The changing nature of health care delivery, the related surge of hospital mergers, and the growth of innovative provider networks have made antitrust's role critical for the health care sector. I, like others at the FTC, believe that conventional antitrust principles not only do but should apply to hospital mergers and the health care field in general. People have recently raised the issue of whether current antitrust tools are appropriate for evaluating competitive issues in health care. Interestingly, several years ago, DOD established an advisory commission to ask a related question -- is antitrust up to the task of evaluating defense mergers? The answer was yes. Moreover, the defense industry was undergoing even more radical consolidation and budget cutting than is occurring in the health care sector. So I have little doubt that, just as we discovered in the defense sector, antitrust law is sufficiently flexible to take into account the particular characteristics of health care markets and the rapid changes occurring in those markets.
Whether the market is for paper clips or pacemakers, notepads or newborn care, Congress has determined -- and our economic history shows -- that competition produces the best mix of price and quality for consumers. Therefore, the goal of the antitrust enforcement agencies is to ensure a competitive marketplace in which consumers will have the benefit of high quality, cost-effective health care and a wide range of choices. Divergence from the competition paradigm, even for laudable goals by the most honorable people, goes against Congress' design. It also risks affecting quality and/or cost to the detriment of consumers. They are the ones who ultimately receive more limited medical treatment or bear greater costs through higher insurance premiums, reduced employee benefits, or increased taxes.
The Efficiencies Revisions to the Merger Guidelines
I would like to take this opportunity to revisit the efficiencies revisions to the Merger Guidelines, which grew out of the FTC's staff report on Competition Policy in the New, High Tech, Global Marketplace. What I would like to examine is whether those revisions are helping us to evaluate efficiencies in hospital mergers, or in mergers generally. I should note that there were hearings that preceded issuance of the staff report, and that provided much of the empirical basis and analytic structure for the report. The testimony about efficiencies in hospital mergers was among the most interesting, controversial, and detailed of the efficiencies contributions we received.
A. What Does the Revised Section Say?
First, let me briefly recap the April 1997 efficiency revisions to the Horizontal Merger Guidelines.(1) Those revisions clarify but do not radically alter how the agencies analyze efficiencies claims in mergers. A clear articulation of policy is important for several reasons. First, it assists the merging parties in explaining and justifying their efficiencies claims. Second, it improves the quality of decision making -- it ensures that the agencies recognize valid efficiency claims and reject those that are unsubstantiated or otherwise not cognizable. Third, it improves the predictability of the process and ensures consistency of analysis, both within and across the antitrust agencies. Central to the revisions is a recognition that cost savings and other efficiencies from a merger can enhance the merged firm's ability and incentive to compete. They should therefore be taken into account by the agencies when analyzing a merger's competitive effects.
The newly revised section first describes how a merger creates efficiencies, which is through a better utilization of existing assets. It also defines what an efficiency is or what effect it has. That is, efficiencies enable the merging firms to achieve lower costs in producing a given quantity and quality than either firm could have achieved without the proposed transaction.(2)
The revisions next discuss how efficiencies may affect a single firm's ability and incentive to compete as well as -- and this is the important part -- affect overall competition in the market.(3) The key insight here is that cost reductions arising from merger-generated efficiencies may be procompetitive because they may tend to enhance the merged firm's incentive to reduce price or increase output. The section gives examples of how cost reductions may affect a firm's incentives. One obvious example is where merger-generated efficiencies enhance competition by allowing two ineffective, high-cost competitors, let's say hospitals, to become an effective, low cost hospital. Perhaps the most direct linkage between efficiencies and market outcomes is that between reduced marginal cost and reduced prices. The revised provision strongly implies that efficiencies affecting marginal cost are more likely to be cognizable than other claimed efficiencies. Lower marginal production costs can occur where firms reallocate production among facilities previously separately owned, where the merged entity reallocates the firms' physical and human assets to exploit scale and scope economies, and where the merging firms have comparative advantages in different stages of production.
