Impacts of Census Differential Privacy for Small-Area Disease Mapping to Monitor Health Inequities

TL;DR

This study evaluates how the US Census Bureau's new differential privacy system affects small-area health data accuracy, finding that while some errors occur, overall health inequity patterns remain reliable, especially in newer DAS versions.

Contribution

It provides the first comprehensive assessment of differential privacy impacts on health inequality metrics using real and simulated data.

Findings

Overall health inequity patterns are preserved despite DAS-induced errors.

Newer DAS versions reduce bias in mortality rate estimates for Black populations.

Errors are larger in early DAS versions but improve in recent updates.

Abstract

The US Census Bureau will implement a new privacy-preserving disclosure avoidance system (DAS), which includes application of differential privacy, on the public-release 2020 census data. There are concerns that the DAS may bias small-area and demographically-stratified population counts, which play a critical role in public health research and policy, serving as denominators in estimation of disease/mortality rates. Employing three DAS demonstration products, we quantify errors attributable to reliance on DAS-protected denominators in standard small-area disease mapping models for characterizing health inequities. We conduct simulation studies and real data analyses of inequities in premature mortality at the census tract level in Massachusetts. Results show that overall patterns of inequity by racialized group and economic deprivation level are not compromised by the DAS. While early versions of DAS induce errors in mortality rate estimation that are larger for Black than for non-Hispanic white populations, this issue is ameliorated in newer DAS versions.