(A World Bank Research Committee funded project (RF-P125203-RESE-BBRSB)
February 2011 - August 2012
Harry Anthony Patrinos, Lead Economist, Human Development Network - Education, World Bank
Dean Karlan (USA), Yale University and Innovations for Poverty Action
Ricardo Godoy (USA), Brandeis University
Tomas Huanca Laura (Bolivia), Innovations for Poverty Action
Miguel Paredes (Bolivia), Innovations for Poverty Action
We evaluate the impact of the program of introducing financial services to newly monetized Indigenous native Amazonian communities in Bolivia. We use a randomized allocation of savings lock boxes to facilitate and increase awareness of saving among the Tsimane'.
In many indigenous communities throughout Latin America, traditional economies based on barter and reciprocity are rapidly monetizing. This is especially true in the Amazon basin, where the construction of new roads and encroachment by cattle ranchers and colonist farmers make it possible for native Amazonians to have increasing exposure and access to money. Many native Amazonians now earn wages working part-time for loggers, farmers, traders and ranchers, and it has become easier to sell products for cash in markets in nearby towns. Many governments have even introduced schools to remote villages, and wage-earning teachers along with them. In countries such as Bolivia, the government pays parents to send their children to primary school and pays the elderly social security benefits, and the benefits have reached many of the most remote native Amazonian communities. Thus a culture of saving and tools to help them accumulate lump sums of money for larger purchases and emergencies, however, do not exist in many of the communities. Instead of introducing these communities to debt and to the more advanced financial literacy that it requires, here we propose making simple tools available to help them save lump sums of money for useful purchases and emergencies.
Poor people have many occasions when they need more cash than they have on hand. They need lump sums of money to cope with emergencies, to cover life-cycle events, and to capitalize on investment and business opportunities. The poor use a wide variety of financial tools to manage their cash flow needs. Recent randomized evaluations of microfinance concluded that it is far from the panacea for ending poverty that many believed it to be. An evaluation of formal microsavings showed that there is a high demand for formal savings among the poor and that access to savings accounts has positive impacts on productive investment levels and expenditures for women and allows them to smooth income and cope with shocks. Several studies explore the way to provide savings products for the poor. One promising type of savings product is commitment savings. A randomized evaluation of commitment savings found that the product had a strong positive impact on savings for the group offered the product when compared to the randomly chosen control group that was not offered the product. Recently piloted savings lock boxes with or without programmed alarms among randomly selected clients of a commercial microfinance bank in La Paz. Preliminary evidence suggests both higher deposit frequencies and higher savings balances of those clients who were assigned the lock boxes (without alarm) compared to the control group (those assigned no lock box). Anthropologists have written little about saving among native Amazonians. As a result, our understanding of the topic is limited. The proposed experiment will allow us to assess the role of these countervailing forces toward the adoption of savings.
Methods and Data
To answer our research questions we use a randomized evaluation based on earlier pilot of ways to facilitate and increase awareness of saving among the Tsimane'. The study will include 1100 members of 50 communities. By randomly assigning households to either the treatment or the control groups, any difference between the groups that we observe can be attributed to the treatment itself. The project will rely on two surveys of the sample participating households across the selected communities. The initial survey (pre-intervention) will be launched within the first two months of the project, in conjunction with the beginning of the treatment.