Spotlight

INTEREST OF AMICI CURIAE^[1]

The International Center for Law & Economics (ICLE) is a nonprofit, non-partisan global research and policy center aimed at building the intellectual foundations for sensible, economically grounded policy. ICLE promotes the use of law and economics methodologies to inform public policy debates and has longstanding expertise in the interpretation and proper implementation of the U.S. antitrust laws.

Amici also include 12 scholars of antitrust law and economics at leading universities and research institutions across the United States. Their names, titles, and academic affiliations are listed in the Addendum. All possess expertise in, and collectively have conducted copious research on, antitrust law and economics.

Amici have an interest in the proper development of antitrust jurisprudence and believe the district court’s decision, if left to stand, would undermine the fundamental goal of the antitrust laws: the protection of market competition.

ARGUMENT

Federal courts have tremendous remedial powers under the U.S. antitrust laws. See Herbert Hovenkamp, Antitrust and Platform Monopoly, 130 Yale L.J. 1952, 2005 (2021) (“Antitrust’s provisions for public equitable relief are extremely broad, with no explicit restriction on the nature of the relief.”). They may prevent mergers or acquisitions ex ante or unwind them ex post, and they may prohibit all manner of concerted or unilateral action in commerce. Not surprisingly, firms often attempt to coopt these powers for their own advantage, petitioning courts for remedies that will aid them in competing against existing or potential rivals.

For that reason, the U.S. Supreme Court has repeatedly cautioned that “[t]he antitrust laws . . . were enacted for the ‘protection of competition, not competitors.’ ” Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477, 488 (1977) (quoting Brown Shoe Co. v. United States, 370 U.S. 294, 320 (1962)).^[2] Across decades of decisions, the Court has consistently recognized that this is the very “purpose of the antitrust laws.” Leegin Creative Leather Prods., Inc. v. PSKS, Inc., 551 U.S. 877, 906 (2007).

This bedrock principle is a veritable mantra in antitrust jurisprudence. And yet it is flatly inconsistent with the decision below. Plaintiff FuboTV is a seller in a market the district court defined as the “live pay TV market.”^[3] Defendants Disney, Fox, and Warner Brothers Discovery have proposed to enter that market and compete with Fubo through a joint venture called Venu Sports. Venu will combine the three defendants’ sports programming into a live-streaming offering for consumers who desire live sports programming. That programming is also available on Fubo. But because Venu will include only sports content, it will retail for a lower price than Fubo. In the words of the district court, “[t]he target consumer will—for the first time—be able to subscribe to a vast array of the sports content he or she wants, without paying for entertainment content he or she does not.” O&O at 19. By enjoining the JV’s entry into the market, the district court protected Fubo from having to compete with this new and attractive offering and therefore inverted the fundamental antitrust principle articulated time and again by the Supreme Court: It allowed Fubo to use antitrust law to protect itself—a competitor—from competition. In particular, the district court (1) ignored the antitrust injury requirement and allowed Fubo to coopt the antitrust laws to insulate itself from competition, (2) flouted Supreme Court limits on price squeeze claims, (3) misapplied an inapposite and likely abrogated case involving a joint venture injunction, (4) cynically condemned a standard ancillary and necessary joint venture restraint, and (5) presumed economically irrational behavior would fill the competitive gap created by the injunction barring Venu’s entry. This Court should reverse.

I. FUBO WILL NOT SUFFER ANTITRUST INJURY FROM THE PROPOSED JOINT VENTURE

The Supreme Court imposes a threshold requirement on all private antitrust plaintiffs to ensure they cannot coopt the antitrust laws to protect themselves from competition: they must establish that they have suffered “antitrust injury,” which is “injury of the type the antitrust laws were intended to prevent and that flows from that which makes defendants’ acts unlawful.” Brunswick, 429 U.S. at 489. Because the antitrust laws were intended to prevent competition-reducing behavior in the market, each antitrust plaintiff must demonstrate that its complained-of harm is a result of diminished—not enhanced—competition. This requirement applies even if the plaintiff is, like Fubo, seeking only injunctive relief. Cargill, 479 U.S. at 113 (“[W]e conclude that in order to seek injunctive relief under § 16, a private plaintiff must allege threatened loss or damage ‘of the type the antitrust laws were designed to prevent and that flows from that which makes defendants’ acts unlawful.’ ”) (citing Brunswick, 429 U.S. at 489).

The proposed joint venture—the only conduct the district court condemned—creates no antitrust injury for Fubo. Fubo complains that the JV will prove so popular among Fubo’s customer base that the JV will win significant business from Fubo, preclude its hoped-for—but never realized—profitability and threaten its continued viability. But those effects would be harms to an individual competitor, not competition, and they stem from increased, not diminished, competition. Indeed, Fubo’s asserted injury resembles that alleged in Brunswick. The plaintiff there brought a claim under Section 7 of the Clayton Act—the statute under which the district court condemned the JV here—alleging that the challenged merger reduced its profits by causing it to face greater competition. Brunswick, 429 U.S. at 488. Such injury, the Court held, is not antitrust injury and cannot sustain a Section 7 claim. Id.

Astoundingly, the district court’s order never even mentions the term antitrust injury. The closest reference to the concept—a prerequisite to any private antitrust enforcement action—appears in a footnote:

The JV Defendants also argue that any harm to Fubo from the JV is the result of legitimate competition at work. See Opp. at 54–55. However, since the Court herein finds that Fubo is likely to succeed on its Section 7 claims, it is also likely that any such competition posed by the JV is contrary to the antitrust laws.

O&O at 56 n.38.

To the extent this brief and passing remark is intended to address the antitrust injury requirement, it fundamentally conflicts with the Supreme Court’s holding in Brunswick. The district court appears to have reasoned that a Section 7 plaintiff may maintain its action if it shows that a violation of Section 7 injured it (or is likely to do so). But Brunswick expressly rejected that reasoning. The Supreme Court there took as given that the challenged merger both violated Section 7 and harmed the plaintiff. Brunswick, 429 U.S. at 484 (“[Defendant] does not presently contest the Court of Appeals’ conclusion that a properly instructed jury could have found the acquisitions unlawful. Nor does [defendant] challenge the Court of Appeals’ determination that the evidence would support a finding that had [defendant] not acquired these centers . . . [plaintiffs’] income would have increased.”). Nevertheless, the Court concluded that the plaintiff did not suffer antitrust injury and could not maintain its claim. Id. at 488–89. Brunswick thus demonstrates that proving a Section 7 violation and resulting harm is not sufficient to establish antitrust injury.

Because the district court did not require Fubo to demonstrate antitrust injury, and because Fubo’s anticipated injury from the enjoined JV results from an enhancement of competition in the live pay TV market, the district court committed reversible error. This Court should vacate the injunction, reverse the district court, and reaffirm the long-established principle that the antitrust laws protection competition, not competitors.

II. FUBO COMPLAINS OF A PRICE SQUEEZE BUT CANNOT ESTABLISH THE NECESSARY PREREQUISITES TO LIABILITY

By the district court’s own admission, the conduct it enjoined—defendants’ launching of the JV—would increase competition in the live pay TV market and benefit consumers. O&O at 51 (“The JV will offer consumers an option to receive their must-have live sports content at a fraction of the cost of what current [sellers in the live pay TV market] can offer.”); id. at 19 (“The target consumer will—for the first time—be able to subscribe to a vast array of the sports content he or she wants, without paying for entertainment content he or she does not.”). The district court maintained, though, that the JV cannot be assessed in isolation. When examined in light of defendants’ alleged practices of licensing their sports content only when bundled with their non-sports content, the court reasoned, the JV is likely anticompetitive. Id. at 45–46 (“[B]undling has been uniformly and systematically imposed on each distributor in the live pay TV industry except the JV, preventing any other distributor from offering a multi-channel sports-focused streaming service. . . . [T]he JV is the vehicle through which the JV Defendants will capitalize on this opportunity, to potential anticompetitive effects.”).

The problem with imposing liability on that basis is that the anticompetitive conduct the district court purported to identify is a “price-squeeze,” and the court did not find (because Fubo could not establish) the necessary prerequisites for condemning such a practice.

A price squeeze may result “when a vertically integrated firm sells inputs at wholesale and also sells finished goods or services at retail.” Pac. Bell Tel. Co. v. Linkline Commc’ns, Inc., 555 U.S. 438, 442 (2009). As the Supreme Court has explained:

If that firm has power in the wholesale market, it can simultaneously raise the wholesale price of inputs and cut the retail price of the finished good. This will have the effect of ‘squeezing’ the profit margins of any competitors in the retail market. Those firms will have to pay more for the inputs they need; at the same time, they will have to cut their retail prices to match the other firm’s prices.

Id.

This precisely describes the conduct Fubo attacked: defendants participate in the upstream wholesale market as licensors of television programming. By licensing their sports content only on a bundled basis (so that licensees must also pay to license unwanted non-sports programming), they allegedly raise the effective price of the sports programming that is a necessary input for a live sports streaming service. Then, because they license their sports content to their own JV on an unbundled—and thus cheaper—basis, they charge lower prices in the downstream retail market, pressuring rival sports streaming services to cut their retail prices. This pattern of behavior squeezes the profits of the defendants’ retail competitors, including Fubo.

In Linkline, the Supreme Court set forth the requirements for a plaintiff that seeks to establish antitrust liability for an alleged price squeeze. The Court held that the plaintiff must prove either (a) that the defendant owed and breached an antitrust duty to deal with the plaintiff on particular terms in the wholesale market or (b) that the defendant engaged in predatory pricing in that it: (i) charged a below-cost price in the retail market and (ii) was likely to recoup its losses from such below-cost pricing through future supracompetitive pricing once its competition was eliminated or weakened. Id. at 452 (“If there is no duty to deal at the wholesale level and no predatory pricing at the retail level, then a firm is certainly not required to price both of these services in a manner that preserves its rivals’ profit margins.”).^[4]

Fubo did not show it was likely to prove either a duty to deal or predatory pricing. With respect to a duty to deal, the district court repeatedly emphasized that it was not concluding that defendants had an antitrust duty to license their sports programming on an unbundled basis.^[5] See O&O at 4 (“The Court need not, and does not, reach the question of the legality of bundling at this stage of the case.”); id. at 45 (“[T]he Court need not (and does not) determine the legality of programmers’ bundling practices in order to decide this Motion. . . . ”); id. (“[W]hether bundling is itself illegal under the antitrust laws is not a question currently before the Court.”); id. at 55 (“If bundling (as to Fubo specifically or as a general industry practice) is to be struck down as an antitrust violation, it should come only after a full trial on the merits.”). With respect to predatory pricing, Fubo produced no evidence, nor did it argue, that the JV would charge below-cost retail prices and ultimately recoup its losses by charging monopoly prices once its rivals were extinguished. Thus, Fubo’s challenge to the “one-two punch” of the JV’s entry into the live pay TV market coupled with defendants’ bundled licensing practices—an effective price squeeze claim—is legally deficient.

The district court rejected this reasoning on two grounds, each of which is unsound. First, it contended that defendants’ alleged misconduct should not be parsed into its component parts—the individual defendants’ bundling practices and the JV’s entry into the retail market at a low price point—to analyze the legality of each separately. O&O at 37. This approach, the district court said, would disregard the Supreme Court’s 1962 instruction that “ ‘plaintiffs should be given the full benefit of their proof without tightly compartmentalizing the various factual components and wiping the slate clean after scrutiny of each.” Id. (quoting Cont’l Ore Co. v. Union Carbide & Carbon Corp., 370 U.S. 690, 699 (1962)). But in 2009, the Supreme Court mandated precisely this sort of “compartmentalizing” approach for price squeeze claims such as that advanced by Fubo. Linkline, 555 U.S. 438. In Linkline¸ the Court held that “[i]f there is no duty to deal at the wholesale level and no predatory pricing at the retail level, then a firm is certainly not required to price both of these services in a manner that preserves its rivals’ profit margins.” Id. at 452.

The district court also asserted two reasons why neither Linkline nor one of the cases upon which its holding rested, Trinko, apply to Fubo’s claims. (Trinko narrowly constrained the circumstances in which an antitrust duty to deal exists. 540 U.S. at 408–10.)^[6] The district court first claimed that Linkline and Trinko are irrelevant here because they “involved actions brought under Section 2 of the Sherman Act, not Section 7 of the Clayton Act, and primarily allege specific technical per se violations of the Sherman Act not applicable here.” O&O at 36. It then stated that “courts have been clear that the ‘no duty to deal’ defense raised by the JV defendants in reliance on Linkline and Trinko is not a defense to concerted actions.” Id. at 37 (citing Buccaneer Energy v. Gunnison Energy Corp., 846 F.3d 1297, 1309 (10th Cir. 2017); In re Dealer Mgmt. Sys. Antitrust Litig., 680 F. Supp. 3d 919, 1004 (N.D. Ill. 2023)). Neither assertion is availing.

With respect to the first, the court was wrong to claim that Linkline and Trinko “primarily allege specific technical per se violations of the Sherman Act.” O&O at 36. The complained of conduct in Trinko was a unilateral refusal to deal on particular terms; in Linkline, it was a price squeeze. Neither behavior is per se illegal under the Sherman Act. Moreover, it is irrelevant that the claims in those cases were based on Section 2 of the Sherman Act rather than Section 7 of the Clayton Act. As explained above, the theory of harm underlying Fubo’s Section 7 claim is that the launching of the JV will produce an illegal price squeeze. See supra § II. The legality of price squeezes must be assessed under Linkline. And when, as here, price squeeze defendants have not engaged (and are not expected to engage) in predatory pricing in the downstream market, the legality of their behavior turns on whether they possessed and breached an antitrust duty to deal on particular terms in the upstream market. Linkline, 555 U.S. at 452. That matter is governed by Trinko. 540 U.S. at 408–11. Defendants’ reliance on Linkline and Trinko is thus not “inapt,” as the district court asserted. O&O at 36.

Nor is the distinction between concerted action and single-entity conduct relevant under these circumstances, because the district court enjoined a proposed joint venture under Section 7 of the Clayton Act. Texaco Inc. v. Dagher, 547 U.S. 1, 6 (2006) (“When persons who would otherwise be competitors pool their capital and share the risks of loss as well as the opportunities for profit . . . such joint ventures [are] regarded as a single firm competing with other sellers in the market.” (citations omitted)). Regardless, courts have not held that Trinko’s limits on an antitrust duty to deal do not apply whenever a claim involves any sort of concerted conduct. One of the cases the district court cited in support of that proposition held merely that Trinko does not apply to one type of concerted conduct: concerted refusals to deal (i.e., agreements among firms not to deal with a particular rival). Buccaneer, 846 F.3d at 1309 (observing that “Trinko simply does not speak to claims, like those here, alleging concerted refusals to deal”).^[7] The other is a district court case that simply repeats the point. Dealer Mgmt., 680 F. Supp. 3d at 1004 (“True, the ‘general right to refuse to deal with competitors’ applies only to unilateral refusals.”) (citing Buccaneer, 846 F.3d at 1309).

The refusal to deal component of Fubo’s claim—its assertion that each defendant refuses to license its sports content on the “skinny” basis Fubo would prefer—alleges unilateral, not concerted, refusals. So the cases cited by the district court are inapplicable.

Fubo’s challenge to the totality of defendants’ sports-content practices—each defendant’s alleged unilateral practice of refusing to license such content on an unbundled basis, coupled with the defendants’ collective entry into the downstream sports streaming market at a low price point—is a price squeeze claim. Because Fubo has not shown a likelihood of proving either an antitrust duty to deal in the wholesale licensing market or predatory pricing in the retail sports streaming market, it is unlikely to succeed on the merits of its underlying action. This Court should therefore vacate the district court’s injunction.

III. COLUMBIA PICTURES, WHICH TURNED ON AN EXCLUSIVITY PROVISION NOT PRESENT IN THE PROPOSED JOINT VENTURE, IS INAPPOSITE

The primary precedent the district court relied on in enjoining the launch of Venu was United States v. Columbia Pictures Industries, Inc., 507 F. Supp. 412 (S.D.N.Y. 1980). In that case, the court enjoined four major U.S. film producers from launching “Premiere,” a joint venture that would enter the growing pay television market to compete with existing services like HBO, Showtime, and The Movie Channel. Id. at 434. The four joint venturers would combine their film content, distribute it via television to subscribers, and share revenues. Id. at 419–20. According to the district court here, Premiere “present[ed] a scenario strikingly similar to this case” in that it was conceived “[a]t an analogous time of rapid change in the television and film industry,” was comprised of participants “that controlled just over half” of the content needed for a service of its sort, and enabled participants “to capture a share of the burgeoning pay TV movie channel market.” O&O at 36.

Columbia Pictures was decided long before Trinko recognized the significant limits of an antitrust duty to deal. But there is another outcome-determinative difference between the enjoined Premiere joint venture and the one at issue here. With Venu, the defendant joint venturers will be free to—and will—license their sports programming to rival distributors; those rivals will not be denied any content available to the JV. Id. at 22. With Premiere, by contrast, the joint venture was to have the exclusive right to distribute the four participants’ films for a nine-month period. Columbia Pictures, 507 F. Supp. at 419 (“The joint venture agreement provides that Premiere is to have certain films distributed by the movie company venturers available to it exclusively for a nine-month period, before those films are shown on the existing satellite-fed network programming services.”).

The district court here downplayed this difference between Premiere and Venu, observing in a footnote that “[t]he decision in Columbia Pictures did not rest solely on the anticompetitive effects of this ‘exclusivity’ provision.” O&O at 36 n.30. A fair reading of Columbia Pictures, however, suggests that the nine-month exclusivity provision was essential to the court’s decision to enjoin Premiere. Most notably, the court concluded that the exclusivity provision was probably a per se illegal group boycott. Columbia Pictures, 507 F. Supp. at 428, 429. In addition, the court repeatedly stressed the centrality of the exclusivity provision to the joint venturers’ economic interests and to the anticompetitive effect of the venture. Id. at 420–21, 430–32.

The Columbia Pictures court rightly fixated on the exclusivity provision of the Premiere joint venture because that feature is what rendered the venture anti-, rather than pro-, competitive. Had it launched, Premiere would have hobbled its rivals by denying them access to inputs they needed for success. Their harm would have resulted from an effort to reduce competition—i.e., to make them less competitive. With Venu, which includes no exclusive licensing commitments from its participants, rivals’ access to needed inputs will not change and any harm they suffer will be the result of increased competition—the entry of a new streaming service offering a product consumers are demanding. As the Ninth Circuit recently observed, there is a difference “between anticompetitive behavior, which is illegal under federal antitrust law, and hypercompetitive behavior, which is not.” FTC v. Qualcomm, Inc., 969 F.3d 974, 982 (9th Cir. 2020). Holding back one’s competitors, as Premiere did, is anticompetitive. Entering their market with a superior offering, as defendants seek to do through Venu, is not.

IV. THE DISTRICT COURT RELIED ON AN ANCILLARY NONCOMPETE AGREEMENT THAT IS IRRELEVANT TO FUBO’S CLAIM

In concluding that Fubo was likely to succeed in proving that the JV would harm competition, the district court made much of a non-compete agreement among the defendants. According to the court, that agreement “forbids the JV Defendants from ‘owning any form of equity interest, including a revenue-sharing or profit-sharing interest, in a commercial venture, where the focus of the commercial venture is the operation of a sports-centric [live-streaming service] similar to the JV Platform for a period of three (3) years from the Launch Date.’ ” O&O at 47 (quoting PX289 at 17). The court concluded that the non-compete agreement has “significant ‘anticompetitive potential’ [and] warrants scrutiny even in the absence of incipient monopoly.” O&O at 49 (quoting Copperweld Corp., 467 U.S. at 769).

