/ 14 October 2010

Census 2011: How it works

The census questionnaire for the 2011 South African census has already been drawn up and is currently being beta tested by Statistics South Africa (Stats SA). Questions include things like the names, ages and educational level of each person in a household, and whether they have any disabilities or chronic illnesses. Some questions are geared at finding out more about maternal and infant mortality.

It will also ask about the type of housing people live in and what sort of access to water, sanitation and refuse removal they have. The questions will try to gauge the economic status of each household and will ask about involvement in agricultural activities, even if it is on a small scale, and whether people keep livestock such as chickens or sheep.

Over the next 12 months, the organisation will recruit, employ and train 120 000 field workers and 30 000 supervisors. This will be done in each province according to Stats SA’s HR policies.

On the evening of October 9 and 10 next year, field workers will fan out across the country to gather information from all South Africans in the country. This will cover:

  • 14-million households.
  • institutions where people reside, including hospitals, prisons, army barracks and university residences.
  • transit points such as border posts, ports and airports.
  • areas where homeless people are known to reside.

Undercounting
Field workers will be hard-pressed to avoid undercounting during this census. There are many reasons for undercounting during a census. Some people may be out of the country or in transit. Others who live in “hard-to-count” areas may simply not be available on the days set aside for the census-taking.

It’s not just deep rural areas that may be deemed hard to count. Census-takers often have difficulty accessing households in gated communities or counting people from highly mobile populations. People who do not have valid work permits or who fear censure from the police or revenue service may also deliberately avoid the census-takers. The homeless and the very young are also among those who are often miscounted.

Census-takers can be recognised by their special uniforms, yellow bibs and a special identification card. People will be able to call a helpline to check the field worker’s ID, which will be stored in a database. In addition, Stats SA will distribute posters of the field worker designated to work in a specific area so that they can be easily identified.

Members of the public can choose to fill in the questionnaire themselves. Alternatively, they can be administered by the census-taker.

Census taking is governed by the Statistics Act of 1999, which states that all information provided to census-takers must be held in the strictest confidence. Complying with census-takers is also obligatory under the law. Anyone who refuses entry to a census-taker or obstructs the work of a census-taker is also liable for a fine of up to R10 000 or six months in prison.

Mini-survey
The information is aggregated to statistics and is stripped of all data that would allow individuals, organisations or businesses to be identified before it is released into the public domain or to other state institutions.

After that, a two-week period is allowed in which every questionnaire that is filled in is returned to Stats SA and all staff are withdrawn from the field.

Thereafter, an independent quality and methodology unit will return to the field to carry out a mini-survey in each province to check whether the information is accurate. It is at this stage that undercount figures are often uncovered.

Meanwhile, the questionnaires are scanned and put through a vetting process in which errors are also uncovered and either discarded or corrected. For example, a two-year-old child may have been listed as an infant instead of a child. This can be easily corrected.

Over the next year, the data is analysed and the final census produced.