Changes in Health Insurance Decisions by Medium-Sized Employers after the Implementation of the Employer Mandate

The employer shared responsibility provisions of the Affordable Care Act imposed penalties on employers that did not provide affordable insurance to their full-time employees if any of those employees received subsidies for coverage obtained from a Marketplace. These provisions were implemented in 2016 for medium-sized employers (those with 50-99 full-time equivalent employees). In this study, we investigate the impact of these provisions on medium-sized employers’ health insurance decisions using newly available longitudinal employer data from the Medical Expenditure Panel Survey – Insurance Component (MEPS-IC) longitudinal survey.

The 2013–14 MEPS-IC longitudinal survey only included employers with fifty or fewer employees, the 2014–15 survey included employers with 100 or fewer employees, and the 2015–16 survey included employers of all sizes. A recent study using data from the first two longitudinal surveys revealed a substantial amount of churn among small and medium-sized employers with respect to employers’ health insurance decisions. For example, 27.1 percent of medium-sized employers that did not offer insurance in 2014 began offering it in 2015 while only 3.0 percent of those offering insurance dropped it (Vistnes et al, 2017).

Since both the 2014-2015 and 2015-2016 waves of the MEPS-IC longitudinal survey contain data on firms with 50-99 employees, we are able to examine the before and after effects of the employer shared responsibility provisions on such firms, an understudied area. We begin by estimating difference-in-difference models to evaluate medium-sized employers’ decisions to change their offer status. We use employers with fewer than 50 workers as a control group in these analyses since longitudinal data are available for firms of this size in both the 2014-2015 and 2015-2016 periods and small employers should not have been affected by the shared responsibility provisions. Since employers that offered insurance in both years may have changed their behavior with respect to workers’ eligibility for coverage, we also examine employers’ requirements regarding the number of hours per week that employees needed to work in order to be eligible for insurance. In addition, we examine any resulting impact of these or other changes on employers’ eligibility rates (#eligible/#employed). Finally, we examine any changes in coverage rates (#enrolled/#employees) to analyze the net effect of these changes on the percent of workers at an establishment covered by insurance through their employer.

We supplement our analyses of the longitudinal MEPS-IC data with cross-sectional difference-in-difference models to test the sensitivity of our results to additional control groups, such as those with 25-49 and 100-200 employees. As in our longitudinal analyses, we refine the firm size categories to account for the presence of part-time workers to approximate the number of full-time equivalent employees in the workforce. In both sets of analyses, we also investigate whether there are differences in employer responses that vary by workforce wage distribution, since in order for employees to receive a tax credit through the Marketplace they must have income in a range to receive subsidies.