Although the court was right to “scrutin[ize]” the noncompete agreement and assess its “anticompetitive potential” (as all features of a challenged joint venture should be evaluated for anticompetitive potential), this noncompete agreement is almost certainly procompetitive. O&O at 49. The agreement does not preclude the defendants from licensing their sports content—even on an unbundled basis—to any rival. Id. at 48 (“This non-compete agreement does not prevent the JV Defendants from licensing their programming to other [live pay TV distributors]. . . .”). All it requires is that they not invest in a sports-streaming service that competes with their own joint venture. Such a commitment prevents each individual joint venture participant from free-riding on the JV’s efforts to develop the market for comprehensive sports-only live streaming services (e.g., by introducing the novel service to consumers). Absent the noncompete, one of the JV defendants could rely on all the joint venturers to share those market development costs and then take a stake in a new rival that would not need to incur them. From the earliest days of federal antitrust law, non-compete agreements aimed at preventing such free-riding and thereby facilitating the creation of a venture have been deemed reasonable, and thus legal, as ancillary restraints. See United States v. Addyston Pipe & Steel Co., 85 F. 271, 280 (6th Cir. 1898), aff’d, 175 U.S. 211 (1899) (observing that “[r]estrictions in the articles of partnership upon the business activity of the members, with a view of securing their entire effort in the common enterprise, were, of course, only ancillary to the main end of the union, and were to be encouraged”); id. at 281 (“[C]ovenants in partial restraint of trade are generally upheld as valid when they are agreements . . . by a partner pending the partnership not to do anything to interfere, by competition or otherwise, with the business of the firm. . . .”).^[8]

Because the defendants continue to license their content to others, the non-compete agreement does not eliminate competition but rather fosters the JV’s ability to compete effectively. In short, there is nothing nefarious about the limited non-compete agreement among the Venu venturers.

V. THE COMPETITIVE SCENARIOS THAT THE DISTRICT COURT ASSUMED WOULD RESULT FROM ITS BLOCKING OF VENU ARE ECONOMICALLY IMPLAUSIBLE

In enjoining the launch of Venu, the district court stripped consumers of a novel offering for which there is significant demand: a sports-focused, live-streaming service featuring a vast array of content at a low price point. Insulating Fubo from competition by such a service, and denying it to consumers, could be justified only if preventing competition were likely to generate greater competition and enhance consumer welfare in the future. But the rosy competitive future the district court assumed would result from its barring of Venu is implausible.

The district court asserted that “the existence of the JV itself incentivizes the JV Defendants to prevent and suppress other potential sports-focused bundles from meaningfully competing” because “[t]he JV Defendants know the unique value of their unbundled sports programming and are aware that any competitor offering such unbundled live sports will devalue the JV.” O&O at 49. Implicit in those remarks is the assumption that if Venu is enjoined, other live-streaming services offering the same sports-only content are more likely to enter the market. It is highly unlikely, though, that enjoining Venu would result in the entry of a streaming service offering sports-only content as extensive as that available on Venu and at a similar price point.

The district court suggested that such an offering could come about in two ways: by all the defendants’ deciding to license their sports content on an unbundled basis to another distributor or by a defendant’s launching its own sports-only streaming service featuring its sports content and that licensed on an unbundled basis from the other defendants. Id. at 50 (“For such a competitor to emerge, in all likelihood one or more of the JV Defendants would have to be involved in launching it, whether by [1] agreeing to fully unbundle their sports channels for another distributor, or [2] launching a [streaming] service themselves that would feature their own sports channels alongside licensed sports channels from other programmers.”).

Both of those scenarios are improbable.

A. A Low-Priced, Comprehensive Sports Streaming Bundle Is Unlikely to Result from Defendants’ Licensing Their Sports Content to Another Distributor on an Unbundled Basis

The defendants each hold exclusive rights to high-demand sports content and can therefore earn supracompetitive profits from licensing their sports programming.^[9] They could capture such profits by licensing their sports programming at very high prices. They have chosen instead to charge lower prices for their sports content but to require that licensees also license, at profit-generating prices, other content they control. Such bundling allows defendants to capture the profits their valuable sports programming makes possible—and to which they are entitled, see supra note 9—while subsidizing less popular content.

The district court assumed that if Venu is blocked, the defendants may eventually respond to consumer demand for a comprehensive, sports-focused live-streaming service by licensing their sports content on an unbundled basis to a third-party distributor. O&O at 50 (hypothesizing that defendants may “fully unbundle their sports channels for another distributor”). But given that bundling is the current means by which defendants extract transactional surplus and capture the profits the law permits them to earn, it would be economically irrational for them to unbundle their sports content without simultaneously raising the price of their unbundled sports content. If they unbundled their sports content but charged more for it, a third-party could offer a sports-focused live-streaming service, but only at a high price reflecting the high cost of its content. The alternative offering—if one came to pass—would therefore lack an essential feature of Venu: its relatively low price.

In reasoning that the blocking of Venu may lead defendants to license their sports content to a third-party distributor on an unbundled basis and at prices that would allow the distributor to match Venu’s price, the district court assumed defendants would act against their own economic interests. Antitrust law, though, rejects theories that assume economically irrational behavior. See, e.g., Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 588–95 (1985) (granting summary judgment for defendants because plaintiffs’ theory of conspiracy assumed economically irrational behavior). It also requires courts to assume rational behavior in hypothesizing scenarios that would exist “but for” the conduct under review. See, e.g., Dolphin Tours, Inc. v. Pacifico Creative Serv., Inc., 773 F.2d 1506, 1510–11 (9th Cir. 1985) (Plaintiffs “must presume the existence of rational economic behavior in the hypothetical free market.”); Murphy Tugboat Co. v. Crowley, 658 F.2d 1256, 1262 (9th Cir. 1981) (“[E]conomic rationality must be assumed for all competitors, absent the strongest evidence of chronic irrationality.”); United Food & Com. Workers Local 1776 v. Teikoku Pharma USA, 296 F. Supp. 3d 1142, 1179 n.42 (“[I]n construct[ing] but-for world scenarios, there is a presumption of economical rationality.”). Because it would be economically irrational for defendants to respond to the blocking of Venu by unbundling their sports content without raising its price, this Court should not credit the district court’s assumption that enjoining the launch of Venu would likely result in third-party offerings matching Venu’s quality (i.e., its comprehensive sports coverage) and price.

B. No Individual Defendant Is Likely to Offer a Live Sports-Streaming Bundle Matching Venu’s Comprehensive Content and Relatively Low Price

The second purportedly superior outcome the district court assumed might result from its blocking of Venu—individual defendants’ launching of their own sports-only live-streaming services featuring their own offerings and those licensed from the other defendants—is similarly implausible, at least at prices as low as Venu’s. That is because of a pricing dynamic known as “double marginalization.” As a threshold matter, two of the three defendants had already entered or announced their own sports streaming services in addition to Venu, and there is no record evidence that the third was going to contemplate entry without Venu.

Double marginalization tends to result when two firms that separately sell complements that are used in combination to produce a “downstream” product or service—for example, the separate bits of sports programming that must be combined to create a comprehensive sports-streaming service—both possess power over pricing. See generally Makan Delrahim, ‘Harder, Better, Faster, Stronger’: Evaluating EDM as a Defense in Vertical Mergers, 26 Geo. Mason L. Rev. 1427, 1429–30 (2019) (explaining the economics of double marginalization).^[10]

Potential consumers of the downstream product—in this case, the comprehensive package of sports programming—mainly care about the total price they pay for the product (or package) that they seek. That is true if the consumer is purchasing the complete package from a single seller or cobbling the desired package together from multiple sellers. If a single firm controlled all the complements, it would set that combined price at a level that maximizes its profits.

When the component parts of an offering are sold by separate sellers who each have power over pricing for the component they sell, each has an incentive to price its own component at a level that will enable it to capture as much of the available profits on the combined offering as possible. But if each component seller takes that tack, the combined price of the separately sold components will exceed the profit-maximizing price of a single offering that combines them. If the separate component sellers were to combine, the combination would have an incentive to set a lower aggregate price for the components—one that would maximize the sellers’ profits on the combined offering. Such a move would benefit the sellers and consumers, who would enjoy lower package prices. See United States v. AT&T Inc., 310 F. Supp. 3d 161, 197 (D.D.C. 2018), aff’d, 916 F.3d 1029 (D.C. Cir. 2019) (explaining how merger of complement producers benefits consumers by eliminating double marginalization). In short, both individual component sellers (here, the three defendant programmers) and consumers benefit from the elimination of double—or here, triple—marginalization.

Because the second competitive scenario the district court envisioned contemplates each defendant’s separately pricing its sports content, it would entail triple marginalization. Each defendant’s attempt to set its individual price so as to maximize its share of the combined price consumers would be willing to pay for a Venu-like bundle would then result in a combined price that would exceed Venu’s. Thus, any bundle that consumers could put together of separate live sports-streaming services launched by each individual defendant would likely be more expensive than Venu (because of triple marginalization) or less comprehensive (if the consumer decided to cobble together less than all the high-priced content offered by Venu). Indeed, it is likely that the defendants recognized this pricing problem and arrived at the Venu joint venture as the solution. There is no basis for assuming that any other hypothesized arrangement would effectively solve the problem.

In the end, then, the district court’s enjoining of Venu sacrifices a competitive “bird in the hand”—a comprehensive, low-priced live sports-streaming service whose entry would enhance competition in the live pay TV market and benefit consumers—for a highly speculative “bird in the bush” of economically implausible future sports-streaming offerings. Because such an outcome would harm rather than further the public interest, this Court should vacate the district court’s injunction.

CONCLUSION

The antitrust laws should remain singularly focused on the protection of market competition, and this Court should resist efforts by individual firms to coopt its remedial powers to insulate themselves from the competitive process. Accordingly, for the foregoing reasons, the Court should reverse the decision below.

[1].       Under Federal Rule of Appellate Procedure 29(c), amici curiae state that no party’s counsel authored this brief in whole or in part, and no party or its counsel made a monetary contribution intended to fund the preparation or submission of this brief. No person other than amici or their counsel contributed money that was intended to fund preparing or submitting the brief. All parties have consented to amici’s filing of this brief.

[2].       See also United States v. Phila. Nat’l Bank, 374 U.S. 321, 367 n.43 (1963) (quoting Brown Shoe, 370 U.S. at 320); Copperweld Corp. v. Indep. Tube Corp., 467 U.S. 752, 767, n.14 (1984) (quoting Brunswick, 429 U.S. at 488); Cargill, Inc. v. Monfort of Colo., Inc., 479 U.S. 104, 110 (1986) (quoting Brunswick, 429 U.S. at 488); Atl. Richfield Co. v. USA Petroleum Co., 495 U.S. 328, 338 (1990) (quoting Brown Shoe, 370 U.S. at 320); Brooke Grp. Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 224 (1993) (quoting Brown Shoe, 370 U.S. at 320); Leegin Creative, 551 U.S. at 906 (quoting Atl. Richfield, 495 U.S. at 338).

[3].       Amici take no position here on whether the district court correctly defined the relevant market. For reasons stated below, even if the district court’s definition of the relevant market its correct, the preliminary injunction it entered is unwarranted. August 20, 2024 Op. & Order (“O&O”) ECF 2.

[4].       Note that with respect to the upstream duty to deal, the Court held that “a firm with no duty to deal in the wholesale market has no obligation to deal under terms and conditions favorable to its competitors.” Linkline, 555 U.S. at 450–51. And with respect to predatory pricing in the downstream market, the Court held that “to prevail on a predatory pricing claim, a plaintiff must demonstrate that: (1) ‘the prices complained of are below an appropriate measure of its rival’s costs’; and (2) there is a ‘dangerous probability’ that the defendant will be able to recoup its ‘investment’ in below-cost prices.” Id. at 451 (citing Brooke Grp., 509 U.S. at 222–24).

[5].       This is no surprise. As the district court observed, bundling of television programming by content providers is a ubiquitous and long-standing practice. O&O at 2 (“These bundling requirements are not unique to Fubo’s contracts with the JV Defendants; bundling has been a pervasive industry practice for decades. . . . ”); id. at 10 (“Bundling has been an industry-wide practice for at least four decades. Bundling is ubiquitous because in many cases, at least some subset of consumers enjoy having ready access to hundreds of channels and doing so on a less-expensive basis than they otherwise would if they paid for each channel individually.”). Prior antitrust challenges to such bundling have failed, see Brantley v. NBC Universal, Inc., 675 F.3d 1192 (9th Cir. 2012), and the practice generates efficiencies. See Thomas A. Lambert, The Efficiency of Cable Bundling, Truth on the Market (July 10, 2011) (https://truthonthemarket.com/2011/07/10/the-efficiency-of-cable-bundling/); Thomas A. Lambert, Appropriate Liability Rules for Tying and Bundled Discounting, 72 Ohio St. L.J. 909, 950–53 (2011); Yannis Bakos & Eric Brynjolfsson, Bundling Information Goods: Pricing, Profits, and Efficiency, 45 Mgmt. Sci. 1613 (1999). There is also no logical stopping point to a “Thou shalt offer TV content on an unbundled basis” rule. Must a programmer license individual television shows (e.g., Seinfeld only)? Individual seasons (e.g., only Seinfeld season eight)? Individual episodes (e.g., only episode 138: “The Little Kicks”)? Individual scenes within episodes (e.g., the one where Elaine dances)? As the Supreme Court has cautioned (quoting Professor Phillip Areeda), “No court should impose a duty to deal that it cannot explain or adequately and reasonably supervise. The problem should be deemed irremediable by antitrust law when compulsory access requires the court to assume the day-to-day controls characteristic of a regulatory agency.” Phillip Areeda, Essential Facilities: An Epithet in Need of Limiting Principles, 58 Antitrust L.J. 841, 853 (1989). Verizon Commc’ns Inc. v. Law Offs. of Curtis V. Trinko, LLP, 540 U.S. 398, 415 (2004).

[6].       Trinko is highly relevant to Fubo’s claim that defendants owed some antitrust duty to license their sports content on an unbundled basis. Because the district court did not assess the legality of defendants’ bundling practices, amici do not discuss Trinko in detail here.

[7].       In Buccaneer, two natural gas firms refused to transport gas for a rival using their jointly owned pipelines. 846 F.3d at 1302, 1306. The refusal to deal was itself concerted conduct. Id. at 1306 (“Buccaneer contends Defendants’ agreement to deny it reasonable access to the RM System was a concerted refusal to deal that violated § 1 of the Sherman Act.”).

[8].       See also Engine Specialties, Inc. v. Bombardier Ltd., 605 F.2d 1, 11 (1st Cir. 1979) (agreement that “neither of the parties to the joint venture will compete with it” is “not offensive in and of itself”); Princo Corp. v. ITC, 616 F.3d 1318, 1336 (Fed. Cir. 2010) (observing that “ancillary restraints” that are often “important to collaborative ventures [include] agreements between the collaborators not to compete against their joint venture”); In re HIV Antitrust Litig., 656 F. Supp.3d 963, 993 (N.D. Cal. 2023) (observing that “[c]ourts and other authorities have recognized that free riding is a legitimate concern when people or entities embark on a joint venture” and concluding that noncompete provisions among collaborating drug companies “may have facilitated the collaboration[]” because “they arguably prevented a collaborator from free riding on the efforts of the joint venture”); Philip Areeda & Herbert Hovenkamp, Antitrust Law: An Analysis of Antitrust Principles and Their Application (CCH) 2213c2 (2018) (observing that: “[J]oint venturers may have quite legitimate reasons for restraining members’ competitive business outside the venture. Most such concerns apply to some variation of the free rider problem.”).

[9].       Antitrust law permits—indeed, encourages—firms that have gained monopoly power legitimately to earn supracompetitive profits. See Linkline, 555 U.S. at 454 (“[A]ntitrust law does not prohibit lawfully obtained monopolies from charging monopoly prices.”); Trinko, 540 U.S. at 407 (“The mere possession of monopoly power, and the concomitant charging of monopoly prices, is not only not unlawful; it is an important element of the free market system. The opportunity to charge monopoly prices—at least for a short period—is what attracts ‘business acumen’ in the first place; it induces risk taking that produces innovation and economic growth.”).

[10].      The basic concept of the double marginalization pricing externality was introduced by Cournot in 1838, Augustin Cournot, Researches into the Mathematical Principles of the Theory of Wealth (Nathaniel T. Bacon, trans., 1897), and formalized by Sonnenschein in 1968, Hugo Sonnenschein, The Dual of Duopoly Is Complementary Monopoly: or, Two of Cournot’s Theories Are One, 76 J. Pol. Econ. 316 (1968). For an informative video explanation of the double marginalization problem, see Marginal Revolution University, Double Marginalization Problem, https://www.youtube.com/ watch?v=7MPdKMeGcv8.

Executive Summary

Privately run digital-payments systems are helping to bring informal market transactions and unbanked people into the formal economies of many countries—in the process, improving livelihoods and fostering innovation and economic growth.

The success these private systems experienced in expanding financial inclusion has led some governments to create their own systems for digital payments. “India Stack,” for example, combines a national ID and associated database (Aadhaar), an interoperability framework for real-time payments (UPI), and a set of standards for sharing financial data (DEPA). Brazil, meanwhile, has focused on the middle part of that stack, with the nation’s central bank, BCB, creating its own real-time payment system (Pix).

The results of these interventions have been mixed. Where they have supported natural market developments (as is largely the case with Aadhaar), they have had a positive impact. But where they have competed directly with market actors, they have created distortions that may inhibit more effective market solutions. For example:

• Restrictions on interchange fees imposed by UPI have favored market participants who are able to monetize the operation of payment networks in other ways, such as Google (which monetizes primarily through search advertising) and Walmart (which monetizes through the sale of goods).
• Similarly, subsidies from the BCB and Pix’s restrictions on interchange fees crowd out private-sector competition and undermine innovations that might lead to increased financial inclusion.

In some cases, governmental attempts to force a shift toward digital money have impeded financial inclusion. India’s “demonetization” in 2016 harmed M-Pesa, contributing to greater financial exclusion. Arbitrary governmental regulations, including restrictions on the number of WhatsApp Pay users, have also likely inhibited both competition and financial inclusion.

The key lesson is that the private sector is well-placed to identify opportunities for digital payments and to implement solutions appropriate to local consumer culture and other contingent facts. Meanwhile, governments should focus on delivering genuine public goods, such as identity verification, and avoid the temptation to participate in digital-payments infrastructure.

I. Introduction

Digital-payment systems are expanding access to financial services around the world. In many countries, they are helping to bring the previously unbanked into the formal economy. In so doing, they undoubtedly improve livelihoods and foster innovation and economic growth. But not all digital-payment systems are alike.

Early digital-payment systems were largely private-sector initiatives. Notable among these are: M-Pesa in Kenya; Venmo in the United States; AliPay and WeChat Pay in China; and MobiKwik and Paytm in India. These emerged as bottom-up responses to the felt needs of consumers and merchants.

