Association for Public Policy Analysis & Management

In the United States, health insurance is closely related to employment. Many Americans only have access to affordable health insurance through their employer or the public health insurance programs. Americans without access to public or employer-provided insurance can purchase health insurance through the individual, non-group market, but that market is believed to face adverse-selection pressures which limit its availability to the public.

Job losses caused by the financial crisis have been going on since December 2007, and accelerated in September 2008. As a result, the number of people with health insurance decreased to 253.6 million in 2009 from 255.1 million in 2008. This is the first year that the number of people with health insurance has decreased since 1987.The nation- wide employer coverage decreased from 58.5% in 2008 to 55.8% in 2009. In the contrast, since people who lost their health insurance along with their jobs can apply for government health insurance programs, such as Medicaid and Medicare, the number of people covered by government health insurance increased to 93.2 million in 2009 from 87.4 million in 2008. However, these programs vary from state to state, depending the on financial budget and specific eligibility of those states. For example, Massachusetts is one of the states that experienced significant expansion of the Medicaid program, while Tennessee had a decline in the Medicaid coverage during the Great Recession.

Medicaid may have large effect on labor supply especially during the time when people lost their job due to the downturn of the economy. On one hand, if some un- employed individuals get coverage of Medicaid, it might create disincentives to hunting for new jobs. On the other hand, those uninsured and unemployed people might have stronger eagerness to find a job compared with those who is covered by Medicaid. But they might have poor health outcome due to the inaccessibility to medical care and would eventually have difficulty going back to the job market.

In this paper, I plan to exploit the across-state variation in exposure to the Medicaid

enrollment. I intend to use the sharp differences in eligibility among different states to estimate difference-in-difference models, which compare Medicaid expansion states and Medicaid non-expansion states during the Great Recession. Second, I will focus on the particular sub-population–unemployed childless adults which are potentially not eligible for Medicaid in non-expansion states. Relative to previous work, I believe that the policy change and large scale of the job losses leads to clear results of the effect of Medicaid on labor supply.