The Revised (But Uncorrected) Version of the Klobuchar Bill
The revised version of the Klobuchar bill (S.2992), aimed at prohibiting a few large tech companies from self-preferencing their services and forcing them to grant their rivals access to vital proprietary assets, fails to correct the critical deficiencies of the bill’s original version. This report provides a legal analysis of the new, yet uncorrected, pitfalls of the Klobuchar bill.
Senator Klobuchar (D-MN) has introduced a revised version of the antitrust bill (S.2992), which prohibits a few tech companies from promoting their products and engaging in certain other practices. Unfortunately, the new version of the American Innovation and Choice Online Act (AICOA) fails to address the bill’s fundamental flaws.
An increasing number of commentators and experts have warned against the considerable unintended consequences of the antitrust bills being debated in Congress, including AICOA. The widespread opposition notably includes many that support antitrust reforms. For example, USC law professor Erik Hovenkamp, who has argued in favor of pro-enforcement reform of antitrust laws, has stated that the AICOA bill (among others) does “not represent a judicious effort to fix the problems.” He has argued that these bills “are much more likely to diminish” both competition and innovation.
While it contains some positive amendments, the revised version of AICOA still fails to address five significant deficiencies:
1. The arbitrary scope of what is considered a covered entity
2. The prohibition of pro-consumer practices
3. The barriers to defendants’ establishment of an affirmative defense
4. The over-availability of temporary injunctions
5. The initiation of enforcement action prior to issuance of guidelines that clarify the bill’s many vague provisions
The arbitrary line-drawing exercise of the bill’s initial version remains concerning. However, in a positive step, the revised version removes the flawed distinction between a publicly traded and privately controlled company to determine which companies must comply with the bill’s prohibitions. However, Section 2 of the bill, reverting back to the bill’s original language, defines a “covered platform” based on the number of users and market capitalization of an entity, irrespective of whether the company is publicly traded or private.
The scope of the covered platform is problematic because it retains controversial quantitative criteria based on size rather than market dominance. In other words, nondominant platforms may fall within the ambit of the AICOA solely due to antibigness bias. For example, the bill would likely prohibit Google from promoting its cloud solutions, although Google Cloud has a nondominant market position, with only 10 percent of the market share.
In a positive step, the revised version removes the flawed distinction between a publicly traded and privately controlled company to determine which companies must comply with the bill’s prohibitions.
Indeed, under the bill, companies with more than $550 billion of market capitalization are covered, irrespective of their market positions and in the absence of evidence of abuse of their market positions. Selecting firms based on the vagaries of the stock market makes little sense. Market power and anticompetitive behavior should be the prerequisite for prohibiting the way entities can compete and innovate. The bill hinders fair and pro-consumer competition from nondominant firms by ignoring these factors instead of size thresholds.
Moreover, the revised version of the Klobuchar bill worsens the arbitrariness of the already controversial line-drawing exercise in the bill’s previous version. For example, Section 2(9)(B) of the revised version now explicitly excludes telecommunications companies from the bill’s ambit, stating that a covered online platform:
“Does not include a service by wire or radio that provides the capability to transmit data to and receive data from all or substantially all Internet endpoints, including any capabilities that are in cadential to and enable the operation of the communications service.”
In other words, large telecommunication companies that increasingly compete with tech companies for online and digital content are exempted from the bill’s ambit.
Many other incumbents also enjoy carve-outs in the new version. For example, the banking and payment industry is no longer covered under the bill, leaving out powerful companies such as banks and financial payment companies operating online that compete with the tech platforms targeted by the bill. Such carve-outs demonstrate the arbitrariness of the bill’s ambit, which encourages unfair competition by treating rivals in similar situations differently.
The bill’s carve-outs demonstrate the arbitrariness of the bill’s ambit, which encourages unfair competition by treating rivals in similar situations differently.
The fact that the revised bill version removes the ability of covered platforms to compete with potential incumbents underscores that the bill’s drafters were more concerned with reducing competition from large and disruptive companies than they were in promoting competition on the merits. When it comes to the deplorable line-drawing exercise of the Klobuchar bill, the revised version worsens the arbitrariness regarding the platforms that are covered by it.
The bill prohibits pro-consumer and pro-innovation practices without justification. The list and substance of the allegedly “unlawful conduct” described in Section 3 of the bill remain unchanged, despite the fact that much of this conduct has long been considered pro-consumer and pro-innovative. The result is consumers will suffer from new prohibitions that limit the covered platforms’ ability to offer products and services favorable to consumers.
