Financing Smallholder Agriculture : An Experiment with Agent-Intermediated Microloans in India

Recent evaluations of traditional microfinance loans have found no significant impacts on borrower incomes or productive activities. We examine whether this can be remedied by (a) modifying loan features to facilitate financing of working capital needs of farmers, and (b) delegating selection of borrowers for individual liability loans to local trader-lender agents incentivized by repayment-based commissions. We conduct a field experiment in West Bengal where this design (called TRAIL) was offered in randomly selected villages. In remaining villages a more traditional design (called GBL) was offered, wherein five-member groups applied for joint liability loans with terms otherwise similar to TRAIL loans. TRAIL loans increased cultivation of potatoes (the major cash crop in the region) and farm incomes by 17-21%, whereas GBL loans had insignificant and highly dispersed effects. We argue this was because TRAIL agents selected borrowers that were low-risk and highly productive, whereas the GBL scheme attracted farmers that were riskier on average and highly heterogeneous in terms of productivity. TRAIL loans also achieved higher repayment and take-up rates, and lower administrative costs