No Longer Locked In

As state health insurance exchanges open in October 2013, changes in the way Americans obtain health insurance are on the horizon. The historical link between employment and access to health insurance will be weakened, as many individuals will no longer need to be employed to secure affordable health insurance. Coupled with this, individuals and families may be eligible for subsidies, depending on their income. The debate on the extent of labor market impacts has begun.

In an effort to quantify possible impacts from this change, researchers Craig Garthwaite, Tal Gross and Matthew Notowidigdo assessed the labor market implications from the largest public health dis-enrollment in US history. In particular, they attempt to quantify the effects of “employment lock”, wherein individuals seek employment primarily for access to health insurance. To do this, they examine the employment impact of the 2005 Tennessee cut to Medicaid where approximately 170,000 individuals were dis-enrolled.

Tennessee’s state Medicaid agency had been running a deficit because of a 1994 change to the program, which eliminated the income requirement and extended eligibility to individuals who were “uninsured” or “uninsurable.” By 1995, under this change, 40 percent of enrollees had incomes above 100 percent of the federal poverty line, 6.3 percent had incomes between 200 and 400 percent, and 1.3 percent had incomes above 400 percent. When the 2005 cuts took place, individuals who had benefitted from the state’s Medicaid expansion were the first to have their eligibility removed, many of whom were childless adults.

Taken together, the socioeconomic and demographic characteristics of individuals who were affected by the 1994 expansion in the Tennessee program are similar to those who will be most impacted by the subsidies and access to insurance to be provided by the Affordable Care Act (ACA).

To assess the labor market impacts of the cuts in public insurance coverage, the researchers utilize data from the Current Population Survey (CPS), which they restricted to individuals between 21 and 64, without a college degree, sampled from Tennessee and other southern states. From this data, information on an individual’s insurance status, source (public or private), and labor market activity is studied.

Following the cuts in 2005, a difference in difference model based on the CPS data suggests that adults in Tennessee were a statistically significant 4.6 percent less likely to have public health insurance. A larger impact was seen when estimating the effect on childless adults, who were estimated to be 7.3 percent less likely to have public health insurance.

The researchers also use difference in difference and triple-difference models to quantify estimates of changes in labor supply. They compute a 2.5 percentage point increase in employment rates after the dis-enrollment. For childless adults, the effect is again more pronounced, with a 4.6 percentage point increase in the group’s employment rate. Delving further, the researchers also estimate the changes in hours worked, and find that among the newly employed, almost all worked more than 20 hours per week, and a plurality worked more than 35 hours per week. To shed light on the source of this increase in labor market activity, the researchers also analyze wages. They observe that wages decreased by 1.8 percent for all childless adults over this time period, which would be consistent with an increase in labor supply, rather than labor demand.

These figures indicate that individuals who were previously eligible for Medicaid appear to have become employed in order to secure health insurance. The researchers estimate that following the changes, there was 4.2 percentage point increase in childless adults with private insurance through their employer and a 3.4 percentage point increase in non-employment private insurance. This increase in employment and access to private insurance is most prevalent in the 40-64 age group, adding further support to the theory that securing health benefits was this group’s primary concern when re-entering the labor force. his effect will likely be most pronounced in the labor force participation rate, as individuals who can secure health insurance from a source other than their employer will likely drop out of the employment market altogether.

Nevertheless, policy makers ought to take caution in extrapolating these results to the entire country, especially given that this study only used data from the CPS, which was a small subset of the 170,000 individuals who lost coverage. Additionally, labor market conditions following the Great Recession are likely different from the conditions found in Tennessee and other southern states in 2005. However, it is undeniable that changing the uniquely American link between health insurance and employment will have effects on the labor market. Policy makers looking at enrollment statistics and labor market activity in the coming years should be mindful of this changing dynamic.

Feature Photo: cc/(Chandrachoodan Gopalakrishnan)