Hospital merger cases are won or lost on geographic market definition. The Third Circuit’s recent finding that it is appropriate to define geographic markets based on patient location will likely incentivize the FTC to define such geographic markets more frequently in future hospital merger litigations. We consider the implications of defining a geographic market based on patient location and highlight a key shortcoming of this approach: since virtually any candidate geographic market based on patient location likely passes the Merger Guidelines’ Hypothetical Monopolist Test, any such conclusion is essentially meaningless and addresses an issue largely irrelevant to whether a proposed merger is likely anticompetitive. Consequently, the FTC’s reliance on patient-based markets could erode a key advantage that the FTC currently enjoys in hospital merger litigations: the courts’ willingness to endorse the Merger Guidelines’ presumption that mergers that sufficiently increase concentration are anticompetitive.
By Ken Field & Steven Tenn[1]
I. INTRODUCTION
It is axiomatic that hospital merger cases are won or lost on geographic market definition. Of course, many antitrust economists and lawyers will tell you that enforcement decisions turn on competitive effects analysis and defining the relevant geographic market is mostly an afterthought. But the government has been wildly successful in convincing the courts to adopt the Horizontal Merger Guidelines, so market definition is now step one in court and the Hypothetical Monopolist SSNIP Test and concentration presumptions are now entrenched law. Moreover, as the government and courts love to remind us, no merger that triggers those presumptions has ever been saved by claimed efficiencies.
Given the strength of the presumption and the government’s remarkable success relying on it, why would the Federal Trade Commission now risk undermining the Hypothetical Monopolist Test by defining a patient-based geographic market in recent litigation? Perhaps they are merely acknowledging the decreasing importance of market definition at the agencies but, more likely, they may be overcorrecting for their one loss in the past two decades and inadvertently risking their ability to succeed over the next two decades. But we are getting ahead of ourselves.
II. BACKGROUND ON GEOGRAPHIC MARKET DEFINITION IN HOSPITAL MERGERS
A. Merger Guidelines’ Hypothetical Monopolist Test
As the 2010 Horizontal Merger Guidelines (“Merger Guidelines”) explain, the Federal Trade Commission (“FTC”) and the Department of Justice’s (“DOJ’s”) Antitrust Division usually define geographic markets based on the Hypothetical Monopolist Test.[2] This is an iterative approach where one starts with a candidate geographic market and then tests whether a hypothetical monopolist of that candidate market would be able to profitably impose a small but significant non-transitory increase in price (“SSNIP”), often taken to be 5%. If so, then the candidate market is a relevant geographic market within which the proposed transaction can be analyzed. If not, one expands the candidate market and then repeats the analysis until the candidate market is sufficiently large to pass the test.[3]
A key motivation for using the Hypothetical Monopolist Test to define the relevant geographic market is that doing so results in a market that is sufficiently large that competition within it matters in the sense that, by construction, the elimination of all intra-market competition would result in at least a small but significant price increase (absent mitigating factors such as efficiencies and entry/repositioning).[4]
The FTC and DOJ typically define geographic markets based on supplier location when price discrimination based on customer location is not possible.[5] When price discrimination based on customer location is possible the agencies may instead delineate geographic markets based on customer location.
When analyzing hospital mergers, the FTC generally focuses on payers representing commercial patients.[6] For commercial patients, hospital prices are typically negotiated via bilateral bargaining between hospital systems and payers, who contract for provider networks on behalf of their members. It makes little sense to apply the Merger Guidelines’ discussion regarding price discrimination by customer location because payers are a hospital system’s customers and the location of a payer’s headquarters (or other physical location) is irrelevant as it does not impact competition. As an alternative, one can potentially use the locations where payers’ members reside, since payers consider the preferences of their members when negotiating contracts with providers.
Rarely, if ever, do negotiated prices between a hospital system and a payer vary depending on where an individual patient resides. Rather, the same negotiated price applies to all of a payer’s members covered by a negotiated contract. When a hospital system negotiates with a payer, both sides generally know where the payer’s members reside (although the hospital system’s information may be more limited than the payer’s). Consequently, while direct price discrimination based on member location is likely not possible in the hospital context, negotiated prices indirectly incorporate member location. Since negotiated prices reflect the aggregate preferences of all a payer’s members covered by a given contract, price discrimination is likely possible only at the payer-contract level rather than for a particular subgroup of those members.
