Disclosure risk estimation in household sample surveys

Introduction: This article aims to estimate the disclosure risk - the probability of discovering the identity of the respondent unit in a disseminated database - from the Continuous National Household Sample Survey public use file. Method: The estimation was carried out using a probabilistic model, more specifically, the Benedetti and Franconi model, colloquially called the "Italian approach". Results: It was observed that although most records have a very low disclosure risk, there are some that require greater attention regarding their dissemination, due to the high risk. This occurs even for more aggregated geographical dissemination areas. Conclusions: The estimates of disclosure risk presented indicate that Statistical Disclosure Control techniques are fundamental tools to help information producers in their mission to guarantee the confidentiality of information.