Professor David Hyman

September 6, 2006

Statement of David A. Hyman Professor of Law and Medicine University of Illinois Hearing on "Examining Competition in Group Health Care" U.S. Senate Committee on the Judiciary September 6, 2006

Thank you for inviting me to speak with you today. I am a professor of law and medicine and the Galowich-Huizenga Faculty Scholar at the University of Illinois, where I direct the Epstein Program in Health Law and Policy. I spent three years as Special Counsel at the Federal Trade Commission ("FTC"), where I coordinated and was principal author of the joint report issued by the FTC and Department of Justice on health care and competition law and policy.1 My academic interests focus on the financing and regulation of health care, and I have written numerous articles on these subjects - including several on the specific issues that you are considering today. The title of today's hearing - "Examining Competition in Group Health Care" actually encompasses a wide array of issues.

These issues include:
? The role of federal and state regulation of health care delivery;
? The role of federal and state regulation of health insurance;
? The role of the federal tax subsidy for employment-based health insurance;
? The role of employers in structuring the marketplace for health care financing and delivery;
? The role of Medicare in structuring the marketplace for health care delivery - and to a lesser extent, the marketplace for health care financing;
? The extent to which information is available to consumers and policy makers about the cost and quality of health care services;
? The optimal strategy for addressing agency problems in employer-based health insurance;
? The role of antitrust law as applied to health care providers and insurers

Given our time constraints, I will focus on a much narrower subject - whether there is evidence of market power in health insurance markets - and what, if anything, we should do about it. My remarks are informed by the several days of hearings we devoted to this subject when I was at the FTC, and the two chapters in the final report on health insurance and competition law.2 They are also informed by the academic literature that has appeared on the subject, and the multiple reports that the American Medical Association has issued decrying consolidation in the market for health insurance - most recently in April, 2006 (using 2005 data).3 I should note that my views on this specific subject are laid out in some detail in my 2004 Health Affairs article, titled Monopoly, Monopsony and Market Definition: An Antitrust Perspective on Market Concentration Among Health Insurers.4 The article was co-authored with Bill Kovacic - then General Counsel, and now a Commissioner on the Federal Trade Commission.5 1 Federal Trade Commission and Department of Justice, Improving Health Care: A Dose of Competition (2004). 2 Id. 3 American Medical Association, Competition in Health Insurance: A Comprehensive Study of U.S. Markets - 2005 Update (April, 2006), available at http://www.ama-assn.org/ama1/pub/upload/mm/368/compstudy_52006.pdf. 4 David A. Hyman & William Kovacic, Monopoly, Monopsony and Market Definition: An Antitrust Perspective on Market Concentration Among Health Insurers, 24 Health Affairs 25-29 (Nov./Dec. 2004). 5

Of course, in writing this article, Commissioner Kovacic was conveying his own views, and not those of the Commission, or of individual Commissioners. 2
Obviously, the backdrop for this hearing is the complaints of health care providers about disparities in bargaining power in dealing with insurance companies. This setting should give pause, for two distinct reasons. First, disparities in bargaining power are simply not the same thing as monopsony (buyer-side monopoly) power. Indeed, equal bargaining power is very much the exception in most markets. However, as long as those markets are reasonably competitive, there is no particular reason to get unduly exercised about bargaining disparities. Indeed, markets can work well with significant disparities in bargaining power, as long as they are reasonably competitive. To pick a few non-random examples, there are huge bargaining disparities in the markets for retail consumer goods, car rental and purchase, and air travel, but these markets are all sufficiently competitive that these bargaining disparities just don't matter to consumers - nor should they.

Second is the simple fact that the complaints come from providers -- and not consumers. In health care, providers have long set the terms of trade, including generous compensation without regard to the quality or value of the services they provide. There has been a dramatic shift in bargaining power over the past several decades in many markets away from health care providers and toward purchasers. It is far from clear the rest of us should be much concerned with that trend - again, as long as the market for health insurance is reasonably competitive. The sellers of a service have a natural tendency to conflate what is good for them with what is good for society - but the interests of consumers are sufficiently at odds with those of providers that we should generally discount provider complaints about disparities in bargaining power - an insight that flows naturally from the maxim that the purpose of antitrust is to protection competition, not competitors.6

Let me now turn to the evidence offered by providers in support of their position. Essentially, they make two claims: there have been a host of mergers among insurance companies that have resulted in the emergence of insurers with a national presence; and there are high Herfindahl-Hirshman ("HHI") indices in individual states and metropolitan areas.7

I address each of these points in turn. Over the past several decades, health insurance markets have moved from markets overwhelmingly dominated by nonprofits (primarily Blue Cross) that operated only in single states to a market with several large national insurers that operate in multiple states, and nonprofits that continue to operate in single states. These national insurers cover millions of Americans in multiple states. Does the emergence of national insurers indicate that we have a problem in the market for health insurance? The short answer to that question is no.