B. Hackensack-Englewood Litigation
In hospital merger challenges, the FTC has most often delineated geographic markets based on hospital location.[7] In its latest hospital merger litigation, however, in which the FTC challenged Hackensack Meridian Health’s (“Hackensack’s”) proposed acquisition of Englewood Healthcare Foundation (“Englewood”), the FTC’s economic expert’s geographic market was based on patient location.[8] This geographic market included all hospitals that serve residents of Bergen County, New Jersey, where Englewood and two of Hackensack’s hospitals are located. During the litigation, the FTC claimed that a geographic market based on hospital location that included only Bergen County hospitals also passes the Hypothetical Monopolist Test, and that both hospital and patient-based approaches result in valid geographic markets under the Merger Guidelines.[9] As part of their decision granting the FTC’s request for a preliminary injunction, the district court judge concluded that commercially insured patients in Bergen County comprise a relevant geographic market for analyzing the proposed merger.[10] On March 22, 2022, the Third Circuit Court of Appeals concluded that the district court did not error in defining this patient-based geographic market.[11]
III. CONSIDERATIONS WHEN DELINEATING HOSPITAL OR PATIENT-BASED GEOGRAPHIC MARKETS
Relying on the Hypothetical Monopolist Test to delineate the geographic market is a meaningful exercise only to the extent this approach addresses a question relevant to assessing the competitive impact of a proposed merger. As explained below, this test addresses a very different question depending on whether the geographic market is based on hospital or patient location. The Hypothetical Monopolist Test applied to a geographic market based on hospital location addresses a question analogous to the competitive effects analysis of a proposed merger. In contrast, the Hypothetical Monopolist Test applied to a geographic market based on patient location addresses a question largely irrelevant to understanding the competitive impact of a proposed merger. We believe this distinction should be a key consideration when geographic markets are defined for proposed hospital mergers.
A. The Hypothetical Monopolist Test Applied to a Geographic Market Based on Hospital Location Addresses a Question Analogous to the Competitive Effects Analysis of a Proposed Merger
When the geographic market is based on hospital location, the Hypothetical Monopolist Test addresses whether a merger between the merging parties and other geographically proximate hospitals is likely to result in a meaningful price increase (absent mitigating factors such as efficiencies and entry/repositioning). That is, the Hypothetical Monopolist Test addresses the same issue as the competitive effects analysis described in the Merger Guidelines,[12] but for a larger merger that includes both the merging parties and additional competitors. Since a larger merger will result in greater anticompetitive effects than a smaller merger (all else equal), satisfying the Hypothetical Monopolist Test provides a necessary, but not sufficient, condition for the proposed merger to result in at least a small but significant price increase. Thus, the market definition exercise can be viewed as an intermediate step towards addressing the ultimate issue of whether a proposed merger is likely anticompetitive.
This relationship between market definition and the competitive effects analysis provides a rationale for the Merger Guidelines’ presumption that a merger that sufficiently increases concentration in a relevant antitrust market is likely anticompetitive, where the Merger Guidelines define a sufficient increase in concentration as a change in the Herfindahl-Hirschman Index (“HHI”) of market concentration of at least 200 points and a post-merger HHI of at least 2500.[13] If a geographic market based on hospital location satisfies the Hypothetical Monopolist Test then, by construction, sufficient consolidation between the hospitals located within the delineated market will result in meaningful anticompetitive effects (again, absent mitigating factors such as efficiencies and entry/expansion). This close relationship between the market definition and competitive effects analyses provides a logical basis for the Merger Guidelines’ structural presumption.
