AT&T Shellacs the Government in Time Warner Merger Case

Professor Randy Picker of the University of Chicago Law School offers an early take on yesterday’s AT&T-Time Warner decision.

The US government got its clock cleaned yesterday as, in a comprehensive 172-page opinion, Judge Richard Leon denied the government’s effort to enjoin the proposed AT&T-Time Warner merger. Even though the case is clearly driven by a modern-media landscape defined by Netflix, Amazon, Facebook, and Google, the actual case as litigated is quite old-fashioned: the government had a burden of proof and it repeatedly failed to meet it.

On October 22, 2016, AT&T announced that it was proposing to buy Time Warner for $108 billion. Merger aficionados will remember that on January 10, 2000, AOL and Time Warner announced a merger to create what they described as “the world’s first fully integrated media and communications company for the Internet Century.” Many marriages end in divorce as did this one: Time Warner spun off AOL on December 10, 2009.

That merger was a so-called vertical merger and that is the framing here as well for the AT&T-Time Warner deal. In a more traditional horizontal merger—think the proposed T-Mobile-Sprint merger—the merging firms compete with each other directly and the central competition concern is that the merger will boost market power and thereby harm consumers. In contrast, in a vertical merger, the firms don’t compete directly but instead operate in complementary businesses (think Oreos and milk).

AT&T is an infrastructure firm that delivers communications. It owns physical wires for delivering video; a satellite video distributor (DirecTV); and, perhaps most importantly, it has wireless phone and data infrastructure serving more than 100 million subscribers. Time Warner is a content company with valuable video assets, including, of relevance to this case, HBO and the Turner properties (TBS, TNT, CNN, and more). These are complementary assets, not competitive assets.

US antitrust law approaches vertical mergers quite differently than horizontal mergers. Vertical mergers can create substantial efficiencies that directly benefit consumers. Indeed, in this case, the government’s chief expert economic witness believed that the proposed merger would create $352 million in EDM—that’s elimination of double marginalization, not electronic dance music—in benefits for consumers of the merged entity. Double marginalization arises when two firms that deal with each other vertically both exercise market power and eliminating that can be a key benefit of a vertical merger.

Given these possible benefits, vertical mergers are rarely litigated—indeed, Judge Leon said that this was the first one in 40 years—but they are instead resolved through some sort of deal with the government. For example, the 2011 merger between Comcast and NBC Universal was resolved with an agreement on so-called behavioral remedies that imposed restrictions on the merged entity requiring nondiscriminatory access to content. The government didn’t want Comcast to tilt competition by giving its own cable properties access to, say, NBC while denying it to other competitors. Think of this remedy as a kind of case-specific version of network neutrality. It is thought that a Comcast-NBCU style deal was put on the table in this case but was rejected by the government in favor of full litigation.

This merger seems to be driven by the fact that AT&T and Time Warner are data poor. The content business is based on two revenue streams: direct payments for content—think HBO and Netflix—and fees for advertising. The front edge of competition online is with firms that are data rich and are therefore highly data driven. Netflix, Amazon, and Hulu are content producers that are more vertically integrated and gather rich data sets from their customers that they can use to shape the content that they produce and, in Amazon’s case, use throughout Amazon’s broad collection of businesses. And of course Facebook and Google are the two leading data-driven advertising firms.

AT&T and Time Warner seem to be cut off from that data world. Judge Leon notes that Turner actually tried to negotiate for access to data about how consumers were watching its shows and couldn’t get meaningful data. AT&T’s traditional video infrastructure seems not to lend itself to a data approach. AT&T’s video future is not the legacy businesses, even though they may provide a nice if declining revenue stream, but instead moving to a wireless video world with its 100 million subscribers. The merger seems to be a move by the data poor firms to better compete with the data rich ones (Facebook and Google on advertising and Netflix and Amazon on content).

The government’s theory of the case was that if AT&T bought Time Warner it would raise the price of the Turner channels and HBO to other MVPDs (multichannel video programming distributors). The merged firm would make more money that way and if instead another cable company just dropped Turner channels given the price increase, the customers of the cable company might leave it and switch to an AT&T property like DirecTV. The merged firm would benefit either way.

Judge Leon basically rejected every aspect of that theory. The first critical problem was that there was no credible evidence that the Turner properties would actually “go dark,” meaning that for an extended period of time they would be dropped from distribution. That just had not been the history of the industry. It is just too important to maximize the number of viewers with access to the channels. And that meant, in Judge Leon’s view, that the threat to raise prices was an empty threat.

The question then became how we should think about bargaining in the face of that threat. The government’s economics expert attempted to quantify that using a Nash bargaining model and a series of data assumptions, but Judge Leon didn’t really buy any of it. As to the theory, his conclusion was “that the Government has failed to provide sufficient evidentiary support to show the Nash bargaining theory accurately reflects post-merger affiliate negotiations ….” And he found the government’s data on harm wanting for a variety of case-specific reasons, which, given that the government had the burden of proof on possible harms from the merger, was fatal. And the defendants also offered evidence from prior media merger cases suggesting the bargaining leverage theory was not borne out and the government didn’t meaningfully challenge that evidence.

There were other issues presented, but with the heart of the case dead, Judge Leon denied the government’s effort to block the proposed merger. He also urged the government not to use a possible appeal to block the merger.

It is important to note what this case is and is not. This case is not a vertical merger involving so-called foreclosure, meaning a situation where it is expected that competitors of the merged firm will lose access to a key input controlled by the merged firm. The evidence in this case suggested that withholding the Turner channels was an empty threat. An actual vertical foreclosure case could look quite different.

Notwithstanding the glimpses of the present and future of content and advertising, this comes across as a very old-fashioned case. The government had the burden of proof on showing consumer harm from the proposed merger and it is clear that Judge Leon didn’t think that the government had come close to meeting that burden here.

The data might look quite different in a new situation and we are likely to see other vertical mergers soon. I don’t expect another 40 years to go by before we see the next vertical merger case litigated, but with this loss—and the opinion is something of a beatdown—you would not expect the government to rush in to litigate another one of these quickly.

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 necessarily those of the University of Chicago, the Booth School of Business, or its faculty. For more information, please visit ProMarket Blog Policy.