Book Review:
Antitrust Economics at a Time of Upheaval:
Recent Competition Policy Cases on Two Continents
Edited by John Kwoka, Tommaso Valletti, and Lawrence White
Antitrust Economics at a Time of Upheaval [hereinafter, Upheaval ] is a collection of eighteen essays by economists discussing various antitrust matters, either investigations, government enforcement actions, or lawsuits. These eighteen matters took place in either the United States, Europe, or (in a few cases) both. The matters are helpfully grouped in three categories—eight merger matters, eight monopolization/dominance matters, and two matters involving allegedly anticompetitive agreements.
Overall, the book is an excellent resource for demonstrating the economics perspectives of government enforcers and their attempts to grapple with emerging issues concerning innovation, the high-tech sector, and the proper role of antitrust in society. Before discussing in detail the case studies in the book, with a particular focus on a few of them, this review begins with a few general comments.
First, the very concept of the book—a compendium by economists about the economics behind antitrust matters—serves as a powerful demonstration of the extent to which antitrust law has been dominated by economics. Although the United States side of the Atlantic has been hardest hit, the effect in both the United States and Europe is unmistakable. On both sides of the Atlantic, economic concepts once played second fiddle (or no fiddle at all) in many antitrust cases. This rise of economic thought in antitrust law is arguably traceable to the University of Chicago revolution led initially by academics and perhaps best remembered by Robert Bork’s famous tome, The Antitrust Paradox. As the authors note in the introduction, “this ultimately shifted antitrust from a largely legal doctrine to an economic efficiency-oriented policy.” But the “new economic learning” reflected in Upheaval is largely a reaction to the Bork revolution and reflects a more nuanced economics of a somewhat different sort. In contrast to Bork’s work, which largely supported a “keep-the-government’s-hands–off-of-business” approach to antitrust enforcement, the new economic learning develops an economic basis for more robust antitrust enforcement. But the very fact that economists are writing a book to advocate views that are more enforcement-oriented than those of the old Chicago School shows how the Chicago School triumphed. It now goes without saying that the bulk of antitrust issues turn on economic analysis. To tweak Justice Elana Kagan’s tribute to Justice Antonin Scalia: We are all [economists] now.
Second, a major theme in the book is that the more nuanced “new economics” approach seems to have gained considerable traction in enforcement agencies on both sides of the Atlantic but has struggled to gain as much traction with the courts. For example, the book cites the judge in the Sprint/T-Mobile merger who derided the party’s “competing crystal balls” that simply “cancel each other out as helpful evidence”—a sort of classic battle-of-the-experts stand-off. The court was much more moved by the testimony of industry witnesses.
Third, it is important to note that virtually all of the eighteen essays in the book are written by economists who were actually employed or hired by parties in the cases (most were written by economists hired by government enforcers). Thus, it is wise to read through the various pieces with a dose of skepticism given their source. Human nature sheds a bit of objectivity after sweating to develop persuasive arguments in support of a particular position. Although some of the essays are more objective than others, one should not come to Upheaval expecting to get a clear-eyed neutral economic analysis or to find equal treatment of competing views. To take just one example, one group of authors, who assisted the government in an antitrust challenge that was successful in a lower court but unsuccessful on appeal, speaks admiringly of “the pertinent parts of the District Court’s decision, which closely tracks our analysis” and the corresponding “errors made by the Appeals Court” when it reversed the trial court’s decision. They then complain that the appeals court “failed to understand” a “textbook proposition” and was hoodwinked by the “specious academic arguments” advanced by their opponents when “facing a novel business practice they found difficult to understand.” The coup de grace? “Additional economics training for appellate judges,” the writers announce, “would help reduce the incidence of such errors.”
Fourth, given that Upheaval features eighteen discrete and detailed case studies apparently written individually and without much coordination, it is a bit challenging to keep sight of the common themes. To their credit, the editors—all of whom incidentally have served as government antitrust economists themselves—make a commendable stab at this in their six-page introduction. That introduction does a fine job of lifting out a few major themes. But the book could have used a more detailed summation at the end coming back to the themes and tying them into the detailed case studies.
