The California Law Revision Commission (CLRC) is currently reviewing proposed amendments to the state’s antitrust statutes, particularly the Cartwright Act. As made clear in a recently published memo, a major goal of the effort is clearly to distance California from the perceived constraints of federal antitrust law that limit liability for single-firm conduct under Section 2 of the Sherman Antitrust Act.
California intends to achieve this by strategically overturning specific U.S. Supreme Court decisions and departing from the error-cost framework that has traditionally shaped federal antitrust analysis. The cumulative effect of California’s proposed amendments—but particularly the section on single-firm conduct—would be to align California antitrust law more closely with EU competition law, where dominant firms must actively give a leg up to competitors (see the International Center for Law & Economics’ comments to the CLRC proceeding here and here, and ICLE President Geoff Manne’s August 2024 presentation to the commission here).
Abandoning the Error-Cost Framework
U.S. antitrust law traditionally operates within an error-cost framework, weighing the risks and costs of mistakenly condemning procompetitive behavior (Type I errors) versus mistakenly allowing anticompetitive behavior (Type II errors). Historically, there has been a preference for avoiding Type I errors, based on the view that the costs of chilling beneficial conduct are often more significant and harder to correct than the costs of permitting harmful conduct.
The reasoning for prioritizing avoidance of Type I errors echoes arguments like that given by Judge Frank Easterbrook. In a nutshell, Easterbrook argued that procompetitive conduct that is deterred by legal action has no constituency to advocate for its re-legalization, and improperly deterred firms have no standing to sue. Thus, if a court wrongly condemns a beneficial practice, those benefits may be lost permanently, whereas if a court wrongly permits a harmful practice, the market itself often works to correct the situation over time (e.g., through new entry attracted by high profits).
In stark contrast, the proposed legislative findings accompanying the CLRC memo explicitly state that courts should “liberally interpret California’s antitrust laws” and “be mindful that California favors the risk of over-enforcement of antitrust laws over the risk of under-enforcement.”
This position rests on the unsubstantiated assumption that the risk of underenforcement currently outweighs that of overenforcement. The contention is based on the idea that, because something feasibly might go wrong, there is no need to establish (or to even seriously inquire about) whether heightened antitrust scrutiny and expanded enforcement are genuinely warranted.
But the “more up-to-date economic analysis” that downplays the risk of Type I errors is, in large part, purely theoretical—rooted in abstract possibility theorems that ignore practical complexities and the real-world institutional context in which enforcement decisions are made.
For example, despite decades of such theoretical models, particularly regarding rational predation, little of this “post-Chicago School” learning has been widely incorporated into antitrust law, and there is virtually no empirical evidence of systematic problems for the conduct they question.
Intuition-driven concerns —i.e., vague suspicions that something must be amiss—pressed into service of a presumptively interventionist agenda do not provide a sufficient basis to abandon the cautious approach embedded in the error-cost framework, or support the assumption that underenforcement is a greater risk than overenforcement. That is as true for California as it is for the rest of the United States.
Targeting Key Supreme Court Precedents
The CLRC memo’s authors seek to overturn fundamental presumptions in U.S. antitrust law by aiming to nullify three key Supreme Court decisions: Trinko, Amex, and Brooke Group. This approach is misguided on both legal and economic grounds.
Nullifying Trinko: Implications for Refusals to Deal
U.S. and EU law differ significantly on the question of refusals to deal, with the United States strenuously limiting enforcement. In seeking to dispense with Trinko, however, the CLRC’s proposed amendment would move the Cartwright Act closer to the European Union’s approach. Under the EU’s lower threshold for liability, free riding is the norm, especially after the mind-boggling Android Auto ruling rendered by the European Court of Justice (see here and here).
As with other aspects of antitrust law, the U.S. approach to refusals to deal is rooted in the error-cost framework’s preference for avoiding Type I errors. Trinko held that enforced sharing “may lessen the incentive for the monopolist, the rival, or both to invest” and would require courts to act as “central planners, identifying the proper price, quantity, and other terms of dealing,” a role for which they are ill-suited.
The Aspen Skiing decision laid down one of the few exceptions, but the Court noted that it considered the case to be “at or near the outer boundary” of Section 2 liability, emphasizing the significance of a firm ceasing a voluntary and presumably profitable course of dealing. This, the Court noted, pointed to anticompetitive foreclosure as the only economically rational explanation.
This highlights a key feature of American antitrust law concerning refusals to deal; namely, that U.S. antitrust law generally does not apply the “essential facilities” doctrine—i.e., one company’s right to access the facilities of another. Indeed, as the Court held in Trinko, “we have never recognized such a doctrine.”
By contrast, EU competition law is much more interventionist, with refusals to deal being a staple of the system. While conditions for liability exist (indispensability, elimination of competition, blocking the emergence of a new product), they have been significantly relaxed in practice. For example, in Microsoft, showing a limitation of technical development was sufficient to prove that the refusal impeded a “new product” (here). Furthermore, the recent Android Auto ruling discarded indispensability, finding “mere convenience” sufficient to create an access obligation on a dominant firm.
Thus, whenever denying access to an app could reduce its appeal to consumers—and the platform is, in principle, designed to host third-party apps—access must be granted. In effect, this means that, once a dominant company (noting that, in the EU, dominance has been found with market shares as low as 39.7%) opens its platform to third parties, it largely surrenders control over access, retaining discretion only on the narrowest security-related grounds.
The CLRC’s proposal that liability should not turn on whether the defendant treated particular parties differently in exercising exclusionary conduct (including refusals to deal) is a further move away from effects-based analysis and toward the European model. By removing the protection for unconditional refusals to deal, it risks applying a standard meant for discriminatory conduct to all exclusionary behavior, thereby impinging on companies’ freedom of contract, encouraging free riding, and chilling innovation.
