Herbert Hovenkamp applauds the Biden administration’s antitrust authorities for intervening in labor markets and more robustly challenging mergers between competitors. However, the next administration should clarify in its guidance that the objective of stronger antitrust enforcement must focus on lowering prices, increasing output, and removing any restraints on innovation.
Editor’s note: This article is part of a symposium that asks how the next presidential administration and its antitrust agencies should reorient competition policy for the next four years. Contributions from Herbert Hovenkamp, John Kwoka, Steven Salop, and Ginger Zhe Jin and Liad Wagman can be read here as they are published in the coming days.
Where did the Biden administration’s antitrust agencies perform well and where could they have done better? Much of public antitrust enforcement is discretionary. The Federal Trade Commission and Department of Justice Antitrust Division (the Agencies) must triage among priorities and available resources. They perform best in markets prone to collusion, with rigid market shares, stagnant technology, little new investment or entry, prices that are persistently high in relation to costs, and where private restraints on competition appear common. Enforcement in these markets produces prompt and visible benefits. The horror to be avoided for the Agencies is the failed litigation against IBM in 1969. This was largely a challenge to IBM’s rapid innovations that made computers faster, cheaper, and more integrated. After thirteen years of litigation (1969-1982) the government acknowledged failure and dropped it. One federal court observed, “to the best of our knowledge no litigation has taken so much time and involved such expense.”
The Biden administration antitrust policy has a mixed report card. It deserves praise for suing RealPage for algorithmic manipulation of rentals rates. The litigants will likely dispute the Sherman Act’s narrow and restrictive “agreement” requirement, but the case promises a valuable payoff in lower housing costs as well as insight into the use of technology to manipulate prices.
The administration was also wise to target employer restraints on labor. The labor share of production has been falling for decades. Wages lag behind economic growth. Collusion (anti-poaching) among employers occurs frequently and even some federal judges fail to appreciate its consequences. Labor markets are excessively subject to employee mobility restraints such as noncompete agreements, and mergers can harm workers as much as consumers. In these cases antitrust remedies can produce visible improvements: cheaper housing, better wages and working conditions, and greater employee mobility.
By contrast, tech has been among the fastest growing parts of the United States economy. Entry by new firms is more robust than in other parts of the economy. It has brought lower costs and prices to many areas of business and is highly innovative. Of the top-300 patent acquirers in the U.S., Alphabet is #8, Apple #9, Microsoft is #17, Amazon #18, and Meta is #41. Tech firms have well-educated and well-paid work forces. They rank high in consumer satisfaction. Collusion does not appear to be a particular problem, although joint agreements require oversight. These firms compete with one another aggressively in developing new technologies.
Just as with all industries, Big Tech needs monitoring for anticompetitive restraints, but it should not be singled out and there should be no bias in favor of structural, or “breakup,” remedies. Amazon gets bad press for high fees to third-party sellers, but compared to what? Startup businesses who sell on Amazon have a lower failure rate than new businesses generally. In tech as everywhere else, it is always important to ask, “who is getting harmed?” and “how, and what are the alternatives?”
Presidential candidate Kamala Harris has recently received pushback for attacks on price gouging in grocery prices. Some of the rhetoric was inartful, but the underlying concerns are important. She should stick with them and emphasize antitrust’s role in mitigating price gouging. Since its inception, antitrust policy has been focused on ensuring lower prices, higher output, and asset mobility. While Harris’s “opportunity economy” slogan speaks differently to different people, it invokes market-driven values: competition, asset mobility, and low entry barriers and switching costs. These are areas where smart antitrust policy can be effective.
Lessons from the administration’s biggest success: Google Search
Perhaps the Biden administration’s biggest antitrust success is the 2024 Google decision. The case represents a good use of agency enforcement resources. Google was found to have a monopoly position in Google Search and paid device and browser makers for default search engine status on their products. Given Google’s high market share, that was not a particularly difficult case under existing law.
One big accomplishment in the search case was the Justice Department’s use of behavioral economics, a sub-field that has been clawing for antitrust recognition for years but has seldom attained it. The issue was exclusive dealing, which under the traditional, or neoclassical, view refers to prohibitions but not to “defaults.” The Clayton Act applies to requirements that a buyer “not use or deal in” a competitor’s goods. A default is much softer. A firm such as Apple obligates itself initially to direct traffic to Google Search, but users are free to substitute a different search engine. As the court observed, however, in most cases they do not. The defaults are “behaviorally” exclusive, particularly on small devices, even if they are not exclusive as a matter of contract law.
The court also affirmed that monopoly power attaches to products, not to big firms. This is noteworthy because “monopoly” is an often misapplied term. Microsoft is bigger than Google by market cap. But its Bing search engine struggles with a market share of around 6%. Consistent with this decision, Microsoft can continue making Bing the default search engine on its Edge browser and devices or even form a search engine joint venture with Apple or someone else. The decision also held that Google did not have monopoly power in general search advertising. The Supreme Court has always based antitrust power inquiries on a firm’s position in a particular product. Two-sided platforms pose problems because market shares often differ on the two sides. Even in tech, however, antitrust policy should focus on market-dominating products and anticompetitive acts, not simply on big firms.
