Other Research

Microdata Review 2016

This study was undertaken for the Department of Planning, Monitoring and Evaluation (DPME) at the Presidency, with funding from the PSPPD programme.

The objective of this study was to raise awareness amongst the research and policy community in South Africa of the multitude of valuable microdata sources in existence in the country. The study covered survey, census and administrative microdata. The study was designed to update an earlier report that had been published in 2007.

With regards to survey and census data, the research team sought to compile a comprehensive list of datasets that are available in the public domain or that can be requested from the data owning organisation. Selected items of metadata were also reproduced for each survey and census dataset and users were directed to the relevant source material should users require more detail information or wish to request access to a particular dataset. Much of the metadata for survey and census datasets was reproduced from the main national data repositories, such as DataFirst and NESSTAR, and referenced accordingly.

With regards to administrative data, the research team were required to adopt a different approach due to the general lack of publicly available information concerning administrative data in South Africa. The research team therefore established a number of lines of personal communication with key data experts across selected government departments in order to discuss South African administrative data. This process of communication highlighted a number of valuable administrative datasets that are currently being collected for operational purposes and the extent to which these datasets are being utilised for research purposes. Three case study administrative datasets were identified: SOCPEN data on social grant recipients (from SASSA); SA-SAMS and LURITS data on school-age pupils (from DBE); and recorded crime data (from SAPS). Each case study dataset is discussed in further depth within the project report. The authors are grateful to SASSA for providing a copy of the SOCPEN database for the purpose of this project which enabled a series of descriptive statistics to be presented illustrating the database structure and content.

The main recommendation from the project was that the process of documenting microdata resources should be formalised as an ongoing strategic priority rather than a process of only periodic updating. This is particularly the case in respect of administrative data, where success depends on building and maintaining personal contacts with data experts across government.