Title: Decentralized Targeting of Agricultural Credit Programs: Private Agents or Local Governments?

Speaker: Sujata Visaria, Hong Kong University of Science and Technology

Time: November 30th, 2018 13:30–15:00

Venue: Conference Room 106B, Zhonghui Building (IESR, JNU College of Economics)

About the speaker:

Sujata Visaria is an Associate Professor in the Department of Economics at the Hong Kong University of Science and Technology. She has a PhD from Columbia University, and worked at Boston University for four years before moving to HKUST. Her research focuses on how the enforcement of credit contracts affects micro-level outcomes in developing countries, the problems that small farmers face in marketing agricultural produce, and explores alternative ways of microcredit beneficiary selection that targets productive borrowers. Sujata Visaria is a Faculty Affiliate at the HKUST Institute for Emerging Market Studies, the Bureau for Research and Economic Analysis of Development (BREAD), and the Small and Medium Enterprises (SME) Initiative, Innovations for Poverty Action (IPA). She is also a member of the Board of Directors of the Asian Migrants Credit Union. Sujata Visaria's work has appeared in such journals as Journal of Development Economics, American Economic Review, Econometrica, American Economic Journal: Applied Economics, and others. 

Abstract:

This paper reports results from a field experiment conducted in rural West Bengal, India that compares alternative approaches to delegate the selection of beneficiaries for an agricultural credit program. In both approaches, a local agent recommends borrowers for individual liability loans, and is incentivized by commissions that depend on repayments. In one approach (TRAIL) the agent is chosen randomly from local traders; in the other (GRAIL) the agent is appointed by the local government. We find that TRAIL loans had higher take-up rates, and created significantly larger impacts on production of high-value cash crops and farm incomes. Loans in both schemes had similar repayment rates of 95%. The TRAIL agent recommended more productive farmers than the GRAIL agent did. More importantly however, the two agents also engaged differently with treated farmers. In our theoretical model, the TRAIL agent's private role as a middleman causes him to interact more intensively with more able farmers and increase their productivity. Instead due to his political motives, the GRAIL agent interacts more intensively with the less able farmers to lower their risk of default. In turn this lowers their productivity. The data support the predictions of this model.