Are Health Insurance Markets Competitive?

New working paper from the NBER, by Dafny:

Although the vast majority of Americans have private health insurance, researchers focus almost exclusively on public provision. Data on the private insurance sector is extremely difficult to obtain because health insurance contracts are complex, renegotiated annually, and not subject to reporting requirements. This study makes use of a privately-gathered national database of insurance contracts agreed upon by a sample of large, multisite employers between 1998 and 2005. To gauge the competitiveness of the group health insurance industry, I investigate whether health insurers charge higher premiums, ceteris paribus, to more profitable firms. I find they do, and this result is not driven by cross-sectional differences across firms or plans: firms with positive profit shocks subsequently face higher premium growth, even for the same healthplans. Moreover, this relationship is strongest in geographic markets served by a small number of insurance carriers. Further analysis suggests profits act to increase employers' switching costs, and insurers exploit this inelasticity where they have sufficient bargaining power. Given the rapid industry consolidation during the study period, these findings suggest healthcare insurers possess and exercise market power in an increasing number of geographic markets.

The existence of price discrimination is interesting, as well as the observation that it seems to be geographically expanding market power. (I wonder if the reason behind such an expansion rhymes with the words "shmate shmegulation.") I still stop short of associating price discrimination with a direct lack of competition, especially because health insurance does not seem to be a particularly profitable industry itself.

However, if these markets are not competitive and lagged employer profit margins (see Table 2) are positively correlated with insurance premiums, then this is all the more reason to not encourage/subsidize employer-based health care, correct? Firms' profit margins are more easily attainable than household income shocks, price discrimmination would be more difficult. (The author actually points the policy implications in the direction of antitrust and reduced private sector expansion, but I'm left wondering what the public premiums are comparable in the ceteris paribus cases.)