Remittance Responses to Temporary Discounts: A Field Experiment among Central American Migrants

The PIs conducted a randomized controlled trial to measure the impact of a transaction fee discount on remittances from Salvadoran and Guatemalan migrant workers in the Washington, D.C. area. The PIs also included an additional element intended to measure the impact of remittances on education in a migrant's home country; however, the analysis is primarily focused on the impacts of the discount. After a baseline survey, the participants were given an envelope containing their random assignment to one of the control or treatment groups.

The intervention included two treatments. For the first, the Discount treatment, participants received a discount card for $3.01 off of the transaction fee for all remittances sent from that location for the next ten weeks. The discount was only applicable to remittances sent to the previously-designated primary remittance recipient (PRR). For the second, the Education Information treatment, participants received an information sheet about the benefits of secondary and tertiary education in their home country. The treatments were cross-randomized in a two-by-two experimental design, yielding a control group and three treatment groups: Education Information Only, Discount and Education, and Discount Only.

The PIs found that the Discount Only treatment had economically and statistically significant positive effects on the number of remittances and total amount remitted. These effects persisted for up to 20 weeks after the discount had ended, suggesting that participants were not inter-temporally substituting future remittances to take advantage of the discount. The Discount and Education treatment had similar but smaller effects, and only for remittances to the PRR were the effects statistically significant. The Education Information Only treatment had no statistically significant effect on any outcomes.

2012-12-15

2013-06-30

Use of the discount, Number of remittances sent, Total amount of remittances

Participants were recruited from existing customers of the partner money transfer company at five locations in the Washington, D.C. area. Potential individuals were approached after they had sent a remittance. Individuals were eligible if they had sent the remittance to their previously designated primary remittance recipient (PRR), if they had sent it through the partner company, and if they were from either El Salvador or Guatemala. Those that accepted were given a $5 credit usable in that location for purposes not related to remittances.

Treatment was randomly assigned to unique ID numbers in blocks of 32, stratified by the location at which they were to be handed out. ID numbers were printed on envelopes which contained the assigned treatment. The envelopes were given, in the order that they were received, to participants after they had completed the baseline survey. The experiment was double-blinded; project surveyors did not know which treatment participants were assigned to until after they had opened the envelope. The PIs used an OLS specification with indicator variables for each of the three treatment groups. Controls include stratification cell fixed-effects for each block of 32 envelopes.

Randomization done by computer in blocks of 32, stratified by location

Randomization at the individual level in blocks of 32, stratified by location

No

946 individuals

946 individuals

Control group: 232 individuals; Education information: 230 individuals; Discount only: 247 individuals; Discount and Information: 232 individuals

We study the impacts on remittances of offering migrants temporary discounts on remittance transaction fees. We randomly assigned migrants from El Salvador and Guatemala 10-week remittance transaction fee discounts, and assess impacts using administrative transaction data and a post-experiment survey. Temporary discounts lead to substantial increases in the number of transactions and total amount remitted during the discount period. Surprisingly, these increases persist up to 20 weeks after expiration of the discount. We find no evidence that the discounts cause migrants to shift remittances from other remittance channels, or to send remittances on behalf of other migrants. These findings are consistent with naïveté on the part of migrants regarding remittance recipients' reference-dependent preferences.

Ambler, Kate and Aycinena, Diego and Yang, Dean, Remittance Responses to Temporary Discounts: A Field Experiment Among Central American Migrants (September 2014). NBER Working Paper No. w20522. Available at SSRN: https://ssrn.com/abstract=2502684

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2502684