On February 10, in Saint Alphonsus Medical Center-Nampa v. St. Luke’s Health System (St. Alphonsus), the U.S. Court of Appeals for the Ninth Circuit handed the Federal Trade Commission yet another provider-merger win. The appellate court affirmed an Idaho district court’s decision that St. Luke’s’ acquisition of a large primary-care physician (PCP) practice violated Section 7 of the Clayton Act. Add this to the FTC’s wins in three previous hospital-merger decisions—its 2007 decision Evanston Northwestern Health Care striking down a hospital merger in the north Chicago suburbs, the 2012 federal district court decision in FTC v. OSF Health Care System preliminarily enjoining the merger of two Rockford, Illinois hospitals, and the 2014 Sixth Circuit’s decision in FTC v. ProMedica Health System decision affirming the FTC’s decision that the merger of two Toledo, Ohio hospitals was unlawful—and it’s obvious that the antitrust environment for mergers of competing providers has become much less hospitable than in the past. Some believe this quite ironic in light of the Affordable Care Act, which appears to encourage greater integration and coordination among health care providers.
In St. Alphonsus, the FTC, Idaho Attorney General, St. Alphonsus Medical Center, and a surgery center challenged St. Luke’s acquisition of the Saltzer Medical Group, the largest and most prestigious group of PCPs providing services in Nampa, Idaho (and in the state). Nampa, Idaho’s second largest city with about 85,000 residents, is located about 20 miles west of Boise, the state capital and largest Idaho city. St. Luke’s employed seven PCPs in Nampa and acquired an additional sixteen through the Saltzer acquisition. St. Alphonsus employed nine, and there were several solo practices. According to the district court, the transaction resulted in St. Luke’s employing about 80 percent of the PCPs in Nampa.
St. Alphonsus and the surgery-center plaintiff challenged the transaction primarily on vertical grounds—that after the merger, the acquired Saltzer physicians would admit most or all their patients to St. Luke’s, foreclosing St. Alphonsus and the surgery center from a substantial percentage of patients and thus revenues, significantly weakening their ability to compete and increasing St. Luke’s’ market power in the market for hospital services. The FTC and state, on the other hand, challenged the acquisition on horizontal grounds, claiming that the loss of competition between Saltzer and the St. Luke’s physicians resulting from the acquisition would increase St. Luke’s bargaining power and thus permit it to increase its reimbursement for adult PCP services. The district court agreed with the FTC and state and ordered St. Luke’s to divest the Saltzer physicians without reaching St. Alphonsus’s vertical claim.
The Ninth Circuit affirmed, finding that the district court applied the correct principles of law and made no clear errors in its findings of fact. The court began by noting that antitrust merger analysis is forward looking: the court must predict the merger’s likely effect on competition in the future. To do so, it applies a burden-shifting framework. The FTC must first define the relevant market and then sustain an initial burden to present a prima facie case that the transaction is likely to substantially lessen competition. It can do so, according to the court, “simply by showing a high market share,” although it can introduce other evidence of likely anticompetitive effects as well. If the FTC succeeds, a rebuttable presumption arises that the merger is unlawful—a “prima facie case”—and the burden of going forward shifts to the defendants to rebut the presumption.
The parties agreed that the relevant product market was limited to adult PCP services. They disagreed, however, about the scope of the relevant geographic market. The FTC argued that it was limited to Nampa, while the defendants claimed it encompassed a significantly larger area. The district court and Ninth Circuit agreed with the FTC. Noting that the relevant geographic market is the area “‘where buyers can turn for alternative sources of supply,’” the court applied the hypothetical monopolist methodology for defining relevant markets as prescribed in the government’s Horizontal Merger Guidelines and recent case law. Under that framework, an area (and the suppliers therein) constitutes a relevant geographic market if a hypothetical monopolist of the relevant product—here, adult PCP services—likely could and would profitably raise prices by a non-significant amount for a significant period of time. If the price increase would be unprofitable because too many of the hypothetical monopolist’s customers would switch to more distant sellers, the market would be expanded until it included only those sellers who, acting jointly as a hypothetical monopolist, could profitably raise price because too few customers would switch to yet more distant suppliers.
Applying that methodology to the facts, the court held that a hypothetical monopolist of adult PCP services in Nampa could profitably raise the price of those services to health plans. The court relied on the strong preference of patients for local providers, testimony from health plans that Nampa PCP participation in their networks was crucial for a competitively viable network, and that some 65 percent of Nampa residents used PCPs in Nampa, as opposed to PCPs outside of Nampa. Based on these facts, the court explained that a hypothetical monopolist of PCP services in Nampa could profitably demand and obtain significantly higher reimbursement from health plans, meaning that the relevant geographic market was limited to Nampa. Worth noting is that under hospital-merger decisions in the 1990s when courts defined geographic markets based primarily on patient-flow statistics, the 35 percent outflow of Nampa residents likely would have been sufficient to require expansion of the relevant geographic market beyond Nampa. Here, the outflow of Nampa residents to Boise PCPs was significant, but the court explained that many of those residents worked in Boise and would not affect the ability of Nampa PCPs to raise prices. A price increase by Nampa PCPs would not induce other Nampa residents (or health plans serving them) to flock to Boise.
