Abstract
Anonymisation inevitably changes data, so agencies need systematic ways to measure how much analytical quality is lost. This paper proposes quality (information-loss) indicators for statistical disclosure methods that quantify the gap between original and protected data across the statistics that matter to users. The indicators are demonstrated in a detailed case study on the Structure of Earnings Survey, a key European business statistic. Published in the Journal of Official Statistics, the work gives a practical framework for documenting and defending the utility of anonymised official data.
statistical disclosure control data utility quality indicators R