A necessary, but not sufficient, requirement for a hospital merger to result in meaningful anticompetitive effects is that patients face meaningful travel costs to be treated at more distant hospitals.[14] If this were not the case, and patient travel costs were sufficiently low, then it would always be possible for patients to find an attractive alternative provider since the United States contains numerous high-quality hospitals. Consequently, in this situation economic theory indicates that payers would respond to the preferences of their members and would choose to exclude the merging parties (or, analogously, the hypothetical monopolist of a candidate geographic market based on hospital location) from their provider networks rather than accept a small but significant price increase. While geographic proximity is not the only dimension of hospital differentiation that matters to patients, it is fundamental in the sense that meaningful anticompetitive effects from a hospital merger cannot arise in the absence of meaningful patient travel costs.
Patient willingness to incur travel costs is accounted for in the analysis when the Hypothetical Monopolist Test is used to define a geographic market based on hospital location. If patient travel costs are sufficiently low, and consequently hospitals outside of the candidate geographic market are viewed by payers (and their members) as relatively attractive alternatives to the merging parties, then economic theory indicates that a hypothetical monopolist of the hospitals within the candidate market would not be able to raise price by a small but significant amount.[15] In contrast, if patient travel costs are sufficiently high, then payers (and their members) will not view hospitals outside of the candidate geographic market as attractive alternatives and consequently a hypothetical monopolist of the hospitals within the candidate geographic market would be able to raise price by a small but significant amount. In this manner, the Hypothetical Monopolist Test applied to a candidate geographic market based on hospital location accounts for patient travel costs, a fundamental driver of whether anticompetitive effects from the proposed transaction are likely to occur.
B. The Hypothetical Monopolist Test Applied to a Geographic Market Based on Patient Location Addresses a Question Largely Irrelevant to Understanding the Competitive Impact of a Proposed Merger
The Hypothetical Monopolist Test addresses a very different question when market definition is based on patient location. The thought experiment is whether a hypothetical monopolist that is the only potential provider of hospital services for a given patient population would be able to increase price by a small but significant amount. For example, in Hackensack-Englewood the FTC’s economic expert considered patients residing in Bergen County, rather than the entire area from which the merging parties attracted patients, which included other portions of New Jersey and New York.
The assumption that the hypothetical monopolist is the only potential provider of hospital services to a target patient group is a remarkably broad proposition. It implies that the hypothetical monopolist not only owns the set of hospitals located close to where those patients reside, but also the next closest alternatives to those hospitals, and the next closest alternatives after that, and so on until the hypothetical monopolist owns all hospitals that the target patient group views as potential choices. Thus, this test involves a thought experiment involving a much higher degree of hospital consolidation than any real-world merger.
An implication of this is that patient travel costs are not a relevant consideration when the geographic market definition is based on patient location. As discussed above, the thought experiment in this formulation of the Hypothetical Monopolist Test involves a merger between every hospital that could potentially treat the target population at issue, regardless of where those hospitals are located. Consequently, patient travel costs are largely irrelevant because there are no other providers that patients could travel to as an alternative to being treated by the hypothetical monopolist.
Since the hypothetical monopolist is assumed to own all potential hospitals from which the target patient group may choose, those patients are left with only a single alternative to the hypothetical monopolist: the “no treatment” option of forgoing hospital services altogether. For patients considering certain elective procedures, it is conceivable (but not necessarily likely) that choosing not to be treated is a reasonable alternative to being treated by the hypothetical monopolist of hospital services. For many hospital services, including services involving life threatening situations, the “no treatment” option is not a viable option. Consequently, for virtually any patient population one might consider defining a geographic market around, it is likely the case that the payers of that population would rather accept a small but significant price increase rather than forgo hospital services for their members altogether.
This implication can be readily observed from the Willingness to Pay analysis that has been used in prior hospital merger litigations, most recently by the FTC’s economic expert in Hackensack-Englewood.[16] Willingness to Pay captures how the exclusion of one or more hospital systems from a payer’s provider network impacts the attractiveness of that provider network from the patients’ perspective (which the payer takes into account when it negotiates prices with providers). As typically estimated, Willingness to Pay to include a given hospital system becomes arbitrarily large as that hospital system’s share approaches 100 percent.[17] This has important implications for the Hypothetical Monopolist Test when applied to a candidate market based on patient location, since the hypothetical monopolist of that candidate market has, by construction, a 100% share.[18] Consequently, Willingness to Pay to include the hypothetical monopolist in the payer’s provider network is arbitrarily large, which according to economic theory would dramatically improve the bargaining position of the hypothetical monopolist when negotiating with a payer and allow it to increase price by at least a small but significant amount.[19] This implies that any candidate market based on patient location likely passes the hypothetical monopolist test, independent of the available data or the facts of the case.
