Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We conduct a field experiment in India comparing two ways of delegating selection of microcredit clients among smallholder farmers to local intermediaries: a private trader (TRAIL), versus a local–government appointee (GRAIL). Selected beneficiaries in both schemes were equally likely to take up and repay loans, and experienced similar increases in borrowing and farm output. However farm profits increased and unit costs of production decreased significantly only in TRAIL. While there is some evidence of superior selection by ability and landholding in TRAIL, the results are mainly driven by greater reduction of unit production costs for TRAIL treated farmers than GRAIL treated farmers of similar ability or landholding. We develop and test a model where the TRAIL agents’ role as middlemen in the agricultural supply chain enabled and motivated them to offer treated farmers business advice, which helped them lower unit costs.