*961 This is an antitrust case arising under § 7 of the Clayton Act, 15 U.S.C. § 18 . The Federal Trade Commission and the State of North Dakota moved to enjoin Sanford Bismarck's acquisition of Mid Dakota Clinic, P.C., alleging that the merger would violate the Act. The district court 1 granted a preliminary injunction after determining that the plaintiffs were likely to succeed in proving that the acquisition would substantially lessen competition in four types of physician services in the Bismarck-Mandan area. The companies appeal, and we affirm.

Sanford is an integrated healthcare system operating in North Dakota and several other States. In the Bismarck-Mandan region, Sanford operates an acute care hospital and multiple clinics. The company employs approximately thirty-seven adult primary care physicians, five pediatricians, eight OB/GYN physicians, and four general surgeons.

Sanford's two main competitors in the Bismarck-Mandan region are Mid Dakota and Catholic Health Initiatives St. Alexius Health. Mid Dakota is a multi-speciality physician group that includes approximately twenty-three adult primary care physicians, *962 six pediatricians, eight OB/GYN physicians, and five general surgeons. Catholic Health employs eighty-eight physicians, the majority of whom are hospitalists; five are adult primary care physicians.

In North Dakota, there are three leading commercial insurers: Blue Cross Blue Shield North Dakota, Medica, and Sanford Health Plan. Blue Cross is the largest, accounting for 61% of the North Dakota health insurance market in 2016. Blue Cross has a participation agreement with every general acute care hospital in the State and with approximately 99% of practicing physicians. Sanford and Medica accounted for 31% and 8% of the 2016 market, respectively.

In 2015, Mid Dakota offered itself for sale, and both Catholic Health and Sanford submitted purchase proposals. Mid Dakota initially executed a letter of intent with Catholic Health, but after Catholic Health terminated the deal, Mid Dakota began negotiations with Sanford. The two entities executed a term sheet in August 2016 providing that Sanford would acquire the assets of Mid Dakota. Ten months later, they signed a stock purchase agreement in which Sanford agreed to purchase the outstanding capital stock of Mid Dakota. If the companies merge, then Sanford will have the following market shares in the Bismarck-Mandan region: 99.8% of general surgeon services, 98.6% of pediatric services, 85.7% of adult primary care physician services, and 84.6% of OB/GYN physician services.

The Federal Trade Commission and North Dakota Attorney General brought an action seeking to enjoin the merger. Section 7 of the Clayton Act provides that no person engaged in commerce and subject to the jurisdiction of the FTC shall acquire the stock or assets of another person if "the effect of such acquisition may be substantially to lessen competition." 15 U.S.C. § 18 . The FTC alleged that Sanford's acquisition of Mid Dakota would contravene this proscription and sought an injunction under 15 U.S.C. §§ 26 and 53(b). The complaint asserted that Sanford's plan "to purchase [Mid Dakota's] assets through two separate transactions" would "violate Section 7 of the Clayton Act by substantially lessening competition." After a four-day evidentiary hearing, the district court found that the plaintiffs were likely to succeed on the merits of their claim. The court therefore issued a preliminary injunction.

We review the district court's grant of a preliminary injunction for abuse of discretion. Lankford v. Sherman , 451 F.3d 496 , 503 (8th Cir. 2006). "An abuse of discretion occurs where the district court rests its conclusion on clearly erroneous factual findings or erroneous legal conclusions." Id. at 503-04 . A district court may enjoin a proposed merger 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." FTC v. Tenet Health Care Corp. , 186 F.3d 1045 , 1051 (8th Cir. 1999) (quoting 15 U.S.C. § 53 (b) ).

To evaluate the FTC's likelihood of success on the merits, the district court employed a burden-shifting method endorsed by the D.C. Circuit in United States v. Baker Hughes Inc. , 908 F.2d 981 , 982-83 (D.C. Cir. 1990). Under this approach, the plaintiffs must first present a prima facie case that the merger will result in an undue market concentration for a particular product or service in a particular geographic area. That showing creates a presumption that the merger will substantially lessen competition. The burden of production then shifts to the defendant to rebut the presumption, and, on a sufficient showing, back to the plaintiffs to present additional evidence of anticompetitive effects. The ultimate burden *963 of persuasion remains at all times with the plaintiffs.

The companies argue that the district court improperly shifted the ultimate burden of persuasion to the defendants when it required them to produce rebuttal evidence that "clearly shows" that no anticompetitive effects were likely. The district court cited United States v. Philadelphia National Bank , 374 U.S. 321 , 83 S.Ct. 1715 , 10 L.Ed.2d 915 (1963), where the Court said that "a merger which produces a firm controlling an undue percentage share of the relevant market, and results in a significant increase in the concentration of firms in that market, is so inherently likely to lessen competition substantially that it must be enjoined in the absence of evidence clearly showing that the merger is not likely to have such anticompetitive effects." Id. at 363 , 83 S.Ct. 1715 (emphasis added). The D.C. Circuit in Baker Hughes reviewed later decisions that used the term "show" instead of "clearly show," and concluded that the Supreme Court, without overruling Philadelphia National Bank , "has at the very least lightened the evidentiary burden on a section 7 defendant." 908 F.2d at 990-91 .

