It is clear from the economics in the government’s complaint against the AT&T-Time Warner merger that the harms to competition articulated by the Department of Justice cannot be remedied to any significant degree with behavioral commitments on the part of the merging parties.



Photo by Bill Bradford, via Flickr [CC BY 2.0]

Assistant Attorney General Makan Delrahim, head of the Antitrust Division of the Department of Justice, has publicly taken a position against behavioral remedies in antitrust cases. He argues that behavioral remedies require ongoing oversight of an industry, including in fast-moving areas like pricing, that the Division is not structurally organized to perform. Indeed, many would argue in addition that the antitrust laws, by virtue of being law enforcement, are an explicit alternative to regulation. If the Division is not very effective in policing a consent decree over time, it will not be protecting consumers from anticompetitive conduct. AAG Delrahim argues that rather than using ineffective remedies, the Division should either require a structural solution, such as an asset divestiture, to cure an otherwise anticompetitive merger or simply block the transaction.


There are aspects of this position that I agree with and admire. It is difficult for the Department of Justice to police behavioral remedies over time, especially in complicated and changing markets. The agency does not have the expertise, procedures, or personnel needed to be a good regulator, and the antitrust laws were not designed to create a regulator. Moreover, the division is fully occupied analyzing current transactions. The mechanism for seeking change or punishment, should the environment shift or the parties fail to comply with the consent decree, is relatively slow and costly. Small, harmed competitors do not have the time or capital to defend themselves through that process. The argument that it is cleaner to simply require divestitures or block anticompetitive mergers makes a lot of sense.


This position is even more sensible in the context of the AT&T-Time Warner merger and the government’s suit to block it. Consider the concerns laid out in the government’s complaint. One major concern is that the combined firm will raise the price of its valuable content (or refuse to sell it at all) to competitors. Other concerns are that the combined firm would withhold content from online video distributors (OVDs) and establish contracts with large vertically integrated rivals that lessen price competition. Some commentators have argued that these harms can be remedied with limitations on the behavior of the merged firm. Indeed, AT&T has put forward a proposal along these lines that would bind the combined firm to offer rival distributors baseball-style arbitration in the event of a disagreement over fair market value of content. AT&T points to the behavioral remedies imposed by the division on Comcast’s merger with NBCU as an example of how competition can be preserved. It is therefore interesting to compare the two transactions along these lines.


Comcast was large enough to have the incentive and ability to foreclose competitors when it acquired NBCU in 2011, which is why the DOJ challenged that merger. The government agreed to a behavioral remedy that attempted to limit Comcast’s leverage over nascent rival OVD distributors. The consent decree required that Comcast offer an OVD similar sales terms to what Comcast was giving other multichannel video programming distributors (MVPDs).


A second benchmark laid out in the consent decree is the price agreed with the OVD for similar content from another source. These provisions represent two familiar ways of defining a comparable price. One is a price for the content at issue being paid by a buyer who is satisfied (the MVPD) and is not complaining (unlike the OVD). That option was available to the FCC in the case of Comcast-NBCU. Another is the price of similar content sold to the complaining buyer (OVD). The second of these requires some analytical work to be useful because the alternative content will not be identical. It will not be equally valuable, or have the same demographics, or be part of a similar series, or have the same rights (e.g. windowing). These differences mean that the price of the similar content can’t be used without modification, and parties will argue about how to carry out the adjustment. This second benchmark is not entirely straightforward to implement even if a specialist agency such as the Federal Communications Commission has formally reviewed the transaction and can oversee the consent decree, as was true in Comcast-NBCU but not AT&T-Time Warner.


However, the first benchmark price available to the FCC in Comcast-NBCU was markedly better than any such construct would be for AT&T. Comcast earned a significant fraction of its revenue from selling NBC content to most other cable systems in America. Critically, almost all of those other cable systems did not compete with Comcast because their geographic footprints did not overlap (with a few small exceptions). Thus, there was essentially no possibility of attracting consumers to Comcast from Cablevision, for example, by raising Cablevision’s costs through high-priced NBC content. Cablevision’s consumers physically could not purchase from Comcast; all that the higher price would do is lower Comcast’s profits from selling to Cablevision (assuming NBC was already profit-maximizing).


One can see that in this setting there is little incentive for Comcast to change the price of NBC content, assuming NBC had chosen profit-maximizing prices already, when selling to other cable systems. Therefore, the price of content to other large and sophisticated cable systems could create a benchmark that concerned customers could bring to the attention of regulators. And there were concerned customers such as DISH, DirecTV, and OVDs who argued that Comcast would have an incentive to raise the price of NBC content when selling to them because of their status as distribution competitors. Indeed, DirecTV filed comments with the FCC explaining their concerns about being charged higher prices by the combined firm or being denied content. These comments are quoted in the division’s complaint against AT&T.