Efficiencies also may enhance competition through innovations or quality improvements that affect a firm's ability to compete. Even reductions in fixed costs, which do not directly determine short-term pricing, may count as efficiency gains.(4) For example, by lowering the merged firm's long-run average incremental costs, fixed cost reductions can make a firm lean and mean, enabling it to exert a stronger, more destabilizing presence in the market. Moreover, fixed costs may in fact be variable ones over the medium term or if a relatively large rather than small change in output is involved. To the extent that fixed cost efficiencies are claimed in hospital mergers, however, I caution that you must demonstrate to us the link between fixed cost reductions and enhanced competition.
The new efficiencies section also adopts a definition of merger-specific that is analogous to the standard set forth in the Intellectual Property Guidelines. Thus, the agencies will consider only those efficiencies likely to be accomplished with the merger and unlikely to be accomplished without either the merger or another means having comparable anticompetitive effects. The Guidelines make clear that the focus is on practical alternatives that the merging firms face in their business situation. They also noted that the agencies will not demand a hypothetical, merely theoretical, less restrictive alternative.(5)
The new section carefully defines how the parties must substantiate the claimed efficiencies. Evidence of efficiencies is usually, and occasionally uniquely, in the hands of the parties. For enforcers, it tends to be more difficult to confirm through third parties than other facts relevant to merger analysis. And human nature is such that merging firms tend to be very sanguine about projecting efficiencies. But often their expectations (even when entirely in good faith) surpass what is achieved in reality. The agencies therefore require that merging firms substantiate their claims so that we can verify by reasonable means (1) the likelihood and magnitude of each efficiency, (2) how and when each will be achieved (and the costs of doing so), (3) how each will enhance the merged firm's ability and incentive to compete, and (4) why each one is merger-specific.(6) Moreover, cognizable efficiencies may not arise from anticompetitive reductions in output or service.(7) In sum, the firms bear a heavy burden in demonstrating their efficiency claims.
The section then addresses the critical question: how do the agencies incorporate efficiencies into their analysis of a merger? The overall standard is that a merger will not be challenged if the cognizable efficiencies are of such a type and magnitude that the merger is not likely to be anticompetitive in any relevant market. Thus, the agencies evaluate whether the procompetitive incentives from efficiencies are sufficient to counteract the anticompetitive incentives brought about by increased concentration. There is a sliding scale aspect to this analysis -- the greater the potential adverse competitive effects, the greater the cognizable efficiencies must be. This is not a one-to-one comparison, however. Efficiency-based cost savings in dollars per unit will often have to be appreciably larger than the anticipated price increase to prevent the likely exercise of market power. Accordingly, the section notes that when the potential adverse competitive effect of a merger is likely to be particularly large, the cognizable efficiencies need to be extraordinarily great to prevent the merger from being anticompetitive.(8)
Lastly, the revised section states two rules of thumb that are based on the agencies' experience in evaluating efficiency claims. The first rule indicates when efficiencies are most likely to make a difference, which is when the likely adverse competitive effects, absent the efficiencies, are not great. The second rule indicates when efficiencies are least likely to matter --- they will almost never justify a merger to monopoly or near-monopoly.(9)
I would now like to turn to three recent merger cases, two in the health care field and one outside it, where efficiency defenses were evaluated with differing results.
B. Recent Cases with Efficiencies Claims
The first case, FTC v. Butterworth Health Corporation, 946 F. Supp. 1285 (W.D. Mich. 1996), aff'd, 121 F.3d 708 (6th Cir. 1997), was briefed and decided, at least at the district court level, before the revision of the Guidelines' efficiencies section. In that case, the FTC sought to preliminarily enjoin the merger of the two largest hospitals in the Grand Rapids, Michigan, area. The district court found that the FTC had established that the merger would result in a significant increase in concentration in the markets for primary and general acute inpatient hospital services. In the judge's words, it would "produce an entity controlling an undue percentage share of each of those markets."(10) Nevertheless, the court concluded that the best interests of the public would be served by permitting the hospitals to achieve the efficiencies that allegedly would result from the proposed merger and that would enable the board of directors of the combined entity to establish world-class health facilities in West Michigan.(11)
How could the court have reached this conclusion? The court primarily reasoned that the hospitals' non-profit status would somehow mitigate the merged entity's combined market share of approximately 65%.(12) But in doing so, the court also rather uncritically accepted the hospitals' efficiencies defense. Interestingly -- even without benefit of the revised Guidelines -- the court started by asking the right question. Citing University Health, the judge noted that "[i]n order to overcome the presumption arising from the FTC's prima facie case that the proposed merger would substantially lessen competition, defendants 'must demonstrate that the intended acquisition would result in significant economies and that these economies ultimately would benefit competition and hence, consumers.'"(13) But the judge never answered whether the proposed merger would benefit, or at least not substantially lessen, competition.
