Digital technology is rapidly emerging as integral to the future of development policy, and particularly to financial services. Recent years have seen rapid digitization in developing countries, from mobile money platforms to biometric smart cards. These innovations are largely perceived as beneficial, saving users on both sides of the transaction effort and time, and often enabling better monitoring. The use of mobile money to streamline payments is only becoming more commonplace. Despite this, relatively little is known about the effects of switching to mobile money. We propose to undertake a randomized control trial to evaluate the effects of digital repayment for microfinance, looking both at barriers to take-up and the effects of digital repayment on the integrity of the microfinance joint liability model.
There are two main channels through which digital repayment might negatively affect repayment behavior and default rates. First, digital repayment will result in reduced observability of loan repayments by other group members. Second, digital repayment will make meeting attendance less mandatory, which might decrease the social cohesion of microfinance groups. We assess both channels by experimentally manipulating whether microfinance clients in some groups repay digitally, and observe repayment performance, the quality of observation of the loan performance of other group members, and social interactions by group members.
We also randomly assign the selection mechanism, creating three groups: a control group that still repays with cash, an individual choice group where individual borrowers are each allowed to opt in or out of digital repayment, and a digital repayment group where all borrowers in that group are asked to repay digitally. This randomization also allows us the unique opportunity to assess the impact of digital repayment on those who would not choose it willingly, allowing insight into an important dimension of impact heterogeneity. We believe this would be the first study to identify the causal effects of digital repayment on a population which would not have chosen it, yielding crucial insight into the effects of increasing levels of digital repayment penetration.
We also experimentally test whether giving participants a receipt which explicitly confirms their payment to BRAC microfinance on a given date increases use of and confidence in the digital repayment system. This part of the experiment is designed to test whether adding a simple but costly confirmation receipt can help overcome knowledge and confidence barriers that inhibit take up of digital finance.