NEW DIRECTIONS AT THE FTC:
EFFICIENCY JUSTIFICATIONS IN HOSPITAL MERGERS AND VERTICAL INTEGRATION CONCERNS
PREPARED REMARKS OF COMMISSIONER CHRISTINE A. VARNEY
FEDERAL TRADE COMMISSION
HEALTH CARE ANTITRUST FORUM
HOLIDAY INN -- MART PLAZA
May 2, 1995
The views expressed in these remarks are my own, and do not necessarily reflect those of the Federal Trade Commission or any other individual Commissioner. Thomas N. Dahdouh provided assistance in preparing this speech.
I appreciate the opportunity to speak before such a group of sophisticated antitrust practitioners in the health care area. The health care industry presents a continuing challenge to antitrust enforcers as we try to understand the implications of the managed care revolution. I believe that competition has generally played a positive role in ensuring that managed care and other innovative health care delivery systems have a chance to compete in these markets. The dynamism of these markets, however, requires antitrust enforcers to keep fully abreast of all the changes in these markets. We know a great deal about health care markets, but we need to know more in order to make sure that we have the analysis right.
Today, I would like to discuss a few of the new directions the Federal Trade Commission is taking in its enforcement agenda concerning the health care industry. First, I would like to discuss our evolving approach to efficiencies justifications in hospital merger analysis. Then, I want to discuss vertical integration issues in the health care industry.
I. Efficiencies Justifications in Hospital Merger Analysis
I think the FTC, under Chairman Pitofsky s leadership, will be emphasizing efficiencies justifications in examining hospital mergers. I agree that efficiencies justifications, as a matter of prosecutorial discretion, should play an important role in our analysis. I intend to take a more in-depth look into efficiencies justifications in future hospital mergers. I believe efficiencies are important in hospital merger analysis, however, my comments here concern prosecutorial discretion only and do not conflict or detract from the legal standard for an efficiencies defense.
There are two unique aspects of hospital mergers that virtually mandate that antitrust enforcers carefully examine asserted efficiencies before determining whether to challenge a particular hospital merger. The first relates to the peculiarities of the industry. The second relates to the peculiarities of the 1992 Horizontal Merger Guidelines analysis as applied to hospital markets. Let me explain both of these in some detail.
First, largely as a result of generous cost-plus reimbursements by Medicare and other third-party payers and, for a time, outright subsidies for new hospitals, hospitals in the 1960's went on a construction binge. In the last ten years, payors have moved away from retrospective cost reimbursement to prospective schedules of allowable prices or capitated payments, putting cost-containment pressures on hospitals. These cost containment measures, along with technological advances, have shortened in-hospital treatment time and forced more outpatient based treatments. As a result, many hospitals have excess capacity, making it difficult to recover the costs of previous capital expenditures. Under these conditions, efficiency arguments will, of course, appear to be more facially plausible.
Second, there are unique properties to hospital merger analysis that suggest that antitrust enforcers will be examining efficiencies justifications more often than perhaps in other contexts. As you know, the Merger Guidelines analysis generally proceeds in a linear fashion, first examining product market and geographic market, then moving on to competitive effects of the merger, proceeding through entry analysis and ending finally with efficiencies. It may seem, then, that, if a product market or a geographic market is not established, the analysis ends at that point, and no further examination of competitive effects or efficiencies is necessary. In a litigation where the plaintiff bears the burden of proof on product market or geographic market, it may make sense to proceed in such a lockstep fashion. 1 However, when antitrust enforcers are examining a merger in their prosecutorial discretion, such a linear examination could prevent full examination of all the competitive significance of a merger. Just to give you some examples in the Merger Guidelines, two theories of competitive harm -- price discrimination and unilateral effects from a merger of two products that are close substitutes -- are usually not viewed in such a lockstep fashion. In those instances, the anticompetitive story helps to inform the relevant geographic or product market analysis. For example, enforcers examining a possible unilateral effects story of goods that are close substitutes are in essence looking at both competitive effects and product market at the same time.
