BEAM, Circuit Judge.
Tenet Healthcare and Poplar Bluff Physicians Group, Inc., doing business as Doctors’ Regional Medical Center (collectively, Tenet) appeal the district court’s order enjoining the merger of two hospitals in Poplar Bluff, Missouri. After a five-day hearing, the district court granted a motion for a preliminary injunction filed by the Federal Trade Commission (FTC) and the State of Missouri. The district court found a substantial likelihood that the merger would substantially lessen competition between acute care hospitals in Poplar Bluff, Missouri, in violation of section 7 of the Clayton Act, 15 U.S.C. § 18. We reverse.
I. BACKGROUND
Poplar Bluff is a city of 17,000 people in southeastern Missouri. It is located in Butler County, which has a population of 40,000. It is the largest city in several counties and has numerous major employers and manufacturing operations. Sike-ston, Missouri, and Cape Girardeau, Missouri, both towns with populations of over 40,000 are forty and sixty miles away from Poplar Bluff. The population in the area surrounding Poplar Bluff is concentrated in Scott and Stoddard Counties, which lie between Poplar Bluff and Cape Girardeau. Poplar Bluff is within a few hours’ drive of several large metropolitan centers including St. Louis, Missouri, Memphis, Tennessee, and Jonesboro, Arkansas.
Tenet Healthcare Corporation presently owns Lucy Lee Hospital in Poplar Bluff. Lucy Lee is a general acute care hospital that provides primary and secondary care services.
Lucy Lee has 201 licensed beds, 185 of which are staffed. It operates ten outpatient clinics in the surrounding counties. Its average daily census was 75 in 1994, 76 in 1995 and 104 in 1996. Doctors’ Regional Medical Center in Poplar Bluff is presently owned by a group of physicians. It is also a general acute care hospital providing primary and secondary care services. It has 230 licensed beds, of which 187 are staffed. Its average census in 1994 was 106, in 1995 was 99, in 1996 was 95 and in 1997 was 77. It also operates several rural health clinics in the area. Though profitable, both hospitals are un
derutilized and have had problems attracting specialists to the area.
Tenet recently entered into an agreement to purchase Doctors’ Regional for over forty million dollars. Tenet plans to operate Doctors’ Regional as a long-term care facility and to consolidate inpatient services at Lucy Lee. It plans to employ more specialists at the merged facility and to offer higher quality care in a comprehensive, integrated delivery system that would include some tertiary care.
Pursuant to the HarNScott-Rodino Act, 15 U.S.C. § 18a, the hospitals filed a premer-ger certification with the FTC. Shortly thereafter, the FTC filed a complaint alleging that the hospitals’ merger would lessen competition for primary and secondary inpatient hospitalization services in the area. The FTC sought to enjoin the merger. The district court held a five-day hearing on the motion for preliminary injunction. Both parties presented testimony by market participants and experts.
The evidence adduced at the hearing shows that Lucy Lee and Doctors’ Regional are the only two hospitals in Poplar Bluff, other than a Veteran’s Hospital. The combined service area of these hospitals covers eight counties and an approximate fifty-mile radius from Poplar Bluff.
There are also’several other hospitals in the surrounding area. Regional hospitals which
offer
the same or a greater range of services as provided by Lucy Lee and Doctors’ Regional are located in Sikeston (Missouri Delta Medical Center), Cape Girardeau (Southeast Missouri Medical Center and St. Francis Hospital), St. Louis (Barnes Jewish Hospital) and Jonesboro, Arkansas (St. Bernard’s Hospital). There is also another Tenet-owned facility in Jonesboro (Methodist Hospital) and a Tenet-owned regional hospital in Kennett, Missouri (Twin Rivers Medical Center). In addition, there are smaller rural hospitals located in the nearby towns of Dexter (Dexter Memorial Hospital), Ellington (Reynolds County Memorial Hospital), Do-niphan (Ripley County Memorial Hospital) and Clay County, Arkansas (Piggott Community Hospital). Each of the smaller hospitals have fewer than fifty beds and provide only primary care.
