Revisiting Guidance on MFN Terms
In 2012, Deputy Assistant Attorney General Fiona Scott-Morton gave a speech that cautioned against “contracts that reference rivals,” including most-favored nation clauses (“MFNs”) and some forms of volume discounts. While recognizing that these kinds of contracts can generate efficiencies in some cases, she advised companies to avoid using them whenever possible because of their enhanced risk of raising antitrust concerns. That risk was pointedly illustrated by a series of cases brought by the Antitrust Division at this time that featured MFNs in a particularly bad light.
Chief among these cases was U.S. v. Blue Cross Blue Shield of Michigan (BCBSM), where the MFNs that BCBSM obtained from Michigan hospitals ensured not only the same price but, in some cases, a better price than any rival. Suddenly, MFNs had become a dangerous type of contract provision with no clear guidance on how to tell the good from the bad.
At about the same time, however, in U.S. v. Comcast Corp. et al. (settling issues with the Comcast/NBCU merger), the Antitrust Division had not only recognized that MFNs could be part of a pro-competitive arrangement, it essentially relied on an MFN concept for a major component of the settlement agreement: under the settlement, the merged entity would have to offer MFN-type arrangements to other content purchasers.
Indeed, the Antitrust Division’s use of an MFN in Comcast suggests that the advice to avoid these kinds of contract terms whenever possible is incomplete. Economists have established that MFNs can alternately promote or restrain competition and can alternatively enhance or detract from market efficiency. So, the question is whether economists can also offer any practical guidelines on whether an MFN in a particular case is likely to be pro- or anticompetitive — or, more to the point, how an MFN is likely to be viewed if it cannot simply be avoided altogether.
We believe that there are some appropriate rules of thumb for when MFNs are likely to be pro- or anticompetitive. Every negotiation will have its own nuances, but we suggest that a good starting point is to ask where the incentive to offer a better price is expected to come from and what source of power makes the MFN an effective constraint against following such incentives. To illustrate how the answers to those questions are likely to affect antitrust risk, we will re-examine these MFN cases after trying to generalize some of the results in the economics literature.
II.ECONOMIC THEORY AND AMBIGUITY
The academic literature on MFNs is, like the literature on most economic topics, helpful but incomplete from a practical line-drawing perspective. Each paper shows that, under its own specific assumptions about market structure, MFNs become more or less likely to result in higher or lower prices, to achieve particular outcomes, or to increase or reduce the number of competing firms. The literature does not attempt to completely map all possible situations nor to establish a definitive bright line between the positive and negative aspects of an MFN. And, to be clear, we are not going to attempt that either. But, we do hope that by illustrating what these theories have in common and where they differ, we can suggest a framework that may help distinguish the facts in specific cases.
A.A Simplified Anticompetitive Story
The simple story of an MFN’s anticompetitive effect is that it makes one party hesitant to negotiate lower prices or better quality with later trading partners. In the typical case, a seller agrees with a first buyer (usually a distributor or retailer) not to offer any better terms to additional buyers. Lower prices and higher quality are generally a good thing, so presumably, an MFN must be bad. In this simple story, we can determine how bad by looking at the price or quality improvements that were rejected. Or, perhaps we layer on some more complications for that calculation. What about the price and quality terms that never happen because the second buyer knows the MFN exists and so never proposes them? Or, to continue this train of thought, what if potential competitors never even attempt to enter the retail market or to explore better ways to distribute the product because it is not worth the investment to become a competitor?
In each of these examples, the seller commits through the MFN to protect the first buyer from future competition. It is obvious why that is good for that first buyer, but presumably, the seller would be better off with more competition among buyers. So, a natural question is why the seller would sign away the prospect of a better future for itself. Typically, the explanation is that the seller needs or values a deal today enough to make this choice rational. That is, the first buyer has enough market power that it can choose to spend some of that power today on extending it into the future. Or, to generalize, one of the hallmarks of the anticompetitive story is that it depends on current market power that is at some risk of being lost.
B.A More Complicated Pro-competitive Story
The pro-competitive story of an MFN is more complicated as it focuses on what the inability to have an MFN would do to negotiations between the original parties. To see this side, it is necessary to think of the MFN less as a constraint on what future buyers might do as rivals and more as a constraint on what the seller would do once the first buyer is locked into a contract. If it is reasonable to fear that a seller will change what it is willing to do for later buyers once it has that first contract, no sane buyer would want to be the first to sign. Potentially, the seller cannot get any buyers to sign and cannot get its product to market at all. More likely, it has to give every buyer terms that assume the seller will operate in the worst possible way towards all of them. In either scenario, an MFN allows the seller to establish from the outset a commitment that defines what the terms are for all buyers — essentially defining the product it wants to offer and not letting the fact of multiple buyers over time define the product down to a lowest common denominator.
