Federal Trade Commission v. Freeman Hospital

BEAM, Circuit Judge. The United States Federal Trade Commission (FTC), sought a preliminary injunction from the district court to prevent the merger of two hospitals in Joplin, Missouri, contending the transaction would have anticompeti-tive effects in violation of Section 7 of the Clayton Act, 15 U.S.C. § 18. The district court denied the requested relief and the FTC appeals. Because the FTC has failed to meet its burden for obtaining preliminary injunctive relief, we affirm. I. BACKGROUND The city of Joplin, Missouri, a community of approximately 40,000 people, is located in the southwest corner of Missouri near the confluence of four states. Joplin is situated just five miles from the Kansas state line, forty miles from the northwest comer of Arkansas, and twelve miles from the northeast corner of Oklahoma. Joplin is also located within a few hours of several larger metropolitan centers, including Kansas City, Missouri, Springfield, Missouri, and Tulsa, Oklahoma. Joplin is currently home to three general acute care hospitals. The largest of these hospitals is St. John’s Regional Medical Center, a nonprofit, allopathic hospital with 331 beds. The other two hospitals are significantly smaller. Freeman Hospital, also a nonprofit allopathic institution, houses 158 beds. Oak Hill Hospital is a 96-bed hospital and the only hospital in Joplin with an osteopathic, rather than an allopathic, orientation.1 Surrounding Joplin are several smaller communities, many of which have their own acute care hospitals ranging from eighteen to 161 beds. In February 1994, Freeman Hospital and Oak Hill Hospital (the Hospitals) agreed to consolidate their assets and form a new nonprofit organization, Health SouthWest Alliance of Missouri, Inc. Oak Hill had been experiencing financial difficulties, and its trustees believed a merger with Freeman would strengthen Oak Hill’s financial standing and enable it to better compete in a changing health care market. Accordingly, on July 21,1994, pursuant to the Hart-Scott-Rodino Act, 15 U.S.C. § 18a,2 the Hospitals filed a premerger notification form with the [263] FTC outlining the transaction. On August 19, 1994, the FTC requested more information about the transaction. It sought to determine whether the proposal violated Section 7 of the Clayton Act, 15 U.S.C. § 18, which prohibits certain mergers and acquisitions if “in any line of commerce ... in any section of the country, the effect of such acquisition may be substantially to lessen competition.” The Hospitals promptly complied with this request. Under the Premer-ger Notification Act, however, the FTC’s issuance of a request for more information suspended the statutory waiting period for completing the transaction3 and prevented the Hospitals from merging as planned. While the Hospitals awaited the FTC’s clearance for the merger, the FTC suggested that Oak Hill’s trustees determine whether any other entities were interested in acquiring Oak Hill. Oak Hill subsequently solicited other offers for the hospital to be submitted by early January 1995. Two for-profit health care companies submitted bids for Oak Hill. A group of physicians from Joplin also expressed interest in acquiring it. In January 1995, however, Oak Hill announced that it had decided to pursue its earlier arrangement with Freeman Hospital rather than accept another bid.4 Shortly after Oak Hill’s announcement, the FTC filed a complaint alleging that the Hospitals’ consolidation would lessen competition for acute care inpatient hospital services in the Joplin area. Pursuant to Section 13(b) of the Federal Trade Commission Act (FTC Act), 15 U.S.C. § 53(b),5 the FTC sought a temporary restraining order and a preliminary injunction to enjoin the Hospitals from merging pending an administrative determination of whether the merger would violate Section 7 of the Clayton Act. The following day, the parties appeared and presented arguments to the district court on the FTC’s request for a temporary restraining order. After hearing the arguments, the district court orally denied the motion for a temporary restraining order, stating: I’m denying the TRO because based upon the pleadings filed here, I don’t feel that the Federal Trade Commission has shown sufficient factual basis that they are entitled to a TRO.... I don’t think you’ve got any business being in here. I don’t see how the Federal Trade Commission can claim there is lack of competition when there [are] four or five hospitals in the area, and reducing it by one is not going to wipe out competition. It looks to me like Washington D.C. once again thinks they know better what’s going on in southwest Missouri. I think they ought to stay in D.C. Temp.Restraining Order Hrg.Tr. at 27-29 (Febr. 22, 1995). On February 28, 1995, without holding an additional evidentiary hearing, the district court issued a written order reiterating its oral denial of the temporary restraining order and denying, in addition, the FTC’s application for a preliminary injunction. [264] By order of March 1, 1995, in the midst of various procedural maneuvers, this court considered the FTC’s notice of appeal of the district court’s decision. We noted that the district court had not held an evidentiary hearing on the issue of the preliminary injunction, and declined to address the substantive issues raised by the parties without a fully developed record. Accordingly, we entered a stay order, retained jurisdiction of the appeal and remanded the matter to the district court for an evidentiary hearing. A. Preliminary Injunction Evidentiary Hearing On March 23 and 24, 1995, the district court held an evidentiary hearing to determine whether to grant the FTC’s requested preliminary injunction. Each side was permitted to present three witnesses and offer additional deposition testimony and exhibits. A voluminous record resulted, in which the parties attempted to establish the Hospitals’ area of competition and to predict the competitive effects of the proposed merger. At this hearing, both sides relied heavily on expert testimony. Dr. Keith Leffler, an economist at the