We conclude that there was no legal error by the district court here. The court followed the analytical framework of Baker Hughes , and specified that "[t]he FTC has the burden of persuasion at all times." While Baker Hughes adverted to the Supreme Court's shift in terminology since the 1960s, the D.C. Circuit also recognized that "[t]he more compelling the prima facie case, the more evidence the defendant must present to rebut it successfully." Id. at 991 . In the context of this case, where the plaintiffs presented strong evidence of monopolization or near-monopolization in each service line, it was necessary for the defendants to make a strong presentation in rebuttal. We are not convinced that the quotation from Philadelphia National Bank , read in the context of the district court's order as a whole, shifted the burden of persuasion to the defendants or required the defendants to do anything more than produce evidence showing that the FTC's prima facie case "inaccurately predicts the relevant transaction's probable effect on future competition." Id.

Turning to the district court's findings of fact, we review the court's determination of the relevant market for clear error. FTC v. Lundbeck, Inc. , 650 F.3d 1236 , 1239 (8th Cir. 2011). The district court first defined the four relevant product markets as adult primary care services, pediatric services, OB/GYN physician services, and general surgeon services, and defined the relevant geographic market as the Bismarck-Mandan area. The district court employed the "hypothetical monopolist test," which is commonly used in antitrust actions to define the relevant market. See Saint Alphonsus Med. Ctr.-Nampa Inc. v. St. Luke's Health Sys., Ltd. , 778 F.3d 775 , 784 (9th Cir. 2015). The test asks whether a hypothetical monopolist could impose a "small but significant nontransitory increase in price" in the proposed market. Id. If consumers would defeat the price increase by switching to products outside of the proposed market, then the market definition is too narrow and must be redefined. Here, the court found that "commercial health insurers would accept a hypothetical monopolist's [small but significant nontransitory increase in price] rather than market a health insurance plan in the Bismarck-Mandan area that did not include Bismarck-Mandan area physicians providing adult PCP services, pediatrician services, OB/GYN services, and general surgeon services." The court's determination was supported by empirical analysis of claims data and testimony of representatives from North Dakota's three largest *964 insurance companies, including Sanford Health Plan-the insurance provider within Sanford Health's integrated care system. The representatives each testified that an insurance plan's network must include each of these services to be competitive in the Bismarck-Mandan area.

The companies argue that the district court failed to account for Blue Cross's dominant position in the market: a provider in North Dakota, they argue, would not be able to impose a price increase on Blue Cross. In determining the relevant market, however, the question is not whether a monopolist would increase prices on an insurer, but whether the insurer "will shift from one product to the other in response to changes in their relative costs." SuperTurf, Inc. v. Monsanto Co. , 660 F.2d 1275 , 1278 (8th Cir. 1981). When applied to a merger between health care providers, the hypothetical monopolist test evaluates whether an insurer could avoid a price increase by contracting with physicians who offer services that are outside of the proposed service markets or who are located in a region outside of the proposed geographic market. See FTC v. Penn State Hershey Med. Ctr. , 838 F.3d 327 , 342 (3d Cir. 2016) ; Horizontal Merger Guidelines § 4.1.3 ("The hypothetical monopolist's incentive to raise prices depends both on the extent to which customers would likely substitute away from the products in the candidate market in response to such a price increase and on the profit margins earned on those products.") (emphasis added). Blue Cross's alleged bargaining power would not impact its ability to find substitute physician services when facing a price increase; the undisputed testimony was that there are no functional substitutes to a plan offering adult primary care services, pediatric services, OB/GYN physician services, general surgeon services in the Bismarck-Mandan area. The court thus did not clearly err in defining the relevant market.

The district court next found that the government made a sufficient prima facie showing because "[t]he changes in [Herfindahl-Hirschman Index] in each of the four physician service lines are well above the Merger Guidelines' threshold for presumption that the proposed transaction is likely to enhance market power." In rebuttal, the companies raised four principal arguments: (1) market concentration has no relationship to bargaining power in the North Dakota healthcare market, (2) Catholic Health was poised to enter the market to compete with Sanford after the merger, (3) merger efficiencies offset the potential to harm consumers, and (4) Mid Dakota's weakened condition justified the merger. The district court evaluated the evidence in support of these contentions and found that the FTC's evidence still carried the day.

The companies first argued that the ordinary presumption that increased market concentration will lead to increased prices does not apply to the North Dakota healthcare market, because Blue Cross is a dominant buyer that sets reimbursement rates using a statewide pricing schedule. Even if a provider has a monopoly or near-monopoly in one region, they argue, the provider would be unable to increase Blue Cross's reimbursement rates, so the merger would not impact prices. The district court viewed this contention as a "powerful buyer defense," and evaluated whether (1) Blue Cross, as a powerful buyer, could use its leverage to sponsor entry to the market, or (2) whether Blue Cross would be able to obtain lower prices from alternative suppliers after a merger. See Horizontal Merger Guidelines § 8. Finding that neither was likely, the court rejected the proffered defense.

