How Does Regulation of the Small Group Market Affect Firm Self-Insurance and Implications for the Affordable Care Act

The Affordable Care Act (ACA) imposes adjusted community rating provisions on small group market policies (along with other coverage and generosity requirements) beginning in 2014, and expands the size of firms affected by these regulations to firms with up to 100 workers in 2016 (currently 50). However, firms can avoid these regulations, yet still comply with the employer mandate, by self-insuring their health benefits. This option is likely to be more attractive to firms with healthier workers standing to benefit more from avoiding the community rating provisions, leading to concern of adverse selection into and higher premiums in the SHOP exchanges and small group market. In this context, we exploit existing state-level variation in small group market rating regulation to evaluate the extent to which stronger community rating has resulted in selective self-insurance among healthier firms.  

We obtain firm-level self-insurance status and other characteristics, including industry and state, from a restricted use version of the 2008-12 KFF/HRET Annual Employer Health Benefits survey and restrict our primary analyses to firms with 25-50 workers subject to state small group regulation. We run probit models of firm self-insurance on firm characteristics and state regulations, including an indicator of whether industry is an allowable premium rating factor, and find no effect in the entire sample of firms. We then stratify the sample of firms into high and low expected spending industries, using variation in mean annual expected insured expenses for workers across industries (constructed using MEPS data), and find heterogeneous effects of rating by industry on firm self-insurance. Premiums being allowed to vary by industry is associated with a 5.3% (p<.05) lower likelihood of firm self-insurance among firms in low expected spending industries, and a 3.4% (p<.05) higher likelihood of firm self-insurance among firms in higher expected spending industries. That is, when premiums cannot reflect heterogeneity in expected expenditures by industry, firms in lower expected spending industries are more likely to self-insure while firms in higher expected spending industries are less likely to self-insure, consistent with adverse selection of higher spending firms into the regulated market. As a control, we run the same set of analyses on firms with 51-100 workers (which are not subject to the regulations) and find no effects on firm self-insurance among the full or stratified samples. We also control for other state insurance regulations including benefit mandates, premium tax rates, and the presence of a high risk pool and find limited effects of these regulations. Interpreting these findings in the context of the forthcoming ACA-based reforms suggests that concerns related to adverse selection into the regulated small group market via selective self-insurance of healthier firms are indeed warranted, and that state or federal policy levers such as penalties for delayed entry of firms into the fully-insured market, restrictions on the sale of low-attachment stop-loss policies to small firms, and/or broader risk adjustment should be considered.