University of Washington, appeared for the FTC. Dr. Leffler testified that the Hospitals were part of a market for acute care inpatient hospital services which encompassed Joplin, Missouri, and areas located within a twenty-seven-mile radius of the city.6 To reach this conclusion, Dr. Lef-fler applied the Elzinga-Hogarty test, a method devised by professors of economies Kenneth G. Elzinga and Thomas F. Hogarty to analyze patterns of consumer origin and destination and to identify relevant competitors of the merging entities. See United States v. Rockford Memorial Corp., 717 F.Supp. 1251, 1266 (N.D.Ill.1989), aff'd, 898 F.2d 1278 (7th Crr.), cert. denied, 498 U.S. 920, 111 S.Ct. 295, 112 L.Ed.2d 249 (1990). The first prong of the Elzinga-Hogarty test requires a determination of the merging hospitals’ “service area,” the area from which they attract their patients.7 To determine the service area, Dr. Leffler examined the zip codes of patients discharged from the three hospitals in Joplin. After accounting for approximately eighty percent8 of the patient population at those hospitals, he arranged them in order of distance from Joplin and developed a preliminary map of the service area. Within that area were other hospitals, and Dr. Leffler performed the same analysis for those hospitals to complete a map of the collective service area of all the hospitals. The second step in the Elzinga-Hogarty test requires an analysis of where patients within the collective service area currently go to receive their health care.9 In undertaking [265] this analysis, Dr. Leffler again used patient zip code information and determined that approximately ninety percent of hospital admissions of people residing in the proposed service area were admissions to hospitals within the service area. Relying on his test results and statements of various market participants regarding the nature of the health care market in southwest Missouri, Dr. Leffler delineated his geographic market spanning twenty-seven miles in every direction from Joplin, Missouri, and encompassing seven acute care facilities. To this market, Dr. Leffler applied the Herfin-dahl-Hirsehmann Index (HHI), an economic index designed to provide a numerical measure of concentration in a given market. Dr. Leffler found that within his proposed market the HHI indicated a highly concentrated industry which presumptively raises antitrust concerns. The Hospitals’ expert, Dr. William Lynk,10 presented a much different picture of the geographic market in which the Hospitals compete. Dr. Lynk also performed the first prong of the Elzinga-Hogarty test to determine the Hospitals’ “service area,” but his method differed from Dr. Leffler’s approach in two important respects. First, Dr. Lynk used a ninety percent inclusion criterion to account for a greater percentage of patient business. Second, he arranged the zip codes contributing patients to the service area not by distance from Joplin but by economic importance to the hospitals, defined in terms of number of patients contributed. Dr. Lynk did not complete the second prong of the Elzinga-Hogarty analysis, as it was his opinion that an accurate picture of where patients in the service area seek health care could only be presented with admission data from all four states in the area. Unfortunately, complete admission data was available only from Missouri and not from Oklahoma, Arkansas, or Kansas. Dr. Lynk testified that the absence of complete data from these states rendered a reliable analysis of patient outflow impossible. To substitute for the second component of the Elzinga-Hogarty test, Dr. Lynk applied principles of geographic proximity, reasoning that if patients from communities surrounding Joplin would travel to Joplin for their health care, they could as easily drive that same distance to another town if Joplin’s health care costs rose. Dr. Lynk’s finalized market consisted of a thirteen-county area which included seventeen hospitals located up to fifty-four miles from Joplin. In this area, he determined that the market share owned by Health SouthWest Alliance would be no cause for antitrust concern. Both sides also presented the testimony of individuals familiar with Joplin’s health care market and with the proposed merger. That testimony focused on current perceptions of the geographic market in which the merging hospitals compete and on the fingering interest of other health care companies in acquiring and maintaining Oak Hill Hospital as a competitor in the Joplin area. B. The District Court’s Opinion The district court denied the FTC’s request for preliminary injunctive relief. Recognizing the necessity of first defining a geographic market within which the competitive effects of the merger could be evaluated, the court began its inquiry by examining the expert testimony. The court found several flaws in Dr. Leffler’s analysis. Initially, the court criticized Dr. Leffler’s use of a lower inclusion percentage to establish the Hospitals’ service area. The court also questioned the reliability of the data used by Dr. Leffler because the source from which he obtained his patient admission information had quoted two different patient admission totals for the same year. Finally, the court discounted Dr. Leffler’s attempt to analyze patient outflow according to the Elzinga-Hogarty test, since he lacked complete information about patient admissions in Oklahoma, Arkansas, and Kan [266] sas. Although the court acknowledged that the FTC did not rely solely on Dr. Leffler to develop its market, it characterized the statements of market participants as “informal, off-the-cuff remarks” which could not substitute for economic analysis. Federal Trade Comm’n v. Freeman Hosp. & Tri-State Osteopathic Hosp. Ass’n, No. 95-5015-CV-SW-1, mem. op. at 12, 1996 WL 462226 (W.D.Mo. June 9, 1995) (hereinafter Mem. op.). In making its findings regarding the relevant market, the district court gave significant weight to Dr. Lynk’s testimony. It found Dr. Lynk’s construction of the Hospitals’ service area by patient contribution numbers more “economically sound” than Dr. Leffler’s method. Id. at 14. It also