Saturday, January 3, 2015
Moving Research Applications into Circulation: Putting Knowledge to Work
The final session of the IMFTI conference focused on research applications. "Putting Knowledge to Work" was moderated by discussant Michael Joyce of TNP2K (the Indonesian National Team for the Acceleration of Poverty Reduction) who noted that Indonesia had already opened electronic accounts for a million people receiving government-to-person transfers. He also thought that IMTFI research would be important in ensuring the success of such programs to get people to actually use the "next generation" full suite of tools realized by leading service providers in places such as Kenya with services such as M-Shwari that had interest-based saving, credit, and more sophisticated models than existing microcredit paradigms.
The first paper on "Effectiveness and Challenges of Using Mobile Money Service in the Implementation of the Social Assistance Grants for Empowerment Programme: A Case of Kiboga District in Uganda" by Julius Okello and Dorothy Massa of the African Institute for Strategic Research Governance and Development emphasized a G2P case in the Kiboga District of Uganda. Okello explained that SAGE (Social Assistance Grants for Empowerment) was the first major social protection initiative by the government of Uganda and how it was an arm of expanding social protection under the Ministry of Gender, Labour and Social Development. It received funding for launching its services from DFID, Irish Aid, and UNICEF, and it is also supported by Maxwell Stamp, a UK based consulting firm. Okello explained that SAGE was a five-year pilot program which started in 2011 in Kiboga, Kaberamaido, and Kyenjojo but was later rolled out to another 11 districts. Researchers noted that Kiboga has a population of 148,606 people, of which 4,808 citizens, both male and female, participate in SAGE. SAGE targets elderly citizens, those from intact heritage cultures, and other vulnerable populations.
According to Okello, the MTN mobile money service was launched in the country in March 2009 and has registered 1,553,770 users since. He asserted that this translates into an adoption rate of 64,740 persons per month. On average the relative market share of MTN Mobile Money, ZAP, and M-Sente between March 2009 and February 2010 was 89.6 percent, 9.1 percent, and 1.3 percent, respectively. Given the history, coverage, and capability of MTN Uganda, the government opted to hire MTN mobile services to remit funds to SAGE beneficiaries. The use of mobile money in Uganda is still relatively new, according to Okello, however.
Okello's study used a mixed methods approach that combined quantitative and qualitative information gathering and was inspired by Creswell & Plano-Clarks' 2011 manual for researchers. The Kiboga project was designed to consult about three hundred informants. Quantitative methods included regression analysis and involved statistical figures, tables, and graphs, while the group's qualitative methods emphasized detailed explanation with evidence about their informants' livelihoods. He showed a complex circuit of stakeholders in the project, who included the government of Uganda, the Bank of Uganda, development partners, The Ministry of Gender, Labour, and Social Development, the Uganda Communications Commission, MTN, MTN district agents, the district coordinators of ESP, and the SAGE beneficiaries themselves. Because Uganda is one of the worst countries in the world for corruption, according to Okello, the hope is that this system of checks and balances will protect SAGE constituents. He also worried about potential complaints from telecommunication companies that might feel "locked out" of participation.
For elderly participants without cellphones, MTN created a variety of devices and services, including this yellow box that performed all of the functional financial operations of MTN transfer services. Yet MTN has relatively limited coverage outside of the capital, so money sometimes cannot reach the beneficiary and becomes diverted to other parties through corruption and fraud. Unfortunately researchers found that the elderly are often targets who unwittingly share pin numbers because they are reluctant to travel the distance to the point of contact with mobile money agents or are too gullible when told that machines have broken down. In concluding Okello argued that the elderly could still be vibrant economic actors capable of launching businesses, and he closed with hopeful stories of growing sunflowers, poultry, pineapples, and piglets and an anecdote about how ritual slaughter is combined with the seeming windfall of mobile money.
Another G2P program was featured in "Paying Conditional Cash Transfer Programs in Bank Accounts" by Enrique Seira of ITAM-QFD, and the panel shifted its focus to Mexico, where it was also possible to analyze characteristics of debit card owners and their behaviors as consumers. Seira noted that in Mexico debit card owners have three times more tertiary education, three times more savings, and more trust in general in society and institutions. (One Seira survey question asked if informants thought that a lost wallet would be returned to measure such trust.) Debit card owners tend to live in municipalities with 20% more ATMs per capita. In contrast, Oportunidades “Debit” beneficiaries, who were not self-selected by their decision to get a card and used accounts relatively little, as described in this earlier IMTFI panel. Seira noted that they also had significantly less education, and although about 45% claimed to have savings, only about 10% had "formal" savings. This amount of savings represented about 1/50 the numbers common in the general population. A comparison of late vs early debit card owners in Oportunidades showed that they were similar, which was good for measuring the impact of debit card expansion, as the cards became able to work in ATMs and be accepted in stores.
Seira claimed that the aspect that influenced the value and use of the account was its convenience, although a savings account with no branch nearby and no card able to pay in POS devices was not very useful. Unfortunately, this was how Oportunidades (Mexico’s main CCT program) operated pre-2009. Of about 1 million savings accounts, none had ATM/Debit card features. In 2009, Oportunidades started awarding Debit ATM cards in a staggered fashion. Expansion was decided by Oportunidades at the local level (not by bank branches).
