A Randomized Impact Evaluation of the Introduction of Mobile Banking in Mozambique

The recent introduction of mobile banking services in rural areas of Mozambique created the opportunity to undertake a randomized impact evaluation of the introduction of this technology in rural and traditionally unbanked areas. Aware of the importance of urban-rural remittances to previous successful cases in similar contexts (namely M-PESA in Kenya), this project focuses on assessing the impact of the mobile banking technology on the level of urban-rural transfers within households and consequent household economic behaviour. Our work focuses in the rural areas of the Gaza, Inhambane and Maputo southern provinces of Mozambique. The data collected so far presents evidence supportive of a great potential for the adoption of mKesh (the only existing Mozambican mobile banking service to date) by the rural populations. Indeed, there is a large incidence of rural-urban migration from the southern Mozambican provinces to Maputo city, and urban-rural remittances are sent most commonly using bus drivers, which charge high remittance fees and are not generally perceived as safe. The alternative of using the bank as a remittance channel, despite its general level of safety, may involve some unexpected hidden costs due to the special characteristics of the rural areas. In communities where few individuals hold a bank account, bank transfers may involve a compensation to the account holder and (perhaps most importantly) a loss in confidentiality of the transfer. Furthermore, we find high objective money and time costs related to travelling to and using the closest bank branch. But, most importantly, the lack of financial literacy seems to be the major constraint to the usage of financial services. An additional objective of the survey we designed was to accurately measure the determinants of mobile banking adoption, namely trust in the different remittance channels. The evidence collected shows both a high level of trust on banking transfers in general and in the mobile phone service provider, mCel, which grounds our expectation of a high level of trust towards the mobile banking service, mKesh. In addition, in terms of the agents that will deliver the service in the communities, we found that although shopkeepers in general do not inspire high levels of trust, the mKesh agents for our study were recruited according to the exact characteristics that inspire the highest levels of trust from the community members. Looking forward, we expect to observe enough take-up and usage of the mKesh service as to promote financial literacy, via not only our intervention, but mainly the learning that results from direct usage of the mKesh service. This improved financial literacy should contribute to intensifying demand for traditional banking services, as is the objective of the Central Bank. This rise in demand may promote additional investment by the traditional financial sector in rural areas in a way that facilitates economic growth in these areas.