Are For-Profit Insurers Different? Leemore Dafny Northwestern University and NBER

Roughly half of private health insurance is provided by for-profit corporations. There is little evidence on whether and how for-profits differ from not-for-profits in this important industry. We explore the association as well as the causal effect of ownership status on one key measure: price (or premiums). Using a detailed panel dataset on employer-sponsored insurance from 1998-2009, we first show that forprofit premium levels are approximately 5 percent higher than non-profit levels, controlling as best as possible for differences in the underlying insured populations and several features of benefit design. We then study the effect of conversions of Blue Cross and Blue Shield affiliates in 11 states from nonprofit to for-profit status on premiums in their respective market areas. On average we find no impact, but there is important heterogeneity across markets. In markets where BCBS had high (low) market share prior to conversion, premiums actually increased (decreased) a nontrivial amount. The results imply that the effect of ownership status on premiums is sensitive to local market structure. The healthcare industry in the United States is an amalgam of not-for-profit, for-profit, and government entities. The health insurance sector is no exception: the non-elderly insured are split roughly equally across insurers of these ownership types. While there is an extensive theoretical and empirical literature examining the impact of ownership form on performance in the hospital sector, there is comparatively little research on the relationship between ownership form and outcomes related to health insurance. Prior studies focus on differences between government-provided insurance and private insurance, rarely distinguishing between private insurance offered by for-profit (FP) companies (such as Aetna and United Healthcare) and private insurance offered by not-for-profit (NP) companies (such as Blue Cross Blue Shield of Arizona and Harvard Pilgrim Health Care). This research gap came to fore during the national healthcare debate of 2009-2010, which culminated in the passage of The Patient Protection and Affordable Care Act (PPACA). During the debate preceding PPACA, policymakers considered adding a “public option” to the set of choices available for purchase by individuals who did not qualify for public insurance by virtue of income, age, or health status. After the public option failed to muster sufficient political support, the concept of providing federal seed money to form “not-for-profit cooperatives” was floated as an alternative. Formally introduced as the Consumer-Owned and – Oriented Plan (CO-OP), Senator Kent Conrad (R-ND) asserted that co-ops “will focus on getting the best value for customers, rather than maximizing plan revenues or profits.” The proposal implicitly maligned for-profits, notwithstanding the lack of evidence about how and whether they behave differently from not-for-profits. A total of six billion dollars was included in the final bill to help establish CO-OPs by July 1, 2013. 1 “Basic Facts & Figures: Nonprofit Health Plans,” Alliance for Advancing Nonprofit Healthcare, accessed 8/15/2010 at http://www.nonprofithealthcare.org/resources/BasicFactsAndFigures-NonprofitHealthPlans9.9.08.pdf A notable exception is Town, Feldman and Wholey (2004), which examines conversions of non-profit HMOs to for-profit status between 1987 and 2001. The authors find no short-term effect of conversions on premiums or profits of converting firms. We discuss this paper at greater length in the following section. 3 “FAQ about the Consumer-Owned and –Oriented Plan (CO-OP),” accessed 7/15/2010 at http://conrad.senate.gov/issues/statements/healthcare/090813_coop_QA.cfm. 4 The decision to structure the new not-for-profits as co-ops similarly implies that traditional NP insurers are not providing maximal value for enrollees. 5 Source: http://www.kff.org/healthreform/8061.cfm