Blog

In many emerging economies, the spread of mobile money has transformed access to financial services, especially for the unbanked. According to the Global Findex Database 2021, financial inclusion has increased by 50% over the past decade.¹

This trend is driven by the adoption of mobile money, especially in Sub-Saharan Africa. Despite this remarkable increase, important barriers to access persist. Lack of a mobile phone has been identified as a key source of financial exclusion, especially for women.²

In response, the World Bank calls for strengthening global efforts to improve mobile phone access to “increase account ownership of hard-to-reach populations.”³

Our study explores the use of a targeted intervention to increase smartphone ownership among female, non-mobile phone owners in Blantyre, Malawi.⁴

Here we provide preliminary insights regarding the impact of the program on demonstrated gains in financial inclusion.

Study Design

We targeted 1,500 married women in Blantyre who did not personally own a mobile phone at the time of recruitment. These women predominantly came from low-income households, with only 30% of respondents reporting household ownership of a mobile phone at baseline. This figure falls well below the country average of 64.6%.⁵

Participants were randomly assigned to one of four treatment groups: control; cash placebo; individual smartphone treatment; and couples smartphone treatment. (See graphic for treatment details.)

The key difference between the two groups that received smartphones was the participation, or lack thereof, of participants’ spouses in the smartphone distribution and training. Compliance was high—more than 95% of couples attended the couples training administered in June and July 2021.

About nine months after the program distribution, we conducted in-person surveys to test women’s uptake and use of digital financial services. In addition to survey questions on self-reported uptake of mobile money,⁶ we included a behavioural measure of mobile money use. Specifically, at the end of the midline survey, we offered participants a micro-grant to be paid on the spot. Either 1500 Malawi Kwacha (~US $1.87) was distributed in cash, or 3000 Malawi Kwacha (~US $3.75) was sent as a mobile money transfer. We then recorded which mode participants chose; and, if they chose mobile money, whose account the money was transferred to. Our expectation was that if a participant was both fluent in the use of mobile money and possessed their own mobile money account, they would opt for mobile money to be sent to their own digital wallet. We thus consider this a useful measure of an individual’s realized financial inclusion—whether or not a woman has access to her own transaction account that she uses to send and receive payments and store money.⁷

A couple participates in the couples smartphone training program in Blantyre, Malawi.

Findings and Implications

Overall, 64% of participants at midline chose to receive the micro-grant via mobile money. As demonstrated in Figure 1, women in the individual and couples treatment groups—which received smartphones plus training—were significantly more likely to opt for mobile money than women in the cash placebo and control groups.

But among those who chose mobile money, only around 50% had the grant sent to their own digital wallets, suggesting that many participants rely on others’ financial inclusion to access the funds. To more directly probe the effects on individuals’ own financial inclusion, we ran a second analysis, in which the outcome is a binary variable coded as ‘1’ if a participant chose mobile money and had it sent to one’s own digital wallet or ‘0’ otherwise (a woman either did not choose mobile money or if she did, she sent it to someone else’s wallet). Also reported in Figure 1, we observe even stronger treatment effects from smartphone conditions on realized financial inclusion. Specifically, women in the individual and couples treatment groups were three times more likely to choose mobile money and have the funds sent to their own digital wallets than women in the cash placebo and control groups. Additionally, we observe significant effects on an index of self-reported measures of mobile money use.

Strikingly, despite the sizable cash grant provided to the cash placebo group, representing more than 10% of Malawi’s GDP per capita in 2021, we observe no statistically significant difference between the cash placebo and control groups on realized financial inclusion. This points to the importance of mobile technology and appropriate training for individual uptake and use of digital financial services vis-à-vis lump-sum transfers.

The left panel in Figure 1 illustrates the mean rates of choosing mobile money versus cash in the micro-grant offer across the four study conditions. The right panel reports mean rates of observed financial inclusion—in which one chooses mobile money in the micro-grant and has it sent to her own mobile money account, versus choosing cash or mobile money and having it sent to someone else’s account.

One important caveat, however, is that participants in our midline survey generally reported that their use of mobile money was quite low. Overall, 85% of women said that they did not personally receive any mobile money transfers the previous month; even fewer women initiated transfers. This suggests that, despite their advances in mobile money technical efficacy and realized financial inclusion, smartphone recipients were not applying these capabilities—at least not in the form of sending and receiving remittances.

Interestingly, we observe no meaningful differences in the choice of mobile money over cash between our individual treatment and couples treatment groups. Impacts on participants’ technical efficacy, as measured by mobile money use and control, were similar in both groups. These results suggest that in the sample, couples training was no more effective than individual training at prompting women to choose mobile money over cash. The subject of future analysis will be how and whether the couples’ treatment affected household and community norms around smartphone use and women’s empowerment. Stay tuned!

References:

1.  The Global Findex Database 2021: Financial inclusion, digital payments, and resilience in the age of COVID-19. The World Bank. Retrieved form https://www.worldbank.org/en/publication/globalfindex#sec3
2. For example, in Malawi, the country of focus of our research, women are one-third less likely to personally own a mobile phone than men, according to a nationally-representative survey by Afrobarometer at the end of 2019.
3. The Global Findex Database 2021: Financial inclusion, digital payments, and resilience in the age of COVID-19. The World Bank. Retrieved form https://www.worldbank.org/en/publication/globalfindex#sec3
4. Our study was implemented in partnership with the Institute of Public Opinion and Research (IPOR) based in Zomba, Malawi and the Girls Empowerment Network (GENET) based in Blantyre, Malawi. We are grateful for funding from a 2019 Bill & Melinda Gates Foundation, Grand Challenges Call to Action Award, the University of Texas at Austin, and BIGD’s Women’s Economic Empowerment and Digital Finance (WEE-DiFine) initiative.
5. Afrobarometer, “Summary of Results: Afrobarometer Round 8 Survey in Malawi, 2020.” Retrieved from https://www.afrobarometer.org/wp-content/uploads/migrated/files/publications/Summary%20of%20results/malawi-afrobarometer_r8_summary_of_results-10may20_fin.pdf
6.  Survey questions cover: whether the participant has a MM account; personal use of mobile money to save; mobile money preferred financial instrument; strength of preference for mobile money; frequency of mobile money use; count of mobile money services used; mobile loans taken out over past year; times sent mobile money in past month; and times received mobile money in past month.
7.  World Bank, “Understanding Poverty. Topics: Financial Inclusion. Retrieved at https://www.worldbank.org/en/topic/financialinclusion/overview

Philip Roessler, William & Mary; Tanu Kumar, Claremont Graduate University; Peter Carroll, University of Michigan; Boniface Dulani, University of Malawi, Chancellor College; Daniel Nielson, University of Texas at Austin