Get by with a little help from your peers

Can your peers help you save more? For members of the informal sector, who don't have the benefit of direct deposit or formalized commitment mechanisms, peers may help.

I just finished reading the working paper, "Peers as a Savings Commitment Device: Evidence from a Field Experiment among Low-Income Micro-Entrepreneurs in Chile." It discusses results from a randomized field experiment that tests whether peers can act as a commitment device to help microfinance clients save more. The sample population for this experiment is a group of low-income entrepreneurs who are already clients of Fondo Esperanza (FE), a microfinance institution in Chile. Up until the study, FE only offered credit to its clients, per Chilean regulatory requirements.

The research team and FE connected with a local bank and then offered 196 groups of FE, roughly 2700 clients, a no frills savings account via FE at the bank. The research team randomly sub-divided the 196 groups into three categories. The first group received the option to open a no frills savings account, with a 0.3% real interest rate (average for Chile) (control group); the second received the option to open a no frills savings account with an accountability structure where peers monitor and act as a commitment device. Specifically the members, "have the option to publicly announce to the group their savings goal for the coming credit cycle (approximately 3 month), and their corresponding weekly savings goal. Subsequently, in each group meeting it will be discussed who complied with their own savings goal and who did not. Those who complied are asked to show a deposit slip to prove it and receive a sticker in a booklet. Those who collect enough stickers receive a diploma as a non-monetary award," (treatment #2). A third group of clients is offered the option to open a no frills savings account but the real interest rate is significantly higher than the normal account, 5.0% compared to 0.3% real interest rate (treatment #2). This second treatment was introduced to understand the effect of interest rate differentials, and to compare that impact with the impact of a peer commitment device.

The results (calculated via Intention to Treat), demonstrated that members of the peer-based commitment treatment make about three times more deposits and had an average account balance that was 65% higher than those in the control group. The effects were particularly profound for group members who has assessed themselves as better-than-average at reaching their goals; this group's average monthly balance was 2.9 times higher in the peer treatment than with a regular savings account.

As written by the authors, "These findings indicate that peer groups may be an important mechanism to help people overcome self-control problems, particularly in areas where formal commitment devices are not available, and that individuals benefit most from joining commitment-groups where their peers are slightly less apt than themselves at reaching the shared self-control objective."

They found that for the 5% real interest rate treatment group, the average account balance was 48% higher than those in the control group, but the difference was not statistically significant. They extrapolated from the findings of treatments #1 and #1 to state that the peer commitment device was equivalent to a 6.5% increase in interest rate.

I thought about this paper and its implications for the microfinance sector in India, where microfinance institutions do not typically offer savings. The self help group bank linkage programme, however might be a more relevant receptacle of learnings from this study, as savings precede borrowing. To my knowledge most SHGs have a prescribed amount that members must save in the initial months but after their first loan, it becomes a bit more arbitrary. I wonder how the dynamics of peer-commitment factor since they had had to save publicly at first.

Much of this is fairly intuitive to microfinance practitioners, who in effect leverage peer-commitment that is more formal than this experiment, for group repayment and group liability. Some research has shown that group repayment meetings, even when the loans are individual, still produces high repayment through peer-commitment and monitoring.

How do you think microfinance practitioners could leverage this research to improve practice?