Lucy Lee’s and Doctors’ Regional’s patient bases are composed primarily of patients who are covered by Medicare and Medicaid and thus remain largely insensitive to price differentials. Most of the remaining patient admissions at Lucy Lee and Doctors’ Regional are covered by health insurance, under a plan administered by a managed care organization.
These organizations include health maintenance organizations (HMOs)
and pre
ferred provider organizations (PPOs).
Hospitals are willing to discount their stated rates to managed care payers in order to entice the managed care entity to send its enrollees to that hospital. Managed care organizations have had a presence in Poplar Bluff for approximately fifteen years. Most employers in the Poplar Bluff area either subscribe to or administer a PPO. Both Lucy Lee and Doctors’ Regional have entered into discount agreements with numerous managed care entities and employers.
The hospitals in Gape Girardeau, on the other hand, refused to negotiate with managed care plans until recently, when, at the insistence of area employers, Southeast Missouri Hospital entered into a discount arrangement with HealthLink, a managed care organization. Healthcare prices in Cape Girardeau have historically been significantly higher than prices in Poplar Bluff. However, there is also a perception of higher quality service at Cape Girar-deau hospitals. Since the entry of managed care in the Cape Girardeau market, there has been some reduction in prices. In fact, the HealthLink managed care contract per diem rate in Cape Girardeau is close to that offered by Poplar Bluff hospitals. Cape Girardeau hospitals now have outreach efforts, including advertising, in Poplar Bluff.
Market participants, specifically, employers, healthplans and network providers testified that they had negotiated substantial discounts and favorable per diem rates with either or both Lucy Lee and Doctors’ Regional as a result of “playing the two hospitals off each other.” These managed care organizations and employers testified that if the merged entity were to raise its prices by ten percent, the health plans would have no choice but to simply pay the increased price. They testified that they perceive it is essential for the plans to include a Poplar Bluff hospital in their benefit packages because their en-rollees would not travel to other towns for primary and secondary inpatient treatment. ' They stated that their employees and subscribers find it convenient to use a Poplar Bluff hospital; are loyal to their physicians in Poplar Bluff and would not be amenable to a health benefit plan that did not include a Poplar Bluff hospital.
The evidence shows that patient choice of hospitals is determined by many variables, including patient/physician loyalty, perceptions of quality, geographic proximity and, most importantly or determinatively, access to hospitals through an insurance plan. Managed care organizations have been able to influence or change patient behavior with financial incentives in other healthcare markets. This practice is known as “steering.” Representatives of Poplar Bluff managed care entities testified, however, that they did not believe such efforts would be- successful in the Poplar Bluff market. They testified it would be unlikely that they could steer their subscribers to another hospital, or could exclude the merged Poplar Bluff entity in the event of a price increase, in spite of the fact that such tactics had been successful in other markets. They did not regard.the Cape Girardeau hospitals as an alternative to Poplar Bluff hospitals because the Cape Girardeau hospitals were more costly. Witnesses conceded, however, that employees had been successfully “steered” to other area hospitals in the past. Several employers testified that they could successfully steer their employees to Missouri Delta Hospital in Sikeston, Missouri. The representative of one large employer testified that the large employers could prevent price increases through negotiation based on their market power and that the merged entity would provide better quality healthcare.
Lucy Lee and Doctors’ Regional obtain ninety percent of their patients from zip
codes within a fifty-mile radius of Poplar Bluff. In eleven of the top twelve zip codes, however, significant patient admissions — ranging from 22% to 70% — were to hospitals other than those in Poplar Bluff. There is no dispute that Poplar Bluff residents travel to St. Louis, Memphis, and Jonesboro for tertiary care. The evidence also shows, however, that significant numbers of patients in the Poplar Bluff service area travel to other towns for primary and secondary treatment that is also available in Poplar Bluff.
