| dc.description.abstracteng | Over the last 10 years mobile money has emerged as an important innovation with a potential to increase financial access in developing countries. Mobile money is making an important contribution to financial inclusion in many ways. First, it’s increasing access to financial services to a large number of people, who are effectively excluded from the commercial banking system by virtue of the fact that they live too far from banks or do not have sufficient funds to meet the minimum deposit requirements to open a bank account. Second, mobile money agents are now located in remote locations where it would not be economically viable to open a bank branch, allowing mobile money services to be extended into remote locations. Households are now able to receive remittances and social grants through mobile money. Furthermore, mobile money is now being used to facilitate access to insurance, credit and savings even for poor households in remote areas. Access to basic financial services through mobile money could enhance the ability of rural households to invest in their livelihoods and improve welfare. Because of its potential to contribute to financial access especially for the poor in developing countries, mobile money has generated immense interest among policy makers and researchers. Despite the importance of mobile money on the lives of rural households, there is limited empirical evidence on the drivers of mobile money adoption and its broader welfare effects on adopting households.
This dissertation comprises three related essays on the adoption and welfare effects of mobile money (a new financial innovation). In the first essay, we analyse whether social networks affect the adoption of mobile money. In the second essay, we analyse whether the use of mobile money technology has an effect on household food security. In the third essay, we analyse the effect of mobile money on household welfare and poverty. In the second and third essays, we specify two alternatives of the treatment variable. We used a dummy variable for mobile money use as well as intensity of using mobile money. We address these questions using cross-section data collected from rural households in Uganda. Uganda offers an interesting case study because the use of mobile money is growing rapidly. About 27% of the population in Uganda is using mobile money and this has surpassed 20% of the population with bank accounts. In addition, over 70% of the population resides in rural areas where banking infrastructure is underprovided.
In the first essay, we analyse the effect of social networks on the adoption of mobile money. In developing countries, financial information services are underprovided, limiting household’s ability to adopt new financial innovations. In most cases, households rely on social networks to get information on new innovations. Interestingly, no previous studies have systematically analysed how social networks affect mobile money adoption. We contribute to this limited literature by using unique social interactions household level data collected through the random matching within sample approach. A conditional logistic regression model is estimated controlling for household characteristics, correlated effects, and other possible information sources. We find that information exchange within social networks helps disseminate information about mobile money and has enhanced its adoption. An increase in the exchange adopters within household social network increases the odds of adopting mobile money by 77% points. However, the structure of the social network appears to have no effect on mobile money adoption. We also analyse the effect of social networks, separately for poor and non-poor households. Our results show that social network effects, and in particular the size of exchange adopters appear to be more pronounced for non-poor households. These findings have important policy implications for the diffusion of mobile money. In particular, they suggest that exchange within social networks help disseminate information about mobile money. The adoption of mobile money is likely to be increased if promotion programs reach more social networks. Furthermore, mobile money promotion programs need to reach the poor, because our evidence suggests that the poor may be trapped in information-poor networks and thus social network multiplier effects will most likely not automatically work in their case. Improving rural household’s access to informal information channels is particularly important in developing countries, where formal information institutions are lacking. There is a need for policy makers, mobile money service providers and extension to strengthen and utilize informal institutions to disseminate information about mobile money.
In the second essay, we analyse the effect of mobile money on household food security. Mobile money technology is growing rapidly in developing countries. However, empirical studies of the broader welfare effects of the technology on rural households are limited. This essay contributes to the emerging literature on mobile money in several ways. First, we study the effects of mobile money on food security – an issue high on the global policy agenda. This has not been addressed in previous literature. Second, unlike studies that use one measure of food security, we contribute methodologically by using two measures of food security. We account for potential endogeneity of mobile money by using treatment effects and instrumental variable regression techniques. Controlling for other factors, use of mobile money decreases household relative food insecurity by 0.20 index points. Furthermore, we find that use of mobile money and intensity of using mobile money increases monthly food expenditure per adult equivalent by 9% and 1.4% points respectively. We conclude that mobile money positively affects food security. Interventions and strategies to improve household food security should consider the promotion of mobile money among rural households in Uganda and other developing countries.
In the third essay, we extend our analysis and study the effect of mobile money on household welfare and poverty. We measured welfare using household consumption expenditure. We measured poverty using the consumption poverty and multidimensional poverty index. We estimate instrumental variable and endogenous switching probit regressions to control for the potential endogeneity of mobile money. Model results show that use and intensity of using mobile money increases consumption expenditure by 10% and 2% points respectively. Based on consumption poverty, our model estimates highlight that use of mobile money reduces the probability of being poor among users by 10% points. Furthermore, among non-users the probability of being could be reduced by 5% points if they were to use mobile money. In addition, higher intensity of using mobile money reduces the likelihood of being poor by 1% point. The use of mobile money reduces multidimensional poverty by 0.67 index points. We find that a unit increase in the intensity of using mobile money reduces multidimensional poverty by 0.07 index points. Our results for the effects of mobile money are robust to alternative specifications of the treatment and poverty outcome variables. The results confirm that mobile money is welfare-enhancing and reduces household poverty. Policy interventions to improve household welfare and reduce poverty should embrace the promotion of mobile money among rural households in Uganda. In the last part of this dissertation (Chapter 5), we conclude and summarize the key study findings. We derive policy recommendations and elaborate the study limitations and considerations for future research. | de |