In virtually any hospital merger, the key competitive issue is whether payers view other providers as being sufficiently close alternatives for the merging parties, not whether payers view the “no treatment” option as a sufficiently close alternative to their members being treated at a hospital. For this reason, applying the Hypothetical Monopolist Test to a candidate market based on patient location does not address a question relevant for understanding the competitive implications of a proposed merger.
This weak relationship between the question addressed by the Hypothetical Monopolist Test applied to a geographic market based on patient location and the question addressed in a competitive effects analysis of a proposed transaction provides little support for applying the Merger Guidelines’ structural presumption to such geographic markets. More generally, the logical implication of any candidate market based on patient location likely passing the Hypothetical Monopolist Test is that such a finding is essentially meaningless. This is a significant disadvantage of defining hospital geographic markets based on patient location, and potentially explains why the FTC has, historically, more frequently chosen to rely on geographic markets based on hospital location.
C. Patient Migration and Substitution to Outside the Market
When the Hypothetical Monopolist Test is used to define a geographic market based on hospital location, a candidate market that passes the test may not closely correspond to the geographic area of focus of the merging parties or other industry participants.[20] It is well recognized, and explicitly acknowledged by the Merger Guidelines, that antitrust markets may not correspond to how industry participants define “markets” in other applications.[21] Nonetheless, this may be perceived as a weakness of the approach, with the merging parties in hospital merger litigations often arguing that the FTC’s geographic market is gerrymandered or otherwise fails to consider practical realities.
In particular, applying the Hypothetical Monopolist Test to a candidate market based on hospital location often results in geographic markets that are not self-contained. That is, a meaningful fraction of patients living in the geographic market may receive care from hospitals outside the geographic market, and a meaningful number of patients living outside the geographic market may receive care from hospitals located within the geographic market. Relatedly, a significant fraction of patients whose first choice is a hospital in the defined market may have as their second choice a hospital located outside of the market. That is, hospitals outside of the defined market may be substitutes for hospitals inside the market. It is widely recognized that it is not necessary for a geographic market to be self-contained for it to pass the Hypothetical Monopolist Test.[22] Rather, it is only necessary for the set of hospitals located within a candidate market to be sufficiently close substitutes. Nonetheless, such markets may suffer from bad optics and, consequently, may be challenging for the FTC to defend in litigation.
This perceived limitation can potentially be avoided by defining markets based on patient location. By construction, there is no patient inflow or outflow from the geographic market because it includes the entire target patient population regardless of where they receive care. Similarly, there is no substitution to outside of the market because it includes all hospitals that treat the target patient population regardless of where those hospitals are located.
For example, in the Hackensack-Englewood matter the merging parties were located relatively near New York City, and a meaningful fraction of patients residing in Bergen County, New Jersey were treated at hospitals located in the state of New York or in other counties in New Jersey. Had the FTC’s economic expert defined a geographic market consisting of hospitals located in Bergen County, then that market would be subject to the critique that it excludes significant competitors outside of Bergen County. By defining a geographic market consisting of patients residing in Bergen County, the FTC’s economic expert avoided such criticism since the market includes all hospitals that treat patients residing in Bergen County, including hospitals located in New York or other counties in New Jersey. This approach allowed the FTC’s economic expert to appear to be more conservative even though, as discussed above, a geographic market based on any patient population is likely to pass the Hypothetical Monopolist Test.
D. Inferences from Market Shares and Concentration
Regardless of whether a geographic market is based on hospital or patient location, it is often useful to consider shares and concentration from both approaches since they measure different aspects of patient preferences. Shares based on hospital location speak to the available hospital options and choices of patients with a preference for receiving care in a given geographic area. Of course, since hospitals located outside of the geographic area are excluded, such shares do not speak to the hospital options and choices of those who prefer to receive care outside that geographic area. Similarly, shares based on patient location speak to the available hospital options and choices of patients residing in a given geographic area, but do not reflect the hospital options and choices of other patients since they are excluded from the share calculation.
