Association for Public Policy Analysis & Management

Lower socioeconomic status (measured by baseline wages or educational attainment) is a well-established risk factor for longer disability duration following occupational injury. Long-term work disability and receipt of federal disability benefits are also more common among workers with low educational attainment. Little is known, however, about how differences between low-wage and high-wage workers in return to work after workplace injury may contribute to the socioeconomic gradient in disability risk.

This study examines how the risks of permanent disability and labor force exit following workplace injury vary with wages, worker demographics, injury type, and employer characteristics (including industry, firm size, turnover, and employment growth rates). Using an administrative dataset containing the population of workers' compensation claims in California linked to longitudinal earnings histories and rich employer characteristics, we compare the risk of adverse post-injury outcomes across the wage distribution and across different work settings for demographically similar workers with similar injuries.

To obtain credible estimates of earnings losses, we used matching to construct a comparison group of control workers for each injured worker: the control workers are workers at the same employer as the injured worker with similar job tenure and earnings to the injured worker, but who do not file a workers' compensation claim. This approach provides a credible counterfactual for the post-injury employment and earnings of each injured worker, enabling us to estimate the causal impact of injury in isolation from confounding factors such as firm-level unobservables, labor market conditions, and reversion to the mean as workers transition out of employment for reasons other than disability. We have used these methods in prior work to estimate average earnings losses and the relationship between injury severity and earnings losses; the contribution of this study is to analyze how post-injury outcomes (including permanent disability, earnings loss, employment, and job separation) vary with workers' economic status prior to injury, and which employer characteristics might moderate the socioeconomic gradient in post-injury outcomes.

Our preliminary findings confirm the existence of a strong socioeconomic gradient in the severity of economic losses following workplace injury: higher pre-injury earnings are associated with better employment and return-to-work outcomes three years after the injury across the entire earnings distribution. However, it is likely that some portion of the socioeconomic gradient in post-injury outcomes is explained by the fact that low-wage workers have different characteristics (e.g., younger and more female) and are concentrated in certain industries (e.g., health care, retail, and accommodations and food services) or in certain types of employers (e.g., those with higher turnover rates).

We are currently estimating regression models and using reweighting methods to test whether the wage gradient in outcomes is driven by these compositional differences, as well as exploring whether there are particular occupations, groups of employers, or types of injuries where low-wage workers have better or worse outcomes relative to observably similar high-wage workers. These findings will help policymakers and researchers set priorities for future research and improve the targeting of return-to-work interventions intended to improve post-injury outcomes for low-wage workers.