Last week's big news was that the Department of Justice and a number of states have challenged Aetna's proposed merger with Humana. As in any merger challenge, the way that the relevant product and geographic markets are defined is crucial to the court's determination of whether the merger should be enjoined. According to the DOJ's (and other states') complaint, this merger would combine one of the two largest insurers of Medicare Advantage (Humana) with the fourth largest (Aetna). This, the complaint alleges, would further concentrate 364 county Medicare Advantage markets that are already highly concentrated among a small number of firms. Aetna's principal defense is that insurer share of Medicare Advantage (“MA”) is of no antitrust relevance given that consumers have the option of enrolling in traditional Medicare (“TM”). Therefore, says Aetna, TM and MA plans are not separate product markets and the merged entity should be evaluated for its effects in a larger Medicare market combining MA and TM. Given the significance of this issue, 20 prominent health economists have analyzed the relevant market definition in a July 8, 2016 letter [full text also included below] to the Florida Attorney General, who has joined the DOJ in this merger challenge. The economists' letter concludes that "MA is not in the same relevant market as TM and, therefore, that the merger will cause a serious increase in concentration that raises a competitive concern." It will be interesting to see how the issue of market definition plays out in the litigation.
The full text of the letter follows:
July 8, 2016
Attorney General Pam Bondi
Office of Attorney General
State of Florida
The Capitol PL-01
Tallahassee, FL 32399-1050
Re: Pending Merger of Aetna with Humana: Erroneous Market Definition
Dear Attorney General Bondi:
We, the undersigned, submit to you this letter in our capacity as economists with expertise in the subjects of antitrust, competition policy, and health economics. We are concerned that the Florida Office of Insurance Regulation (OIR)’s recent decision to approve the Aetna-Humana merger is based on an erroneous belief that Medicare Advantage (MA) is in the same relevant product market as traditional fee-for-service Medicare (TM). Based on the commercial realities of the market and on scientific evidence from economic research, we believe that the MA is not in the same relevant market as TM and, therefore, that the merger will cause a serious increase in concentration that raises a competitive concern. Below we sketch out the rationale for our position.
The OIR’s finding is stated in the Consent Order:
the OFFICE finds that Medicare Advantage, the private market product, competes directly with Traditional Medicare. Accordingly, when considering the impact of the acquisition, the private market is only a portion of the Medicare market. When analyzed as the combination of the public and private markets, the Medicare market on a statewide basis is not highly concentrated, and the impact of the proposed acquisition affects the concentration by only a minimal amount (Florida OIR Consent Order, 2016, p. 5).
This finding is pivotal because the OFFICE also finds empirically that the MA market is already highly concentrated in most counties and MSAs and that the merger causes most of the moderately concentrated counties and MSA to become highly concentrated. Further, the increase in concentration is more pronounced in the more populous areas (Florida OIR Report, 2016, pp. 12, 15). Therefore, concluding that the merger would not affect competition requires the finding that MA is in the same relevant market as TM. As a result, the overall decision of the OFFICE is that:
the OFFICE finds that the proposed acquisition is not likely to be hazardous or prejudicial to the insureds of the insurer or the public and that the acquisition would not substantially lessen competition in this state or tend to create a monopoly therein (Florida OIR Consent Order, 2015, p. 9).
To the contrary, we believe that the nature of the products and economic research leads to the conclusion that MA is not in the same relevant market as TM and, therefore, that the merger raises competitive concerns.
A. MA is Substantially Different than Traditional Fee-for-Service Medicare
Most MA plans are HMOs. In 2015, 64 percent of MA beneficiaries were in HMOs. In return for reduced choice of providers and utilization review, the Medicare beneficiary enrolled in an HMO obtains more complete coverage. Over the long term, MA plans have been steadily increasing in share, attracting 31 percent of Medicare beneficiaries by 2015 (Jacobson, Damico, Neuman and Gold, 2015, fig. 2|Newhouse and McGuire, 2014). Traditional fee-for-service Medicare is a very different type of plan than MA plans. It has no panels and no serious utilization review. Indeed, fee-for-service Medicare is the only surviving large-scale example of traditional indemnity insurance in the U.S.