The evidence shows that the healthcare industry is rapidly changing. The emergence and growth of managed care — a system in which a third party monitors healthcare resources and expenditures— has had a large impact on healthcare. This monitoring has caused a corresponding decline in the number and length of inpatient admissions. Many procedures that formerly required a hospital stay are now performed on an outpatient basis. Another trend has been growth of outreach efforts such as rural clinics to extend the service area of a hospital. Patient loyalty to a certain doctor has diminished as patients’ out-of-pocket expenditures have increased.
The FTC presented the expert testimony of an economist, Dr. Lawrence Wu. He concluded that a merger between the two Poplar Bluff hospitals would be anticom-petitive. Dr. Wu testified that he had performed a statistical analysis on hospital admissions data and had also relied on testimony of market participants such as employers, health plans and healthcare network providers to support his conclusion. Dr. Wu’s analysis was based- on the “Elzinga-Hogarty” test, which analyzes the number of patients coming into and leaving a proposed market as a means of testing whether the proposed market constitutes a geographic market for antitrust purposes.
Dr. Wu excluded certain patients from his analysis, however, including those who traveled to St. Louis for any reason, and those admitted to hospitals by specialists. He reasoned that those patients most likely sought services that were not available in Poplar Bluff and thus the loss of those patients to other hospitals could not be considered in his analysis.
Dr. Wu further testified that he had not performed any specific “critical loss” analysis, although he conceded that such an analysis was an important element in evaluating a proposed market. A “critical loss” analysis would identify the threshold number of patients who, by seeking care at other hospitals, could defeat a price increase by making it unprofitable. The purchasing behavior of these patients or “marginal consumers” would discipline or constrain any potential price increase by a merged entity. Dr. Wu concluded that there would be no critical loss in reliance on the statements of market participants that they would not switch hospitals in the event of a price increase.
Tenet presented the testimony of another economist, Dr. Barry Harris. He con-, eluded that the proposed merger would not harm competition. Dr. Harris had performed a “critical loss” analysis and had concluded that, if the merged hospital were to raise prices, enough patients would leave the merged hospital and seek care at an alternative hospital to render the price increase unprofitable. Dr. Harris presented evidence, based on the hospitals’ financial data, that the loss of only a few commercially insured patients to other hospitals would make a five percent price increase unprofitable. Dr. Harris also testified that it was likely that enough par tients would, in fact, switch to defeat such a price increase. To reach this conclusion, Dr. Harris relied on a zip-code-by-zip-code “contestability analysis.” This showed that many commercially insured patients already sought treatment at hospitals outside Poplar Bluff for services that were available in Poplar Bluff hospitals. Dr. Harris limited his analysis to treatment for
those Diagnostic Related Groups (DRGs) that were available in either of the Poplar Bluff hospitals.
Dr. Harris testified that in many of the zip codes comprising the FTC’s proposed market, more than twenty percent of patients were using hospitals other than Lucy Lee and Doctors’ Regional for services they could have obtained at the Poplar Bluff hospitals.
In rebuttal, the FTC presented the testimony of Alan Bruce Steinwald. Based on an analysis of DRG data, he opined that patients seeking care outside Poplar Bluff were seeking a more sophisticated level of service than that available in Poplar Bluff. He relied on average-length-of-stay and cost-per-admission data to support this conclusion. Significantly, Steinwald’s conclusion was based on an average that could have been skewed by a high or low number at either end. Steinwald conceded that some patients went to other hospitals for services that were available in Poplar Bluff, but was unable to quantify those patients.
Based on the evidence, the district court enjoined the merger. It found that the statistical evidence presented in the case failed to establish the relevant geographic market. Relying on anecdotal evidence, “confirmed by common sense,” the district court concluded that the geographic market proposed by the government was appropriate and that “[a]t some point, a hospital ceases to become a practical alternative for general acute care because of distance.” The district court thus concluded that the merger would be anticompetitive.
