Impact of bank accounts on migrant savings and remittances: Evidence from a field experiment

I design a randomized field experiment specifically to identify the causal effect of having a U.S. bank account on migrants' saving and remittance behavior. I find that migrants in the treatment group were 81 percent more likely to get a matricula consular and 38 percent more likely to open a U.S. bank account relative to the control group. Also, total savings rose and remittances fell as a fraction of income. The composition of savings changed as well, with the increase in savings staying in the U.S. The changes in savings and remittance behavior are especially pronounced for migrants who reported in the baseline interview that they did not have control over how remittances are spent in Mexico. Migrants in the treatment group who reported not-having control over how remittances are spent in Mexico were 61.5 percentage points more likely to open a U.S. savings account, earned ;These findings suggest the following. First, the unitary model of the household does not apply to these migrant households. If it did, then merely getting a U.S. savings account should not have changed the allocation of migrant earnings among remittances, Mexican savings and U.S. savings. Second, that income increases due to the U.S. bank account, but only among subjects in the treatment group who report a lack of control over how remittances are spent in Mexico, suggests that the household is not efficient. Finally, that there is an impact on savings even among subjects in the treatment group reporting control suggests that banking the unbanked population can raise savings. This study expands our knowledge of how banking the unbanked affects savings and remittances, and additionally has important implications for how to model decision-making in migrant households.