Deindustrialization and Disability: Social-Structural Predictors of Physical Disabilities among Working-Age Americans

Abstract

Objectives: Concerns about rising chronic illness and recent declines in life expectancy have led to a surge in research examining physical disabilities among working-age Americans. I extended this research in a few ways. First, I examined trends in disabilities among working-age Americans over time to determine whether physical disabilities have continued to increase. Second, I examined whether the changing labor market due to deindustrialization was associated with physical disabilities among working-age Americans at the county-level. Third, because states vary in the robustness and comprehensives of public assistance benefits, I examined whether state-level characteristics and public assistance benefits moderated these associations. Fourth and finally, I examined whether race and ethnicity moderated the association between labor market changes and disability. Methods: I used IPUMS National Health Interview Survey (IPUMS-NHIS) data linked to restricted NHIS data to obtain county-level data on disability and other important sociodemographic characteristics. Because labor markets vary within states, I used county-level data. I linked these datasets to the U.S. Census Bureau's County Business Pattern data to examine how labor market changes were associated with disability over time. I also linked these datasets to the University of Kentucky Center for Poverty Research's National Welfare Data to examine whether state-level characteristics and public benefits affect the strength of these associations. I examined the age-adjusted rate of physical disabilities at the national level over a long period to determine whether disabilities have continued to increase. Next, I used a state and county fixed effects OLS regression to examine the associations between labor market changes and disability at the county-level. I then incorporated a variety of sociodemographic characteristics aggregated at the county-level to examine whether county-level characteristics confounded the associations. Finally, I incorporated interaction effects for state-level characteristics, public assistance benefits, and race and ethnicity to determine whether these variables moderated the associations of interest. Results: Some physical disabilities among working-age Americans remained stable over time after adjusting for the age composition of the population, and others increased. Of note was the increase in ADL limitations, the average number of functional limitations, and disabilities among younger Americans. Labor market conditions, and to a lesser extent, labor market polarization was associated with physical disabilities among working-age Americans. A one percent increase in goods-producing industries was associated with lower rates of disability at the county level, and a one percent increase in service-providing industries was associated with higher rates of disability, net of sociodemographic characteristics. The state-level interactions effects indicated that higher goods-producing industries paired with higher unemployment was associated with higher rates of disability whereas higher goods-producing industries paired with higher minimum wage was associated with lower rates of disability. Finally, race and ethnicity did not moderate the association between labor market conditions/changes and disability at the county level. Discussion: The increase in more severe disabilities, the average number of comorbid disabilities, and disabilities among younger Americans demonstrates a concerning trend that may affect the strength of the labor market and the health and wellbeing of American workers. I found that labor market conditions, and to a lesser extent labor market polarization, was associated with physical disabilities among working-age Americans net of sociodemographic characteristics. Interventions to strengthen the labor market should focus on maintaining goods-producing industries in the U.S., increasing minimum wage, and implementing job-training programs specifically for people with disabilities. Future research should examine the potential mechanisms through which service-sector jobs are associated with disability. Finally, I found that race and ethnicity did not moderate the association between labor market conditions/polarization and disability. This result may be due to low statistical power associated with difficultly examining between group variations at the county-level. Future research might examine racial disparities in these associations at the state or metropolitan statistical areas level to achieve higher statistical power.