Issues with Nonparametric Disclosure Risk Assessment

Whenever an agency releases microdata files containing sensitive infor- mation about a sample of individuals, the major concern must be to protect the identity of the participants from what is known as disclosure risk. An intruder, in fact, could match the published file with previously available information and dis- close the identity of the participants. One of the approaches to tackle this type of risk is the statistical quantification of it by means of the estimation of the number of sample uniques that are also population uniques, i.e. the individuals whose risk of being disclosed is the highest. In this paper, we review the literature about paramet- ric and nonparametric estimation of this measure and present new potential research questions.