Anomalies in the recent census do not necessarily mean that the figures themselves are wrong, writes Phil Harrison.
Since the release of Census 2011, critics have pointed to peculiarities in the data, thus questioning the veracity of the results. The most surprising "anomaly", and the one with the most significance for future planning, is the surge in the national fertility rate after the marked decline of 1996-2001.
We need to know whether this is the result of inaccuracies in the data, or a real social phenomenon. The latter is possible, given the success in recent years of South Africa's programme to reduce mother-to-child HIV transmission. Demographers and social scientists must debate the matter and reach a conclusion with some urgency: it is serious if there is doubt about the accuracy of the database underpinning national and local planning in South Africa.
As Philip de Wet pointed out ("Invasion of the young white women", November 16), demographically speaking the structure of the white population in South Africa is of little consequence overall; it comprises a small minority of the total. De Wet then points out an apparent discrepancy in the numbers of white women and men to question the results of the census as a whole. The growth in the female population between 2001 and 2011 significantly exceeded that of the male population and, because a sudden influx of single white women is improbable, the census must be flawed.
But there is an alternative explanation. It is simply that large numbers of white men were missing on the days of the census. No, they were not necessarily up to any shenanigans. The possibility is that thousands of them were not in the country.
The data shows that gender variance in population numbers differs by age cohort. In a "normal" population profile, there is a slight preponderance of males among children and among the aged there is a female majority. The surprise is the preponderance of females in the working-age categories.
Three-quarters of the missing men in this category are from the urbanised provinces of Gauteng and the Western Cape and most would appear to be in the higher socioeconomic status bands. There are about 68 000 absent males. Where were they? The answer, I suggest, is to be found in the shifting work patterns of white men. The work environment has become increasingly constrained, partly because of employment equity, and white men have become more geographically and contractually flexible.
In the past they may have emigrated with their families, but prospects for permanent jobs in traditional destinations such as Australia, New Zealand and Britain have dimmed. And so they have taken up work elsewhere, especially in Africa and the Middle East, where their skills are in high demand.
A quick scan of LinkedIn profiles and the adverts of international recruitment agencies provides an initial indication of the extent of this phenomenon. There are different categories of transnational migrants.
First, there are the employees of South African firms that have aggressively expanded into Africa (MTN, Vodacom, Shoprite, SABMiller and Standard Bank).
Second, there is a growing number of mining and prospecting firms employing white males in Africa's rapidly developing mineral, gas and oil sectors. These firms attract men willing to work in remote and often risky places. Here South Africans have a competitive advantage over Europeans and North Americans, working in these places as geologists, engineers, metallurgists, finance and human resources officers, logistics directors, project managers and security consultants.
Third, the South African presence in the Middle East is considerable and there are large numbers of expatriates in Saudi Arabia, Kuwait, Qatar, Bahrain, Oman and the United Arab Emirates (Dubai and Abu Dhabi). These expatriates hold a wide diversity of jobs in middle-level and senior positions in many sectors. The numbers are uncounted, but one publication claimed, perhaps with some exaggeration, that there were 50 000 South Africans in Dubai alone.
The next category is South Africans working as contractors for companies in high-risk locations such as Afghanistan, Iraq, Somalia and Libya. Many work for private security companies and are part of networks linked to the former South African Defence Force. A 2005 United Nations report on Iraq counted about 10 South African companies sending personnel to Iraq and noted that more than 5 000 of the estimated 30 000 private military personnel in the country were from South Africa.
Finally, white South Africans have been offered farms in countries such as Mozambique, Congo-Brazzaville, Zambia and Georgia. Congo-Brazzaville offered 10-million hectares of derelict state-owned farm land to members of the Transvaal Agricultural Union and nearly 100 farmers took up the offer.
Not all these South African expatriates are white or male, but there is clearly a preponderance of white males in all these categories. It would seem they are South Africa's new migrant workers, albeit a highly privileged class of migrants.
This observation illustrates the danger of jumping to the conclusion that a peculiarity in the census reports necessarily implies the data is flawed. Pali Lehohla, the statistician general, may be correct when he says that time would be better spent trying to understand what caused the anomalies than in questioning the census.
Professor Phil Harrison holds the South African research chair in development planning and modelling at the University of the Witwatersrand and is a member of the national planning commission