preferred Dr. Lynk’s geographic proximity analysis over Dr. Leffler’s attempt to conduct the second prong of the Elzinga-Hogarty test with incomplete data. Therefore the district court analyzed the potential competitive effects of the Freeman-Oak Hill merger using Dr. Lynk’s geographic market for context. It found that based upon the more likely correct market proposed by Dr. Lynk, various market factors indicated that the merger would not produce anticompetitive effects. Based upon these findings, the district court held that the FTC had failed to demonstrate a likelihood of success on the merits of its Clayton Act claim. It also determined that the balance of the equities weighed in favor of allowing the Hospitals’ merger to proceed. Accordingly, the district court denied the FTC’s request for a preliminary injunction. This appeal followed. II. DISCUSSION A. FTC Authority Over Nonprofit Hospitals As an initial matter, the Hospitals urge us to address the threshold issue of the FTC’s authority to challenge the Hospitals’ merger. The Hospitals argue the FTC has no authority to request a preliminary injunction against the Hospitals because the FTC lacks jurisdiction under Section 7 of the Clayton Act to challenge asset acquisitions by nonprofit hospitals. Section 7 of the Clayton Act provides, in relevant part: [N]o person subject to the jurisdiction of the Federal Trade Commission shall acquire the whole or any part of the assets of another person ... where in any line of commerce ... in any section of the country, the effect of such acquisition may be substantially to lessen competition.... 15 U.S.C. § 18. The Hospitals argue that to determine those persons “subject to the jurisdiction of the Federal Trade Commission,” we must turn to the FTC Act, 15 U.S.C. §§ 41^46, 47-58, the original charter of the FTC. In Community Blood Bank of the Kansas City Area, Inc. v. Federal Trade Comm’n, 405 F.2d 1011, 1022 (8th Cir.1969), we interpreted the jurisdictional provision of the FTC Act and held that it expressly excludes from the FTC’s jurisdiction “nonprofit corporations without shares of capital, which are organized for and actually engaged in business for only charitable purposes, and do not derive any ‘profit’.” Thus, the Hospitals argue, Freeman and Oak Hill are not “subject to the jurisdiction of the Federal Trade Commission” under the Clayton Act and the FTC has no authority to challenge their consolidation. In response, the FTC relies on Federal Trade Comm’n v. University Health, Inc., 938 F.2d 1206 (11th Cir.1991) and United States v. Rockford Memorial Corp., 898 F.2d 1278 (7th Cir.), cert. denied, 498 U.S. 920, 111 S.Ct. 295, 112 L.Ed.2d 249 (1990), two cases which address the FTC’s jurisdiction over asset acquisitions by nonprofit hospitals. In University Health, the Eleventh Circuit rejected the argument that the FTC’s authority to challenge a potential violation of Section 7 of the Clayton Act was defined by the FTC Act. Instead, that court held that the FTC’s jurisdiction to enforce the Clayton Act is determined by Section 11 of that same Act. Section 11 provides no exemption for nonprofit hospitals. University Health, 938 F.2d at 1214-15. The Seventh Circuit echoed that sentiment in Rockford Memorial. We do not believe Community Blood Bank forecloses this issue. Community Blood Bank addressed only the FTC’s jurisdiction to challenge certain entities’ potential violations of the FTC Act; it did not speak to the relationship between the FTC Act and the [267] Clayton Act. Further, we are persuaded by the reasoning of the Eleventh and Seventh Circuits regarding the reach of the FTC’s authority under Sections 7 and 11 of the Clayton Act, and find that the FTC may challenge the merger of Freeman and Oak Hill Hospitals. Nevertheless, we hold that the FTC is not entitled to a preliminary injunction in this case. B. Standard of Review We review the denial of preliminary injunctive relief under a deferential standard. As we explained in Federal Trade Comm’n v. National Tea Co., 603 F.2d 694, 696 (8th Cir.1979), “[t]he denial of a preliminary injunction may be reversed only if the trial court abused its discretion or based its decision on an erroneous legal premise.” The question in this case, therefore, is not whether this court would have granted the injunction in the first instance but whether, after a thorough review of the record, the proof unmistakably establishes clear error or an abuse of discretion. See Love v. Atchison, 185 F. 321, 331 (8th Cir.), cert. denied, 220 U.S. 618, 31 S.Ct. 721, 55 L.Ed. 612 (1911). C. The Merits We now consider whether the district court applied an incorrect legal standard or abused its discretion in denying the FTC’s request for a preliminary injunction. Section 13(b) of the Federal Trade Commission Act, under which the FTC seeks relief, provides that a preliminary injunction may be granted “[u]pon a proper showing that, weighing the equities and considering the Commission’s likelihood of ultimate success, such action would be in the public interest.” 15 U.S.C. § 53(b). Thus, the district court was charged with making two separate inquiries, one examining the likelihood that the FTC would ultimately succeed on the merits of its antitrust action and one addressing the competing equities advanced by the parties. We address each issue in turn. 1. Likelihood of Success on the Merits In order to demonstrate a likelihood of ultimate success for purposes of obtaining a preliminary injunction under Section 13(b), the FTC must raise “ ‘questions going to the merits so serious, substantial, difficult and doubtful as to make them fair ground for thorough investigation, study, deliberation and determination by the FTC in the first instance and ultimately by the Court of Appeals.’” National Tea, 603 F.2d at 698, quoting Federal Trade Comm’n v. Beatrice Foods Co., 587 F.2d 1225, 1229 (D.C.Cir.1978). In National Tea, we rejected the Commission’s argument that it need only show a “fair or tenable chance of ultimate success on the merits” in order to qualify for injunctive relief. 