The Census Bureau recently removed several pages on disclosure avoidance from its website. These materials—co-authored by Census Bureau experts and PRB staff—provide essential background for understanding the Bureau’s methods to protect respondent confidentiality in the decennial census and periodic surveys, and what is at stake in current debates about census data policy.
PRB is making some of these materials available here to ensure they remain accessible to researchers, data users, policymakers, and the public. These publications represent one set of resources from a much larger body of disclosure avoidance documentation produced by the Census Bureau—including technical handbooks, webinars, demonstration data, and FAQs—much of which is also at risk of becoming unavailable.
Together, the report and briefs below explain what disclosure avoidance is, why it matters, how the Census Bureau's approach evolved for the 2020 Census, and what the changes mean for different data products. They are intended for a broad audience—from first-time data users to advanced researchers.
Disclosure Avoidance for the 2020 Census: An Introduction
This foundational report provides a comprehensive overview of how and why the Census Bureau applied new disclosure avoidance techniques to the 2020 Census, and the key implications for data users. It covers the history of disclosure avoidance, the shift to differential privacy, and guidance for working with the new data products.
📄 Download PDF
Brief 1: Disclosure Avoidance and the 2020 Census Redistricting Data
The first in the series. Provides key information about how disclosure avoidance was applied to the 2020 Census redistricting data—the data used to redraw congressional and state legislative districts. 📄 Download PDF
Brief 2: Why the Census Bureau Chose Differential Privacy
Explains the Census Bureau's decision to adopt differential privacy as its disclosure avoidance framework for the 2020 Census, including the vulnerabilities of older methods and the results of the Bureau's database reconstruction simulation using 2010 Census data. 📄 Download PDF
Brief 3: Disclosure Avoidance and the 2020 Census: How the TopDown Algorithm Works
A technical explanation of the TopDown Algorithm—the computational system used to implement differential privacy for 2020 Census tabular data products. 📄 Download PDF
Brief 4: Disclosure Avoidance Methods for the Detailed Demographic and Housing Characteristics File A (Detailed DHC-A): How SafeTab-P Works
Describes the SafeTab-P algorithm used to protect detailed race and ethnicity population tabulations in the Detailed DHC-A file. 📄 Download PDF
Brief 5: Disclosure Avoidance and the 2020 Census: How SafeTab-H Works
Describes the SafeTab-H algorithm used to protect detailed race and ethnicity household tabulations in the 2020 Census. 📄 Download PDF
Brief 6: Disclosure Avoidance for the Demographic and Housing Characteristics File (DHC) and Guidance for Data Users
Explains how disclosure avoidance was applied to the DHC file and provides practical guidance for data users on interpreting and working with the protected data. 📄 Download PDF
Brief 7: Disclosure Avoidance and the Supplemental Demographic and Housing Characteristics File (S-DHC): How PHSafe Works
Describes the PHSafe algorithm used to protect the Supplemental DHC file, which provides additional demographic detail at the national and state levels. 📄 Download PDF
These materials were produced through a partnership between the Population Reference Bureau (PRB) and the U.S. Census Bureau's 2020 Census Data Products and Dissemination Team. They are provided here as a public resource.