Open Letter to Sens. Durbin, Grassley, Klobuchar, and Lee Regarding the American Innovation and Choice Online Act (S.2992)

Dear Senators,

I write in relation to the markup of the American Innovation and Choice Online Act (S.2992). While the bill has positive elements regarding the ability of covered platforms to advance legitimate business justifications to explain some practices, it creates unintended consequences that will harm consumers and stifle innovation. Instead of advancing S.2992 or adopting unsatisfactory manager’s amendments on a profoundly misguided bill, senators should consider three fundamental flaws lying at the heart of S.2992.

1. Size Thresholds Create Unfair Competition

First, while antitrust enforcement traditionally sanctions anticompetitive conduct irrespective of a company’s size, the proposed bill sanctions size irrespective of the conduct at stake. For example, it prohibits common business practices only because the covered platform performs these practices, not because they are anticompetitive. Size thresholds inevitably generate considerable unintended consequences. This creates the conditions for unfair competition, not for fairer competition.

Antitrust laws are generally triggered when market dominance generates anticompetitive conduct. But market dominance needs to be established first. The bill ignores this fundamental requirement, creating perverse consequences. 

For instance, the bill will stringently regulate, say, Apple Music, which has only 16 percent of the global market for subscription music streaming; Amazon Music, which holds 13 percent of that market; or Google’s YouTube Music with 8 percent of market share. But it leaves the Swedish company Spotify, the dominant player with 32 percent market share, entirely unregulated. This example illustrates that the bill does not tackle potential abuses arising out of market dominance. It merely lashes out large online companies based on flawed quantitative criteria even though these large companies are small entrants and valuable disruptors in markets dominated by large incumbents.

Similar bizarre consequences could result from the bill’s application to financial payments companies, which fiercely compete with incumbent financial institutions and banks to offer cheaper and better financial payments solutions. The bill may prevent Apple Pay, Google Pay, Amazon Pay, or Facebook Pay from competing against incumbents such as PayPal, American Express, Visa, MasterCard, and other significant banks nationally and internationally. The bill’s distinction between rivals does not foster fair competition. It underpins unfair competition at the expense of consumer benefits and market disruptions.

Also, the quantitative criteria outlined in Section 2(5) of the bill fail to provide the necessary legal certainty market actors expect. It is unclear whether the criteria laid down in Section 2(5)(B)(ii) and Section 2(5)(C)(ii) are cumulative or alternative. Either way, the consequences are regrettable. If the criteria are cumulative, the size threshold would target only a handful of companies, leaving major online platforms unregulated against fair competition principles. If the criteria are alternative, the size threshold would include many companies that are in a growth trajectory and do not yet represent tech leaders when compared with global competitors. Again, either way, the size thresholds are unclear and irremediably prone to considerable unintended consequences for consumers who use the targeted products and for American competitiveness.

2. Prohibitions of Common Business Practices Harm Consumers and Innovation

Second, the prohibited practices generally are pro-competitive. It would be ironic for a bill that aims at fostering competition to ban pro-competitive and pro-innovative practices by which companies can effectively outcompete rivals for the benefit of consumers and innovation. The bill creates decreased competition and stifled innovation, not increased competition and bolstered innovation.

For instance, self-preferencing is a widespread business practice in all sectors of the economy. It is mainly pro-competitive because it ensures that disruptors can use their initial market position to compete with the incumbents in other markets, and it is pro-innovative because it forces companies not merely to imitate rivals but, more importantly, to innovate through quality or price differentiation. The proposed bill may prevent Microsoft from displaying LinkedIn profiles in Outlook emails, prevent Facebook from helping consumers exchange on its marketplace, prevent Google from helping scholars find articles via Google Scholar, prevent Apple from pre-installing any app on its smartphone and prevent it from prohibiting the safety-decreasing practice of sideloading, and prevent Amazon from promoting Whole Foods’ products on its platform. These examples, among many others, the harmful unintended consequences of prohibiting common business practices both beneficial to consumers and spurring from massive innovative efforts.

3. Per Se Rules of Illegality Frustrate the Rule of Reason

Third, should the covered platforms be subject to the unfair competition enabled by the bill’s size thresholds, and should the covered platforms be subject to efficiency-decreasing prohibitions, they will have virtually no argument to put forward to rationally justify their practices. Indeed, the bill frustrates the fortunate balancing exercise of the rule of reason inherent to antitrust analysis and may even thwart the rule of law.