Understanding why requires a brief review of some basics of economics and antitrust law. Antitrust law focuses on the problem of market power. Market power is when sellers (or buyers) have the ability to profitably maintain prices above (or below) competitive levels for a significant period of time. When sellers exercise market power, it is called "monopoly." When buyers exercise market power, it is called "monopsony." Both monopoly and monopsony decrease consumer welfare. Health insurers are both buyers of medical services (from 6 Brown Shoe Co. v. United States, 370 U.S. 294, 320 (1962) ("Taken as a whole, the legislative history illuminates congressional concern with the protection of competition, not competitors, and its desire to restrain mergers only to the extent that such combinations may tend to lessen competition.") 7 See AMA, supra note 3, at 1. 3
providers) and sellers of insurance (to consumers), so they can raise both monopsony and monopoly concerns. Absent direct evidence of anticompetitive effects (e.g., higher prices, lower outputs, and lower quality), antitrust analysis of market power generally begins with the identification of relevant product and geographic markets and calculation of the shares of market participants and concentration ratios.

With this background, it is easy to see why the raw number of Americans that are covered by a particular national insurer is effectively irrelevant to an inquiry into market power. The starting point for analysis should be the market share of these insurers in particular geographic markets - not the total number of Americans who receive health insurance from national insurers. What of the HHI indices, the second basis for the AMA's position?

The HHI, which forms the analytical foundation for the FTC/DOJ merger guidelines, represents the sum of the squares of the market share of individual competitors in the market. In a market with a single seller, the HHI is 10,000. The FTC/DOJ merger guidelines provide that an HHI below 1000 corresponds to an "unconcentrated" market; an HHI between 1000 and 1800 corresponds to a "moderately concentrated" market, and a HHI above 1800 corresponds to a "highly concentrated" market. The HHI is used as a screening tool to assess whether a proposed merger is more or less likely to have anticompetitive consequences. The merger guidelines provide that different presumptions apply, depending on the extent of post-merger market concentration and the increase in HHI that will result from the merger. For example, a merger that results in an unconcentrated market "ordinarily require no further analysis" because it is unlikely to have adverse competitive effects, but where the post-merger HHI exceeds 1800, it is "presumed that mergers producing an increase in the HHI of more than 100 points are likely to create or enhance market power or facilitate its exercise."8 There is no question that the reports prepared by the AMA have gotten much more sophisticated and more comprehensive over time.9 Unfortunately, these reports fail to address the fundamental problems that have beset their analysis from the outset.

The first problem is that high HHIs do not demonstrate that market power exists or is being exercised. HHIs are a screening tool. The purpose of the HHI is to raise or lower our index of suspicion about the likelihood of market power being created or exercised in the context of evaluating a proposed merger - not to establish that market power exists or will exist.

Second, even if it could be shown that a health insurer actually has market power, the issue for antitrust purposes is whether the insurer has obtained or maintained that power through improper means. Absent such evidence, the sole fact that a market is concentrated is unlikely to attract the interest of an antitrust enforcer. With one exception, high levels of concentration have never been thought sufficient, taken by themselves, to merit an antitrust challenge. In the late 1970s, the FTC briefly flirted with using a "no-fault" theory of antitrust liability to de-concentrate various industries without proof of improper conduct. The FTC dropped this approach after developments in the case law and overwhelming criticism from antitrust experts led the FTC to conclude that 8 FTC/DOJ Merger Guidelines, 1.51 (1992), available at http://www.ftc.gov/bc/docs/horizmer.htm. 9 For example, the first study analyzed 40 metropolitan areas, while the latest study analyzes almost 300 metropolitan areas. AMA, supra note 3, at

1.no-fault cases would receive a hostile reception in the courts.10 No competition agency has sought to revive this strategy in the intervening twenty-odd years. Third, the HHI is used to calculate market concentration only after the scope of the product and geographic market is determined. The validity of the HHI as a screening tool depends entirely on proper definition of the relevant market. As Judge Richard Posner has observed, "the definition of the market in which to measure the market shares of the merging parties and their competitors is critical; given enough flexibility in market definition a surprising number of innocuous mergers can be made to appear dangerously monopolistic."11

