Association for Public Policy Analysis & Management

Increasing the costs of providing health insurance affects labor markets, including wages, labor input composition and working time as well as group insurance coverage. To keep total compensation the same, increased group insurance costs may be offset by lower wages. Additionally, employers would also like to substitute part-time workers who are exempted from group coverage for full-time workers, or increase working time per worker instead of adding workers. This paper presents new evidence on the impact of rising health insurance costs on working time using variation in mental health parity law establishment at both the state and federal levels.

     Mental health parity is a mandated health benefit that prohibits insurers from discriminating between coverage for mental and physical health care. Traditionally, the benefits for mental health have had more restrictions than physical health in group insurance plans, thus parity drives up the cost of providing group coverage substantially. At present, many states have passed their own mental health parity laws for non-self-insured firms. The federal government also took steps to promote more comprehensive coverage for mental health by passing the “Mental Health Parity Act” in 1996 (MHPA1996) and “Mental Health Parity and Addiction Equity Act” in 2008 (MHPAEA2008) for all kinds of group plans. MHPAEA2008 is still effective under the context of the Affordable Care Act. With this set of laws, I am able to estimate the response of working time to rising health insurance costs by exploiting these exogenous cost shocks.

     To identify the effect of state parity on working time, I merge state by year policy variation to the individuals in the Current Population Survey March supplements from 1992 to 2010. My results suggest that, for 25-64 year old workers in the private sector, there is a 1.4 percent increase in weeks worked last year in the parity states over this period. For workers of 35-44 years old, state parity also increases their probability of working part-time by 1.8 percentage points. Their average hours worked are decreased by 1.7 percent by this employment composition change. Since self-insured firms are exempted from state mandated regulation, I also compare the effect of parity law on small firms and large firms as the latter are more likely to be self-insured. I find that state parity increases the working weeks 1 percent more for workers in small firms than in large firms.

     I then examine the effects of federal parity laws to see whether they confirm my previous findings. The passage of MHPA1996 and MHPAEA2008 can be treated as “reverse experiments” by providing the existing state parity a set of controls. A before-and-after comparison of states that already had their own parity and states that did not shows that MHPA1996 has little effect on working time. However, MHPAEA2008, which is a stronger federal law than MHPA1996, increases weeks worked by 1.7 percent. The findings on both state and federal parity pass a pre-trend check, and are robust to the inclusion of state, year, and region-by-year effects as well as state-specific trends.