Many beneficiaries of social welfare programs around the world receive benefits in cash or by check. Can distributing welfare benefits through electronic transfers directly into bank accounts help low-income individuals enter the formal financial sector? We evaluate how transitioning a social welfare program from cash distribution to electronic transfers impacts recipients’ access to their funds, as well as their savings, debt and subjective well-being.
A., Claudia, Abhijit Banerjee and Esteban Puentes. 2017. "Electronic Transfers of Public Subsidies to Bank Accounts (Chile Cuenta)." AEA RCT Registry. April 27. https://doi.org/10.1257/rct.2081-1.0.
We evaluate a financial inclusion intervention in Chile. The intervention consist in offering an electronic payment of subsides to families of the main anti-poverty program in Chile (Programa Puente). We use bank accounts of BancoEstado (the only estate-owned commercial bank in the country) and compare this offering with the usual form of payment which is done through checks.
Intervention Start Date
2012-11-01
Intervention End Date
2013-10-31
Primary Outcomes (end points)
Savings, Debt, Consumption
Primary Outcomes (explanation)
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
The Chilean Ministry of Social Development, which runs the Puente social welfare program, is working with Banco del Estado, a state-owned but independently run bank, to electronically deposit social welfare payments into beneficiaries’ bank accounts. As the program rolled out, researchers randomly selected 3,232 Puente beneficiaries in Santiago to participate in the study. Three quarters of these beneficiaries were randomly selected to receive an offer for a bank account and direct deposit of their welfare payments into the account. The remaining beneficiaries continued to receive their payments by check, serving as a comparison group.
Experimental Design Details
Not available
Randomization Method
The randomization was made at an individual level. The stratification variables were (1) age of the beneficiary, (2) score in the social security card, (3) time they have been participating in the program and (4) municipality.
Randomization Unit
Individual
Was the treatment clustered?
No
Sample size: planned number of clusters
3,232 families
Sample size: planned number of observations
3,232 families
Sample size (or number of clusters) by treatment arms
810 control; 2,422 treatment (total 3,232)
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)