Instead, the court arrived at a general lump conclusion that the proposed merger would achieve in excess of $100 million in capital expenditure avoidance and operating efficiencies, largely accepting the two hospitals' claim that if the merger were allowed, they could avoid approximately $100 million of capital expenditures in their ongoing "medical arms race" and over $68 million in operating efficiencies in the 5 years after the merger.(14) The judge did so by incorrectly discounting the FTC's testimony because it was, in his view, not an independent efficiencies study but a "mere critique."(15) He did not scrutinize whether each claimed cost saving was truly merger-specific, or whether it reasonably could have been achieved without the merger. Nor did he ask how the claimed efficiencies would enhance the merged hospital's ability to compete. No doubt, virtually every horizontal merger has the potential to achieve cost savings through reductions in capital expenditures, administrative overhead, or duplicative services. But another term for duplication of services is competition. Rather than inquire how the savings would improve the merged hospital's ability and incentive to compete, the court simply assumed that the hospital's non-profit status and Community Commitment would ensure that savings would be passed on to consumers.(16) This is an especially troubling assumption given that the hospitals had recently enjoyed very high profit margins, which tends to undermine the conclusion that they would pass on future savings to consumers.
At the end of the day, the court literally threw up its hands. It noted: "[b]ecause measuring the efficiencies of a proposed transaction is inherently difficult and because both sides' estimates are clearly based in some measure on speculative self-serving assertions, . . . the Court finds it neither appropriate nor necessary to engage in a detailed evaluation of the competing views."(17) This abdication in part stemmed from the court's overly sanguine view of nonprofit hospitals in general and its belief that hospital boards drawn from the community are a superior proxy for consumers' interests than is competition. In a sense, I can not entirely blame the court. Efficiencies claims are difficult to evaluate and until the 1997 revisions, I do not think it was entirely clear how a decision maker should wrestle with these claims. I hope that future courts will benefit from the revisions' recognition that, other things being equal, as market shares increase, the cost savings necessary to prevent a price increase not only increase but do so at an increasing rate. Thus, in the words of the Guidelines, "when the potential adverse competitive effect of a merger is likely to be particularly large," the cognizable efficiencies need to be "extraordinarily great" to prevent the merger from being anticompetitive.(18) Perhaps most telling for the Butterworth situation is the Guidelines' admonition that "efficiencies almost never justify a merger to monopoly or near-monopoly."(19)
Let's examine another recent case, United States v. Long Island Jewish Medical Center, No. 97-3412 (E.D.N.Y. Oct. 23, 1997), decided after the efficiencies revision to the Guidelines. There, the Department of Justice sought to enjoin the merger of Long Island Jewish Medical Center, which consists of an acute care adult hospital, a children's hospital, and a psychiatric hospital, and North Shore Health Systems, which is made up of nine hospitals, including North Shore Manhasset. According to the court, the main hospitals involved in the merger, North Shore Manhasset and Long Island Jewish, are two of the premier hospitals on Long Island and were "fierce competitors."(20) The real battle in this case was whether the relevant product market was the bundle of acute inpatient primary and secondary care services that anchor hospitals provide to managed care plans, as Justice contended. The court rejected Justice's market definition as too narrow and found the relevant market to be the more traditional one for general acute inpatient hospitals services.(21) The court ultimately relied on (1) testimony of a representative of a large managed care organization that the merger would reduce prices, (2) the hospitals' stipulation to the state attorney general that they would not raise prices for two years, and (3) the hospitals' low percentages of total numbers of "patient days" for the geographic market to find that the merged entity would not have an undue share of the market and that a price increase was unlikely.(22) The court stated that it reached this conclusion despite the fact that the hospitals are the two premier teaching hospitals in the geographic market, are direct competitors, and are sought after by managed care companies.(23)
The court seemed to conclude that a price rise was unlikely regardless of whether any efficiencies were realized. Nevertheless, the court went on to balance the reduced competition from the merger against the projected efficiencies. As in Butterworth, the court started with the right question: "with regard to the so-called 'efficiencies defense,' the defendants must clearly demonstrate that the proposed merger itself will, in fact, create a net economic benefit for the health care consumer."(24)