A quick summary of the themes identified in the introduction will help define the remainder of this review. First, the introduction highlights the “new body of academic work” that has arisen in reaction to the ascendance of the Chicago School. Second, the introduction notes that the Chicago-school approach had some negative effects: (i) weakened merger enforcement over time, involving both fewer challenges and the use of behavioral remedies (as opposed to more effective structural ones); and (ii) weak antitrust enforcement against the emerging tech giants—whether in merger enforcement or restraints on “questionable business practices” such as “self-preferencing, misuse of data, tying and foreclosing practices.” Third, the introduction observes that antitrust policy in Europe and the United States diverged somewhat, with European enforcers being less enamored with Chicago-School efficiency arguments than their U.S. counterparts (and, perhaps more importantly, U.S. courts), leading to greater concern in Europe over matters such as predatory pricing, vertical mergers, innovation concerns, so-called “conglomerate effects,” and the activities of high-tech powerhouses generally. Fourth, in both jurisdictions, the cutting edge issues concern the related concept of innovation and the growing dominance of rapidly developing technology.
The rest of this review applies these themes to the more significant groupings of case studies set out in the book—mergers and monopolization/abuse of dominance.
Of the eight merger cases featured in Upheaval, four were in the United States, three were in Europe and one was in both. About half of the merger cases were resolved through divestitures or other fixes demanded by government enforcers, two were abandoned after being disapproved by the European Commission or the UK Competition Markets Authority, and four were challenged by the U.S. government, with the court siding with the enforcers in one of the four.
What do we learn from the eight merger essays? The introduction notes that in many respects mergers are analyzed similarly in both the United States and the European Union and both jurisdictions utilize public guidelines or guidance in this area. But the introduction promises that the case studies will probe “new issues or issues that had long been overlooked” in these jurisdictions.
First, we learn that merger challenges are more likely to succeed in Europe than the United States. For example, the very first essay concerns the proposed 2018 acquisition of Farelogix by Sabre Corp., which involved the sales of technology services to airlines. The U.S. Department of Justice (DOJ) challenge to the merger was rejected by a federal court. Yet days later, the UK Competition and Markets Authority concluded that the merger would substantially lessen competition, and the merger was abandoned within weeks. More generally, of the four U.S. mergers detailed in the book, three of them (Sabre/Farelogix, Sprint/T-Mobile, and AT&T/Time Warner) resulted in judicial rejection of the challenge. And the only case in which the government prevailed (Anthem/Cigna), narrowly survived a 2-1 appellate court opinion where the dissent was penned by Judge Brett Kavanaugh, who now sits on the U.S. Supreme Court. A fifth U.S. merger involved a challenge by the State of California to a hospital merger that was resolved by a conduct remedy limiting future pricing. By contrast, the EU case studies involve two successfully challenged mergers—Sabre/Farelogix and a proposed supermarket merger—and other mergers that were allowed to go forward only after the parties agreed to structural remedies by selling off troubling overlaps—Dow/Dupont, Bayer/Monsanto, and GE/Alstom. Perhaps the differences in result here are best explained by the authors of the Sabre piece: in Europe, the relevant competition authority administratively “reaches its own decisions, instead of having to take cases to court.” As noted below, the greater role of non-specialist courts in U.S. merger policy means that the complex economic models characteristic of the “new economic thinking” might fare better in an administrative system like Europe’s than before generalist judges.
Second, many of the merger cases reported fulfill the introduction’s promise to explore the problem of innovation in dynamic (mostly high-tech) industries. But if there is a theme on this point it is that there is no established economic method for predicting the innovation effects of a merger. The basic concern is easy enough to grasp: one way that companies compete is by innovating to develop new and better products and services. So, the removal of competition might potentially lessen the incentive to innovate (i.e., one does not have to “keep up with the Joneses” if the Joneses have moved away). But how does one truly measure the potential loss of innovation? Or determine whether the risk of innovation lost is great enough to block a potential merger? Or determine in any particular case whether a merger might increase innovation by providing more resources to an innovative firm?