TM provides unrestricted choice of provider, but it exposes the beneficiary to the risk of high out-of-pocket responsibilities. In 2013-14, 16 percent of Medicare beneficiaries faced out-of-pocket responsibilities that exceeded 20 percent of their annual income (Schoen, et. al. 2016, p. 14). Purchase of a private Medicare supplemental policy (“Medigap” coverage) reduces the risk of high out-of-pocket responsibilities, but a high cost. MA insurance, on the other hand, leads to less risk of high out-of-pocket responsibilities. In MA plans, the average out-of-pocket maximum was $5,014 per year per beneficiary in 2015 (Jacobson, Damico, Neuman and Gold, 2015, fig. 9). In addition, most (87 percent) MA plans cover pharamaceuticals, where TM does not (Medicare Advantage 2016). TM enrollees can obtain drug coverage through Medicare Part D, at an additional expense.
A beneficiary could piece together a plan of TM+Part D+Supplemental Insurance to approximate the breadth of coverage of MA plans. However, that pieced-together plan would be much more expensive in both premiums and out-of-pocket expenses than MA. A recent study of Miami-Dade County data showed that the average pieced-together plan would have monthly premiums of $318 v. $88 for MA. Average monthly out-of-pocket expenditures would be $409 v. $182 (Sinaiko, Afendulis and Frank 2013, pp. 206-207).
Further, MA utilization control for hospitals appears to be quite strict and effective. A recent study found that when MA beneficiaries had to switch to fee-for-service Medicare, their hospital utilization and costs rose substantially (Duggan, Gruber and Vabson 2015). This shows that MA utilization review had a large impact. In summary, there are large functional and financial differences between MA and TM.
B. MA Beneficiaries Differ from those in Traditional Fee-for-Service Medicare
Economic research shows that Medicare MA beneficiaries differ from those in TM in important respects. MA beneficiaries are healthier than those in TM. One way to see this is to compare the past health care expenditures of beneficiaries who switch from TM to MA to those who remain in TM. Various studies in recent years have found expenditures to have ranged from 20 to 45 percent lower among switchers than the average in the TM population (See Gruber, 2016, p. 7) for more discussion on this point). Further, the small percentage (3 percent) of beneficiaries who switch in the opposite direction (from MA to TM) are sicker than those who remain (Brown et. al. 2014, pp. 3356, 3357). Further, MA beneficiaries are more accepting of restrictions on provider choice in order to reduce costs than TM beneficiaries are (Hu 2005, pp. 1, 3).
C. Switching from MA to Traditional Fee-for-Service Medicare is Rare
Consumer behavior shows that beneficiaries view MA plans as quite different from traditional fee-for-service Medicare. MA enrollees who were involuntarily terminated because their plan left the market overwhelmingly (95 percent) sought another MA plan (Sinaiko and Zeckhauser 2015, p. 12). Voluntary switching from MA to TM is quite rare, only 3 percent per year nationally (Brown et. al. 2014, p. 3357). The corresponding figure for Miami-Dade County is similar at 5 percent (Sinaiko, Afendulis and Frank 2013, p. 209). These facts alone cast serious doubt on whether MA and TM are in the same product market.
D. The Growth of MA at the Expense of Traditional Fee-for-Service Medicare is Irrelevant
Over time, MA plans have grown at the expense of traditional fee-for-service Medicare. But, that does not imply that they compete closely enough to be considered to be in the same market. The MA growth represents a slow shift to a new organizational form and incentive system that is favored by a (slowly) growing number of Medicare beneficiaries. As indicated above, switching between TM and MA is very low. For a historical analogue, consider the slow grow of automobile sales at the expense of horse-drawn carriages in the early 20th Century. The availability of horses did not constrain the pricing of automobiles.