Similarly, Robert Pitofsky, former Chairman of the FTC, has observed that "knowledgeable antitrust practitioners have long known that the most important single issue in most enforcement actions - because so much depends on it - is market definition."12 The market definition process was farcically described by Professor (and Nobel laureate) George Stigler: Consider the problem of defining a market within which the existence of competition or some form of monopoly is to be determined. The typical antitrust case is an almost impudent exercise in economic gerrymandering. The plaintiff sets the market, at a maximum, as one state in area and including only aperture-priority SLR cameras selling between $ 200 and $250. This might be called J-Shermanizing the market, after Senator John Sherman. The defendant will in turn insist that the market is worldwide, and includes not only all cameras, but also portrait artists and possibly transportation media because a visit is a substitute for a picture. This might also be called T-Shermanizing the market, this time after the Senator's brother, General William Tecumseh Sherman. Depending on who convinces the judge, the concentration ratios will be awesome or trivial, with a large influence on his verdict.13 At first glance, it might seem intuitively appealing to use states and metropolitan areas as geographic markets. An individual state is clearly a relevant parameter for regulatory purposes, and a considerable amount of data is available on a state-by-state basis.

However, there is no evidence that individual states constitute relevant geographic markets for health insurance - and there is considerable evidence to the contrary. Indeed, the AMA's study of market concentration expressly cautions that "state-level data can be very misleading because in many states, health insurers do not compete on a statewide basis."14 It is much more plausible to evaluate markets at the metropolitan level, but once again one must demonstrate that a particular metropolitan area is a market, and not simply assume it. Bluntly stated, if an entire state or metropolitan area is not a relevant 10 Timothy J. Muris, Improving the Economic Foundations of Competition Policy, Jan. 15, 2003, available at http://www.ftc.gov/speeches/muris/improveconfoundatio.htm. 11 RICHARD A. POSNER, ANTITRUST LAW: AN ECONOMIC PERSPECTIVE 125 (1976). 12 Robert Pitofsky, New Definitions of Relevant Market and the Assault on Antitrust, 90 COLUM. L. REV. 1805, 1807 (1990). 13 George J. Stigler, The Economists and the Problem of Monopoly, in THE ECONOMIST AS PREACHER AND OTHER ESSAYS 38, 51 (1982). 14 AMA, supra note 3, at 3. 5
geographic market, the existence of high HHIs in that state or metropolitan area has no competitive (or probative) significance. Fourth, it is important to distinguish lawful managed care contracting from unlawful monopsony behavior.