Even apart from these specialized competitive theories, geographic and product market definition should not be viewed as on-off switches. In particular, in hospital merger analysis, there may be instances where the evidence on geographic and product market definition is too inconclusive to render an informed judgment as to whether they have been established. We can obtain a great deal of information about the relevant geographic market in hospital mergers: information about patient inflows and outflows, physician admitting privileges, views of third-party payers, and internal planning documents of hospitals. However, such information is not always available, is sometimes not completely reliable, or is too inconclusive. Patient data does not provide us with a dynamic sense of changes in the flow of residents into and out of area hospitals over time 2 and suffers because it often includes tertiary care (such as complicated open heart surgery) that is excluded from the relevant product market definition. Evidence from third-party payers may not be reliable, either because of bias problems or because of lack of specific knowledge about the particular geographic area in question. Consequently, in certain situations, we may conclude that there is weak evidence supporting a particular geographic market. My view is that, at that point, we should not end the analysis -- we should go on to consider competitive effects and, hence, any efficiencies resulting from the merger, before concluding our analysis.
By taking a holistic approach to hospital mergers, I believe that we have a better understanding of all of the competitive ramifications of a hospital merger. Such an approach does require antitrust enforcers to balance the weak nature of the geographic market, in situations such as I have just described, with the magnitude of competitive effects and likely efficiencies resulting from the merger. But such a view more accurately reflects the complex reality of spatial competition: geographic markets often taper off, with competition at the fringes, and may often be affected by the fast-changing nature of the health care industry itself. These complexities also demand a careful review of efficiencies.
In the past, some antitrust enforcers have advanced overly strict standards for parties to meet in order to prevail on an efficiencies justification. They have stated that efficiencies should be shown with clear and convincing evidence and that the parties must in all cases prove that the fruits of those efficiencies will be passed on to consumers before they will be accepted as valid. It has also been suggested that efficiencies from planned consolidations which have not been publicly aired to the affected communities should be given less weight. The 1992 Merger Guidelines require only significant net efficiencies that cannot reasonably be achieved ... through other means. Efficiencies evidence is generally controlled by private parties and hence it makes sense to place the burden on parties to show efficiencies. However, a clear and convincing evidence requirement could be viewed as so strict that no one can practically meet it. 3 Similarly, Chairman Pitofsky has advanced the view, with which I wholeheartedly agree, that the requirement of a pass through to consumers is a killer qualification. As he notes, [t]he only sure way of making such a showing would be to prove that the merger is taking place in a near perfectly competitive market, and therefore that market forces would require that the efficiency be passed on to consumers. But if that were the case -- if the market were that competitive -- the merger would not have been a matter of concern in the first place. 4
I would instead take the view that where a merger does not impose a severe threat to competition, efficiencies should be presumed to flow to the benefit of consumers. A third killer qualification, in my view, would be any requirement that planned consolidations, which the parties are asserting will lead to efficiencies, be publicly aired to the affected communities before they are considered valid by antitrust enforcers. While such public airing may help increase confidence in the parties predictions of efficiencies, failure to do so should not prejudice an otherwise sound showing of likely efficiencies. Industries such as hospitals, which provide vitally needed services to the community but which suffer from chronic overcapacity, often need to make painful decisions involving layoffs and shifts in the provision of services. Such plans may, however, be efficiency-enhancing, by providing similar quality of care at lower cost. If such plans are publicly released, they could generate community or employee opposition to the merger that is not based on antitrust concerns. By requiring that such plans be publicly released, antitrust enforcers could unwittingly inject nonantitrust considerations into merger review analysis or, in effect, chill hospital merger proponents from advancing such efficiencies arguments, out of a fear that public airing of consolidation plans could bring pressure to rescind the merger.
By contrast to these killer requirements, the Merger Guidelines approach is a wholly defensible one. The Guidelines look at net efficiencies, permitting examination of the (sometimes) negative effects on efficiency of a merger so that we can get an overall sense of whether the merger will benefit consumers. Moreover, the focus on merger-specific efficiencies assures that there are no less restrictive alternatives -- such as joint venturing or contracting -- whereby the parties could achieve equivalent or comparable savings. Of course, the greater the competitive effect is, the more significant efficiencies will need to be before antitrust enforcers will decline to challenge the merger. Some further fine-tuning in the standard for efficiencies justifications may be necessary. 5 In the future, I expect Commission staff will be taking a closer look at the proper standard for examining efficiencies justifications.
