A methodology for calculating comparable wages: an application to direct care workers

This paper develops a methodology for estimating comparable wages and applies the methodology to home health and personal care aides. In contrast to regression-based approaches that use individual-level data, our methodology exploits published occupational mean wage and employment estimates from the Bureau of Labor Statistics’ Occupational Employment and Wage Statistics program and an extensive set of job skills, tasks, and working conditions from the Occupational Information Network (O*NET). The key to our approach is the construction of occupation-specific proximity weights that capture how similar the required skills and tasks of any occupation are to those of direct care workers. Analysis of the data indicates that in our current application our approach dominates the regression-based approach, and under plausible assumptions, this will also be the case in other applications. We find the estimated wage gap, defined as the percent difference between comparable and observed wages of direct care workers, increased from 2.4% in 2019 to nearly 10% in 2022.