18.01  We have explored the acquisition and use of personal data to obtain significant market power. Data-driven mergers and network effects, as we have seen, are two important ways to obtain power. A third way is data-driven anticompetitive exclusionary and predatory behaviour, our focus here.

18.02  This chapter cuts through the rhetoric of the virtues of monopolies and inapplicability of antitrust in dynamic industries. The harm from data-driven monopolies can be greater and longer lasting. The chapter explores the issue of path dependence and competitive portals, namely critical inflection points when antitrust can make a key difference. Moreover, Big Data can provide some dominant firms with a ‘Nowcasting Radar,’ which makes them more dangerous than past monopolies. Thus another signpost of progress is when competition authorities, to protect society from these harms, become more proactive in identifying and challenging anticompetitive data-driven strategies that monopolies use to maintain or attain their power.

18.03  Some within the tech industry praise monopolies:

A monopoly like Google is different. Since it doesn’t have to worry about competing with anyone, it has wider latitude to care about its workers, its products and its impact on the wider world. Google’s motto—‘Don’t be evil’—is in part a branding ploy, but it is also characteristic of a kind of business that is successful enough to take ethics seriously without jeopardizing its own existence. In business, money is either an important thing or it is everything. Monopolists can afford to think about things other than making money; non-monopolists can’t. In perfect competition, a business is so focused on today’s margins that it can’t possibly plan for a long-term future. Only one thing can allow a business to transcend the daily brute struggle for survival: monopoly profits.

(p. 278) … the world we live in is dynamic: We can invent new and better things. Creative monopolists give customers more choices by adding entirely new categories of abundance to the world. Creative monopolies aren’t just good for the rest of society; they’re powerful engines for making it better.¹

18.04  Some large tech firms have joined the chorus for non-intervention in dynamic markets. They raise the risk of false positives. The Four Horsemen of the Apocalypse arising from governmental intervention are chilling innovation; chilling investment; protecting competitors, rather than competition; and public choice theory of how companies manipulate competition authorities to punish rivals or erect entry barriers. For some, these Four Horsemen are so scary, that the company gets a pass, whenever it can show any efficiency from its anticompetitive behaviour (even if the efficiency is as little as one dollar, euro, yen, or pound).

18.05  There is no empirical support that monopolies—whether in dynamic or static industries—are generally good for society.² As Professor Baker notes, the claim that monopoly enhances incentives to innovate ignores important ways that greater competition enhances these incentives. Thus firms often increase research and development investment in response to greater investment by their rivals. The claim also ignores the ability of firms exercising market power to restrict, deter, or eliminate new forms of competition through exclusionary conduct. ‘To relax antitrust rules on the rationale that one firm is enough for competition, in rapidly changing high-technology markets or otherwise, would undermine innovation incentives under the guise of protecting them’.³ Ensuring a multiplicity and diversity of independently innovating firms can promote the search for new problem solutions, safeguard the effectiveness of competition as a process of parallel experimentation and mutual learning, and enable faster adaptation to exogenous shocks.⁴

(p. 279) 18.06  The US and EU competition laws seek to prevent mergers that create monopolies. But in neither jurisdiction are monopolies guilty for being a monopoly: ‘it is not in itself illegal for an undertaking to be in a dominant position and such a dominant undertaking is entitled to compete on the merits’.⁵ To be liable, the monopoly must monopolize, that is engage in unfair anticompetitive practices. The EU law is arguably broader than the US law in that the dominant firm has ‘a special responsibility not to allow its conduct to impair genuine undistorted competition on the common market’.⁶ Under EU law, unlike in the US, a dominant firm could be liable for charging an excessively high price.⁷ Nonetheless, in both jurisdictions, the emphasis is on ‘safeguarding the competitive process’ and ensuring that dominant firms ‘do not exclude their competitors by other means than competing on the merits of the products or services they provide’.⁸ Thus, both the US and EU are mindful that ‘what really matters is protecting an effective competitive process and not simply protecting competitors’.⁹ The expectation is that competitors ‘who deliver less to consumers in terms of price, choice, quality and innovation will leave the market’.¹⁰

18.07  Nor is there any empirical support that many companies with market power are now cowering in fear of improving their products so as to not raise antitrust concerns. Google and Apple, for example, were caught colluding. They agreed with their competitors not to poach each other’s employees in order to depress wages.¹¹ So if they are violating antitrust’s per se illegal offences, one wonders how seriously they are worried about abusing their dominant position. Moreover, as we shall see, the US, unlike the EU, has been less willing to intervene in technology markets and pursue dominant companies for potential abuse.

18.08  When competition authorities investigate new, technologically dynamic markets, one response is that the competition laws are inapposite. Microsoft argued—unsuccessfully—in the 1990s that the competition laws were ill-suited for its fast-moving tech market. The argument is that competition comes from surprising sources, so the monopolist needs to continually innovate for fear of the next competitive threat. Few, if any, anticipated Facebook’s and Google’s significance, when the US Department of Justice (DOJ) prosecuted Microsoft in the 1990s, just as few anticipated Microsoft’s dominance when the DOJ prosecuted International Business Machines Corporation (IBM) in the 1970s. If innovation comes from unexpected sources, then competition officials and the public need not worry about monopolistic abuses, since the next disruptive innovator is currently tinkering away in some lab, garage, basement, or attic.

