Reports are circulating that the FTC is having doubts about its antitrust case against Google. This is not surprising. There are many hazards of a case against Google, including the difficulty in identifying consumer harm, the logical paradox of "search neutrality," and the inappropriateness of expanding "duty to deal" case law. And of course, even if the FTC can overcome these problems, the question of prospective regulation moving forward remains.
Over two years ago, I discussed the idea of a government imposed search regulator. The notion appeared so farfetched that some commentators insisted this idea is a straw man, a ploy to dismiss the arguments made by Google's competitors by exaggerating their claims. A Wired article came to the same conclusion, commenting about the notion of search neutrality with, "It's hard to tell if [search neutrality] was ever a serious proposal, since it was most often deployed by ISPs as a sort of reductio ad absurdum against network-neutrality proposals." But it was clear the complaints levied by Google's competitors could only lead to one result and I posed questions to help others realize what road we were on:
- How are you going to regulate search?
- Are you going to demand that every provider follow the same algorithm?
- Are you going to allow a regulator to see the algorithm and verify that it is fair?
- What constitutes "fair" search results?
- Will the regulations governing search be static or dynamic? If static, how do we anticipate innovation? If dynamic, who chooses what changes are appropriate?
- re we only going to target Google? What happens if Bing surpasses Google and becomes "dominant"? Do they trade places?
- Where should search engines direct consumers?
Regulating search seems like regulating thought, a simply impossible task and a venture not worth pursuing.
And then there are the institutional problems. We have no search regulator (and few in the halls of Congress would welcome creating another regulator). The FTC is investigating Google but it is an enforcement agency, not a sectorial regulator, and simply lacks the resources to regulate search in any fashion, much less a meaningful one. When the FTC brings an enforcement action its remedies are inevitably self-executing, since the FTC cannot police a market. And to make the FTC the national nanny of search would clearly be beyond its capability, and would be an extraordinarily ill advised use of limited government resources.
Last week, FairSearch (comprised entirely of Google competitors, including Microsoft, Expedia, Hotwire, Foundem, and TripAdvisor) published its "proposed remedies" for the Federal Trade Commission and European Commission to follow in their efforts in "restoring and preserving competition in search and search advertising... which can help ensure continued innovation and opportunity across online markets while reducing the need for competition agencies to monitor and enforce behavioral restrictions." Unsurprisingly, they call for an awful lot of monitoring without actually ever answering the difficult remedy questions that are an absolute predicate for government intervention.
FairSearch breaks its proposal into three categories: Behavioral Remedies, Structural Remedies, and Monitoring and Enforcement Mechanisms. Marvin Ammori of the New America Foundation already debunked thoroughly the proposed behavioral remedies, and I see no need to repeat his effort. Instead, I will focus on what FairSearch has termed Structural Remedies and Monitoring and Enforcement Mechanisms.
Let's start by remembering what the law instructs on the limits of a court or antitrust agency to remedy an alleged illegal monopoly. Courts have appropriately instructed that a remedy must be narrowly focused to cure the competitive harm. That's why the typical remedy in a dominant firm case is behavioral -- to prevent the specific anticompetitive conduct. The DC Circuit in the U.S. v. Microsoft case rejected structural relief, advising "[a]bsent some measure of confidence that there has been an actual loss to competition that needs to be restored, wisdom counsels against adopting radical structural relief."
As former Assistant Attorney General and FairSearch counsel Tom Barnett has instructed, structural remedies "should be used sparingly... and only where there is a clear, significant causal connection to the acquisition and maintenance of monopoly power, and even then only after determining that alternative remedies would not satisfactorily achieve the proper goals of a Section 2 remedy."
Despite their claim to limit "the need for competition agencies to monitor and enforce" the industry at the conclusion of this investigation, FairSearch's proposed remedies are precisely the type of radical structural relief rejected by the antitrust agencies and the courts.