603 F.2d at 698. Such a standard runs contrary to congressional intent and reduces the judicial function to a mere “rubber stamp” of the FTC’s decisions. Id. Because Congress expected courts to use independent judgment in reviewing preliminary injunction applications under Section 13(b), we have adopted a more stringent standard. Id. The FTC observes that the district court did not refer to the National Tea standard in concluding the FTC had not shown a likelihood of success on the merits of its antitrust action.11 We find, however, that although the district court did not expressly articulate that standard, it clearly expected no more from the FTC than National Tea requires. [268] The district court did not, as the FTC argues, expect the FTC to “prove” the exact parameters of the relevant market or predict the precise competitive impact of the merger. Instead, the district court examined the record for evidence raising “serious, substantial, difficult and doubtful” questions about the competitive implications of the merger. Because we find that the district court did not base its determination of likely success on the merits on an erroneous legal premise, we proceed to evaluate whether the district court abused its discretion in holding that the FTC had not presented sufficiently reliable evidence of likely success to warrant a preliminary injunction. The FTC’s task in proving a Section 7 case, and thus in showing likely success in a preliminary injunction proceeding, is twofold. The FTC must present evidence of (1) a relevant market within which (2) the effect of the acquisition in question may be to substantially lessen competition. 15 U.S.C. § 18; see also Brown Shoe Co. v. United States, 370 U.S. 294, 324, 82 S.Ct. 1502, 1523, 8 L.Ed.2d 510 (1962). The FTC argues that the district court erred on both of these issues. a. The Relevant Market The determination of the relevant market is a “necessary predicate” to a finding of a Clayton Act violation. United States v. E.I. du Pont de Nemours & Co., 353 U.S. 586, 593, 77 S.Ct. 872, 877, 1 L.Ed.2d 1057 (1957). Without a well-defined relevant market, an examination of a transaction’s competitive effects is without context or meaning. It follows, then, that identification of a proper market is a “necessary predicate” to the FTC’s task of raising “questions going to the merits so serious, substantial, difficult and doubtful” as to warrant a preliminary injunction under Section 13(b).12 A relevant market consists of two separate components: a product market and a geographic market. See Morgenstern v. Wilson, 29 F.3d 1291, 1296 (8th Cir.1994), cert. denied, — U.S. —, 115 S.Ct. 1100, 130 L.Ed.2d 1068 (1995). The parties agree that the product market in this case consists of “acute care inpatient hospital services.” The parties possess widely divergent views, however, of the geographic market within which the merging hospitals compete. As noted earlier, the FTC has proposed a geographic market which includes Joplin, Missouri, and areas spanning twenty-seven miles in all directions from Joplin. The Hospitals, on the other hand, have proposed a thirteen-county market including communities over fifty miles from Joplin. We need not fully resolve this conflict at this juncture, for in a Section 13(b) preliminary injunction action it is the FTC’s burden to identify a geographic market within which a challenged transaction raises “serious, substantial, difficult and doubtful” questions of antitrust concern. See, e.g., Morgenstern, 29 F.3d at 1296. A geographic market is that geographic area “to which consumers can practically turn for alternative sources of the product and in which the antitrust defendants face competition.” Id. In order to meet its burden, the FTC is required to present evidence addressing the critical question of where consumers of acute care inpatient hospital services could practicably turn for alternative sources of the product should the Hospitals’ merger be consummated and Joplin hospital prices become anti-competitive. The FTC has failed to produce such evidence. Two recent cases show the importance of adequately discerning practicable alternatives for consumers of the relevant product. In Morgenstem, a cardiac surgeon brought suit under the Sherman Antitrust Act alleging that members of the Nebraska Heart Institute monopolized the market for cardiac [269] surgery in Lincoln, Nebraska. Morgenstern proposed a geographic market including Lincoln and twenty-six counties located up to 200 miles from Lincoln but excluding Omaha, Nebraska. Id. That market was based largely on expert testimony detailing the residences of cardiac surgery patients in the Lincoln and Omaha heart surgery programs and demonstrating the counties that supplied patients to each program. Id. In rejecting Morgenstern’s proposed market in favor of a market including Omaha, we noted that the record addressed only where Nebraska residents actually went, as opposed to where they could practicably go, for cardiac surgery services. Id. We held that Morgenstern’s failure to present evidence on practicable alternatives for cardiac surgery patients was fatal to his ease. The plaintiffs’ case in Bathke v. Casey’s General Stores, Inc., 64 F.3d 840 (8th Cir.1995), suffered from a similar defect. In Bathke, gasoline retailers located in sixty-seven different small Iowa towns brought suit against a multi-state retailer of gasoline for alleged predatory pricing activities. The plaintiffs sought to establish each small town as a separate geographic market of competition for gasoline. Id. at 344. The plaintiffs compiled testimony from complaining businesses who stated they considered in-town stores their competitors and testimony from consumers who expressed a preference for buying gas in the towns in which they lived. Id. at 345. The record contained no evidence, however, about where consumers could practicably travel in order to avoid purchasing gas at a Casey’s store. Lacking evidence on that issue, we affirmed the district court’s entry of summary judgment against the plaintiffs. In this case, the district court held that, like the plaintiffs in Morgenstern and Bath-ke, the FTC failed to produce sufficient evidence