On appeal, Tenet argues that the district court erred in its analysis by improperly shifting the burden of proof to Tenet and by failing to address the critical market definition question — where could consumers of hospital services practicably turn in the event of a price increase?
II. DISCUSSION
A preliminary injunction may be granted in an antitrust case if the FTC shows 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). In order to demonstrate such a likelihood of ultimate success, “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.”
Freeman Hosp.,
69 F.3d at 267 (quotation omitted). A showing of a fair or tenable chance of success on the merits will not suffice for injunctive relief.
See id.
Section 7 deals in probabilities not ephemeral possibilities.
See United States v. Marine Bancorporation Inc.,
418 U.S. 602, 622-23, 94 S.Ct. 2856, 41 L.Ed.2d 978 (1974).
The determination of a relevant market is a necessary predicate to the finding of an antitrust violation.
See Freeman Hosp.,
69 F.3d at 268. Without a well-defined relevant market, a merger’s effect on competition cannot be evaluated.
See id.
at 268 n. 12. It is thus essential that the FTC identify a credible relevant market before a preliminary injunction may properly issue.
See id.
A relevant market consists of two components: a product market and a geographic market.
See id.
at 268. The parties agree that the relevant product market at issue in this
case is the delivery of primary and secondary inpatient hospital care services. They disagree, however, on the relevant geographic market.
A geographic market is the area in which consumers can practically turn for alternative sources of the product and in which the antitrust defendants face competition.
See id.
Market share must be established in a well-defined market.
See Flegel v. Christian Hosp.,
4 F.3d 682, 689 (8th Cir.1993). A properly defined geographic market includes potential suppliers who can readily offer consumers a suitable alternative to the defendant’s services.
See Bathke v. Casey’s General Stores, Inc.,
64 F.3d 340, 346 (8th Cir.1995). Determination of the relevant geographic market is highly fact sensitive.
See Freeman Hosp.,
69 F.3d at 271 n. 16. The proper market definition can be determined only after a factual inquiry into the commercial realities faced by consumers.
See Flegel,
4 F.3d at 690. A monopolization claim often succeeds or fails strictly on the definition of the product or geographic market.
See Morgenstern v. Wilson,
29 F.3d 1291, 1296 (8th Cir.1994).
The government has the burden of proving the relevant geographic market. See
United States v. Connecticut Nat'l Bank,
418 U.S. 656, 669, 94 S.Ct. 2788, 41 L.Ed.2d 1016 (1974);
Morgenstern,
29 F.3d at 1296 (stating it is the FTC’s burden to establish that a specified area constitutes a relevant geographic market). To meet this burden, the FTC must present evidence on the critical question of where consumers of hospital services could practicably turn for alternative services should the merger be consummated and prices become anticompetitive.
See Freeman Hosp.,
69 F.3d at 269. This evidence must address where consumers could practicably go, not on where they actually go.
See id.
at 268;
Bathke,
64 F.3d at 346 (articulating the test as the distance “customers will travel in order to avoid doing business at [the entity that has raised prices]” rather than the distance customers would travel absent a price increase); and
Morgenstern,
29 F.3d at 1297.
The FTC proposes a relevant geographic market that essentially matches its service area: a fifty-mile radius from downtown Poplar Bluff. It is from this service area that the two hospitals obtain ninety percent of their patients. A service area, however, is not necessarily a merging firm’s geographic market for purposes of antitrust analysis.
See Bathke,
64 F.3d at 346 (noting that “trade area” and “relevant market” are precisely reverse concepts). The FTC’s proposed geographic market includes four other hospitals: a Tenet-owned regional hospital in Kennett, Missouri, and three rural hospitals. The FTC contends that its evidence shows that the merged entity will have a post-merger market share of eighty-four percent of this geographic market.
