STATISTICAL COMMISSION and COMMISSION OF THE ECONOMIC COMMISSION FOR EUROPE EUROPEAN COMMUNITIES CONFERENCE OF EUROPEAN STATISTICIANS EUROSTAT

A major concern of statistical agencies is to protect the confidential information contained in data sets that have to be released. The problem is not trivial since the agency has to find an optimal balance between the needs of the users and the needs of the data providers. Ideally the statistical agency should provide maximum information to the users, preserving the privacy of the individual entities represented in the data. The sub-field of statistic concern with such a problem is usually referred to as statistical data protection. In this paper I present a decision-theoretic approach to data disclosure problems. The approach is innovative because (i) it offers a theoretical framework to develop optimality criteria for the choice of the best form of data release, (ii) it recognizes the different perspectives of the statistical agency and of the users of the data in assessing the extent of disclosure and the quality of the users inference (data utility) associated with different forms of data release. This leads to new measures of the extent of disclosure and data utility and a new interpretation of the notion of inferential disclosure currently used in statistical confidentiality.