The relaxed acceptance of affirmative defenses, while helpful, remains insufficient; many socially beneficial practices will remain prohibited under the current regime. The affirmative defenses allow for proportionate enforcement of the bill’s prohibitions, permitting consideration of efficiency considerations raised by the covered platforms. The initial version laid down a high threshold for such affirmative defenses to be acceptable in courts. Initially, for a defendant to successfully claim an affirmative defense of one of the prohibited practices, the bill required them to establish by a preponderance of the evidence that the conduct was “narrowly tailored,” “nonpretextual,” and “reasonably necessary” to further either of the following legitimate goals to:
▪ “prevent a violation of, or comply with, federal or state law;
▪ protect safety, user privacy, the security of nonpublic data, or the security of the covered platform; or
▪ maintain or substantially enhance the core functionality of the covered platform.”
Under Section 3 of the new version of the bill, an affirmative defense is acceptable if the “defendant establishes that the conduct was reasonably tailored and reasonably necessary, such that the conduct could not be achieved through materially less discriminatory means” to achieve the same abovementioned legitimate goals identified in the bill.
The relaxed acceptance of affirmative defenses, while helpful, remains insufficient; many socially beneficial practices will remain prohibited under the current regime.
The removal of the requirement for the affirmative defense not to be nonpretextual is positive but insufficient. The previous requirement that arguments by the defense be nonpretextual to be accepted wrongly required delving into the subjective intent of the company, which would have faced difficulties in convincing a judge that actions were not intended to undermine competition under the guise of furthering one of the three legitimate goals. For instance, any product improvement leading a covered platform to discriminate against, say, low-quality business users in the name of improving users’ data privacy can quickly be challenged by the impacted business users as being a “pretextual” improvement designed to undermine competition.
To illustrate, let’s say a data privacy requirement imposed by a covered platform on a message service provider led a provider to be demoted or rejected from the platform. Under the prior version of the bill, this messaging service provider could have successfully argued that such requirements were pretextual for the covered platform to thwart competition and engage in anticompetitive behavior. The covered platform would have been left with a difficult defense since nonpretextual intentions would have remained a highly subjective assessment left to judicial discretionary power. Given the potential impact of such a provision on product improvements, it is wise that the bill’s revised version dropped such a requirement. The requirements of “reasonably tailored,” reasonably necessary, and “not achieved through materially less discriminatory means” refer to a more objective assessment of the arguments raised by the defendants, away from the controversially subjective assessment of the notion of nonpretextual.
The shift from “narrowly tailored” to reasonably tailored is also positive. The change grounds the affirmative defenses in the notion of reasonableness, which is commonsensical and highly judicious, as judges apply the criterion of reasonableness as a general principle of law. Moreover, the reasonably tailored notion echoes the least restrictive/discriminatory means. These requirements underscore the need for the defendant to articulate the different alternatives available to achieve one of the three legitimate goals and the extent to which the alternative chosen to achieve a specific goal was the least restrictive (i.e., least discriminatory) means among the other options available.
These requirements also force the defendant to justify the proportionality (or reasonableness) of the invoked affirmative defense and for the judges to review such proportionality. They are demanding for defendants since they must justify choices considering multiple policy alternatives. Still, these requirements nevertheless represent improvements in the efficiency and reasonableness of the affirmative defenses, as they re-focus the justificatory process on objective arguments (i.e., proportionality and reasonableness) rather than on subjective and discretionary arguments (i.e., nonpretextual claims).
However, the improvements made to the regime of affirmative defenses under the bill’s revised version remain insufficient. Indeed, Section 3(b)(4) of the bill’s revised version now makes explicit that “the defendant has a burden of proving an affirmative defense under this subsection by a preponderance of evidence.” This evidentiary requirement, stated in the earlier version of the bill less intelligibly, clarifies the difficulty for defendants to claim an affirmative defense successfully. Indeed, since “preponderance of evidence” suggests a 51 percent chance of the claim to materialize, it appears that defendants cannot successfully rely on the likelihood of the legitimate goals to materialize to convince an enforcer or a judge to grant a defendant an affirmative defense.
The revised version refocuses the justificatory process of affirmative defenses on objective arguments (i.e., proportionality and reasonableness) rather than on subjective and discretionary arguments (i.e., nonpretextual claims). However, the bill’s improvements remain insufficient.
