#econtwitter Happy to share that our paper “Psychological and social motivations in microfinance contracts: theory and evidence” (with @Sanjit Dhami and Ali al-Nowaihi) is now out in the Journal of Development Economics. Open-access link: doi.org/10.1016/j.jdev… 1/10

Two key features of the Grameen-I microfinance contract were joint liability (JL) and public repayment of loans. JL was dropped in the Grameen-II contract, but public repayment was retained. Both contracts delivered high and comparable repayment rates. Why? 2/10

We use the framework of psychological game theory to model a two-period microfinance game, and propose guilt aversion in a JL contract & shame aversion under public repayment of loans as the main psychological drivers of effort choice. 3/10

We conduct a 2 x 2 lab-in-the-field experiment in Pakistan with 400 microfinance borrowers. Along one dimension, we vary the liability structure (IL or JL) and along the other, the method of repayment, private-individual basis (I) or public in a group (P). Four contracts 4/10

Results: Private repayment: conditional on receiving a private signal about the partner’s first-order positive beliefs, the average effort level in the JL contract almost doubled relative to IL,confirming the role of guilt aversion. 5/10

Public repayment: conditional on receiving a public signal about the social group’s first-order normative beliefs, the effort levels are comparable under IL & JL contracts,the liability structure does not matter. Shame aversion is adequate to discipline borrower behaviour. 6/10

Under IL, public repayment relative to private repayment increases effort by 60%, again confirming our shame-aversion hypothesis. Under JL, the private versus public repayment contrast shows that shame trumps guilt in explaining borrowers’ effort choices. 7/10

Our results have two implications for contractual choices by banks. (1) Under public repayment, given the more restrictive borrowing conditions in the JL contract, the bank may prefer the IL contract; shame aversion suffices in this case. 8/n

(2) Under private repayments, the bank may prefer the JL contract to the IL contract due to the higher repayment rate in the first period and the higher take-up of loans in the second period in the JL contract; guilt aversion is effective in this case. 9/10

2nd-period effort decisions do not show the predicted end-game effects. Avg. 2nd-period effort levels are not lower than the avg. 1st-period effort levels in the JLI, ILP and JLP contracts. This inconsistency with the std optimization approach suggests a form of anchoring. 10/10