Tina Van Rossem, Vrije Universiteit Brussel and Ghent University
Patrick Deboosere, Vrije Universiteit Brussel
Isabelle Devos, Ghent University
Because people do not always reside in the municipality where they are registered, there are important differences between the de jure population, i.e. the legal population of a territorial area, and the de facto or the actual population living in the area. This discrepancy causes difficulties determining the correct population at risk for the construction of demographic measures. Nineteenth and twentieth century mortality figures for local populations in Belgium for instance can be distorted because of the de jure notation of the population and the de facto notation of deaths in the original sources. In this article we develop a method to determine the bias of unregistered numbers of (temporarily) absent and present people on nineteenth century mortality figures. We use data on the de facto and de jure deaths to estimate the amount of unregistered attendees and absentees in Belgian municipalities. By applying this estimation method to the mortality figures of the four largest Belgian cities around 1900, we demonstrate the need to control for these numbers for the interpretation of mortality figures, and especially for the comparison of mortality figures of different areas.
Presented in Session 9. Health and mortality in the past