For instance, if the defendant legitimately wants to increase the security of its platform in a way that would inadvertently harm rivals, the likelihood that such increased security may not materialize with a 50 percent chance will lead enforcers and judges to reject the defendant’s affirmative defense claim at the benefit of those harmed rivals. In other words, the likelihood for the defendants to demonstrate “by a preponderance of evidence” that the actions they undertook genuinely and effectively pursued legitimate goals remains incredibly low: Defendants are likely to fail to show the positive effects of their actions since these positive effects may materialize long after (if ever) the adverse effects would occur. This time disconnect between tangible, short-term adverse effects of the prohibited actions on rivals and the likely, long-term positive effects of the defendants’ actions to pursue one of the legitimate goals identified by the bill will necessarily disfavor the defendants at the benefit of the likelihood of the success of the enforcers’ complaint.
In addition to the fact that affirmative defenses’ claims must demonstrate by a preponderance of the evidence that “the conduct was reasonably tailored and reasonably necessary, such that the conduct could not be achieved through materially less discriminatory means,” Section 3(b)(2) requires that some actions (those identified in paragraphs 4 to 10 of Section 3(a)) must also demonstrate that the blamed action “has not resulted in and would not result in material harm to competition.” However, the bill fails to define “harm to competition” in this context. Does it mean harm to competitors? Does it mean harm to the competitive process? Does it mean harm to competition by evidence of consumer harm? Or does it mean harm to competition by evidence of stifling innovation? Section 3(C)(ii)(II)(bb) could provide some insight into the drafters’ conception of harm to competition since the bill allows for temporary injunctions against an action that “materially impairs the ability of business users to compete with the covered platform operator”—meaning materially harming competition. Therefore, the harm to competition could mean harm to competitors, thereby creating a risk of stifling rather than promoting competition contrary to the bill’s stated objectives. There are many possible meanings of harm to competition, each yielding dramatically different outcomes.
More disturbingly, the notion of harm to competition suggests that the action needs to be anticompetitive (or else, a pro-competitive action cannot be said to create a harm to competition). But how can action by a covered platform be deemed to be anticompetitive if the covered platform does not have a market (let alone monopoly) power? Indeed, since the bill does not define the market and does not require a showing of market power by the covered platforms, enforcers can conclude that nondominant platforms have engaged in anticompetitive actions (i.e., creating harm to competition) absent the enjoyment of any market power. This reverts to the fundamental pitfall of the bill, which is to regulate by size rather than by banning abuse of market power in accordance with the tradition of antitrust laws. Consequently, the affirmative defenses’ requirement of not materially harming competition will have a chilling effect on the ability of nondominant companies regulated by the bill to compete and innovate since, although not enjoying market power, their actions are susceptible to being considered as harming competition, thereby barring the defendant from successfully claiming an affirmative defense.
While the revised version narrows the ability to obtain a temporary injunction, it does not do enough to address deficiencies in the prior version. Section 3(C)(ii)(II) (aa) provides for temporary injunctions against the covered platforms when “there is a plausible claim, supported by substantial evidence raising sufficiently serious questions going to the merits to make them fair ground for litigation, that a covered platform operator violated this Act.” This represents an improvement from the previous version, which stated that temporary injunctions were available when “there is a plausible claim, supported by evidence, that a covered platform operator took an action that would violate this Act,”
This change positively heightens the evidentiary threshold for a judge to grant a temporary injunction against a covered platform from “evidence” to “substantial evidence.” This heightened evidentiary requirement is laudable since mere evidence of self-preferencing and other prohibited actions will abound given the triviality of the business practices prohibited in the bill. Consequently, substantial evidence suggests a strong pattern of actions on “sufficiently serious questions” as to the fairness of the competition in a given market. This evidentiary change could help prevent frivolous claims wherein rivals weaponize antitrust laws to the detriment of the competitive process.
Section 3(C)(ii)(II)(cc) retains the ability to obtain temporary injunctions that “would be in the public interest.” This excessive vagueness fails to provide the necessary legal security for covered platforms to compete without the threat of being subject to unpredictable temporary injunctions from judges.
More importantly, the provision now clarifies that only materialized violations of the bill can lead to temporary injunctions. The previous version enabled temporary injunctions for the mere likelihood of violations of the act. In other words, the mere prospect of a covered platform to self-preference with, for instance, default settings in forthcoming innovations could have led enforcers to seek temporary injunctions against that platform. Temporary injunctions would have become readily available against the covered platforms whenever their envisaged actions had a remote effect of potentially violating the bill’s prohibitions. Such overreach would certainly have stifled innovation and encouraged a status quo rather than disruptions by the platforms at the expense of both innovation and consumer welfare. The revised version of the bill helps address this problem by ensuring that only actual (not potential) violations of the bill’s prohibitions can lead to temporary injunctions.