II. Vertical Integration Issues
I also believe that vertical integration issues are going to play an increasingly significant role in enforcement actions at the Federal Trade Commission. First, I want to discuss a major vertical merger enforcement action at the Commission -- Lilly s acquisition of one of the largest pharmacy benefit management companies, PCS -- and then discuss in more general terms issues of vertical integration.
The emergence of prescription drug benefit management programs ("PBMs") is a significant aspect of the managed care revolution. PBMs provide managed prescription drug programs to managed care providers, corporations, labor unions, retirement systems, and federal and state employee plans and other plan sponsors. PBMs typically select participating pharmacists, drug manufacturers and suppliers, administer point of sale claims processing systems, negotiate quantity discounts with pharmaceutical manufacturers and pharmacists, administer plan record keeping and payments systems, and maintain quality control. In the past year, pharmaceutical manufacturers have acquired some of the largest PBMs -- Medco/ PAID, PCS and DPS. These acquisitions are controversial because a manufacturer's acquisition of PBMs may threaten the PBM's independence and, as a result, its ability to create purchasing efficiencies and secure lower prices for plan sponsors and their subscribers. 6 Critics worry that manufacturers will utilize captive PBMs to foreclose competitor's products from the market and that manufacturer-ownership of PBMs will facilitate collusion in the pharmaceutical manufacturing and PBM markets.
The Commission decided to challenge Eli Lilly s proposed acquisition of PCS Health Systems, the largest PBM in the country. The Commission s complaint alleged that the acquisition might foreclose non-Lilly products from the PCS formulary and that PCS would be eliminated as an independent negotiator of pharmaceutical prices with manufacturers. 7 The complaint also alleged that the acquisition could facilitate collusion through reciprocal dealing, coordinated interaction, and interdependent conduct among Lilly and other vertically integrated pharmaceutical companies. In addition, the complaint alleged that the acquisition could raise entry barriers by effectively requiring an competitor to enter at more than one level. The complaint finally alleged that the acquisition would likely increase prices, diminish quality, and reduce the innovation incentives of other pharmaceutical manufacturers.
Lilly offered to enter into a consent order with the Commission with respect to the acquisition, which the Commission provisionally accepted. The Commission is currently considering whether to make the order final. Of course, questions still remain as to the overall competitive effect of PBM ownership by pharmaceutical companies. As then-Chairman Steiger and I said in our public statement regarding Lilly, "[w]e are concerned about the overall competitive impact of vertical integration by drug companies into the pharmacy benefits management market." Our hope is that through monitoring this proposed order and through analysis of these evolving markets, the Commission can better assess all of the ramifications of vertical integration in these markets.
B. Vertical Integration Issues in Health Care Markets
Lilly reflects the Commission's recognition that vertical integration may cause competitive problems. There has been a dramatic increase in vertical integration in other areas of the health care industry. For example, many hospitals and/or health plans are affiliating with medical clinics and physician practices. Professor Flynn of the University of Utah, in commenting on the recent Columbia/HCA acquisition of Healthtrust, notes that: the nature of competition in health care markets has shifted from competition between individual hospitals or small groups of hospitals to competition between integrated networks; from competition between independent entrepeneur doctors to competition between group practices and integrated HMO s; from competition between clinics and stand alone outpatient facilities to competition between captured insurance subsidiaries of health care networks offering employers complete coverage packages for employees from the cradle to the grave for services provided by -- and only by -- networks. 8
Provider networks are expected to achieve significant economies of scale and scope resulting in lower costs of providing care. These include economies of shared services, coordination economies, and improved organizational incentives for the delivery of cost-effective care. 9 I believe -- and I want to emphasize -- that these vertical arrangements will generally provide procompetitive benefits.
However, competitive harm may result from some types of vertical mergers. For example, vertical mergers can result in competitive foreclosure through the control of necessary upstream inputs, by either making it impossible for competitors to obtain these inputs or by raising their costs in doing so. Statement 9 of the Health Care Policy Statements cautions that the formation of a multiprovider network may foreclose the development of similar, competing networks if the network ties up the providers of a necessary input by means of exclusive contracts. 10 This result could also occur in a hospital s acquisition of physician practices, particularly if the hospital thereby acquired a large percentage of primary care physicians who could in turn control most referrals to specialists.
