Comparative National Elections Project, South Africa 2015
Public Opinion Survey
This dataset is the election survey conducted in South Africa by Robert Mattes of the Democracy in Africa Research Unit at the University of Cape Town in 2015. The survey collects data using standard questions from two international election studies, the Comparative National Elections Survey (CNEP), and the Comparative Study of Electoral Systems (CSES). The Comparative National Elections Survey is coordinated by the Mershon Center for International Security Studies at Ohio State University (https://u.osu.edu/cnep/). The Comparative Study of Electoral Systems is a collaborative program of research among election study teams from around the world, run by the Center for Political Studies and GESIS, Leibniz Institute for the Social Sciences, in Germany, and the University of Michigan in the US (http://www.cses.org/). The South African study includes additional questions. The study and the earlier 2004 CNEP for South Africa are part of a series of South African surveys conducted by DARU, called the South African National Election Study.
Kind of Data
Sample survey data
Unit of Analysis
v1: Edited, anonymised data for public distribution
This version of the dataset was completed in 2016.
CNEP surveys collect basic demographic data, and data on personal discussion networks, media use, political information sources, socio-political values and attitudes, and voting behavior. CSES surveys collectdata on distributional politics and social protection, mobilization, voter preferences, party and candidate contact, and personal contacts intended to influence voter choice.
This survey has national coverage.
The lowest level of geographic aggregation covered by the data is settlement type.
The universe of the study is citizens in South Africa.
Producers and sponsors
Democracy in Africa Research Unit
University of Cape Town
Open Society Foundation for South Africa
South African National Research Foundation
Mershon Center for International Security Studies, Ohio State University
Center for Political Studies, Institute for Social Research, University of Michigan
The survey used a random, nationally representative, stratified, area probability cluster sample. Primary sampling units were census enumerator areas (EAs) selected as a random sample, with probability proportionate to population size. All EAs were stratified by 1) Province, 2) Urban/Rural and 3) Race. Within each EA, a skip interval of 10 dwellings to select a household was used. That is, walking in a designated direction away from the start point, selecting the 10th household for the first interview, counting dwellings on both the right and the left (and starting with those on the right if they are opposite each other). Once the household was chosen, the interviewer randomly selected an individual respondent within the household to be interviewed (altering gender quota). The total number of household in the sample was 1300.
Deviations from the Sample Design
If the household was vacant, if the household refused to participate, if the selected person refused to be interviewed, or if the selected respondent is not available after two callbacks, interviewers were instructed to move to the next house in the walk pattern (i.e. every tenth house). They were not permitted to substitute within a household
The response rate for the survey that the CSES Module appeared in was 34%.
Two types of weights are included in the data: Those benchmarked to South Africa's 2014 midyear population estimates, and those benchmarked to sample size n=1300. People living in Northern Cape province, and coloured and Indian race group members were oversampled, and then weighted downward. Other variables (e.g. household size) were also used to adjust for any discrepancies between the original and realized sample. Weights use information on Urban/rural, race, age, total people in the household and total people over the age of 18.
Dates of Data Collection
Data Collection Mode
Ethics approval for data collection for the CNEP South Africa 2015 study was granted by the Humanities Ethics Review Committee at the University of Cape Town on the 23rd of January 2015.
Data Collection Notes
Citizen Surveys undertook data collection for the survey.
The survey collected data using core questionnaires from both the Comparative National Elections Project and the Comparative Study of Electoral Systems, as well as a separate questionnaire developed by DARU.
The data was checked and cleaned by the original team at the Democracy in Africa Research Unit. DataFirst undertook further cleaning on the data, including the consolidation of data documentation.
Democracy in Africa Research Unit (DARU). Comparative National Elections Project South Africa 2015, fourth phase [dataset]. Version 1. Cape Town: DARU [producer], 2016. Cape Town: DataFirst [distributor], 2017. https://doi.org/10.25828/fb7s-zj78