Now let us consider the different case of AT&T selling Time Warner content and subject to a similar behavioral decree to sell at a “market” price. What sale represents a benchmark price? The arbitrator needs to find a transaction price to a buyer that does not compete with AT&T so that the price remains undistorted. (And this is a conservative benchmark. Even if this condition holds, AT&T may still want to distort this price upward if it realizes a large fraction of its other revenue is based on that one/few transactions.) The arbitrator is looking for buyers whose customers cannot be attracted to any AT&T distribution product should AT&T charge rivals higher prices for Time Warner content. Let us work through the main buyers of content. Cable systems compete with AT&T’s DirecTV for almost all households nationwide; the AT&T U-Verse wired product is another competitor of cable systems in a significant fraction of the country. The complaint describes how cable systems, therefore, are buyers that could be subject to foreclosure or raising rivals’ costs by the combined firm. DISH competes with DirecTV and is therefore also potentially subject to foreclosure. All mobile phone companies—which are increasingly used for cellular video access—compete with AT&T’s mobile phone service and similarly could be foreclosed against. OVDs compete with content arriving over U-Verse, DirecTV, and AT&T’s cellular network.


In short, it is hard to think of any significant future purchaser of Time Warner content that does not already compete with AT&T today. The complaint describes why, in this vertical setting, it is economically logical that AT&T will have an incentive to increase the price of attractive content like HBO in order to raise the cost of rivals and thereby drive their rivals’ customers to AT&T products. Without any net neutrality rules, there is only the 1992 Cable Act that would allow the FCC, should it so desire, to constrain the combined firm from following this strategy, and that rule provides no protection in the broadband OVD segment. And in the absence of a readily observed benchmark price, it will be difficult and contentious to establish “reasonable” prices and conditions at which AT&T should sell its content, and that regulated price will always reflect market conditions from years earlier. No regulator, no matter how competent, will be able to solve this problem.


The second concern in the complaint is the incentive the combined firm will have to slow the attractiveness and adoption of OVD sources of full-length professional video content, e.g. Netflix, Sling, Hulu, and others. The complaint explains that Time Warner content is critical for such services to be attractive to consumers. However, OVD subscription video is a competitor to the video subscriptions sold by DirecTV and U-Verse, and the nascent OVD operation of HBO, HBONow. The complaint argues that the combined firm would have incentive to withhold Time Warner content from OVD distribution. While monitoring any withdrawal from existing agreements is fairly straightforward, it is nearly impossible to determine what agreements over new content would have been struck between Time Warner and OVDs in the absence of the merger. The complaint provides the reasons that, going forward, a free-standing Time Warner would have a greater incentive to sell content to OVDs.


The third major concern raised in the complaint is the problem that if two large distributors each sells content to the other, and each agrees that—due to their size and market importanceno other distributor will obtain better content terms, then those terms essentially become a floor for the whole industry. This type of pricing parity clause is called an “MFN,” and it was common in earlier years of a cable industry formed from distinct, small territorial monopolies. However, deployed by large vertically-integrated firms, MFNs can be used as a tool for coordination across a whole market where satellite and OVD distribution also offer nationwide competition.


With MFNs between the two largest vertically integrated distributors, it’s very unlikely another distributor will be able to negotiate a lower price for that content. Moreover, no rival content providers will be able to cut prices to other distributors without offering the same discount to the largest two distributors, which makes such price discounts very unlikely. The government’s complaint describes this potential behavior of AT&T and Comcast and raises the concern of what economists call “coordinated effects,” an overall lessening of competition in an industry. Coordinated effects can increase due to the increased concentration in the market, and it is not obvious how one would limit it with a behavioral remedy. Instructing the merged firm not to choose high prices in a setting where it knows its largest rival is choosing high prices is likely to be ineffective. Instructing the merged firm not to behave symmetrically with respect to its rivals seems unlikely to be either effective or prudent.


Before taking action, Division staff would customarily evaluate the efficiencies generated by the transaction, and would usually have information and assistance in that process from the merging parties. Since the Division ultimately decided to file a complaint, we can infer that the efficiencies put forward were insufficient to offset these competitive concerns.


A second option in a transaction with anticompetitive effects is to agree on a remedy. A structural remedy would typically involve divesting an asset chosen in such a way as to eliminate the competitive harm. For example, if AT&T divested DirecTV the combined firms’ overlap with cable systems would be much lower. However, it is clear from the economics in the government’s complaint that the harms to competition articulated by DOJ cannot be remedied to any significant degree with behavioral commitments on the part of the merging parties. Combined with AAG Delrahim’s dislike of behavioral remedies, it is thus not surprising that a lawsuit emerged as the outcome of the government’s merger review. It will be interesting to follow the arguments of both sides as they appear in court. There has been significant progress in the theory of foreclosure and in the ability to empirically measure its extent in media markets in the many years since the last significant government-litigated vertical merger. One excellent development that might result from the case is bringing some of that new economics into antitrust policy and precedent.


Disclaimer: The ProMarket blog is dedicated to discussing how competition tends to be subverted by special interests. The posts represent the opinions of their writers, not those of the University of Chicago, the Booth School of Business, or its faculty. For more information, please visit ProMarket Blog Policy.