In the field…with mobile applications

An economist working in Africa might seem an unlikely author for an article of interest to the telecom industry.
I write this piece from near the Sudan-Uganda border, where my work would not be possible without the newest mobile technology. My research…

An economist working in Africa might seem an unlikely author for an article of interest to the telecom industry.

I write this piece from near the Sudan-Uganda border, where my work would not be possible without the newest mobile technology. My research centers on the risks taken by the poorest farmers in Africa, particularly those here in Uganda and in surrounding regions.

When I was still a Masters student at the University of Chicago, I volunteered through the Bankers Without Borders program and began working with Grameen Foundation. As I began to study microinsurance – very small insurance policies meant to guard poor, small-plot farmers against threats such as drought – I began to realize that few, if any, large-scale studies had been completed on the actual risk behaviors of these farmers. Without knowing how farmers make decisions and try to avoid or mitigate risk to their farms, it’s nearly impossible to understand how insurance products might be built to protect these farmers.

I started designing a method to measure over fifty factors of a farmer’s life around the time I began as a doctoral student at the London School of Economics, but I quickly realized that deploying a large-scale study with just a few research assistants would be logistically impossible. The project sat on my desk for a few weeks as I experimented with various sample sizes and ran ideas past colleagues, each option being either too expensive or simply impractical. I estimated the sample size needed to truly get a sense of the farming conditions across an area like Uganda would have to at least include 4,000 farmers across two regions (one highland, one lowland). If I took two research assistants to Uganda with me and visited a different farm each day, five days a week, for six hours on each farm, and if I paid my research assistants ten pounds per hour for these six hours plus an average travel time of 2 hours each way, the project would take over fifteen years and would cost over US$1m – and that would only cover the salary of my two research assistants! Clearly, there had to be a better way.

Then, I was on a call with Grameen discussing their work in the field and possible sample sizes and everything seemed to fall into place.

The Community Knowledge Worker (CKW) network is a web of hundreds of people throughout Uganda equipped with mobile phones. It is quickly growing, thanks to a US$4.7m expansion grant from the Bill & Melinda Gates Foundation. The program also created an alliance with MTN Uganda, the country’s dominant mobile operator. Grameen’s work with Google now included the use of Android-based HTC handsets for the CKWs, allowing them to run advanced apps.

With Jennifer F. Helgeson – another economist from the London School of Economics – and an application programming team, we began working to turn the survey tool into a mobile application. In all, over thirty versions of the survey were developed until the final version was ready to be loaded into the mobile phones. Because many of the CKWs are in remote locations and rarely travel to the capital, Kampala, the application was made available for download via their handsets.

The final version includes other ways of testing farmers’ attitudes toward risk-taking, including looking at the riskiness of crops the farmers have planted in the past and asking the farmers to play games that involve flipping coins and rolling dice. By looking at how farmers behave when faced with uncertainty in these contexts, we get a better sense of how farmers might make decisions about an upcoming, uncertain growing season.

Jenn and I flew to Kampala from London and worked with other members of Grameen’s team to finalize the application programming, loading it on a series of HTC devices running Android. With a working application, we drove several hours to Kapchorwa in the east of the country to meet with a group of CKWs. What we found when we arrived, however, was beyond anything we could have hoped for.

The CKWs were eager to discuss the new tool and we spent an afternoon discussing the application. After a few hours of role-playing where we asked each other the questions and critiqued the application’s translation to the local language, we were ready to make the final edits. We drove back to Mbale, a few hours away, and worked directly with the programmers to make overnight changes to the application.

Next, we drove to Oyam, an area in the north-central region of Uganda to meet with farmers and get their impressions of the issues covered by the application. In an afternoon focus group in the shade of mature citrus trees, we learned about the problems facing farmers in Oyam. Indeed, Oyam’s farmers seemed substantially worse off than those in Kapchorwa. The 200 kilos of maize a small-plot farmer might grow in either place only produces 200 shilling per kilo here. The 40,000 shillings a farmer might earn for his harvest is only equal to about sixteen U.S. dollars. Farmers were eager to hear about the new survey tool, the availability of insurance, and our interest in their farming challenges.

The next day, we returned to Oyam to meet with CKWs there. Oyam is a very poor area and a difficult place to be a farmer. Oyam falls victim to drought one out of every two or three years and experiences bizarre weather, especially given its equatorial location – a huge hailstorm wiped out nearly 100% of its crops two years ago. Three such hailstorms struck Oyam in the last decade. We spent the entire afternoon working with translators and listening to CKWs to make the MHST as easy as possible to administer in the local language. As in Kapchorwa, the local language in Oyam has no word for “insurance,” so explanations and examples had to be devised. One hypothetical involved a bowl of apples, but someone pointed out that many in Oyam have never seen an apple; the fruit was changed to a locally-abundant fruit in the final version.

I took notes as the final translation, phrasing, and other edits were completed. We left Oyam and drove to Gulu, a city marred by warfare. Several of Gulu’s buildings still show evidence of shelling and gunfire, and the local airfield is in desperate need of repair We used Orange’s newly-installed 3G network in Gulu to take a few Skype calls and to run through email before beginning the five-hour drive back to Kampala.

Once back in Kampala, I sat down with Charles, our programmer, to make the final edits to the mobile phone application. After over eight hours of work, the MHST application was ready to be downloaded by over a hundred CKWs and taken to over 5,000 farmers in Uganda. Without the MHST, many of these farmers would never have been able to tell their stories or influence the engineering of financial products – now, they have a voice in the design of a generation of new insurance products that may protect them from the next drought, flood, or disease.

This project, if successful, has the potential to revolutionize how mobile technology is used in the developing world. Further, it will reshape the boundaries of feasibility in using mobile technology in fieldwork and academic settings. What was once gadgetry aimed solely at a developed-world audience is now in use throughout northern Uganda by some of the world’s poorest people. What would have been a study of a hundred farms ten years ago can now be a survey of over 5,000 farms, thanks to several parallel evolutions of communication technology, from the handset to the application to database. An application that would have cost millions to develop only a few years ago can now be built on a small budget in an office building in Kampala, Uganda and nearly instantly downloaded and installed on over one hundred phones, some of them in villages without a single paved road.

I’m excited to be part of the movement to use these technologies in research historically dominated by wandering academics with notebooks. And, by using local people, local social networks, locally-built software, and local languages, I believe we’ll learn far more about these farmers’ lives than ever would have been possible with an economist from London and two research assistants.

About the author:

Karl is an economist and legal scholar whose current research focuses on issues facing poor farmers. His pieces and comments on issues ranging from economic theory to gentrification to securities law to technology have been featured in academic journals, online media, and the Oprah Winfrey Show. He is a PhD candidate at the London School of Economics and Political Science, a researcher at the University of Chicago, and a blogger at www.globalpolicyjournal.com. He divides his time between Uganda and the United Kingdom.