The success these private systems experienced in expanding financial inclusion has led some governments to create their own systems for digital payments. “India Stack,” for example, combines a national ID and associated database (Aadhaar), an interoperability framework for real-time payments (UPI), and a set of standards for sharing financial data (DEPA). Brazil, meanwhile, has focused on the middle part of that stack, with the nation’s central bank, BCB, creating its own real-time payment system (Pix).

To better understand the costs and benefits of government interventions, this brief investigates the economics of mobile digital payments, their effects on financial inclusion, and the consequences of specific government interventions. Section II begins with a discussion of the core problems that payment systems must address, providing a law & economics framing. Section III describes the development of digital-payment platforms in Kenya, China, India, and Brazil. Section IV puts the evidence presented in section III into broader context and draws conclusions for policymakers.

II. The Economics of Mobile Digital Payments and Financial Inclusion

This section briefly considers the economics of payments, focusing on the two core problems that all payments systems must address: counterparty risk and the two-sided nature of the payments markets.

A. Trust and Counterparty Risk

In all transactions, there is a risk that a counterparty will not perform the promised action. For example, in the sale of any good, the seller may fail to deliver that good; the good may not function as described; or the buyer may fail to make payment. This is particularly problematic in the context of transactions among parties who are unknown to one another, or who are separated by time and space, as is the case in many online transactions.

One way to address counterparty risk is through the law of contract. In common-law jurisdictions, buyers may bring breach-of-contract actions against sellers if the goods purchased are damaged, defective, or otherwise not fit for purpose (or some equivalent term). Meanwhile, sellers may bring breach-of-contract actions against buyers who fail to make payment as agreed in a timely manner.

Often, however, the cost of legal action is large relative to the value of a transaction. As such, while contract law may function as a backstop, it is usually more efficient to resolve such disputes through alternative means. In the online marketplace, entrepreneurs have developed several important technologies that serve to facilitate trust and otherwise address counterparty risk; these include user ratings, security protocols and fraud-detection systems, payment gateways, and online escrow systems. The digital-payment systems discussed in Section III typically rely on some or all of these technologies. (The Appendix includes a more detailed discussion of the various ways counterparty risk has been addressed in the context of online transactions.)

B. Two-Sided Markets

Payment systems are an example of what economists call “two-sided markets”: buyers are on one side, sellers on the other, and the payment system itself acts as the platform that facilitates interactions between them.[1] The primary challenge faced by any two-sided market is to ensure that there are sufficient participants on both sides for it to be self-sustaining. Conceptually, this is known as a cross-side network effect.

In the case of payment systems, if too few sellers accept a particular form of payment, buyers will have little reason to adopt it. Likewise, if too few buyers hold a particular form of payment, sellers will have little reason to accept it. This applies equally to other two-sided markets, such as shopping malls. For customers to want to go to a mall, there must be stores that they want to visit. And for stores to want to locate in a mall, they need to know that there will be customers who will visit.

Two-sided markets also must cover the operational costs. In the case of shopping malls, this means the costs of building and operating the mall, including ongoing maintenance, lighting and heating/cooling of common areas, and various services, usually including security. In the case of payment networks, it includes the development and maintenance of technology needed to facilitate secure transactions, verify identity, and deter fraud.

Platforms typically address these problems simultaneously by charging differential fees and offering incentives. Thus, malls typically charge store owners (one side of the market) rent and service fees sufficient to cover both capital and operational costs. Those store owners thus effectively subsidize access to the mall for customers (the other side of the market), who are generally free to walk around enjoying the mall’s climate and security even if they do not ultimately make any purchases. This highlights a feature common to most two-sided markets: it is usually efficient for one side to subsidize the other, thereby maximizing the value to all participants. Typically, the side that provides the subsidy has a lower price elasticity of demand (i.e., it is less sensitive to changes in the cost of platform services).[2]

Meanwhile, to encourage popular retailers to locate in the mall in the first place, mall owners may offer discounted rent to early occupants and/or specific retailers who are especially likely to attract customers (such as hotels and restaurants with a loyal following).[3] This highlights another feature of two-sided markets: it may also be efficient for some participants on one side to subsidize other participants on the same side, again in order to maximize the value to all participants.

Like malls, payment networks typically facilitate cross-network subsidies. Merchants typically place a higher value on the ability to accept a specific payment type than consumers place on the ability to use that payment type. In other words, merchants have lower price elasticity of demand for the payment network. Because of this, the subsidy typically flows from the merchant to the consumer.

In some cases, payment networks also exhibit own-side subsidies. For example, smaller-ticket items may be subject to proportionally lower fees—a merchant-side subsidy. Meanwhile, some consumers may receive an initial rewards bonus and/or pay a lower fee for the first year they use a certain payment type (such as a credit card)—a consumer-side subsidy.

III. Mobile Digital-Payment Platforms

While Venmo and Zelle may be better known in the United States, the first cellphone-based payment networks first emerged informally in Tanzania and some other sub-Saharan African countries in the early 2000s, allowing users to send cellphone airtime minutes from one phone to another.[4] These informal networks inspired the development of more formal systems, such as M-Pesa. In China, meanwhile, mobile payments emerged from two different sources: an online payment system (Alipay) and a cellphone-messaging service (WeChat). India’s early leader in mobile payments, Paytm, grew out of a cellphone-charging platform.

A. M-Pesa and the Mobile-Payments Revolution in Africa

In 2006, Kenya had 41 retail banks with 400 branches and 600 automated-teller machines (ATMs) for a population of 36 million.[5] By comparison, the United States at the time had 7,523 banks with 72,958 branches and 400,000 ATMs for a population of 296 million. On a per-capita bases, that’s nearly 20 times the number of banks and branches, and 80 times as many ATMs.[6]

Unsurprisingly, Kenyans—especially those in rural areas—were not major users of the banking system. A survey undertaken in 2006 found that just 18.9% of Kenyan adults had access to a bank account or other formal financial-services provider, such as insurance, while a further 7.5% had accounts with semi-formal savings and credit organizations and/or a microfinance institution. Thus, no more than about a quarter of the adult population had access to a formal or even semiformal financial-services provider.[7] Of the remainder, slightly less than half of the population (35.2% of those surveyed) had access to informal financial services, while the rest (38.4%) were considered “financially excluded.” The proportion of adults with access to electronic payments was even smaller, with less than 10% having a credit card.[8]

Three years before that survey was undertaken, Nick Hughes, then head of social enterprises at Vodafone, learned of the use of air-time credits as an informal currency in Tanzania, and believed it should be possible to create a more formal system to facilitate microcredit and microinsurance—services that were popular among development groups at the time. [9] Hughes put together a proposal to do just that and submitted it to the UK government’s Department for International Development (DFID). The DFID awarded Vodafone £1 million for a pilot program, which Vodafone matched with a combination of cash and personnel time.

For the pilot project, which launched in October 2005, Hughes selected Commercial Bank of Africa (CBA), which would hold customer funds in a trust account; Faulu Kenya, a microfinance institution that would provide loans; and Vodafone’s 40% owned subsidiary Safaricom.[10]

Hughes called the system M-Pesa: “pesa” is Swahili for money and M stands for mobile, so M-Pesa is “mobile money.” The pilot project proved to be a wild hit, though not primarily for microfinance loans. It was mainly used to send money from one person to another, including people who were not part of the pilot. For example, it was used for business transactions, for remittances, and to store money overnight.[11]

In 2007, Safaricom formally launched M-Pesa at-scale. That meant onboarding thousands of agents, who receive funds deposits from accountholders, and whose identity they verify using a national ID card or passport. The agents add those funds to the system, as well as pay-out funds that are withdrawn. By 2012, there were more than 30,000 M-Pesa agents. As of October 2023, there were 160,000 M-Pesa agents in Kenya alone, and 500,000 across all the markets in which M-Pesa operates.[12]

M-Pesa began as a text-based system with limited functionality. As smartphones became more prevalent, it expanded to include a wider range of financial services. In 2011, Safaricom partnered with CBA to enable users to create a linked bank account, called M-Shwari, that enables M-Pesa users to save and borrow.[13] By 2015, more than 10 million Kenyans had opened M-Shwari bank accounts; for many, this represented their first such formal bank account.[14] CBA also launched similar products in Tanzania, Uganda, Rwanda, and Cote d’Ivoire.[15]

Using a mobile app or web interface, it is also now possible to pay money into M-Pesa electronically by linking to a bank account.[16] Since 2022, M-Pesa has had a partnership with Visa, which enables M-Pesa accountholders to establish a virtual account that can then be used to make international purchases online.[17]

According to the most recent FinAccess report, M-Pesa helped to expand financial inclusion in Kenya from 26% of the population in 2006 to 84% in 2021.[18] It has also facilitated economic growth and, according to a study published in Science by MIT economists Tavneet Suri and William Jack, has helped to lift hundreds of thousands out of poverty.[19]

M-Pesa has been so successful that it has expanded to the Democratic Republic of Congo (DRC), Egypt, Ghana, Lesotho, Mozambique, and Tanzania. It now boasts more than 51 million users and $314 billion in annual transactions.[20] Other mobile operators have also followed suit, both in Kenya and across Africa, with Orange providing its “Orange Money” service to more than 80 million customers in 17 countries in Africa and the Middle East.[21]

B. China

The success of mobile-payments systems in Africa has been much vaunted. But by some metrics (e.g., rate of adoption, value of transactions, and broad economic effects), it pales in comparison with the success of the two primary mobile-payments systems that have developed in China: Alipay and WeChat Pay.

1.  AliPay

In 2003, there were only 3 million credit cards issued in China, implying that less than 5% of adults had a card.[22] Meanwhile, China had around 20 million internet-connected computers and 60 million internet accounts.[23] The vast majority of Chinese internet users thus lacked a reliable means to make online purchases. This posed significant problems for online auctions and other consumer-to-consumer marketplaces, where there is an inherent trust deficit between buyer and seller—and associated counterparty risk.[24]

To address this problem, in 2003, online retailer Alibaba introduced an escrow service, called Alipay, for its consumer-to-consumer subsidiary Taobao.[25] When making a payment, funds are initially deposited in Alipay’s escrow account and only released to the seller once both parties are satisfied that the transaction is complete.[26] Initially, users would fund their Alipay personal digital wallet by linking it to their bank account.[27] Over time, Alipay has expanded the ways in which wallets can be funded to include debit cards, credit cards, bank transfers, and other digital wallets.[28]

Alipay’s mobile wallet, launched in 2008, had grown to roughly 100 million users by 2013. Today, it has more than 900 million users in China and about 1.3 billion users around the world.[29]

2. WeChat Pay

WeChat Pay (officially “Weixin Pay”) is a mobile wallet and payments system launched in 2014 with the express intention of enabling users of the WeChat messaging app to send money to one another.[30] To drive adoption, WeChat introduced the ability to share virtual “red envelopes” with family and friends over the Lunar New Year holiday—replicating an ancient tradition.[31] This proved enormously popular, with 20 million red envelopes sent in the first year, and 3.2 billion sent in the following year. Over time, WeChat’s owner Tencent has built an ecosystem around WeChat Pay, including mini-programs that enable the user to hail rides or pay for movie tickets.[32] As of 2022, about 60% of WeChat users use these mini-programs.[33]

Unlike AliPay, WeChat Pay does not provide escrow services, and is thus less well-suited to certain kinds of online transactions. It works well, however, both for in-person payments to merchants and transfers to other individuals (including gift giving). WeChat Pay has approximately 1.1 billion users worldwide.[34]

3. Discussion

AliPay and WeChat Pay’s precursor (TenPay) both began life as mechanisms to enable payments within existing ecosystems of products and services. AliPay’s pivot to mobile payments proved an enormous success, motivating Tencent to create its own mobile-payments system. Over time, both systems have gradually expanded to provide payment services outside their own ecosystems. Payments using either platform can be made at an increasingly large number of merchants using a range of technologies, including quick-response (QR) codes for in-store or peer-to-peer (P2P) payments.[35]

For nearly a decade, AliPay and WeChat Pay have been engaged in fierce competition. As access to credit cards in China has grown, this competition has expanded to include China Union Pay. As of 2022, AliPay and WeChat Pay were by far the most used means of making online payments (see Figure 1).

FIGURE 1: Most Used Online-Payment Services in China (2022)

SOURCE: Daxue Consulting[36]

In 2015, Tencent and Alibaba established banks associated with their payment networks. WeBank (Tencent) and MyBank (Alibaba, now part of Ant Financial) have since grown rapidly. Ant Financial reports that, as of June 2022, MyBank had provided loans to more than 49 million small and medium-sized enterprise (SMEs), of which 80% had not previously had a bank loan.[37] WeBank, meanwhile, reports having more than 370 million individual clients and more than 4 million micro-small-and-medium-sized-enterprise (MSME) clients.[38]

In July 2023, the Chinese authorities permitted AliPay and WeChat Pay to integrate payments made using Mastercard, Visa, American Express, and Discover.[39] This, combined with increasing adoption of China Union Pay cards domestically, has led to an increase in the use of AliPay and WeChat Pay as mobile-payment gateways, in contrast to their role as digital wallets.

The emergence and expansion of digital payments in China has undoubtedly resulted in improvements in access to financial services, both directly (since payments are a financial service) and indirectly, by enabling tens of millions of businesses and hundreds of millions of individuals to hold bank accounts and thereby access related services, including loans.

C. India

India’s mobile-payment system has grown rapidly over the past decade. The market’s key players are MobiKwik, Paytm, PhonePe, and Google Pay. This section discusses those services, as well as WhatsApp Pay and M-Pesa, which have been less successful in India.

A major factor affecting the development of mobile payments in India has been the national digital ID (Aadhaar) and the Unified Payment Interface (UPI), both part of what is now referred to as “India Stack.”[40] I plan to look at India Stack in more detail in a future piece, but it is worth noting the following:

1. Aadhaar was developed by the Unique Identification Authority of India (UIDAI), a government authority established in 2009, in order to enable every resident of India to have a unique digital identity that could be instantaneously verified. The goal was to thereby reduce identity theft and related fraud, and facilitate more rapid onboarding to banks and other financial entities that require proof of identity for know-your-customer (KYC) purposes.[41] UIDAI began enrolling Indian residents in Aadhaar in 2010 and, by the end of 2015, had enrolled more than 800 million people.[42] As of February 2023, 1.38 billion Indian residents are enrolled—about 96% of the population.[43]
2. UPI was developed in 2016 by the National Payments Corporation of India (NPCI), a public-private partnership, as an open-source interoperable application programming interface (API) that facilitates real-time transfers between individuals with accounts at participating banks who have integrated the API into their smartphone apps.[44]

1. MobiKwik

Among the first mobile-payments systems in India, MobiKwik was initially established in 2009 by husband-and-wife team Bipin Preet Singh and Upasana Taku.[45] In 2012, it launched a digital wallet feature and, in 2013, the Reserve Bank of India awarded it a license for that wallet.[46] MobiKwik grew organically, aided by significant investments from Bajaj Capital and Sequoia, and signed up 1.5 million merchants and 55 million registered users by 2015.[47] It has continued to grow in the nearly decade since, albeit at a slower rate, recording around 4 million merchants, 140 million registered users, and 35 million active monthly users in 2023.[48]

MobiKwik generates revenue from commissions and advertisements from its Zaak payment-gateway franchise subsidiary,[49] as well as loans—including short-term credit, buy-now-pay-later, and personal loans—and investment advice.[50]

2. Paytm

Paytm began life in 2010 as a prepaid mobile-charging platform, but pivoted to payments in 2014 with the introduction of a mobile wallet. Paytm now generates revenue from three main sources: payment services, financial services, and “commerce and cloud.” Payment services—the core of its business, contributing 58% of its revenue in Q3 2023[51]—arise from users making payments from mobile wallets, debit cards, and credit cards, for which Paytm charges merchants a fee (merchant-discount rate) that ranges from 0.4% to 2.99% of the transaction amount, depending on the payment type (for small to medium-size businesses).[52] Financial services—including loans, investments, and insurance—accounted for 22% of Paytm’s revenue in the most recent quarter, while “commerce and cloud” services—Paytm’s marketplace that features more than 100,000 merchants—accounted for 20% of revenue.

This combination of offerings has allowed Paytm to become a highly effective multisided market. A year after launching its mobile wallet, Paytm had 100 million registered accounts, making it the early leader in the field in India.[53] The user base grew by 50% in 2016,[54] likely in part as a result of the government’s decision to cancel the majority of its banknotes.[55] It now has approximately 100 million active users.[56] While Paytm’s wallet remains the most popular digital wallet in India, its share of the mobile-payments market has declined as a result of competition from PhonePe and Google Pay.

In addition to offering payments, Paytm provides a range of financial services, including personal and business loans. It also operates its own ecommerce platform.

3. PhonePe

Created in 2015 by former Flipkart executives Sameer Nagim and Rahul Chari, PhonePe is India’s market-leading mobile-payments platform and was acquired by Flipkart in 2016 prior to its launch.[57] Flipkart itself—founded in Bengaluru in 2007—was purchased by Walmart in 2018.[58] In December 2022, PhonePe was separated from Flipkart, though both remain majority owned by Walmart.[59]

By launching as a wholly owned subsidiary of Flipkart, India’s largest online marketplace, PhonePe was able to leverage the marketplace’s then 100 million users (it has since grown to 400 million),[60] enabling it rapidly to acquire customers. PhonePe also differentiated itself from Paytm by primarily targeting middle-class consumers and associated merchants in second-tier cities.[61] PhonePe was also an early adopter of the Unified Payments Interface (UPI), a real-time payments solution that facilitates account-to-account transfers at no cost to either party.

Since separating from Flipkart, PhonePe’s offerings are now remarkably similar to those of Paytm. It does, however, have a larger user base (approximately 200 million monthly active users, double Paytm’s 100 million) its historic connection to Flipkart enables it more effectively profit from low-margin, high-volume activities.

4. Google Pay

Google Pay, which launched in India in 2017 (originally under the brand Tez), is now the second most popular mobile-payments system in India.[62] Google Pay does not act as a wallet in India, but is essentially a payment gateway that works exclusively with UPI.[63]  Google Pay has built a large Indian customer base due largely to brand recognition. Google is able to monetize Google Pay through advertising and its local online marketplace.[64] Google has also been granted a license to operate as a payment aggregator.[65]

5. WhatsApp Pay

Meta launched a pilot program for WhatsApp Pay in India in 2017, but did not fully launch until 2020 and has been subject to limits on its number of users.[66] Despite operating the most popular messaging service in the country, with more than 400 million users, WhatsApp Pay has not taken off. Indeed, its peak market share was about 0.4%, and it has since fallen to about 0.1%.[67]

WhatsApp Pay’s failure in India is likely due in no small part to regulatory delays in launching and the NPCI-imposed limits on the number of users—initially 40 million, but increased to 100 million in 2022.[68] The delay in launching meant that PhonePe and Google Pay were able to gain the upper hand on WhatsApp, becoming the predominant UPI-based wallet and payment gateways, respectively. The limit on signups, meanwhile, has forced WhatsApp to impose arbitrary restrictions on which customers can adopt its payment service, which effectively undermines the self-organization that underpins P2P-transaction networks.

