Piloting a Measure of Segregation at the Census Tract Level: Associations with Place and Racial/Ethnic Disparities in Life Expectancy

Abstract

This study considers residential segregation as a critical driver of racial/ethnic health disparities and introduces a proxy measure of segregation that estimates the degree of segregation at the census tract level with a metric capturing the overrepresentation of a racialized/ethnic group in a census tract in relation to that group’s representation at the city level. Using Dallas, Texas as a pilot city, the measure is used to investigate mean life expectancy at birth for relatively overrepresented Hispanic, non-Hispanic white, non-Hispanic Black, and Asian census tracts and examine for significant differences between mean life expectancy in relatively overrepresented census tracts and that group’s mean life expectancy at the state level. Multivariable linear regression analysis was utilized to assess how segregation measured at the census tract level associates with life expectancy across different racialized/ethnic groups, controlling for socioeconomic disparities. This study aimed to expose the need to consider the possibility of neighborhood mechanisms beyond socioeconomic characteristics as an important determinant of health and draw attention to the importance of critically engaging the experience of place in examinations of racial and ethnic health disparities. Multivariable linear regression modeling resulted in significant findings for non-Hispanic Black, non-Hispanic white, and Asian groups, indicating increased census tract-level life expectancy for Black and white residents in highly segregated census tracts and decreased life expectancy for residents of tracts in which the Asian community is overrepresented when compared to state means. Unadjusted models demonstrated socioeconomic inequities between first and fourth quartile census tracts and pointed to the importance of mixed methods in health disparities research and the importance of including the voice of community members to account for places of daily lived experience and people’s relationships with them.