Tenet, on the other hand, proposes a relevant geographic market that encompasses a sixty-five mile radius from downtown Poplar Bluff in addition to Barnes Hospital in St. Louis. The proposed area includes sixteen hospitals in addition to those in the FTC’s proposed geographic market. The area includes a regional hospital in Sikeston, Missouri, and two large hospitals in Cape Girardeau, Missouri, which is sixty miles northeast of Poplar Bluff. Notably, population is concentrated in the counties east of Poplar Bluff, meaning that many people live or work closer to Cape Girardeau and Sikeston than to Poplar Bluff.
Tenet argues that the evidence relating to “marginal consumers” shows that the merged entity would be unable to raise prices without causing the “critical loss” of enough patients to make the increase unprofitable. This “critical loss” approach is in fact employed by the FTC in its own Horizontal Merger Guidelines which are used by the FTC to ascertain a relevant geographic market in exercising its prose-cutorial discretion to challenge a merger.
See
Department of Justice, Federal Trade Commission, Antitrust Division, 1992 Horizontal Merger Guidelines, 57 Fed.Reg. 41552 § 1.21.
The question before us is whether the FTC provided sufficient evidence that the proposed merger will result in the merged entity possessing market power within the relevant geographic market. Because we conclude that the FTC produced insufficient evidence of a well-defined relevant geographic market, we find that it did not show that the merged entity will possess such market power. The FTC’s failure to prove its relevant geographic market is fatal to its motion for injunctive relief.
The district court found that statistical evidence did not establish either the geographic market proposed by the FTC or the market proposed by Tenet.
It nonetheless found, relying on anecdotal evidence, that the merger would likely be anticompetitive. Our review of the record convinces us that the district court erred in several respects. The evidence in this case falls short of establishing a relevant geographic market that excludes the Sike-ston or Cape Girardeau areas. The evidence shows that hospitals in either or both of these towns, as well as rural hospitals throughout the area, are practical alternatives for many Poplar Bluff consumers.
In adopting the FTC’s position, the district court improperly discounted the fact that over twenty-two percent of people in the most important zip codes already use hospitals outside the FTC’s proposed market for treatment that is offered at Poplar Bluff hospitals.
The district court also failed to fully credit the significance of the consumers who live outside Poplar Bluff, particularly those patients within the FTC’s proposed geographic market who actually live or work closer to a hospital outside that geographic market than to either of the Poplar Bluff hospitals. If patients use hospitals outside the service area, those hospitals can act as a check on the exercise of market power by the hospitals within the service area.
See Freeman Hosp.,
69 F.3d at 264 n. 9. The FTC’s
contention that the merged hospitals would have eighty-four percent of the market for inpatient primary and secondary services within a contrived market area that stops just short of including a regional hospital (Missouri Delta in Sikeston) that is closer to many patients than the Poplar Bluff hospitals, strikes us as absurd. The proximity of many patients to hospitals in other towns, coupled with the compelling and essentially unrefuted evidence that the switch to another provider by a small percentage of patients would constrain a price increase, shows that the FTC’s proposed market is too narrow.
We question the district court’s reliance on the testimony of managed care payers, in the face of contrary evidence, that these for-profit entities would unhesitatingly accept a price increase rather than steer their subscribers to hospitals in Sikeston or Cape Girardeau. Without necessarily being disingenuous or self-serving or both, the testimony is at least contrary to the payers’ economic interests and thus is suspect.
In spite of their testimony to the contrary, the evidence shows that large, sophisticated third-party buyers can do resist price increases, especially where consolidation results in cost savings to the merging entities. The testimony of the market participants spoke to current competitor perceptions and consumer habits and failed to show where consumers could practicably go for inpatient hospital services.
The district court rejected the Cape Gir-ardeau hospitals as practicable alternatives because they were more costly. In so doing, it underestimated the impact of nonprice competitive factors, such as quality. The evidence shows that one reason for the significant amount of migration from the Poplar Bluff hospitals to either Sikeston, Cape Girardeau, or St. Louis is the actual or perceived difference in quality of care. The apparent willingness of Poplar Bluff residents to travel for better quality care must be considered. As the district
court
noted, healthcare decisions are based on factors other than price. It is for that reason that, although they are less expensive, HMOs are not always an employer’s or individual’s choice in healthcare services.
