In an unprecedented development, Stats SA says it has been advised by its technical experts not to release key data from the 2022 census, including figures on mortality, fertility, employment and household income.
The exclusion of the data amplifies the concerns previously raised by UCT demographers about the integrity of Census 2022 and potentially hampers the work of policymakers, researchers and government departments that routinely use the information.
A census is a population count at a given point in time. Stats SA conducted the most recent census in February 2022, during the Covid-19 pandemic, and released the main results in October 2023. A follow-up report released earlier this week said it excluded data on household income, employment, mortality, and fertility after evaluation by its subject-matter specialists and technical experts, which identified reporting and coverage biases.
“When variables that have been released routinely in the past with very few exceptions are not released at all, it leads us to question again whether the census is indeed fit for purpose,” said UCT demographer Tom Moultrie, co-author of an analysis of Census 2022 published by the Medical Research Council in July. It concluded Census 2022 was so flawed it could not be relied on for planning budgets and investments or monitoring government programmes.
There was a huge 31% undercount, but the census contained many implausible anomalies, the authors said. The 2011 census had an undercount of 14%, which was already high by international standards: the 31% undercount was the highest recorded to date by the UN Population Division.
Moultrie said it was important to understand what went wrong with Census 2022 before planning began for the next census, due to take place in 2031.
“A 31% undercount in anyone’s book is a sign it was not a good census. That we are seeing data not being released is another sign this wasn’t a good census,” he said.
Stats SA previously stood by the integrity of Census 2022, and attributed the high undercount to the challenges of running a census during the coronavirus pandemic and the rush to get the job done due to the Treasury’s refusal to let it roll unspent census funds over from 2021/22 to the next fiscal year.
On Wednesday, Stats SA deputy director-general for population and social statistics Solly Molayi said the exclusion of data on fertility, mortality, income and employment did not undermine Census 2022 in any way.
“If we were not solid in terms of our numbers, we would not [have] released it. The Statistics Council, an independent body, has done its own assessment; they concluded the numbers are solid,” he said.
Technical input had been provided by Stats SA staff, the Statistics Council and external consultants. Their reports are not in the public domain, and Molayi declined to name the external advisers. “We have to protect our technical experts. We are the face of the census.”
Stats SA also stood by its midyear population estimates for 2023, published in July, said Molayi. Census 2022 was just one of the inputs used for this population projection, which also drew on administrative data such as the birth and death registration records held by the department of home affairs, Molayi said.
Wits governance expert Alex van den Heever said it was vital for Stats SA to resolve the issues raised by Moultrie and his co-author, Rob Dorrington, and explain in detail why so much data was being withheld.
“There should not be a mysterious black box in the middle of the conversation. It’s not a matter of national security: it is entirely technical and there is no reason for it not to be in the public domain,” he said.
Moultrie and Dorrington are not the only researchers to question the results of Census 2022. Last year the Gauteng City-Region Observatory said many of the census findings on Gauteng were surprising and bewildering.
“The census data, if correct, suggests a remarkable shift in the fortunes of Johannesburg relative to other Gauteng municipalities and other major cities in SA,” it said in a report published soon after the census was released.
“Johannesburg added only 368,631 people between 2011 and 2022. This was a mere 0.7% annual average compound growth per year. By comparison, Tshwane saw 3% annual average compound growth, and Emfuleni, historically a very low-growth municipality, grew at 2.5% per year,” it said.
Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.
Please read our Comment Policy before commenting.