FairSearch's Proposed "Structural" Remedies
1.Requiring Google to license data that it has obtained in significant part through its improper actions
Antitrust law deals with this proposed remedy efficiently. Since the landmark Supreme Court case of Verizon Communs. v. Law Offices of Curtis V. Trinko, LLP, a market leader, even a dominant firm, has no duty to deal with its rivals unless there was a prior relationship or the dominant firm is withholding an otherwise publicly accessible good. There is no qualifying prior relationship, as Google's competitors' placement on Google search results was not related to Google's provision of data. Second, the data in question is a valuable commodity acquired "at considerable expense and effort" just like Verizon's telecommunications services.
Preemption through case law is just the beginning. Think about how many questions must be answered to consider implementing this "structural" remedy. What data is FairSearch alluding to? Which competitors have a right to access? And how would we implement this access? Who makes these decisions?
Let us not forget most important question posed by this proposed remedy -- what is "improper actions?" Who determines what is improper? Is the measurement net consumer welfare? Harm to a specific competitor? Shrinking of the industry? Would certain data be exempt from this de facto compulsory license if Google could show it was not obtained through improper actions?
We simply cannot consider illusory "remedies" that provide no solution while introducing a mountain of new problems and questions.
2.Requiring Google to divest its vertical products that have benefited from Google's abuses
Again, there is little legal support for such a substantial remedy. The law provides that a court can prohibit certain practices it has found anticompetitive; it does not permit a punitive remedy that tries to root out all the alleged benefits of any alleged conduct.
Moreover, just as the first proposed option, this structural remedy poses more questions than solutions. Vertical search is not a product. It is not a standalone good, but rather a method of perceiving the only good being provided to consumers -- information. But, arguendo that a vertical product is a definable asset, this suggestion still is froth with problems:
Who would gain control of the divested product? Would only products that the government can identify as benefitting from Google's abuses be divested? What constitutes an abuse sufficient to force divestiture? Once divested, would Google need to continue servicing the vertical product for its recipient-competitor?
The point is not that Google has YouTube, or Google News, or Google Places. The point is that Google is able to provide information accurately and reliably to people who want it. Stripping them of a vertical "product" does not strip them of the business acumen they have accumulated, and does not mean they would struggle to provide this service. On the other hand, FairSearch could be advocating not only a divestiture, but also a restriction from the provision of certain types of information. I struggle to see how this would be in the best interest of the consumer. And this can't be what FairSearch truly believes is best for competition.
FairSearch's Proposed Monitoring and Enforcement Mechanisms
The FairSearch proposals really are the worst of both worlds. They are simultaneously over-burdensome and ineffective. It would be impossible for the government to adequately train and staff a "Technical Monitor" to review Google's algorithms and software updates. This is the type of solution that someone with absolutely no knowledge or understanding of the complexities of software or search functionalities conjures up.
The current Director of the Bureau of Economics for the FTC Howard Shelanski has commented about the need to carefully assess the costs of any remedies. "Pure conduct remedies would clearly involve enforcement costs... so it is unclear why a structural remedy that also entails conduct restrictions would not also have such costs." He then explained that the only way to avoid a remedy that is more costly than the harm it is trying to prevent is 1) tailor a remedy to the alleged anticompetitive act, and not to conduct that is alleged but not established, or possible but not yet occurring; 2) minimize administrative costs 3) adopt a remedy that prepares for the possibility for technical change; and 4) ensure the remedy does not encourage rent-seeking (as Commissioner Rosch defines it "petitioning of the responsible agency by market participants for action against others, namely, their rivals, in order to obtain a market advantage or to advance a personal agenda.").
FairSearch's proposed remedies fail all of these tests, and the harm from these proposals would certainly outweigh any benefit. In the end, these proposed solutions fall short of both logic and the law. They are simply a wish list for competitors, desperately asking the government to make the successful firm give up its valuable assets and restrict its ability to compete. The most pressing hurdles to a government case remain unresolved.
Co-authored by Brendan Coffman
David Balto is a former policy director of the Federal Trade Commission, attorney-adviser to Chairman Robert Pitofsky, and antitrust lawyer at the U.S. Department of Justice. He has been a senior fellow at the Center for American Progress and has worked with the International Center on Law and Economics, both of which receive funding from many organizations including Google. Mr. Balto has also published research and authored scholarship for Google on technology policy topics.