In then determining whether the FTC proved a prima facie case, the court, interestingly, relied primarily on the post-merger level of market concentration and the extent to which the transaction increased that level as measured by the Herfindahl-Hirschman Index or HHI. This is interesting because when the concern from a merger is whether the loss of competition between the merging parties will permit them, regardless of the actions of other firms in the market, to raise prices (so-called “unilateral effects”), there is general agreement that market concentration provides little insight into the merger’s likely effect. Rather, market concentration is important when the concern is “coordinated effects”—concern that the merger may result in the merged firm and others coordinating their competitive conduct. In St. Alphonsus, all agreed that only a unilateral effect was the potential problem with the transaction.
In any event, the court noted that the district court had found that the post-merger HHI was 6,219 and the transaction increased the HHI by 607—far exceeding the 2,500 HHI and 200 point increase that, under the Merger Guidelines, result in a rebuttable presumption of illegality. The court explained that “[t]he extremely high HHI on its own establishes the prima facie case.” Also interesting is that while the court stated that a prima facie case can result from “showing a high [post-merger] market share,” it did not even mention the market share resulting from the Saltzer transaction.
To its credit, however, the court examined other factors in addition to the HHI in assessing the FTC’s case. In particular, it noted that the Saltzer and St. Luke’s physicians were each other’s closest and most direct competitors, which is the single most important variable in determining whether a horizontal merger will have adverse unilateral effects; all else being equal, the more substitutable the merging parties are for one another compared with their substitutability with other competitors, the greater the likelihood of unilateral effects. The district court had found that if Nampa residents using St. Luke’s physicians could not use them, 50 percent would switch to Saltzer physicians rather than to other PCPs, and if Saltzer patients could not use that group, a third of them would switch to St. Luke’s physicians. In addition, the Ninth Circuit cited bad St. Luke’s documents indicating St. Luke’s belief that the acquisition would increase its clout with health plans, health-plan testimony that the transaction would provide St. Luke’s with the ability to increase reimbursement because plans had to include at least one of the groups in their plans for a viable network, and the experience in another Idaho town when St. Luke’s acquired a practice and successfully negotiated higher reimbursement for its physicians. The district court had cited an additional factor—that St. Luke’s could increase the facility fee for ancillary services provided by the employed group above that charged when the group was separate. The Ninth Circuit rejected that finding because ancillary services were a different relevant product market than adult PCP services, and the district court provided no findings on the merger’s effect on St. Luke’s market power in that market.
Having found that the FTC proved a prima facie case, the focus shifted to the defendants’ rebuttal; some of the court’s discussion there is troubling. Defendants can rely on a potential plethora of factors to argue that the government’s statistics provide an inaccurate prediction of the merger’s likely effect on competition. Four, however, are most important: (1) Evidence that the government defined the relevant market too narrowly, leading to erroneously high post-merger concentration and market share figures; (2) low entry barriers; (3) substantial efficiencies; and (4) acquired-firm financial weakness.
The Ninth Circuit affirmed the district court finding of significant entry barriers, and the defendants made no financial weakness argument. Rather, they focused on geographic market definition, an argument which, as noted before, they lost, and on claimed efficiencies from the transaction. Their efficiencies argument implicated the Affordable Care Act; they argued that the Act encouraged substantial integration and cooperation among providers to incentivize a switch from fee-for-service to value-based delivery of care and reimbursement. More specifically, they argued that the merger would provide the large core of physicians necessary for risk-based contracting and improve care through an integrated electronic records system. The district court agreed with them that the merger would likely improve patient outcomes, but held “that there are other ways to achieve the same effect[s] that do not run afoul of the antitrust laws and do not run the risk of increased costs”—i.e., that the efficiencies were not “merger specific.”
The Ninth Circuit seemed quite skeptical of efficiencies arguments in general. First, it indicated that there is some question whether efficiencies are even a cognizable argument in favor of a merger, although it did note that some courts and the FTC have indicated they are. Second, it noted that even if so, “none of the reported appellate decisions have actually held that a § 7 defendant has rebutted a prima facie case with an efficiencies defense” and that “the parameters of the defense remain imprecise.” Third, the court noted that merger cases are sufficiently complex “without adding to the judicial balance a prediction of future efficiencies.” And it might have added the age-old antitrust problem of figuring out a practical way, once the efficiencies are predicted, of balancing them against the effects of the increased market power from the merger to determine which effect predominates.
The court held that if there is an efficiencies defense (and it assumed there was for purposes of the case), the claimed efficiencies must relate solely to the merger’s effect on competition, not some other value. For example, the court explained that even if the merger might enable St. Luke’s “to better serve patients” or improve the delivery of health care to patients, that was not a cognizable argument in support of the merger. Rather, the court seemed to suggest that only efficiencies that lower the cost of providing care and lower the price charged health plans count.