6. M-Pesa

M-Pesa entered the Indian market in 2011, hoping to repeat its success in Kenya. Despite attracting more than 400 million Indian subscribers, many of them in rural areas, Vodafone was only able to onboard 8.4 million subscribers and eventually shut down the operation in 2019.[69] In an interview in 2020, the former head of M-Pesa, Michael Joseph, identified four factors that explained the failure:[70]

• Difficulty recruiting agents: In Kenya, Safaricom had invested heavily in building a large network of M-Pesa agents across the country, including in remote villages. This allowed easy cash deposits and withdrawals. But in India, Vodafone struggled to find shops and merchants to serve as last-mile M-Pesa agents in rural areas.
• Lack of understanding among consumers: In contrast to Kenya, poor and unbanked Indian consumers needed more handholding and awareness-building to adopt M-Pesa, which required significant time, money, and labor from Vodafone.
• Competition from well-funded startups targeting more profitable market segments: Meanwhile, mobile-payment startups such as Paytm targeted urban middle-class and affluent consumers. They could link to bank accounts, unlike M-Pesa, which dealt in cash for the unbanked. Paytm also had significant investor backing.
• Demonetization: The 2016 Indian demonetization benefited digital payments that were linked to bank accounts, but it harmed M-Pesa, which focused primarily on the unbanked.

7. Discussion

The veritable explosion in mobile payments in India over the past decade has undoubtedly contributed to improvements in financial inclusion. India’s primary mobile-payments apps—MobiKwik, Paytm, PhonePe, and Google Pay—have sought to leverage their platforms to maximize value for both sides of the market (merchants and consumers) in different ways:

• MobiKwik (Zaakpay), Paytm, PhonePe, and Google Pay provide payment-gateway and/or aggregator services.[71]
• MobiKwik, Paytm, and PhonePe offer users wallet features and seek to supply additional financial services, such as insurance and investment.
• Paytm, PhonePe, and Google Pay generate revenue through online marketplaces.
• All the apps (but especially Google Pay) generate revenue from advertising.

As noted in the introduction to this section and throughout, UPI likely contributed to the rapid adoption of mobile payments, especially for PhonePe and Google Pay, which were built around it. The fact that UPI requires parties to have a bank account has likely created incentives to open such account, thereby increasing financial inclusion. The provision of welfare support through e-RUPI vouchers that are delivered over UPI also likely increases financial inclusion.[72]

Some aspects of UPI, however, may also have hindered more widespread adoption of mobile digital payments and inhibited financial inclusion. For example, NPCI currently prohibits card and app operators from charging merchants for most transactions made using UPI (fees of up to 1% are permitted only for prepaid debit cards and pre-funded mobile wallets). Those card issuers and app operators who lack existing means to monetize their relationships with customers are therefore likely to find it difficult to break into the market using UPI, as they cannot easily offer rewards and other inducements to customers to on-board and use their card or app.

The Payments Council of India estimates that its members lose 55 billion rupees ($660 million) annually as a result of the zero MDR on UPI and RuPay transactions.[73] This is effectively a transfer from banks to the companies whose apps monetize UPI transactions. India’s government partly offsets this loss through a subsidy to UPI participants of between 15 and 25 billion rupees.[74] But experience with other systems that impose restrictions on payment-transaction fees suggests that banks will seek to recover these losses from other fees.[75] To the extent that such additional fees fall on lower-income accountholders, the effect on financial inclusion is likely negative. (I plan to explore this and related issues in more detail in a future brief on India Stack.)

D. Brazil

The first mobile-payments system in Brazil appears to have been Oi Paggo, a closed-loop short-messaging-service (SMS) “credit card” system founded in 2004 and formally launched in 2006 as an app on the Oi mobile-phone service (it was subsequently purchased by Oi).[76] In 2008, Oi Paggo reportedly had around 1 million users, but does not appear to have achieved mass adoption.[77]

Over the course of the past dozen years, several other entrants have built successful digital mobile-payments systems. According to a recent Survey by Statista, the top five most-used mobile-payments systems in Brazil at the end of 2023, based on proportion of respondents who had used the payment online or in stores in the previous year, were: PicPay (50%), PayPal (46%), Mercado Pago (44%), PagBank (31%), and Google Pay (25%).[78]

Before discussing these mobile-payment solutions, it is worth briefly noting that, in 2021, the Central Bank of Brazil (BCB) introduced its own real-time payment system, Pix, and required all of Brazil’s larger banks to implement it for all accountholders. As a result, Pix has had a significant effect on mobile digital payments in Brazil.

1. PicPay

PicPay was founded in 2012 as a digital wallet and initially launched in the city of Vitória, where it has gained significant traction, with approximately 75% of the population using it.[79] It subsequently launched in Rio de Janeiro and São Paulo. In 2022, PicPay—owned by closely held J&F Investments—obtained a banking license via J&F’s bank, Banco Original, enabling it to expand its offerings and lower its funding cost.[80] PicPay has more than 30 million active users and is accepted at more than 10 million merchants.[81] It generates revenue from a range of product lines, including loans, credit cards, investments, insurance, and a crypto-currency exchange.[82]

2. PayPal

Originally conceived as a payment system in its own right, PayPal functions both as a payment gateway—processing card payments—and as a wallet. In Brazil, PayPal also has a Pix integration, which allows users to send and receive payments to their PayPal account using their Pix keys.[83]

3. Mercado Pago

A subsidiary of the online marketplace Mercado Libre, Mercado Pago is able to leverage the customer base of that market, offering a full range of payments and banking services, including checking accounts, deposit accounts, loans, debit cards, and credit cards.[84] Mercado Pago operates in several countries and reported having nearly 65 million accountholders in 2023.[85]

4. PagBank and PagSeguro

PagSeguro was founded in 2006 as an online payment system by Universo Online (UOL), Brazil’s largest internet-content and services provider.[86] In 2013, PagSeguro diversified into point-of-sale (POS) payments and then into banking, offering a range of services. PagBank now has around 15 million active Brazilian users and 7.7 million merchants.[87]

5. Google Pay

In Brazil, Google Pay operates primarily as a payment gateway, facilitating transactions made using debit and credit cards whose (tokenized) information is stored on a user’s phone and transmitted to POS machines using radio-frequency identification (RFID).[88]

6. Discussion

The above-mentioned survey suggests that there is considerable multi-homing for mobile digital payments in Brazil, with the same consumers using several different mobile-payments apps. It also seems plausible that these apps have contributed to financial inclusion. As Figure 2 shows, in the five years from 2017 to 2022, there was a dramatic shift away from the use of cash toward a combination of card and digital mobile payments.

FIGURE 2: POS Payments in Brazil by Method

SOURCE: Statista

Meanwhile, the proportion of Brazilians with accounts at a financial institution rose from 70% in 2017 to 84% in 2021. Some have suggested that this was the result of Pix’s introduction, but that appears unlikely, as Pix was only introduced at the end of 2020. Two other factors seem better candidates: first, the growing use of digital payments and associated onboarding by financial-technology firms (fintechs), such as PagBank; second, some government subsidies offered in response to the COVID-19 pandemic required Brazilians to open bank accounts.

A study by Americas Market Intelligence (commissioned by Mastercard) found that, during the COVID-19 pandemic: “Brazil reduced its unbanked population by an astounding 73%.”[89] The study was based on research conducted between June and August 2020 and was published in October 2020, the month before Pix launched. It described the implementation of state and federal programs that Brazil launched in response to the pandemic:

• The “Coronavoucher” program distributed emergency funds to low-income informal workers exclusively via the state-owned bank Caixa Econômica Federal (CEF). Applications for funds could only be made via CEF’s Caixa Tem smartphone app, and funds were distributed using the same app. As of Aug. 5, 2020, 66 million people had received Coronavouchers via the Caixa Tem app. Of those, 36 million were previously unbanked.
• Merenda em Casa (“snack at home”), a program run by state governments, distributed funds to low-income families with children at public schools to help pay for food while schools were closed due to COVID-19. The program distributed funds via PicPay and PagBank’s PagSeguro, both private-sector payment apps.[90]

Following the launch of Pix, the BCB-run RTP program was made available to clients of Caixa Tem, PicPay, and PagBank.[91] As a result, previously unbanked individuals who had become banked because of the Coronavoucher and Merenda em Casa programs were able to obtain and use Pix keys to send and receive payments.

IV. Conclusions and Policy Implications

At one level, every successful payment system solves the same set of problems—in essence: creating a self-sustaining network that facilitates the transfer of funds, while limiting counterparty risk. But as this study shows, each system does so in ways that are economically, legally, and culturally contingent.

In Kenya—and, subsequently, many other countries—M-Pesa provided a means for millions of people without ready access to banks to send and receive money over long distances. To do so, it created a network of tens of thousands of agents providing on- and off-ramps for cash. Over time, this has gradually transformed into a broader ecosystem that includes banking and other financial services.

In China, AliPay provided a solution to counterparty risk for internet-based transactions in an economy where alternative solutions, such as credit cards, were not widely available. AliPay was able to leverage its reputation for reliable and secure payments to facilitate transactions using increasingly widely available smartphones. Meanwhile, WeChat Pay was able to leverage its network of hundreds of millions of chat users to establish an effective mobile-payments ecosystem and associated markets for goods and services.

In India, MobiKwik and Paytm both offered digital wallets as an alternative to card payments for middle-class consumers in an economy where credit cards were tightly regulated. But over time, and especially with the introduction of UPI, all of the mobile-payment systems’ business models have converged around offering a range of financial products and markets.

In Brazil, an early attempt to introduce a mobile digital platform predicated on the credit-card model seems to have foundered, but later entrants—notably PicPay, PayPal, Mercado Pago, PagSeguro and other players—have been able to build and monetize payment networks by offering payment-gateway services and a range of adjacent products.

Public policy has played a substantial role in the development of all these mobile-payments ecosystems. In some cases, the effects have likely been positive. For example, India’s Aadhaar digital ID has dramatically reduced the cost of proving the identity of hundreds of millions of people, thereby facilitating quicker and less expensive access to bank accounts. The distribution of welfare payments via payment apps in India and Brazil also likely increased adoption at the margin.

On the other hand, restrictions on transaction fees in India (zero MDR on UPI and RuPay) and Brazil (extremely low interchange fees permitted on Pix) have almost certainly made it more difficult to reach some potential users. In the absence of adequate merchant-transaction fees, the only way to monetize mobile-payments systems is through advertising or by selling other products. Operators of mobile-payments systems might encourage some participants to use their platform by offering discounts on certain products and services. For example, in India, Google offered cashback incentives for use on apps within its own (Android) ecosystem.[92] While there is nothing wrong with this, it is less likely to attract marginal users who are not significant consumers of Android’s app-based services than would simply offering unencumbered cashback rewards.

It is also worth noting that limiting MDR fees on UPI and interchange fees on Pix does not necessarily grant merchants a free ride. They will often have to pay hefty per-transaction fees to participate in online markets, with those fees paid to the parent company (e.g., Google) rather than the payment-service provider (in this example, Google Pay). Moreover, since mobile-payments-system operators have incentives to keep customers within their ecosystem through internal rewards programs, brick-and-mortar merchants may find that they lose customers.

At base, these policy failures stem from the fact that payments markets do not cease being two-sided simply because governments choose to intervene in them. The operators of digital mobile-payments systems are highly aware of this fact. Consider this note from PayPal:

PayPal began with a vision of an online world where consumers could safely, securely, and easily use digital payments to buy what they need and want from businesses. Over the past 24 years, our two-sided network has grown significantly, seamlessly connecting shoppers and retailers across the globe.

Our platform now connects businesses, who look to us to help them navigate the digital economy, with shoppers, individuals who increasingly value flexibility and ease in where and how they shop and manage their daily financial lives. [93]

Meanwhile, PicPay notes:

We now have a more comprehensive portfolio of products and services, further enhancing our value proposition beyond the digital wallet and day to day payments services, becoming a much broader two-sided financial ecosystem.[94]

Interventions in two-sided markets change the participants’ incentives. For example, when MDR or IF rates are subject to price controls, system participants look for other ways to monetize their products (or exit). Fintech companies that operate mobile digital payments have found ways to monetize consumers through advertising and/or online marketplaces. But for banks that don’t also operate mobile digital-payments systems, participating in payment apps for which they cannot charge fees on most transactions, such as UPI and Pix, is often both unavoidable and unfortunate. It is unavoidable because accountholders now expect to be able to use these apps for payments. It is unfortunate because it adds cost, while bringing in minimal IF revenue at best. Banks therefore look for other ways to monetize those accounts, which can include increasing account-maintenance fees or hiking borrower interest rates and/or other fees. Such actions typically harm the poorest consumers, who are least able to afford additional account fees.

Following enactment in the United States of the so-called “Durbin amendment,” which imposed price controls on debit-card IFs, covered banks raised account fees and increased the minimum balance required to maintain an account. In response, it appears that many lower-income consumers closed their bank accounts.[95] In other words, in this and likely other cases, price controls on payment-network fees appear to have had negative effects on financial inclusion.

Furthermore, it is unclear whether the systems established by Brazil and India are sustainable in their current form. One problem is that both rely on government subsidies. In the case of UPI, these come from the central government and must be financed by taxpayers. In the case of Pix, they come from the central bank (BCB), and are thus presumably paid, in part, out of fees from the banks and other organizations that are licensed by BCB. Because such subsidies are not directly related to the system’s performance, the system operator’s incentives are not necessarily aligned with users’ incentives (banks, fintechs, consumers). This may result in underinvestment in security (as arguably has been the case with Pix) and in other necessary ongoing upgrades.

Finally, potential conflicts of interest arise when a payment system’s operator is also the regulator of other payment systems, especially when there are no apparent mechanisms to address this conflict. For example, BCB both operates Pix and regulates payment-service providers, but does not appear to even acknowledge that this may be a conflict.[96] Such concerns were evident Brazil when WhatsApp proposed to introduce a payment app in the country, but the BCB prevented it from doing so until after it had launched Pix.[97]

Looking at changes in the proportion of adults with an account at a financial institution from 2011 to 2021 (Figure 3), the clear winners out of the four countries studied here are China and Kenya. While India and Brazil have also seen significant improvements, the growth has been less dramatic. In part, this is likely because both Brazil and India started from a higher base, but policy also likely played a role. It is probably too soon to see an effect from Pix in the data (Brazil only launch that system at the end of 2020) but UPI and demonetization appear to have had a detrimental effect on financial inclusion in India, with the proportion of adults having an account at a financial institution falling slightly after 2016

FIGURE 3: Share of Adults with an Account at a Financial Institution (2011 to 2021)

SOURCE: World Bank

Governments should not distort the market in these ways. Quite the opposite: they should be as neutral as possible. And they should limit themselves to the production of genuine public goods, such as courts and identity registers. Doing so will enable participation, competition, and innovation, which will drive financial inclusion.

Appendix: Trust and Counterparty Risk in Online-Payment Systems

As noted in Section I, entrepreneurs have developed several important technologies to improve trust and otherwise address counterparty risk in the online world, including user ratings, security protocols and fraud-detection systems, payment gateways, and online escrow systems. The following subsections discuss these technologies.

A. User Ratings as a Means to Enhance Trust

In February 1996, eBay founder Pierre Omidyar introduced the “feedback forum” on the site in order to encourage buyers and sellers to provide constructive feedback to one another. Accompanying the forum’s launch, he included the following note:

Most people are honest. And they mean well. Some people go out of their way to make things right. I’ve heard great stories about the honesty of people here. But some people are dishonest. Or deceptive. This is true here, in the newsgroups, in the classifieds, and right next door. It’s a fact of life. But here, those people can’t hide. We’ll drive them away. Protect others from them. This grand hope depends on your active participation. Become a registered user. Use our feedback forum. Give praise where it is due; make complaints where appropriate. [98]

The forum became extremely popular and Omidyar credits it with much of eBay’s subsequent success, because it enabled users to generate trust with one another.[99]

Such ratings systems have now become a common feature of e-commerce sites; indeed, they are a more-or-less ubiquitous and expected feature of larger sites. Moreover, the granularity of such ratings arguably makes them superior to other forms of regulation.[100]

B. SSL and Credit-Card-Security Protocols

From the perspective of merchants, one of the advantages of accepting credit-card payments is the credible commitment made by card issuers (embedded in the agreement those issuers make with the networks) that they will pay the merchant (via the acquiring bank, in the case of bank-issued cards) as long as the merchant has undertaken the necessary authentication.

Early internet-based transactions, however, posed some novel problems for card networks. In particular, there were security concerns on both sides of the market: issuers and cardholders were concerned about the risk that card details would be stolen and used fraudulently, while merchants and acquirers were concerned that they would be presented with stolen card details that, if used, would result in liability through chargebacks. For online payments to work required the development of solutions that would enhance security and trust on both sides.

Part of the solution came with the development in 1994 of the “secure sockets layer” (SSL) by Marc Andreessen and his team at Mosaic/Netscape.[101] SSL allowed for encrypted information to be sent between a web browser and the host server, preventing that information from being stolen. It was the precursor to modern online browsers, which use transport layer security (TLS) to transmit encrypted information. The incorporation of a padlock sign in a browser’s uniform resource locator (URL) helped to reassure users that their connection was secure, and information could be transmitted without fear of theft.

Another part of the solution came with card networks’ introduction of additional security protocols. Initially, individual payment networks developed their own standards, but these were superseded in 2004 with the development and implementation of an initiative known as the Payment Card Industry Data Security Standard (PCI DSS), which since 2006 has been an independent entity that includes all the major card networks.[102]

C. Credit-Card Guarantees and Chargebacks

Another important mechanism by which credit cards increase trust is card issuers’ commitment (as part of their agreement with networks, in the case of four-party cards) to guarantee payment to merchants on the condition that the merchants themselves meet certain criteria, including undertaking authentication checks. Meanwhile, card issuers may initiate a chargeback against a merchant on behalf of a cardholder in cases of fraud or theft (e.g., the cardholder’s information was stolen) or if the merchant failed to meet a cardholder’s legitimate expectations (e.g., it failed to provide goods or services in a timely fashion; the goods were damaged, defective, or missing parts; or the cardholder was charged an incorrect amount).[103]

D. Online-Payment Gateways as Trusted Intermediaries

Another important piece of the two-sided-market puzzle for online markets was solved in 1999 with the launch of PayPal. Originally conceived as a payment system in its own right, PayPal’s most important function (at least, until recently) has been as a payment gateway—a trusted intermediary that processes card payments.[104] PayPal is itself also a two-sided market, with merchants on one side and consumers on the other.[105] But it’s more complicated: in 2009, it partially opened up its APIs to third parties, thereby enabling others to build interoperable payments systems that would permit true peer-to-peer payments.[106]

E. Escrow and ‘Holds’

Parties to transactions frequently use trusted intermediaries to hold funds until specific conditions have been met, as documented in an escrow agreement. In the United States, escrow had traditionally been used mainly for transactions involving high-value idiosyncratic items, such as real-estate properties and paintings. This also appears to be the case thus far with online escrow services, such as those offered by escrow.com. The likely explanation is that escrow adds complexity and delays to a transaction; where financial remedies are adequate in the event of nonperformance, it is not an efficient solution.

Another reason escrow has not been used more widely for online purchases is that payment networks’ dual-message systems perform a similar function at much lower cost, enabling payments to be put on hold by the merchant until they are satisfied that the client will not issue a chargeback.[107] Some online payment gateways have similar “hold” systems that are very similar to escrow.[108]

But escrow has emerged as an important means to facilitate online commerce in China, where simple financial remedies facilitated by credit-card companies were not available to the majority of consumers. Within that context, online marketplace Alibaba established Alipay in 2003, which was initially an escrow-based system in which a purchaser’s funds were held by Alibaba until the purchase was delivered. As explored in Section 3, this proved to be a phenomenally successful strategy.