The Anthem/Cigna essay notes that “[t]here is no general economic theory that a merger necessarily promotes or retards innovation.” The authors observed that predicting the result in any case demands knowing the specific market context.” Innovation is often driven by “contestability” i.e., can a smaller firm potentially win business away from established incumbents with a low-price structure by innovating? A larger incumbent firm might be more likely to settle as a “second mover,” perhaps adopting successful innovations after they have been developed and successfully implemented by an upstart. In this case, a merger between an upstart and a larger incumbent might reduce incentives to innovate. But a merger might cut the other way and encourage innovation if it increased the likelihood that the merged firm could retain the benefits of its innovation for itself or if the merger combined complementary assets. The government pointed out that Cigna had been a leader in introducing new insurance products and services (e.g., wellness programs, accountable care). Government experts used a framework to describe strategic factors that influence innovative output and concluded that the merger would diminish innovation. The Court ultimately agreed.
The authors in the Sabre/Farelogix essay note the difficulty of merger challenges in “dynamic industries.” Traditional quantitative tools are of limited value because they are inherently backward looking, whereas the innovation problem centers on predicting a dynamic future. For example, in Sabre, enforcers worried that the merger “could lead to an overall decrease in innovation on the booking service market.” As the “little engine that could,” Farelogix was a reliable innovator, and its innovations spurred Sabre to innovate. Although the concern that eliminating Farelogix would reduce innovation was based on “economic logic,” the authors conceded that there were “no data to conduct any meaningful econometric analysis” and that the enforcers instead constructed their case from the firms’ internal documents. Likewise, the European Commission decisions in both Dow/Dupont and Bayer/Monsanto were animated in part by concern over future innovation. The authors opine that the Bayer case enhanced the “prominence” of innovation competition, extending the concern beyond products that were already far along in the development pipeline (as in GE/Alstom). This obviously permits consideration of somewhat more-speculative competitive effects. Finally, there was some discussion of “reverse-killer” acquisitions, as opposed to the more traditional “killer” acquisition. This would seem to be an acquisition of an emerging competitor that would obviate the need for the acquiring firm to innovate to develop new or better products or services—i.e., by simply acquiring the product or service already developed by the innovation of an emerging competitor.
The Bayer/Monsanto authors propose an innovation roadmap and “limiting principles” for specific cases, considering such factors as whether innovation is an important parameter of competition; the number of significant innovators and the possibility of entry; whether the merging parties are close competitors or have similar innovation capabilities; and whether the target could likely “appropriate” (i.e., benefit from) the results of innovation (normally if IP protection is available).
The authors observe that especially given the ubiquity of mergers of somewhat smaller players in highly concentrated markets, the perennial question is whether (say) a 4-3 merger where the two smallest of the big four join forces (as happened in Sprint/T-Mobile), the result will be to make the merged entity a more significant competitor of the Big Two or whether it will facilitate a cozy oligopoly. Would a merger of “mavericks” defang them as mavericks or instead create a “super maverick?” Note that the DOJ settled its challenge to the merger by requiring the transfer of certain assets to DISH, which would become a new fourth player in the mobile wireless services market. Ultimately, in the lawsuit by the Attorneys General, the court determined that a merged entity would likely compete more aggressively rather than engage in coordination and that DISH, as a new entrant, would be a more influential competitor than Sprint would have been had it remained in the market. (The merging parties also made certain commitments about their post-merger conduct—a fact that critics who favored structural remedies seized upon in attacking the settlement.) In the Anthem/Cigna essay, the authors conclude that, although the courts decided the case based primarily on an assessment of likely static effects based on a 4-3 merger, “we believe …that future courts that weigh healthcare consolidation should give greater consideration to dynamics. Let us not lose sight of the forest for the trees.”
Most merger challenges have focused on unilateral effects—the possibility that the merged firm will be able to unilaterally raise prices or reduce output. Of course, another potential negative result of a merger is increased coordinated effects i.e., the likelihood that the thinned-out number of competitors remaining will coordinate their conduct. According to the authors of the Sprint/T-Mobile piece, although increased coordinated effects is surely a theoretical possibility, “economists have [difficulty] quantifying the extent to which a merger would increase the likelihood of coordination.”