If this is what the court meant, it is troubling. It seems tantamount to holding that improvements in quality and other non-price competitive variables fail to count as “efficiencies,” a position that other courts, as well as the FTC itself, reject. While quality improvements may not lower costs or actual prices charged health plans (indeed, they often do the opposite), they can result in lower quality-adjusted prices, which should be the variable of importance. If, for example, the merger permitted the merged entity to raise prices five percent and increased quality in the same proportion, it can be argued that the merger did not permit the merged firm to raise price. Not every benefit to consumers from a merger can be measured empirically and quantified, but that should not mean that these benefits are irrelevant to the analysis. Somewhat more amorphously, if it could be reasonably estimated that quality improvements from the merger would save X number of lives, should the value of those lives not be considered? And don’t hospitals compete on the basis of quality and other non-price factors as well as on price? The FTC must think so because a number of its complaints in provider-merger challenges, including that in St. Alphonsus, allege that one adverse effect of the transaction will be to “dampen the combined entity’s incentive to improve or continue offering high quality services.” Similarly, the speeches and articles of FTC officials emphasize their concern with quality as well as price. In sum, improvements in the quality of its services increase a firm’s competitive strength.
In any event, the Ninth Circuit agreed with the district court that even if the claimed quality improvements could be considered in support of the transaction, they were not merger-specific as required by the Merger Guidelines and case law. There was no evidence, according to the court, that St. Luke’s needed more than the physicians it already employed to negotiate risk-based contracts, and the sharing of an electronic records system could be accomplished contractually rather than through a merger—and without the anticompetitive effects likely from the merger.
All that seems true, but what appears overlooked is that, all else being equal, the tighter the parties are integrated, the more effective, efficient, and likely these programs would probably be. This increased efficiency is, in general, a major reason for integration, whether vertical or horizontal. All else equal, the greater the integration, the greater the efficiencies. Would this likely marginal increase in efficiencies from a merger over some less integrative structure justify the transaction here given the facts of the government’s case? Probably not, but courts should recognize that in estimating likely efficiencies, the tightness of the form of integration counts.
Finally, the defendants argued that in ordering relief, the district court should have adopted a conduct remedy rather than ordering divestiture, a structural remedy. The Ninth Circuit noted that the defendants, in convincing the district court not to issue a preliminary injunction early-on in the litigation, had assured the court that later divesture, if ordered, would not be a problem. The court also cited the usual Section 7 relief principles—that divestiture is the preferred remedy for unlawful consummated transactions and that any doubts about remedy favor the government. It seemed particularly concerned about the future need for government monitoring or “entanglement in the market” absent divestiture, but it also cited the limited duration of conduct remedies, after which the merged firm is free to exercise whatever market power the merger provides, and problems in drafting a consent order in sufficient detail to cover all later possibilities.
What can we conclude about the Ninth Circuit’s St. Alphonsus opinion?
First, its result is probably correct. Although the court relied on market concentration in concluding that the FTC proved a prima facie case in a unilateral effects case, there was sufficient other evidence to show, all else being equal, that St. Luke’s was likely to profitably raise reimbursement levels for adult PCP services. St. Luke’s post-merger 80 percent market share, although not mentioned by the court, bolsters that conclusion.
Second, notwithstanding the court’s skepticism about the relevance of efficiencies in merger analysis, it is clear today that they are relevant and, in some circumstances, can trump the government’s prima facie case. The court was correct in pointing out the difficulty of balancing market-power and efficiency effects from the transaction, but this problem is no different than that in balancing procompetitive and anticompetitive effects in Sherman Act Section 1 rule-of-reason cases, and is certainly no excuse for avoiding the task completely.
Third, the court was incorrect in suggesting, if not holding, that quality improvements don’t count as efficiencies. Hospitals do compete on the basis of quality, and the benefits of this type of efficiency are clearly passed on to one type of hospital customer—its patients.
Fourth, the court was probably correct in holding that St. Luke’s’ claimed efficiencies were not merger-specific in the sense that they could be achieved, at least in large part, by means other than a merger. But the court should have recognized that a merger would maximize the likelihood of the achievement and effectiveness of the quality benefits and that this should be considered in subjectively balancing the merger’s procompetitive and anticompetitive effects. Its doing so, however, would likely not have changed the result in the case.
Fifth, the court was correct that divestiture was the appropriate remedy. But the question is a closer one in physician markets than in mergers involving other markets because there is always some possibility that the acquired group may not be able to reassemble in a competitively viable form after divestiture. The cure may be worse than the disease if, for example, the acquired physicians leave the market.
Sixth, the decision and other prior hospital-merger decisions show that the Affordable Care Act provides no explicit or implicit exemption from the antitrust laws. This is as it should be if one believes that increased market power leads to higher prices when a major goal of the Affordable Care Act is to decrease the cost of health care.
And seventh, problematic provider mergers of competitors face a rough road today compared to the period prior to the Evanston Northwestern Healthcare hospital-merger decision in 2007, and the legal advice of attorneys needs to reflect this fact.