Mobile banking and remittances among garment workers in Bangladesh: An experimental study

• Mobile money has been shown to have a positive impact on household poverty reduction. However, little experimental evidence exists on the impacts of mobile money on household financial and risk management.
• We used a randomised experiment to study the adoption, impacts, and perceptions of mobile money in Bangladesh.
• We found that mobile money adoption has positive impacts on health and education outcomes in rural households, and positive impacts on savings.
• We also found evidence for the importance of peer or family influence on adoption decisions, particularly for women, and evidence that households may make different decisions when using mobile money, but that this operates differently depending on the context.

Rapidly adopted in most developing country contexts, mobile technologies have the potential to serve as a broad-distribution platform for other services and products. The extent of their welfare implications remains unknown, and our study aimed to provide information to stakeholders about the possible benefits of adoption for poor and migrant households, as well as ways of boosting adoption among these households.

We found that adoption decisions are significantly influenced by peer effects and pro-social messaging. In particular, we randomised whether migrants received pro-social messaging and whether they made adoption decisions as “second-movers” with respect to their households. We found that pro-social messaging had significant effects on bKash adoption rates. We also found significant effects of being a “second-mover” for women.

Second, we found evidence for positive impacts of bKash use on health outcomes and educational outcomes for children, and also decreases in loan-taking, in rural households. However, the effects for migrants were more mixed. bKash use is marginally significantly associated with greater retention of formal employment but with decreases in health status among migrants.

Finally, we found evidence that in the urban sample, individuals demonstrate a greater willingness-to-pay for products when primed to think about mobile money. By contrast, in the rural sample individuals demonstrated a reduced willingness-to-pay when primed to think about mobile money.