Any census of the population comes with major challenges, given that it runs ten years apart. Those who participated in the first may not be available for the second and, in that respect, tacit knowledge is lost.
Technology changes also severely affect census planning and running because it impacts the process flow and system integration. When paper-based systems were deployed, the system was very linear.
But what remained very crucial was setting the date so that all streams converge to that date. Once that is agreed upon, user consultations begin in earnest in order to ground the framing of the concepts and what the national enquiry is all about. Then comes the testing of concepts and crafting these into specific questions that seek to get one and only one answer, be it yes or no, male or female, age as in two years old, employed or unemployed, or a pensioner.
When it comes to where you are on census night, again, only one place can be recorded as the place where you slept at nightfall and woke up from by daybreak.
However, technological advances have rearranged the steps of conducting a census and order. For instance, under a key entry processing environment, which has dominated census data processing for at least several centuries until scanning was possible, the first step in data processing was coding of the questionnaire to ready it for data entry through the keyboard.
But scanning technology reverses that entire order wherein the starting point is data entry and then coding, subsequently. With hand-held devices, the three phases of data collection, data entry and coding are merged.
Of course, the first is data capture, coding and data entry. However, all these operations happen all at one point, at data enumeration. So the writing of responses, which used to be captured on paper, are now merged into a single instrument, a hand-held electronic device. This is the case under internet-based census undertaking. The census taker is eliminated in the value chain.
But what does not change, however, are the interesting questions that worry politicians. The standard answer of a census taker is that a census should be run when the environment in the country is calm with limited movement. A census is a snapshot and not a videography. That is why in the fixing of the date, it is tied to where did you sleep on that night, the census night.
In one of the consultations for the Census 2011, we invited politicians. And they came and asked “innocent” questions. But there is nothing innocent in a question any politician asks.
Inkosi Patekile Holomisa, a member of Parliament, asked his “innocent question”. He asked, “Statistician-General, why don’t you hold a census in December?”
The book answer is, “Inkosi, the rules of the game are only in times of calm you can run a successful census.”
He says: “ I see. But what is the true population of Gauteng? If you measured it in December, then the Eastern Cape would receive its appropriate share of revenue, is that not so?” This was met with laughter.
Another such moment was my presentation of the Census results. Two professors who were on the Statistics Council were on record protesting the release of the results of the 2011 Census. They wanted the release to be delayed so that they could beat the numbers until they cried.
I told them where to jump off and released the results based on all quality measures and statutory requirements demanded by the office of the statistician-general and the Statistics Council per the Statistics Act.
I then started the journey of presenting results in provinces. A birdie had said the Western Cape would object and raise issues about their results.
Then-Premier Hellen Zille asked an interesting question after the presentation. Her excitement was quite palpable, more like a child waking up to toys on Christmas morning.
“So, statistician-general, you are saying the Western Cape grew the fastest in population numbers, and you are also saying, despite this, the Western Cape increased the proportion of people who accessed the services the best across all nine provinces? Am I correct?,” Zille said.
I replied: “The people, through these numbers, say so, Premier Zille.”
You could see the ear-to-ear smile of content in the evidence, but, more importantly, the calculus was directed towards the 2014 national elections, without a doubt.
As for the protesting professors, a later release of a longitudinal study from the surveillance sites run by Wits confirmed the authenticity of fertility levels as observed in the census, which was a major bone of contention.
Technical matters are complex, but the questions of politicians are complicated.
Dr Pali Lehohla is the director of the Economic Modelling Academy, a Professor of Practice at the University of Johannesburg, a Research Associate at Oxford University, a board member of Institute for Economic Justice at Wits and a distinguished Alumni of the University of Ghana. He is the former Statistician-General of South Africa.
BUSINESS REPORT