This paper reviews different approaches to estimating differences in economic outcomes across groups defined by race, with a focus on whether and how covariates such as education are incorporated in these analyses. My review focuses on the economics literature, but also draws lessons from research on health disparities. Using data from the Current Population Survey and the National Health Interview Survey, I present two examples of estimating Black-white differences in retiree health insurance and cost-related problems with access to medical care. The examples illustrate the importance of estimating Black-white gaps separately for men and for women, whether or not controls for education and other characteristics are included in the model.
Estimating Racial Disparities in Economic Outcomes: An Application to Employer-sponsored Retiree Health Insurance and Access to Care
Published: 2023
Abstract
Downloads
Key Findings
- The question of what covariates to include in a model of differences across groups defined by race is not new.
- There is no generally accepted answer to what covariates to include. Rather, the research question to be answered should dictate the choice of covariates.
- Estimating separate models of Black-white differences for men and for women yields more interesting empirical results than estimating a single model for both men and women. The nature of the racial gaps varies by gender.