Employer-Provided Health Insurance Costs and the Demand for Older Workers
by Owen Davis, New School for Social Research
The aging of the U.S. workforce makes the employment prospects of older workers a top public policy concern. While researchers have devoted substantial attention to the drivers of older workers’ labor supply, far less is known about demand for older workers (Allen, 2019). Employers see older workers as offering positive qualities such as know-how and people skills (Munnell et al., 2006). Yet older workers also face age discrimination (Lahey, 2008; Neumark, 2021) and the perception that they cost more (Munnell et al., 2006).
One suspected barrier to hiring and retaining older workers is health costs. The costs of employer-provided health benefits have grown sharply in recent decades and are now equivalent to about to 12% of total money wages (BLS, 2021). Health expenses borne by employers rise nonlinearly with employee age, peaking between ages 55 to 64 (Burtless and Koepcke, 2018). This increases the marginal cost of employing older workers, all else equal. Although economists and policy analysts have long expressed concern that the system of employer-sponsored health insurance weakens demand for older workers (e.g., Rappaport and Morrison, 1984), the topic has attracted relatively little empirical analysis. Numerous open questions remain: Do higher health costs reduce older workers’ hiring prospects and job security? Do older workers’ wages fall—or fall more—when premiums rise? Do employers adjust along other margins, such as the type or generosity of health plans?
In this study, I will examine the effects of rising health insurance costs on older workers’ labor market outcomes in the U.S. using matched employer-employee administrative records, 1996-2014. By enhancing the Linked Employer Household Dynamics dataset with health insurance premium data from mandatory Form 5500 filings on employee benefits, I will create a first-of-its-kind dataset that includes nationwide firm-level data capturing the type of health insurance offered, per capita health premiums, employee contributions, employment growth, age composition, and workers’ earnings and labor market transitions.
With this novel dataset, I will test the impacts of rising health premiums on the hiring, retention, and earnings of older workers in comparison with their younger coworkers. The richness of the administrative data allows me to control for state and local variation, industry heterogeneity, and firm revenue shocks. To address potential endogeneity in premium growth, I will exploit exogenous variation in insurance costs stemming from hospital mergers.
A supplemental analysis makes use of a new crosswalk linking the employers of respondents in the Health and Retirement Study to Census administrative records. While premiums in this part of the study will be imputed, the analysis could help validate and contextualize the findings from the broader analysis. In particular, detailed questions into the employment transitions and workplace experiences of HRS respondents may provide an additional window into the effects of health costs on older workers’ labor market outcomes.