Abandoning Amex: Turning Back the Clock for Two-Sided Markets
The Supreme Court’s decision in Amex is considered uniquely important for the antitrust analysis of firms in the modern platform economy (see here). In Amex, the Court held that, in cases involving two-sided markets, plaintiffs must show harm on both sides of the market.
The CLRC’s proposed amendments would reverse Amex by suggesting that showing harm on only one side of a multisided market should suffice to prove antitrust liability. This would be fundamentally misguided. The economics of two-sided markets require considering both sides, as there is no meaningful economic relationship between benefits and costs on each side when considered in isolation.
Assessing harm on only one side will arbitrarily include (or exclude) users and transactions, and thus result in incorrect inferences about market power. Indeed, evidence of a price effect on one side can be consistent with neutral, anticompetitive, or procompetitive conduct: distinguishing among these in any meaningful way requires assessing price and quality on both sides.
Adopting an approach that permits liability based solely on harm to one side of a platform risks benefiting certain business users at the expense of consumers. This would further distance California antitrust law from the consumer-welfare standard, edging it toward a model where liability hinges on nebulous concepts of “power,” hunches, and the enforcers’ discretion in choosing which group of businesses to favor.
The critique of Amex stems from a broader opposition to vertical integration. Like other critics, the CLRC identifies the lenient treatment of vertical arrangements under U.S. antitrust law as a significant barrier to increased enforcement. U.S. law, however, favors vertical arrangements for a valid reason based on the error-cost framework. Empirical evidence consistently indicates that, while firms do employ vertical restraints, these practices seldom result in net anticompetitive effects. In many cases, in fact, they actually promote competition (here and here).
Forfeiting Brooke Group: Risking Harm to Procompetitive Pricing
The evidentiary standard for proving predatory pricing is deliberately high in U.S. antitrust law, which aims to promote consumer welfare by encouraging high output and low prices—not by protecting competitors or artificially maintaining an arbitrary, fixed number of firms in the market. Aligned with the error-cost framework, the law errs on the side of underenforcement, recognizing that penalizing low prices risks deterring the efficient, pro-competitive behavior that antitrust laws are meant to encourage. This high threshold also helps to prevent rivals from exploiting the legal system to shield themselves from more efficient competitors—a misuse of the law that would distort public policy for private gain.
The U.S. approach is encapsulated in the Brooke Group decision. Brooke Group set two strict conditions for predatory pricing claims:
- Prices must be below some measure of incremental costs; and
- There must be a realistic prospect of recouping these losses.
The recoupment requirement is crucial because, without it, predatory pricing leads to lower aggregate prices and enhances consumer welfare. Economic learning makes clear that, without a likelihood of recoupment, predatory pricing is not a rational business strategy (see here).
In the short term, punishing such cases would likely harm consumers. Over the medium to long term, it would surely discourage efficient conduct—particularly when that conduct happens to disadvantage competitors. This would be a paradoxical outcome for a law intended to protect competition.
By contrast, the EU standard is much laxer and more likely to injure consumers. In the EU, authorities must prove prices are below average variable cost (presumed predatory) or between average variable and average total cost (predatory if part of a “plan to eliminate competition”). Crucially, recoupment does not need to be shown.
By dispensing with Brooke Group, the CLRC proposal effectively recommends shifting California predatory-pricing law toward the European model. But such a standard lacks a basis in economic theory or evidence. The European approach reflects “structuralist considerations” far removed from consumer-welfare concerns. The fear is that dominant companies could, even to consumers’ benefit, create more concentrated markets presumed to be detrimental. This approach leaves less room to analyze concrete effects and is more prone to false positives, ignoring both consumer benefits from lower prices and the chilling effect on aggressive pricing. There is no basis for enshrining such an approach in California law.
Conclusion
The proposed changes to California’s single-firm-conduct provision represent a sharp departure from the established U.S. antitrust framework. By rejecting caution against Type I errors and seeking to overturn key Supreme Court precedents, California risks deterring pro-competitive behavior and ultimately harming consumers and innovation.
These reforms appear driven by a single unexamined assumption: that there must be more antitrust enforcement. But tearing down settled doctrine and abandoning the error-cost framework simply to increase the number of infringements—particularly when based on a vague, unquantified intuition that current enforcement is insufficient—is neither sound law, sound economics, nor sound policy. Who, after all, can credibly determine the “right” amount of enforcement? If there was “more” enforcement in the past, perhaps it reflected not a better equilibrium, but socially costly over-litigation, economic ignorance, ambiguous standards, and imprecise rules.
A way out of the hollow “more” versus “less” enforcement—or “pro” versus “anti” business—debate is to recognize that legal standards and precedent exist for a reason. They are the product of decades of legal and economic debate, litigation, and incremental refinement. Experimental arguments have been raised, tested, and discarded—and, occasionally, they have moved the needle.
This evolution doesn’t make the current standards perfect. But the rules that have emerged from this ongoing process of argumentative refinement offer a degree of predictability and coherence in a field where market behavior is often ambiguous, and where the outcomes of pro-competitive and exclusionary conduct can be nearly indistinguishable. This ambiguity is even more acute in so-called “digital markets,” where innovation is fast-moving, variable costs are often minimal or zero, and the risk of misjudging business practices is especially high.
Rather than pursue divergence for its own sake, California should preserve alignment with the broader U.S. framework—one grounded in the consumer-welfare standard, effects-based analysis, and the prudent caution of the error-cost approach.
