Adult mortality is a partial determinant of the growth and structure of a population. The indirect estimation of adult mortality levels is based on parental survival questions in censuses and surveys. But responses to the question may not be taken at face value because there are often errors in the responses. It is therefore necessary to evaluate the quality of the data arising from the responses before estimating indicators of adult mortality from the data. The parental survival questions were included in the 2022 South Africa’s census. This study provided an assessment of the quality of the responses to the parental survival questions and estimated adult mortality indicators from the responses based on Census 2022 10% sample. The results indicate that the attributed percentages of the unspecified to the parental survival questions were high nationally (about 14%) and provincially – highest in Gauteng and the Western Cape. These magnitudes of unspecified are out of sync with those from previous censuses as well as with those from the 2021 General Household Survey.

The trends in adult mortality based on the Census 2022 10% sample are inconsistent with the trends derived from the 2011 South Africa’s census nationally. However, the results suggest that life expectancy at age 15 years improved from 45.6 years in 2004 to about 48.5 years in 2015 among adult males. The corresponding improvement among adult females was from about 51.7 years in 2004 to about 55.5 years in 2015 nationally. These are plausible improvements. The results at provincial level suggest that over the years adult male mortality levels were highest in KwaZulu-Natal compared to the other provinces with the Western Cape having the lowest levels whereas adult female mortality levels were highest in the Free State compared to other provinces and lowest in the Western Cape.

The results have implications for different sectors. For example, pension funds administrators/operators may be under-paying monthly benefits to some pensioners that choose certain type of annuity in retirement in South Africa if the pension funds do not take into consideration improvements in life expectancies, variation in life expectancies by province and by sex.

There are several limitations in this study due to the unavailability of fertility data in the 10% sample provided by Statistics South Africa. A key limitation of the estimates provided in this study is, that they give only a partial view of mortality in the South African population from Census 2022 since children dead of children ever born were not included in the 10% sample.

Author: Emeritus Professor Eric O Udjo (PhD, London) 2025