods: remote sensing, randomized controlled trial, panel surveys
Partners: Timap for Justice, Sierra Leone’s National Mineral Agency, World Resources Institute
Funding: AFOSR