Vertical mergers or integrations can be anticompetitive in several ways, as reflected in the 1984 Merger Guidelines. 11 First, an industry can become so highly vertically integrated that "two level" entry becomes necessary -- that is, an entrant into either the downstream or upstream markets finds it necessary to enter both markets. 12 If such two-level entry is more risky, difficult or time-consuming than entry into one of the markets alone, a merger that increases vertical integration could increase barriers to entry and thus be anticompetitive. 13
Second, the 1984 Merger Guidelines recognize that vertical integration could foreclose a competitor in the downstream market from purchasing needed supplies: that is, that the integrated firm could use its position as a supplier to disadvantage unintegrated competitors and thereby cause competitive harm, either by restricting supplies or increasing prices. 14
Third, vertical integration can facilitate collusion in either the downstream or upstream market. Integration of a supplier and a purchaser may create opportunities to inappropriately monitor the upstream supplier's competition. 15 Or vertical integration may involve the purchase of a particularly disruptive downstream buyer. 16 By eliminating a buyer who played one upstream firm off of another, such a merger may facilitate collusion in the upstream market.
Theorists have sought to refine the conditions under which vertical integration may be anticompetitive. Professors Riordan and Salop have developed theories of "raising rivals' costs," where an integrated company may increase the costs of its rivals in either the upstream or downstream market.17 One concern about the "raising rivals' costs" theory is that harm to competitors does not always result in harm to competition itself, that is, it may not adversely affect consumer welfare. This concern is exacerbated in the health care industry because arrangements of integrated providers can provide tremendous procompetitive benefits. Thus, in any of these theories, a showing of likely consumer injury should be required before a vertical merger or integration is challenged -- that is, a likely increase in quality- adjusted price or likely decrease in output as a result of the vertical integration. 18
These vertical concerns may apply in the health care area. Consider the following hypothetical. The market has only two hospitals, the dominant one with 70% of the market and a smaller one with 30%. Both of these hospitals operate HMOs. Both of these hospitals have also created networks of physicians associated with the hospital. There are also two other significant physician groups in town. The dominant hospital now wants to acquire its own network of physicians as well as the two independent groups, turning all of them into employees. This vertical integration would leave the bigger hospital in control of roughly 60% of the primary care doctors in the area. Although the network affiliated with the smaller hospital already has 30% of the primary care doctors, those doctors are currently free to contract with others to provide services. One competitive concern -- a vertical one -- is that, as a result of the acquisition, the dominant hospitals lock on primary physicians may disadvantage other health care plans seeking to compete with the hospital s HMO. Indeed, this hypothetical has a double whammy because the entity with 60% of the primary physician market also has a dominant position in the hospital market. Another competitive concern is whether, as a result of this vertical integration, collusion would be likely in one of several horizontal markets.
The main competitive issue to resolve will be whether competing health care plans have adequate alternative means to contract with a sufficient supply of hospital services and primary care physicians at competitive prices or whether these inputs are so foreclosed by this vertical integration that other health care plans costs increase and consumer welfare is reduced. In examining such a situation, we would normally interview other health care plans to see if they are concerned about the possibility of foreclosure or an increase in their costs. We would also examine whether the dominant hospital, in order to remain profitable, would have continuing incentives to offer both its hospital services and its primary physician services to other health care plans at competitive prices.
In this hypothetical, one possibility is that, if the employment contract with the hospital is a sufficiently lucrative one, other health care plans may need to offer expensive inducements to entice the physicians to end the relationship and contract with them, thus effectively raising rivals costs and increasing barriers to entry. Indeed, because the relationship is an employment one rather than an exclusive contract, the costs of luring away primary physicians may be quite high. On the other hand, the remaining nonexclusive primary physicians may be sufficient for a health care plan to contract with, without raising its costs of doing business or entering. Or, if physicians do not find the employment contract sufficiently lucrative or are philosophically opposed to becoming a mere employee of a hospital (as one might imagine), this acquisition may not result in a competitive problem at all -- physicians may simply walk away from the deal.
