Phillip de Wet looks at some of the puzzling anomalies that cast doubt on the veracity of the figures released this month by Statistics South Africa.
Sudden surge in babies and toddlers
During the next seven years, nearly 5.7-million children will be entering primary schools, an increase of slightly more than a million compared with those who entered school in the past seven years.
Between them, the new group will require more than 26 000 extra classrooms and teachers, not to mention books and other tools of education, which will stretch the capabilities of provincial education departments to breaking point – if these extra children exist.
The very large, entirely unexpected and as yet unexplained number of children between the ages of zero and four reflected in last year's census count is one of the anomalies that have led at least two prominent demographers to question the accuracy of the entire count.
The analysis of other data released by Statistics South Africa this month has revealed other counterintuitive and troubling trends, including a large influx of young white women, even as people born in Europe and North America are leaving South Africa in massive numbers.
But a sudden surge in the number of babies and toddlers could have the biggest impact, requiring either a very rapid building of schools if it is accurate, or, if it is not, potentially leaving tens of thousands of newly provisioned classrooms as white elephants.
Outcome reflects reality
Statistics South Africa has consistently defended the integrity of the census data, saying that its process meets the highest of standards and the outcome reflects reality. Where the results conflict with expectations, statistician general Pali Lehohla has said, time would be better spent trying to understand what caused the anomalies than questioning the census.
With much of the census data now public, though, there are still vocal dissenters.
"With what we now have, the reported age distributions would imply a total increase in the fertility rate of 17% between 2002 and 2008," said Tom Moultrie, a demographer at the University of Cape Town, who first raised concerns about the census.
"That is demographically implausible."
It is also diametrically opposite to the assumption used at various levels of government for future planning for many years that there will be a small but steady decline in the birth rate.
The school requirement implications of the census are most dire for Gauteng, in which the primary school population will jump by 44%. This will require about 9100 more classrooms and teachers in the next seven years, according to a Mail & Guardian analysis.
North West will have the second-highest percentage increase in primary school populations, with nearly 100000 more children entering the school system, although KwaZulu-Natal will see the second-highest growth in actual numbers, which will require about 3800 new classrooms.
In the past two years, the Gauteng department of education delivered 1600 prefabricated classrooms to schools in the province a year, it said in early November. It expects to hire 750 new teachers in 2013.
Invasion of the young white women
South Africa's white population is tiny relative to the whole, but some demographers consider that a blessing in terms of checking the accuracy of models, forecasts and surveys, exactly because of its small size. That white families tend to live in formal housing and are disproportionately urbanised also make them more likely to be accurately counted.
In 2011, the numbers for younger white women stack up strangely.
Comparing the census results for 2011 and 2001 shows unexpected patterns, particularly in the female white population.
If both counts are accurate, then the number of white women between the ages of 20 and 39 increased by 30299 during the decade.
This is a net number. Factoring in deaths and emigration over the same period, more than 40 000 young white women moved to South Africa from abroad during the decade, increasing the size of the group by more than one-tenth.
The increase cannot be explained by age misreporting, which is a known problem with census counts. The only part of the young, white, female population that didn't see an increase in their numbers since 2001 are those who were between 25 and 29 during the 2011 count, and this group decreased by only 2.6%.
Could foreign-born young white women be streaming into the country?
Shortage of men
Not according to the census data. A question on the region of birth shows massive declines between 2001 and 2011 in the number of people who say they were born in North America or Europe.
That leaves one possible explanation, other than a problem with the data: a big influx of former expatriates. Are young white women who were born in South Africa but left the country with their families before 2001 coming back in record numbers?
Although possible, it seems unlikely because of the shortage of men.
According to Homecoming Revolution, a non-profit organisation that encourages expatriates to return to South Africa and has dealt with tens of thousands of potential returnees since 2003, its gender split is 51% female and 49% male.
This does not match the census findings at all. Compared with the increase of more than 30 000 young white women, a net amount of only 4068 young white men was recorded during the same 10 years.
If young white female expatriates were indeed streaming back into the country, they were leaving their brothers behind.
The flow back to Europe and the US
If 2011 census data is taken at face value, then people born in Europe and North America have been streaming out of the country since 2001 in extraordinary numbers, largely denuding South Africa of anyone born in Britain or Germany.
A comparison with the 2001 census shows a 77.6% decline in the number of people counted who said they were born in North America, whereas those born in Europe or the United Kingdom decreased by 68.5%.
Although the region of birth category in the census results refers to broad geographic regions, a comparison with a small subset of sample data indicates that every person born in Germany or the UK left this country during the past decade. Foreign missions to South Africa have limited information on their citizens resident in South Africa, but such a trend is not showing up in their data.
"Based on the number of Americans who have self-selected to register with the US consulates throughout South Africa, we estimate that the number of US citizens resident in South Africa has increased since 2004," said embassy spokesperson Jack Hillmeyer this week.
The 2011 census data shows a disturbingly high rate of poor answers to the question of region of birth. More than two million people, or nearly 4.2% of the total population, were defined as "unspecified".
However, this high rate of uncertainty does not seem to have affected more expected trends. A comparison with the 2001 census shows that in 2011 South Africa had 77.4% more people who were born in the Southern Africa region and the number of immigrants or visitors from elsewhere on the continent had jumped by 217.4%, for a total born-north-of-the-Limpopo-River count of more than 1.35-million people.
The number of people originally from Asia also more than doubled to 84 976.