On September 27, 2016, the US Court of Appeals for the Third Circuit handed an important victory to the Federal Trade Commission (FTC) and the Commonwealth of Pennsylvania in a closely watched hospital merger case involving two Harrisburg-area health systems: Penn State Hershey Medical Center (Hershey) and PinnacleHealth System (Pinnacle). Earlier in 2016, the US District Court for the Middle District of Pennsylvania denied the FTC’s motion to enjoin the proposed merger pending conclusion of an administrative proceeding. The lower court held that the agency failed to define a relevant geographic market and thus could not demonstrate a likelihood of success on the merits as required for the grant of an injunction.
The Third Circuit reversed, finding that the lower court erred in both the formulation and the application of the proper legal test to determine the relevant geographic market. The court’s decision was based on its conclusion that geographic markets in hospital merger cases are to be defined based principally on evidence from payors, and that payors had made clear that Hershey and Pinnacle needed to remain separate to ensure competitive bidding for payor contracts. The court further found that the government met its burden for a preliminary injunction and that the defenses advanced by the merging parties were inadequate to overcome the presumption that the merger would be anticompetitive. The decision provides clear guidance on the appropriate tests for determining the geographic markets in hospital mergers, while also suggesting that efficiencies claimed in many hospital transactions may face increased scrutiny in future cases.
Hershey and Pinnacle (collectively, the Hospitals) both operate in the Harrisburg, Pennsylvania, area. Hershey, located in the town of the same name, is an academic medical center that offers basic care but also provides more advanced, specialized services. Pinnacle is a three-campus health system, with two facilities located in Harrisburg and a third in Mechanicsburg. It provides primary and secondary services and a limited range of more complex services. The two systems notified the FTC of their proposed merger in April 2015.
Following an eight-month investigation, the FTC filed an administrative complaint alleging that the proposed merger would substantially lessen competition for general acute care services sold to commercial insurers in a four-county area surrounding Harrisburg (the Harrisburg Area). Together with the Commonwealth of Pennsylvania, the agency subsequently filed suit in the Middle District of Pennsylvania seeking a preliminary injunction pending the resolution of the FTC’s administrative proceeding.
Following five days of hearings, the district court concluded that the government failed to properly define the relevant geographic market. Instead of focusing, as the government did, on the location of hospitals to which commercial health plans turn to create their provider networks, the district court emphasized the substantial in-migration of patients from counties outside the Harrisburg Area. Given this broad patient draw area and the district court’s view that patients in rural central Pennsylvania will typically travel significant distances for services, the court concluded that the government’s proposed geographic market was too small and unreasonably excluded important competitors.
Although this failure to define a relevant geographic market was itself fatal to the government’s motion, the district court also opined on several other arguments raised by the parties. First, the court described long-term contracts with two key insurers as “extremely compelling” evidence of the merger’s potential effects on competition. The district court stated it “simply [could not] be blind to this reality when considering the import of the hypothetical monopolist test advanced by the Merger Guidelines.” Second, the district court found that repositioning by other health systems would constrain the merging health systems. Finally, the district court credited various efficiencies claimed by the merging parties, including capital savings from a planned bed tower that would no longer be needed and an improved ability to engage in risk-based contracting.
After the district court denied the government’s motion for a preliminary junction, the government timely appealed to the Third Circuit.
Three Flaws in the District Court’s Geographic Market Analysis
The Third Circuit soundly rejected the district court’s geographic market analysis. Noting that the definition of the relevant geographic market is an essential prerequisite of any merger analysis, the Third Circuit observed that one common method—and the standard agreed upon by the district court, government and Hospitals—used to define geographic markets is the hypothetical monopolist test, as described in the Merger Guidelines. Under this test, if a hypothetical monopolist could impose a small but significant non-transitory increase in price (SSNIP) in the proposed market, then the market is properly defined. Although the district court properly identified this test, the Third Circuit concluded that the lower court erred in both its formulation and its application of the test. Reasoning that a district court commits legal error where it applies an incomplete economic analysis or an erroneous economic theory to the facts that constitute the relevant geographic market, the Third Circuit reviewed the district court’s analysis de novo.
The court identified three distinct errors in the district court’s analysis. First, the district court failed to formulate the hypothetical monopolist test correctly. According to the district court, the key analysis was to delineate the geographic area where “few patients leave . . . and few patients enter.” This attention to patient flows explains the emphasis in the district court’s decision on the substantial in-migration to Hershey from counties outside the Harrisburg Area. However, as the Third Circuit held, this focus was “inconsistent with the hypothetical monopolist test. Rather it [was] one half of a different test utilized in non-healthcare markets to define the relevant geographic market: the Elzinga-Hogarty test.” While the court acknowledged that many courts formerly used this test, subsequent empirical research had discredited the test for determining geographic markets in the hospital sector. The court also noted that even if patient flow data were consistent with the hypothetical monopolist test, the district court did not consider patient outflow data. Undisputed evidence presented by the government showed that 91 percent of Harrisburg residents receive general acute care services in the Harrisburg Area.
Second, the district court ignored the “commercial realities” of the health care market when it focused on patient, rather than insurer, reactions to a SSNIP. Acknowledging the views of the FTC and other courts, the Third Circuit agreed that health care markets are characterized by a two-stage model of competition. First, hospitals compete to be included in a health insurer’s provider network; second, in-network providers compete to attract members of an insurer’s plans. The court found that insured patients are relatively insensitive to price fluctuations in contracts between hospitals and insurers, and cited the Hospitals’ own study showing that only 2 percent of respondents considered out-of-pocket costs in choosing a hospital. As a result, the Third Circuit found that an analysis of the effect of the SSNIP must ordinarily focus on health insurers. While the court cautioned that analysis of patient reactions may be appropriate in some circumstances, it held that the district court erred in completely disregarding the role that insurers play in the health care market.