18.09  Dynamic industries, as the US Court of Appeals for DC Circuit observed in Microsoft, do present challenges for courts and the agencies:

As the record in this case indicates, six years seems like an eternity in the computer industry. By the time a court can assess liability, firms, products, and the marketplace are likely to have changed dramatically. This, in turn, threatens enormous practical difficulties for courts considering the appropriate measure of relief in equitable enforcement actions, both in crafting injunctive remedies in the first instance and reviewing those remedies in the second. Conduct remedies may be unavailing in such cases, because innovation to a large degree has already rendered the anticompetitive conduct obsolete (although by no means harmless). And broader structural remedies present their own set of problems, including how a court goes about restoring competition to a dramatically changed, and constantly changing, marketplace.¹²

18.10  But these challenges, the DC Circuit added, did not mean that the competition laws no longer play an important role in curbing abuses by dominant firms in dynamic markets; nor should courts assume this in assessing the case’s merits.¹³

18.11  One problem is that the dynamic disruption, especially in markets characterized by network effects, can take a long time. Suppose someone devises a better search engine or social network. With data-driven network effects, the innovations of ‘one’ will not immediately convert the many.

(p. 281) 18.12  To illustrate, Microsoft benefits from network effects in the personal computer operating system market.¹⁴ It still commanded in 2015 a worldwide share of 90.85 per cent in that market. The threat to Microsoft’s core business, which is built around personal computers, comes not from other PC operating systems, but from mobile phones and tablets. Benefitting from several data-driven network effects, Google continues to dominate the search market, despite Microsoft’s reportedly investing in 2010 over USD 4.5 billion into developing its algorithms and building the physical capacity necessary to operate its rival search engine Bing. Facebook identifies numerous competitors, including Google+ and other, largely regional, social networks that have strong positions in particular countries.¹⁵ Nonetheless, benefitting from several network effects, Facebook continues growing. In December 2014, it had on average 890 million daily active users, an 18 per cent increase from December 2013; on average, 745 million daily active users accessed Facebook from a mobile device, a 34 per cent increase from December 2013.¹⁶ Both Google’s and Facebook’s average revenue per user (ARPU) has steadily increased in the US, Canada, Europe, and Asia.¹⁷ Google’s and Facebook’s dominance on the free side enable both companies to significantly profit from their advertising platforms: Google’s net income in 2014 was USD 14.4 billion.¹⁸ Facebook’s net income was USD 2.925 billion.¹⁹

18.13  As new products or technologies are introduced and existing products evolve, the dominant platforms may become subject to additional competition. Nonetheless, monopolies can persist, even in tech industries. There is no reason why consumers should suffer from monopolistic abuses while waiting for the next big thing.

18.14  Another assumption is that potential competitors will materialize and veer the market towards its natural competitive equilibrium (and what the market would have looked like absent the monopolistic restraints). So, yes, the monopoly temporarily causes harm, but dynamic disruption will shepherd us to the new competitive equilibrium where we would have been (but for the monopoly).

(p. 282) 18.15  This may be true for homogeneous goods, where price is the key parameter of competition. The monopoly elevates the price of bolts, entrants lower the price. The bolts remain the same, but their price differs. In tech industries, the assumption goes further: innovators will materialize whatever competition officials and monopolists do, or do not do, and bring us the innovations that would have occurred but for the monopoly. Indeed, the race may be for the entire market, not just a piece of it, so the incentives to derail a monopoly are said to be even greater.

18.16  The debate is akin to assuming that we have recreated all the cultural knowledge that was lost when the Library of Alexandria was destroyed. This was debated in Tom Stoppard’s play, Arcadia:

THOMASINA: … the enemy who burned the great library of Alexandria without so much as a fine for all that is overdue. Oh, Septimus!—can you bear it? All the lost plays of the Athenians! Two hundred at least by Aeschylus, Sophocles, Euripides—thousands of poems—Aristotle’s own library! … How can we sleep for grief?

SEPTIMUS: By counting our stock. Seven plays from Aeschylus, seven from Sophocles, nineteen from Euripides, my lady! You should no more grieve for the rest than for a buckle lost from your first shoe, or for your lesson book which will be lost when you are old. We shed as we pick up, like travellers who must carry everything in their arms, and what we let fall will be picked up by those behind. The procession is very long and life is very short. We die on the march. But there is nothing outside the march so nothing can be lost to it. The missing plays of Sophocles will turn up piece by piece, or be written again in another language. Ancient cures for diseases will reveal themselves once more. Mathematical discoveries glimpsed and lost to view will have their time again. You do not suppose, my lady, that if all of Archimedes had been hiding in the great library of Alexandria, we would be at a loss for a corkscrew?²⁰

18.17  The exchange raises the issue of path dependence and competitive portals, namely critical inflection points when antitrust can make a difference. Under Septimus’s perspective, high tech products and services are like bolts. Path dependencies play a minor role; we eventually arrive at the same competitive equilibrium and enjoy the same innovations (with or without the monopoly). So monopolies, like the destruction of the Library of Alexandria, do not impose any long-term harm. Thomasina is less confident. For her, under an evolutionary economic process, ‘chance plays a significant role’, and ‘small, random (and therefore unpredictable) events may have severe long-run consequences’.²¹ We may not recover what was lost (by the fire or abuses of a dominant firm).

