I struggle to understand health care antitrust issues. It’s not for lack of trying. It’s just that I find much writing by lawyers to be difficult to parse. Not so the 2010 Boston University Law Review article by Erica Rice titled “Evanston’s Legacy: A Prescription for Addressing Two-Stage Competition in Hospital Merger Antitrust Analysis.” Even with a title like that, it is very readable. For those interested in antitrust, I recommend reading the whole thing.
Back-story: The 2007 case of the Evanston Northwestern Healthcare Corporation (ENH) reversed government’s ten-year-plus losing streak in challenging hospital mergers. The case was notable for several other reasons. It was a rare post-merger challenge, coming years after the year 2000 consummation of ENH’s merger with Lakeland Health Services in Northern Illinois. More importantly for Rice’s analysis, the court focused on “first-stage” competition–that between hospitals and managed care organizations–and not “second stage competition”–in which hospitals compete for patients. That is, the ENH case was adjudicated in the context of the balance of power between hospitals and insurers, a focus of this blog over the past year.
Rice’s article is in three parts. I’ll mostly focus on Parts I and II, which provide an overview of antitrust enforcement and reviews some basic concepts in analysis of anticompetitive effects of hospital mergers.
With respect to mergers, the Agencies’ [FTC’s and DOJ’s] antitrust work generally occurs before a merger is consummated. [… F]irms of sufficient size wishing to merge must notify the Agencies of their intentions. […] [Over the next] thirty days, the Agencies will decide whether to challenge the merger as anticompetitive. […]
If the Agencies determine, according to the [1992 Horizontal] Merger Guidelines, that the merger would lead to anticompetitive effects, they can go to a federal court and request a preliminary injunction. Often the threat of a preliminary injunction will cause the parties to abandon the proposed merger [… or if] the court grants the injunction, the parties usually abandon the merger plans. […]
[T]he  Health Care Guidelines defined a “safety zone” for acute care hospital mergers that will generally not be challenged under antitrust laws. The safety zone applies if one of the hospitals involved in the merger has an average of fewer than one hundred beds and fewer than forty inpatients at a time over the past three years. The safety zone does not apply if that hospital is less than five years old.
If a merger falls outside the safety zone, it will be evaluated under the five analytical steps in the Merger Guidelines, which are applicable to all industries. The five steps are: 1) market definition, measurement and concentration; 2) the potential adverse competitive effects of mergers; 3) entry analysis; 4) efficiencies; and 5) failure and exiting assets.
These five steps provide, roughly, an outline for Part II of Rice’s article. Below I focus on market definitions and some of the arguments used to rebut the presumption of anticompetitive effects, those based on entry, efficiencies, and failing firms.
Market Definition(s). The first step of analysis toward a determination of anticompetitive effects is to define the market. That has two parts, definition of the product market and definition of the geographic market. What is the product market for a hospital? Is it all inpatient and outpatient services? Or are those two separate markets? In FTC v. Butterworth Health Corp., the court ruled that the choice between inpatient and outpatient care is made on the basis of “medical judgement” and not cost. Therefore, they are not substitutes and do not belong to the same product market. A related product market question is whether anchor (or “must have” or “star”) hospitals really compete in the same market as other hospitals.
I’ve discussed ideas pertaining to geographic market definition in two prior posts, so I’ll skip it here. I’ve also written extensively on the market concentration (mis)measurement used by DOJ/FTC, Herfindahl index or Herfindahl-Hirschman index (HHI), so I’ll skip that too.
Rebuttals. A merger that causes an HHI increase of more than 100 points in a highly concentrated market (HHI above 2500, according to revised Horizontal Merger Guidelines) is presumed anticompetitive. Firms can rebut this presumption in several ways. If there are no significant barriers to entry than the high HHI will draw competitors. Thus, the merger is unlikely to have anticompetitive effects. The hospital industry has large regulatory barriers to entry so this is not a reasonable defense in that industry.
A second defense hinges on efficiencies. Rice writes,
A merger is not anticompetitive if it exploits efficiencies and passes the benefits on to consumers. […] Over time, courts have become more willing to accept hospitals’ claims that savings and efficiencies will lead to benefits for consumers that will far outweigh any anticompetitive effects of a merger. [… E]fficiencies defenses have been hugely successful for hospitals, allowing them to realize a seven-case winning streak against the government in hospital merger antitrust enforcement actions.
A final defense is that one of the firms involved in a merger would otherwise fail (and meets other criteria).
The remainder of Rice’s article focuses on the ENH case specifically and draws lessons in several areas. In particular, she notes the need for a focus on how hospital prices are set,
[T]he ALJ [administrative law judge] in Evanston explicitly differentiated between the two stages of competition, identifying the “relationship between hospital and managed care organizations” as “first stage competition” and the “relationship between patients and hospitals” as “second stage competition.” Importantly, the ALJ recognized that competition at the second stage is not based on price because prices are set at the first stage when hospitals contract with MCOs. Thus, hospitals compete to attract patients through other nonprice factors. Because the underlying question in any antitrust merger analysis is whether the merged entity will be able to exploit market power to raise prices above the competitive level, the ALJ rightly determined that the “critical concern” was the merger’s effects on first stage competition.
The problem with ignoring the complexity of hospital competition is that antitrust analysis differs depending on which stage of competition the analysis is based. For example, “both product and geographic markets may differ between the first and second stages of competition . . . , and the effect of a hospital merger, may differ across the two stages.” Clearly, which stage is used can greatly affect the outcome. A reduction in competition at one stage does not necessarily imply a reduction at the other. Going forward, courts should clearly explain who the relevant consumer is and at which stage potential anticompetitive effects are being evaluated.
To me, the careful distinction between the implications of competition between hospitals and insurers for those entities versus for consumers is critical. In particular, one cannot conclude consumer harm from a concentrated hospital market or a concentrated insurer market. As I’ve written many times, it’s the balance of power between the two that is relevant (as in the figure in this post). Moreover, as Rice points out, market definitions may differ in the two stages, further challenging our ability to draw inferences about one market from characteristics of the other.
In cases where illegal market power is found, guidelines for allocating damages among the entities at various levels of competition (direct and indirect payers) have yet to be established. Now throw in the fact that the health care market landscape will shift as consumers pay more for care (consumer directed health plans) and hospitals and their associated integrated delivery system are at higher risk for costs and have even high market power (ACOs), and antitrust regulators and courts have a very complex market to parse.
If you’ve read this much, you’ll likely enjoy reading Rice’s full article. It’s worth it, and I left out a lot.