PHSafe: Disclosure Avoidance for the 2020 Census Supplemental Demographic and Housing Characteristics File (S-DHC)

TL;DR

This paper presents PHSafe, a differential privacy-based algorithm that adds noise to census data to protect individual privacy while maintaining data utility, specifically applied to the 2020 Census S-DHC.

Contribution

The paper introduces PHSafe, a novel privacy-preserving algorithm based on discrete Gaussian noise, with theoretical privacy guarantees and practical implementation details.

Findings

PHSafe satisfies zero-concentrated differential privacy.

The algorithm effectively balances privacy and data utility.

Implementation on Tumult Analytics demonstrated practical viability.

Abstract

This article describes the disclosure avoidance algorithm that the U.S. Census Bureau used to protect the 2020 Census Supplemental Demographic and Housing Characteristics File (S-DHC). The tabulations contain statistics of counts of U.S. persons living in certain types of households, including averages. The article describes the PHSafe algorithm, which is based on adding noise drawn from a discrete Gaussian distribution to the statistics of interest. We prove that the algorithm satisfies a well-studied variant of differential privacy, called zero-concentrated differential privacy. We then describe how the algorithm was implemented on Tumult Analytics and briefly outline the parameterization and tuning of the algorithm.