The IDASA Opinion Survey was a series of opinion polls conducted prior to the 1999 election. They aimed to provide as complete a picture as possible of South African's views of the political, social and economic developments in the country since 1994. The survey was run by IDASA, Markinor and the SABC, along with the Electoral Institute of South Africa to increase credibility of the results and to utilise the best range of data collection and analysis expertise available.
Several key findings of the surveys had important impacts on the electoral process: The surveys confirmed the existence of a large proportion of citizens who lacked the correct identity documents to register. The increased public debate after the release of these results was followed by an extension of the registration process as well as the creation of Temporary Registration and an intensified public advertising campaign. The results helped identify the issues that ordinary voters think are most important (ie job creation, crime-reduction, education, housing, the economy and health care). This helped re-orient the way that the media covered the campaign and focussed government and opposition attention on these key issues. The surveys also revealed the existence of quite a sophisticated electorate and who differentiate between a range of dimensions of political performance and who look to real-world events such as the economy, government performance, and how the country is doing to help them make political choices.
Kind of Data
Sample survey data
Unit of Analysis
Households and individuals
v1.1: Edited, anonymised data for public distribution
Version 1 (the original dataset) was missing questionnaires for waves 2, 3 and 4 of this dataset. These were subsequently sourced from Markinor, the fieldwork company, and provided to DataFirst.
Version 1.1 of this dataset (this current version) includes these questionnaires.
The survey recorded South African's views of the political, social and economic developments in the country since 1994. The polls covered key issues related to the conduct of free and fair elections, voter participation, and other economic, political and partisan trends.
The survey has national coverage.
The lowest level of geographic aggregation for the data is metro area.
The universe for the survey was all adult South Africans who would be of voting age in 1999.
Producers and sponsors
Institute for Democracy in South Africa
Electoral Institute of South Africa
The sample was drawn using a multi-stage, area stratified probability sampling methodology. The sample was stratified by province, population group and community size (metro, city, large town, small town, village and rural). The metropolitan areas were also stratified into "Formal" and "Informal" settlements, to ensure good representation. All qualifying members of a household were listed and one respondent selected randomly. Three attempts were made to interview this person, before substitution with another randomly selected household. 14% of originally selected respondents refused to be interviewed. The sample was representative of the universe from which it was selected. Results can, therefore, be projected to the universe (South Africa as a whole).
Dates of Data Collection
Survey part 1
Survey part 2
Data Collection Mode
Data Collection Notes
Fieldwork for part one of the survey was conducted from 1 to 30 September 1998. Fieldwork for part 2 of the survey was conducted from 5 October to 11 November 1998.
Institute for Democracy in South Africa (IDASA). IDASA Opinion Survey 1998 [dataset]. Version 1.1. Cape Town: IDASA [producer], 1999. Cape Town: DataFirst [distributor], 2012. DOI: https://doi.org/10.25828/fg8x-je72