A Problem with Market Share

The Complaint alleges that the combined entities would have over 30% of the market, and that they are two of the three largest. ¶¶7, 29. This allegation sounds meaningful until you do the math.

At $2.2 billion and $2.3 billion in revenues, respectively, Amedisys and UnitedHealth would have approximately the same market share. ¶¶29-30. If combined Amedisys and UnitedHealth would have 30% market share, each would have around 15% if their revenues were even. Since they are the first and third largest competitors in the market, no other competitor would have more than 15%. To find the smallest number of competitors you can have based on those allegations, you divide the 70% remaining by 15%, and you get around 5 competitors of about the same size. If you have more than 5 competitors, the market becomes less concentrated. Worst case scenario, this is a 7-6 merger.

The main problem with the Complaint is that the allegations only go to how much UnitedHealth and Amedisys compete against each other. The Complaint alleges almost nothing about the other competitors. It could very well be that UnitedHealth competes vigorously against the other five competitors as well. It’s simply not plausible that because they compete aggressively against each other the merger will harm competition because the remaining competition, which is not addressed, could very well be sufficient. You can have atomistic markets where two competitors compete vigorously against each other for any number of reasons, but any attempted price increase by a merged entity would be defeated by the remaining competition.

And if the geographic market is in fact local, a large national footprint would not be necessary to provide services within that market, making it plausible that small local providers do in fact compete with the larger providers. If the allegation that “[m]any of Defendants’ smaller, local competitors lack the resources to invest in larger workforces and programs, such as local quality improvement coordinators, that create these advantages,” ¶39, is true, one would expect the market shares to be much higher and the geographic market much larger. Only larger competitors would be able to service these markets. The small geographic market allegation, ¶58, therefore contradicts the suggestion that large national providers only compete against other large national providers.

DOJ also acknowledges there is presently price competition in the market. “Amedisys, for example, acknowledges that rates with Medicare Advantage plans are ‘driven down by price competition.’” ¶42. The vast majority of hospice care is paid for by the United States government. It is implausible to allege the government does not have buyer power or could not resist the combined entity’s ability to raise price or lower quality in a market with at least 6 participants.

According to the Complaint, UnitedHealth believes “Amedisys does a lot of things that we do not do—if they get a foothold in [the] county, they will likely push us out.” ¶33. This allegation suggests the two are not close substitutes and that their products are differentiated reducing the chance the merger would harm competition.

A Problem with HSR

The HSR Act forbids companies from consummating transactions before they have filed notification and observed the applicable waiting periods. The Agencies can extend the initial waiting period by issuing a Second Request. When the parties are in “substantial compliance” they certify so to the Agencies which begins a second waiting period. The Agencies must sue under Section 7 to enjoin parties to an anticompetitive transaction. If there is no injunction in place when the second waiting period expires, the parties may consummate their transaction without violating HSR. You don’t have to have a reportable transaction under HSR to violate Section 7.

The Complaint alleges that Amedisys certified substantial compliance in December 2023. ¶14. The Complaint also alleges that Amedisys knew that it was not in substantial compliance but submitted the certification anyway, and so violated Section 7A.

Section 7A(g)(1) provides that if a person fails to comply with “any provision” of Section 7A they will be liable for the HSR fine. 7A(e)(2)(B) requires a statement of noncompliance if the parties have not substantially complied. So presumably the DOJ is arguing that by failing to provide the statement of noncompliance, Amedisys has failed to comply with a “provision” of Section 7A and is therefore subject to fines.

The HSR Act prohibits the consummation of a reportable transaction before the termination of the applicable waiting period. DOJ is suing to permanently enjoin the consummation of the transaction. ¶102(b). The transaction has therefore not been consummated. Moreover, if parties have failed to produce something required, under Section 7A(g)(2), the DOJ may seek an order from a US District Court ordering compliance, extending the waiting period or other equitable relief.

There is no affirmative duty to make a notification under the HSR Act. It simply provides parties may not consummate reportable transactions until they have notified the agencies and the appropriate waiting periods have expired. If there is no affirmative duty to file, there can be no affirmative duty to submit a notice of substantial compliance or to state the reasons for noncompliance. Parties just cannot consummate the transaction before they do so. And DOJ’s only remedy if it was harmed by the premature certification at all is a (g)(2) action to compel production of the statement of noncompliance (or the missing materials). Since Amedisys did not consummate the transaction, Amedisys did not violate the HSR Act by submitting a certification that proved to be inaccurate or failing to state reasons for noncompliance, and is therefore not subject to a fine.

Next Steps

The parties in Amgen/Horizon filed an answer that looked a lot like a motion to dismiss. And ultimately they came up with a consent that gave the parties the vast majority of their transaction. I suspect that the inclusion of the HSR count here was to make Amedisys look bad to the court. In reality, I think it makes the DOJ look bad because the DOJ actually does not state a claim upon which relief may be based. Moreover, because it doesn’t, the HSR Count serves as an invitation to move to dismiss the HSR Count and take a stab at the weakness of the Substantive Section 7 Count as being implausible. Motions to dismiss are very useful in antitrust suits that do not involve Section 7 because they offer the parties the opportunity to shape the court’s view of the counterarguments. In the past, parties have been unwilling to do so because courts granted the agencies a lot of leeway and because they were usually sufficiently plead. The lesson of Amgen/Horizon is, I think, that sometimes you should file a motion to dismiss a merger complaint.

The Complaint was also filed late in the Biden Administration. It's entirely possible that the next administration will dismiss it with prejudice. If I were the parties, I would still move to dismiss.