The options for expanding access to census microdata were the focus of discussion at a Webex expert group meeting (EGM) convened by ECLAC Subregional Headquarters for the Caribbean in Port of Spain, Trinidad and Tobago on 23rd November 2015, with participation from representatives of national statistical offices (NSOs), representatives of data archives and some United Nations agencies.
The EGM examined how Caribbean census microdata can be made more widely available to researchers. It considered techniques for anonymizing microdata and options for the dissemination of anonymized data sets, including the technical, legal and administrative arrangements which are necessary to protect the confidentiality of individual census records.
The need for training of staff of the NSOs in the subregion in the methods of data anonymization was also emphasized and ECLAC Caribbean stands ready to provide technical assistance to Member and Associate Member Countries in this regard. To achieve wider dissemination of Caribbean census microdata, the EGM recommended greater cooperation among data providers, such as the NSOs, and data archives, such as the Derek Gordon Databank at the Mona Campus of the University of the West Indies.
The term ‘microdata’ refers to data for individual persons, households, or entities. For censuses, microdata constitute information collected on individuals and households. These data sets are usually confidential by law and are thus not disseminated. However, they contain useful information for researchers.
Census microdata can be used for research and planning in areas such as health, fertility, housing, transportation, education, employment, migration and regional development. Therefore, providing controlled access to anonymized census microdata means that the valuable information that is collected can be put to use for the benefit of all.
Census microdata are among the most useful to social researchers because of the detailed information they contain and over the last 25 years, statistical offices worldwide have increasingly sought to meet the demand from researchers for greater access to statistical microdata.
The risk of identity disclosure, however, constitutes a major obstacle to the release of census microdata. This risk is amplified for Caribbean countries because of their small populations. To address this, statistical techniques have been developed to anonymize microdata, thus enabling their use in social research and policy making.
Caribbean countries and overseas territories generally publish traditional census reports, and some have developed online data tabulation tools using ECLAC’s REDATAM software. However, while Caribbean statisticians have agreed on the principle of microdata dissemination, progress in the release of actual microdata has been slow. The outcome of the EGM is expected to provide a framework for Caribbean countries in anonymizing their census microdata and in disseminating the anonymized data set for use by researchers.