Beyond the dynamic innovation issue associated especially with the tech industry, the discussions of the merger cases presented a number of other issues. Among those of interest:
Of the eight monopolization/dominance essays, six of them focused on Europe and only two of them—FTC/Qualcomm, the “no license-no chips” case, and Ohio/ American Express, a two-sided market case which in fact was not a monopolization case at all—arose in the United States. The types of exclusionary conduct/abuse of dominance on which the cases focus span a broad range: licensing arrangements; tying; excessive pricing; predatorily low pricing; attacks on Google’s practices for Search and Android, which allegedly leveraged power in one market to foreclose competition in another; Germany’s case attacking Facebook’s privacy policies; and a discussion of European platform pricing parity cases involving hotel-room sales. This review will touch on a number of them.
The Google Leveraging Cases. Google has the honor of having two of the chapters in Upheaval devoted to enforcement actions against it—Google Search and Google Android. Both actions were brought by the European Commission, and both invoked leveraging theories. In Google Search, the Commission found that Google abused its dominance and fined it for self-preferencing its own Google shopping results over those of competing comparison shopping sites. The abuse finding was based on a leveraging theory: Google leveraged its dominance in search to artificially strengthen its position in comparison shopping. This was the first leveraging decision against Google and also broke new ground in applying antitrust laws to the dynamic internet search industry. (U.S. enforcers had decided not to bring a similar enforcement action.) Google unsuccessfully argued that decreases in traffic to alternative comparison-shopping sites were driven, not by its actions, but by shifting consumer preferences towards “merchant shopping platforms such as Amazon and E-bay.” Google also challenged the decision as improperly refusing to apply the stringent standards for a “refusal to supply” case (the European version of the essential facilities doctrine). The European Union countered the typical Chicago-School argument against leveraging—i.e., that a monopolist has only one monopoly profit to squeeze from its product—by pointing out that since Google had decided to provide Google Search free of charge to its customers, the only way that it could monetize its monopoly was to leverage its power in search into comparison shopping or similar ventures.
In Google Android, the European Union fined Google and issued a cease and desist order prohibiting it from using restrictive provisions in its contracts with OEMs of mobile devices. Basically, the various agreements allowed the OEMs to install Google Play (Google’s app store) only if they pre-installed Google Search as the default search engine on the device. As in the ongoing DOJ case against Google in the District of Columbia, the European Union was concerned about Google’s distribution practices for its Search product. In essence this “tying arrangement” (with Google Play as the tying product and Google Search as the tied product) was said to have protected Google’s search monopoly and leveraged its extension into search on mobile devices. The Commission found against Google, the European General Court once again largely affirmed, and (once again) a final appeal is still pending with the European Court of Justice.
Google contended that it was these arrangements that enabled it to make the Android operating system for mobile devices both open source and free. Setting Google Search (and Chrome) as defaults did not prevent end users from substituting other search engines or browsers. The European Union pointed to “consumer inertia” as guaranteeing that the default position would have lasting competitive effects. Competing search apps never obtained appreciable market share on devices where Google Search was preinstalled as the default.
The authors describe the case as the “first time that economic theories of exclusion were adapted to a multisided environment, in which the service that was supplied was free to end-users and monetization occurred on the advertising side.” Economic theory posited that such bundling increased the costs of competitors attempting to go head-to-head with Google. The book notes that in many ways, the case was the analogue to the European Union’s earlier Microsoft Media case (which was itself analogous to DOJ’s epic lawsuit against Microsoft) “updated to the age of multisided and zero price constraints.” The authors argue that both Google cases suggest that the remedies process in the cases has been ineffectual. For example, in Android, Google first responded to the cease and desist by offering the various components under a pricing structure that constituted a de facto tie. Subsequent alternatives also tended to simply reaffirm Google’s dominance (“changes that appeared to create the semblance of competition but that are inherently incapable of having any effect”). Interestingly, the authors opine that relying on consumers to make choices that will dilute Google’s power will not work. And they add that the problem was compounded because Google’s tactic had already worked to secure its position in the mobile market long before the European investigation even began. They conclude that “political” problems might prevent the European Union from imposing a meaningful remedy and explain the drift toward regulatory solutions (with the Digital Markets Act) as the logical next response. In short, they conclude that the wheels of antitrust justice grind too slowly to make a meaningful difference in this rapidly developing technological sector.