I caution you that it is important to look at the reality of the underlying arrangements, rather than concentrating on the labels that the parties place on them. For example, in determining whether contractual arrangements in a physician network joint venture are exclusive, the Policy Statements do not simply look at the label that has been placed on the contract, but rather at a series of practical indicia to determine whether a network is truly non-exclusive, such as the extent of participation in other networks, past departicipation, and the presence of viable alternatives.
Of course, into this mix, we need to examine carefully the procompetitive benefits of the vertical integration. There is a perverse logic at work here that is not easily resolvable. We know that the arrangements that provide the greatest procompetitive, efficiency-enhancing benefits are those that selectively contract so that they can control costs by ensuring that providers have the strongest incentives not to over treat or over utilize. A market with five competing networks, each contracting exclusively with twenty percent of the market s physicians, is likely to realize greater price competition and innovation than one in which each of five plans had contracts with all the doctors in the market. 19 In this hypothetical, for example, the fact that the arrangement fully integrates doctors as staff employees of the integrated provider would seem to suggest that it will provide tremendous procompetitive benefits. The irony is that, the more exclusive the vertical arrangement is, the greater the potential for anticompetitive foreclosure and resulting collusion in the horizontal markets.
This is not an easily resolvable conundrum, but I believe that a full analysis of all aspects of the market -- as we do in a rule of reason analysis -- allows us to reach a considered judgment as to the overall competitive impact of a particular vertical integration. Most importantly, we need to examine how great the foreclosure effect on the market is and see what is happening in the rest of the market. We can also look at other markets to obtain a sense of whether the prices the integrated provider is offering seem relatively competitive. Of course, if the vertical integration is really intended to keep out competition or drive prices up, there may be documentary evidence of intent that may make the competitive effect of the integration more obvious and our job easier. Finally, we need to examine whether the efficiencies are real and whether they are integration-specific: that is, are there other less restrictive alternatives by which the vertically integration could achieve comparable efficiencies?
Past enforcement actions give us some guidance as to when anticompetitive vertical foreclosure is more likely. In 1984, for example, the Department of Justice announced plans to challenge a preferred provider organization that DOJ alleged sought to restrain competition by enrolling nearly 90% of physicians in one market and 50% of physicians in another market as members and forbidding them from contracting with other PPOs or health care delivery systems. 20 By contrast, the FTC, in reviewing a hospital s exclusive contract for radiologists, gave a favorable advisory opinion when the hospital would have only a 26% share of the radiologists market and the contract was only three years old, with either side able to terminate the contract on short notice. 21 Similarly, the First Circuit rejected an antitrust challenge to an HMO plan that offered its physicians higher capitation payments in exchange for the physicians agreement not to participate in any other HMO. 22 Although the opinion is of limited precedential value for the issues we are dealing with here,23 Judge Boudin noted that only 25% of physicians were tied to the HMO and that the exclusivity clause could be changed on one months notice.
Of course, in examining vertical foreclosure, we must be sensitive to differences in geographic area. In rural areas, for example, the number of providers is by nature limited. I would caution against any rule that would, across the board, make it impossible for rural area providers to engage in procompetitive exclusive arrangements. 24 I stress procompetitive arrangements, because I also believe that consumers in rural areas should be able to expect the same procompetitive benefits of the antitrust laws that we afford consumers everywhere. My sense is that much of the special nature of rural markets will be captured by efficiencies arguments and, consequently, that the best way to ensure that we are sensitive to rural concerns is by being sensitive to efficiency concerns.
Although I know that you have a panel later this morning discussing the interesting Marshfield Clinic case, I would be remiss if I did not mention that recent decision here. As I am sure you are aware, the federal district court judge there upheld a jury verdict that Marshfield Clinic and its HMO violated the antitrust laws by allegedly excluding the Blue Cross HMO from the HMO market and by allegedly requiring Blue Cross to pay supracompetitive prices for the services of its employee-physicians and affiliated doctors. 25
In short, I believe that vertical competition issues are going to play an increasing role, particularly as health care markets move toward increased integration in the delivery of health care services. I also believe that antitrust has a role to play in ensuring that vertical integration in health care markets is accomplished in the most procompetitive fashion possible.