Finally, the district court erred by basing part of its analysis of the relevant geographic market on the long-term contracts (at pre-merger prices) that the Hospitals entered into with health insurers. The Third Circuit clarified that the relevant test is not “‘the contractual constraints binding a particular plaintiff,’ but . . . whether a hypothetical monopolist could profitably impose a SSNIP.” The court held the Hospitals’ long-term agreements with insurers “have no place in the relevant geographic market analysis.” The Third Circuit also rejected the district court’s reluctance to make predictions about potential future anticompetitive effects, saying that this is “exactly what we are asked to do.”
Assessing the Government’s Likelihood of Success on the Merits
After rejecting the district court’s analysis, the Third Circuit examined whether the government met its burden to define a relevant geographic market. The court credited extensive testimony by insurers that they could not market health plans in the Harrisburg Area that did not include at least one of the merging hospitals, and that they do not consider hospitals in neighboring counties to be suitable alternatives to Harrisburg-area hospitals. The court also highlighted a natural experiment involving an insurer that unsuccessfully marketed a network that excluded the merging hospitals. Based on this evidence, the court concluded that the government had demonstrated that a hypothetical monopolist could profitably impose a SSNIP in the Harrisburg Area and that the government had properly defined a relevant geographic market. Within this geographic market, the court found a high level of market concentration, resulting in a presumption of anticompetitive effects.
The court also rejected the Hospitals’ argument that payors have sufficient bargaining leverage to defeat a SSNIP because payors can threaten to exclude the Hospitals from their networks. The court argued that whatever leverage the payors have after the merger, they have now, and that the relevant inquiry is whether the merger will cause a significant increase in the Hospitals’ bargaining leverage such that they could profitably impose a SSNIP.
The court next considered whether the Hospitals could rebut this presumption by refuting the existence of anticompetitive effects or demonstrating that any such effects would be offset by extraordinary efficiencies.
The court’s assessment of anticompetitive effects focused on alleged repositioning by competitors. Although the court acknowledged that repositioning within the Harrisburg Area lessened some of the concern posed by the proposed merger, it ultimately held that this would not constrain post-merger pricing. Health insurer testimony concerning the viability of networks that excluded the Hospitals was critical in this analysis.
With respect to alleged efficiencies enabled by the proposed merger, the court expressed substantial skepticism about the very availability of an efficiencies defense. The court nevertheless acknowledged that the Merger Guidelines and other courts of appeal have held that the defense is cognizable. Without taking a position on the availability of the defense, the court concluded that the Hospitals failed to present efficiencies that meet the requirements articulated in the Merger Guidelines.
The Hospitals claimed that the merger enabled them to avoid the expense associated with construction of a 100-bed tower necessary to alleviate capacity constraints. The court agreed that capital savings may play a role in an efficiencies defense but commented that they must “be verifiable and must not result in any anticompetitive reduction in output.” On this measure, the claimed efficiency failed. The court found that Hershey’s own efficiencies analysis showed it required only 13 additional beds to address capacity constraints. Perhaps more importantly, the court found that the ability to forego building the new tower was a reduction in output. Citing the Merger Guidelines, the court found that the FTC will not consider efficiencies that arise from anticompetitive reductions in output or service.
The Hospitals also claimed that the merger would enhance their risk-based contracting efforts. The Third Circuit also rejected this efficiency. First, it noted that the district court concluded that each system was already capable of engaging in risk contracting and therefore the efficiency was not merger-specific. Second, the Third Circuit found that, even if the merger enhanced the Hospitals’ ability to engage in risk-based contracting, they still had to “demonstrate that such a benefit would ultimately be passed on to consumers. It is not clear from the record how this would be so beyond the mere assertion that it would save the hospitals money and such savings would be passed on to consumers.”
This decision is a very strong win for the FTC. The Third Circuit largely adopted the agency’s analytical methods and arguments. Moreover, the decision underlines the importance of market definition and market concentration in litigated merger cases. Once the Third Circuit ruled that the geographic market was much narrower than that suggested by the district court, the Hospitals faced a presumption that their proposed transaction was anticompetitive based on market concentration statistics, and they were unable to rebut this presumption.
The Third Circuit’s opinion offers obvious guidance on geographic market definition in the health care setting. The clarity is significant given the uptick in agency enforcement activity in recent years. In two prior hospital merger cases litigated by the FTC, geographic market definition was uncontested. Yet some recent cases also saw the adoption and elaboration of the agency’s two-tier model of competition, with its emphasis on the importance of health insurers as the “purchasers” of general acute care services. While the Third Circuit emphasized that, in the health care context, considering the effects of price increases on patients may be appropriate, the decision provides very clear guidance on the importance of focusing first on the role of health insurers.
Between announcement of the merger and the litigation, the Hospitals concluded long-term agreements with key payors that restricted the merged system’s ability to engage in anticompetitive practices. While the Third Circuit chastised the district court for weighing the effect of these contracts in its assessment of the relevant geographic market, it offered no guidance as to whether such agreements may mitigate the potential anticompetitive effects of a proposed merger. Health systems pursuing mergers in the future will likely continue to examine contracting options with payors that address upfront potential allegations by the FTC or others that their transaction will lead to anticompetitive effects.
The Third Circuit’s decision may also present challenges for efficiencies defenses in the future. Capital avoidance efficiencies, such as the Hershey bed tower, are common in many health care mergers, but the court’s approach may encourage other courts to view decisions to avoid new construction as an output reduction rather than an efficiency. Similarly, health systems that wish to argue that a merger enhances their ability to engage in risk-based contracting likely will need to demonstrate concrete consumer benefits in order to gain credit for such efficiencies.