The fate of the badly misnamed American Innovation and Choice Online Act, S. 2992 (AICOA), may be decided by the August congressional recess. AICOA’s serious flaws have been ably dissected by numerous commentators (see, for example, here, here, here, and here). Moreover, respected former senior Democratic antitrust enforcers who have advocated more aggressive antitrust enforcement have also come out against the bill. For example, Stanford professor and former Acting Assistant Attorney General for Antitrust Douglas Melamed (who oversaw the Microsoft case for the Clinton Administration) very recently authored an article stressing that AICOA “is likely to impair innovation by the platforms.” The case has ably been made that the perverse welfare-reducing effects of multiple AICOA provisions, which impose inordinate costs (stemming, for instance, from interoperability requirements and prohibitions on “self-preferencing,” “discrimination,” and data usage) and discourage efficient vertical integration (see here), among other defects.
One aspect of AICOA that perhaps has garnered less attention is its affront to the rule of law. That deficiency in and of itself is sufficient to justify the summary rejection of this legislation by the Congress. Let’s examine it more closely.
A core element of the rule of law is that the government should apply the law neutrally to similarly situated entities. This principle is mocked, however, by the AICOA. The AICOA’s convoluted definition of “covered platform,” found in section 2(a)(5)(B) of S. 2992, focuses on rather arbitrary “monthly user,” capitalization, and sales value thresholds. Although the definitional elements were clearly designed to capture only the largest current digital platforms (all American) that have been in the public spotlight – Amazon, Facebook (now Meta), Apple, Google (now Alphabet), and Microsoft (possibly) – companies could fall within or outside the bill’s scope based on unpredictable changes in financial and user data in the future. This would lead to uncertainty as to whether particular firms were covered by the bill. It would also encourage corporate gamesmanship by specific firms as they sought to avoid the AICOA’s reach. As such, business planning would be rendered more difficult and less efficient, and the rule of law would be frayed.
A related rule of law concern is that parties be informed of the conduct they must adopt in order to avoid violating a particular law. Contemporary antitrust law does a far better job than the AICOA in satisfying this concern.
Contemporary American antitrust law has identified a few types of actions that are inherently anticompetitive, and therefore are “per se illegal” under all circumstances (bid rigging and naked horizontal price fixing and market division). Most business behavior, however, is assessed on a case-by-case basis under the antitrust “rule of reason,” which only condemns behavior whose anticompetitive effects outweigh its procompetitive effects. Rule of reason analysis prohibits behavior that inefficiently weakens the competitive process and excludes rivals for no legitimate business reason, and thereby tends to reduce consumer welfare. Mere harm to individual competitors (due say to more efficient or innovative production techniques) is not condemned. Specific enforcement agency guidance through speeches, enforcement actions, and enforcement guidelines have developed over time on a bipartisan basis to clarify what competition on the merits means in particular circumstances. As former Acting Assistant Attorney General for Antitrust Andrew Finch explained in a 2017 speech, enforcers’ emphasis has been giving notice about enforcement principles that allows private parties to reasonably predict the legal consequences of their actions:
[S]tability and continuity in enforcement are fundamental to the rule of law. The rule of law is about notice and reliance. When it is impossible to make reasonable predictions about how a law will be applied, or what the legal consequences of conduct will be, these important values are diminished.
In comparison with existing antitrust law, however, AICOA, does a very poor job of fostering predictability regarding what is prohibited. As Professor Melamed explains, it makes it unclear what it means when it uses the key term “material harm to competition”— whose absence a covered platform must demonstrate in order to avoid liability under the bill. Specifically, as Melamed stresses:
[T]he bill does not include the normal antitrust language (e.g., “competition in the market as a whole,” “market power”) that gives meaning to the idea of harm to competition, nor does it say that the imprecise language it does use is to be construed as that language is construed by the antitrust laws. . . . The bill could be very harmful if it is construed to require, not increased market power, but simply harm to rivals.
Rule of law predictability is further undermined by other ambiguous AICOA terms, which also threaten to harm competition and innovation, as Professor Dan Spulber points out (citations omitted):
The new [proposed platform-related] antitrust laws may have adverse effects on innovation and competition because of imprecise concepts and terminology. The American Bar Association Antitrust Law Section expressed concerns about “ambiguous terminology in the [AICOA] Bill regarding fairness, preferencing, materiality, and harm to competition on covered platforms.” The Section recommended that “these definitions direct attention to analysis consistent with antitrust principles: effects-based inquiries concerned with harm to the competitive process.”
Finally, AICOA also is in tension with the rule of law by placing the onus first on private parties to show that they have not violated the law (have not caused “material harm to competition”) when they have engaged in certain types of specified behavior deemed “problematic” under the bill. This is at odds with the approach under the antitrust rule of reason, in which the government first must show harm to competition before the defendant is required to justify its behavior as having procompetitive welfare-enhancing features. The AICOA’s placing of the initial burden on parties is troublesome, because the particular actions that trigger an initial presumption of illegality (self-preferencing, limitations on competitor access to the covered platform, certain “discriminatory” acts, certain restrictions on interoperability, certain use of nonpublic data, and so forth) are efficient and welfare-enhancing in many situations. Thus, AICOA undoubtedly would lead to the presumptive condemnation of much procompetitive conduct. Platforms that fell just outside AICOA’s coverage would not face this risk, because their similar conduct would be under the rule of reason. In short, the AICOA would lead to disparate treatment of identical conduct by similar firms, based on the bill’s arbitrary jurisdictional line-drawing. In conclusion, the AICOA sows confusion and undermines legal stability, continuity, and predictability. As such, it is an affront to the rule of law and should not be enacted, without regard to its substantive policy merits.