The court also rejected an attempt to require Google to share its products with rivals. The state plaintiffs had claimed that Google unlawfully refused to share its advertising platform, SA360 (“Search Ads 360”), which had evolved out of Google’s acquisition of DoubleClick in 2008. SA360 permitted users to purchase advertising simultaneously across multiple platforms. The Supreme Court’s very restrictive Trinko decision foreclosed the refusal-to-deal claim, however, as Judge Amit Mehta observed in his Google ruling. That portion of the Google case may dampen other antitrust litigation, particularly against Apple. There, the government challenges Apple’s refusal to share technology with others. Apple’s defense relies heavily on Trinko.
Is the Supreme Court duty-to-deal doctrine too restrictive, particularly for digital networks that require cooperation among firms? I believe that it is, but the more immediate question is whether challenging it is worth the enforcement Agencies’ commitment of resources, given the Supreme Court’s current personnel.
The 2024 Google decision addressed only liability. A remedies decision is expected in August, 2025, but the court gave some hints. First, Google had produced an extremely valuable product and had continuously innovated to make it better. But second, it had purchased default status from numerous implementers. Both contributed to its large base and desirability. Sorting them out raises issues about causation. The court quoted the Microsoft case’s observation that if the plaintiff “is seeking only injunctive relief,” the standard for causation is “somewhat relaxed.” By contrast, a structural remedy (breakup) can be “imposed only with great caution,” requiring a “significant causal connection between the conduct and creation or maintenance of the market power.” Further, any structural remedy must be reasonably expected to “restore competitive conditions,” as Microsoft put it. It must provide a realistic promise of lower prices (where relevant), higher output, or a better consumer experience.
A pervasive problem for antitrust remedies is eliminating the bad conduct without killing the effects of the good. If a large firm has economies of scale or scope but also engages in exclusionary practices, a breakup could end both. By contrast, a properly tailored injunction can be limited to the bad practices. If a Google divestiture benefits Microsoft, its principal rival in search, but makes user experiences no better or perhaps worse, it will be judged a failure. One lesson from the successful first stage of the Google litigation is that finding liability is often much easier than providing the right fix. In Section 2 cases in particular, injunctions have fared much better than breakups.
Tech platforms also exhibit other problems. The fact that they can be accessed at home and on smartphones increases the risk of deception and predation, particularly for children or other vulnerable groups. Further, as a communication medium, ecommerce raises problems more appropriately addressed to the Federal Communications Commission. These are rarely antitrust matters and are best left to other agencies and bodies of law.
Revisiting the 2023 Merger Guidelines
Another thing the Agencies should do after the election is write a commentary, or addendum, to the 2023 Merger Guidelines. They have done so in the past. In some ways the 2023 Merger Guidelines were an improvement over its 2010 predecessor. In others they were a step back. Good Guidelines should provide guidance. The 2023 Guidelines do an ample and even excessive job of pointing out harms, but they are not good at providing assurance when mergers are competitively beneficial. Condemning an economically beneficial merger can be just as harmful as failing to condemn a harmful one.
The 2023 Guidelines were wise to set lower thresholds for challenging mergers of competitors. Those in the previous Guidelines were too tolerant of mergers between firms with large market power. That said, the Guidelines do a poor job of relating economic evidence to enforcement thresholds. Numerous economic studies do far better, but the Guidelines did not discuss them. These studies provide useful benchmarks for Agency decisions to challenge. They also note a substantial role for merger efficiencies. Even in concentrated markets, as many as one-third of mergers result in lower prices. The hard part is predicting them. That places a premium on getting it right.
The Guidelines properly discuss the impact of mergers on labor, although they should have included a comment about the relationship between product and labor markets. Mergers that result in higher prices and lower output in product markets typically harm labor markets as well.
The 2023 Guidelines also did well to be less distinctive about “vertical” mergers. Instead, they were grouped with mergers that make it “harder for rivals to compete.” These are better classified as mergers of complements. Then they should have been more forthright in stating that these mergers as a group are far more likely to lead to lower costs and prices, and less likely to produce competitive harm. A few of them can be anticompetitive.
The same thing is true of two categories of mergers first recognized in the 2023 Merger Guidelines: two-sided digital networks and ecosystems. The new Guidelines did well to include them, but the presentation is too one-sided. Most mergers in these two areas produce improvements in operation, cost reductions, or better user experiences rather than competitive harm. They should have been presented in that way. For example, a merger that actually integrates two networks can provide broad benefits from network effects, even if overall market concentration levels are relatively high.
The Agencies should also rescind the 2023 Guidelines’ suggestion of closer scrutiny if an industry shows a “trend” toward concentration or vertical integration. There is no economic support for such a policy. The two dominant explanations for “trends” toward concentration are technological change, which often leads to larger firms with lower costs, and also differential rates of innovation, which tend to favor more innovative firms. In both cases prices go down or product quality or innovation rises even as markets become more concentrated.
As for “trends” toward vertical integration, vertical mergers often enable cost reductions from either technical change or savings in transaction costs. Even the Brown Shoe decision recognized this in the 1960s, approving the lower court’s decision that the vertical merger resulted in “lower prices or better quality for the same price.” When one firm’s vertical merger improves performance in these ways, competition will force others to follow, causing a “trend” toward vertical integration.
Finally, the Merger Guidelines should rationalize merger enforcement by emphasizing that the underlying purpose of all merger law is to identify mergers that threaten higher prices, reduced market output, or threats to innovation. That would give enforcers and litigators a target consistent with long-standing antitrust goals and more in line with what the public is expecting.
These things should be on the next president’s agenda. Given the public’s focus on low prices, high output and unrestrained innovation, few tools are better placed than antitrust for addressing them.
Author Disclosure: the author reports no conflicts of interest. You can read our disclosure policy here.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
