Bert Foer's comments from ABA Masters Course held October 10, 2014.
I am not sure what we mean by high tech. Does the taxi cab industry suddenly qualify as high tech when Uber introduces modern Internet reservation systems? If so, we can talk about whether state regulations of taxis should also apply to Uber—but we won’t. The rise of the Internet has fundamentally reshaped markets, and yet regulation, including its antitrust variety that I think of as “Regulation Lite”, is still playing catch up. Today I have chosen to talk briefly about four major areas, not including Uber, where antitrust law and technology overlap to muddle the legal landscape, namely: Free products; Natural Monopoly; Innovation; and Competition and the First Amendment.
*President, the American Antitrust Institute, www.antitrustinstitute.org. My thanks to Richard Brunell, Randy Stutz, Sandeep Vaheesan, and most of all Geoffrey Kozen, all on the AAI staff, for their help on this presentation.
I. Free Products
Let’s start with the problem of free products. Many network Internet services are offered at a price of zero. Web products, including Facebook and Twitter, dating sites, or Google’s e-mail and search services are often described as free, although something of value such as giving up personal data or subjecting oneself to advertising is usually a form of consideration. It is worth a moment to ask if there may be antitrust concerns with free products.
Antitrust lawyers like to look gift horses in the mouth, so we shouldn’t stop with the recognition that free is often pro-consumer. The most obvious potential downside is of a product being offered for free as part of a predatory strategy, with intent to drive competitors out of business and recoup losses in the long term. But predation, as difficult as it has become to prove, is not the only possible issue. Relatedly, there is a difficult market definition problem in a two-sided market. In some circumstances, the focus must also be on both sides more or less simultaneously, meaning not only on the Internet user’s zero price but also the advertiser’s economics. Working out how to do this analysis is still a hot topic.
Free goods may also involve non-monetary costs that consumers discount—or are even entirely unaware of. The apparently free nature can interfere with the market signals that we value for the allocative efficiency that antitrust is supposed to facilitate. An illustrative example would be Facebook or Google mining user data to better target ads. When shown how much privacy they are losing, many consumers are displeased. This suggests that companies have the potential of competing on their privacy policies, as well as other terms of trade. Yet many users fail to read the “fine print” when using these services, and so are unaware of the implications for privacy. With these two facts in mind, does it make sense to regard privacy guarantees as a form of non-price quality competition, similar to warranties? How would we react to a cartel that conspires to reduce privacy protections, but not to directly affect price? Two days ago, the opposite question occurred to me when Microsoft and 13 other high tech education companies signed a pledge not to violate the privacy of K-12 students. How would the Rule of Reason deal with this if someone raised a horizontal collusion question? I raise the questions, but since these topics are by my definition “hot”, I will not provide answers other than to suggest that if privacy is a quality that can be taken into account in a competition analysis, we may have to take another look at the National Society of Professional Engineers case, in which public safety apparently could not be considered.
II. Natural Monopoly
Traditionally natural monopolies have been recognized in industries where a single firm supplies all demand at the lowest cost, making competition too inefficient. Because such markets gravitate toward a single firm, natural monopolies have traditionally—one might say naturally-- been more heavily regulated than other markets.
Though some Internet businesses do not have the high fixed costs often associated with natural monopoly, like railroads, many strongly resemble them nonetheless. Facebook, Skype, eBay, and YouTube, to name just a few, derive most of their benefits from powerful network effects; but are they candidates for natural monopoly status or otherwise likely to become so powerful as to require a degree of regulation? Companies wishing to create a competing service are faced with the near insurmountable barrier of convincing users to join without already having a substantial number of other network users, with whom they are interested in interacting. The issue is not necessarily whether a properly earned and properly behaving monopoly violates the antitrust laws, but rather what remedy we would be comfortable applying, whether by antitrust or by legislation, as a substitute for competition-- if we allow a durable natural monopoly to come into existence.
For example, if we have strong reason to believe that a company like Google will inevitably become (or already is) a natural monopoly, wouldn’t it be more prudent to stop Google from acquiring additional applications companies now rather than later having to face up to regulating it like a common carrier? Such regulationI may not be effective, given the company’s inherent advantages in capacity to work around behavioral remedies. More importantly, do we really want to put the government in the position of determining what information will reach the public?
There used to be a popular TV show called, ungrammatically, Who Do You Trust? Would we be more trusting of Google controlling access to information, or the Government? Or are these the only options? Is it a matter only of trust, or can Antitrust help out? We undoubtedly need to be cautious about defining anything in industrial organization as inevitable, but at the same time, if we have grave doubts about how we would regulate an information monopolist, perhaps we should give more attention to using antitrust to keep it from growing by acquisition. Today there is huge focus on Google. Tomorrow it may be Amazon. The world of the dominant internet network company is becoming a hot topic.
