UW Researchers Work on Financial Services for the Developing World

In Kenya, sending money to anyone with a mobile phone is easy, fast, and just a normal part of life for everyone from fishermen and rural farmers to wealthy business owners in the cities. But for all the good that Kenya’s M-Pesa system has done in spreading access to the country’s formal financial system, duplicating it elsewhere has proved difficult.

Now, a team of University of Washington researchers is embarking on a multi-year project to build technology that could help solve the complex puzzle of delivering accessible, mobile-phone based financial services to other places in the developing world. Funded by a $1.7 million grant from the Bill & Melinda Gates Foundation, the team is continuing a decade of work at the UW on Information and Communications Technology for Development (ICTD), including the Open Data Kit, a widely used tool for data collection via Android mobile phones by aid agencies and development workers.

Richard Anderson, computer science professor at UW and leader of the new Digital Financial Services Research Group, is quick to acknowledge that technology alone will not replicate M-Pesa.

“We don’t have any illusions that there’s a silver bullet,” he says. “It has to be done in partnership with multiple different areas of expertise.”



Country-specific factors include policy, regulation, cultural and economic practices, and the willingness and capacity of the telecomm companies and banks to implement what amounts to a sweeping transformation of their financial systems. But having technology at the ready could help “smooth the introduction” of a mobile banking system to other countries, Anderson says.

“There’s a host of areas where computing technology and potentially the developing of new toolkits or new software could help make it easier for these systems to be adopted,” he says.

These include usability—balancing the need to enter passwords and PINs for security with a smooth, easy-to-use service—and infrastructure reliability, again with a focus on security and privacy.

“Our hope is that it’s going to be possible to develop a toolkit that makes it feasible to develop these applications that can pass a basic security audit,” Anderson says.

In 2010, I spent a month in Kenya reporting on the country’s renewable energy potential for my previous employer. M-Pesa—“M” is for mobile, “pesa” is Swahili for money—was a breeze to set up on the candy bar phone I purchased on arrival. I used it to pay guides, taxi drivers, and food vendors. I recharged my account at roadside kiosks in rural villages as easily as in a Nairobi shopping mall.

In the remote village of Kisingitini, Kenya, which had cell coverage but no reliable power when I visited in 2010, fishermen used phones to check market prices and make financial transactions using M-Pesa. Photo by Benjamin Romano

In the remote village of Kisingitini, Kenya, which had cell coverage but no reliable power when I visited in 2010, fishermen used phones to check market prices and make financial transactions using M-Pesa. Photo by Benjamin Romano

Beyond ease and convenience, the impact of M-Pesa and mobile phones—now possessed by about two-thirds of people in Africa—is profound. Gates himself expounds on how mobile phones are unlocking a “financial services revolution” in the developing world in a recent essay for Foreign Affairs, also posted to his Gates Notes blog, on farming, poverty, and the future.

“The power of a phone in every pocket is turning out to be extremely disruptive in exciting ways—and the poor finally have a chance to use technology in ways that solve the real problems they face in their lives,” Gates writes. “Mobile phones have recreated the economics of providing financial services to the poor. In an analog era when banking required buildings, piles of paperwork, security guards, and tellers, the cost per transaction was high enough that no company could even conceive of profiting by serving poor people who transacted in tiny amounts. As a result, the poor led their financial lives informally, paying exorbitant amounts in fees and interest to borrow, save, and send money.

“But phones get rid of all that expensive infrastructure. Transaction costs are so low that companies can make money by serving the poor. And in the process of competing for poor people’s business, these companies will develop new financial products that meet poor people’s unique needs…. In short, digital financial services can create one thriving formal economy that includes everyone.”

Anderson and his colleagues are thoughtful about the appropriate role of researchers and academics from a place like Seattle building technology for the developing world.

“One of the very legitimate questions that comes up about the type of work that we do is, ‘Why are you doing the work in Seattle?’,” he says.

Experts from a top-tier research university can provide deep technical expertise, but may not be able to understand how technologies will be used in a given community, and whether or not the implementation they envisioned will be culturally appropriate. “We cannot have illusions about being able to do that work without the deep local ties,” Anderson says.

Security, a recognized shortcoming of many existing mobile banking apps, is an example of something well-suited to the research group, Anderson says. The group includes two security and privacy experts, the UW computer science professors Yoshi Kohno and Franzi Roesner as well as Kurtis Heimerl, a new faculty member with expertise in low-cost cellular base stations, and Joshua Blumenstock from the UW Information School who focuses on data analysis to understand poverty and economic development.

“The flaw often comes when many people are inventing their own security,” Anderson says. “That’s the type of thing that an academic organization has the potential to move forward—developing toolkits that can be then utilized and employed in multiple settings.”

There is also the question of technology licensing. After investigating, vetting, and prototyping technologies at the university, any solution would have to be handed off to partners for further development and deployment in each country. (Anderson says the geographic focus will be on countries in sub-Saharan Africa and South Asia.)

“The Gates Foundation has very strong principles towards open access, and in our mind that’s consistent with an open-source model,” Anderson says. To attract development and deployment partners, that model “has to be a flexible approach to open source that allows commercial use,” he adds.

The Open Data Kit again serves as a template in this regard.

“One of the ways that Open Data Kit is expanding and being successful is a lot of small companies in these countries will pick up the tool, modify the tool, provide local services, and expand the capacity for the tool,” he says. “And, so, with an open-source project, that is a very strong potential.”

Anderson and UW computer science professor Ed Lazowska credit Gaetano Borriello, a colleague who led the Open Data Kit, with helping establish a culture within the department that is accepting of non-traditional research, and specifically research into technologies that address real problems in the developing world.



“Especially for Gaetano, his students really led the way,” Anderson says, noting in particular the UW PhD thesis nearly a decade ago of Tapan Parikh, now a professor at the UC Berkeley School of Information. Parikh’s interest in exploring how emerging technologies could be used in low-resource settings inspired both Anderson and Borriello to shift their focus.

This work continues at the UW despite the loss of Borriello, who died of cancer a year ago. His presence is still strongly felt on the Open Data Kit team, which received a Gates Foundation grant last fall to make further improvements to the software, with help from Nafundi, a company based in Seattle and founded by two of Borriello’s PhD students.

A remembrance of Borriello on the Open Data Kit site reads: “The last thing he asked of us was to continue doing good work.”

Trending on Xconomy