E. MA Pricing is Driven by Concentration in MA
Another approach to market definition is to see if MA pricing and other behavior responds to concentration among MA plans. Recent research indicates that this is the case: where there are fewer MA insurers, premiums are higher. This demonstrates that traditional Medicare is not a serious constraint on MA pricing. If TM were in the same market as MA there should be little relation of MA premiums to the number of MA insurers – traditional Medicare would already act as a competitive constraint on MA pricing, so it wouldn’t matter how many MA plans are in the market. That is not what the research shows.
A recent study of competition in the MA market at the county level finds evidence of market power, stemming from market concentration of MA plans (Song, Landrum and Chernew 2013). MA plans are able to and do charge higher premiums where there is higher concentration among MA plans. Another county-level study shows that integrated MA/hospital plans charge higher quality-adjusted premiums than MA plans that are not integrated (Frakt, Pizer and Feldman 2013). These results would be impossible if TM competed closely with MA plans.
This research relates directly to the bottom line issue with the merger. Higher concentration in MA markets leads to higher MA prices , in spite of the presence of TM.
In sum, economic research suggests that Medicare Advantage insurance, is a separate relevant product market. Evidence shows that traditional fee-for-service Medicare does not much constrain Medicare Advantage price and decisions. Empirical research by the OIR shows that the merger would cause large increases in concentration in many local markets that are already highly concentrated. We urge you to consider this when weighing the evidence on the competitive consequences of allowing this merger.
Brown, Jason, Mark Duggan, Ilyana Kuziemko, and William Woolston. 2014. How Does Risk Selection Respond to Risk Adjustment? New Evidence from the Medicare Advantage Program American Economic Review 104(10) (October): 3335–3364.
Dafny, Leemore S. 2015. Evaluation the Impact of Health Insurance Consolidation: Learning from Experience. Commonwealth Fund Issue Brief (November).
Duggan, Mark, Jonathan Gruber and Boris Vabson. 2015. The Efficiency Consequences of Health Care Privatization: Evidence from Medicare Advantage Exits. National Bureau of Economic Research Working Paper 21650 (October).
Florida Office of Insurance Regulation Report on the Review of Aetna Inc.’s Acquisition of Humana and Affiliates. 2016. (February 12).
Florida Office of Insurance Regulation Consent Order in the matter of the Indirect Acquisition of Human Health Insurance Company of Florida, et al. by Aetna Inc. 2016. (February 15).
Florida Office of Insurance Regulation Report on the Review of Aetna Inc.’s Acquisition of Humana and Affiliates. 2016. (February 12).
Frakt, Austin P., Steven D. Pizer and Roger Feldman. 2013. Plan—Provider Integration, Premiums and Quality in the Medicare Advantage Market. Health Services Research, 46(6), Part 1 (December): 1996-2013.
Frech, H. E. III, 2016. Comments on Selected Issues Re: The Proposed Mergers of Anthem and Cigna and Aetna and Humana, submitted to the Departments of Insurance in California and Missouri (May 19) available at http://www.insurance.ca.gov/01-consumers/110-health/60-resources/upload/FrechReport-FINAL-051716-CA-002.pdf
Gruber, Jonathan. 2016. Report to the Missouri Department of Insurance regarding Competition in the Medical Advantage and Individual Exchange Markets (May 6) available at http://insurance.mo.gov/documents/exhibit-34.pdf
Tu, Ha T. (2005). Medicare Seniors Much Less Willing to Limit Physician-Hospital Choice for Lower Costs. Issue Brief No. 96, Center for Studying Health System Change (June) available at http://www.hschange.com/CONTENT/744/?words=Medicare%20Seniors
Jacobson, Gretchen, Anthony Damico and Tricia Neuman. 2015. Medicare Advantage Enrollment, by Firm. Data Note, Kaiser Family Foundation (July 14).