[1] Todd J. Zywicki, The Economics of Payment Card Interchange Fees and the Limits of Regulation, Int’l. Ctr. L. & Econ. (Jun. 2, 2010), available at https://laweconcenter.org/images/articles/zywicki_interchange.pdf.

[2] Marc Rysman, The Economics of Two-Sided Markets, 23  J. ECON. PERSP. 125, 130 (2009).

[3] See, e.g., Philip Moore, The State of the American Mall: Competitive, Attractive and Here To Stay, CORESIGHT RESEARCH (Jun. 27, 2023), https://coresight.com/research/the-state-of-the-american-mall-competitive-attractive-and-here-to-stay (noting that a new mall in Atlanta features a Nobu hotel and restaurant).

[4] Gunnar Camner, Emil Sjöblom, & Caroline Pulver, What Makes a Successful Mobile Money Implementation? Learnings from M-PESA in Kenya and Tanzania, GSMA (2009), available at https://www.gsma.com/mobilefordevelopment/wp-content/uploads/2012/06/mpesa_case_study9983.pdf (noting that airtime sharing was not observed in Kenya).

[5] See TONNY K. OMWANSA & NICHOLAS P. SULLIVAN, MONEY REAL QUICK: THE STORY OF M-PESA (Guardian Books, 2012); See also Enabling Mobile Money Transfer: The Central Bank of Kenya’s Treatment of M-Pesa, Alliance for Financial Inclusion (2010), available at afi_casestudy_mpesa_en.pdf.

[6] Banks Find Suite: Find Annual Historical Bank Data, FEDERAL DEPOSIT INSURANCE CORPORATION, https://banks.data.fdic.gov/explore/historical?displayFields=STNAME%2CTOTAL%2CBRANCHES%2CNew_Char&selectedEndDate=2022&selectedReport=CBS&selectedStartDate=1934&selectedStates=0&sortField=YEAR&sortOrder=desc (last visited Jun. 19, 2024); Geographical Outreach: Number of Automated Teller Machines (ATMs), Country Wide for United States, FED. RES. BANK ST. LOUIS (2016), https://fred.stlouisfed.org/series/USAFCACNUM (retrieved from FRED).

[7] Financial Access in Kenya: Results of the 2006 National Survey, Nairobi, Kenya: The Steadman Group Res. Div. (2007).

[8] Id.

[9] Omwansa & Sullivan, supra note 6.

[10] Id.

[11] Id.

[12] Experience M-PESA, Safaricom, https://www.safaricom.co.ke/personal/m-pesa/getting-started/experience-m-pesa (last visited Jun. 13, 2024); Driven by Purpose: 15 Years of M?Pesa’s Evolution, McKinsey & Co. (Jun. 29, 2022), https://www.mckinsey.com/industries/financial-services/our-insights/driven-by-purpose-15-years-of-m-pesas-evolution.

[13] M-Shwari FAQs, SAFARICOM, https://www.safaricom.co.ke/personal/m-pesa/credit-and-savings/m-shwari (last visited June 13, 2024)

[14] Tavneet Suri, Mobile Money, 9 Ann. Rev. Econ. 497, 515 (2017).

[15] Id.

[16] M-Pesa and Your Bank, SAFARICOM, https://www.safaricom.co.ke/personal/m-pesa/do-more-with-m-pesa/m-pesa-and-your-bank (last visited Jun. 13, 2024)

[17] M-Pesa Partners with Visa for Virtual Card Payments in Africa, PYMNTS (Jun. 2, 2022), https://www.pymnts.com/digital-payments/2022/m-pesa-partners-with-visa-for-virtual-card-payments-in-africa; About the M-PESA Globalpay Virtual VISA Card, Safaricon, https://www.safaricom.co.ke/mpesaglobalpay/about (last visited Jun. 13, 2024).

[18] 2021 FinAccess Household Survey, Cent. Bank of Kenya, Kenya Bureau of Nat’l Stat., & FSD Kenya (Dec. 2021), available at https://www.centralbank.go.ke/wp-content/uploads/2022/08/2021-Finaccesss-Survey-Report.pdf.

[19] Suri Tavneet & Jack William, The Long-Run Poverty and Gender Impacts of Mobile Money, 354 SCI. 1288 (2016),  https://www.science.org/doi/10.1126/science.aah5309.

[20] M-Pesa, VODAFONE, https://www.vodafone.com/about-vodafone/what-we-do/consumer-products-and-services/m-pesa (last visited Jun. 13, 2024).

[21] Improving People’s Everyday Financial Experience, Orange (Aug. 28, 2020), https://www.orange.com/en/groupe/nous-connaitre/linnovation-utile-et-source-de-progres-pour-tous/improving-peoples-everyday.

[22] Don Lee, China Charges into Credit Cards, L.A. Times (Oct. 22, 2008), https://www.latimes.com/archives/la-xpm-2008-oct-22-fi-chinacredit22-story.html; The card penetration rate was considerably higher in cities—around 18%, according to a contemporaneous Nielsen survey (rising to 22% the following year). See Survey: China Steps into Credit Card Era, CHINA DAILY (Aug. 9, 2004), https://www.chinadaily.com.cn/english/doc/2004-08/09/content_363531.htm.

[23] The Internet Timeline of China 1986~2003, CHINA INTERNET NETWORK INFO. CTR. (Jun. 28, 2012), https://www.cnnic.com.cn/IDR/hlwfzdsj/201306/t20130628_40563.htm#:~:text=On%20January%2016%2C%202003%2C%20China,to%20the%20Internet%20in%20China.

[24] Infra Section II.A and Appendix

[25] China: A Digital Payments Revolution, Consultative Group to Assist the Poor (Sep. 2019), https://www.cgap.org/research/publication/china-digital-payments-revolution.

[26] What Is Alipay?, CHECKOUT.COM (May 18, 2023), https://www.checkout.com/blog/what-is-alipay.

[27] Yichen Zhu & Sarah Hui Li, A Hangzhou Story: The Development of China’s Mobile Payment Revolution (Lee Kuan Yew Sch. Pub. Pol’y, Nat’l Univ. Sing., 2018), available at a-hangzhou-story.pdf.

[28] Payment Methods, ANTOM DOCS (Apr. 24, 2024), https://global.alipay.com/docs/ac/cashierpay/payment_method.

[29] Barry Elad, Alipay Statistics 2023 – Market Share, Facts and Marketing Trends, ENTERPRISE APPS TODAY (last updated Oct. 10, 2023), https://www.enterpriseappstoday.com/stats/alipay-statistics.html.

[30] WeChat Now Supports Payments Between Users and One-Click Payments, Finance Magnates (Jun. 24, 2014), https://www.financemagnates.com/fintech/payments/wechat-now-supports-payments-between-users-and-one-click-payments (Tencent, the owner of WeChat, established a payment system called TenPay in 2005. That failed to take off, but it remains the official licensed payment provider underpinning Weixin Pay).

[31] Eveline Chao, How WeChat Became China’s App for Everything, Fast Company (Feb. 1, 2017), https://www.fastcompany.com/3065255/china-wechat-tencent-red-envelopes-and-social-money.

[32] Matthew Fulco, The WeChat Economy, From Messaging to Payments and More, CKGSB Knowledge (Aug. 28, 2017), https://english.ckgsb.edu.cn/knowledge/article/the-wechat-economy-from-messaging-to-payments-and-more/#:~:text=Matthew%20Fulco%20Authors,August%2028%2C%202017.

[33] All You Need to Know About WeChat Mini-Programs, GOCLICK CHINA (Jun. 28, 2022), https://www.goclickchina.com/blog/wechat-mini-programs-all-you-need-know.

[34] Shubham Singh, 18 WeChat Statistics — Users & Revenue Data, DEMANDSAGE (May 27, 2024), https://www.demandsage.com/wechat-statistics.

[35]  Multiple Payment Methods, WECHAT PAY, https://pay.weixin.qq.com/wechatpay_guide/intro_method.shtml (last visited Jun. 19, 2024); In-Store Payment, ANTOM DOCS, https://global.alipay.com/docs/instorepayment (last visited Jun. 19, 2024).

[36] Payment Methods in China: How China Became a Mobile-First Nation, Daxue Consulting (Jan. 29, 2024), https://daxueconsulting.com/payment-methods-in-china.

[37] ANTGROUP, https://www.antgroup.com/en/business-development/digital-finance-tab-details/mybank (last visited Jun. 18, 2024).

[38] WEBANK, https://www.webank.com/en/characteristic (last visited Jun. 18, 2024).

[39] Laura He, Visa and Mastercard Can Now be Used on China’s Biggest Payment Apps, CNN (Jul. 21, 2023), https://edition.cnn.com/2023/07/21/tech/china-alipay-wechat-pay-international-credit-cards-intl-hnk/index.html.

[40] INDIA STACK, www.indiastack.org (last visited Jun. 18, 2024)

[41] Vision and Mission, UNIQUE IDENTIFICATION AUTH. INDIA, https://uidai.gov.in/en/about-uidai/unique-identification-authority-of-india/vision-mission.html (last visited Jun. 18, 2024).

[42] Aadhaar Verification API: Innovation in Identity Verification, PERFIOS (Dec. 5, 2024), https://www.perfios.com/post/aadhaar-verification-api-innovation-in-identity-verification; India Population 1950–2024, MACROTRENDS, https://www.macrotrends.net/countries/IND/india/population (last visited June 18, 2024).

[43] Enrolment Dashboard, UNIQUE IDENTIFICATION AUTH. INDIA, https://uidai.gov.in/aadhaar_dashboard/india.php (last visited Jun. 18, 2024).

[44] Unified Payment Interface (UPI), NPCI, https://www.npci.org.in/what-we-do/upi/product-overview (last visited Jun. 18, 2024); UPI Live Members, NPCI, https://www.npci.org.in/what-we-do/upi/live-members (last visited Jun. 18, 2024).

[45] MOBIKWIK, https://www.mobikwik.com/about (last visited Jun. 18, 2024).

[46] Mahesh Sharma, Payments Startup MobiKwik Launches Mobile Wallet As India’s Central Bank Acts To End Country’s Cash Dependence, TechCrunch (Sep. 27, 2013), https://techcrunch.com/2013/09/27/payments-startup-mobikwik-launches-mobile-wallet-as-indias-central-bank-acts-to-end-countrys-cash-dependence.

[47] Jon Russell, Indian Payments Startup MobiKwik Nabs $25M From Tree Line, Cisco, AmEx and Sequoia, TechCrunch (Apr. 7, 2015), https://techcrunch.com/2015/04/07/mobikwik-series-b.

[48] MobiKwik Continues Profitable Streak for Second Quarter in a Row, ECONOMIC TIMES (Oct. 5, 2023), https://economictimes.indiatimes.com/tech/technology/mobikwik-continues-profitable-streak-for-second-quarter-in-a-row/articleshow/104183594.cms?from=mdr.

[49] MobiKwik Consolidated Financial Statement, MOBIKWIK (2023), available at https://documents.mobikwik.com/files/investor-relations/statements/zaakpay/zaakpay-financials-sept2023.pdf; financial statements of other subsidiaries available at https://www.mobikwik.com/ir/subsidiary-financials; RBI notice of “in-principle authorization” available at https://www.rbi.org.in/Scripts/bs_viewcontent.aspx?Id=4236; Report on the Audit of Special Purpose Interim Financial Statements, TATTVAM & CO. (Dec. 31, 2023), available at https://documents.mobikwik.com/files/investor-relations/statements/zaakpay/zaakpay-financials-sept2023.pdf; Subsidiary Financials, MOBIKWIK,  https://www.mobikwik.com/ir/subsidiary-financials (last visited Jun. 19, 2024); Status of Applications Received from Online Payment Aggregators (PAs) under Payment and Settlement Systems Act, 2007, RES. BANK OF INDIA, https://www.rbi.org.in/Scripts/bs_viewcontent.aspx?Id=4236 (last updated Jun. 16, 2024).

[50] Id., Res. Bank of India.

[51] Press Release, Paytm’s Earning’s Release for Quarter and Year Ending March 2024, Paytm (May 22, 2024), available at https://paytm.com/document/ir/financial-results/Paytm_Earnings-Release_INR_Q4_FY24.pdf.

[52] Paytm’s Pricing, Paytm, https://business.paytm.com/pricing (last visited Jun. 19, 2024).

[53] Paytm Reaches 100 Million Users, Business World (Aug. 11, 2015), https://www.businessworld.in/article/Paytm-Reaches-100-Million-Users-/11-08-2015-84698.

[54] Patrick Jenkins, Paytm and Transmit Security Win FT Fintech Awards, FINANCIAL TIMES (Nov. 16, 2016), https://www.ft.com/content/1a92d762-ac1b-11e6-ba7d-76378e4fef24.

[55] Justin Rowlatt, Why India Wiped Out 86% Of Its Cash Overnight, BBC NEWS (Nov. 14, 2016), https://www.bbc.co.uk/news/world-asia-india-37974423.

[56] Paytm Surpasses 100 Million Monthly Transacting Users for the First Time in Q3 FY24, Livemint (Jan. 22, 2024), https://www.livemint.com/companies/news/paytm-surpasses-100-million-monthly-transacting-users-for-the-first-time-in-q3-fy24-11705932856486.html.

[57] PHONEPE, https://www.phonepe.com/about-us (last visited Jun. 19, 2024); Charlie Graham, PhonePe, CONTRARY RESEARCH (last updated Dec. 1, 2023), https://research.contrary.com/reports/phonepe.

[58] Press Release, Walmart to Invest in Flipkart Group, India’s Innovative eCommerce Company, Walmart (May 9, 2018), https://corporate.walmart.com/news/2018/05/09/walmart-to-invest-in-flipkart-group-indias-innovative-ecommerce-company.

[59] Sri Deepti, Flipkart Completes Separation from PhonePe, TECH IN ASIA (Dec. 23, 2022), https://www.techinasia.com/flipkart-completes-separation-phonepe.

[60] Alnoor Peermohamed, Flipkart Grows User Base to 100 million, Business Standard (Sep. 22, 2016), https://www.business-standard.com/article/companies/flipkart-grows-user-base-to-100-million-116092100216_1.html; FLIPKART CATAPULT, https://brands.flipkart.com/catapult-about (last visited Jun. 19, 2024).

[61] Graham, supra note 58.

[62] Ingrid Lunden, Google Debuts Tez, A Mobile Payments App for India that Uses Audio QR to Transfer Money, TECHCRUNCH (Sep. 17, 2017), https://techcrunch.com/2017/09/17/google-debuts-tez-a-mobile-wallet-and-payments-app-for-india.

[63] Google Pay API for India & Google Pay & Wallet Console, GOOGLE, https://support.google.com/google-pay-and-wallet-console/answer/10945206?hl=en#:~:text=Google%20Pay%20API%20for%20India%20operates%20on%20a%20unique%20India,used%20as%20your%20UPI%20ID (last visited Jun. 19, 2024). The version of Google Pay implemented in India is thus quite different to the version implemented in the United States and other markets.

[64] Manish Singh, Google Paves Way to Monetize Pay Users’ Data in India, TECHCRUNCH (Mar. 11, 2021), https://techcrunch.com/2021/03/11/google-pay-paves-way-to-tap-pay-users-data-in-india.

[65] Certificates of Authorisation Issued by the Reserve Bank of India Under the Payment and Settlement Systems Act, 2007 for Setting Up and Operating Payment System in India, RES. BANK OF INDIA, https://rbi.org.in/Scripts/PublicationsView.aspx?id=12043 (last updated Jun. 18, 2024).

[66] Manish Singh, WhatsApp Permitted to Extend Payments Service to 100 Million Users in India, TECHCRUNCH, (Apr. 13, 2022), https://techcrunch.com/2022/04/13/whatsapp-permitted-to-extend-payments-service-to-100-million-users-in-india.

[67] Id.

[68] Id.

[69] Press Release, Vodafone India Launches M-Pesa Pay for Merchants and Retailers, Vodafone (Jan. 5, 2017), https://www.vodafone.com/news/inclusion/m-pesa-pay-india; Aman Rawat, Vodafone Winds Up M-Pesa in India After Huge Losses, INC 42 (Jan. 22, 2020) https://inc42.com/buzz/vodafone-winds-up-m-pesa-in-india-after-huge-losses.

[70] Jackson Lott & Mona Sinha, M-Pesa’s Failure in India: Why Couldn’t Vodafone Replicate its Kenyan Success? An International Marketing Case Study, 6 KENNESAW J. UNDERGRADUATE RES. 1, 12-13 (2019), available at https://digitalcommons.kennesaw.edu/cgi/viewcontent.cgi?article=1160&context=kjur (“In Kenya, Safaricom had invested considerable effort and money over the years to build a critical mass of M-Pesa agents. These were essentially small shop owners across the country, including in villages, who could also serve as M-Pesa agents, available to sell mobile airtime as well as enable M-Pesa wallet top-ups and withdrawals, as needed. However, when Vodafone went to villages in India where most migrant workers came from, they did not find much economic activity. Unable to find small shop keepers who could serve as agents, they could not create as good a network of last-mile agents crucial to the service. Moreover, unlike Kenya, poor, unbanked, or underbanked consumers struggled to adopt self-service technologies and needed assistance. Creating awareness and driving behavior change amongst this segment of the population required tremendous resources in terms of time, money, and human capital that Vodafone would have to divert from its core business in India, i.e., cell phone service. Meanwhile, a slew of mobile payment upstarts entered the Indian market. Unlike Vodafone that aimed to service the unbanked and underbanked, these new entrants, like Paytm, reached out to middle class and affluent consumers who wanted the convenience and lived in urban and semiurban areas where agents could be easily appointed. Given the income profile of their target consumers, their technologies could be based on linking their app to the customers’ bank accounts. This demographic was markedly different from the unbanked/underbanked consumers that M-Pesa serviced who used feature/basic phones and dealt in cash to top-up or withdraw cash from their M-Pesa wallets. Moreover, Paytm had the backing of large investors like Soft Bank and Alibaba. One key trigger event for spurring adoption of digital payments in India was the demonetization announcement by the Prime Minister of India. However, a shortage of cash at that time due to high-value currencies being declared defunct meant that a cash dependent system like M-Pesa did not benefit from the surge of mobile payment adopters.”)

[71] Zaakpay and PhonePe have “in principle approval,” see supra note 47; Google has full approval, see supra note 63; Paytm has been asked to revise and resubmit its application, see Press Release, Update on PA License: Paytm Payments Services Receives Extension From RBI for Resubmission of Application & Remains Hopeful of Getting Necessary Approvals, Paytm (Mar. 27, 2023), https://paytm.com/blog/investor-relations/update-on-pa-license-paytm-payments-services-receives-extension-from-rbi-for-resubmission-of-application-remains-hopeful-of-getting-necessary-approvals.

[72] Hiral Thanawala, e-Rupi Vouchers Get a Boost from RBI Monetary Policy, MONEY CONTROL (Jun. 8, 2023), https://www.moneycontrol.com/news/business/personal-finance/e-rupi-vouchers-get-a-boost-from-rbi-monetary-policy-10766151.html.