As an example, consider the licensing of video content. An important feature of this kind of product is that it is “non-rivalrous” — if a content producer sells it to one person (or one distributor), it can just as easily sell it to another person (or another distributor). So, a would-be distributor of such a product would naturally want some assurance from the producer that it will not act on that ability after the distributor’s check is cashed. Perhaps an exclusive or a territorial exclusive works for both parties in this regard and they can move ahead with negotiations. But, perhaps the producer wants to be able to reach distinct but overlapping audiences of customers represented by two or more distributors.
In that case, each would-be distributor may be particularly worried that a later-contracting distributor will offer the seller a share not only of the profits from distributing to that distributor’s customers, but also of the profits that could be made again within the first distributor’s customers. In this scenario, the seller can potentially gain by giving a later distributor a lower price or other terms that would let that distributor offer a better product and win customers away from the first. The producer may not want to have that change in its incentives kick in whenever it has signed a contract, but as long as it does, all the contracts have to reflect that a distributor will not get what it thinks the producer is offering once these later contracts happen. An MFN helps the seller commit to offer all the distributors a common set of terms and avoid this degradation of its product.
Admittedly, a defense of MFN that equates it to an exclusive may appear to just connect one set of confusing legal concepts to another. In this example, though, the exclusives essentially mirror the underlying IP rights that a producer would, presumably, have over its content. Extending that thought, the MFN is essentially maintaining the IP rights’ positive incentives to create something new while allowing that new product to be used by rivals instead of through a monopoly that would have been less efficient.
C.Takeaways from the Two Stories
The positive side of an MFN, then, is its ability to assure every buyer there will be a uniform price and common terms. That assurance is only relevant, though, because of a time-inconsistency in the seller’s behavior. The seller would like to assure the first buyer that it will not degrade the value of their contract in the course of negotiating with future buyers, but all parties predict that without such assurances the seller will do exactly that. Recall, however, that it was the unwillingness to engage in a price reduction for later-arriving buyers that was our simple example of a bad effect. That same mechanism is now the way an MFN becomes good. Is an insistence on uniformity of terms good or bad?
We do not usually have a problem with uniformity. Nobody challenges a grocery store for posting prices on its shelves that apply to everyone. There is not a hue and cry over the fact that this implies it will not entertain haggling. Indeed, the key part of most competition models is the idea that a seller will choose to forego some sales it could have made at a lower price because it wants to make profits on customers who will buy at a higher price — a problem that only exists because the higher or lower price is assumed to apply to everyone. Those models may generate a situation that warrants a challenge in some cases, but the culprit is not generally identified as the practice of setting uniform prices.
So, in distinguishing MFN effects, we need to take this core mechanism as a given. The refusal to accept some future potential deals is just what makes an MFN relevant; it is not in and of itself particularly helpful in determining if the term represents a restraint on competition. To decide what MFN provisions negotiators should avoid, we need to focus on why certain deals are being discouraged and how that effect is being ensured (or who is gaining from it). If the MFN is a product of one firm’s current market power and the deals it precludes are expected to become desirable primarily as a result of that market power weakening, it seems reasonably evident that the MFN likely is a restraint on competition. Conversely, if all of the buyers benefit from the term — recognizing that they may only be equal in an abstract sense of considering them before their order of negotiations is determined — it seems more likely that the enforcement power comes from that commonality of interest rather than some restraint on competition.
The key questions are whether market power led to the MFN and whether it would have come into play without that market power. That is not intended to capture every facet of economic theory in this area. In specific cases, the core mechanics of an MFN may be important or the details of which transactions became possible or were dissuaded may be relevant. And, in some cases, an MFN that originally seemed completely reasonable could become an inefficient drag on an industry’s ability to adapt as the world changes in unforeseen ways. But, in the interest of abstracting enough to have a framework for general guidance, it seems that these are the key questions.
III.CONTEXT FROM CASES
Taken together, the recent cases suggest a more nuanced view of when contracts can pro-competitively reference rivals. The MFN provisions that have been most problematic are ones that are part of an overall contractual arrangement to address the potential that rivals might in their own right introduce a more attractive alternative to the contract at hand. And, the power behind enforcing these terms is the current or expected market power of the firm facing these rivals. In contrast, MFNs that address the potential for differences to arise from time-inconsistency or from general changes in an industry over time seem to be accepted as pro-competitive.