I hope I have given you a sense of the new directions that the Commission is pursuing in both the hospital merger area and in vertical integration issues. Again, I want to stress that I believe we, as government enforcement officials, need to be working with you as closely as possible on these issues. I urge you to contact me and staff at the Commission as the new Chairman begins to tackle these tough issues in health care.
1 Phillip Areeda & Donald F. Turner, IV Antitrust Law 921 (1980).
2 Adventist Health System/West, Dkt. No. 9234, slip op. at 9 (April 1, 1994).
3 Dennis A. Yao & Thomas N. Dahdouh, Information Problems in Merger Decision Making and Their Impact on Development of an Efficiencies Defense, 62 Antitrust L.J. 23, 30 (1993).
4 Robert Pitofsky, Proposals for Revised United States Merger Enforcement in a Global Economy, 81 Geo. L.J. 195, 207-08 (1992).
5 Yao & Dahdouh, supra at n.3, at 27, n.9 (listing various proposals for alternative standards for efficiency justifications).
6 See "The ABCs of PBMs," Drug Topics 67 (Sept. 5, 1994).
7 Eli Lilly & Co., File No. 941-0102 (Nov. 3, 1994).
8 Letter from Professor John J. Flynn to Utah Governor Michael O. Leavitt at 1 (Dec. 22, 1994).
9 For an interesting article that provides an in- depth look at vertical issues, see Thomas L. Greaney, Managed Competition, Integrated Delivery Systems and Antitrust, 79 Cornell L. Rev. 1507 (1994).
10 U.S. Department of Justice and Federal Trade Commission, Statements of Enforcement Policy and Analytical Principles Relating to Health Care and Antitrust (September 27, 1994) at 100-01.
11 U.S. Dep t of Justice, Merger Guidelines, 4 Trade Reg. Rep. (CCH) 13,103 (1984) (hereinafter 1984 Merger Guidelines). The provisions in the 1984 Merger Guidelines regarding non-horizontal mergers have not been superseded by the 1992 Guidelines.
12 1984 Merger Guidelines 4.21.
13 One reason why two-level entry might be costlier is that the minimum efficient scale for two- level entry is different than for one-level entry, which may force the entrant to enter at a much larger scale than it would otherwise. Another is that potential entrants may be less efficient at one stage of production than another.
14 1984 Merger Guidelines 4.212 n.31.
15 1984 Merger Guidelines 4.221.
16 1984 Merger Guidelines 4.222.
17 Michael H. Riordan and Steven C. Salop, "Evaluating Vertical Mergers: A Post-Chicago Approach," 63 Antitrust L.J. 513 (1995). See also Thomas G. Krattenmaker & Steven C. Salop, "Anticompetitive Exclusion: Raising Rivals' Costs to Achieve Power Over Price," 96 Yale L.J. 209 (1986).
18 Riordan & Salop, supra n.17, at 548-50.
19 Greaney, supra n. 9, at 1536.
20 Stanislaus Preferred Provider Organization, Inc., Dept. Of Justice Press Release (Oct. 12, 1984) (no lawsuit was filed because, the press release noted, the organization had decided to dissolve itself).
21 Burnham Hospital, 101 F.T.C. 991 (Feb. 24, 1983).
22 U.S. Healthcare, Inc. v. Healthsource, Inc., 986 F.2d 589 (1st Cir. 1993).
23 The plaintiff had pursued the case solely on a per se theory of liability and, therefore, had refused to offer any evidence in support of a rule of reason challenge to the exclusive arrangement. Under Supreme Court precedent, exclusive dealing contracts are not considered per se unlawful. Tampa Electric Co. v. Nashville Coal Co., 365 U.S. 320 (1961).
24 In Oltz v. St. Peter s Community Hospital, 861 F.2d 1440, 1449 (9th Cir. 1988), the court condemned an exclusive contract for anesthesiologists where the hospital was the only effective source of anesthesia services in the market. The court, however, went to great pains to note that the exclusive contract may have been the result of a conspiracy between anesthesiologists and the hospital to exclusive competitors and, consequently, warned that its decision cannot be read as establishing any rule applicable to other situations involving rural hospitals engaged in exclusive contracts for staff privileges. Id. at 1449.
25 Blue Cross & Blue Shield United of Wisconsin v. The Marshfield Clinic, No. 94-C-137-S, (W.D. Wisc. March 22, 1995).