The caution of Judge Amit Mehta’s remedy in the Google Search case is unlikely to open internet search to competition. Steve Salop recommends several amendments to the remedy that can improve competition without undercutting the revenue that has benefited Google’s partners to date.
Judge Amit Mehta’s remedy in United States v. Google did not prohibit Google from sharing revenue with third-party search distributors like Apple and Mozilla in order to purchase default search status on browsers and devices, despite the Department of Justice successfully showing that the strategy helped Google entrench its monopoly in internet search. One key reason was Mehta’s concern about the hardship that it would place on search distributors who received a share of Google’s monopoly profits. Yet, there is a better way that may have been overlooked or given short shrift. The remedy could have prohibited this conditional revenue-sharing and instead required Google to pay the distributors fees for all the searches they generate, regardless of the search engine used. This alternative remedy would better satisfy the remedial goals for monopolization conduct that has been found to violate Section 2 of the Sherman Act.
The courts have identified several economic goals for remedies in monopolization cases: (i) terminating the illegal conduct; (ii) restoring competition; (iii) denying the monopolist of the fruits of the illegal conduct, and (iv) deterring future violations.
Restoring competition eliminated by monopolization typically requires more than a narrow injunction of the illegal conduct. Undoing the effects of the monopolist’s entrenchment means that additional remedies are needed to “jump start” the restoration of competition. These may be additional injunctions, affirmative behavioral provisions, or even divestitures.
The deterrence goal is intertwined with the others. Deterrence is not served if the monopolist is permitted to continue the illegal conduct. Requiring the monopolist to disgorge some or all of these illegal fruits also serves the goal of deterring future anticompetitive conduct by both this monopolist and would-be future monopolists.
Remedy design in an uncertain world needs to balance the harms from mandating too much versus too little. The Supreme Court has connected judicial humility with doing “too little.” This is clear in Trinko and Alston, as Judge Mehta recognized. But “judicial humility” should not automatically imply inaction. Humility should counsel that when there is great uncertainty, a court should be extra careful not to err on either side. For example, it would not be a good exercise in judicial humility to respond to a series of fatal accidents by autonomous trucks to simply rely on competition among the truck manufacturers to solve the problem.
Judicial humility in antitrust demands that claims of rapid market self-correction should be taken with many grains of salt. In this regard, when there is an entrenched monopolist in a market with network effects and other barriers to entry, rapid market self-correction is highly unlikely. Faced only with a simple injunction, it is more likely that a monopolist will maintain the now-entrenched monopoly power it gained from its anticompetitive conduct for a significant period of time.
The overly optimistic Google Search remedy
The central provisions of the Google remedy involve: (i) an injunction against Google’s “exclusive” default agreements (e.g., an agreement with Apple for Google Search to be the default for all Apple search access points on all Apple devices) and (ii) limited data sharing with qualified competitors to improve their offerings.
An alternative remedy would have prohibited Google from making any revenue-sharing payments. Judge Mehta clearly explained the rationale for this remedy as follows: “The rationale for a payment ban is straightforward: It would pry open the market to competition. The revenue share payments shape the market for general search services in Google’s favor.”
However, the court did not adopt this remedy. It permitted non-exclusive default agreements with revenue-sharing. Google would also be permitted to share revenue for searches voluntarily directed to Google by the search browser or other access point owner, or additional searches facilitated by preferencing Google Search on the device, even if Google Search is not the default search app. This very narrow injunction that prohibits only broadly exclusive default deals is unlikely to restore competition. For example, the remedy would allow Google to purchase all the default positions for (say) 95% of the distributor’s queries, a point emphasized by Erik Hovenkamp and Douglas Melamed in ProMarket. It might even be able to purchase the default position for 100% of the search access points as long as it does so one search access point at a time.
Given Google’s initial advantages and the fundamental point of Richard Gilbert and David Newbery’s seminal model that a monopolist has an inherently greater payoff from preserving monopoly profits than a rival’s payoff from obtaining competitive profits, there are good reasons to expect that Google will have the ability and incentive to do so, as I explained in ProMarket. In fact, Judge Mehta confessed this concern. As he stated:
The court well recognizes what eschewing a payment ban may mean for competition. Due to Google’s massive financial advantage and its superior monetization, distributors will be incentivized to stick with Google because it can pay more, thus leaving in place the very forces that “effectively [have made] the ecosystem exceptionally resist[ant] to change.”
But Judge Mehta was reluctant to adopt this remedy for several reasons. First, he was concerned that the loss of revenue suffered by the search distributors who had deals with Google (Mozilla and other browsers; telecommunication companies like Verizon; and original equipment manufacturers (OEMs) like Samsung) would place a financial hardship on them. Second, he worried that their loss in revenue could lead to less innovation and investment, which could lead to lower consumer welfare. Third, he was concerned that Google would still get more of the searches so that the net effect would be to reduce Google’s costs without improving consumer welfare. Fourth, and likely especially important, the court was cognizant of the Supreme Court’s general advice to do less.
