The U.S. Census Bureau today released the Noisy Measurement File associated with the 2020 Census Demographic and Housing Characteristics File (DHC). The 2020 Noisy Measurement Files are considered research-based statistical products and should not be considered the official 2020 Census counts.

A Noisy Measurement File is the intermediate output of the new 2020 Disclosure Avoidance System’s TopDown Algorithm (TDA). The TDA generates noisy measurements when it applies differentially private noise to each of the tabulations from the confidential data. Because the noise can result in internal and hierarchical data inconsistencies, the TDA completes a final step called “post-processing” to mitigate the inconsistencies. That post-processing corrects those inconsistencies and improves accuracy for lower-level geographies before the tables or Privacy-Protected Microdata Files (PPMFs) are published.  

Noisy measurements help data users analyze the impact of the new 2020 Census Disclosure Avoidance System. The Census Bureau previously released noisy measurements using comparable 2010 redistricting and DHC census data as part of the Demonstration Data Products Suite. In June, we released the first set of noisy measurements for 2020 Census data—the 2020 Census Redistricting Data (P.L. 94-171). Today’s publication completes the release of noisy measurement files for the 2020 Census.

Learn more about how to access and work with noisy measurement files in this recorded webinar from the June 15 Demonstration Data Products Suite release.

Given their large size, noisy measurement files are accessible via external data platforms. In addition to Inter-university Consortium for Political and Social Research (ICPSR) and Harvard DataVerse, we have added a third location for noisy measurement file access: the Registry of OpenData on AWS. Access all available platforms via the link below.

A new paper by Census Bureau researchers published in the Proceedings of the National Academy of Sciences lays out a framework for analyzing disclosure risk assessment methodologies and evaluates current options based on that framework.

As always, we want to ensure our products are meeting your needs. We welcome your feedback and questions at 2020DAS@census.gov.