[73] Roll Back Zero Merchant Discount Rate on UPI, RuPay Debit Card Payments, Industry Body Payments Council of India Writes to Finance Ministry, INDIAN EXPRESS (Jan. 23, 2022),  https://indianexpress.com/article/business/banking-and-finance/merchant-discount-rate-rollback-on-upi-rupay-debit-cards-7737229.

[74] Pratik Bhakta, Fintechs Await Government Word on MDR Subsidy Allocation, ECONOMIC TIMES (Feb. 22, 2024), https://economictimes.indiatimes.com/tech/technology/fintechs-await-government-support-for-promoting-digital-payments-for-current-fiscal/articleshow/107891943.cms?from=mdr.

[75] Julian Morris, Todd J. Zywicki, & Geoffrey A. Manne, The Effects of Price Controls on Payment-Card Interchange Fees: A Review and Update, Int’l. Ctr. L. & Econ. (Mar. 3, 2022), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4063914.

 

[76] Oi Paggo, LINKEDIN, https://www.linkedin.com/company/paggo/about (last visited Jun. 19, 2024); Paggo, WIKIPEDIA, https://pt.wikipedia.org/wiki/Paggo (last visited Jun. 19, 2024), (information unverified because the links are broken).

[77] Oi Paggo: A Disruptive Brasileiro Credit Play, STL PARTNERS (Jun. 2008), https://stlpartners.com/insights/oi_paggo_a_disruptive_brasilei.

[78] Umair Bashir, Most Used Mobile Payments by Brand in Brazil 2023, Statista (Feb 13, 2024), https://www.statista.com/forecasts/1226566/most-used-mobile-payments-by-brand-in-brazil#:~:text=We%20asked%20Brazilian%20consumers%20about,in%20our%20Consumer%20Insights%20tool.

[79] Crescimento do PicPay: 34 Milhões Usam o App No Dia a Dia, PICPAY (Dec. 14, 2021), https://blog.picpay.com/picpay-crescimento/?utm_source=iupana&utm_medium=web&utm_campaign=010822EN.

[80] Fabiane Z. Menezes, PicPay, Latin America‘s Largest Digital Wallet, Is Now A Bank, THE BRAZILIAN REP. (Jul. 13, 2022), https://brazilian.report/liveblog/2022/07/13/picpay-digital-wallet-bank; PicPay 1H23 Earnings Release, PICPAY (Aug. 2023), at 11, https://api.mziq.com/mzfilemanager/v2/d/f4269298-97ef-427d-ac29-04f9d6a6054a/765c070b-b1ca-b07b-d699-b6a7eb7be191?origin=1.

[81] PicPay, supra note 79, at 2.

[82] Id., at 4.

[83] PayPal User Agreement, PAYPAL, https://www.paypalobjects.com/marketing/ua/pdf/BR/en/ua-032122.pdf?locale.x=en_BR (last visited Jun. 19, 2024).

[84] Mercado Libre SEC Proxy Statement (2023), at 6, https://api.mziq.com/mzfilemanager/v2/d/098a2d95-0ea8-4ed5-a340-d9ef6a2b0053/bfba3d21-84f6-7263-8abc-f32f54dc122d?origin=1.

[85] Id., at 7.

[86] Our History, PAGBANK, https://investors.pagbank.com/about-us/our-history (last visited Jun. 19, 2024).

[87] Our History, PAGSEGURO, https://international.pagseguro.com/about-us (last visited Jun. 19, 2024).

[88] Google Wallet Help, GOOGLE, https://support.google.com/wallet/answer/12059326?hl=en-GB&co=GENIE.CountryCode%3DBR (last visited Jun. 19, 2024).

[89] The Acceleration of Financial Inclusion During the COVID-19 Pandemic: Bringing Hidden Opportunities to Light, MasterCard (Oct. 12, 2020), available at https://www.mastercard.us/content/dam/public/mastercardcom/na/us/en/banks-and-credit-unions/other/mastercard-financial-inclusion-during-covid-whitepaper-20201012.pdf.

[90] Vinicius Colares, Bolsa Merenda PagBank 2021: CADASTRO, VALOR e como RECEBER o BENEFÍCIO…, PRONATEC (Apr. 11, 2021), https://pronatec.pro.br/bolsa-merenda-pagbank-2021-cadastro.

[91] Eduarda Andrade, CAIXA Tem Libera Transferência do Auxílio e Bolsa Família por PIX e TED, FDR (Aug. 18, 2021), https://fdr.com.br/2021/08/18/caixa-tem-libera-transferencia-do-auxilio-e-bolsa-familia-por-pix-e-ted; PICPAY, https://picpay.com/pix (last visited Jun. 19, 2024); Pix, PAGBANK, https://faq.pagseguro.uol.com.br/pix/408#rmclgBank (last visited Jun. 19, 2024).

[92] Manish Singh, Google‘s New Plan to Push Google Pay in India: Cashback Incentives in Android Apps, TECHCRUNCH (May 16, 2019), https://techcrunch.com/2019/05/16/google-pay-india-android-cashback.

[93] A Peek Inside PayPal’s Two-Sided Network of Consumers and Businesses, PayPal (Oct. 27, 2022), https://newsroom.paypal-corp.com/2022-10-27-A-Peek-Inside-PayPals-Two-Sided-Network-of-Consumers-and-Businesses.

[94] PicPay, supra note 79, at 3.

[95] Vladimir Mukharlyamov & Natasha Sarin, Price Regulation in Two-Sided Markets: Empirical Evidence from Debit Cards, SSRN (Nov. 24, 2022), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3328579.

[96] Julian Morris, Central Banks and Real-Time Payments: Lessons from Brazil’s Pix, Int‘l. Ctr. L. & Econ. (Jun. 1, 2022), available at https://laweconcenter.org/wp-content/uploads/2022/06/Lessons-from-Brazils-Pix.pdf.

[97] Julian Morris, Pixtopia Is Not Real, TRUTH ON THE MARKET (Jun. 2, 2022), https://truthonthemarket.com/2022/06/02/pixtopia-is-not-real.

[98] Pierre Omidyar, Founder’s Letter, eBay (Feb. 26, 1996), https://pages.ebay.co.uk/services/forum/feedback-foundersnote.html.

[99] Pierre Omidyar – On Innovation, The Henry Ford (Apr. 29, 2010), https://www.youtube.com/watch?v=RKVmsifohgM&t=295s.

[100] Julian Morris, Consumer Protection in the 21st Century, Int‘l. Ctr. L. & Econ. (Feb. 24, 2023), https://laweconcenter.org/resources/consumer-protection-in-the-21st-century.

[101] The original version of SSL was easily hacked, and thus was not launched publicly. Version 2.0, which was more secure, was included in releases of Netscape in late 1994. See Huzaifa Sidhpurwala, Evolution of the SSL and TLS protocols, RED HAT (Nov. 16, 2016), https://access.redhat.com/blogs/766093/posts/2758801; Phillip Hallam-Baker, Crypto Standards vs. Engineering Habits – Was: NIST About to Weaken SHA3?, ianG.org (Oct. 4, 2013), https://www.metzdowd.com/pipermail/cryptography/2013-October/018041.html.

[102] PCI SECURITY STANDARDS COUNCIL, https://www.pcisecuritystandards.org (last visited Jun. 19, 2024).

[103] Troy Segal, What Is a Chargeback? Definition, How to Dispute, and Example, Investopedia (Dec. 15, 2023), https://www.investopedia.com/terms/c/chargeback.asp.

[104] There are now many other online-payment gateways, including Worldpay, Stripe, and Square/Block.

[105] The company even talks about itself as a “two-sided network,” see PayPal, supra note 93.

[106] Sebastian Rupley, PayPal’s (Partially) Open Platform to Usher in New Payment Models & Apps, GIGAOM (Nov. 3, 2009),  https://web.archive.org/web/20110102142203/http://gigaom.com/2009/11/03/paypals-partially-open-platform-to-usher-in-new-payment-models-apps.

[107] Credit/Debit Authorisation Holds Explained, GOCARDLESS (last updated Mar. 2022), https://gocardless.com/guides/posts/credit-debit-authorisation-holds-explained.

[108] How to Resolve Payment on Hold or Unavailable Funds, PAYPAL (Oct. 10, 2023),  https://www.paypal.com/uk/brc/article/funds-availability.

Executive Summary

We appreciate the opportunity to respond to this request for information on corporate consolidation through serial acquisitions and roll-up strategies issued by the U.S. Justice Department (DOJ) and the Federal Trade Commission (FTC) (collectively, “the agencies”). We agree that robust competition across markets is critical to consumer welfare and the U.S. economy. The agencies have important roles to play in protecting competition and consumers against anticompetitive conduct—including, specifically, those mergers and acquisitions that have harmed, or are likely to harm, competition and consumers. We do not gainsay the importance of effective and efficient enforcement of the federal antitrust laws—including, but not limited to, the Clayton Act. Moreover, we recognize the substantial contributions that agency staff have made to such enforcement through systematic economic research and other policy studies.

While serial acquisitions and roll-up strategies merit further study, there is no apparent basis—in either the economic literature or the agencies’ enforcement experience—for any general changes to the procedures or substantive standards by which serial acquisitions are scrutinized. We also have concerns about the RFI’s approach and framing. The inquiry appears to presume that serial acquisitions tend to be harmful, without adequately acknowledging or requesting information on the potential procompetitive effects or efficiencies of such acquisitions. In addition, the broad scope of the inquiry, covering numerous industries and transaction types, may not yield sufficiently focused insights to inform policy initiatives going forward.

Our response comprises, at the highest level of generality, one observation and one recommendation. The observation is that, while serial acquisitions may constitute an important domain of merger scrutiny, neither enforcement experience nor the economic literature support any fundamental changes in procedural or substantive antitrust law or regulation for these types of transactions. Competition policy is not, and should not be, static. At the same time, sound policy reform is a difficult and iterative process, and one that requires a firm foundation in both research and enforcement experience, along with attention to established precedent.

Correspondingly, our overarching recommendation is that the agencies build on the substantial body of research regarding mergers and acquisitions that has been conducted over the course of several decades by agency staff and others. More specifically, we recommend that economic and policy staff at the agencies synthesize the existing body of research at their disposal. To be sure, market developments and developments in research methods and available data might suggest new avenues of research, as well as those in need of significant updates. But a serious, critical synthesis of the available literature will help to sharpen the agencies’ sense of new research demands, just as it will provide a basis from which to contemplate new enforcement initiatives.

This research-focused approach is particularly crucial for understanding serial acquisitions and roll-up strategies, given their complexity and likely variation across industries. In pharmaceuticals, for example, serial acquisitions may be a vital mechanism to bring innovations to market, while in other sectors they might raise more significant or more frequent competitive concerns. Only through rigorous, sector-specific research can the agencies develop a nuanced understanding of these strategies’ competitive effects.

Furthermore, we encourage the agencies to:

1. Develop more focused and productive requests for information on critically important issues in serial acquisitions going forward.
2. Recognize that sound policy reform requires a firm foundation in both research and enforcement experience, along with attention to established precedent.
3. Be cautious about drawing general conclusions about whole industries, business models, or methods of acquisition, and more cautious still in condemning them.
4. Consider the tradeoffs inherent in any new reporting requirements or regulatory approaches, carefully weighing potential benefits against the burdens imposed on businesses and agency resources.

The following sections elaborate on these points and provide specific responses to the RFI questions. Our goal is to encourage a more balanced, evidence-based approach to addressing potential competitive concerns arising from serial acquisitions and roll-up strategies, grounded in rigorous research and analysis.

I. Framing and Conceptual Problems

The role of an RFI should be to collect diverse objective data and perspectives that can help the agencies to better understand complex market dynamics within and across industries. Unfortunately, the current RFI appears to deviate from this purpose. The questions appear designed to elicit responses supporting a predetermined negative view of serial acquisitions, rather than to gather balanced information. It thus creates the unfortunate appearance that the agencies are primarily interested in collecting negative perspectives about serial acquisitions, rather than seeking to learn about their potential—and potentially varied—benefits and harms to competition and consumers.

A. The General Framing Suggests an Answer

The framing of this RFI seems, in many ways, conclusory. It does not suggest a neutral inquiry but, rather, raises significant concerns about bias in the agencies’ approach to serial acquisitions.

The RFI’s language often presumes negative outcomes from serial acquisitions. For instance, the document states that “some companies use serial acquisitions of small firms as a business strategy that can harm competition to the detriment of consumers, workers, and innovation in an industry or business sector without detection by the Agencies.”[1] This statement, presented in the background section, sets a tone that appears to prejudge the issue. It suggests that serial acquisitions are inherently harmful, without acknowledging potential procompetitive effects or efficiencies. After all, only 2% of all mergers subject to premerger notification receive second requests; moreover, a second request is not a complaint, much less a final decision that a proposed merger would be unlawful.[2] Do the agencies believe that the proportion of harmful mergers is higher for small acquisitions or multiple acquisitions by a given firm, whether within or across markets? If not, the presumption of harm seems misguided.

Similarly, several of the RFI’s questions are framed in ways that appear to seek confirmation of harm, rather than objective information. Question 2(c), for example, asks whether respondents have “witnessed any actual or attempted coordination or collusion between competitors that you did not notice prior to the serial acquisitions”[3]—a highly unscientific approach, both as a means of sampling and in that it assumes or suggests causality. Question 3 lists nine specific business practices, all of which are framed negatively, such as “[s]elling products or services below cost with the goal or effect of rivals or driving them out of the market.”[4]

Meanwhile, the agencies’ public outreach around the RFI betrays the same presumption of harm, as well as an apparent quest for enforcement targets, rather than better understanding. The subheading on both the FTC’s and DOJ’s press releases announcing the inquiry explicitly asks for information on harmful acquisitions: “Agencies launch public inquiry to identify serial acquisitions, roll-ups that have harmed competition, consumers, workers, and innovation.”[5] Both press releases also include similarly biased quotes from FTC Chair Khan, referring to “stealth consolidation schemes” and highlighting that “[f]irms can use serial acquisitions to roll up markets, consolidate power and undermine fair competition, all while jacking up prices and degrading quality.”[6] Of course, serial acquisitions, like any acquisition, can have these deleterious effects. But the plain implication in the agencies’ communications is that serial acquisitions (perhaps unlike other acquisitions) inherently do have these harmful characteristics.

As noted by former antitrust enforcers in their response to this RFI, “[a]lthough several questions take a neutral approach, many of them solicit negative information about acquisitions, and not one asks about any benefits.”[7] This imbalance is particularly striking, given the inquiry’s breadth and the fact that, as the former enforcers likewise highlight, the agencies have stated that “the vast majority of mergers are either procompetitive and enhance consumer welfare or are competitively benign.”[8]

This apparent bias in the RFI is disconcertingly reinforced by the agencies’ recent statements to the Organisation for Economic Co-operation and Development (OECD).[9] In those comments, the FTC stated that “[s]erial acquisition strategies which aim to achieve—or actually achieve—high market share, exclusion of competitors, suppression of wages, reduction in innovation, or pricing power may violate Section 2.”[10] The commission also noted that “[t]he Agencies are focused on enforcement against serial acquisition strategies.”[11]

If “may violate,” in this context, suggests mere possibility, it is uninformative: Any given acquisition may or may not run afoul of Section 2, as there is no recognized class of mergers that is per se lawful. But “may violate” here would appear to be more naturally read as a suggestion of suspicion, or a predisposition on the part of the federal enforcement agencies toward viewing serial acquisitions as inherently or likely problematic.

These statements, along with examples of enforcement actions, suggest that the FTC has already formed conclusions about the competitive effects of serial acquisitions and determined its enforcement approach. This predetermined stance would contradict the ostensibly open nature of the current RFI, which purports to seek objective information to inform policy, not to vindicate prior policy suspicions.

As enforcement agencies, the FTC and DOJ have a responsibility to be vigilant and duly skeptical when examining potential anticompetitive practices. Their mandate to protect competition and consumers necessitates scrutiny of business practices that could harm market dynamics. This skepticism is not only appropriate, but essential to their role. The RFI’s framing, however, appears to go beyond reasonable skepticism to approach presumption, if not prejudgment. While it is legitimate for the agencies to investigate practices they believe may be harmful, an RFI is a tool for gathering objective information to inform policy and to help establish enforcement priorities.

Careful information gathering, like systematic research, also requires a degree of skepticism about its own priors, including working hypotheses. The current framing lacks such balance or care, ahead of any specific findings or even preliminary analysis of data. For instance, while it is appropriate to ask about potential harms from serial acquisitions, a balanced inquiry would also—at a bare minimum—seek information about potential benefits or efficiency gains associated with or caused by such acquisitions. As noted, the current RFI’s lack of such balance suggests a predisposition toward finding harm, rather than an open-ended exploration of the complex competitive dynamics surrounding serial acquisitions.

This approach not only risks biasing the information received, but it may also undermine the credibility of the agencies’ ultimate enforcement efforts. Effective enforcement requires not just identifying anticompetitive practices but also understanding the full context in which business strategies operate. By appearing to presume harm from serial acquisitions, the agencies may miss important nuances that could inform more targeted and effective enforcement actions.

These concerns are especially troubling when viewed alongside, for example, the 2024 interagency RFI regarding consolidation in health-care markets[12] and the 2023 proposed changes to Hart-Scott-Rodino Act (HSR) rules.[13] As we pointed out in our comments to those inquiries, the workshop that accompanied the healthcare RFI “seems to have been conclusory by design.”[14] Similarly, the HSR rules would dramatically increase the burden of premerger notification, requiring extensive new information on labor markets, nonhorizontal relationships, and other areas—again, without clear evidence that such information is necessary or appropriate for effective merger screening.[15] Both of these initiatives, like the current RFI, suggest presumptions that mergers are unlawful far more often than they have proven to be. Both also suggest substantially increasing scrutiny and reporting burdens on merging parties without clear evidence that such measures are likely to identify anticompetitive harms.

The leading nature of this RFI’s questions, especially when considered alongside other recent agency actions, raises concerns about the objectivity of the agencies’ approach to serial acquisitions. We urge the agencies to adopt a more balanced, evidence-based approach that considers both the potential harms and benefits of these business strategies. Such an approach would be more consistent with established antitrust principles and more likely to yield insights that truly enhance the agencies’ ability to protect competition and consumer welfare.

B. Conceptual Issues

The RFI also suffers from conceptual issues—mostly concerning the question of what constitutes a “serial acquisition”—that would make subsequent enforcement based on these issues inherently arbitrary.

The RFI’s definition of serial acquisitions as “the same firm consolidating a fragmented market through a number of acquisitions, typically of many relatively small companies” lacks precision and raises several questions.[16] What constitutes a “number” of acquisitions? Over what period should these acquisitions be considered “serial”? How small must the acquired companies be to fit this definition?

Precise numerical answers to these questions are not required at this stage of the agencies’ inquiry, nor are fixed thresholds. Still, without clearer parameters—without some way to cabin the agencies’ inquiry—there is a risk of disparate and unfocused responses to the RFI. Further, a signal of overly broad agency scrutiny could chill legitimate business activity. For instance, two acquisitions over 14 years might be considered “serial” under a broad definition, yet such a pattern hardly suggests a systematic strategy to consolidate a market. Indeed, even within a given industry or sector, a considerably larger number of acquisitions and a narrower time frame may commonly be procompetitive or benign if they range across product or geographic markets.