Seira has planned to do a DID analysis of the effect of ATM/debit on savings in recipients' accounts. He noted that his sample was based on data on account savings for 342,000+ beneficiaries of 308 Bansefi branches in 411 municipalities during a period covering 17 bimesters: January 2009 to October 2011 (5 bimesters pre-treatment and 12 post, max). About 70% of participants received an ATM card during the researchers' sample period. There were three group studied: early switchers, late switchers, and those who had not yet switched. The three groups were distributed geographically across the country and included similar distributions of indigenous groups. Although seemingly counter-intuitive, Seira also posited that debit cards could impact savings behavior positively, because withdrawals could be smaller and done in a more controlled manner. He acknowledged that yesterday's IMTFI session with Jonathan Morduch reached the opposite conclusion that plastic could encourage more irresponsible spending behavior. Subjects in Seira's study ultimately saved 30% more than non-implementers, although this behavior often took six months to manifest itself. It also often followed a pattern of significant variation among users, and research indicated that subjects had relatively low financial literacy regarding ATM use and fees. He observed that new users also frequently checked balances, because they were distrustful about money remaining in the account.
Researchers also found evidence of early implementation problems, especially involving emigration to debit card accounts. Surveys were also given to 100+ personnel from ATM providers, and those serving in early adopting municipalities seemed to have less knowledge about training materials. There were also surveys distributed to 5000+ beneficiaries, which revealed that early adopters (even if they have had more time to learn) needed more help to use ATM card than late adopters (57% vs 50%, 5% significance). While 32% of early adopters knew they had a savings account, only 30% of early adopters knew (10% significant). Late adopters did 35% more ATM purchases per bimester. According to Seira, there seemed to be increasing use and trust with time as well.
The final formal presentation at the IMTFI annual conference was "The Use and Impact of M-Shwari as a Financial Banking Product in Urban and Rural Areas of Kenya" by Jane Mutinda of Kenyatta University and Ndunge Kiiti of Houghton College and Emory University. You can read more about previous IMTFI research produced by Kitii and Mutinda's collaborations here. Mutinda credited Professor Charles Nzioka of Nairobi University, who also participated as an author in their findings presented this year. Monique Hennik of Emory served as a consultant to improve the group's training, because the university students that they used as field researchers came from many different institutions.
Mutinda opened her presentation with a gorgeous and evocative video that explained that the Jua Kali (the word refers to the heat of the sun) were self-employed people working in the informal sector of the Kenyan economy doing 90% of the service work, which included masonry, plumbing, carpentry, metal working, and many artisanal trades. According to experts shown in the film, this sector also employs about 19 million people and is critical for successful development. Even the economic commentators described their own experiences as Kenyans in the bankable population as very difficult, so that someone might have to walk dozens of kilometers to make just one transaction. People in the Jua Kali sector found opening and maintaining accounts particularly challenging, although access to resources, small loans, and savings for small businesses were all among their critical needs. In the film researchers argued that Safaricom needed to do more outreach and education to explain the product, which was less widely in use in rural areas divorced from cosmopolitan populations of users and which privileged English speakers over those who communicated in Swahili or tribal languages.
Kiiti took over the presentation after the video to explain the "why" and the "how" of the project, which built on their earlier work in mobile money in Kenya. They also observed increased use of the M-PESA platform and an expansion of Safaricom bank partners providing mobile money services and products. In addition to M-Shwari, provided in conjunction with the Commercial Bank of Africa (CBA), Kiiti also noted the existence of M-Kesho with Equity Bank and M-Benki with Kenya Commercial Bank. Researchers chose to focus on M-Shwari, which was introduced in 2012, because it allowed subscribers to save and borrow from their phones and to earn interest on money saved. It also provided access to credit and a paperless form of financial transaction. Nonetheless, more needs to be done to reach the unbanked: apparently CBK has estimated that over Kshs. 300 billion ($3.5 billion) sits outside formal banking systems, and millions of Kenyans are still unbanked. She noted that 50% of people using M-PESA still did not have a bank account. Researchers also consulted with Safari.com to attempt to influence policy, although they were wary of possible influence.
Kiiti summarized from research highlights that indicated that M-Shwari loans were "essential in providing quick cash-flow for the Jua Kali businesses." She also cited a number of strong assertions from those in the financial sector about the power of this product, including Michael Joseph, Director of M-Pesa within Vodafone, who argued that "M-Shwari is a 'transformational service'; saving is no longer the privilege of the elite." She justified the study's focus on the Jua Kali, by emphasizing the fact that the informal sector contributes about 18.4% of Kenya’s GDP while creating 74% of all new jobs annually, although researchers were sometimes stymied by a lack of data. She incorporated research done by the Africa Development Bank in her presentation as well, and argued that Kenyans rely on the Jua Kali sector for services and feel strong investments in those communities. She explained that workshops continue to be part of their research study design.
The study trained 15 graduate students and professionals. Students came from six different campuses and were tested on their cultural/language representation (both vernacular & national). Their training in a mixed methods approach included role-playing and practice listening, probing, and note-taking, and they included M-Shwari representatives in the process. 8 counties were and 4 regions were included in their fieldwork. Consent was very important given the nature of the very personal questions asked by researchers about money. 160 Questionnaires were gathered with 10 users and 10 non-users of M-Shwari in each county. Researchers examined sample characteristics, such as access/convenience, usage, gender, and sectors and found trends that involved social status. As the research team moves forward they hope that the link/gap between marketing or promotion and knowledge or application could be better addressed if policy and practice questions receive more attention. Obviously researchers felt that their findings could be very significant for issues of financial inclusion and exclusion in mobile banking in Kenya.