Finally, Judge Mehta wondered if artificial intelligence might cause the market to self-correct. As he noted,
Today, established technology companies are making, and start-ups are receiving, hundreds of billions of dollars in capital to develop GenAI products that pose a threat to the primacy of traditional internet search. … These companies already are in a better position, both financially and technologically, to compete with Google than any traditional search company has been in decades (except perhaps Microsoft).
Thus, he chose to bet the future of competition on his hopes for AI:
These new realities give the court hope that Google will not simply outbid competitors for distribution if superior products emerge. It also weighs in favor of caution before disadvantaging Google in this highly competitive space.
This hope seems more like over-trusting than humble but rational expectations. The remedy is not quite catch-and-release, but it does raise that concern. Despite the limited data-sharing remedy that will give potential competitors, including AI startups, access to Google’s data, AI is unlikely to rapidly catch up to Google. Google has more than a decade-long headstart in creating an entrenched monopoly protected by the network barriers to entry it created plus the detailed search data provided by users. Google’s headstart in consumer searches oriented towards choosing products and sellers might be difficult to be undone unless and until the AIs gain detailed personal information from users about their detailed shopping preferences, and this may take a significant period of time. Meanwhile, Google will continue to collect more information and anticompetitive monopoly profits.
A procompetitive remedial modification
The court’s decision to allow Google to continue to collect and share its monopoly revenues in order to obtain more searches is very worrisome. While Judge Mehta acts like he is being cautious, antitrust commentary makes the strong point that we should rely on competition, not permitting continued monopoly pricing to protect the distributors that were sharing the monopoly profits. As the court recognized, the remedy also will raise barriers to the competitive entry that is needed to restore competition.
With this in mind, I suggest the following modification of the remedy:
First, the court should prohibit Google from making any revenue-sharing or other payment deals that are conditioned on either default search or promotion status.
Second, since the court has expressed concerns about the impact on search distributors (i.e., browsers, telecoms, OEMs), the court should require Google to make payments to the distributors, but where the payments are based on the total number of searches, but not on the distributors’ choice of search engine (either as a default or for voluntary directed search queries). The payment per search can be based on Google’s pre-judgment per search payments to each of these distributors.
Third, if the court wishes to go further to jump-start competition, it could require Google to pay distributors’ a higher per search payment for searches made on non-Google search engines than those made on Google.
Fourth, I also recommend a prohibition on unconditional payments to distributors even for any queries voluntarily directed to Google search by search distributors.
Because Google’s monopoly power gives it greater benefit from deterring competition, these revenue-sharing payments also would encourage searches to be voluntarily directed to Google by the distributors. As a result, they would deter competition from being restored. While this modification is supported by the same reasoning (and the Gilbert/Newbery model) on which the prohibition of exclusive defaults is based, Judge Mehta concluded that the record on the need for this remedy was insufficiently developed, a condition that can be corrected with filings or a new hearing.
There are significant competitive benefits of this modification.
– This modification resolves the district court’s concern about the potential hardship to search distributors without permitting Google to continue to collect monopoly profits and deter competition with conditional revenue-sharing payments. Because the distributors (e.g., Mozilla, telecoms and OEMs) will continue to obtain payments per search, any incentives to pass-on the payments to their customers with lower device price and greater innovation will continue as before.
– By mandating a stronger payment ban injunction, this remedy provision is more likely to jump-start the restoration of competition in two ways. It reduces the distributors’ financial incentive to support Google’s monopoly power by using Google Search. It is also for this reason that I recommend Google be prohibited from offering revenue-shares for voluntarily directed searches that do not involve defaults. (Because Google’s monopoly power gives it greater benefits from deterring competition, these unconditional revenue-sharing payments encourage more searches to be directed to Google and thereby further deter competition from being restored.) The impact of these prohibitions on restoring competition is even stronger if the fourth prong of the proposed modification is included. This provision also supports deterrence by reducing somewhat Google’s incentive to prevent search engine competition.
– By removing the further fruits of Google’s illegal conduct until the market self-corrects, it also significantly strengthens deterrence of anticompetitive exclusionary conduct by Google and others.
In light of the public interest benefits of restoring competition and eliminating the illegal monopoly fruits, requiring these payments would be legally permissible under the broad equity power of the courts. While such unconditional revenue-sharing might not violate Section 2 in a well-functioning market, the prohibition here is being used as a remedy for an adjudicated monopolization case so the standard is different.
This modification could be adopted on remand or a rehearing later in the likely event that AI has not already saved the day. Judge Mehta actually left open this possibility, stating that “the court is thus prepared to revisit a payment ban (or a lesser remedy) if competition is not substantially restored through the remedies the court does impose.”
Author Disclosure: Steve Salop is Professor of Economics and Law Emeritus, Georgetown University Law Center and Senior Consultant, Charles River Associates.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
Subscribe here for ProMarket’s weekly newsletter, Special Interest, to stay up to date on ProMarket’s coverage of the political economy and other content from the Stigler Center.