(p. 283) 18.18  A personal observation are the three adjacent Connecticut rivers: the Housatonic, Shepaug, and Naugatuck. Over time, the persistence of random events—such as a failed company, a leak in a canal, and the growth in demand for brass—accumulated leading to completely different outcomes among the three rivers.²² Today, the bucolic Housatonic and Shepaug Rivers in northwestern Connecticut attract trout anglers, kayakers, and hikers along their heavily wooded banks; the Naugatuck is industrialized. If the history tape was rewound, and random events unfolded differently (say the canal in Falls Village did not leak), the situation today among the three rivers might differ, and walking along the Housatonic may be less inspiring.

18.19  This may be especially true in complex adaptive industries, such as many tech industries. When the competitive portals are open, entry, expansion, or random events during these periods of competitive opportunity can foster significant innovation. On the other hand, a dominant firm may use its market power to close the competitive portals. Thus abuse of dominance, if unchecked, may have greater negative implications beyond that immediate industry and timeframe. The long-run consequences of monopolistic practices may not simply be higher prices, but foregone innovations. So, like some great works lost in the Library of Alexandria that, contrary to Septimus’s belief, were never recreated, so too one price we may pay—when monopolies are unchecked—are innovations foregone.

18.20  Policymakers generally recognize the economist Joseph Schumpeter’s ‘creative destruction’ thesis that capitalism ‘is by nature a form or method of economic change and not only never is but never can be stationary’.²³ Many accept that competition from new commodities, technologies, sources of supply, and organizational structures can be more important than static price competition. The difficulty lies in predicting where and when this innovation will emerge and what motivates this innovation.

18.21  More recent economic theory (building on Schumpeter’s disequilibrium dynamics) has identified the shortcomings of neoclassical economic theory in explaining industries where technological change drives economic growth. As economist (p. 284) Douglass North noted in his Nobel Prize speech, ‘Neoclassical theory is simply an inappropriate tool to analyze and prescribe policies that will induce development.’²⁴ Neoclassical economic theory assumes static equilibrium systems. When subject to an exogenous force (entry, government regulation, new technology, energy crisis), this equilibrium is disrupted temporarily. But neoclassical theory does not explain how markets develop.

18.22  Instead many markets may be complex adaptive systems. Firms, as individuals, make mistakes, readjust, and undertake new strategies. The competitive process ‘is inherently a process of trial and error with no stable end-state considered by the participants in the process.’²⁵ And innovation and dynamic forces may need competitive portals, ie windows of opportunity.²⁶

18.23  To illustrate competitive portals, let us consider whether the former AT&T and IBM monopolies still affect us today. Would our current technological developments exist if AT&T’s and IBM’s monopolies went unchallenged during the 1970s? Can we really say that the level of innovation would be the same today if the DOJ never prosecuted these monopolies?

18.24  The DOJ’s break-up of AT&T is considered one of antitrust’s success stories in unleashing innovation.²⁷ Less clear is IBM. The Reagan administration famously sacked the DOJ’s 13-year-old investigation into the computer monopoly. In a memo to Reagan’s Attorney General, John Roberts (before he was Chief Justice of the US Supreme Court) discussing an upset conservative’s upcoming visit: he ‘will doubtless arrive with many criticisms of the Department for not advancing conservative ideals’. Among the points Roberts mentioned: ‘More reasonable [approach] to antitrust law, epitomized in the dropping of the IBM case’.²⁸ One scholar called the government’s case ‘the greatest waste of resources in the history of antitrust enforcement’.²⁹ Some might say that IBM’s computer dominance (outside mainframes), with or without any antitrust investigation, was destined to be eclipsed—and cite the Wintel combination of Microsoft’s Windows operating system and Intel microprocessors.

(p. 285) 18.25  But one issue is whether the DOJ’s antitrust investigation opened the competitive portal that facilitated Microsoft’s growth. In the late 1960s, IBM controlled about 70 per cent of the computer market. After the DOJ challenged IBM’s practices, particularly its ‘bundling’ hardware and software, IBM changed course. This led to the development of the computer software industry. As IBM’s second president Thomas J Watson, Jr wrote, ‘[p]‌recipitated by a massive antitrust complaint filed against IBM by the Justice Department in January 1969, the company reexamined its practices and decided to stop requiring customers to buy software, services, and hardware as one bundle in June of the same year. This pricing change opened up software markets to independent companies.’³⁰

18.26  A decade later, when preparing to launch its personal computers, the still dominant IBM approached the start-up Microsoft about creating a version of a BASIC computer program. Microsoft suggested that IBM talk to Digital Research, whose CP/M operating system had become the standard for computer hobbyists. One account is that Digital Research’s president apparently disliked the arrogant IBM from his university days and was late in meeting the IBM executives. After the negotiations stalled, IBM returned to Microsoft to create an operating system for its personal computer. When introducing its personal computer, IBM sold the Microsoft operating system for a much lower price than the CP/M-86 system.³¹

18.27  So one cannot assume that with or without antitrust enforcement, Microsoft (or some other operating system) would have become dominant by the 1990s. If anything, the DOJ’s investigation of IBM, it appears, opened a competitive portal, namely IBM’s decision to unbundle software from its computers, which enabled software development to flourish.

