Explaining Mortality Outcomes at the District Level in Ghana, Malawi, and Tanzania

Rachel S. Robinson, American University

There is more variation in health outcomes within countries than between them, yet major studies of life expectancy, maternal mortality, and infant mortality persistently analyze outcomes at the national level. At the same time, research conducted at the individual level of analysis often fails to take into account the characteristics of populations and places that influence the spread of communicable and parasitic diseases. In order to address both these shortcomings, this paper uses unique data measured at the district level along with exploratory spatial data analysis techniques to model the social, physical, and built environmental determinants of health outcomes in Ghana, Malawi, and Tanzania. Such a strategy makes it possible to identify the drivers of good, as well as poor, health at the level of analysis most conducive to programmatic interventions.