on the crucial aspect of the geographic market: where consumers of acute care inpatient hospital services could practicably turn for alternative sources of that product. After reviewing the extensive record in this case, we cannot say the district court’s holding on this issue constitutes an abuse of discretion. The evidence produced by the FTC to support its geographic market consisted primarily of Dr. Leffler’s Elzinga-Ho-garty analysis of zip code data illustrating current patient flow into and out of the three Joplin hospitals and hospitals located within a twenty-seven-mile radius of Joplin. This analysis gives a static, rather than a dynamic, picture of the acute care market in Joplin and the surrounding areas. Even if we did not credit the concerns Dr. Lynk raised about the completeness and correctness of Dr. Leffler’s analysis,13 we would still be forced to conclude that the Elzinga-Hogarty analysis presented in this case did not, by itself, address the decisive question of where consumers could practicably go for alternative sources of acute care inpatient hospital services. Simply put, Dr. Leffler’s theory provided no insight into the future effects of the allegedly anti-competitive merger of the Hospitals. The FTC argues, however, that the evidence it presented to establish the appropriate geographic market consisted of far more than Dr. Leffler’s testimony. The FTC submitted the statements of various market participants to address the question of where consumers would likely go if the merger was consummated. The majority of these participants testified that even if Joplin health care prices increased after the merger due to collusion between the remaining Joplin hospitals, few patients currently traveling to Joplin for care would travel instead to hospitals located outside the FTC’s proposed geographic market. The FTC asserts that the district court abused its discretion in discounting the importance of this evidence to defining the relevant market. [270] We agree that testimony of market participants is relevant to a determination of a proper geographic market, and recognized the importance of exploring their perceptions in Morgenstem. See 29 F.3d at 1297 (detailing testimony of health care providers and hospital administrators regarding the nature of the cardiac surgery market). Our opinion in Bathke, however, clearly demonstrates that the views of market participants are not always sufficient to establish a relevant market, especially when their testimony fails to specifically address the practicable choices available to consumers. In evaluating market participant testimony in Bathke, we stated: Evidence of the competitor’s perspective is not sufficient ... because a geographic market is determined by inquiring into the “commercial realities” faced by consumers. Likewise, evidence of consumers’ actual habits is not enough to establish [a] relevant geographic market under orn-eases. ... [E]ven if we fully credit the testimony from a number of the plaintiffs ... there is still an absence of evidence on a critical question: where those gasoline consumers could practicably turn for alternatives. 64 F.2d at 345-46 (internal citations omitted). The testimony from market participants in this case, like the testimony in Bathke, spoke mainly to current competitor perceptions and current consumer habits and not to the crucial question of where consumers could practicably go to seek alternative acute care inpatient hospital services should Freeman Hospital and Oak Hill Hospital merge.14 The testimony that most closely addressed the question of where patients could practicably turn for alternative sources of acute care inpatient hospital services was the testimony of health care administrators from hospitals located both at the margins of and outside of the FTC’s proposed geographic market. Several of these administrators testified that patients in their area who travel to Joplin for health care go mainly because the Joplin hospitals provide services which are not otherwise available in their local hospitals. The FTC claims that this testimony establishes differences in the quality and range of services available in Joplin from those available at many of the hospitals in Dr. Lynk’s proposed geographic market. These differences, the FTC argues, undercut Dr. Lynk’s assertions that the hospitals in his thirteen-county market are competitors of the Joplin hospitals merely because of their geographic proximity to Joplin and the communities which supply patients to Joplin hospitals. Taking this analysis further, the FTC borrows a page from the Seventh Circuit’s opinion in Rockford Memorial, arguing that it would be “ridiculous” to assume that Joplin residents, in particular, “will be searching out small, obscure hospitals in remote rural areas” if prices charged by the postmerger Joplin hospitals rise above competitive levels. Rockford Memorial, 898 F.2d at 1285. For two reasons, we find no clear error by the district court in refusing to fully credit this testimony and hold that it adequately established acute care consumers’ practicable alternatives. First, we have searched the record and have been unable to find significant economic or statistical data which supports the assertions of the health care providers and the FTC regarding quality and range of services. For example, each patient discharged from a hospital generally receives a Diagnosis Related Group Code (DRG Code) which designates the type of treatment [271] the patient received. The FTC could have compiled and analyzed the DRG records of hospitals potentially in the geographic market to determine whether significant differences exist in the DRGs available in the Joplin hospitals and those offered by other hospitals.15 The FTC might also have introduced evidence illustrating the sophistication of procedures performed on patients traveling from zip codes located outside the FTC’s proposed market, or evidence breaking down patient discharges by cost of care, which may have served as an indicator of the severity of the illnesses of patients migrating to Joplin for care. We hesitate to require particular