To the extent that patients generally prefer to receive care close to where they live, then shares based on hospital location may be quite similar to those based on patient location. If that is the case, then calculating shares based on both approaches and demonstrating the similarity of shares can be quite informative. Conversely, if shares based on hospital location are meaningfully different from shares based on patient location, then it is likely the case that it will be important to address the underlying reasons for that result.
IV. HOSPITAL MERGER ENFORCEMENT GOING FORWARD
The outcome of the Hackensack-Englewood litigation may have a significant impact on the trajectory of the FTC’s hospital merger enforcement program. Specifically, the Third Circuit’s finding that it is appropriate to define geographic markets based on patient location will likely incentivize the FTC to define such geographic markets more frequently in future hospital merger litigations. While the FTC may find this beneficial in the short run, doing so could erode a key advantage that the FTC currently enjoys in hospital merger litigations: the courts’ willingness to endorse the Merger Guidelines’ presumption that mergers that sufficiently increase concentration are anticompetitive.
It has long been debated whether market definition should be deemphasized in merger analysis, and instead greater focus should be placed on the ultimate question of competitive effects. The revised 2010 Merger Guidelines can be viewed as a step in this direction, as they explicitly note that the antitrust agencies need not start their analysis with market definition and that competitive effects can be analyzed without first defining a relevant market.[23]
Nonetheless, the Merger Guidelines’ presumption that a merger is likely anticompetitive if it significantly increases concentration within a properly defined market continues to play a key role in hospital merger litigations. The Merger Guidelines’ presumption shifts the burden of showing that a proposed merger is likely not anticompetitive to the merging parties (rather than the FTC’s burden being to show that this is not the case). This burden shifting is a key reason why the FTC has suffered only a single loss that was not reversed on appeal since the agency rebooted its hospital merger enforcement program two decades ago.[24] Notably, the district court in the FTC’s sole loss found that it had failed to properly define the geographic markets at issue.[25]
By construction, a merger that sufficiently increases concentration in a geographic market based on hospital location, and which passes the Hypothetical Monopolist Test, would allow the combined firm to increase price by at least a small but significant amount (absent mitigating factors). However, this is not necessarily true for geographic markets based on patient location. Since virtually any candidate geographic market based on patient location likely passes the Hypothetical Monopolist Test, any such conclusion is essentially meaningless and addresses an issue largely irrelevant to whether a proposed merger is likely anticompetitive (for the reasons explained earlier in Section III).
This disconnect between market definition and the competitive effects analysis when a geographic market is based on patient location diminishes the value of the market definition exercise. Consequently, the courts may respond to the FTC’s reliance on geographic markets based on patient location by giving greater weight to competitive effects, and less weight to market definition and the Merger Guidelines’ presumption. If so, the FTC may find it more challenging to win hospital merger litigations going forward, since it would become its burden to demonstrate that a merger is likely anticompetitive (rather than the merging parties’ burden being to show that this is not the case). Thus, while hospital systems wishing to merge may view the FTC’s reliance on patient-based geographic markets as a negative development in the short run, since the FTC will have an easier time satisfying the Hypothetical Monopolist Test for such markets, this may ultimately be a pyrrhic victory for the FTC that reduces their ability to block hospital mergers over the longer term.
We conclude by noting that market definition generally plays a less prominent role in the FTC’s investigations of proposed mergers compared to hospital merger litigations. Consequently, the FTC’s potential undermining of the import of market definition and the Merger Guidelines’ presumption by defining patient-based geographic markets is likely to affect primarily whether it will be successful in merger litigations rather than FTC decisions regarding whether to take enforcement action. But, since the FTC takes litigation risk into account when making enforcement decisions, any lessening of its ability to win hospital merger litigations likely would eventually be internalized by the FTC and result in less aggressive FTC enforcement for hospital mergers going forward.