See Blue Cross and Blue Shield United of Wisconsin v. Marshfield Clinic,
65 F.3d 1406, 1412, 1410 (7th Cir.1995) (Posner, J.) (noting “[generally you must pay more for higher quality” and “the HMO’s incentive is to keep you healthy if it can but if you get very sick, and are unlikely to recover to a healthy state involving few medical expenses, to let you die as quickly and cheaply as possible.”) Thus, the fact that Cape Girardeau hospitals are higher priced than Poplar Bluff hospitals does not necessarily mean they are not competitors.
See, e.g., United States v. Archer-Daniels-Midland Co.
., 866 F.2d 242, 246-M7 (8th Cir.1988). The district court placed an inordinate emphasis on price competition without considering the impact of a corresponding reduction in quality.
We further find that although Tenet’s efficiencies defense may have been properly rejected by the district court, the district court should nonetheless have considered evidence of enhanced efficiency in the context of the competitive effects of the merger. The evidence shows that a hospital that is larger and more efficient than Lucy Lee or Doctors’ Regional will provide better medical care than either of those hospitals could separately. The merged entity will be able to attract more highly qualified physicians and specialists and to offer integrated delivery and some tertiary care. In view of “the significant changes experienced by the hospital industry in the recent past and the profound changes likely facing the industry in the near future, ... a merger, deemed anticompetitive to
day, could be considered procompetitive tomorrow.”
United States v. Mercy Health Servs.,
107 F.3d 632, 637 (8th Cir.1997) (citation omitted) (dismissing appeal as moot). The evidence shows that the merged entity may well enhance competition in the greater Southeast Missouri area.
In assessing the “commercial realities” faced by consumers, the district court did not properly evaluate evolving market forces in the rapidly-changing healthcare market. Significantly, it did not consider the impact of the entry of managed care into the Cape Girardeau market. The evidence shows that managed care has reduced prices in Poplar Bluff and in other markets. A similar downward pressure on prices is now being felt in Cape Girardeau, with the recent entry of managed care into that market. The district court also relied on the seemingly outdated assumption of doctor-patient loyalty that is not supported by the record. The evidence shows, and the district court acknowledged, that the issue of access to a provider through an insurance plan is determinative of patient choice. Essentially, the evidence shows that patients will choose whatever doctors or hospitals are covered by their health plan. Undeniably, although many patients might prefer to be loyal to their doctors, it is, unfortunately, a luxury they can no longer afford. Also, the advent of shorter hospital stays and more outpatient procedures has made travel less onerous and thus has broadened geographic markets. As much as many patients long for the days of old-fashioned and local, if expensive and inefficient, healthcare, recent trends in healthcare management have made the old healthcare model obsolete.
The reality of the situation in our changing healthcare environment may be that Poplar Bluff cannot support two high-quality hospitals. Third-party payers have reaped the benefit of a price war in a small corner of the market for healthcare services in Southeastern Missouri, at the arguable cost of quality to their subscribers. Antitrust laws simply do not protect that benefit when the evidence shows that there are other practical alternatives for healthcare in the area. We are mindful that competition is the driving force behind our free enterprise system and that, unless barriers have been erected to constrain the normal operation of the market, “a court ought to exercise extreme caution because judicial intervention in a competitive situation can itself upset the balance of market forces, bringing about the very ills the antitrust laws were meant to prevent.”
United States v. Syufy Enters.,
903 F.2d 659, 663 (9th Cir.1990). This appears to have even more force in an industry, such as healthcare, experiencing significant and profound changes. Under the circumstances presented in this case, the FTC has not shown a likelihood of success on the merits of its section 7 complaint and we find the district court erred in granting injunctive relief.
III. CONCLUSION
The judgment of the district court is reversed, the order enjoining the merger is dissolved and this action is remanded for proceedings consistent with this opinion.