A.U.S. v. Blue Cross Blue Shield of Michigan
The case filed against BCBSM in 2010 presents the most obvious example of how an MFN might be constructed to insulate a firm with substantial market power against the possibility that its rivals build their own market presence and become significant rivals. BCBSM was, and remains, the largest health insurer in the state of Michigan and the largest purchaser of healthcare services in Michigan. BCBSM had entered into contracts with Michigan hospitals that included MFN clauses, entitling BCBSM to prices equal to the lowest price agreed to with any of BCBSM’s competitors, and, in some cases, MFN-plus clauses, entitling BCBSM to prices that were less than the lowest price agreed to with any of BCBSM’s competitors by some agreed-upon amount. Thus, for some hospitals, BCBSM could be assured that its competitors would pay a significantly higher price for the same services. The price gap in this form of MFN is particularly difficult to construe as promoting any form of efficiency. After all, the MFN-plus arrangement specifically prevented uniformity of prices and other conditions. The federal and state prosecutors in the case seem to have placed it squarely on the protection-of-market-power side of the test.
B.Marshfield and Delta Dental
Other cases in health care, involving straight MFNs, have found a more mixed reception in the courts, and are more difficult to parse. For example, in Blue Cross and Blue Shield United of Wisconsin v. Marshfield Clinic, Marshfield had an arrangement with its affiliated physicians in which Marshfield would pay physician affiliates no more than those physicians charge to any other payer. The court determined that Marshfield Clinic’s MFN agreement with physicians was pro-competitive because it indicated the clinic’s attempt to get the lowest prices possible. However, in U.S. v. Delta Dental of Rhode Island, a case that settled before trial, the government argued that an MFN very similar in form to the Marshfield MFN should be considered anticompetitive. Delta Dental’s MFN stated that participating dentists had to give Delta the best rate they gave any other payer. The DOJ’s complaint emphasized that Delta’s MFN term blocked entry by dental plans that sought to put together limited networks of dentists willing to offer lower-priced services for some part of their business in exchange for higher volume — that is, it preserved Delta’s position in the dental insurance market.
What would tip the balance on these cases? Size could certainly play a role: Delta was the dominant insurer in its region, while Marshfield was a considerably smaller player. One implication of size may be that Delta had market power which Marshfield did not. But, size can also speak to the credibility of a reasonable purpose for the MFN. A need for parity with Delta is a very hypothetical exercise. It is hard to imagine that the contract was influenced by the prospect of another insurer equivalent to Delta emerging. In contrast, Marshfield could reasonably anticipate competition with other narrow network plans.
C.U.S. v. Comcast Corp. et al.
The Comcast/NBCU merger stands in marked contrast to these cases. In 2011, after an extensive review of the video industry, the DOJ reached a settlement allowing the merger to close. While asserting some potential for abuse, the settlement explicitly recognized that some exclusive and MFN-style terms were common in the industry and represented a pro-competitive arrangement. Section V.C.3 of the Comcast/NBCU decree expressly allows the merged firm to obtain and enforce such MFNs as long as they are not at odds with the rest of the decree.
In the accompanying Competitive Impact Statement (“CIS”), the government explained that exclusive distribution windows can be important in this industry both to get the best value and, therefore, most incentive to create new content and also as a potentially pro-competitive boost for a new competitor. The CIS discussion assumes an exclusivity term for its exposition. But, as we pointed out above in outlining this kind of situation, the two concepts are linked in this regard. An MFN allows the producer to work with multiple distributors without losing the kind of beneficial incentives to develop content in the first place that are usually ensured by exclusivity.
Another aspect of the Comcast/NBCU decree also points out a different innocuous reason for MFN terms. Hypothesizing that Comcast might have different incentives vis à vis future online distributors than GE had as the licensor of NBCU content, the decree gives these future entrants the right to impose a form of MFN on Comcast. Specifically, if the entrant could obtain carriage agreements with specific content identified as a peer group to the NBCU content, then the merged firm has to offer its own content on no less favorable terms to the entrant. 
The problem that made this style of MFN desirable in the Comcast/NBCU decree is that renegotiation to reflect unfolding factual developments is practically impossible in the context of an agreement embodied in a judicial decree. Absent the MFN-like provision, it may not have been possible to establish an agreement before the uncertainty of the future was resolved.
IV.UNDERSTANDING MFNS AS A REGULATORY TOOL
The regulatory use of an MFN-like provision carries its own risks, one of which can be seen in another prominent example — the Medicaid Drug Rebate Program. This program was designed to rein in state and federal spending on prescription drugs. It conditions coverage of a manufacturer’s drugs on an agreement to pay rebates to the states which effectively lower the price of the drug to either the best price per unit or to a fixed percentage below the average unit price. While this program did address the historic difficulties of state Medicaid programs negotiating prices with pharmaceutical manufacturers and effectively allowed them to free-ride on the negotiations of other payers, it may have inadvertently raised the prices overall. Economic studies of the Medicaid MFN have suggested that it did raise prices.