evidence, especially in litigation involving complex industries such as health care, in light of the broad range of evidence that may be of value in determining a geographic market. Nevertheless, such data would have substantially assisted the FTC in demonstrating where patients could practicably go for acute care inpatient services. Lacking this evidence, it was well within the discretion of the district court to expect more formal analysis before accepting the premise that the hospitals located outside the FTC’s proposed geographic market did not compete with Joplin hospitals because of their limited range and quality of services.16 Second, placing too much emphasis on the allegedly superior range of services available at Joplin hospitals could ultimately lead to a blurring of the product market. If, as the FTC urges, patients from areas extending beyond the FTC’s market come to the three Joplin hospitals to receive specialized, sophisticated services, those services may begin to define a separate product market. Focusing on more specialized care may mean that hospitals in larger cities in the four-state area, such as Springfield, Missouri; Kansas City, Missouri; or Tulsa, Oklahoma, become viable competitive alternatives to hospitals in Joplin. A separate product market may begin to develop, for example, for certain types of tertiary care which, while serious, need not always be treated immediately and may often be scheduled in advance. We perceive problems with discounting as potential competitors those hospitals that compete for a significant number of the merging hospitals’ admission base but do not, because of limited resources, compete for the smaller percentage of patients seeking advanced tertiary care. Nevertheless, it is not necessary for us to decide at this time what percentage of patients the outlying hospitals must be capable of accommodating before they may be included within the geographic market. We hold only that it was not clearly erroneous for the district court to hold that the testimonial evidence of the market participants did not sufficiently demonstrate where consumers could practicably turn for alternative sources of acute care inpatient hospital services. In summary, the FTC’s expert testimony addressed only the question of where patients currently go, rather than where they could practicably go, for acute care inpatient services. The bulk of the testimony from market participants suffered from a similar defect, since it spoke only to present competitor perceptions of the geographic market and present customer preferences in seeking health care. To the extent the market participants addressed the question of practicá- [272] ble alternatives for consumers, we do not read the district court’s opinion to have improperly discounted that evidence. We therefore hold that the district court did not abuse its discretion when it determined that the FTC had failed to meet its burden of establishing the relevant geographic market. b. Anticompetitive Effects The FTC advances several arguments regarding the alleged anticompetitive effects of the Hospitals’ merger on the market for acute care inpatient hospital services. Specifically, the FTC argues that it demonstrated that, within any of the three alternative markets it identified, levels of concentration, high barriers to entry, and probable collusion between the Joplin hospitals raise sufficient antitrust concerns to warrant a preliminary injunction. We have already stated, however, that identification of a relevant market is a “necessary predicate” to a successful challenge under the Clayton Act and thus to establishing a likelihood of ultimate success for preliminary injunction purposes. Because our resolution of this issue is disposi-tive, we need not reach the other issues presented on appeal. 2. Balance of the Equities We now turn to the second inquiry mandated by Section 13(b) of the FTC Act: balancing the equities. Section 13(b) urges a district court to consider public and private equities to determine whether the issuance of a preliminary injunction pending the outcome of an administrative proceeding would be in the public interest. National Tea, 603 F.2d at 697. The district court determined that the primary public interest in this case was “the interest in maintaining competitive hospital prices.” Mem. op. at 31. The district court also recognized that public equities favored strong enforcement of antitrust laws, and that enforcement is made difficult when the FTC must undo a merger after it has been consummated. Against these strong public equities the district court balanced the hardships faced by the Hospitals. The court found that Oak Hill’s financial status was such that it may no longer be in business to complete the merger by the time the FTC concludes its administrative hearing on the merits of the transaction. Ultimately, the court held that enjoining the consolidation of the Hospitals would not be in the public interest. The FTC argues that the district court erred as a matter of law by giving undue weight to the parties’ private interests in consummating the merger. Our eases make clear, however, that a district court may properly consider both public and private equities in undertaking the weighing mandated by Section 13(b). See National Tea, 603 F.2d at 697. It is true that the public equities are to be given greater weight in the balance. Id. In this case, however, once the district court found the FTC had failed to show a substantial threat to competitive hospital prices—a conclusion we hold was not clearly erroneous—the public equities advanced by the FTC were substantially weakened. Moreover, where, as here, the party moving for preliminary injunctive relief has failed to adequately show a likelihood of success on the merits, it faces a particularly heavy burden of demonstrating that the equities nevertheless militate in favor of granting the requested relief. Under the circumstances, we see nothing in the district court’s opinion that would indicate it did not give proper weight to the public equities involved. The FTC further alleges that the district court’s opinion