Discussion: (0 comments)
There are no comments available.
A public policy blog from AEI
The latest on technology policy from AEI, published daily.
View related content: Technology and Innovation
Last week’s announcement that the Department of Justice (DOJ) is suing to stop AT&T from merging with Time Warner echoes a remarkably similar finding in February this year, when the New Zealand Commerce Commission (ComCom) declined to give clearance to a merger between Sky Television, the country’s top pay television provider, and Vodafone, its top mobile operator (and number two fixed-line broadband supplier).
As in the US case, the ComCom expressed concerns that there was “the real chance that the merged entity would leverage its market power over premium live sports content, foreclosing competition in the relevant broadband and mobile service markets in the long term.” It held there was “a real chance that it would have the incentive to use its market power over premium sports rights to supply bundles of pay TV, broadband, and mobile services with which rival TSPs would be unable to effectively compete.” And just as has occurred in the United States, there was considerable dismay among some antitrust aficionados (and even humble academics such as myself and coauthor Petrus Potgieter) as to how it could be found that there were anticompetitive effects in a vertical merger, especially given the high degree of innovation and apparent competition in both the content delivery and broadband markets.
At the heart of the New Zealand decision, and likely underpinning the DOJ’s concerns, as the ComCom identified, is the potential for a firm with control of content rights to extend that market power into the broadband markets where increasingly that content is delivered. Indisputably, as content is subject to copyright, the rights holder has a monopoly over it. But can that market power be extended downstream? Based on Michael Whinston’s famous paper from the days of the plain old telephone systems (POTS), a telco with market power in the upstream network access market (local line rental) could extend that downstream to services provided in a competitive market (e.g., local calling and long-distance) by tying sale of the service to access. As network access could only be purchased with the dominant firm’s calling service, stand-alone calling competitors would be foreclosed.
The parallel appears clear — if content is equivalent to network, then broadband access is equivalent to competitively supplied local calling and long-distance.
But of course, the real comparison is nowhere close to this presumption.
First, and crucial to Whinston’s argument, is that the dominant firm’s services are tied — that is, the only way that the POTS network access can be acquired is by buying the merged firm’s bundle. Even if tying is not used — that is, the network access could be purchased by a consumer separately so the consumers can “make their own bundle” by purchasing a rival’s long-distance service (mixed bundling) — the stand-alone price would be jacked up so as to make the merged firm’s bundle overwhelmingly more attractive. The rival calling service would be uncompetitive with the two components purchased separately, and so foreclosed from the market. The merged firm can then jack up its bundle prices too, knowing that the rivals will not reenter.
Second, and very importantly, the model relies on the two products being perfect complements — that is, calling would be of no use to the consumer without network access and vice versa. And it is here that the parallels to content distribution and broadband access start to break down. The upstream product — a specific set of copyrighted content — is not essential to some (or even most) consumers’ ability to lever value from having a broadband connection. Even if the content provider bundles it with broadband, so long as broadband is available separately (i.e., mixed bundling), there is no requirement for those consumers not valuing that specific content to buy their broadband from the merged firm. They can buy it from any number of other competitive suppliers. Consumers buy the bundle because it gives them more surplus gain, and nobody loses as the products can still be sold stand-alone.
The merged firm can jack up its content prices but not materially affect the market for broadband so long as the majority of consumers can buy their broadband sans the specific content wherever they choose. The merged firm may have some market power over those who do value its content highly, but if they are not the majority of broadband consumers, the likelihood of foreclosure occurring in the broadband market diminishes. And if it has more broadband consumers who do not purchase the content than do (as occurred in the Sky/Vodafone case), then it can’t jack up those prices without harming its profitability either. This is the argument Jay Pil Choi posed in his more nuanced paper, which comes to the conclusion that the foreclosure risk is highly contingent on a range of highly specific market circumstances prevailing. We argue that most of these do not prevail in the current or likely expected future markets for the relevant products.
Choi and Sue Mialon show (separately) that the foreclosure risk diminishes significantly under many of these circumstances currently observed in the markets for content distribution and broadband. The more firms there are selling products in either market, the less likely foreclosure is. The less complementary the products are, the lower the foreclosure risk. The nature of competition in both markets matters — the more intense competition is (the closer substitutes the products are), the more likely foreclosure is (as arguably is the case if all broadband connections are interchangeable). But the less intense competition is, then the less likely foreclosure is — for example, in the monopolistically competitive content markets, Netflix is hardly a substitute for the sports channels. Also, the scope for the market to expand also matters. The foreclosure models assume the market is fully saturated, and the only way for one firm to gain market share is to steal it from another. But where the market can expand — as is the case where consumers can buy more and more new and different content and applications, either alone or in bundles with broadband access, then once again the less likely is foreclosure.
So if all these possible reasons militate against the harms to competition from one firm owning both content and broadband, then what exactly is the DOJ’s proposed factual if the merger proceeds? And what makes it more likely to occur under common ownership than under the contractual arrangements under which the same bundles could be offered (and were in the New Zealand case)? These and a lot more interesting questions will undoubtedly be argued in the ensuing court case. It will be interesting to see how the various protagonists argue their cases — and what in the end is deemed to be legal or illegal.
There are no comments available.
1789 Massachusetts Avenue, NW, Washington, DC 20036
© 2018 American Enterprise Institute