How Do People with Disabilities Fare in Hard Economic Times? Evidence from the Great Recession

It is often asserted that workers with disabilities are the "last hired, first fired," making them particularly vulnerable to job loss and economic hardship during recessions. This hypothesis is intuitively appealing since employers with a sudden decline in labor demand might selectively retain and hire workers with higher perceived productivity. However, there has been little theoretical or empirical examination of how the employment of people with disabilities varies over the business cycle in comparison to similar workers without disabilities.

We show that, notwithstanding the "last-hired, first-fired" hypothesis, the relative impact of the business cycle on employment for disabled and non-disabled workers is theoretically ambiguous. In addition, anti-discrimination laws (such as the Americans with Disabilities Act) could either amplify or dampen the impact of the business cycle on employment for workers with disabilities. The threat of civil penalties for discriminatory firing or layoffs might protect workers with disabilities from job separations, but the law's prohibition against wage discrimination could also make people with disabilities more attractive targets for layoffs if employers perceive people with disabilities as having lower productivity relative to their wages.

We then test the relationship between disability status and economic outcomes over the Great Recession using unique data from the California workers' compensation system on earnings losses for workers after a disabling injury. Empirical research on this question has been limited by the fact that most datasets that capture both economic outcomes and health status rely on individual self-reports of work-limitations, which could be endogenous to current employment status. Similarly, past studies have used repeated cross-sections to examine unemployment rates but have not been able to compare job separations between workers with and without disability. Our data combine detailed clinical information on the nature and severity of injuries with panel data on earnings and employment for both injured workers and matched control workers with the same job tenure and earnings trajectories. Data on all injuries occurring between 2005-2012 allows us to compare economic outcomes for workers with observationally identical injuries over the business cycle. Finally, we use detailed data on employer and job characteristics to measure trends in earnings losses after an injury independent of compositional changes in the injured worker population.

We find that the Great Recession disproportionately worsened economic outcomes for disabled workers. The earnings of permanently disabled workers relative to their matched control workers fell by 9 percentage points (30%) for those injured during the peak of the recession, with little to no recovery for injuries occurring in the subsequent three years. Although nearly half of this decline can be explained by changes in the composition of claims, including a shift in permanent disability claims to workers with more severe injuries, older workers and workers with lower pre-disability earnings, our findings suggest that workers with a disability are indeed more likely to be let go and less likely to be rehired during economic downturns.