I want to now switch gears and talk a little about innovation, which raises topics that were hardly considered antitrust issues just a few years ago.
FRAND: The Problems of Meaning and Remedies
Industry standards are ubiquitous in our everyday lives and more crucial to the economy than ever before. Intellectual property owners push hard to have their patents incorporated into a standard, making them “standard-essential,” sometimes pledging allegiance to fair, reasonable, and non-discriminatory, or FRAND, terms as a condition of adoption of their patent by the standard setting organization. In some instances, patent holders avoid making a FRAND commitment by concealing their patents or using other forms of deception to have their intellectual property unwittingly incorporated into a standard. This so-called patent ambush becomes problematic when, after the standard has been commercialized and entrenched in the market, IP holders suddenly demand a large royalty over and beyond FRAND terms.
In the past three years or so, the Justice Department, the FTC, and the Patent and Trademark Office have made important efforts to address the issue of when the holder of a Standard Essential Patent (or SEP) may seek or obtain injunctive relief or an exclusion order. Progress has indeed come with unusual speed in the antitrust world, such that today there is a widespread consensus that a FRAND commitment means that a SEP holder may not seek an injunction against a “willing” licensee, although there is still dispute about what it means to be a “willing” or “unwilling” licensee.
Somewhat less consensus has evolved on how one determines a FRAND royalty rate, which is obviously highly fact specific. but at least two principles have attracted significant support. One is that the FRAND rate should rest on ex ante incremental value of the patented technology compared to its next best alternative, rather than ex post total market value, because, as in standard patent infringement royalty determinations, the hypothetical royalty negotiation occurs before the technology is incorporated into the standard.
The second principle is that the royalty base should also be the smallest affected component rather than for the entire device. These guides are a good start, though they hardly solve all FRAND licensing issues.
FRAND: The Problem of Succession
A second type of FRAND problem is whether successive owners of Standard Essential Patents are bound by FRAND pledges that their predecessors had made. The FTC takes what seems to be the correct position that FRAND and other licensing commitments should run with patents, like a real estate easement that runs with the land. Be alert to the possibility, moreover, that an owner of a SEP that repudiates FRAND commitments may have even engaged in exclusionary conduct in violation of Section 2 of the Sherman Act. And, arguably, a patent portfolio’s acquisition that enables or facilitates this kind of exclusionary conduct may be challenged as a violation of Section 7 of the Clayton Act. These are hot topics that will likely remain in play for a while.
These intellectual property topics are hot on an international level, too. Earlier this year, for example, the European Commission accepted a binding commitment by Samsung providing that it will not seek injunctions against parties that agree to a two-part licensing framework, which provides for a 12-month negotiation window and, in the event the parties cannot agree after one year, determination of a FRAND royalty by either a court or an arbitrator. More courts around the world are beginning to recognize that injunctions may not be the appropriate remedy for patent infringement suits.
These encouraging developments are part of a larger trend involving the Supreme Court as well as the antitrust agencies and reflect an increased willingness to place patent rights within competition policy, rather than to treat intellectual property as sacrosanct.
Innovation as an Antitrust Defense
Both patents and antitrust are intended to facilitate innovation. But should innovation, by itself, be a defense to an antitrust claim? Can there be predatory innovation? In 2010 the Ninth Circuit issued an opinion in Allied Orthopedic Appliances, finding that any innovative change to a marketed product, no matter how minimal, was a per se defense against exclusion claims. The panel reasoned that it is the place of the market, not the courts, to determine the value of an innovation. While this holding is not binding outside the Ninth Circuit, it created a tidal wave of criticism in the literature. Much of the discussion has centered on attempts to reconcile Tyco with the 2001 D.C. Circuit opinion in United States v. Microsoft Corp., which held that a non-pretextual claim of innovation had to be balanced against the anticompetitive effects of excluding competitors.
I would note a similarity between efforts to declare innovation a defense and the current effort by FTC Commissioner Joshua Wright to have it declared that Section 5 of the FTC Act could not be interpreted in a stand-alone action (i.e., one not based on the Sherman or Clayton Acts) to be violated if the defendant can bring forth even an iota of efficiency proof.
In practice, many district courts, tend either to find that a claim of innovation is a pretextual defense, and therefore rule against the innovator; or they find that it is not pretextual and proceed to permit the purported innovator free rein. The problem with applying this binary test this way is that it obscures the judges’ decision-making and gives too much liberty for anti-competitive strategies.
Once again, balance is needed. There is no question that firms with market power can redesign products in ways that primarily are intended to and in fact do harm competition. This may take the form of redesigns that reduce cross-compatibility or, in the case of pharmaceuticals, promote product hopping, i.e. the extension of a patent through minimal changes. These redesign elements are usually not deemed legally pretextual, in that they offer at least some real benefit to some consumers. But to end the analysis there may allow these minor changes to result in major anticompetitive effects. It is no doubt difficult to quantify the benefits and harms to consumers in a way that allows an objective comparison. But with the current obsession with innovation, we can anticipate more situations of this sort.
