Protecting Census 2021 Origin-Destination Data Using a Combination of Cell-Key Perturbation and Suppression
The UK Office for National Statistics (ONS) is intending to produce outputs involving travel to and from different locations (origins and destinations) in 2021, as they have done for previous Censuses. This data poses a particular challenge for protecting against disclosure risk, as categorising respondents on multiple geographical variables yields very sparse tables. This paper explores the disclosure risk and data utility of one option for protecting this data: applying cell-key perturbation (noise), and suppressing the remaining disclosive values. It finds that these methods provide good protection for the data with considerable loss of utility for outputs at low geographies. Whether this is an acceptable approach will be determined by user feedback.
KeywordsOrigin destination Flow data Cell-key perturbation Suppression
- 1.Fraser, B., Wooton, J.: A proposed method for confidentialising tabular output to protect against differencing. In: Joint UNECE Eurostat Work Session on Statistical Data Confidentiality, Geneva, Switzerland, 9–11 November 2005Google Scholar
- 2.Leaver, V.: Implementing a method for automatically protecting user-defined Census tables. In: Joint UNECE/Eurostat Work Session on Statistical Data Confidentiality, Bilbao, Spain (2009)Google Scholar
- 6.Data Protection Act (1998). http://www.legislation.gov.uk/ukpga/1998/29
- 7.Statistics and Registration Service Act (2007). http://www.legislation.gov.uk/ukpga/2007/18/section/39
- 8.UK Statistics Authority Code of Practice for Official Statistics (2009). https://www.statisticsauthority.gov.uk/wp-content/uploads/2015/12/images-codeofpracticeforofficialstatisticsjanuary2009_tcm97-25306.pdf