Doctoral Dissertation Research in Economics: Debit Cards, Cash Transfers, and Savings: Evidence from Mexico

Tulane University, New Orleans LA

Investigators

Abstract

A significant number of adults in the US and outside lack access to formal financial services such as savings accounts. When available, savings products are used to decrease income volatility, accumulate money for business investments, pay for unexpected health emergencies, and invest in children's education. Saving outside of formal bank accounts can be difficult: cash in hand or under the mattress is more easily spent. Hence, devising means to improve savings can lead to better living standards and economic outcomes Americans and the poor in developing countries. After expanding its large cash transfer program Oportunidades to urban areas, the Mexican government opened bank accounts for urban recipients of the program and paid their benefits directly into these accounts. In 2009, the government began tying debit cards to these bank accounts, which are run by a government bank. This doctoral dissertation project exploits this unique data collection opportunity by using the gradual rollout of debit cards in urban localities in Mexico to identify their causal impact on savings among program recipients. It will use data from a number of sources, including bank account balances, deposits, and withdrawals in program recipients' bank accounts over three years, as well as household survey data on their consumption, income, and assets over seven years. A theoretical model predicts that, by enabling program recipients to access their accounts at the nearest ATM of any bank rather than the nearest government bank branch, debit cards will increase savings through two channels. First, debit cards reduce the non-monetary transaction cost of accessing savings, making saving more attractive. Second, they provide a technology to build trust in the bank, which is initially low among Mexico's poor: while saving, a recipient can repeatedly check her balance at any ATM to make sure the bank hasn't stolen her money. Furthermore, these effects should be greater for those living far from government bank branches but sufficiently close to other banks' ATMs, and this distance effect should be greater for those who have more difficulty saving at home (for example, because the husband takes and spends the saved cash, which is always given to women). The distances between each family and the nearest ATM and nearest government bank branch are key variables to test these hypotheses. Combining Census block codes for families in my panel data with Census block shape files and the geographic coordinates of ATMs and government bank branches, I will estimate the distance from each household to the nearest ATM and government bank branch, and use variation in these variables and the rollout of debit cards to test the model's predictions. The results from this study will provide important insights for the American government and other countries seeking to increase financial inclusion.