/ 22 March 2013

The South African numbers story

Statistician General Pali Lehohla gives a brief history of the census during a Mail & Guardian business breakfast event.
Statistician General Pali Lehohla gives a brief history of the census during a Mail & Guardian business breakfast event.

On March 15 2013, Statistics South Africa (Stats SA) and the statistician general Pali Lehohla held a business breakfast to share more information about the results of Census 2011, in particular the use of small-area statistics.

The South African population has grown by about 10-million from 1996 to 51 770 560 in 2011. Interestingly, Gauteng has become the largest province with 12-million people, followed by KwaZulu-Natal with 10-million people. The Northern Cape has the lowest population count at 1.1-million.

City spaces as centres of growth are reflected by the largest density of people residing near major urban areas. A whopping 40% of the country's population resides in only eight municipalities. The City of Johannesburg's population is larger than that of three provinces and the urban centre contains 9% of the population.

Living arrangements
"You may hear anecdotally that Sandton residents are rich and much better off than Alexandra residents, but we need evidence for that. The census can inform us on many of these questions," says Lehohla.

Living arrangements can be shown in terms of race, male-to-female ratio, age distribution and country of birth, and the statistics offer up interesting questions. In a comparison table between Sandton and Alexandra, Sandton shows a much larger older population. Lehohla says that one of the reasons is that Alex is predominantly a migrant population and elderly people may have gone home.

Employment and education
Data around the unemployment rate and matric pass rates of schools in the Western Cape point to areas of consideration at a policy analysis level. The immediate questions raised are around the reasons for poorer school performances in the lower-income areas and the influencers, and the reasons that some schools perform well in these areas. (The official unemployment rate for South Africa is 29.8%.)

So can you place a value on education? In a graph that compared the percentage of no income in comparison with level of schooling, it became clear that the less schooling a person has, the higher the possibility there is of them having no income.

"Qualifications provide the possibility of income, especially higher education. Education counts," says Lehohla.

Average income by race showed white people earning the highest, followed by Indian/Asian people, coloured people and then black people. If this flies in the face of what you see in the street, Lehohla cautions against anecdotal evidence: "Although you may see lots of black professionals, know that there are income inequalities within the black population itself."

What does the future look like?
Will South Africa be a more equal society in the future? Lehohla says that, in terms of equity, it's difficult to say but he says that the national development plan articulates a desired future.

"The census shows how far off we are from this desired outcome. What we can see is that society is still broken around class and race lines. Class is emerging as a feature but this is not dominant as yet."

The issue of black economic empowerment's continued existence was raised. Lehohla says that although almost one and a half million black people moved into the middle class, they are just a small proportion of the population. Stats SA hasn't run studies on whether black economic empowerment is working or how effective it is, but Lehohla says that it's a study worth pursuing.

Timing of the census
The census is a useful data set but could be improved on through a shorter period between counts. Legislatively, the census should be conducted every five years. However, previously Stats SA wasn't ready at the five-year mark.

The organisation is currently in consultation about the timing, with the Statistics Council advising. Lehohla says that a five-year cycle would reduce the number of surprises produced by changes in the census. However, cheaper and more effective ways of conducting the census need to be found.

Political independence
Regarding political independence, Lehohla is emphatic that interfer- ence is absolutely unacceptable. "Interference is a violation of the fundamentals of official statistics. In South Africa there are no threats and no pressure. We have observed the law, practice and fundamental principles. However, I believe we could strengthen the law even further," says Lehohla.

Citizen engagement
Citizen engagement is a central axis of the national development plan. Lehohla believes that engagement will occur when people have the information.

"In 1996 the census data was sold. This added to the aversion people have to stats," says Lehohla. It also meant that people did not use the information.

"Stats SA needs to provide data that is understandable, that can be engaged with and that is freely available," says Lehohla. The 2011 census data is free.

Certain issues within the country, such as levels of knowledge, education and literacy, play a role when it comes to census dissemination. Lehohla says that using the mass media assumes levels of literacy and that Stats SA needs to find ways of reaching people outside of the mass media. The organisation is currently working on dissemination strategies. Existing programmes are running and include schoolchildren interacting with data and then relating information back to their parents.

Understanding how to use statistics and how to interpret the information is critical. "Previously we were not as articulate at presenting the results, but we have learnt a great deal. Statisticians need to make the information more accessible," says Lehohla.

Development planning relies on evidence-based data and the census plays a critical role. Lehohla believes there will be a more explosive move forward in terms of evidence-based decisions.

Further to that, the census data can be helpful to business in particular with respect to small-area statistics and defining markets.

This article was supplied and approved by the Mail & Guardian’s advertisers. It forms part of a larger supplement.