Population science implications of the inclusion of stillbirths in demographic estimates of child mortality

Similar factors cause both early neonatal mortality and stillbirth, and most stillbirths are preventable. Yet stillbirths are not counted in commonly used summary measures of mortality, such as life expectancy and child mortality. The bioethics literature has recognized that this is paradoxical because it implies that a population in which many babies are stillborn is equally healthy as one in which those same babies survive. However, the demography literature fails to account for stillbirth despite the field’s primary aim of describing population processes. In this article, I show that current thinking about how to measure mortality mischaracterizes population health challenges and contributes to the neglect of stillbirth measurement. To address these shortcomings, I examine stillbirth data from reproductive calendars in Demographic and Health Surveys across 42 countries in Asia, Latin America, and Africa. I find that stillbirth undermeasurement is not uniform across places: the places with worse early life health may have poorer measurement of stillbirth than places with better early life health. Through the calculation of “stillbirth-adjusted” infant and neonatal mortality rates, I demonstrate that it is possible to implement population health measures that include stillbirth. Adjusted rates show that a substantial portion of early-life mortality is left out of traditional demographic mortality measures. These findings have implications for population theory, forcing us to reconsider underlying assumptions in demographic measurement. They also allow for a more complete assessment of the health of societies.