The Long Island court's opinion also reflects an improved attempt to identify cost savings with some specificity and to evaluate whether they were in fact merger specific. The court specifically examined a range of claimed efficiencies, running from the elimination of managerial, administrative and clinical employees, reduced capital expenditures, economies in purchasing medical supplies, to reduced laundry bills. Major savings purportedly would come in the areas of professional services, information systems, and physician support services.(25) The court properly questioned whether many of these savings could be achieved by each hospital acting independently. And it concluded that the hospitals could individually achieve reductions in human resources personnel, claims recovery costs, insurance premiums, interest expense savings, capital avoidance items and faculty savings.(26)
Consistent with the revised Guidelines, the court also noted that the hospitals' expert failed to account for the costs and difficulties of achieving the claimed efficiencies. For example, the hospitals ignored the "start up" and "termination compensation" costs related to firing large numbers of employees.(27) Ultimately, the court concluded that the merger would result in operating efficiencies of $25 to $30 million a year -- a substantial discount from the hospitals' optimistic $75 million, but more robust than the government's projected $6.3 million yearly savings.(28)
The greatest flaw in the court's efficiency analysis is one that pervades the Butterworth opinion as well. In asking whether the cost savings would be passed on to consumers, the court did not focus on competition as critical to that outcome. Instead, the court found the hospitals non-profit status and "genuine commitment to help their communities" to be circumstantial evidence that cost savings would inure to the benefit of consumers.(29) The court also placed weight on the hospitals' agreement with the New York attorney general to pass on $100 million in cost savings to the community over a 5 year period. In the court's mind, this provided "reasonable certainty" that the efficiencies from the merger would ultimately benefit consumers.(30) As I noted with respect to Butterworth, promises and good intentions are admirable, but they do not substitute for competition over the long term.
Finally, I want to discuss a case that does not involve health care but which presents a model application of the revised efficiencies provision. In FTC v. Staples, Inc., 970 F. Supp. 1066 (D.D.C. 1997), we challenged the merger of the first and second largest office product superstores. The court accepted our definition of the relevant product market as the sale of consumable office supplies through office superstores.(31) It also found persuasive our evidence demonstrating that Staples and Office Depot were able to charge higher prices in markets with only one office supply superstore than in markets with two or three superstores, regardless of the presence of other suppliers in those markets.(32) Based on this evidence and the high concentration that would result from the merger, the court concluded that the FTC established a reasonable probability that the merger would have an anticompetitive effect.(33)
The court then considered whether Staples' and Office Depot's claimed efficiencies could rebut the presumption that the merger may substantially lessen competition. The superstores submitted an efficiencies analysis that predicted huge savings of approximately $5 to $6.5 billion over five years. They also asserted that two-thirds of those savings would be passed along to consumers. The court identified several flaws in the analysis, however. First, it found that the claims were not based on "credible evidence" and appeared to be grossly exaggerated.(34) For example, the court noted that the savings claimed in court were almost 500% greater than the figures presented to the stores' Boards of Directors when they approved the transaction and were substantially greater than those claimed in the superstores' joint filing with the SEC.(35) Second, it observed that many of the projected savings were unverified.(36) Third, and perhaps most important, the court found that many of the savings were not merger-specific. For example, the stores claimed that they would achieve the largest savings, over $2 billion, by obtaining better prices from vendors. But this huge estimate was based on a comparison with the cost savings that Staples enjoyed at the end of 1996 rather than the even greater savings Staples expected to receive in each future year as a stand-alone company. As the court astutely noted, to identify merger-specific savings, one must compare the projected (not the past) cost savings of Staples as a stand-alone company with the projected cost savings of the merged company.(37) In addition, both parties to the merger were expanding rapidly by opening new stores, as many as 100 or 150 each per year. Consequently, even if increased buying power could be used to extract better prices from vendors, such savings would have occurred as a result of internal expansion in any event. At most, any merger-related buying efficiencies were limited to those savings that flowed from reaching a larger scale immediately rather than over a period of 3 or 4 years.(38) But those efficiencies would have been temporary and declining. The merger and its anticompetitive effects, in contrast, would be permanent.