18.28  Before the Big Data era, dominant tech firms were less aware of what their customers and rivals were doing (or planning to do). As Chapter 2 discusses, some platforms have a relative advantage in accessing and analysing data to discern consumer trends well before others. As we saw, companies can nowcast, ie, ‘predict the present’ by using search inquiries, social network postings, tweets, etc. Nowcasting (p. 286) can yield a competitive advantage. Hedge funds, for example, are nowcasting to see in real-time how market forces are affecting portfolios, such as how many cars are in the Wal-Mart parking lots across the country.³² In monitoring search queries, Google can predict flu outbreaks well before the government health agencies can. Twitter’s data can help companies identify emerging trends. Google and Apple, in controlling the mobile phone app stores, immediately know when users download rivals’ apps. As the UK competition authority observed,

A number of third party firms also now offer tools and services that enable first parties to gain insights on how their brands and products are being discussed online (sometimes referred to as ‘social listening’, ‘opinion mining’ or ‘sentiment tracking’). By analysing the extent to which they are mentioned in social media content (such as blogs, microblogs, forums, news sites and social network sites), whether trends are positive or negative and why, firms can adjust their marketing activity.³³

18.29  Nowcasting represents a potent data-based weapon, not previously available for monopolies, to monitor new business models in real-time. The data-opoly can use its relative advantage in accessing and processing personal data (such as watching for trends in its proprietary data from posts on a social network, search queries, emails, etc) to quickly identify (and squelch) nascent competitive threats. The dominant firm can acquire entrants before they become significant competitive threats or blunt the entrant’s growth (such as manipulating its search engine results to make it harder to find the company³⁴). For example, Facebook warns its investors that its ‘[p]‌latform partners may use information shared by our users through the Facebook Platform in order to develop products or features that compete with us’.³⁵

18.30  Thus, it is as if the monopoly invented a radar system to monitor in real-time the competitive portals. It can track nascent competitive threats shortly after they take off, and intercept or shoot them down long before they become visible to regulators and others. Moreover, the courts and agencies, if they follow the UK competition authority’s logic in Google/Waze, will find that the distant planes (p. 287) pose potential (yet speculative) threats, and will have insufficient evidence to prove that competition was likely harmed. The monopolist, however, is not troubled by the overall welfare effects in shooting down or intercepting the planes. Granted, the monopolist may damage its reputation, if it acts too brazenly, but reputational concerns generally do not inhibit some monopolies from raising prices. For entrants, there is the potential reward of being acquired. But there are also casualties when the monopoly shoots down others. If the blown-up planes come easier to mind, then some potential entrants (or funders), under the availability heuristic, may amplify the risk, and decide not to leave the runway. And the competition authority cannot force (or incentivize) entrants to fly towards the monopolist armed with this radar.

18.31  The European Data Protection Supervisor asked competition officials to consider (in coordination with privacy and consumer protection officials) the following issues:

• •  first ‘how the control of personal information contributes to market power in the digital economy and the implications for data protection’ and

• •  second ‘the risks to the consumer posed by concentrations and the abuse of market dominance where firms process massive amounts of personal data’.³⁶

18.32  Big Data can help prolong monopolies in at least two ways: data-driven network effects and this unique ‘nowcasting’ radar. In accessing consumer and other market data in real-time, some dominant firms can quickly detect and squelch competitive threats and close competitive portals.

18.33  Competition law, when effectively enforced, can deter exclusionary and predatory practices and keep competitive portals open. After all, it is easier for the creative destruction to breeze through a window screen than topple a concrete wall.

18.34  Big Data, as we saw, can confer power and a durable competitive advantage.³⁷ Data-driven network effects can improve the product’s or service’s quality. Firms like Google thrive (and serve their users) by gathering as much data as possible to personalize search results. At times, consumers benefit from this competitive rivalry and drive to maintain a data-advantage. Companies innovate to expand their (p. 288) platform of services to secure a greater following. No one, for example, questions Google’s investment in technology.³⁸

18.35  But when the stakes are greater, so too are the incentives to engage in unfair practices to tip the market in one’s favour and maintain a monopoly.³⁹ Consequently, competition authorities must be alert to dominant companies’ unfair practices to thwart competitors and disruptive innovators. We outline below several potentially anticompetitive data-driven tactics.

18.36  One historic concern is when a monopoly, through exclusive dealing, deprives its rivals of a needed resource. For example, aluminium producers to extract aluminium from alumina, require a ‘very large amount of electrical energy, which is ordinarily, though not always, most cheaply obtained from water power’.⁴⁰ To foreclose other aluminium producers, Alcoa in its contracts with several hydro-power companies, illegally added covenants binding the power companies not to sell or let power to anyone else for the manufacture of aluminium.⁴¹

18.37  Likewise, the European Commission considers that ‘such input foreclosure is in principle liable to result in anti-competitive foreclosure if the exclusive supply obligation or incentive ties most of the efficient input suppliers and customers competing with the dominant undertaking are unable to find alternative efficient sources of input supply’.⁴²

18.38  Data in our industries is a critical input. Thus one obvious concern would be for a dominant firm to foreclose its rivals’ timely access to critical data. The DOJ had this concern in the Google/ITA Software case, where it prevented Google post-merger from restricting, through exclusive dealing, its rivals’ access to the airlines’ seat and booking class data.⁴³ And in a merger between advertising firms, the European Commission inquired whether the competitors would still have access to ‘big data’ from other providers if post-merger, the merged entity were to develop its own ‘big data’ analytics platform, and deny access to its competitors.⁴⁴

