Statistical Confidentiality and the Dissemination of Restricted-Access Integrated Census Microdata Extracts: The Case of Kenya, 1969-1999

Agnes A Odinga, Robert McCaa

Research output: Contribution to conferencePaper

Abstract

Kenya has one of the richest microdata collections in Africa and, indeed,the world. Beginning in 1969 the Central Bureau of Statistics has not only conducted censuses at regular decennial intervals but also conserved the complete microdata for the two most recent censuses. Before May 2002, when the IPUMS released five percent samples of the censuses of 1989 and 1999, this valuable trove was not readily available to scholars or public policy-makers. Access has been the principal obstacle, not only in the case of Kenya, but for many other countries around the world. An important step in providing broader, if still restricted accessand reaping the benefits to be gleaned from these valuable sourcesis to ensure that the data are anonymized according to the requirements of the corresponding official statistical agency. The IPUMS International project, in cooperation with a group of National Statistical Agencies in Europe, the Americas, Asia , and Africa, is acquiring anonymized census samples of individuals and households from official statistical agencies, and, for others, constructing anonymized samples. This paper summarizes statistical confidentiality issues and, then as a example, discusses the requirements for release of five percent samples drawn from the 1989 and 1999 censuses of Kenya. The results are promising. Of the thirty-six person variables in the 1989 census microdata, five geographic variables are suppressed entirely (because they report finely detailed information on place of residence), and one social variable are suppressed: tribe/ethnicity. For the 1999 sample, religion is added to the list of suppressed variables. While suppression of these variables may disappoint purists who demand total access to the original data, this approach strikes a balance between access and statistical confidentiality, sacrificing some degree of detail to safeguard statistical confidentiality to a maximum, yet still make it possible for researchers to gain access to the vast bulk of the Kenyan census microdata.
Original languageEnglish (US)
StatePublished - 2002

Fingerprint Dive into the research topics of 'Statistical Confidentiality and the Dissemination of Restricted-Access Integrated Census Microdata Extracts: The Case of Kenya, 1969-1999'. Together they form a unique fingerprint.

Cite this