Abuse of Consumers/Abusive Pricing. Four of the monopoly cases involve pricing or consumer abuse matters—one of them involving Facebook’s asserted abuse of consumers by using their private data, two of them involving allegedly unlawful high prices, and one of them involving allegedly low prices.
(1) Facebook
We’ll start with Facebook. In that case, Germany’s Federal Cartel Office (FCO) brought an action against Facebook under the German Competition Act contending that its exploitation of user’s private data amounted to an abuse of dominance. This case—at least in its initial phases—almost qualifies as what some in the United States would call “hipster antitrust”: an attempt to extend antitrust to solve societal problems caused by big companies. This case goes further than any other in Upheaval in attempting to use antitrust to attack evils not conventionally seen as amenable to antitrust treatment. The FCO attacked Facebook’s asserted violation of German privacy law as an abuse of its dominant position in social media—a sort of antitrust bootstrap that might be amenable to using antitrust law to police violations of virtually any consumer-protection law by a dominant firm. Indeed, the authors suggested the FCO defined Facebook’s actions as “excessive” and thus abusive simply because they violated a law (i.e., the privacy law) “governing situations of imbalanced bargaining.” The authors readily concede that such a rule is “not primarily driven by welfare economics” but is instead rooted in a sort of Brandesian notion of “equal chances in the competitive process” for all participants. In other words, even if competition does not exist, dominant firms are required to pretend that it does exist and act accordingly. This doctrine grows out of the same root as the law against excessive pricing. Although the authors make a half-hearted attempt to justify the case on economic grounds, the better explanation, in this writer’s view, is that the FCO abandoned economic analysis in favor of using antitrust broadly to attack socially undesirable activity by a large company. Indeed, the case went even further (according to the essay) by suggesting that a communications portal like Facebook has assumed state-like status and should thus be subject to the privacy restrictions that apply as against the government. The FCO’s decision, however, was sharply rejected by a German appellate court which stressed the lack of traditional “competitive harm.” But that court itself was reversed by the German Federal Court of Justice, which reinstated the decision of the FCO (but with narrower reasoning) and then sent the case to the European Court of Justice, where it awaits further action. To its credit, the German Federal Court of Justice improved on the rationale of the FCO from a traditional competition perspective. Moving away from a privacy standard, it “reframed the abuse” as a sort of “compulsory tying”; consumers who wanted merely a social network experience were also bulldozed (or hoodwinked) into an unsolicited “highly personalized experience that includes extensive data revelation.” They are “pushed into” this network and “not given a choice as to what to consume and what to give for it.” This leads, not only to the abuse of consumers, but also to harm to competing social networks and advertising clients; as Facebook garners more and more consumer data, entry barriers for potential competitors who do not share such access become formidable.
(2) Excessive pricing
Chapter 11 analyzes three cases against the pharmaceutical industry applying the European rule that excessive monopoly pricing can sometimes violate the competition laws. In this respect, of course, European law differs from U.S. law, which does not constrain the excessive pricing by a monopolist. The authors support the application of the excessive-pricing prohibition to these three specific cases, but they seem to be dubious about the “seminal excessive pricing case” in Europe (United Brands) because it links excessiveness to the “economic value” of the product in question. They assert that “economic value” is a meaningless concept to an economist, since “economic value” consists only of what someone is willing to pay. But they defend the application of the excessive pricing prohibition in the three pharmaceutical cases based on economics work defining the appropriate cases to pursue such claims. Here the cases were justified because each of the defendants was protected by high barriers to entry, increased prices dramatically (more than 1000%) in the face of inelastic demand, and any patents (government monopoly protection) of the pharmaceuticals had long ago expired and the offenders were not the innovators who had initially developed the drugs.
