Censuses are principal means of collecting basic population and housing statistics required for social and economic development, policy interventions, their implementation and evaluation. The Post-Apartheid South African government has conducted three Censuses, in 1996, 2001 and 2011.
Kind of Data
Census/enumeration data [cen]
Unit of Analysis
The units of analysis for the South Africa Census 2011 were households and individuals
v2: Edited, anonymised dataset for public distribution
Version 1 of the South African Census 2011 10% sample dataset was provided on CD to DataFirst by Statistics South Africa in May 2014.
Version 1.1 of the South African Census 2011 10% sample was subsequently released on the Statistics South Africa website and it included value labels which were not available in version 1.
Version 1.1 of the SA Census 2011 10% sample dataset did not include the following variables.
H5 - Estimated value of the property
H6 - age of the property
p29 - Industry
p30 - Occupation
p35 - Total surviving and living in the household [children]
p36 - Total surviving and living elsewhere [children]
p37 - Total children no longer alive
There is no explanation in the metadata for the missing H5 and H6 variables. The metadata states that P29 and P30 (Occupation and Industry variables have been excluded due to incomplete coding (p77).
Data for questions P-35 to P-37 were excluded due to poor data quality (p82).
Version 2 is a result of revisions done to the "SA Census 2011 10% Persons" data file that was released by Statistics SA on their website in August 2015. In line with our policy, we give a new release a full version number. The revised version of the "SA Census 2011 10% Persons" data file includes data on employment, occupation and industry which were not available in version 1.1 (see version 1.1 notes above). It is this version of the "Persons" data file, which we have called version 2 which we now make available on this site. Version 2 of the "Persons" data file includes the following nine new variables:
P29A_INDUSTRY - “Industry”
P30A_OCCUPATION - “Occupation”
DERP_INDUSTRY - “Grouped Industry”
DERP_OCCUPATION - “Grouped occupation”
OCCUP_LEV01 - “Occupation major Group”
OCCUP_LEV02 - “Occupation sub-major group”
OCCUP_LEV03 - “Occupation minor group”
OCCUP_LEV04 - “Occupation unit group”
DERP_Sector - “Sector (excludes agriculture from formal and informal sectors)”
Version 2 also includes an additional data file on agricultural households (that is, households involved in agricultural activities) added by Statistics SA in October 2015.
The South African Census 2011 dealt with the following topics:
Household characteristics, including dwellling type, home ownership, household assets, access to services and energy sources;
Individuals' characteristics, including age, population group, language, religion, citizenship, migration, fertility, mortality and economic characteristics of individuals, including employment status and employment activities. Questions on disability were replaced by general health and functioning questions.
The South African Census 2011 has national coverage.
Data at enumeration area is not released with this dataset, to protect respondent confidentiality. A Small Area Layer (SAL) of geography has been created by Statistics SA to fulfil researchers' need for data at more detailed geographic levels. The small area is made up of one or more EA's provided they confirm to certain criteria such as population thresholds, area size, geographical constraints, and land use type. For example in formal areas the population of the small area must be over 300. Enumeration Areas have been combined with neighbouring EAs with the same characteristics to make up these numbers. When the SALs are mapped blank spaces will appear which represent EAs that did not meet the criteria of having the same characteristics as surrounding EAs.
The South African Census 2011 covered every person present in South Africa on Census Night, 9-31 October 2011 including all de jure household members and residents of institutions.
Producers and sponsors
Statistics South Africa
Government of South Africa
The sampling frame for the PES was the complete list of Census 2011 EAs, amounting to 103 576 EAs. The primary sampling units (PSUs) were the Census EAs. The principle for selecting the PES sample is that the EA boundaries for sampled EAs should have well defined boundaries, and these boundaries should correspond with those of Census EAs to allow for item-by-item comparison between the Census and PES records. The stratification and sampling process followed will allow for the provision of estimates at national, provincial, urban (geography type = urban) and non-urban (geography type = farm and traditional) levels, but estimates will only be reliable at national and provincial levels. The sample of 600 EAs was selected and allocated to the provinces based on expected standard errors which were based on those obtained in PES 2001. Populations in institutions (other than Workers' Hostels), floating and homeless individuals were excluded from the PES sample.
The data files in the dataset include Household, Person, and Mortality files. The 10% sample for the Mortality data file was sampled separately and is not the same as the 10% sample for Household file and Person file.
Dates of Data Collection
Data Collection Mode
Data Collection Notes
The enumeration primarily took place over the period 9 October to 31 October 2011.
Three sets of questionnaires were developed for Census 2011:
1. Questionnaire A - the household questionnaire - administed to the population in a household set-up including those households that were found within an institution, such as staff residences
2. Questionnaire B - the population in transit (departing) and those on holiday on reference night (9/10 October 2011). The homeless were also enumerated using this set of questions
3. Questionnaire C - the institutions questionnaire administered to the population in collective living quarters (people who spent census night 9/10 October 2011 at the institution)
A Post-Enumeration Survey was carried out after the census, which used a PES questionnaire.
Estimates of Sampling Error
Comparison of Census 2011 with previous Censuses requires alignment of the data to 2011 municipal boundaries
Questions on disability asked in former censuses were replaced in census 2011 with General health and functioning questions.
Misreporting on general health and functioning for children younger than five years means data for this variable are only profiled for persons five years and older.
The dataset does not have a code list for the “geotype” variable which has 3 values (1,2,3).
Statistics South Africa. South African Census 2011, 10% sample [dataset]. Version 2. Pretoria: Statistics South Africa [producer], 2015. Cape Town: DataFirst [distributor], 2015. DOI: https://doi.org/10.25828/vjy1-tz66