Skip to content

Survey on access to African government microdata for social science research

Literature › Works

Lynn Woolfrey (2010). Survey on access to African government microdata for social science research. 2010 IASSIST Conference: Social data and social networking: connecting social science communities across the globe, Cornell University, Ithaca, NY, June 1-4, 2010. DataFirst, University of Cape Town.
Type conference paper
Scope Comparative and methods
Method survey fieldwork
Period studied 2009–2010
Regions Africa, South Africa
Themes methods data sources
Citation key woolfrey2010

Argument

Most microdata collected in Africa is official data that national statistics offices (NSOs) use in-house and release only as aggregates and reports. A March-April 2009 survey of NSOs in 53 African countries tested whether researchers could obtain microdata on request. Only four NSOs (Botswana, Kenya, Nigeria, South Africa) provided good access. The obstacles named are weak government commitment and policy, confidentiality concerns, poor infrastructure, skills shortages and poor producer-user communication. The author concludes that ministerial-level awareness is needed so that microdata is treated as a national asset and curated by dedicated archives or research data centres.

Evidence and method

Original survey of NSOs in 53 African countries, March-April 2009. Requests were sent by email, fax or online form in the language of each country, with contact details taken from NSO websites. Responses were classified into eight categories. The slides also draw on websites of existing data-sharing projects (IPUMS, NADA, NESSTAR, DataFirst) and one citation (James 2001).

Findings

  • Most African microdata is official data held by NSOs and used mainly in-house. (p. 4)

    Most microdata collected in Africa is official data Microdata used in-house by govt statisticians

  • The survey covered NSOs across the continent, using local languages and avoiding personal contacts. (p. 6)

    A request for microdata was sent to NSOs in 53 African countries

  • The largest group of NSOs did not respond at all. (p. 7)

    No response (25) - Algeria, Burundi, Burkina Faso, Cameroon, Cape Verde, Chad, Cote d’Ivoire, Djibouti, Egypt, Equatorial Guinea, Ethiopia

  • Only four NSOs gave good access, including Statistics SA via NESSTAR. (p. 7)

    Good microdata access – NSOs in Botswana and Kenya emailed microdata files and documentation

  • Some NSOs charged fees, with Senegal charging the most. (p. 7)

    Payment required for microdata – Ghana ($320 for panel study data) and Senegal ($902 per dataset) (2)

  • Mauritius required payment plus a visit to the country. (p. 7)

    the Mauritian NSO required payment ($129.45) and a visit to the country to obtain the microdata files (1)

  • Thirteen NSOs had acquired NADA software yet dissemination was still hampered. (p. 10)

    NSOs in 13 African Countries have acquired the NADA application to manage and share their microdata

  • NSOs' colonial origins as underfunded population-statistics offices still shape them. (p. 4)

    Underfunded by colonial administrations - NSOs still reflect this characteristic today

  • In South Africa a few university projects release microdata, but other African countries lack data archives. (p. 3)

    RDCs and data archives have not been established in other African countries

  • Ad hoc decisions and bureaucracy, justified by confidentiality, block data sharing. (p. 11)

    No established data dissemination policies Decisions on access are ad hoc. Sometimes informal contacts will lead to access