Consider the health officials who hope to vaccinate every child in a district, but have little up-to-date information on the population or how it breaks down by age or gender. Census data differs significantly from the numbers coming from health facilities and development organizations, and it seems every organization is using different geographic boundaries to define the district. In development parlance, this is called “the denominator issue.” It would be hard to overstate how profoundly the denominator issue affects health programs in low- and middle-income countries. Without accurate population and demographic data, health officials don’t know how much medicine to procure, how many health workers are needed, or what dollar figure rests on the bottom line. The result is either untreated populations or wasted resources.
Fortunately, several development organizations are applying new technologies to solve the denominator issue, and with some success. The Digital Impact Alliance (DIAL) is working with government ministries and regulators, MNOs, and data analytics companies to use mobile phone metadata for health, agriculture programs, and disaster response. They note, however, that “the demand for MNO data remains fragmented because cross-sector applications are not well known, leading to ad-hoc approaches” or reliance on project-specific and outdated, unreliable census data. GIS mapping is another promising way to crack the denominator nut. The quality of satellite imagery (and access to it) is rapidly improving, and organizations like Facebook, Maxar, Akros, the UCSF’s Malaria Elimination Initiative, and Columbia University’s Center for International Earth Science Information Network (CIESIN) are applying powerful algorithms to identify homes and villages, layer in demographic predictors, and produce comparatively current, reliable population predictions. The pace of these changes is startling: data providers claim to have mapped Africa’s entire 1.3 billion population with over 90% accuracy. In fact, much of Africa has been thoroughly mapped, and population data – expensive to generate – is being made available (free, or at a reduced cost) to governments and development organizations.
It’s worth noting that the denominator issue is not just a health issue. Knowing how many people live in a given area, and something about their demographics, is important for financial services, taxation, agriculture, infrastructure development, and more. Governments and development programs around the world could utilize population data to deliver a wide range of services more efficiently. Together, these organizations represent a very interesting market for technology companies (for-profit, or non-profit) who can offer a reliable, affordable technological solution to the denominator issue.
