2010 Census Privacy-Protected Microdata for Disclosure Avoidance System Error Analysis

A set of 26 microdata files (1 seed and 25 replicates) intended for estimating the magnitude of error introduced by the 2020 Census Disclosure Avoidance System. The files were created by the U.S. Census Bureau using the 2010 Census Edited File and the TopDown Algorithm to produce Privacy Protected Microdata Files (PPMFs). They contain all Person and Housing Unit attributes necessary to produce Redistricting and Demographic and Housing Characteristics publications for the United States and Puerto Rico.

Use Cases

• Estimating confidence intervals for 2020 Census statistics based on the Approximate Monte Carlo method described.
• Analyzing the scale of errors introduced by the Disclosure Avoidance System based on comparisons between the seed PPMF0 and its replicates.
• Studying demographic and housing characteristics at the Census Block level for the United States and Puerto Rico.
• Researching the application of differential privacy in official statistics using the provided privacy-protected microdata.

Strengths

• Includes 26 distinct microdata files (1 seed PPMF0 and 25 replicates) for robust error analysis.
• Contains geographic detail down to the Census Block level for the United States and Puerto Rico.
• Files are cleared for public dissemination by the U.S. Census Bureau Disclosure Review Board (CBDRB-FY24-DSEP-0002).

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Data on Detailed Race (e.g., Cambodian) is not included, only Major Race categories.

Provenance

Source

United States Census Bureau

Collection Method

Generated by executing the 2020 Decennial Census Disclosure Avoidance System TopDown Algorithm on the confidential 2010 Census Edited File.

Time Range

Based on the 2010 Census.

Freshness

Last update date is unknown; freshness unverified.

Geography

United States and Puerto Rico.