Moreover, the difficulty in precisely defining “serial acquisitions” raises concerns about potential arbitrary enforcement. While wide-ranging requests for diverse inputs may be suitable for preliminary information gathering, clear, objective criteria are essential for rigorous study and more focused investigations and enforcement. Such criteria are also likely important to develop guidance for businesses to understand their obligations, and for consistent application of the law.

II. Do Serial Acquisitions Present Unique Competition Issues?

Many of the activities described as “serial acquisitions” are indistinguishable from normal patterns of business growth and consolidation in maturing industries. As a general matter, it is not clear why a company growing through multiple small acquisitions should be viewed differently than one growing “organically” or through fewer, larger acquisitions. This raises important questions about the agencies’ underlying theory of harm. If the concern is market concentration, this can manifest through various means, not just serial acquisitions. If the concern is the specific process of multiple small acquisitions, it’s unclear why this would be inherently more problematic than other forms of growth.

Recent research by Jonathan Cohn, Edith Hotchkiss, and Erin Tower sheds light on the motivations behind roll-up strategies in private-equity buyouts of private firms.[17] Their study suggests that these strategies are often driven by two primary motives: unlocking growth potential in capital-constrained firms and improving operational performance in underperforming firms. They find that acquired firms often experience significant increases in sales growth and moderate improvements in profitability post-acquisition, supporting the view that these strategies can create value through both growth and operational improvements. These findings suggest that properly executed roll-up strategies can serve legitimate business purposes beyond mere market consolidation.

Given the legitimate business reasons for acquisitions (serial or not), we are aware of no theoretical or empirical grounds on which to suppose that multiple acquisitions are typically anticompetitive. At the same time, there is no reason to suppose that the organic growth of a firm precludes anticompetitive conduct. The competitive effects of growth, whether through acquisition or internal expansion, depend on a variety of factors—including market structure, barriers to entry, and the specific capabilities and assets being acquired or developed. For example, in some cases, serial acquisitions might allow a firm to quickly assemble complementary assets and capabilities, leading to increased innovation and more robust competition. In other instances, organic growth might allow a firm to build market power in ways that are difficult for competitors to challenge.

By identifying such a broad range of serial acquisitions for special scrutiny, the agencies risk deterring transactions that may be procompetitive or competitively neutral. This is particularly concerning given that, as noted in our HSR comments, “the vast majority of mergers are either procompetitive and enhance consumer welfare or are competitively benign.”[18] The costs of deterring beneficial transactions could be substantial, including reduced incentives for startup formation and innovation, decreased liquidity in capital markets, lost efficiencies from beneficial consolidation, and reduced competitive pressure on incumbent firms.

A. Distinctions Between Organic Growth and Acquisitions Are Baseless

The agencies’ RFI appears animated by a fundamental presumption that it is better for firms to “build” new capabilities than to “buy” them.

While eschewing the more contentious form of the statement in their draft guidelines,[19] the agencies assert in the 2023 Merger Guidelines that “[i]n general, expansion into a concentrated market via internal growth rather than via acquisition benefits competition.”[20] The FTC’s challenge of the Meta/Within transaction centers on the agency’s claim that Meta’s decision to buy its way into a new market obviated its role as a potential entrant through organic growth, thereby reducing competition.[21] As John Newman, then-deputy director of the FTC Bureau of Competition, said in criticizing Meta’s acquisition strategy:

Instead of competing on the merits, Meta is trying to buy its way to the top. Meta already owns a best-selling virtual reality fitness app, and it had the capabilities to compete even more closely with Within’s popular Supernatural app. But Meta chose to buy market position instead of earning it on the merits. This is an illegal acquisition, and we will pursue all appropriate relief.[22]

The FTC’s view has been supported by commentators who contend that, despite their ability to grow organically, many incumbent firms have “opportunities to ‘roll up’ (willing) startups to ‘get there faster,’ ‘buying’ instead of expending effort in rival innovation.”[23] Their conclusion is that “[f]oregoing such effort is never good for consumers and society as a whole.”[24] Indeed, these commentators contend that “the general conclusion of the academic literature is that consumers and society are better off when innovative firms are not permitted to merge.”[25]

But this blunt distinction between “competition on the merits” and growth and entry through acquisition is unwarranted, and not well-supported by the literature. As Jay Ezrielev has observed: “There is, however, no basis for concluding that build-or-buy acquisitions harm innovation. Despite providing ‘build’ incentives, restrictions on buying may lead to less building and less innovation.”[26]

Not only do many firms pursue “hybrid strategies,” as mentioned above, but acquisition by itself can constitute competition on the merits for several reasons. First, acquisitions can lead to significant efficiency gains. These may include cost savings, better resource allocation, and enhanced innovation capabilities, which can ultimately benefit consumers. Second, acquisitions may be the most efficient and effective way to increase competition, by facilitating incumbents’ entry into other markets. As Ezrielev notes: “Buying rather than building frees up resources that the acquirer may use for other innovations. Buying also allows firms to expand into new markets faster and with more certainty.”[27] Third, the potential for takeovers can also drive dynamic competition by encouraging managers to perform sufficiently well to avoid becoming targets. This competitive pressure can also lead to overall improvements in industry performance.[28]

Acquisitions may also serve to reallocate and recombine resources in ways that spur innovation.[29] It may be more efficient to acquire a firm with specialized technologies than to develop those technologies internally. Conversely, acquired firms that excel at developing technology may themselves require managerial and other resources of the acquired firm to develop and commercialize their innovations.[30]

Potential acquisition is also a key exit strategy that makes financing startups more attractive. As Joanna Shepherd similarly notes: “Forcing inefficient builders to build would raise costs and discourage more efficient builders from undertaking building projects as these projects would have fewer potential buyers.”[31] As a result, the existence of a robust merger market offers incentives to create new firms in the first place.

As Jay Ezrielev explains:

The main purpose of build-or-buy acquisitions is to expand into a new market (or to expand a firm’s capacity in a market it already serves) in the most cost-effective way. These acquisitions do not weaken incentives to continue ongoing innovations. On the contrary, many acquirers may be looking to accelerate the development of the target’s innovation. These acquirers may be attracted to the target’s technology precisely because it is promising. Acquirers may be uniquely qualified to recognize the true potential of the target’s innovation because of their related expertise.[32]

By the same token, with respect to vertical acquisitions, there is no basis for the oft-repeated claim that anything a firm can accomplish by merger, it can accomplish by contract, nor that the latter is preferable because it avoids increased consolidation and foreclosure incentives.[33] In reality, just like the decision to buy rather than build, the decision to buy rather than contract is often motivated by important distinctions that make the one preferable to the other.[34]

It is true that there are “limits to vertical integration. In particular, there may be limits to his [sic] entrepreneur’s resources or abilities, such that adding an additional transaction within the firm would generate decreasing returns.”[35] But acquisition can also be distinctly preferable to contracting. It can often be more efficient to organize production within a firm: “[F]irms have a fundamentally different production function from separate, additive market-based production—and [] cooperative team-based production could be much more efficient.”[36]

At the same time, at various stages of the industry life cycle, firms exhibit increasing returns to scale, making acquisition relatively more efficient than contracting, depending on the maturity and other attributes of the relevant industry.[37] Meanwhile, contracting becomes less attractive as asset specificity (i.e., specialized investments or assets that are extremely costly to relocate or redeploy) increases.[38] By the same token, as the costs of contracting increase due to complexity and lack of information (and the risks of asset specificity), the costs of the resulting incomplete contracts increase, again making acquisition relatively more efficient.[39]

Of course, the distinction between vertical and horizontal transactions will also often be blurred, and this, too, provides further reason to be skeptical of claims that perceived horizontal “buy-rather-than-build” acquisitions are harmful. In the case of Meta/Within, for example, while the FTC focused on the acquisition’s purported horizontal effects in the virtual-reality (VR) fitness-app market, it is undeniable that the deal simultaneously allowed Meta to rapidly expand its portfolio of VR-app offerings.[40] In this way, it also signaled a commitment to Meta’s development of the broader ecosystem on which all VR-app markets depend. Meta’s primary interest in the transaction (and the real benefit of the deal for consumers and other app developers) may have been for its broader contribution to the development of Meta’s platform, rather than its narrow contribution to Meta’s place in any particular VR-app market.

In much the same way, the distinction between “organic” growth and growth through acquisition is not always clearcut. This is particularly true at a time when antitrust enforcers around the world are increasingly tempted to treat minority shareholdings and key personnel hires as notifiable mergers.[41] Indeed, the concept of the so-called “acqui-hire” provides a compelling argument against a rigid preference for internal growth over mergers. An “acqui-hire” is an acquisition primarily aimed at acquiring the talent of a company, rather than its products, services, or other assets.

But in this regard, the very aim of an aqui-hire is ultimately to spur organic growth. And, indeed, many firms pursue hybrid strategies, combining internal development with strategic acquisitions. Mergers of this sort can be essential for internal growth, as they bring in critical human resources that drive innovation and competitiveness. The agencies should be cautious about drawing sharp distinctions between these growth strategies without robust empirical evidence of their differential competitive effects.

B. Issues Particular to Small, Serial Acquisitions

Certainly, it is possible for a series or sequence of acquisitions to prove anticompetitive, whether those acquisitions fall under or above HSR reporting thresholds. Indeed, even a single acquisition may prove anticompetitive. At the same time, the costs of increased scrutiny must be weighed against the likely benefits. As noted in comments we previously submitted to the FTC,[42] the 2023 proposed changes to the premerger notification rules could lead to between $350 million and $2.23 billion in additional annual compliance costs,[43] with other additional costs imposed on agency staff. Extending similar requirements to smaller transactions would necessarily add to those costs, potentially imposing disproportionately high burdens on relatively small filers. Moreover, reviewing a larger number of small transactions would significantly increase the agencies’ workload, potentially diverting resources from more critical enforcement priorities.

The potential for errors in judgment is particularly high when dealing with small acquisitions and evolving markets. False positives could stifle innovation and efficiency, while false negatives could allow anticompetitive harm. Given that most mergers are procompetitive or benign, an approach that scrutinizes all serial acquisitions risks an unacceptably high false-positive rate.

While we acknowledge that a pattern of acquisitions could potentially raise competitive concerns in specific circumstances, we are skeptical that “serial acquisitions” represent a distinct phenomenon that requires special treatment. We urge the agencies to focus on case-specific analysis rather than creating new, potentially overbroad categories of acquisitions for heightened scrutiny. In the alternative, the agencies should use the information informally gathered in response to the RFI—together with further economic research and enforcement experience—to sharpen our understanding of serial acquisitions and the circumstances under which greater scrutiny of such transactions may be cost-justified. Any changes to the current approach should be based on robust empirical evidence of harm and a careful consideration of costs and benefits.

Private-equity investment can bring substantial benefits. Even the DOJ has previously recognized the potential advantages of private-equity buyers in certain merger-remedy scenarios, though this view has since shifted.[44] Research has shown that private-equity investments can lead to improved operational efficiency and total factor productivity, increased innovation, and enhanced competitiveness through entry and exit in the industry.[45]

For example, Steven Davis and his co-authors find that private-equity buyouts lead to a 2.1 log point increase in total factor productivity at target manufacturing firms over two years post-buyout, compared to controls.[46] The authors also find that buyouts catalyze creative destruction, increasing job creation and destruction rates by 14 percentage points over two years.[47] Relatedly, contrary to concerns about short-term benefits but long-term harms, Lerner et al. found that leveraged buyouts do not sacrifice long-term investments in innovation.[48]

It’s important to note, however, that the effects of private-equity investment can vary significantly depending on the specific industry, firm, and investment strategy involved.[49] The shift in the agencies’ stance towards private equity, particularly in merger remedies, should be carefully evaluated against this nuanced body of evidence to ensure that potentially beneficial investments are not unduly deterred.

III. Is Industry Variance in Serial Acquisitions Evidence of Harm?

The competitive implications of multiple acquisitions can also vary significantly both across and within industries, underscoring the importance of context-specific analysis, rather than broad generalizations or simple numerical thresholds. Just as we cannot look at concentration measures across industries as a uniform indicator of competitive harm, we similarly cannot view the number of acquisitions in isolation as a reliable signal of market harm.

Different industries have distinct characteristics that shape their competitive dynamics, innovation cycles, and efficient scales of operation.[50] These factors critically influence the role and effect of acquisitions within each sector. For instance, in pharmaceuticals, serial acquisitions may be a vital mechanism to bring innovations to market, especially given regulatory hurdles and the substantial costs of adequate clinical trials above and beyond those of primary pre-clinical research.[51] Small pharmaceutical or biotech firms, while excelling in early-stage research, frequently lack the resources for large-scale clinical trials, regulatory approval, and commercialization.[52] Acquisitions by larger entities can bridge this gap, facilitating the entry and delivery of new drug therapies to the benefit of both competition and consumers (namely, patients).

Thus, while some pharmaceutical and bio-tech acquisitions might prove anticompetitive, multiple or serial acquisitions could often, or on average, prove pro-competitive, signaling a vibrant innovation ecosystem rather than anti-competitive behavior. Indeed, as Joanna Shepherd has noted: “In industries in which most innovation originates externally…, analyses should be less concerned with mergers’ impacts on internal innovation, and more focused on whether consolidation will increase demand for externally-sourced innovation and, ultimately, increase aggregate drug innovation.”[53] This perspective underscores the importance of understanding the specific dynamics and needs of different sectors in order to accurately assess the implications of serial acquisitions.

Similarly, in technology sectors, acquisitions of small, innovative firms can accelerate the development and deployment of new technologies.[54] Many tech startups are founded with the intent and expectation of being acquired, and the viability of this type of exit strategy in turn creates incentives for investment and innovation. The fast-paced nature of technological change means that the competitive landscape can shift rapidly, and what might appear to be market consolidation through acquisitions might instead be a series of attempts to keep pace with evolving consumer demands and technological possibilities.

In contrast, in more mature industries with stable technologies, a pattern of acquisitions might warrant closer scrutiny. Even here, however, the specific market conditions matter. In fragmented industries like home services or local retail, roll-up strategies can create efficiencies of scale that benefit consumers through lower prices or improved service quality.[55]

Moreover, the relevant geographic market can vary dramatically across industries, further complicating any attempt to apply uniform standards. A series of local acquisitions in a national market might have minimal competitive impact, while similar acquisitions in a local or regional market could be more significant. This variability in geographic-market definition is well-documented in antitrust literature and case law. For instance, in the health-care industry, hospital markets are often defined locally or regionally, as evidenced by the FTC’s lawsuit against U.S. Anesthesia Partners Inc., which focused on the Houston and Dallas markets.[56]

The dynamic nature of many industries also means that market boundaries and competitive threats can change rapidly. Today’s competitors might be tomorrow’s complements, or vice versa. This fluidity makes it particularly challenging to assess the long-term competitive impact of a series of acquisitions based solely on their number or frequency.

IV. Conclusion

In light of these complexities, we urge the agencies to resist the temptation to rely on simple metrics or across-the-board presumptions about serial acquisitions. Instead, a nuanced, industry-specific approach is necessary to accurately assess the competitive implications of multiple acquisitions. This approach should consider such factors as the pace of technological change, the role of innovation, typical firm lifecycles, efficient scale of operations, and the specific competitive dynamics of each industry.

For example, in many smaller markets, independent providers of hospital-based services (such as anesthesiology) may be highly concentrated on any standard for “highly concentrated” markets. Further research might aid the agencies in examining highly concentrated provider markets to develop filters or screens for provider acquisitions below the HSR filing threshold, so that the agencies might identify and investigate those subthreshold filings that are the most likely candidates for investigations and, depending on the results of those investigations, enforcement actions. Efficient matter-selection tools will be critical to that effort, less the agencies commit scarce resources to small, unpromising investigations and impose undue costs on health-care providers.

By recognizing and accounting for these industry-specific factors, the agencies can more accurately identify truly problematic acquisition patterns, while avoiding undue interference in benign or beneficial market evolution. This nuanced approach is crucial to maintain a competitive economy that fosters innovation and efficiency across all sectors.

[1] Request for Information for Public Comment on Corporate Consolidation Through Serial Acquisitions and Roll-Up Strategies, Fed. Trade Comm’n & U.S. Dep’t of Justice, Docket No. FTC-2024-0028 (May 22, 2024), at 1, https://www.regulations.gov/docket/FTC-2024-0028 [hereinafter “RFI”].

[2] Lina Khan & Jonathan Kanter, Hart-Scott-Rodino Annual Report, Fiscal Year 2022, Fed. Trade Comm’n & U.S. Dep’t of Justice (Jan. 4, 2024), at 23, available at https://www.ftc.gov/system/files/ftc_gov/pdf/FY2022HSRReport.pdf.

[3] RFI at 5.

[4] RFI at 6.

[5] Press Release, FTC and DOJ Seek Info on Serial Acquisitions, Roll-Up Strategies Across U.S. Economy, Fed. Trade Comm’n (May. 23, 2024), https://www.ftc.gov/news-events/news/press-releases/2024/05/ftc-doj-seek-info-serial-acquisitions-roll-strategies-across-us-economy; Press Release, Justice Department and Federal Trade Commission Seek Information on Serial Acquisitions, Roll-Up Strategies Across U.S. Economy, U.S. Dep’t of Justice (May 23, 2024).

[6] Id.

[7] Asheesh Agarwal et al., Former Enforcers Comment on Request for Information on Corporate Consolidation Through Serial Acquisitions and Roll-Up Strategies (Jun. 26, 2024), at 2, available at https://www.regulations.gov/comment/FTC-2024-0028-0214.

[8] Id., at 3 (citing Statement of Ass’t Att’y Gen. Christine Varney, Merger Guidelines Workshops, Third Annual Georgetown Law Global Antitrust Enforcement Symposium (Sep. 22, 2009)).

[9] See Serial Acquisitions and Industry Roll-ups—Note by the United States, OECD DAF/COMP/WD(2023)99 (Dec. 6, 2023), available at https://one.oecd.org/document/DAF/COMP/WD(2023)99/en/pdf.

[10] Id., at 6.

[11] Id., at 7.

[12] Request for Information on Consolidation in Health Care Markets, Docket No. ATR-102, Dep’t Justice, Dep’t Health & Human Servs., & Fed. Trade Comm’n (Mar. 5, 2024), https://www.regulations.gov/docket/FTC-2024-0022; Daniel J. Gilman & Geoffrey A. Manne, ICLE Comments Re: Request for Information on Consolidation in Health Care Markets, Int. Ctr. Law Econ. (Jun. 5, 2024), https://laweconcenter.org/resources/icle-comments-re-request-for-information-on-consolidation-in-health-care-markets.

[13] Premerger Notification Rules, 88 Fed. Reg. 42178, (Jun. 29, 2023), (to be codified at 16 C.F.R. Parts 801 and 803); Brian Albrecht, Dirk Auer, Daniel J. Gilman, Gus Hurwitz, & Geoffrey A. Manne, Comments of the International Center for Law & Economics on Proposed Changes to the Premerger Notification Rules, Int. Ctr. Law Econ. (Sep. 27, 2023), https://laweconcenter.org/resources/comments-of-the-international-center-for-law-economics-on-proposed-changes-to-the-premerger-notification-rules.

[14] Gilman & Manne, supra note 12, at 6.

[15] Albrecht et al., supra note 13, at 7-9.