On appeal, the companies dispute the district court's characterization of their argument as a powerful buyer defense , and *965 complain again that the district court shifted the burden of persuasion. Yet however the submission is described, the district court expressly placed the burden of persuasion on the plaintiffs, and found that despite the market power of Blue Cross, there was a relationship between market concentration after the merger and bargaining leverage. A Blue Cross representative testified that Sanford, after the proposed merger, would indeed have the power to force Blue Cross to choose between raising prices or leaving the Bismarck-Mandan region. And there was evidence that Blue Cross in the past was forced to modify contract terms with a near-monopoly provider in another area of the State. The district court did not clearly err in crediting this evidence and finding that it outweighed the testimony of the companies' expert that the merged company would be unlikely to extract higher reimbursements from Blue Cross.

The companies also argued that Catholic Health, a competitor of Sanford, was poised to enter and compete in the Bismarck-Mandan market. They contend that Catholic Health's entrance would counteract the anticompetitive effects of the merger. Entry of competitors into a market can offset anticompetitive effects, however, only if the entrance is "timely, likely, and sufficient in its magnitude, character, and scope to deter or counteract the competitive effects of concern." Horizontal Merger Guidelines § 9; see also FTC v. Cardinal Health, Inc. , 12 F. Supp. 2d 34 , 55 (D.D.C. 1998). The court found that Catholic Health would not be able to enter the market quickly after the merger. Catholic Health's president testified that the company faced difficulties recruiting physicians in the Bismarck-Mandan area. Although the president testified that the company could recruit doctors to enter the market in the short term, Appellants' App. I, at 174, he also explained that it would take up to twice as long to establish a name and reputation that could compete with Sanford. Id . at 156. On appeal, the companies point to testimony that Catholic Health intended to enter the market and had recruited a top physician in Bismarck. But the district court did not clearly err in giving more weight to Catholic Health's testimony that it could not timely compete with Sanford in the Bismarck-Mandan market, and in finding that entry of this competitor would not come soon enough to offset anticompetitive effects of the merger.

The companies further assert that the district court erred in evaluating the quality efficiencies that would be generated by the merger. In the district court, the companies pointed to five consumer benefits that would be available to more consumers in the Bismarck-Mandan region after the merger: (1) Imagenetics, a program integrating genetic medicine into primary care, (2) behavioral health therapists embedded into primary care clinics, (3) cancer care trials and cancer care outreach to communities outside the Bismarck-Mandan area, (4) a combined and customized electronic medical record system that would better integrate and coordinate patient care, and (5) recruitment of subspecialists to the area.

For these efficiencies to counteract anticompetitive effects, they must be independently verifiable and derived specifically from the merger: "[T]hey must be efficiencies that cannot be achieved by either company alone." FTC v. H.J. Heinz Co. , 246 F.3d 708 , 721-22 (D.C. Cir. 2001). The district court found, however, that only the Imagenetics program was merger-specific. This putative benefit was not enough to offset the anticompetitive effects of the merger, the court found, because "[e]fficiencies almost never justify a merger to monopoly or near-monopoly." Horizontal *966 Merger Guidelines § 10; see also Saint Alphonsus , 778 F.3d at 790 .

The companies argue that the other four proposed efficiencies were also merger-specific, but the district court's finding was adequately supported by the record. The FTC's expert reported that Mid Dakota could make the other quality gains without the merger: he provided evidence that patient demands-not practice size-drive physician recruitment, that a combined electronic medical record system was neither required nor certain to integrate and coordinate patient care, and that Mid Dakota and Sanford already provided community outreach services and could expand those services without the merger. Sanford's own executive admitted that Mid Dakota could employ a behavioral health therapist without the merger. The companies argued in the district court that Mid Dakota did not offer these quality improvements, but the relevant issue was whether Mid Dakota was capable of developing them without the merger. On the record as a whole, the district court's finding on merger-specific efficiencies was not clear error.

The district court also understood the companies to raise a "weakened competitor" defense, asserting that dim long-term prospects of Mid Dakota justified its merger with Sanford. The court rejected this argument based on sufficient evidence that Mid Dakota was financially healthy. Mid Dakota increased revenues in the three years before the lawsuit; physician compensation was 32% above the national average; and minutes from a Mid Dakota shareholder meeting in 2015 shows that the motivation to sell was high share value, not concern about long-term viability. The companies argue that they did not raise a freestanding "weakened competitor defense," but merely urged the district court to consider the financial status of Mid Dakota in the context of its arguments about efficiencies and potential entry to the market by Catholic Health. Assuming that is true, the district court's findings on those asserted efficiencies were not clearly erroneous.

The judgment of the district court is affirmed.