Felipe Dizon is a PhD candidate in Agricultural and Resource Economics at the University of California, Davis.
In 2013 alone, donors pledged $31 billion to support financial inclusion. In the past, microcredit and insurance programs received all of the attention, but the focus has recently shifted towards improving the capacity of the poor to save. In 2010, for example, the Gates foundation pledged $500 million to support microsaving programs.
The impacts of microsaving programs seem to be worth the costs. For example, in Nepal, research has found that simply providing access to formal savings accounts increases welfare. Furthermore, driven by insights from behavioral economics and advances in digital technologies, the costs to microsaving interventions are decreasing. Commitment saving devices, such as those that have been tested in Kenya, Malawi, and the Philippines, are able to increase savings and improve welfare. Savings products are being integrated into mobile money platforms such as M-Pesa in Kenya and Tigo mobile money in Rwanda, thereby dramatically lowering transaction costs. Beyond improvements in individual welfare, positive spillovers of various microsaving programs on one’s social network have been documented in Nepal and in Malawi.
But, as with every program, broader welfare effects must be evaluated most carefully. In a study with vulnerable women in Kenya, we investigate whether negative spillovers of microsaving programs occur and whom these affect.
We randomly assigned participants to mobile money savings account treatments, which also included setting savings goals and receiving weekly SMS reminders on these goals. We find that access to this microsaving program reduced interpersonal transfers sent and received to cope with negative shocks, also known as risk-sharing. And, likely as a result of the combination of the reduction in risk-sharing and positive spillovers of savings that are documented in the literature, we find that the net effect of the microsaving program on the food security of risk-sharing connections is zero. Others have similarly found that an increase in bank access reduced risk-sharing in India.
This reduction in risk-sharing is consistent with theoretical findings wherein access to savings exacerbates moral hazard problems inherent in risk-sharing agreements. Informal risk-sharing agreements primarily require trust — I provide you with support when you experience shocks, and I trust that you will provide me with support when I in turn experience shocks. Trust makes up for the absence of written binding contracts. Access to savings, however, reduces this trust, because it increases the incentive to break one’s promise to provide support.
There is a silver lining. In our study, the reduction in risk-sharing occurs only between pairs where one individual had access to the program, but not both individuals. Ensuring universal access to a savings program might be the straightforward solution.
We must carefully consider who is excluded from a microsaving program. Commitment saving programs that target present-biased individuals exclude those who do not exhibit such time preferences. Mobile money based programs exclude those without access to mobile technologies. Microsaving programs can attempt to include as many people as possible. Or, social safety nets can be put in place to ensure that when microsaving programs exclude individuals, those excluded are protected through other means.
Similar to many other microsaving programs discussed above, the intervention in our study had direct benefits to program recipients, such as increased savings. By and large, the gains to microsaving programs are likely to outweigh the negative spillovers we uncovered. But we can do better by identifying possible welfare losses and minimizing such losses if they exist.