National Getis-Ord Gi*statistics datasets for select populations, 2019-2023 American community survey 5-year estimates

TL;DR

This paper provides precomputed segregation statistics for U.S. populations from 2019 to 2023 to help researchers study local-level racial and ethnic segregation.

Contribution

The paper introduces precomputed Gi*statistics and R code for calculating them, making segregation analysis more accessible.

Findings

Gi*statistics for the U.S. from 2019–2023 American Community Survey data are provided for multiple population groups.

R code is shared to calculate Gi*statistics for years prior to 2023, enabling broader analysis.

Abstract

Racial or ethnic residential segregation is the physical separation of individuals into different neighborhoods by race or ethnicity. This selective sorting process has a large impact on the social, economic, political, and health outcomes of populations, with more marginalized populations having worse outcomes. Segregation measures seek to quantify this process and characterize how different population groups are spread across geographical areas. Residential segregation has traditionally been measured at the metropolitan or county level, but a growing number of population health studies have quantified it at the neighborhood or “local” level. Many such studies have used racial or ethnic composition as a proxy for local residential segregation. However, racial or ethnic composition does not take into account the larger context surrounding the neighborhood individuals live in The Gi*statistic is the most commonly used spatial local segregation measure in the literature. However, it is hard and time consuming to compute, potentially discouraging researchers from its use. Thus, we provide Gi*statistics for the entire US for 2019–2023 American Community Survey 5-year estimates, for numerous population groups. We also provide R code that can be used to calculate the Gi*statistic for years before 2023.

Figures