may have been driven by its personal bias against the FTC and the work of administrative agencies in general.17 We are disturbed any time a district court expresses personal opinions regarding the parties or their motives during a formal judicial proceeding, and we agree that the district court was injudicious in its remarks about the FTC. Our affirmance of the district court’s ultimate decision is in no way intended to express our approval of such a practice. [273] Nevertheless, there is nothing in the district court’s remarks which suggests “such a high degree of favoritism or antagonism as to make fair judgment impossible.” Liteky v. United States, — U.S. —, —, 114 S.Ct. 1147, 1157, 127 L.Ed.2d 474 (1994). While we do not condone the district court’s comments, we have reviewed the record and the district court’s opinion and find no clear error or abuse of discretion. III. CONCLUSION The district court did not err in holding that the FTC had failed to carry its burden of showing a likelihood of success on the merits of its antitrust action or in holding that the balance of the equities favors denying a preliminary injunction. We therefore affirm the district court’s order denying the FTC an injunction, dissolve our order of March 15, 1995, and deny as moot all other pending motions. Footnotes [1] Allopathy and osteopathy are distinguished by their medical philosophies. Allopathy is a system of therapy in which "diseases are treated by producing a condition incompatible with or antagonistic to the condition to be cured or alleviated.” Dorland’s Illustrated Medical Dictionary 48 (28th ed. 1994). Osteopathy, on the other hand, is "based on the theory that the body is capable of making its own remedies against disease and other toxic conditions when it is in normal structural relationship and has favorable environmental conditions and adequate nutrition." Id. at 1202. Historically, prescription drugs and surgery were not accepted osteopathic treatment methods; osteopathy instead focused on manipulation of the musculoskeletal system to correct faulty structure. See Eatough v. Board of Medical Examiners, 191 N.J.Super. 166, 465 A.2d 934, 936 (1983). Today, osteopathy utilizes "generally accepted physical, medicinal, and surgical methods of diagnosis and therapy” in addition to traditional osteopathic treatments. Dorland’s Illustrated Medical Dictionary at 1202. An osteopathic physician carries the title of D.O. (Doctor of Osteopathy) rather than M.D. See Eatough, 465 A.2d at 936. [2] The Hart-Scott-Rodino Act, 15 U.S.C. § 18a, provides, in pertinent part: Except as exempted pursuant to subsection (c) of this section, no person shall acquire, directly or indirectly, any voting securities or assets of any other person, unless both persons ... file notification pursuant to rules under subsection (d)(1) of this section [3] Under 15 U.S.C. § 18a(b)(l), a waiting period of thirty days, triggered upon receipt of the required notification by the FTC, must expire before an acquisition may be completed. [4] Oak Hill's Request for Proposals listed several criteria by which bids for the hospital would be evaluated, including: (1) bidder's financial strength; (2) bidder's strategic position in the Missouri health care market; (3) appointment of the Oak Hill professional staff, with full privileges, to the new hospital’s medical staff; (4) continuation of the osteopathic educational programs at Oak Hill; (5) continuation of significant acute care services at the Oak Hill campus; and (6) assumption of Oak Hill's existing bond indebtedness. The trustees testified that according to this criteria Freeman's proposal was superior to the other bids. [5] Section 13(b), 15 U.S.C. § 53(b), provides: Whenever the Commission has reason to believe— (1) that any person, partnership, or corporation is violating, or is about to violate, any provision of law enforced by the Federal Trade Commission, and (2) that the enjoining thereof ... would be in the interest of the public— the Commission ... may bring suit in a district court of the United States to enjoin any such act or practice. Upon a proper showing that, weighing the equities and considering the Commission’s likelihood of ultimate success, such action would be in the public interest ... a temporary restraining order or a preliminary injunction may be granted.... [6] Dr. Leffler prepared an analysis for three alternative geographic markets, but at the preliminary hearing expressed his preference for the twenty-seven-mile radius market. [7] Elzinga and Hogarty label this element of their test the "LOFI," or "little out from the inside" measurement. In the hospital context, a LOFI statistic signifies the percentage of patients in an area's hospitals who reside in the area and, concurrently, the number of patients immigrating to those hospitals from outside the area. See Rockford Memorial Corp., 717 F.Supp. at 1266; see generally Kenneth G. Elzinga & Thomas F. Ho-garty, The Problem of Geographic Market Delineation in Antimerger Suits, 18 Antitrust Bulletin 45 (1973) {Market Delineation I); Kenneth G. Elzin-ga & Thomas F. Hogarty, The Problem of Geographic Market Delineation Revisited: The Case of Coal, 23 Antitrust Bulletin 1 (1978) {Market Delineation Revisited). [8] Each expert spoke to the proper inclusion criterion which should serve as a cutoff point for designating an accurate service area. The inclusion percentage defines the degree of inclusiveness of the service area, or the degree to which the area accounts for a hospital's business. The lower the inclusion percentage becomes, the greater the patient business which goes unaccounted for in the analysis. In their original article, Elzinga and Hogarty expressed satisfaction with a seventy-five percent benchmark. Market Delineation I, supra note 7, at 75. Subsequently, however, they endorsed a ninety percent inclusion criterion as the more appropriate cutoff. Market Delineation Revisited, supra note 7, at 2. Dr. Leffler testified that he chose eighty percent as a benchmark because of his assessment, based on the testimony of market participants, that twenty percent of the patients traveling