Undercounting Controversies in South African Censuses

Jeremy J. D. Gumbo, University of the Witwatersrand

The main purpose for census taking is to obtain an accurate population count often used for directing policy formulation and resource allocation in a given country. South Africa’s last three censuses have largely failed to achieve this, as high undercount rates have been consistently recorded. National undercount rates have been 10.6%, 17%, and 14.6% for censuses 1996, 2001 and 2011 respectively. Such high undercounting has triggered controversies especially around population counts arrived at by census authorities using Post Enumerative Survey (PES) when estimating and adjusting for undercount. In this study we applied various demographic techniques to access the accuracy of adjusted population counts in these censuses. These included; growth rate analysis, graphical cohort analysis, age ratios, and sex ratios. Findings from the various analyses suggest coverage errors in these censuses. However, such distortions could also result from content errors, and/or population changes through migration.