Notwithstanding the positive changes to the injunctive relief provisions, serious problems remain. Section 3(C)(ii)(II)(cc) retains the ability to obtain temporary injunctions that “would be in the public interest.” This excessive vagueness fails to provide the necessary legal security for covered platforms to compete without the threat of being subject to unpredictable temporary injunctions from judges.
The bill fails to ensure that any enforcement action takes place after the bill’s enforcement guidelines have been adopted, which is particularly problematic given the vagueness of numerous requirements. Indeed, given the bill’s high legal uncertainty and its effect on chilling innovation and competition, the enforcement guidelines should be published before, not after, the bill’s enforcement. The indeterminacy of what constitutes harm to competition under this bill, the vagueness surrounding the characterization of the prohibited actions in a specific context of digital service or device, and the legal uncertainty as to how the enforcers will interpret the legitimate goals of the affirmative defenses underscore the need for additional guidelines before enforcement.
How can a covered platform introduce an innovative product integrated into its digital ecosystem without incurring the legal risk that such introduction would disproportionately favor the platform’s product over those of business users and that no affirmative defense will be available? Also, how can a covered platform preserve the existing integration of two different services without incurring the risks of falling within one of the bill’s prohibited actions by failing to demonstrate that when the services were integrated, such integration was the least discriminatory means available? The legal risks are so high, and the legal certainty is so low, that enforcement guidelines are necessary before, not after, any enforcement action.
The revised version of the bill fails to address five critical deficiencies and, in some cases, makes these problems worse.
In conclusion, the revised version of the bill fails to address five critical deficiencies and, in some cases, makes these problems worse. These five aspects are the bill’s scope:
1. Arbitrary line-drawing exercise to define the bill’s ambit
2. Inclusion of prohibited practices that are pro-consumer
3. Insufficient affirmative defenses
4. Excessive acceptance of temporary injunction
5. Its legal vagueness
The bill’s revised version not only fails to address the fundamental flaws of the bill’s initial version, but, in some cases, makes existing problems worse. The bill should either be fundamentally amended to prevent such flaws from harming competition, innovation, and consumers, or abandoned, leaving any needed enforcement up to antitrust authorities.
About the Author
Dr. Aurelien Portuese (@aurelienportues) is the director of ITIF’s Schumpeter Project on Competition Policy. He focuses on the interactions of competition and innovation for antitrust purposes. Portuese holds a Ph.D. in law and economics from the University of Paris II (Sorbonne), an M.Sc. in Political Economy from the London School of Economics, and an LL.M. from the University of Hamburg. Portuese’s forthcoming book, Antitrust Populism, will be published by Oxford University Press.
The Information Technology and Innovation Foundation (ITIF) is an independent, nonprofit, nonpartisan research and educational institute focusing on the intersection of technological innovation and public policy. Recognized by its peers in the think tank community as the global center of excellence for science and technology policy, ITIF’s mission is to formulate and promote policy solutions that accelerate innovation and boost productivity to spur growth, opportunity, and progress.
For more information, visit us at www.itif.org.
. American Innovation and Choice Online Act, S.2992, 117th Cong. Amendments available at: https://www.klobuchar.senate.gov/public/_cache/files/b/9/b90b9806-cecf-4796-89fb-561e5322531c/B1F51354E81BEFF3EB96956A7A5E1D6A.sil22713.pdf
. See, for instance, Aurelien Portuese, “Open Letter to Sens. Durbin, Grassley, Klobuchar, and Lee Regarding the American Innovation and Choice Online Act (S.2992),” January 19, 2022, https://www2.itif.org/2022-ITIF-letter-S2992.pdf
. See, for instance, ITIF, “Schumpeterian Takes on Pending Antitrust Bills,” https://itif.org/publications/2022/01/21/schumpeterian-takes-pending-antitrust-bills/
. “Breaking Big Tech Bad,” Wall Street Journal, June 5, 2022, https://www.wsj.com/articles/breaking-big-tech-bad-senate-judiciary-committee-bipartisan-antitrust-bill-facebook-meta-google-twitter-11654113174
. Erik Hovenkamp, “Proposed Antitrust Reforms in Big Tech: What Do They Imply for Competition and Innovation?” CPI Antitrust Chronicle, forthcoming (2022), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4127334
. Felix Richter, “Amazon Leads $180-Billion Cloud Market,” Statista, February 8, 2022, https://www.statista.com/chart/18819/worldwide-market-share-of-leading-cloud-infrastructure-service-providers/