It is common for providers to treat disparities in bargaining power as prima facie evidence of anticompetitive behavior. This is silly. Managed care plans and other health insurers can legitimately lower provider prices by increasing competition among providers or engaging in other activities that lower the costs of provider services. By engaging in hard bargaining, insurers lower the cost of coverage - which directly benefits consumers. Because one purpose of managed care is to lower prices closer to a competitive level, it can be extremely difficult to determine when a managed care purchaser is exercising monopsony power - necessitating a lengthy, fact-intensive investigation that is prone to error. The more general point, as Table 1 reflects, is that any system for deciding whether monopsony power is being exercised will generate four kinds of the results: true positives (cell 1), false positives (cell 2), false negatives (cell 3), and true negatives (cell 4). Table 1: A Typology of Monopsony Evaluation Did the Enforcer Determine There Was Monopsony? Was There Monopsony? Yes No Yes True Positive (1) False Negative (3) No False Positive (2) True Negative (4) True positives and true negatives occur, respectively, when an enforcer correctly determines that there was monopsony, or correctly determines that there was not monopsony. False positives and false negatives occur, again respectively, when an enforcer determines there was monopsony even though there wasn't, or determines there wasn't monopsony even though there was. True positives and true negatives are correct results. False positives and false negatives are mistakes. The goal for an antitrust enforcer is to maximize the number of cases in cells 1 and 4, and minimize the number of cases in cells 2 and 3, while simultaneously minimizing the costs associated with the system of antitrust enforcement. The more difficult it is to distinguish between true positives and true negatives (let alone false positives and false negatives), the more expensive and error-prone the system is likely to prove. The AMA is understandably interested in minimizing the number of false negatives - but doing so is likely to increase the number of false positives - and consumers will directly bear the costs of those erroneous decisions. Before starting down this path, we should ask ourselves whether the game is worth the candle - particularly when there is precedent (written by then-Judge and now Supreme Court Justice Stephen Breyer) indicating that Aa legitimate buyer is entitled to use its market power to keep prices down," as long as the prices are not below incremental cost or predatory.15 15 Kartell v. Blue Shield of Massachusetts, 749 F.2d 922, 927-931 (1st Cir. 1984).
6
Finally, for all the complaints we have heard from providers about monopsony power and market concentration, it is striking how little we have heard from employers on the subject. At the FTC/DOJ health care hearings, panelists representing employers testified that health insurance markets in most geographic areas enjoy healthy competition, with multiple health insurer competitors offering multiple product options.16 Most employers can self-insure, and avoid most of the problems that might otherwise result from health insurance market concentration. Of course, these hearings were held several years ago, and it is certainly possible that employers in particular markets in particular states might express different views now - but that is where we should be looking if we want to get a reading on the likelihood insurers are exercising monopsony power. Where Should We Go From Here? My skepticism about the issue of monopsony does not mean that I think all is well in the health care sector of the economy. Let me summarize a few specific reforms that would help improve the status quo. ? Improve transparency of price and quality information, and use incentives to improve quality The quality of American health care is not what it should be.17 One important step in addressing this problem is improving the transparency of price and quality information. There have been promising preliminary steps in this direction, but there is much more to be done. The same goes for payment-for-performance ("P4P").18 ? Fix the tax subsidy The tax subsidy for employment-based health insurance is the source of considerable horizontal and vertical inequity.19 Although we need to be careful not to destabilize the existing system, it is long past time to experiment with various ways of eliminating these inequities. ? Lower barriers to entry in health insurance The AMA believes that state regulation is an important barrier to entry in the health insurance market.20 I concur with that assessment. The question is how best to fix the problem. One possible strategy is to allow small businesses to form Association Health Plans in order to get the benefit of ERISA preemption. Another possible strategy is to use regulatory federalism to create a national market in health insurance. As with the tax 16 Hyman & Kovacic, supra note 4, at 27. 17 David A. Hyman & Charles Silver, The Poor State of Health Care Quality in the U.S.: Is Malpractice Liability Part of the Problem or Part of the Solution?, 95 CORNELL L. REV. 893-993 (2005) 18 David A. Hyman & Charles Silver, You Get What You Pay For: Result-Based Compensation for Health Care, 58 WASHINGTON & LEE L. REV. 1427-1490 (2001); Arnold M. Epstein, Paying For Performance in the United States and Abroad, 355 NEW ENGL. J. MED. 406 (2006); Robert Galvin, Pay For Performance: Too Much of A Good Thing? A Conversation with Martin Roland, Health Affairs Web Exclusive (2006), available at http://content.healthaffairs.org/cgi/content/abstract/hlthaff.25.w412. 19 David A. Hyman & Mark Hall, Two Cheers For Employment-Based Health Insurance, 2 YALE J. HEALTH POL'Y, L. & ETHICS 23-57 (2001). 20 See AMA, supra note 3, at 2. 7
subsidy, although we need to be careful not to destabilize the existing system, it is long past time to experiment with various ways of addressing unnecessary state-created barriers to entry in the markets for health insurance and health care delivery. ? Malpractice Reform The academic consensus on the performance of the malpractice system is considerably at odds with the terms of the political debate that has been waged over the issue in the past few years.21 Although there is certainly a broad range of views on the best way to fix the malpractice system, no serious academic thinks that a cap on non-economic damages is going to address the pathologies of the existing system. If we want to improve the performance of the health care system, malpractice reform needs to be part of the discussion.22 Conclusion Let me close with a concrete example of the problem with the AMA's approach to the issue of monopsony power. I am one of two people at the University of Illinois College of Law that teaches health law, and the only person who does empirical research on medical malpractice. If one treats the College of Law as the relevant geographic market (and we are the only law school within 125 miles), the HHI for health law is 5,000 and the HHI for medical malpractice is 10,000. These are staggeringly high - but utterly meaningless HHIs. I can assure you that I don't have any market power in dealing with my dean with regard to my salary and teaching package. The obvious point is that unless the product and geographic market is correctly defined, high HHIs are simply irrelevant to what we actually care about - and even if these markets are properly defined, the HHI is only a screening test that calls for further investigation. For these reasons, the kindest thing one can say about the charge of monopsony, at least based on the current record, is the old Scottish verdict, "not proven." 21 David A. Hyman & Charles Silver, Medical Malpractice Litigation and Tort Reform: It's the Incentives, Stupid, VAND. L. REV. (forthcoming, 2006). 22 I note that the Senate Committee on Health, Education, Labor, and Pensions held a hearing on "Medical Liability: New Ideals for Making the System Work Better For Patients" on June 22, 2006. See http://help.senate.gov/Hearings/2006_06_22/2006_06_22.html.
8