The medium-term effects of gaining or losing a cash transfer on poverty

Andrea Molina-Vera and Hessel Oosterbeek

This paper uses a regression discontinuity at the eligibility threshold to estimate the effects of receiving a monthly (unconditional) cash transfer during a period of five years on poverty. We exploit a change in eligibility criteria that took place in Ecuador in 2009. This allows us to separately estimate effects for women who received the transfer before and lost it, and for women who did not receive the transfer before and gained it. The 40th percentile of the wealth index is the program's eligibility threshold; our results therefore apply to the richest of the poor. We find that receipt of the cash transfer has a significantly negative effect on the score of the 2014 wealth index. Women who gained the transfer in 2009 are worse off than women who continued not getting it, and women who kept the transfer are worse off than women who lost it. Effect sizes are largest for women who were young or unmarried at baseline. Further results suggest that part of the effects are due to recipients being less likely to be married and therefore being less likely to have a spouse who works. This points to a potential trade-off between poverty alleviation and women being independent from their spouses. We find no effects on female labor supply or fertility.

I want to read this! (PDF)