18.39  As the US Court of Appeals for the Eleventh Circuit noted in 2015, a monopoly can violate section 2 of the Sherman Act when its exclusive dealing programme deprives smaller rivals of ‘distribution sufficient to achieve efficient scale, thereby raising costs and slowing or preventing effective entry’.⁴⁵ So too a dominant data-driven company can use exclusionary tactics to prevent rivals from achieving the minimum efficient scale.⁴⁶ Scale, as Chapter 12 discusses, can be especially important in data-driven industries, such as search and search advertising. We saw, for example, how increasing the volume of both ‘tail’ and popular search queries can improve the quality of the search algorithm’s results. In unfairly preventing smaller rivals and potential entrants from accessing critical data, the dominant firm can use the network effects (learning-by-doing, scope, and spill-over effects) to widen the quality gap over rivals, attract more users and advertisers, and expand its platform.⁴⁷

18.40  The Federal Trade Commission (FTC) Bureau of Competition staff, from the released portions of its inadvertently produced report, recommended suing Google for several unfair practices. (The FTC Commissioners instead closed the investigation after Google voluntarily agreed to change some practices.) The FTC legal staff discussed the competitive significance of data and ‘substantial scale effects’ in the Internet search, search advertising, and search syndication markets.⁴⁸ One alleged anticompetitive practice was Google’s use of exclusivity provisions to prevent its rival Microsoft from achieving scale, including the volume of search queries it received. Google used contractual restrictions, according to the FTC legal staff, to deny Microsoft critical scale and impair its ability to compete effectively in the markets for general search and search advertising.⁴⁹

(p. 290) 18.41  One can access a search engine in various ways, such as the browser one uses. Twenty companies (including AOL), the FTC legal staff found, account for 90 per cent of all search query volume. To steer users to its search engine, a search engine provider (like Google, Microsoft, or DuckDuckGo) can enter into distribution agreements with these entry points, namely hardware manufacturers, independent software vendors, and Internet service providers, ‘to distribute toolbars and establish default settings that direct user searches to [its] search engine’.⁵⁰ Google, the FTC legal staff reported, had exclusive or restrictive agreements with four of the top five companies, and 12 of the top 20.⁵¹ Google, for example, is the default engine on Apple’s Safari Internet browser. Google reportedly paid Apple USD 82 million in 2009, and USD 1 billion in 2013 and 2014 for this partnership.⁵² Google’s internal documents, the FTC legal staff found, showed that ‘Google’s interest in renewing deals with some of its largest syndication customers may have been, in part, to keep Microsoft from gaining scale’.⁵³ Interestingly, Amazon decided it was in its long-term interest to funnel some query volume to Microsoft’s Bing, even if it was losing money on each query.⁵⁴ One wonders why others did not do this. Perhaps, as the European Commission generally noted from its market investigation, the distributors’ major concern was Google’s bargaining power.⁵⁵

18.42  We saw in Chapter 9 how the regulated French energy monopoly GDF Suez was using its vast customer database to target customers in the unregulated market with deals on gas and electricity.⁵⁶ Since 2007, French gas customers could opt for the regulated tariffs, which only the incumbent operator GDF Suez offered, or the ‘market’ offers, which GDF Suez and its new rivals offered. In making its market offers, GDF Suez had an unfair advantage over its rivals. It was using the data it collected as a regulated monopoly to target customers with customized offers based on their usage. The personal data in question was commercially valuable. (p. 291) With the data a company could precisely locate gas consumers, identify their consumption level, and propose offers better suited to their profile.⁵⁷ The data was unavailable to the monopoly’s competitors. Nor could the competitors replicate this data. Moreover, the database was not the ‘product of a specific innovation that GDF Suez may have introduced’ but was ‘merely inherited from its former status as monopolistic gas supplier’.⁵⁸ GDF Suez was found in 2014 to have abused its dominant position.

18.43  In a similar case, the Belgian Competition Authority in 2015 fined the National Lottery EUR 1,190,000 for abusing its dominant position when launching its sports betting product Scooore!⁵⁹ Here, too, the defendant used the data it collected as a legal monopoly in organizing public lotteries to enter the sports betting market. The monopoly used the contact details of persons registered in its database to email them about launching Scooore! The National Lottery ‘did not acquire these contact details following competition on the merits but in the context of its legal monopoly’.⁶⁰ Nor could competitors reproduce the data ‘at reasonable financial conditions and within a reasonable period of time’.⁶¹

18.44  In both cases, the dominant firm acquired the data through its legal monopoly. Monopolies in other industries may distinguish these cases. For example, they might have amassed the data following competition on the merits, such as an innovative app. That would be a closer call. But if the firm used other unfair tactics to attain or maintain its dominant position, then arguably using the valuable consumer data from its illegally maintained or attained monopoly is not competition on the merits. Even here, the competition authority must show why competitors could not reproduce the data under reasonable financial conditions and within a reasonable time period.

18.45  To maintain its data-advantage and prevent rivals from attaining scale, a monopoly may make it harder for its customers to leave. Customers, if they are locked-in, will continue to supply the monopoly (rather than its competitors) with data. The basic (p. 292) premise is that as the time and cost needed to switch products or services increase, the greater the customer is locked-in, the harder it will be for rivals to attract users and achieve scale. This is especially the case where consumers cannot readily predict the long-run costs in using that platform or its quality levels over time.⁶²

18.46  Network effects, as we saw, can increase users’ switching costs. For example, users concerned over Facebook’s privacy policies may want to switch to another social network. But unless they can get their friends, family, and acquaintances to switch, they will likely stick with Facebook (if they continue using a social network). This lock-in effect, by itself, does not violate the competition law. Other users’ utility can increase as more join the social network, as they have more people to befriend online.

18.47  But a firm can abuse its dominant position by undertaking additional actions, the net effect of which is to increase users’ switching costs and the firm’s power. One way, in the European Commission’s case against Microsoft, is to reduce the interoperability with other systems or platforms.⁶³ Likewise, in its Facebook/WhatsApp investigation, the Commission inquired, among other things, whether:

• •  users of the consumer communications apps were locked in to any particular physical network, hardware solution, or anything else that needed to be replaced in order to use competing products;

• •  the parties controlled and limited the portability of users’ data; and

• •  the parties had any means to preclude competitors from recreating a user’s network on the parties’ applications.⁶⁴

18.48  Presumably, if the answer was yes, the risk of anticompetitive unilateral conduct increases. Facebook and WhatsApp users, the European Commission concluded, could easily port their contact data to other texting apps. (Moreover, texting data, the Commission found, had little long-term value.) But if consumers invested a lot of time and effort in the service, such as a homepage with photos, timeline, updates, etc, and the dominant firm blocked customers’ ability to port their data, when data portability was technologically feasible, that would raise antitrust concerns.

18.49  The switching cost can be as subtle as setting the app or service as the default option, thereby requiring consumers to opt out each time they want to use another service.

18.50  We saw in Chapter 14 why platforms are worth billions, while apps are worth millions. Apple and Google have significant power in effectively controlling the respective mobile operating systems iOS and Android.⁶⁵ Both Google and Apple have business interests in ‘targeted advertising’ and ‘run the two biggest services, by revenue, for putting advertisements on mobile phones’.⁶⁶ Google especially relies on personal data for maintaining a competitive advantage for advertising.⁶⁷ As such, they have a greater incentive to prevent the personal data being diverted (as well as individuals’ using rival apps). To maintain and secure fresh sources of valuable data, Apple and Google have greater opportunities to introduce their own applications and foreclose rival applications on their smartphone platforms. Thus there is a greater risk of exclusionary behaviour.⁶⁸

18.51  Competition authorities are sensitive to vertical integration by a dominant platform operator (ie where it also becomes a seller on its platform). The platform’s incentives now change, as it may earn greater profits by steering users and advertisers to its own products and services to the detriment of rival sellers (and contrary to consumers’ wishes). The platform has a ‘frenemy’ relationship with the independent application developers.⁶⁹ The platform and independent apps are friends—in that both (p. 294) benefit as more users and complementary software developers are attracted to that platform, as opposed to rival platforms. Such will be the case when it is costly or time-consuming for independent software developers to customize, promote, and update their apps across multiple platforms or where one platform imposes greater restrictions on functionality, terms of sale, advertising, etc. The platform operator, however, is also competing with the independent software developer’s app, and thus an enemy.⁷⁰ As the Organisation for Economic Co-operation and Development (OECD) warned, the platform owner ‘may seek to exclude third-party applications developers, either to protect its own vertically integrated applications subsidiary or to prevent the emergence of a potentially competing platform’.⁷¹

18.52  Thus in its Facebook/WhatsApp investigation, the European Commission inquired whether:

• •  the parties controlled any essential parts of the network or any mobile operating system; and

• •  the parties’ applications were pre-installed on a large base of mobile phones, tablets, or PCs, and if so whether ‘status quo bias’ could potentially affect consumers’ choices.⁷²

18.53  Likewise, Facebook in 2015 warned investors of the risk of the dominant mobile platforms inhibiting Facebook’s apps or preferring their own programs or services.⁷³ So did Twitter, LinkedIn, Yelp, and smaller online platforms, like Coupons.com. They all noted their dependence on the Apple and Android mobile platforms.⁷⁴ They recognize that web usage is increasingly shifting to mobile platforms such as smartphones and other connected devices.⁷⁵ Their business growth and success depend on their interoperability with the popular mobile operating systems that they do not control.⁷⁶ So one significant business risk is if the mobile super-platforms—Apple (p. 295) and Google (and to a much lesser extent Microsoft)—change the mobile operating systems that degrade the functionality of the independent apps and online platforms—like Twitter, Yelp, or Coupons.com—or give preferential treatment to their own similar services or competitive services.

18.54  Facebook identified several potential anticompetitive measures by the mobile super-platforms including:

• •  degrading the independent app’s functionality,

• •  reducing or eliminating the independent app’s ability to distribute its products,

• •  giving preferential treatment to competitive products, or

• •  limiting for any app whose revenues are primarily from advertising its ability to deliver, target, or measure the effectiveness of ads, or imposing fees or other charges related to its delivery of ads.⁷⁷

18.55  In its 2012 Annual Report, Facebook warned that ‘[c]‌ertain competitors, including Google, could use strong or dominant positions in one or more markets to gain competitive advantage against us in areas where we operate including: by integrating competing social networking platforms or features into products they control such as search engines, web browsers, or mobile device operating systems; by making acquisitions; or by making access to Facebook more difficult.’⁷⁸

18.56  Facebook, given its apps’ strong consumer appeal, has less to fear than smaller, lesser known apps. To make it harder for consumers to access and use the smaller independent apps on their smartphones, the super-platform could:

• •  degrade the independent app’s functionality by having it run slower than the operating system’s app,⁷⁹

• •  reduce or eliminate the independent app developer’s ability to distribute its app by making it harder for consumers to find the app on its search engine or app store,⁸⁰

• (p. 296) •  limit a competing app’s revenue stream by excluding the app from its online wallets, such as Apple Pay and Google Wallet, or

• •  give preferential treatment to its own products, by pre-loading its app on the smartphone, having it on the opening screen,⁸¹ or integrating its own products into its other popular products, including its search engine and the operating system.