IV. Competition and the First Amendment
I’m turning to the last area I’ll discuss today. While antitrust previously had some run-ins with the First Amendment, the conflicts appear to be gaining in importance here, while parallel concerns are rising in Europe, where a human rights treaty is increasingly being brought to bear on competition issues. The transformation of the First Amendment into a corporate shield against regulation is very much a hot topic in high technology because it is being driven by developments in cutting edge information industries that blur the line between production and speech. Consider patent trolling, information products, and the network neutrality debate.
Patent Trolls and Letters Threatening Litigation
The general public understands intellectual property rights as a system to protect valuable inventions and foster future innovation. However, in recent years, we’ve become aware of the growth of patent assertion entities (PAE’s), more popularly known as “patent trolls,” -- a class of patent holders whose strategy is to extract royalties from manufacturers and users of technology products by acquiring patents and aggressively enforcing them. Patent lawsuits can be immensely expensive, so even if the putative defendant in an infringement claim knows that the case has little merit, most threatened targets end up agreeing to pay a generous settlement payment. This can waste valuable resources, raise operating costs, distract from innovation, and may even drive some small companies out of business.
Patent trolls have recently defended their legitimacy by claiming that “cease and desist” letters are free speech protected by the First Amendment, and outside the reach of antitrust under the Noerr-Pennington doctrine. PAEs argue, for example, that if they send form letters to hundreds of companies, threatening to sue for infringement of unidentified patents for components of a product used by the target, the letter must be deemed protected speech, and therefore cannot be the subject of an antitrust suit even if it is used tactically to squelch competition. The argument that the threatening letters are preparation for litigation and litigation is protected speech has been accepted in some courts, although others recognize that repetitive petitioning can be an abuse of process that vitiates the immunity.
With the growth of the Internet, more and more of the economy is being dominated by the information product sector. The producers of information products, such as search engines or data mining companies, market the creation and organization of information; because of economies of scale and scope, many information product providers often have immense market shares that could allow anticompetitive practices to flourish. The informational nature of their product, however, has led several to claim absolute protection from antitrust scrutiny under the First Amendment.
Google has in the past deployed the defense that its search algorithms, and the results they output, are a form of speech and are absolutely protected, barring antitrust claims based on manipulation. Other companies have also adopted similar defenses. Nielsen Media Research argued that their television measurements are protected opinions, and even a company compiling information on doctors’ prescription practices successfully claimed its output was protected speech.
So how do we resolve this? Some commentators have argued that allegedly manipulated algorithms should be considered deceptive commercial speech and thus outside of First Amendment protection. It is unclear that courts will accept those arguments since the commercial speech doctrine, which provides important but reduced free speech protection to commercial entities, is usually restricted to truthful advertising claims. Indeed, recent decisions possibly upholding the seemingly unlimited First Amendment rights of corporations may be in the process of absorbing the concept of commercial speech as a second-rate citizen. If a corporation can have religious freedom to ignore a statute of universal applicability, how can a corporation’s right to speak (not to mention, to give money to politicians) be restricted? On the other hand, one recent decision, American Meat Institute v. U.S. Dep.’t of Ag. provides insight as to how a corporation’s information products might still be regulated; the D.C. Circuit found that mandated country-of-origin disclosure notices on meat —government-mandated speech-- could help inform consumer choice, and so were permitted. A similar argument that links transparency to necessary market signals could be made in other circumstances.
The House of Representatives recently held hearings asking if antitrust was up to the task of dealing with network neutrality; if the FCC enters the field, expect a strong speech defense to any suits challenging inequitable practices. For example, Verizon recently raised a First Amendment claim against being classified as a Type II common carrier, arguing that to do so would “violate the First Amendment rights of broadband Internet access providers, who use their platform to ‘engage in and transmit speech’ – namely, their own Internet services and others’ content services.” Verizon argues that broadband “no less than printing presses or newspaper stands, cannot be conscripted for speech on the government’s terms.” It seems unlikely that Verizon would not make a parallel First Amendment claim to control over which web pages device users can access on its network were it to face either antitrust scrutiny or top-down regulatory oversight. Thus, defining the Constitutional rights inherent in corporate personhood is likely to become an even hotter topic.
Many of these issues we’ve been discussing come at the crossroads of antitrust and other discrete fields of law—privacy, intellectual property, the First Amendment. Today’s antitrust practitioner, like the U.S. enforcement agencies, must increasingly devote time and concentrated energy to maintaining awareness of both technological change and what is happening at the margins of the antitrust field.