Abstract

Joint-liability is maybe the most distinctive feature of microfinance contracts in developing countries. Yet, very little evidence exists on the impact of joint-liability contracts as compared to individual lending contracts. On the one hand, theory claims that joint-liability plays a key role in mitigating agency problems and thus enhancing repayment rates, especially when borrowers lack collateral. On the other, experimental evidence has shown mixed, sometimes contradictory results, highlighting major pitfalls like harsh social sanctions and peer pressure. We contribute to the debate on the relative merits (and weaknesses) of joint-liability by showing that, under certain conditions, joint-liability may not be able to solve adverse selection problems. We build a model in which a risk-neutral lender offers both individually and jointly-liable contracts, but has limited funds and limited knowledge about borrowers' quality. In this case, joint-repayments can be used as a signalling device of borrowers' type. Our model shows that if the lender allows for competition among borrowers, risky borrowers may have an incentive to reveal a higher joint-repayment level than safe borrowers. In other words, joint-liability may increase adverse selection. We test our predictions in a small experimental environment.