U.S. — The Trump administration released an order titled "Disclosure Avoidance for Statistical Products" on June 4. The order forbids "any use of noise infusion" for statistical products and states that "Coarsening shall be the preferred category of Disclosure Avoidance methods for all statistical products." It further specifies that "Suppression shall be permitted as a last resort, only to be used when coarsening is prohibited by law or would substantially defeat the accuracy or usability of a statistical product."
Noise infusion is a technique that adds random values to a dataset to protect privacy. Coarsening is the process of grouping, rounding, or reporting data in ranges. Suppression involves redacting information or not releasing data entirely.
Beth Jarosz, a senior fellow at Georgetown University's Massive Data Institute and vice president of the Association of Public Data Users, said, "Because 'coarsening' (grouping, rounding, reporting in ranges) and suppression are the only not-prohibited tools named in the order, it means that to keep information safe, the Census Bureau and BEA need to group small things (like small communities or small business types) into larger ones, or they need suppress the data completely." She added, "Small industries may get rolled into bigger industry categories. Small counties may get rolled into county groups or not reported at all." Jarosz noted that when the agencies gather data, they promise respondents confidentiality. "When the Census Bureau and Bureau of Economic Analysis gather data, they promise respondents that they will keep responses confidential," she said. "By taking away tools that those agencies use to protect privacy and confidentiality, people may question whether or not Census and BEA can live up to that promise."
The Population Association of America, Council of Professional Associations on Federal Statistics, Association of Public Data Users, Inter-university Consortium for Political and Social Research, and Association of Population Centers released a joint statement on June 17. The statement asserted, "This order subverts processes developed over decades to foster transparency and public trust and creates a scenario in which there will either be less privacy for our personal information, or less usable data, or both." Steve Pierson, Director of Science Policy for the American Statistics Association, said the order "handcuffs the Census Bureau and the Bureau of Economic Analysis in terms of the techniques they can use for protecting the privacy of respondents."
Lynda Kellam, Leader of the Research Data and Digital Scholarship team at the University of Pennsylvania Libraries, said, "Regarding the datasets that used noise infusion, it is unclear how this policy will impact public access." She added, "The policy is intended to be retroactive, raising concerns that data might be removed, but how that will play out is uncertain." Multiple webpages related to noise infusion and differential privacy on the Census Bureau's website were removed following the order; most of those pages have since been restored.
forum Comments (0)
No comments yet. Be the first to comment.