Document Type

Discussion Paper

Date of Paper

Summer 6-7-2023


Disagreements over business deals, land boundaries, or loan non-repayment are very common sources of disputes, and courts are congested in developing countries. We evaluate the effects of the government introducing formal “village courts” (VCs) in rural Bangladesh using a randomized controlled trial with a pre-analysis plan. VCs were designed to improve dispute resolution and relieve the pressure on congested district courts. We find that VCs are more of a substitute for the ubiquitous informal dispute resolution mechanism (DRM) called shalish. VCs become functional and active in treated areas, but shalish remains the primary preferred DRM in both treatment and control areas. The elected leaders in charge of implementing VCs are also involved in settling shalish cases, and the potential of VCs is limited by the constraints on their time. There is some substitution from shalish to VC, but the district court congestion and downstream village economic activities, social dynamics, and political attitudes remain unaffected. The decentralized, diffuse shalish system is a better fit given the aggregate demand for dispute resolution, and it will not be possible for VCs to supplant this informal institution without additional investments in human resources to enhance VC capacity.


We thank Sarder Asaduzzaman, Nick Beresford, Sirazul Haq, Sudipto Mukherjee and the UNDP Bangladesh team for support, and Chris Blattman, Mark Rosenzweig for comments. We thank Mehrab Ali, Muhammad Bin Khalid, Sara G´omez-Mesa, John Mori, Proma Saha, and Mohammad Riad Uddin for excellent research assistance, as well as IPA Bangladesh for research support. This project was approved by Yale University IRB (Protocol ID 2000032185) and Innovations for Poverty Action IRB (Approval 13964). For financial support, we thank 3ie (grant number DPW1/1100).


A randomized-controlled-trial registry and pre-analysis plan are available at: