In the second of a series of articles on the historic antitrust lawsuits against Google and Facebook, I look at what the endgame may bring for the government plaintiffs and highlight the challenges they face along the way. I also preview the broader legal reforms and regulatory changes that may follow if competition laws reveal themselves unable to reach the conduct at issue in these cases.
For an overview of the antitrust lawsuits against Facebook and Google, and what they mean as a practical matter for tech businesses, check out the first article in the series. (Update: A recent amendment to the Texas-led complaint against Google has added an allegation that the company’s plan to eliminate third-party cookies on its Chrome browser locks out potential ad tech rivals who depend on the technology to track users and target them with ads. The government claims this will give Google a competitive advantage over these rivals because it will be able to continue to collect user data on Chrome and its other products and services.)
Predicting how government antitrust cases will end is tricky business. Their fates rest in the hands of the shape-shifting enforcers who bring them, the varied courts who render the decisions and hear the appeals, and the fluid markets upon which their facts turn. In a variety of ways, the ground beneath a major antitrust case can shift and send everyone in unpredictable directions. Then there is the fact-heavy inquiry that these cases call for, the evidence getting culled and weighed by a single judge. Overlay all of that with a body of case law that is slow to incorporate novel legal theories, and the unpredictability and uncertainty of the antitrust lawsuits against Google and Facebook start to become clear.
Though by no means a futile endeavor, these are tough cases for the government. My skepticism in this regard is nothing new (as I previewed last summer here). But the detailed complaints that have now been filed by the government plaintiffs offer a deeper insight into the major legal roadblocks ahead.
The government’s cases may be found to revolve around a difficult antitrust “duty to deal” claim
Although the word “duty to deal” is never uttered in the complaints, core aspects of the government’s antitrust lawsuits against Google and Facebook center on the notion that the tech giants have an obligation to work with their competitors and business partners operating within their digital ecosystems.
With Facebook, the duty to deal underpins the exclusionary conduct at the heart of the case: that selectively cutting off access to its API (needed to send/receive data to/from Facebook’s platform) to third-party app developers made it difficult for potential social media rivals to emerge. With Google, it lies in allegations of failing to make its ad tools interoperable with third-party ad services (including potential competitors) or precluding their access to user data and connectivity to auctions on par with what it granted its own vertically-integrated ad tech tools. Other claims that hint of a refusal-to-deal include allegations that Google disadvantaged search result positioning or display of third-party vertical shopping search services (also known as “aggregators”) that competed with its own Google Shopping.
These various allegations could be framed by the plaintiffs as seeking to impose on Google and Facebook a duty to deal with its rivals. Yet none of them are, and there’s good reason for that. Decades of attack on the so-called “essential facilities” doctrine in economic academia, a skepticism adopted into the law by federal courts (including the Supreme Court) rendering decisions in key antitrust cases, has all but eliminated the legal basis for arguing that a monopolist has a duty to deal with its rivals or business partners in any but the narrowest of circumstances.
The government will try to avoid this legal roadblock by framing its allegations as not ones based on the defendants granting access to essential online facilities but rather as involving other forms of “exclusionary” conduct that face slightly less hostile legal precedent. The defendants will try to recast the plaintiffs’ allegations as seeking to impose a duty to deal with its rivals, bringing as many claims as possible under the more trying legal standards. Given the present composition of the federal courts (including the Supreme Court) and a recent trend of reining in the reach of monopolization law, that strategy could very well succeed. That, in turn, could strongly influence the government’s prospects.
The government faces challenges in showing that Google and Facebook harmed competition (anti-competitive effects).
In the the government antitrust lawsuits against Google and Facebook, as any alleging monopolization, it is not enough to show that a company has attained a monopoly or near-monopoly. It must be shown that the company did something to increase or maintain its market power by eliminating competition. The accused monopolist has to have undermined the competitive process itself, which is shown by the government proving that its conduct had anticompetitive effects” on the market.
Harm to competition is shown by reference to a so-called “consumer welfare standard” which focuses on how output or price levels in the market have been impacted. The practical effect of this is that competitive harm is generally seen as occurring only when a dominant firm engages in exclusionary conduct aimed at its rivals which lowers output or increases prices charged to consumers. Otherwise, anti-competitive effects are unlikely to be found to exist—no matter what other harms can be shown (whether to markets, society, political systems, or otherwise). One consequence of this standard is that harm to consumers reigns supreme, while harm that is felt only by competitors takes a back seat. What’s bad for a competitor is good for customers, so the reasoning goes.
Competition law’s singular focus on price and output-related harms to consumers poses problems for the government plaintiffs in the antitrust lawsuits against Google and Facebook. One of them is clear right off the bat: the defendants’ plethora of innovative products are free to end users and very popular with them. So as far as consumers (end users) are concerned, the government is left having to argue non-traditional forms of competitive harm that are more difficult to nail down, such as a slower rollout of product innovations or the absence of sufficient choices for users.
The problem for the plaintiffs is that such non-traditional iterations of anti-competitive effects, though not expressly precluded by antitrust statutes, lack well-grounded precedent in the applicable case law interpreting them. These novel theories of harm also raise evidentiary challenges, as the requisite proof may evade the government when it comes to anti-competitive effects that are not prone to economic measurement and analysis. Price and output, after all, are quantitative–and even those elements prove highly elusive for economists seeking to analyze competitive effects in antitrust cases. (I’ve written about that here.)
One non-traditional theory of competitive harm to consumers in the government’s antitrust lawsuits against Google and Facebook involves allegations that the defendants’ exclusionary acts harmed competition by slowing or preventing the roll-out of products and services with better protections for user privacy. For example, the Facebook complaints allege that the company “degraded privacy protections” on the products of the companies it acquired, resulting in “reductions in quality and variety of privacy options” and loss of “consumer choice” in data protection. The DOJ complaint against Google alleges that its exclusionary practices “harmed consumers by reducing the quality of general search services (including dimensions such as privacy, data protection, and use of consumer data).” The Texas-led complaint goes even farther, alleging that privacy is the very arm with which Google restricts competition (claiming that it “uses privacy concerns as an excuse to advantage itself over its competitors” and engages in “privacy fixing” with competitors).
Whether antitrust and data privacy should overlap or be kept as distinct bodies of law is a live debate in competition policy circles and academia. A tension between the two often arises. (I’ve written about that here.) The lawsuits against Google and Facebook are the first major attempt by US agencies to wed antitrust and data protection concerns, and so they would seem on track to force federal courts to weigh in on the important issue with precedent-setting decisions.
But the antitrust cases against Google and Facebook do not only allege competitive effects on the basis of exclusionary conduct that imposed harms directly on consumers. The government also claims that exclusionary practices aimed at other market participants–in particular rivals–operating within the tech companies’ digital ecosystems had the effect of eliminating competition. For example, Google is alleged to have engaged in exclusionary conduct that allowed it to “charge advertisers more than it could in a competitive market” and that the company “has charged supra-competitive fees and degraded quality in the publisher ad server market.” Facebook, for its part, is alleged to have “deprived advertisers of the benefits of additional competition” such as “lower advertising prices.”
These claims of anticompetitive effects have the benefit of more clearly involving the sorts of increases in prices or decreases in output that monopolization cases usually involve. But to the extent they base their claims on harms imposed on rivals, the government will need to make a connection to competitive effects imposed on the market at large which ultimately result in harm to end users. (More on how this requirement has derailed recent tech-related antitrust cases can be found here.) Failing a showing of such a linkage to consumer harm, the government’s claims of anticompetitive effects will face an uphill battle under the existing case law.
The government will have to rebut showings of how Google and Facebook made markets more competitive (pro-competitive effects).
The government’s troubles will not end in trying to convince a court that they present legally sufficient—and factually supported—theories of how defendant’s conduct harmed competition. They will also have to respond to a mountain of evidence and argument put forth by Google and Facebook about the ways in which their conduct has benefitted competition and achieved pro-consumer outcomes in the market.
The assessment of these so-called “pro-competitive effects” (the counterweight to anticompetitive effects) will be critical to the outcome of the antitrust lawsuits against Google and Facebook. The so-called “rule of reason” framework governing monopolization claims requires courts–conceptually, at least–to “weigh” the harmful effects to the market caused by the conduct in question against any legitimate beneficial effects to the market also resulting from it. If that sounds difficult, that’s because it is. (I’ve written about the challenges this poses in antitrust cases here and here.)
The practical effect of this framework is that a defendant able to point to legitimate, reasonable examples of pro-competitive justifications for its conduct wins the case so long as the plaintiff is not able to show those justifications to be either false or otherwise outweighed by anticompetitive effects.
When it comes to end users, Facebook will point to its social media network’s immense popularity, growth, and innovation. It will also probably argue that bringing WhatsApp and Facebook under one roof benefited consumers with more integrated and seamless functionality. Google will also point to the popularity of its products and services. It will further likely argue that the preferential placement of its search engine on devices or of its shopping aggregator at the top of search results made for a more seamless user experience. It could also stake claim to having helped create the vast ad-based machinery that generates endless online content available to users at no out-of-pocket cost. And the steps it has taken to limit third-party browser cookies on Chrome could be defended as a competitive response to similar restrictions on Apple’s Safari browser and as a pro-privacy benefit to users.
With respect to advertisers and publishers, Google and Facebook alike will be able to argue that their innovations in online advertising revolutionized the industry and are responsible for generating the bulk of revenue which sustains the business models feeding free online content to users. In particular, they could point to the technology they developed for gathering and analyzing user data and then using it to target ads to the consumers most likely to engage with them. On top of that, they developed the complex online auctions that connect those who want to buy ad space with those who want to sell it. With Google, in particular, where the very structure of the company seems to underpin the government’s allegations of monopolization, the defense will likely focus heavily on how its vertical integration across the entire “ad stack” has generated efficiencies and lowered fees for advertisers and publishers.
The government plaintiffs—who bear the ultimate burden of proof—in the antitrust lawsuits against Google and Facebook will have to overcome strong arguments that the alleged exclusionary conduct was justified by its pro-competitive aims. This is as true for the claims that concern harm to competition at the end-user level (people who run searches on Google or have a Facebook account) as it is for the claims concerning business-user level (advertisers and publishers using the ad platforms offered by Google or Facebook).
And that’s the situation if the case law is read favorably for the government. There is some recent precedent which the courts could rely upon to make things even more difficult for the plaintiffs. A recent Supreme Court ruling overturning an antitrust case brought by the DOJ against American Express, for example, could allow for a free-for-all competitive effects balancing exercise that links up the various interconnected markets of the vast digital ecosystems that Google and Facebook operate in. (I cover that and other related rulings here)
Under such an approach, the court could take into consideration pro-competitive effects felt by end users when determining the net impact of anticompetitive effects felt by business users. For example, benefits to Facebook’s users could be weighed against harms caused to advertisers. Or, benefits for Google’s search users weighed against harms to its publishers. If the courts apply this sort of open-ended balancing test, Google and Facebook will surely benefit, as the government’s strongest evidence of harm to competition will likely involve rivals while the companies’ strongest arguments for healthy competition are likely to involve end users.
The plaintiffs have to show that it was the exclusionary conduct, and not something else, that harmed competition (causation).
Another tricky aspect of proving harm to competition will involve establishing “causation”. This is a legal question not often at the heart of a dispute in monopolization cases but which may play an outsized role in the antitrust lawsuits against Google and Facebook.
In the course of analyzing whether competition was harmed by the actions of Google and Facebook, the courts will have to be convinced that it was the alleged conduct, and not some unrelated intervening factor, that caused the anti-competitive effects in the market (by enhancing or maintaining the defendants’ monopoly positions).
It is not well-established in the law what precise causation standard should be applied. It could require showing that harmful effects to competition would not have occurred “but for” the exclusionary conduct of the defendant. Put another way, plaintiffs (bearing the ultimate burden of proof) might have to show that competition would have been stronger if not for the defendant’s actions. This is an especially tough bar to meet in difficult-to-predict digital markets.
Alternatively, the courts could apply a causation standard that requires a less definitive showing of the linkage between the exclusionary conduct and the observed deleterious market effects. For example, in a seminal antitrust case against Microsoft brought by the DOJ, a federal appeals court held that judges may “infer causation when exclusionary conduct is aimed at producers of nascent competitive technologies … [or] producers of established substitutes” because the “defendant is made to suffer the uncertain consequences of its own undesirable conduct.” The Supreme Court has not weighed in on this issue.
Which causation standard the courts apply in the antitrust lawsuits against Google and Facebook (and different courts could apply different standards) could have a major impact on their outcomes. Google and Facebook could have strong arguments for why their conduct was neither the sole (or even primary) cause of the anticompetitive effects observed in the market, nor, for that matter, their alleged dominance in the market. They could argue that any effects seen in the market as well as their market positions are not due to the alleged exclusionary conduct but rather are the result of innate conditions of the digital ecosystems in which they operate that are outside of their control.
There is nothing controversial about the notion that platform markets such as the ones Google and Facebook operate in have unique traits that drive them towards near-monopoly conditions and make a strong market position, once obtained, difficult for rivals to un-do. Economists and policymakers agree on both this fact and its root causes (such as the “network effects” dynamic by which users flock to common services which, in turn, further entrenches an incumbent’s advantage). The controversial part is how to factor that into the competition analysis, a question not well tested by courts. But it’s an issue the judges hearing the antitrust lawsuits against Google and Facebook will have to grapple with if they are to avoid outlawing the mere possession of a monopoly—a ruling the antitrust laws would not support—as opposed to one obtained by competition not on the merits.
The antitrust lawsuits against Google and Facebook could face political headwinds and judicial delay.
Antitrust cases last many years. That makes those brought by government plaintiffs, especially ones like the Google and Facebook lawsuits which are brought by officials who are elected (State Attorneys General) or politically appointed (the Assistant Attorney General of the DOJ’s Antitrust Division) vulnerable to political change. An aggressive case can be whittled down over time if support for its prosecution wanes, especially if its resolution is subject to a negotiated settlement.
In the end, who is in charge of a case at its later stages can have a bigger impact on the outcome than who was in charge when it was filed.
State-led cases are especially prone to fragmentation. If local interests prevail, defendants under investigation can take a divide-and-conquer strategy that pits key states against each other in settlement negotiations (for example, if they vie over concessions that send investment dollars or jobs to their states). Politics and diverging views can derail federal cases, too. Microsoft’s prospects in the last big federal tech antitrust investigation waxed and waned with three Presidential administrations. Google avoided an antitrust lawsuit during the Obama administration thanks at least in some part, it has been reported, to its lobbying efforts.
For now, a loose cross-ideological consensus exists within American politics that antitrust laws should be used to rein in Big Tech. If it holds, the antitrust lawsuits against Google and Facebook have decent prospects for making it to some final resolution—whether a court win, a court loss, or an out-of-court settlement. But the political parties’ views could morph over the long lives of these cases and introduce divergent interests. Political divisions already seem to have caused different states to bring separate state-led lawsuits against Google. Add to that the uncertainty of who will control Congress and the Presidency when it comes time to negotiate a settlement or take a case to trial, and it quickly becomes clear that political headwinds could emerge that slow these cases down or outright kill certain aspects of them.
Tensions between federal and state interests can also send antitrust cases sideways. The most recent example is the T-Mobile/Sprint merger, where a group of states filed suit in federal court to oppose a merger that the DOJ had agreed to permit subject to certain conditions. (The states lost their case.) Presently, the FTC has filed a case against Facebook, and the states their own. With Google, a mix of states joined the DOJ case, but others joined one of two separate state-led lawsuits. The potential for conflicts and diverging interests in the litany of antitrust lawsuits against Google and Facebook is perhaps unparalleled in any previous major case.
Finally, even if some or all of these cases get resolved in court, a drawn out appeals process adds to the delay and uncertainty. At any step along the way, a hostile court can un-do everything with a few lines. Most recently, a decades-long antitrust battle by the DOJ against American Express was tossed out by the Supreme Court. And the FTC’s five-year battle with Qualcomm was thrown out by an appeals court. Both cases, notably, involved decisions expressing skepticism about monopolization claims, which is what the government plaintiffs allege in the current antitrust lawsuits against Google and Facebook.
It’s all about the remedy, stupid!
To speak about remedies last is to bury the lead in any analysis of a monopolization case like the antitrust lawsuits brought against Google and Facebook.
An antitrust plaintiff can have all the grandest intentions when filing a lawsuit, but it’s the remedy achieved at the end that determines how that case will impact competition, market conditions, and the parties to the case. The halls of antitrust’s history are full of well-intentioned, well-founded, ground-breaking cases that ended with a whimper when it came to the remedy.
In the antitrust lawsuits against Google and Facebook, the plaintiffs threaten drastic relief, including breaking up the companies. Whether this is posturing for a settled remedy or the true aim of the lawsuits, antitrust laws set a high bar—and little in the way of recent precedent—for the courts to order divestitures of these businesses. And since we can presume that neither company is going to agree to its own disintegration in a settled resolution, any attempt to impose such relief would probably get litigated all the way up to the Supreme Court.
The break-up remedy may very well fail when judged on the merits. For one thing, there is the big open question underlying much of the analysis in these cases: to what extent do the observed outcomes reflect the natural state of things for digital platform markets? This goes back to the issue of causation—whether it is the defendant’s conduct that caused the alleged dominance and deleterious competitive effects—but now in the context of fashioning a remedy. Because if it turns out that the ecosystems that Facebook and Google operate in naturally tend towards the observed market outcomes—not through exclusionary conduct but by the preferences of its customers or the nature of how the products are developed and offered—then breaking them up might do more harm than good. Market players might then expend time and resources to reconstruct the same conditions again. (This a criticism of the last major effort in antitrust to break up a company, when AT&T’s local phone service business was forcibly divested from its long-distance business, only to see the “Baby Bells” reabsorbed into AT&T in the decades that followed.)
Moreover, divestiture does not seem to best fit the theories of the cases brought by the plaintiffs. In the Google cases, so-called “conduct” remedies—such as forced rules or codes of conduct in the fair treatment of auction participants, prohibitions against imposing restrictions on the use of competing ad services, requirements for sharing data and access to key auction systems, cease and desist orders against implementing restrictions on third-party user tracking or paying devices manufacturers for default search placement—would seem more germane to the alleged unlawful conduct than would “structural” relief. In the Facebook cases, conduct relief such as forced interoperability of its services with third parties (including competitors), making user data portable to competing platforms, and perhaps restrictions on collecting and using data across different services, would align more closely with the antitrust theories currently pled by the plaintiffs.
If Google and Facebook are successful at avoiding the worst-case scenario of having divestitures imposed on them, that may be enough for them to “win” on the remedy endgame. That is because even if they cannot fully control what the terms of the relief might be, they can control a lot about its timing. The reality is that given the tedious pace of antitrust litigation, the historic scale of these massive cases, and the resources available to the companies to defend themselves, Google and Facebook can, if they want to, buy themselves many years before having to sit down at the negotiation table to hash out an arrangement.
The further out they can push those negotiations, the more time the defendants will have to adapt their businesses in advance to any changes forced upon them by conduct remedies. The ability to put off the ultimate day of reckoning, therefore, helps the companies smoothen the transition to a new state of play. It also gives them time to make changes that improve their prospects in the remedy stage of the litigation. This could include, for example, taking steps to moot the need for certain injunctions by changing their businesses (or even the industry at large). This has already happened, for example, with some of the claims brought against Facebook, as industry initiatives for data portability and interoperability create new market norms which may soften the blow of the relief negotiated with the government or imposed by a court in the pending antitrust lawsuits.
Another example of how the defendants could benefit from biding their time in a drawn out antitrust litigation is to make changes to their businesses that would make the legal basis for certain relief more difficult. Facebook, for example, has been accused of rushing to more fully integrate WhatsApp into the Facebook platform as a way to head off a potential forced divestiture in the government’s antitrust cases (or by Congressional act).
The tactic is a familiar one for companies facing antitrust liability in the deal context. Merging parties who, having consummated a transaction which they think could be challenged under the antitrust laws, may quickly move to “scramble the eggs” by fully combining their companies in ways that might make it difficult for a government agency or court to later impose a divestiture as a remedy. In the antitrust cases against Google and Facebook, any remedy that could un-do product integration that has benefitted the user experience will be a tough sell for the plaintiffs to make to the court (given, as discussed above, that consumer welfare is the gold standard for enforcement of the antitrust laws).
The government lawsuits against Google and Facebook as catalysts for legal reforms: antitrust and beyond.
When all one has is antitrust, everything looks like an antitrust problem.
Although brought under the antitrust laws, some of the arguments at the core of the Google and Facebook lawsuits ultimately raise issues more befitting of industry regulation or agency rulemaking. Issues in the Facebook complaints surrounding user privacy and fair treatment of third-party services, which are an awkward fit for the modern antitrust framework, may be better suited for a national data privacy law or digital consumer protection rules. Allegations in the Google complaints of “conflicts of interest” and “insider trading” raise issues that would seem more suitable for industry regulation akin to what is seen in financial sectors. The Texas complaint is surprisingly explicit in analogizing its claims to financial and securities crimes, which exist only because industry-specific rules and regulations (not the antitrust laws) make them unlawful. With no such regulations yet in place in the tech sector, the complaint could be characterized as aspirational but lacking an existing legal basis.
This isn’t just a matter of semantics. These are technical challenges facing the government’s legal claims that will bolster the defense’s arguments that the conduct at issue is not the sort that competition laws were intended to reach. Of course, this is not to preclude the possibility that the cases against Google and Facebook become a turning point in antitrust that leads to new legal standards befitting of a modern economy. Competition laws only evolve if plaintiffs–most often, the government–are willing to push the envelope and are able to convince courts to accept new ways of proving a violation existing laws. But that can be a challenging and slow process. More often, legal rules and standards are methodically shaped by courts over many cases spanning several decades. The precedent established by any one case can be interpreted narrowly by future courts, resulting in limited general applicability to an industry at large. Market-wide change from antitrust enforcement comes very slowly.
For large-scale shifts in how market players act, new legal frameworks would likely have to emerge. And so the biggest mark left by the antitrust lawsuits against Google and Facebook might turn out to be as catalysts for broader legal reforms. The EU, for its part, already went down this road with Big Tech and has reached the conclusion that antitrust has limited power over industrywide conduct in tech markets (for that, it is crafting industry regulations that will govern data privacy, data use/sharing, fair treatment of platform users, and so on). In the US, potential reforms of existing antitrust laws were set out in a recent report published by the U.S. House Antitrust Sub-Committee, discussed in recent follow-up hearings, and partly incorporated into the proposed Competition and Law Enforcement Reform Act.
The recommended reforms include reversing Supreme Court rulings that limit the reach of monopolization laws, expanding merger laws to block more deals involving nascent competitions, and even to legislatively break up certain companies. Identifying the gaps in antitrust could also lead to new laws, regulatory frameworks, and even agencies being established to deal with such issues as data privacy, treatment of platform users and competitors, and conflicts of interest. (I cover this in more detail here and here.)
So while the government lawsuits against Google and Facebook will no doubt be historic, it may be for ushering in legal and regulatory changes outside the competition realm rather than as antitrust trials-of-the-century that result in major precedent-setting wins for the plaintiffs.