Controversially, covered platforms will not be able to argue that a given practice increases consumer welfare or is innovative as a legitimate justification of the practice. This violates decades of antitrust jurisprudence and lays down the basis for courts’ efficiency-decreasing and anti-innovation sanctions.

Additionally, Section 3(b) provides for insurmountable thresholds to obtain an “affirmative defense.” Indeed, besides the need to comply with federal or state law, Section 3(b) provides minimal considerations that not related to efficiency or innovation for the first three practices that the bill prohibits. The “affirmative defenses” of Section 3(b)(2) applicable to the remaining practices that the bill prohibits are so stringent that they are virtually out-of-reach for covered platforms. For example, the covered platform will have to demonstrate that the blamed practice “was narrowly tailored, could not be achieved through less discriminatory means, was nonpretextual, and was necessary to” regulatory compliance or safety objectives. Because the covered platforms may engage in a practice because of user considerations and not with proportionality considerations, it will always be possible to retrospectively speculate that alternative “less discriminatory means” were available, thereby deterring the platform from innovating in the first place given the real risks of prosecutions and litigations. The chilling effect of these insurmountable affirmative defenses will inevitably lead covered platforms to refrain from innovating, disrupting incumbents, and serving consumers nimbly and efficiently.

Consequently, because these affirmative defenses are virtually unavailable to covered platforms since they are impossible to meet, the bill lays down rules of per se illegality rather than rules subject to a balancing exercise inherent to the rule of reason that has dominated antitrust enforcement for more than a century now. Indeed, the bill discriminates against a handful of companies and imposes onto them ex ante rules of per se illegality despite the considerable amount of literature and evidence advocating for a strengthening of the rule of reason and the avoidance of any per se rule of illegality. In short, the bill takes a precautionary approach to innovative companies and innovative practices in the age of a fierce race for global innovation leadership. It is thus the wrong approach at the wrong time. Manager’s amendments aimed at potentially including Chinese platforms within the bill’s remit will fail to minimize the numerous harms that the bill intrinsically generates. Amendments will fail to stop offering Chinese app developers access to U.S. platforms—such access that this bill generously and dangerously provides them.

Finally, given these three fundamental flaws of the bill (i.e., size thresholds, prohibition of pro-competitive practices, and rules of per se illegality), there are two potential outcomes: Either courts and agencies will aggressively enforce the bill, in which case the unintended consequences described will unfold at the expense of American consumers and innovation, or courts and agencies will refrain from applying the obligations contained in the bill, in a similar manner to the poorly enforced Robinson-Patman Act of 1936, because the statutory obligations and prohibitions appear utterly disconnected from market reality and consumer considerations. Should they choose the latter alternative, it will be no better solution. So, senators should refrain from adopting the bill: Disapplied laws harm the legislature, the credibility of lawmakers, and ultimately the institution of Congress.

Montesquieu once wrote, “useless laws weaken the necessary laws.” Should the current bill be disapplied, the useless bill would weaken antitrust laws. Montesquieu added that “it is sometimes necessary to change certain laws. But the case is rare, and when it happens, they should be touched only with a trembling hand.” There is no apparent trembling hand in the present case, but rather an overconfident hand drafting provisions bound to harm American consumers and innovation.

Because these three flaws, among others, are core to the bill’s structure, it appears that any further amendment may be vain, as manager’s amendments already demonstrate. For, amendments would fall in either of the following categories: i) insufficiently changing the structure of the bill, thus leaving untouched the unintended consequences of the bill, or ii) sufficiently changing the structure of the bill, therefore voiding the bill of any substance, thereby leading to a disapplied law which would weaken the whole antitrust regulatory framework. Either way, amendments would fail to address the bill’s flaws.

Antitrust reforms should instead seek to strengthen antitrust agencies’ capabilities with increased resources. Reforms also should ensure that dynamic competition is privileged over the static competition. Otherwise, innovation and consumers will be collateral damage. To promote dynamic competition, antitrust reforms need to revolve around “dynamic antitrust” principles—namely, innovation-based antitrust enforcement.


Aurelien Portuese
Director, Schumpeter Project on Competition Policy
Information Technology and Innovation Foundation
700 K Street NW, Suite 600, Washington, DC 20001

cc: Members of the U.S. Senate Committee on the Judiciary