News24

Stats Council defends Census data

2012-11-01 16:43

Johannesburg - The Census results are accurate, the Statistics Council, which monitors censuses conducted by Statistics SA, said on Thursday.

"We stand by our decision. The Census results are accurate, and in fact they are very timely," council chairperson Howard Gabriels told reporters.

"They are fit for use in policy-making and decision-making."

Two consultants involved in the Census, academics Tom Moultrie and Rob Dorrington, have expressed concern that the Census contained inaccuracies and had been prematurely released.

They also voiced reservations about figures relating to the age distribution of the population in the Census, and provincial population estimates.

Gabriels rejected criticism that the findings had been rushed, and said the release had in fact been delayed by a few weeks.

Although some of the findings had been surprising, the data was consistent internally and with other sources.

The Census, which was released on Tuesday, showed that the number of people aged between nought and four was much larger than the cohort of people aged between five and nine.

Griffith Feeney, a consultant who worked with the council on the Census, said the data was consistent with a population estimate calculated with the 2001 Census results, the number of births registered after 2001, and the number of births during that time.

"The agreement is not perfect... but it suggests the pattern is real."

A perfect agreement would have raised suspicions as this should never occur, he said.

Good-quality data could be expected to produce results consistent with other data sources, while poor data would be expected to return inconsistencies and wide variance, said Feeney.

Council consultant Eric Udjo said the anomaly in age distribution could be explained by inaccuracies in previous surveys.

"There is nothing like a perfect census anywhere in the world. There will always be errors," he said.

"The 2011 Census was better at capturing the nought to five age group compared to other censuses," he said.

However, it was possible that there was some over-reporting, as people did not always correctly report even their own ages.

Comments
  • ahenriquesfernandes - 2012-11-01 16:51

    SA Motto: "Doesn't matter, Got paid"

      elgh.elgh - 2012-11-01 17:15

      The credibility of the Census came into question the moment we heard that you need to complete the form in pencil. I also know of a few (about 6) that never got counted/completed a form. How many others did not get counted and how many of the Census forms got manipulated.

      louis.olivier.988 - 2012-11-01 17:25

      Howard Gabriels let me a white taxpayer enlighten you, dork. The census documents were completed in pencil and there were occasions where your peolple were altering the census forms. Any contract under law signed in pencil is a worthless piece of paper. Thus the conclusion is how can we trust anything this government spews out. It is merely a stick to beat the whites with again. Hence all the noise made about the higher income of the whites. I know this will not be the last we hear of it. Funny driving down the highway I noticed that the majority of top of the range expensive motor vehicles were driven by AA's. ANy one else notice that? And there were a lot more AA drivers than whites. So take your census and sell that to the anc circus as most of us do not believe the Sh.t you and your circus spew out.

      justice.ubuntu - 2012-11-01 19:15

      @elgh.elgh... i live in the southern suburbs of cape town... i saw neither hide nor hair of any census takers... if they missed our house i would bet they missed a helluva lot more houses around the country... no way in hell can these figures be accurate...

      lsfreak - 2012-11-02 02:24

      Yeah, pencil on a legal document :D When we filled in our form we sat and watch how the forms were changed :D We just stood and laughed, yet now they say this and that and we still laugh. Man it great in SA :D :D

  • bathandwa.ngoma - 2012-11-01 16:53

    i knew that these statistics were not accurate and were cooked. A good timing to release them though

  • julian.booyens - 2012-11-01 16:53

    Hands up, who was not counted?

      kala.bafazi - 2012-11-01 17:02

      And who was counted twice?

      parys.fotograaf - 2012-11-01 18:54

      About 90 hands went up around me ..................in Angola. Do they count?

  • leon.lipschitz - 2012-11-01 17:16

    Forget it

  • thembaletu.mbangcolo - 2012-11-01 17:40

    There is a lot ignorant commentator's here, it is rather sad. It was a pencil used, i did not declare my income, i was not counted. If any of you knew what qualitative research is, these are common occurences, this is why you take a sample - you cannot count 50 million people, not everyone will declare their income and not every house will be interviewed - that is why social science has something called a STASTISTICAL inference (which is why YOU ALL NEED to go study Statistics) and maybe research Methodology, the levels of ignorance may improve.

      benji.engelbrecht - 2012-11-01 18:13

      although your comment may sound well thought-out the simple fact is that it wasnt a "sample", it was a census. the entire point of a census IS to count 50 million people (or however many). so it must be impossible to count 20 million votes using that logic... also, your argument that everyone should go study Statistics doesnt make sense. the census is used by people who didnt study Statistics to make decisions... as per the aritcle: "They are fit for use in policy-making and decision-making."

  • bcluley - 2012-11-01 17:56

    Its actually quantitave research that they conducted! Impossible to conduct QUALItative research on such a large sample.

  • bcluley - 2012-11-01 18:00

    Oops.it is qualitative my bad

  • danie.smit.587 - 2012-11-01 18:08

    Did SSA deliberately try to undercount Afrikaans by reusing and rubbing out info on Afrikaans forms , many times?

  • Gerhard1Marais - 2012-11-01 18:26

    The census is not correct as a data trend analysis coordinator i can state that the information for the census was indeed rigged. The personal that conducted the census across South Africa was not trained adequately as there is no official standard for training on census counting. Investigate ok....

  • Carlton - 2012-11-01 18:30

    They did cook the books, the original figures showed there are 48 mil SAns inculding foreigners, SAs population is actually shrinking. They just want to prove that SAs belongs in BRICS.

  • parys.fotograaf - 2012-11-01 18:51

    What else could they say, but then, we do not have to believe anything they say either.

  • laurens.bussmann - 2012-11-01 19:21

    "some over-reporting because some people could not report their age right??" Well... Isn't that the same with people who claimed they are between 110 and 120 years old?? The oldest living person is 116. Do you really think some old dementia-ridden person will be able to report his age right?? My grandpa who was 102 was swearing to me on his deathbed, that he was Jerry, from Tom&Jerry the cartoon show. The age distribution, fitted into a a graph would not represent the typical growth-graph. Yes the base is broad, because that is the youngest generation, but take that in relation to the generation that is able to bear children... There are not enough "adults"/"men" to produce this amount of offspring, deducting the amount of babies who have been born into this world due to excessiv force and trauma. Please, whoever you are who made the Census 2011, reconsider the mistakes and do not blame it on the mistakes previous Censuses' have made, but rather do it right yourself!!

  • andile.vabasa - 2012-11-01 19:38

    I don't know bout the salaries, average white person earning R360 000. That is R30 000 a month. that is BS, very few people get that kind of salary. I would say those on top earned higher than this but def not average no way.

      wesley.bischoff - 2012-11-01 19:47

      True that. I earn 6 times less than that, R5000 a month. By no means wealthy.

      rohann.geldenhuys - 2012-11-01 23:43

      Lets just read the heading of the graph again..."average annual HOUSEHOLD income by population group of household head". This "statistics does not refer to individual income, but a household in total. This if you have 3 working adults in one household this would increase the average.

  • maryna.wallis - 2012-11-01 20:25

    yip this household also not counted and when contacted them to report this was not interested wot a waste of money and time as per usual in new SA

  • gary.desousa.7 - 2012-11-02 09:30

    I heard on radio this morning that there is a report which brings the results into dispute but they will not release it???? Reports,tapes all secret now??

  • pages:
  • 1