The regulatory use of MFN-like provisions illustrates our point that MFNs can resolve important difficulties with negotiation. When renegotiation is costly, but long-term contracts are desirable, MFNs may make contracts possible that would otherwise require exclusive arrangements, or would simply not occur. For example, MFNs have been common in natural gas contracts, where producers and pipeline owners typically commit to contracts with very long terms. In these markets, a typical MFN would guarantee that a pipeline operator will pay the best price in a region to a wellhead producer. The economic literature’s discussion of efficiency gains from such an arrangement illustrates why private parties rely on MFN terms in this context rather than renegotiation. Pipeline operators must make costly long-term investments in capacity in order to carry gas, while wellhead producers must also make long-term investments. These investments make long-term contracts desirable lest either side be subject to a hold-up of its investments. However, neither side wants to be locked into a contract that pays less than the market rate, which will likely vary over time as market conditions fluctuate in unpredictable ways. MFNs allow pipeline operators and producers to address this uncertainty of future market conditions within the long-term contract that they both desire. In this sense, the parties are both using an MFN to regulate their trading partner (prevent hold-up of their investments) and the term is generally innocuous unless it has the kind of wider-market distorting effect seen in the Medicaid example.
The Medicaid example should stand as an important caveat to the potentially pro-competitive uses of MFNs. If the set of transactions to which an MFN links prices is not large enough and independent enough from the set being indexed, the MFN can lead to negative outcomes for all market participants. While this is generally thought of as a regulatory problem, largely because renegotiation in commercial contexts is generally assumed to be easier, there are also commercial situations that can use MFNs in this way and, potentially, present this same issue.
All of this goes to illustrate that the government’s approach to MFNs in Comcast/NBCU represents a delicate balance between some positive and potentially negative effects of the terms. This is also reflected in the consent decree itself as a number of safeguards are built in to ensure that the usual and customary contract terms do not become a vehicle for accomplishing indirectly the restraint of new distribution rivals that the government had tried to prevent with its own form of MFN.
Taken together, the cases suggest that the government appreciates the pro-competitive possibility of MFNs. And, the pattern is, hopefully, somewhat understandable once you look beyond the mechanics of an MFN as a commitment not to offer somebody outside the current negotiation a better price.
For companies trying to make practical decisions, the key questions about the incentives and purposes embodied in specific MFN contracts suggest negotiators examine their own motives. If the point of seeking an MFN is to ensure parity and that what is being agreed to today will not be undermined tomorrow, that is more likely to be acceptable. If the point is to ensure an advantage relative to rivals or to firms that have not yet entered, it is more likely to pose a problem. And, if the MFN term is common in an industry because every firm has the same sort of interest in parity, it probably derives its enforcement power from a much more benign source than if it is enforced by the risk of disrupting terms to one particularly significant market participant. It is also worth keeping in mind the possibility that an MFN can create an issue unintentionally. Negotiators may want to take care to ensure that they can revisit the terms periodically, if that can be done without risking the commitment value.
The views expressed in this article are solely those of the authors, who are responsible for the content, and do not necessarily represent the views of Cornerstone Research. The article was first published by Competition Policy International.
 809 F.Supp.2d 665 (2011).
 808 F.Supp.2d 145 (D.D.C. 2011).
 The incentive to lower price can be a bit confusing here because price determines both the share of profit for the seller and the cost (or competitiveness) of the buyer. It works because the same profit is being “shared” multiple times. But, to make that clearer, imagine this is a market where the seller would like to charge both a franchise fee and a per-unit wholesale price. After collecting the first buyer’s franchise fee, the seller would like to maximize the franchise value (and the associated franchise fee) of a second buyer. It can do so by setting a lower wholesale price so the second buyer undercuts the first and wins all the customers that they both reach. Of course, then the whole process could be repeated with a third buyer undercutting the second and so on.
 809 F.Supp.2d 665 (2011).
 65 F.3d 1406 (7th Cir. 1995).
 943 F. Supp. 172 (R.I. 1996).
 808 F.Supp.2d 145 (D.D.C. 2011).
 It may be interesting to note that the settlement between the government and Comcast reverses the usual form of an MFN. It is a commitment on Comcast’s part to meet any terms an entrant can obtain with certain other providers rather than a commitment to obtain the same terms from all providers or to offer all entrants the same terms. Each of these variations is a commitment contingent on the strategic choices of rivals. The set of rivals (and the firm making the commitment) may differ but, under the right circumstances, any of these variations of contracts that reference rivals could generate the same ambiguity of effects for the relevant set of rivals as we have presented for an MFN.