Another case that involved allegedly excessive pricing is the FTC’s case against Qualcomm. Although the authors do not define it as such, their article—and the Ninth Circuit decision—makes clear that the thrust of the case was to argue that Qualcomm’s no license-no chips policy for OEMs was suspect because it amounted to a surcharge exceeding a “reasonable royalty” and circumvented Qualcomm’s FRAND commitment for its standard essential patents. The authors assert that this case differs from other cases where a monopolist charges a high price because “this case relies on Qualcomm’s promise to license its SEP’s on FRAND terms.” In other words, Qualcomm was extracting too much money for the licensing of its SEP patents. It may seem odd that a U.S. case was premised on excessive pricing by a monopolist (since the general law in the United States is that a monopolist does not violate the antitrust laws by pricing too high) but this reflects an ongoing debate over whether this general principle is loosened when a company enters into a FRAND commitment in connection with a standard-setting process (or after otherwise deceiving others involved in that process). Some courts (and economists) have urged that excessive pricing in defiance of a FRAND commitment amounts to anticompetitive excessive pricing. Others—and now including the Ninth Circuit—disagree. The DOJ (under a Republican administration at the time) also weighed in against the FTC’s position and in favor of Qualcomm and the position adopted by the Ninth Circuit. The authors never really explain why an alleged violation of what amounts to a contractual commitment not to charge unreasonable prices should be transformed into an antitrust violation.
(3) Predatory pricing
The final pricing case also involves Qualcomm. This time, however, Qualcomm was accused of engaging in excessively low (predatory) pricing. Once again brought by the European Commission, the case highlights the European Union’s less favorable (to defendants) AKZO I and II approach to a dominant firm accused of predation—unlike in the United States, no likelihood of recoupment is required and (although both jurisdictions demand a price-cost assessment) there is a theoretically greater willingness to consider cost measures higher than average variable cost. Even so, this was the first predatory pricing case the Commission had brought in a decade. In Qualcomm, the economists used Long Run Average Incremental Cost (LRAIC), a measure of cost higher than Average Variable Cost but lower than Average Total Cost, which combined how much Qualcomm paid outside vendors to manufacture its chips and certain product-specific development costs.
The authors describe this as the first time that European enforcers had relied upon pricing that was above Average Variable Cost. The economic basis for this selection was predicated on an “equally efficient competitor” analysis and Average Avoidable Cost work by Professor Baumol. On the recoupment issue—although recoupment is not required under European law—the authors note that Qualcomm’s documents reflected an understanding that normal recoupment would not be possible. The goal of the predation, however, was to nip an incipient competitor in the bud before it could expand into a broader and more lucrative market in competition with Qualcomm. The authors note that this “selective-targeting” style of predation “is not typically captured by economic models of predatory pricing” and “has been explored only sparsely in the economic literature.” They tied the economic rationale to one of three modern economic models of predation—namely “inflicting losses on the prey which will then induce its investors to withdraw their support” because they are somehow unable to “distinguish losses due to predatory attacks from losses due to inefficiency or underperformance.”
In sum, Upheaval provides a many-pronged journey through the economic views of multiple areas of antitrust (mostly from the government enforcement side). The analyses will prove interesting to those who want to learn more about those matters and think about some of the broader issues presented, especially the best ways to use antitrust to foster innovation in the modern economy. Notwithstanding Upheaval’s efforts to be forward-looking and to focus on the future, it seems that enforcement trends are moving so quickly that the book might have difficulties similar to the government enforcement actions it covers. By the time the cases are resolved (or by the time the essays covering them are written and published), the industries involved, the technologies upon which they are built, and the practices under challenge have evolved and moved on at an ever-increasing rate. Which raises the big question, which is perhaps beyond the scope of Upheaval but certainly is pregnant throughout: are competition laws and the enforcement mechanisms that now exist up to the task?