18.57  These concerns are real. One example is Bankrate Inc. Its website allows users to compare online the rates of over 300 financial products, including mortgages, credit cards, automobile loans, money market accounts, certificates of deposit, checking and ATM fees, home equity loans, and online banking fees. As the Internet’s ‘leading aggregator of financial rate information’, Bankrate, according to its website, ‘continually surveys approximately 4,800 financial institutions in all 50 states in order to provide clear, objective, and unbiased rates to consumers’.⁸² During the fourth quarter of 2015, Google began testing a competing service called Compare Credit Cards. Google’s search engine displayed its own service more prominently on credit card-related search results than Bankrate’s service.⁸³ The fallout was significant. First, Google’s actions ‘adversely affected’ Bankrate’s ‘Credit Cards segment growth and profitability’.⁸⁴ Second, Bankrate’s stock price, after this news was released, declined 48 per cent in one day, a record drop for the 40-year old company.⁸⁵ Bankrate’s stock plummeted even though Google earlier announced that it was terminating its Compare services, including Compare Credit Cards.⁸⁶ Investors were still jittery. Bankrate, along with many other companies, depends on Google’s search engine to attract a significant portion of visitors to its website. As one analyst commented, ‘Bankrate faces an uncertain future, in our opinion, as its ability to maintain low-cost traffic and consistent monetization appears threatened by changes over which it has little control’.⁸⁷

18.58  The browser war between Microsoft and Netscape in the 1990s is another example. The DOJ challenged several actions Microsoft took in integrating its Internet Explorer browser into its Windows operating system. In technologically binding its browser to Windows, Microsoft, the district court found, both prevented original (p. 297) equipment manufacturers from pre-installing other browsers and deterred consumers from using them.⁸⁸ The US Court of Appeals for the DC Circuit affirmed that Microsoft’s commingling of its browser and operating system code violated section 2 of the Sherman Act.⁸⁹ So, too, the super-platform can abuse its dominant position by fusing its app with its operating system code, when it does not achieve any real integrative benefits, but helps maintain its data-advantage and monopoly by reducing users’ likelihood of using competing apps.

18.59  Moreover, data-driven exclusionary conduct may unite some within the Chicago and post-Chicago Schools. University of Chicago professor Dennis Carlton is a member of the Chicago School. Like others in the Chicago School, Carlton is generally sceptical about antitrust enforcement directed towards exclusionary conduct by a monopolist. Carlton, however, accepts that there is a legitimate role for antitrust in refusal to deal cases in certain situations.⁹⁰ Significantly, he has argued that antitrust enforcement is appropriate in dynamic industries (such as the computer industry) where network effects are present and where scale is especially important to the ability to compete. He argues that:

in a dynamic model, the cost of being small initially can be magnified in later periods, especially with assumptions about network dependencies, importance of installed base, or scale economies. In those settings, strategic behavior designed to keep a rival small initially can yield later significant competitive advantage.⁹¹

18.60  Carlton finds these conditions satisfied (and enforcement appropriate) in Lorain Journal, where the owner of a local newspaper, which was the major local advertising vehicle, responded to the entry of a local radio station by refusing to deal with customers who advertised on the radio.⁹² Carlton notes that most commentators have viewed the case as suggesting that radio and newspapers are substitutes, but it is better to view them initially as complements for some advertisers—ways of reaching different demographic groups. However, over time, radio could grow into a substitute. So, Carlton suggests, the exclusionary conduct was ‘designed to so limit the size of the radio station that it could not survive as a vigorous competitor later on’.⁹³

18.61  Similarly, Carlton argues that the government’s Microsoft cases were appropriate. The first case involved de facto exclusive dealing by Microsoft, which required computer manufacturers to pay Microsoft a licence fee based not on how many (p. 298) computers they shipped with the Windows operating system but based on how many computers they shipped in total. The second case involved contracts with computer manufacturers that either required or created incentives for exclusivity in browsers. Carlton notes that these cases, similar to Lorain Journal, are properly viewed as limiting potential rivals to the operating system monopoly from attaining efficient distribution.⁹⁴

18.62  The European Commission in 2015 opened a formal investigation involving Google’s Android. Although Android is an open-source mobile operating system, which others can freely use and develop, Google controls the operating system through its licensing agreements. As the Commission stated, the ‘majority of smartphone and tablet manufacturers … use the Android operating system in combination with a range of Google’s proprietary applications and services. In order to obtain the right to install these applications and services on their Android devices, manufacturers need to enter into certain agreements with Google.’⁹⁵ The European Commission is investigating whether Google has

• •  ‘illegally hindered the development and market access of rival mobile applications or services by requiring or incentivising smartphone and tablet manufacturers to exclusively pre-install Google’s own applications or services’;

• •  ‘prevented smartphone and tablet manufacturers who wish to install Google’s applications and services on some of their Android devices from developing and marketing modified and potentially competing versions of Android (so-called “Android forks”) on other devices, thereby illegally hindering the development and market access of rival mobile operating systems and mobile applications or services’; and

• •  ‘illegally hindered the development and market access of rival applications and services by tying or bundling certain Google applications and services distributed (p. 299) on Android devices with other Google applications, services and/or application programming interfaces of Google’.⁹⁶

18.63  Besides Android, the Commission is investigating several other Google business practices:

1. (i)  The use by Google without consent of original content from third-party web sites in its own specialized web search services.

