In one sense the decision of the Third Circuit in Eisai v Sanofi-Aventis on 4 May was fairly unremarkable. Like those before it in Concord Boat, Allied Orthopedic, Southeast Missouri Hospital and others, the court rejected Eisai’s allegation that Sanofi’s loyalty discounts were an anticompetitive de facto exclusive dealing arrangement. Commentary immediately following the decision tended to see the case as another nail in the coffin of loyalty discount liability. Yes, the court’s finding that Sanofi’s discounts didn’t foreclose competition was unequivocal. But there was enough in the decision to suggest that the door remains open to liability in future cases with stronger facts.
The conduct at the centre of the case involved a loyalty discount programme for Sanofi’s Lovenox, the market leading anticoagulant drug used to treat deep vein thrombosis. Under the programme, Sanofi offered hospitals discounts based on the amount of the drug they bought. When a hospital bought below 75% of its total purchases of anticoagulants, it received a flat 1% discount. Above 75%, the hospital would receive increasingly higher discounts – from 9% to 30% – depending on the share of its purchases with Sanofi. However, the hospitals were not contractually obliged to purchase from Sanofi and there was no suggestion that Sanofi would terminate supply if a hospital didn’t engage with the programme.
Eisai, which had been trying to upset Sanofi’s market leading position with its own anticoagulant, Fragmin, complained that the loyalty discounts had foreclosed competition for hospitals involved with Sanofi’s programme.
Key to Eisai’s claim was that Sanofi’s loyalty discounts had the effect of bundling each customer’s contestable demand for Lovenox (the units that the customer is willing to switch to rival products) with the customer’s incontestable demand for Lovenox (the units that the customer is less willing to switch to rival products). Hospitals were said to have needed at least some volume of Lovenox because, unlike other anticoagulants on the market, it was also certified to treat severe forms of heart attack. In order to get higher discounts on their incontestable demand for Lovenox, customers had to also buy Lovenox to meet their contestable demand. This bundling mechanism of foreclosure was argued to be the same as that which led to liability in LePage’s v 3M, except there an ‘incontestable’ product was bundled with a ‘contestable’ product.
Eisai argued that the bundling meant that hospitals had no choice but to buy Lovenox to meet their contestable demand because they couldn’t afford to lose the higher discounts on the incontestable demand. This absence of meaningful choice to switch contestable demand to a rival’s product was said to be the same as that which was central to the liability findings in Dentsply and ZF Meritor. In those cases, the defendants were found to have threatened to terminate supply if customers didn’t meet all (or most) of their demand with the defendants’ products. Because the defendants’ products were a ‘must have’ for at least some of customers’ demand, customers had no choice but to exclusively deal with the defendants.
The Third Circuit didn’t see it the same way, saying that “the threat of a lost discount is a far cry from the anticompetitive conduct at issue in ZF Meritor and Dentsply”. “Lovenox customers did not risk penalties or supply shortages for terminating the Lovenox Program or violating its terms”. Accordingly, “Lovenox customers had the ability to switch to competing products. They simply chose not to do so”.
Having failed to prove foreclosure of hospital customers, Eisai’s challenge failed.
In this case, the Third Circuit saw an important conceptual difference between (1) arrangements where continuing supply of a product for which a customer has incontestable demand is conditional on the customer also meeting its contestable demand with that product; and (2) arrangements where getting higher discounts on a product for which a customer has incontestable demand is conditional on the customer also meeting its contestable demand with that product.
The decision showed scepticism that the latter arrangement could deprive customers of meaningful choice to switch its contestable demand to a rival’s product. However, despite its scepticism, the court clearly limited its reasoning to the facts of the case, rather than making a general statement that loyalty discounts could never foreclose competition. For example, after discussing the foreclosure evidence before it, the court concluded that it “cannot credit Eisai’s bundling claims, at least not on the facts before us” [emphasis added]. On the other hand, there were some statements that indicate that it saw the logic in the bundling theory advanced by Eisai. The court remarked that Eisai’s expert “certainly explains why, in theory, a customer might hesitate to switch from Lovenox to one of its lower priced competitors”, but held that the facts didn’t support the theory in this case. And in an earlier statement the court found that “[e]ven if bundling of different types of demand for the same product could, in the abstract, foreclose competition, nothing in the record indicates that an equally efficient competitor was unable to compete with Sanofi” [emphasis added].
Furthermore, the Third Circuit rejected Sanofi’s argument that Eisai’s suit must be dismissed under a standard predatory pricing price-cost test, as it was “not persuaded that Eisai’s claims fundamentally relate to pricing practices”. This part of the decision is significant because of the almost insurmountable challenges a plaintiff would face in trying to prove that a loyalty discount arrangement resulted in predatory pricing.
So, while the Third Circuit wasn’t ultimately prepared to find that Sanofi’s discount programme foreclosed competition in this case, what it said on the way to reaching that conclusion suggests liability for anticompetitive loyalty discounts remains a possibility in future cases with stronger facts.
I would argue that the facts don’t need to be much stronger to find foreclosure. The Third Circuit’s claim that the threat of a lost discount is a “far cry” from the threat of termination of supply that led to liability in Dentsply and ZF Meritor seems to be an overstatement, and one which fails to pick up on the potentially similar economic consequences of the two types of conduct. When a customer is faced with the threat of losing supply of a ‘must have’ product unless it agrees to satisfy its contestable demand with that product, the economically rational choice is to secure its survival by acquiescing. In certain circumstances, the same may be true where a customer is faced with the loss of a large discount on a ‘must have’ product unless it agrees to satisfy its contestable demand with that product. If that customer competes in a downstream market where differences in input costs are critical to survival, and its rivals are receiving a large discount under a loyalty programme, then the customer really will have no choice but to also participate in the loyalty programme.
The case that finally makes it through the door that the Third Circuit has left open will likely be one where there is strong evidence that customers wanted to switch some of their demand to a rival’s product but couldn’t because they risked survival by doing so. In Eisai v Sanofi-Aventis, evidence that a few dozen hospitals (out of 6000 in the US) wanted to switch some demand but couldn’t wasn’t strong enough for the court.
Despite the outcome in Eisai v Sanofi-Aventis, the Third Circuit’s decision suggests we should keep watching out for the case that breaks new ground in the antitrust treatment of loyalty discounts.
Richard Davidson is Assistant Director at the United Kingdom Competition and Markets Authority. The views in this article are his own and do not reflect those of the Competition and Markets Authority.