[1] Ken Field is an antitrust partner and co-chair of Jones Day’s global Health Care and Life Sciences Practice. Steven Tenn is a vice president in the Antitrust and Competition Practice at Charles River Associates. Both authors have significant experience working on healthcare mergers and other antitrust matters, including transactions involving hospitals. The views and opinions expressed herein are those of the authors and do not reflect or represent the views of Jones Day, Charles River Associates, or any of the clients or other organizations with which the authors are affiliated. Both authors worked on behalf of a hospital system that was involved as a third party in the FTC’s investigation and litigation of the Hackensack-Englewood matter discussed in this article. We thank Greg Vistnes for providing helpful comments.
[2] U.S. Department of Justice and Federal Trade Commission’s 2010 Horizontal Merger Guidelines (August 19, 2010).
[3] In this article, we focus on high level conceptual issues and do not address the specifics of how the Hypothetical Monopolist Test may be implemented. For such discission, see Steven Tenn & Sophia Vandergrift, “Geographic Market Definition in Urban Hospital Mergers: Lessons from the Advocate-NorthShore Litigation,” Antitrust Source (December 2017).
[4] Under the Merger Guidelines, such mitigating factors are analyzed separately from the market definition and competitive effects analyses. Merger Guidelines, § 9 and 10.
[5] Merger Guidelines, § 4.2.
[6] Most hospital mergers are investigated by the FTC rather than the DOJ.
[7] In Cabell-Huntington-St. Mary’s and Hackensack-Englewood, the FTC defined a geographic market based on patient location. In all other hospital merger challenges since 2015 the FTC defined a geographic market based on hospital location.
[8] United States District Court for the District of New Jersey Opinion, FTC v. Hackensack Meridian Health, Inc. and Englewood Healthcare Foundation (August 4, 2021) at 35.
[9] Answering Brief of the Federal Trade Commission, FTC v. Hackensack Meridian Health, Inc. and Englewood Healthcare Foundation (October 29, 2021) at 24-39.
[10] United States District Court for the District of New Jersey Opinion, FTC v. Hackensack Meridian Health, Inc. and Englewood Healthcare Foundation (August 4, 2021) at 44.
[11] United States Court of Appeals for the Third Circuit Opinion, FTC v. Hackensack Meridian Health, Inc. and Englewood Healthcare Foundation (March 22, 2022) at 20.
[12] Merger Guidelines, § 6.
[13] Merger Guidelines, § 5.3.
[14] Travel costs refer to any costs associated with a patient receiving care in a different location, including both direct time and monetary costs incurred as well as any patient preferences associated with receiving care in different locations.
[15] Intuitively, in the absence of travel costs one could treat all hospitals in the United States as being collocated, resulting in a plethora of hospital choices that would prevent meaningful anticompetitive effects from arising.
[16] Memorandum in Support of Federal Trade Commission’s Motion for a Preliminary Injunction, FTC v. Hackensack Meridian Health, Inc. and Englewood Healthcare Foundation (March 22, 2021) at 25-26.
[17] See, e.g. Cory Capps et. al., “Competition and Market Power in Option Demand Markets,” Rand Journal of Economics (2003).
[18] We assume that shares are measured only for hospital services and exclude the “no treatment option” from the share calculation. This is a common approach when considering hospital services.
[19] Steven Tenn, “Introduction to the Economic Analysis of Hospital Mergers,” Newsletter for the Economics Committee of the ABA Section of Antitrust Law (2019).
[20] Competition in hospital markets is often modelled in two stages. In the first stage payers and providers negotiate prices, and then in the second stage hospitals compete to attract patients. A hospital’s geographic area of focus may differ depending on which stage of competition it is considering.
[21] Merger Guidelines at 8.
[22] See e.g. Kenneth Elzinga & Anthony Swisher, “Limits of the Elzinga–Hogarty Test in Hospital Mergers: The Evanston Case,” International Journal of the Economics of Business (2011).
[23] Merger Guidelines at 7.
[24] The FTC abandoned its challenge to the merger between Jefferson Health and Albert Einstein Healthcare Network following an adverse district court decision. See, https://www.ftc.gov/legal-library/browse/cases-proceedings/181-0128-thomas-jefferson-university-matter.
[25] United States District Court for the Eastern District of Pennsylvania Opinion, FTC v. Thomas Jefferson University et al. (December 8, 2020) at 60-61.
