Harnessing Big Data for the Greater Good

Max Richman is a data scientist focused on building data products and data teams for global development. In addition to his fulltime job at GeoPoll, he is a lead volunteer with DataKind, an organization dedicated to using data science in the service of humanity. He kicked off the second day of the SEEP 2014 conference by explaining that big data is largely designed to anticipate the interests of upper-middle class consumers in the Western world, which includes only 10 percent of the world’s population. He proposed that instead of marketing accessories and luxuries to people who already live in comfort, data can be more meaningfully harnessed by focusing on finding solutions for the 90 percent of the world’s population who are in need of livelihood improvements.

On a panel entitled “The ‘Big Data’ Revolution: Risks and Opportunities for Inclusive Market Development,” Richman shared his views along with Julie Peachey of the Grameen Foundation, Greg Chen from CGAP, and Mike Angus from MasterCard Advisors. The panel discussed ways to use new sources of information to adapt service delivery and improve approaches to inclusive market development.   

Harnessing Data to Increase Financial Inclusion

Following Richman’s introduction, Angus took the stage to discuss how mobile phone data can be harnessed to improve financial inclusion. MasterCard recognizes that many people in developing markets are unable to access credit because they do not have a history with financial institutions, Angus pointed out. Some payment history data may exist, such as utilities, or tuition payments, but that data is not easily accessible or comprehensive. Angus noted that mobile phones are more widely used in low-income countries than bank accounts. He further explained that MasterCard developed models that look for evidence of stability based on mobile phone data to predict financial stability and ability to repay.   

Angus provided a second example where MasterCard developed a program to help ensure that welfare grants were being appropriately used. In 2012, the South African Social Services Agency (SASSA) shifted disbursement of welfare grants from cash to MasterCard debit cards. Analysis of card transaction data identified 850,000 fraudulent grants in the first year of the program. The SASSA’s effectiveness could be improved by increasing accountability on the part of welfare recipients and reducing fraud. 

Using Data to Improve the Livelihoods of Smallholder Farmers

Next, Peachey provided an example of how data can be used to improve the lives of smallholder farmers.  She described the Grameen Foundation’s experiences harnessing data to empower smallholder farmers to provide the best, most informed, and most capable production services. The project conducted surveys through mobile phones in Uganda, Colombia, and Guatemala to develop a profile for farmers and better understand their needs. The profiles included information on: access to financial services, food security status, socio-economic information, productivity information, and a Progress out of Poverty (PPI) evaluation. The project used this information to develop a productivity development plan and advise farmers on which crops provided the most viable business opportunities. For example, the project found that a high percentage of coffee farmers live in poverty, which was attributed to coffee rust disease.  Not only did the project use this information to advise coffee farmers to diversify crops, but in Uganda the project also designed an early warning system to track and predict the spread of coffee rust. This data could be used to alert the Ministry of Agriculture of potential outbreaks so that they could intervene. Overall, the project found that the small costs incurred by data collection were offset by improvements in productivity throughout the value chain.

In a conversation with the Knowledge Driven Agricultural Development Project (KDAD), Alberto Solano elaborates on the Grameen Foundation's experience using data on smallholder farmers to ensure that programs more effectively meet smallholder needs. 

Steps toward Data-Driven Development

Richman and Peachey both noted that most organizations are falling short of fully taking advantage of their in-house capacity to harness data. Peachey noted that data analysis is “in its infancy.” Richman added that this is nothing to be ashamed of, given that even cutting-edge data scientists are still learning how to use data. Richman advised that once organizations take full stock of their data, they should cross-purpose internal staff to develop a strategy for harnessing data. He advised that the best way to get ideas on data strategy is to connect with local technology communities and attend hack-a-thons. If the international development sector is to become a data-driven sector, data-driven decision-making and data sharing across organizations will have to become more integrated into program operations.