How Do Health Insurer Market Concentration and Bargaining Power with Hospitals Affect Health Insurance Premiums?

The US health insurance industry is highly concentrated and employer-sponsored health insurance (ESI) premiums are high and rising rapidly. While much attention has focused on the possible link between these two phenomena, the net effect of health insurance market concentration on premiums is ambiguous because, while market power may enable insurers to impose higher markups on premiums charged to employers, it may also result in stronger bargaining leverage with hospitals and lower negotiated payment rates, which could, in turn, be passed through to employers in the form of relatively lower premiums. We use a nationally representative dataset of ESI premiums from 2006-11 to empirically investigate these potentially offsetting effects by constructing two unique measures of insurance concentration representing the market in which insurers sell policies to employers distinctly from that in which insurers negotiate with hospitals. We restrict our analyses to premiums for fully-insured policies sold to smaller employers and measure the competition in the market for selling these policies to employers as the Herfindahl-Hirschman Index (HHI) constructed for fully-insured policies at the CBSA level (from the InterStudy data). Alternatively, we construct the HHI for the market in which insurers bargain with hospitals by aggregating insurer enrollment using the entire commercial book of business (i.e., fully- and self-insured plans) to either the CBSA or HRR level. We also use the AHA data to construct HHIs for hospital systems at the CBSA or HRR level. We merge these market-level variables to a restricted-use version of the KFF/HRET Annual Employer Health Benefits survey and conduct plan-level OLS regressions of the logged annual premium on these market concentration measures and controls. 

We find that premiums are indeed higher for plans in more concentrated markets in which insurance is sold to employers and lower for plans in more concentrated markets in which insurers negotiate with hospital systems. Additionally, we find higher premiums for policies sold to employers located in more concentrated hospital markets. Moreover, these relationships with market concentration are larger where the other market is more competitive. For instance, compared to a market with competitive insurers and hospitals, a market with competitive insurers and concentrated hospitals has 8.3% (p<.01) higher premiums, while a market with both concentrated insurers and hospitals 6.2% (p<.05) higher premiums. These findings are robust to alternative market definitions.  

Our results suggest that there are indeed offsetting effects of insurance market concentration on premiums and that not just the levels but also the relative balance of insurer and hospital market concentration have important implication for ESI premiums. We interpret these findings to suggest that recent legislative efforts targeted toward increasing competition in the US health insurance industry may be misdirected if they do not also target the level of concentration among hospital markets and/or regulate price negotiations in some way. Additionally, our findings have important antitrust implications, as they suggest that higher hospital prices resulting from higher hospital market concentration are passed through to employers in the form of higher premiums, particularly in more competitive insurance markets.