2. (ii)  Agreements that oblige third-party web sites (‘publishers’) to obtain all or most of their online search advertisements from Google.

3. (iii)  Contractual restrictions on the transferability of online search advertising campaigns to rival search advertising platforms and the management of such campaigns across Google’s Adwords and rival search advertising platforms.⁹⁷

Moreover, the Commission in 2015 issued its statement of objections over Google degrading the quality of its search results by systematically favouring its own comparison shopping products in its general search results page.⁹⁸

18.64  It bears noting that these allegations have not been proven in court. The Commission’s open investigations as of early 2016 have not reached statement of objections or formal action, and even the statement of objections are preliminary, with Google having the right to respond. Our point here is not Google’s potential liability, but to illustrate the types of abuses by dominant firms that touch on Big Data. To adequately assess these claims, the competition authority and court must understand the competitive significance of the four ‘V’s—volume, variety, velocity, and value—of data, the data-driven network effects, and how these data-driven strategies may help companies attain and maintain their dominant position and leverage their power across markets.

18.65  Lastly, competition authorities must respond swiftly to prevent data-opolies from benefitting from their unfair data-driven practices. As we saw, data-driven network effects increase firms’ incentives to resort to unfair tactics. As the benefits from illegality increase, so too must the magnitude and probability of punishment increase to deter the anticompetitive behaviour. Otherwise, monopolization pays.

(p. 300) 18.66  In the US, monopolization pays. The DOJ criminally prosecuted more persons in one year under the Migratory Bird Treaty Act (227 in 2012)⁹⁹ than it has civilly and criminally prosecuted monopolies over the past 35 years (13 since 1980).¹⁰⁰ Between 2005 and 2014, the DOJ opened only 19 monopolization investigations, and brought only one case (in 2011).¹⁰¹ Thus a monopoly has more to fear about its wind turbine killing a golden eagle¹⁰² than its executives killing off a competitor.

18.67  In the US, executives conceivably could go to jail for monopolization. Over the past 50 years, Congress has increased the maximum criminal fines and term of incarceration for Sherman Act violations. From a misdemeanour, the criminal penalties now stand as a felony with up to ten years’ imprisonment and a fine up to USD 100 million for corporations and USD 1 million for individuals.¹⁰³ The Sherman Act does not delineate which conduct should be criminally or civilly prosecuted; this has been left to the DOJ’s discretion. The DOJ, however, has not criminally prosecuted firms or individuals for violating section 2 since the 1970s.¹⁰⁴ Since the Reagan administration, the DOJ has criminally prosecuted only horizontal, per se illegal agreements among competitors, such as price-fixing, bid rigging, and customer and territorial allocations. Nor has the FTC brought many monopolization cases.¹⁰⁵

18.68  The antitrust fines likely represent a fraction of the monopoly profits. This is especially so, when dominant firms can avoid antitrust liability for their abuses in jurisdictions like the US. Class action antitrust lawsuits, under the recent Supreme Court decisions, are harder to bring. If there is a problem with class action settlements in (p. 301) antitrust cases, the American Antitrust Institute found, ‘it is that plaintiffs sometimes settle strong cases for too little, not weak cases for too much’.¹⁰⁶

18.69  While running for president Barack Obama criticized the Bush administration for having ‘what may be the weakest record of antitrust enforcement of any administration in the last half century’.¹⁰⁷ Obama noted that ‘in seven years, the Bush Justice Department has not brought a single monopolization case’.¹⁰⁸ Obama promised to ‘reinvigorate antitrust enforcement’ and ‘step up review of merger activity.’¹⁰⁹ Now with his second term coming to an end, the same criticism has been made about his administration.¹¹⁰

18.70  Many tech firms’ business models depend on collecting and monetizing consumer data. Several network effects can enable the company to become so firmly entrenched, so dominant in a given market, that it has both the ability and incentive to squelch competition, including by mavericks who challenge that data-dependent business model. When that happens, the incentive to innovate and take on that data-opoly is diminished. Consumers, even though they continue to get many apps and services for free, are nonetheless harmed, including the loss of technology that advances their privacy interests.

18.71  Although the EU is more active in investigating abuse of dominance cases, this cannot be left to one jurisdiction. Monopolization pays today. The incentives to abuse a dominant position, given the network effects, are even greater in data-driven industries. So, too, are the opportunities, especially for data-opolies with the nowcasting radar or controlling a critical platform, like smartphones. If the competition authorities ignore data-driven exclusionary and predatory conduct, then we will likely see more industries dominated by a few firms. Thus another signpost of progress is when the US and other jurisdictions investigate and swiftly prosecute data-driven abuses.