Medicare Advantage Fact Sheet. (2016) Kaiser Family Foundation (May 11) available at http://kff.org/medicare/fact-sheet/medicare-advantage/
Newhouse, Joseph P. and Thomas G. McGuire. 2014. How Successful Is Medicare Advantage? Milbank Quarterly 92(2) (June): 351-394.
Schoen, Cathy, Claudia Solís-Román, Nick Huober, and Zachary Kelchner. 2016.
On Medicare But At Risk: A State-Level Analysis of Beneficiaries Who Are Underinsured or Facing High Total Cost Burdens. The Commonwealth Fund Issue Brief (May).
Sinaiko, Anna D. and Richard Zeckhauser. 2015. Persistent Preferences and Status Quo Bias Versus Default Power: The Choices of Terminated Medicare Advantage Clients. Working Paper, Harvard University.
Sinaiko, Anna D. Christopher C. Afendulis and Richard G. Frank. (2013). Enrollment in Medicare Advantage Plans in Miami-Dade County: Evidence of Status Quo Bias? Inquiry 50(3): 202–215.
Song, Zuri, Mary Beth Landrum and Michael E. Chernew. 2013. Competitive bidding in Medicare Advantage: Effect of benchmark changes on plan bids. Journal of Health Economics 32(6) (December): 1301-1312.
Spiro, Topher, Maura Calsyn and Meghan, O’Toole. 2016. Bigger is not Better: Proposed Insurer Mergers are Likely to Harm Consumers and Taxpayers. Center for American Progress (Jan. 21).
Adam Atherly, Professor of Health Systems, Management and Policy, University of Colorado
Susan Athey, The Economics of Technology Professor, Graduate School of Business, Stanford University
Thomas C. Buchmueller, Professor of Business Economics, Ross School of Business, University of Michigan.
Kathleen Carey, Professor of Health Law, Policy and Management, Boston University
Patricia M. Danzon, Professor, Health Care Management, The Wharton School, University of Pennsylvania
Martin Gaynor, Professor of Economics and Health Policy, Carnegie Mellon University
Joshua Gottlieb, Assistant Professor of Economics, University of British Columbia
Ban Handel, Assistant Professor of Economics, University of California, Berkeley
Vivian Ho, Baker Institute Chair in Health Economics, Rice University
Richard Hirth, Professor and Chair of Health Management and Policy, University of Michigan
Neale Mahoney, Assistant Professor of Economics, University of Chicago, Booth School of Business
Keith Marzilli Ericson, Assistant Professor of Markets, Public Policy, and Law, Boston University
Thomas G. McGuire, Professor of Health Economics, Department of Health Care Policy, Harvard Medical School
Stephen T. Parente, Professor of Health Finance, University of Minnesota
Mark V. Pauly, Professor of Health Care Management, Business Economics and Public Policy, University of Pennsylvania
Steven D. Pizer, Associate Professor of Health Economics, Northeastern University
Mark A. Satterthwaite, A.C. Buehler Professor in Hospital and Health Services Management, Professor of Strategy, and Professor of Managerial Economics, Northwestern University
Richard M. Scheffler, Professor of Health Economics and Policy, University of California, Berkeley
Frank A. Sloan, Professor of Health Policy and Management, and Economics
Amanda Starc, Associate Professor of Strategy, Kellogg School of Management, Northwestern University
 For a fuller discussion of some of these issues and also for a discussion of additional factors suggesting that MA is a separate market see Gruber (2016). For an excellent discussion of the effects of past insurance mergers see Dafny (2015). For a general discussion of competitive concerns and market definitions relevant to this merger and also to the proposed Anthem-Cigna merger, see Frech (2016).
 See also, (Spiro, Calsyn and O’Toole, 2016). They conclude that when Humana offers a MA plan in the same county as Aetna, Aetna’s premium is lower than in counties where Humana does not offer a plan.