In a refreshing contrast to Butterworth and Long Island, the court also questioned the claim that the merged store would pass two-thirds of its savings on to consumers. Evidence showed that Staples' historic pass-through rate was only 15-17%.(39) Logically, the increased concentration would, if anything, cause that rate to fall. Accordingly, the court concluded that Staples and Office Depot did not rebut the presumption that the merger would substantially lessen competition.(40)
This analysis is consistent with the revised Guidelines. Given that the merger's likely anticompetitive effect was so great (a 13% price increase at retail in many markets), and the credible efficiencies so modest, it is not surprising that the claimed efficiencies would not be passed on to consumers in an amount sufficient to counteract the merger's anticompetitive impact.(41) Moreover, in cities where Staples and Office Depot operated, but Office Max was not present, the merger would have led to monopoly or near monopoly. As the revised Guidelines note, in those circumstances efficiencies will almost never justify a merger.(42)
The court's rigorous analysis of the claimed efficiencies presents an excellent model for evaluating such defenses. First, the court evaluated the evidence based not solely on the companies' "best case" litigation posture but in comparison with the companies' own statements to their directors and shareholders and their previous pricing behavior. This helped counteract the artificial and speculative nature of the experts' efficiencies analysis by comparing it to what the companies' say and do in the real world. Second, the court grasped the concept of a merger-specific efficiency: it is not just any efficiency after the merger but one that is not reasonably achievable by other, no more anticompetitive, means. Third, the court did not allow the companies to rely on unverified assertions of savings but instead required documentation of the claims. Finally, and most importantly, the court did not accept at face value the rosy prediction that virtually all the efficiencies gains would be passed on to consumers. Instead, it looked to the firms' historical record and the market context to get a more realistic prediction of future behavior.
My hope is that we should soon see the same analysis appearing in hospital cases in the future. The revised Guidelines establish a framework for parties to follow in making efficiencies claims, for the agencies to follow in evaluating and verifying those claims, and for courts to follow in assessing the credibility and competitive impact of those claims.
* The views expressed here are those of the author, and not necessarily of the Federal Trade Commission or any Commissioner
1. U.S. Dep't of Justice and the Federal Trade Comm'n, Horizontal Merger Guidelines (1997).
2. Horizontal Merger Guidelines, § 4, para. 1.
3. Id. § 4, para. 2.
4. Id. § 4, para. 2.
5. Id. § 4, para. 3.
6. Id. § 4, para. 4.
7. Id. § 4, para. 5.
8. Id. § 4, para. 6.
9. Id. § 4, para. 7.
10. FTC v. Butterworth Health Corporation, 946 F. Supp. 1285, 1294 (W.D. Mich. 1996), aff'd, 121 F.3d 708 (6th Cir. 1997).
11. 946 F. Supp. at 1302.
12. Id. at 1295-96.
13. Id. at 1300.
14. Id. at 1301.
18. Horizontal Merger Guidelines, § 4, para. 6 (emphasis added).
19. Id. § 4, para. 7.
20. United States v. Long Island Jewish Medical Center, No. 97-3412, slip op. at 18 (E.D.N.Y. Oct. 23, 1997).
21. Id. at 41-42.
22. Id. at 52-54.
23. Id. at 55.
24. Id. at 59.
25. Id. at 61.
26. Id.at 63.
27. Id. at 61.
28. Id. at 63-64.
29. Id. at 64.
31. FTC v. Staples, Inc., 970 F. Supp. 1066, 1080 (D.D.C. 1997).
32. Id. at 1077.
33. Id. at 1082-83.
34. Id. at 1089.
35. Id. at 1089.
36. Id. at 1089-90.
37. Id. at 1090.
38. Horizontal Merger Guidelines, § 4, para. 3 n.35.
39. 970 F. Supp. at 1090.
41. Cf. Horizontal Merger Guidelines, § 4, para. 6.
42. Id. § 4, para. 7.