to Joplin come for specialized services and therefore were not representative of the Joplin hospitals' main patient base. [9] This analysis is known as the "LIFO” or "little in from the outside” measurement. LIFO signifies the percentage of hospital patients from a particular area who remain in that area for services, and thus determines whether patients make frequent use of hospitals outside the ser [265] vice area. If patients utilize hospitals outside the area, those hospitals can act as a check on the exercise of market power by the hospitals within the service area. See Rockford Memorial Corp., 717 F.Supp. at 1266. [10] Dr. Lynk is an economist with the Lexecon Inc. in Chicago. [11] The district court borrowed language from several different cases in setting forth its standard for determining the FTC’s likelihood of success. The court stated: Pursuant to Section 7 of the Clayton Act, the FTC must demonstrate a likelihood that the effect of the proposed merger "may be substantially to lessen competition, or to tend to create a monopoly” within the relevant product and geographic markets. 15 U.S.C. § 18. To establish a prima facie case, the FTC must show that this acquisition will produce "a firm controlling an undue percentage share of the relevant market, and [would] result[ ] in a significant increase in the concentration of firms in that market." United States v. The Philadelphia National Bank, 374 U.S. 321, 363, 83 S.Ct. 1715, 1741, 10 L.Ed.2d 915 (1963). At bottom, the FTC's case must "satisfy some minimum threshold of persuasiveness and be better than the defendant's case.” United States v. Rockford Memorial Corp., 898 F.2d 1278, 1286 (7th Cir.), cert. denied, 498 U.S. 920, 111 S.Ct. 295, 112 L.Ed.2d 249 (1990). Mem. op. at 27-28. [12] The FTC argues that the National Tea standard requires only that it present "serious, substantial, difficult and doubtful” questions about the relevant market at the preliminary injunction stage. We find this suggestion incongruous. The “questions going to the merits” to which National Tea refers are questions regarding the ultimate merits in a Clayton Act case—the lessening of competition. Without a well-defined relevant market, a particular transaction’s effect on competition cannot be evaluated. In other words, without a properly-defined market no National Tea "question” exists to be answered. It is therefore essential that the FTC identify a credible relevant market before a preliminary injunction may issue. [13] The FTC criticizes the district court’s decision to afford greater weight to Dr. Lynk's analysis and ultimately adopt his proffered market. We find this criticism meritless. It is axiomatic that a district court has the discretion to evaluate the credibility of expert witnesses and accept the testimony it finds most plausible. See, e.g., Fox v. Dannenberg, 906 F.2d 1253, 1256 (8th Cir.1990). Moreover, the district court "adoptfed]” Dr. Lynk’s geographic market only for context in order to address the competitive effects of the merger. See Mem. op. at 14. Such an approach may have been unnecessary, see discussion infra part 11(C)(1)(b), but it was not an abuse of discretion. [14] We realize that in the case of health care, the term "consumers” often means not individual patients but large purchasers of health care such as managed care coalitions or third-party payors. Where, as here, these larger purchasers testify that only a significant price increase of approximately fifteen to twenty percent would cause them to steer their members to outlying hospitals, that testimony should be given some weight, since the terms of managed care and third-party insurance contracts necessarily have some bearing on where patients can practicably turn for their health care. Similarly, when patients' physicians do not have and are not able to obtain privileges at certain hospitals, some patients may be limited in the hospitals to which they can practicably turn. Nevertheless, we do not read the district court’s opinion to have ignored the FTC’s evidence on these issues. The district court merely indicated that it required more than that evidence in order to accept the FTC's proffered geographic market. Such a decision is well within the purview of the district court, which possesses the sole discretion to decide when evidence raises "serious, substantial, difficult and doubtful” antitrust questions. [15] Even Dr. Leffler testified that, "ideally” he would like to examine a breakdown of services by DRG Code in order to separate patients who traveled to Joplin for tertiary care from those who went for primary and secondary care. Prelim.Hrg.Tr. at 88 (March 23, 1995). [16] The FTC's reliance on Rockford Memorial is misplaced. The Supreme Court has repeatedly emphasized that the definition of a geographic market is highly fact-driven and therefore different in each case. See, e.g., Brown Shoe, 370 U.S. at 336-37, 82 S.Ct. at 1529-30 ("Congress prescribed a pragmatic, factual approach to the definition of the relevant market....”) Moreover, a look at the district court’s opinion in Rockford Memorial reveals that the record in that case did contain statistical evidence which sought to illustrate the comparability of services at the Rockford hospitals proposing to merge and those in the outlying counties. In Rockford Memorial, evidence was presented demonstrating the DRGs available at the hospitals involved, and a breakdown of admissions by cost was undertaken in an effort to show the severity of illnesses of patients coming from outside Rockford to Rockford hospitals. Rockford Memorial, 717 F.Supp. at 1265. Such information lends significant credibility to broad assertions that the quality of care and range of services at one hospital is not comparable to that of another. [17] In addition to expressing its dismay during the temporary restraining order hearing over the FTC's interference with southwest Missouri's health care market, the district court concluded the preliminary injunction hearing by observing, "[t]he last vestige of denial of due process lies in these administrative agencies.” Prelim.Hrg.Tr. at 393 (March 24, 1995).