[16] RFI at 1.

[17] Jonathan B. Cohn, Edith S. Hotchkiss, & Erin M. Towery, Sources of Value Creation in Private Equity Buyouts of Private Firms, 26 Rev. Fin. 257 (2022).

[18] Albrecht et al., supra note 13, at 5.

[19] See Draft Merger Guidelines, U.S. Dep’t of Justice & Fed. Trade Comm’n (2023), available at https://www.ftc.gov/system/files/ftc_gov/pdf/p859910draftmergerguidelines2023.pdf (The assertion in the draft guidelines was: “The antitrust laws reflect a preference for internal growth over acquisition.”)

[20] 2023 Merger Guidelines, U.S. Dep’t of Justice & Fed. Trade Comm’n (2023), available at https://www.ftc.gov/system/files/ftc_gov/pdf/2023_merger_guidelines_final_12.18.2023.pdf.

[21] See Press Release, FTC Seeks to Block Virtual Reality Giant Meta’s Acquisition of Popular App Creator Within, Fed. Trade Comm’n (Jul. 27, 2022), https://www.ftc.gov/news-events/news/press-releases/2022/07/ftc-seeks-block-virtual-reality-giant-metas-acquisition-popular-app-creator-within (“Meta is a potential entrant…. But instead of entering, it chose to try buying Supernatural. Meta’s independent entry would increase consumer choice, increase innovation, spur additional competition to attract the best employees, and yield other competitive benefits. Meta’s acquisition of Within, on the other hand, would eliminate the prospect of such entry, dampening future innovation and competitive rivalry.”).

[22] Id.

[23] Cristina Caffarra, Gregory S. Crawford, & Tommaso Valletti, “How Tech Rolls”: Potential Competition and “Reverse” Killer Acquisitions, CPI Antitrust Chron. (May 2020), at 2, available at https://www.competitionpolicyinternational.com/wpcontent/uploads/2020/05/CPI-Caffarra-Crawford-Valletti.pdf.

[24] Id.

[25] Id., at 4.

[26] Jay Ezrielev, Antitrust Policy and Legal Standards for Build-or-Buy Decisions, CPI Columns US & Canada (Mar. 2024), at 2, https://www.pymnts.com/cpi-posts/antitrust-policy-and-legal-standards-for-build-or-buy-decisions.

[27] Id.

[28] Henry G. Manne, Mergers and the Market for Corporate Control, 73 J. Pol. Econ. 110 (1965).

[29] See generally Carl Shapiro, Competition and Innovation: Did Arrow Hit the Bull’s Eye?, in The Rate and Direction of Inventive Activity Revisited (Josh Lerner and Scott Stern, eds., 2012), at 361–404. Shapiro calls this the “Synergies Principle: ‘Combining complementary assets enhances innovation capabilities and thus spurs innovation.’” Id., at 365. As he explains: “The Synergies principle emphasizes that ?rms typically cannot innovate in isolation. The quest for synergies is especially important in industries where value is created by systems that incorporate multiple components, as in the information and communications technology sector. The Synergies principle is directly relevant for competition policy since procompetitive mergers and business practices allow for the more e?cient combination of complementary assets.” Id.

[30] See, e.g., Joanna Shepherd, Consolidation and Innovation in the Pharmaceutical Industry: The Role of Mergers and Acquisitions in the Current Innovation Ecosystem, 21 J. Health Care L. & Pol’y 1 (2018).

[31] Id.

[32] Ezrielev, supra note 26, at 3.

[33] See, e.g., Steven C. Salop, The AT&T/Time Warner Merger: How Judge Leon Garbled Professor Nash, 6 J. Antitrust Enforcement 459, 468 (2018).

[34] See generally Geoffrey A. Manne, Kristian Stout, & Eric Fruits, The Fatal Economic Flaws of the Contemporary Campaign Against Vertical Integration, 68 Kansas L. Rev. 923 (2020).

[35] Id., at 939 (citing Ronald Coase, The Nature of the Firm, 4 Economica 386, 394 (1937)).

[36] Id., at 940.

[37] See George Stigler, The Division of Labor Is Limited by the Extent of the Market, 59 J. Pol. Econ. 185, 188 (1951); Manne, Stout, & Fruits, supra note 32, at 941 (“[I]n Stigler’s model, the degree of vertical integration is a dynamic interaction among the firm’s economies of scale as well as the extent of the firm’s market and the market for its inputs.”).

[38] See Oliver E. Williamson, The Economic Institutions of Capitalism 85-102 (1985); see also Paul L. Joskow, Vertical Integration, 55 Antitrust Bull. 543, 563 (2010).

[39] See Benjamin Klein, Robert G. Crawford, & Armen A. Alchian, Vertical Integration, Appropriable Rents, and the Competitive Contracting Process, 21 J.L. Econ. 297, 306-09 (1978).

[40] FTC v. Meta Platforms Inc., U.S. Dist. LEXIS 29832 (2023);

[41] See, e.g., Ryan Browne, Microsoft’s Hiring of Staff from AI startup Inflection Referred for UK Merger Probe, CNBC (Jul. 16, 2024), https://www.cnbc.com/2024/07/16/microsoft-hiring-of-ai-inflection-staff-referred-for-uk-merger-probe.html; Brandon Vigliarolo, UK Antitrust Cops Thrust Probe into Microsoft, Inflection AI Merger, The Register (Jul. 16, 2024), https://www.theregister.com/2024/07/16/uk_competition_officials_kick_off.

[42] Albrecht et al., supra note 13.

[43] Id., at 7-8.

[44] See Bernard A. Nigro Jr. & Harrisson C. Kummer, Unraveling the Roll-Up: Private Equity’s Misunderstood Investment Strategy, CPI Antitrust Chron. (Jun. 2024). The authors discuss the benefits of private-equity investment, citing the DOJ’s 2020 Merger Remedies manual (now withdrawn), which acknowledged that “in some cases a private equity purchaser may be [the] preferred” remedies buyer. See Merger Remedies Manual 24, Dep’t of Justice Antitrust Division (Sep. 2020, withdrawn Apr. 2022), https://www.justice.gov/atr/page/file/1312416/dl. Nigro & Kummer also reference a recent court decision that rejected DOJ arguments against a private-equity divestiture buyer, finding the buyer “well-positioned to maintain, and perhaps even improve upon” competition. United States v. UnitedHealth Group Inc., 630 F. Supp. 3d. 118, 137 (D.D.C. 2022).

[45] See Shai Bernstein et al., Private Equity and Industry Performance, 63 Mgmt. Sci. 1198 (2017).

[46] Steven J. Davis et al., Private Equity, Jobs, and Productivity, 104 Am. Econ. Rev. 3956 (2014).

[47] Id.

[48] Josh Lerner et al., Private Equity and Long-Run Investment: The Case of Innovation, 66 J. Fin. 445 (2011).

[49] Steven J. Davis, John C. Haltiwanger, Kyle Handley, Ben Lipsius, Josh Lerner, & Javier Miranda, The (Heterogeneous) Economic Effects of Private Equity Buyouts (Nat’l Bureau of Econ. Research, No. 26371, 2019), https://doi.org/10.3386/w26371.

[50] See generally Stigler, supra note 35.

[51] See Shepherd, Consolidation and Innovation in the Pharmaceutical Industry, supra note 30.

[52] See Jeffery M. Drazen et al., Drug-Development Challenges for Small Biopharmaceutical Companies 376 New Engl. J. Med. 469 (2017), (“Small biopharmaceutical companies often encounter important challenges in designing and implementing clinical development programs. In a context in which only approximately 10% of clinical programs result in drugs that achieve regulatory approval, small-company clinical programs may have an even lower rate of success than that of large companies owing to limited internal experience in clinical development and limited infrastructure, which may also affect manufacturing and clinical supply. However, these challenges are largely overshadowed by limited resources and funding, which in turn fuel demand for short timelines owing to the need to demonstrate progress to investors. As such, these companies must focus their resources on small, less-costly development programs for very specific targets and often must spearhead new approaches to testing new products in order to survive.”).

[53] Joanna Shepherd, Understanding Innovation Markets in Antitrust Analysis, Truth on the Market (Mar. 30, 2017), https://truthonthemarket.com/2017/03/30/understanding-innovation-markets-inantitrust-analysis-ag-biotech-symposium.

[54] See discussion of the Meta/Within merger, infra Section III.A.

[55] See Stigler, supra note 35, at 190.

[56] Fed. Trade Comm’n v. U.S. Anesthesia Partners, Inc., et al., No. 2010031, Complaint Filed (S.D. Tex. Sep. 21, 2023).

Abstract

AI has taken the world by storm, and competition law is no exception. Policymakers, academics, and commentators are struggling to make sense of how to apply competition law principles to burgeoning AI markets. The question is spurred by an impending sense that inaction is likely to lead to monopolistic outcomes that will later be impossible to revert. What is feared is that AI will become dominated by a few large technology companies and, more spuriously, that these will be the same companies that already control vast swathes of the so-called digital sphere. In other words: it will make big tech even bigger.

One difficulty with this narrative, however, is that, strictly speaking, there is no such thing as an “AI market”because AI is not a unitary, monolithic technology. In this chapter, we argue that the first step to ensuring that antitrust law stays relevant in the age of AI is developing a principled approach to defining AI relevant markets; one that is legally, economically, and technologically sound. Relevant market definition is central to antitrust law because it is the starting point of most, if not all antitrust law cases.A relevant product market is typically comprised of all those products which consumers view as substitutable and which can therefore be said to compete against each other. By delineating the boundaries of competition between firms, relevant product market definition alerts of the presence of market power and, by extension, of the likelihood of anticompetitive effects.Despite its limitations and despite not being an end in itself,  relevant product market definition is, and likely will remain, the main tool for thinking about the contours of competition between firms for the foreseeable future.

We suggest how this can be done by grasping the internal heterogeneity of AI and by understanding what makes AI similar to HI and HI-powered tasks, thus eschewing simplistic narratives about AI’s supposed ubiquity and uniqueness that are bound to impede antitrust law from discharging its social role, which is to protect competition for the ultimate benefit of consumers.

 

Roughly this time last year, I was writing an International Center for Law & Economics (ICLE) issue brief that considered online age-verification and parental-consent laws from a law & economics perspective. The resulting paper, “A Coasean Analysis of Online Age-Verification and Parental-Consent Regimes,” found that the major U.S. Supreme Court cases on age verification and parental consent implicitly followed a Coasean least-cost-avoider analysis, especially with respect to the First Amendment’s “least-restrictive means” test. I noted there that federal district courts appeared to be applying similar analysis in reviewing state-level online age-verification and parental-consent laws, and finding them to be likely unconstitutional.

Since then, federal courts continue to find state laws that impose age-verification and parental-consent requirements on social-media companies likely violate the First Amendment. Courts (see here, here, and here) have consistently applied some type of heighted First Amendment scrutiny, ultimately concluding that the available technological and practical means of avoiding online harms are less burdensome on speech than the laws in question.

In light of these rulings, then, it is peculiar that federal lawmakers continue to consider laws that would impose age-verification and parental-consent mandates. One such recent proposal from Rep. John James (R-Mich.) would apply both types of mandates to digital app stores. The law would make app stores liable if they fail to verify users’ ages or get verifiable consent from minors’ parents before allowing them to use the app store, download an app, or make any purchases within an app. James argued at a recent markup hearing that his proposal—offered as an amendment to H.R. 7890, the Children and Teens’ Online Privacy Protection Act—would be constitutional, even in light of recent rulings suggesting such mandates at the application layer are not. Despite courts rejecting this exact analogy, he compares the obligation to a convenience store asking for identification before kids can buy alcohol or cigarettes and argues that app stores should similarly be required to “age gate.”

Below, I will consider whether app stores are lower-cost avoiders of online harms to minors when compared either to social-media companies or to minors and their parents. I will then offer a brief analysis of the First Amendment questions at-play.

Read the full piece here.

Everyone in Europe, and across the international competition-law sphere, seems to have their own interpretation these days of former Italian Prime Minister and European Central Bank President Mario Draghi’s recent report “The Future of European Competitiveness” (a.k.a., the “Draghi report”). And, of course, those various interpretations, unsurprisingly, inevitably match the interpreter’s policy preferences.

This is not necessarily the fault of those commenting on the report. As The Economist notes: “Mr. Draghi’s recommendations are so numerous that policymakers will be able to pick and choose from among them.” But despite these somewhat equivocal prescriptions, the report makes at least one thing crystal clear: the EU’s regulatory zeal is not the success story its proponents make it out to be.

This is most evidently true in the area of digital markets, where the EU has enacted several sweeping regulations in recent years: the General Data Protection Regulation (GDPR), the Digital Services Act (DSA), the Digital Markets Act (DMA), and the AI Act. As I explain below, the Draghi report ought to be received by other jurisdictions as an admonition to be cautious about following the European example when it comes to digital markets regulation.

Read the full piece here.

Robust property rights and calibrated regulations have long been key ingredients of the U.S. life-sciences ecosystem, facilitating the development of new drugs that require substantial investments in research, development, and commercialization. Recent legislative proposals, however, threaten to undermine this system under the guise of addressing drug-pricing concerns.

Of particular concern in the current legislative landscape are S. 2780, the Medication Affordability and Patent Integrity Act, and S. 2305, the Biosimilar Red Tape Elimination Act, both of which are currently scheduled to be considered in a Sept. 26 markup session of the U.S. Senate Health, Education, Labor, and Pensions (HELP) Committee. While ostensibly designed to improve drug affordability and bolster market competition, these bills risk disrupting the delicate balance that has made the United States a global leader in pharmaceutical innovation.

Read the full piece here.

The Korea Fair Trade Commission (KFTC), the nation’s competition authority, announced Sept. 9 that it had abandoned plans for comprehensive platform regulation modeled after the European Union’s Digital Markets Act (DMA) or Section 19a of Germany’s Competition Act. The proposed Korean regulation would have involved an ex-ante designation process, alongside stringent prohibitions. The KFTC noted that it would instead pursue amendments to Korea’s competition law, the Monopoly Regulation and Fair Trade Act (MRFTA), to better address platform-related issues within the traditional competition-law framework.

Read the full piece here.

In April, we published a short piece in Notice & Comment on 5 key monopolization cases in the tech sector. In it, we presented the results of an informal poll of economists with expertise in antitrust. The poll asked them to rate the strength of the government’s cases by providing both stand-alone ratings and relative ones. Here, we present the results of a follow-up poll in the wake of Judge Amit Mehta’s decision in the Google search case. Judge Mehta’s decision is that of a federal district court and, hence, a fact on the ground; it is also, in our view, a substantial and largely thoughtful one—which is not to say that we agree with its findings of liability. For those reasons, we thought it worth asking whether the decision had changed any minds about the cases.

The short story is this: plus ça change, plus pretty darn similar. Reported views vary across both respondents and cases but, on average, our respondents do not think that the government has brought very strong cases.

Read the full piece here.

A decisão do juiz Amit Mehta contra o Google no caso antitruste do Departamento de Justiça dos Estados Unidos (DOJ) sobre a sua ferramenta de busca foi recebida com grande alarde por muitos.

Embora acreditemos que a decisão seja errônea, alguns especialistas a consideram uma vitória histórica para o DOJ, especialmente seu líder, Jonathan Kanter, que a considera “o Monte Rushmore dos casos antitruste“. Mas essa perspectiva não compreende a história do antitruste em setores de tecnologia e a importância dessa decisão específica.

Read the full piece here.

The trial of the U.S. Justice Department’s (DOJ) “adtech” antitrust lawsuit against Google kicked off Sept. 9 in U.S. District Court in Alexandria, Virginia. In a nutshell, the DOJ (joined by 17 states) argues that Google illegally monopolized key digital-advertising technologies through a variety of anticompetitive tactics. But the DOJ will find it difficult to prove that Google’s actions harmed competition and consumers in this market. Rather than furthering the public interest, this sort of lawsuit is far more likely to deter innovation in dynamic markets, to the detriment of consumers and the American economy.

Read the full piece here.

The Federal Trade Commission (FTC) issued comments Sept. 11 in support of a proposed U.S. Department of Agriculture (USDA) rule that “seeks to clarify the scope of what constitutes unfair practices under the Packers and Stockyards Act (PSA), which assures fair competition and fair trade practices to protect farmers, ranchers, growers, and consumers.”

In the abstract, “fairness” is a fine thing—or, at least, a fair one. And covered entities ought to obey the law. But why is the commission chiming in to deny the importance of competitive harm in USDA regulations implementing the PSA?

Read the full piece here.

With the 3rd U.S. Circuit Court of Appeals’ recent decision in Anderson v. TikTok, it’s time to revisit the interplay between the First Amendment’s right to editorial discretion, Section 230 immunity, and children’s online safety in the context of algorithms.

As has been noted many times, the use of algorithmic recommendations is ubiquitous online. And the potential harms to children receiving bad recommendations is significant, as well, as the underlying facts of the TikTok case show. But there are also countervailing speech and consumer-welfare concerns that arise if algorithmic recommendations are not protected under the law.

Ironically, insofar as the 3rd Circuit is right about algorithmic recommendations constituting first-party speech (and thus not receiving immunity under Section 230), this could mean that online platforms that use algorithmic recommendations have a strong First Amendment defense against products-liability claims.

Read the full piece here.

The U.S. Justice Department (DOJ) and several states filed suit late last month against the property-management software firm RealPage Inc. for its “unlawful scheme to decrease competition among landlords in apartment pricing and to monopolize the market for commercial revenue management software that landlords use to price apartments.”

While this is not the first case to deal with so-called “algorithmic collusion,” it may nonetheless prove noteworthy to the extent that it signals a new approach that competition enforcers might take to such cases, and to anticompetitive agreements more generally.

Read the full piece here.

Congress intended the Infrastructure Investment and Jobs Act’s (IIJA) ambitious $42.5 billion Broadband Equity, Access, and Deployment (BEAD) program to bridge America’s digital divide by subsidizing infrastructure buildout to areas that are either unserved or underserved by broadband internet. With projects administered by the 50 states, five insular territories, and the District of Columbia, the program was never going to be a “quick fix.”

Read the full piece here.

The U.S. Justice Department (DOJ) and the Federal Trade Commission (FTC) are advancing two major antitrust cases that will have significant implications for the American public.

The DOJ, joined by eight states, announced Aug. 23 that it was suing RealPage Inc. for an “unlawful scheme to decrease competition among landlords in apartment pricing and to monopolize the market for commercial revenue management software that landlords use to price apartments.” This case could have economywide ramifications for businesses’ use of algorithms to set prices.

Read the full piece here.

We all know that smoking kills, and many smokers know that switching to vaping is the best way to get their nicotine fix without the cancer risks of burning tobacco. The UK government is so convinced that vaping is the best off-ramp from smoking that they’ve sent vaping kits to more than a million smokers.

Read the full piece here.

TL;DR

Background: The Federal Communications Commission (FCC) issued rules late last year to prevent “digital discrimination” in broadband access. But the rules currently face legal challenges, and several decisions handed down in the U.S. Supreme Court’s most recent term may affect how these challenges play out.

2021’s Infrastructure Investment and Jobs Act (IIJA) required the FCC to issue rules prohibiting discrimination based on factors like income, race, and ethnicity. The resulting rules are broad, covering broadband deployment, pricing, internet speeds, marketing, and more. The 8th U.S. Circuit Court of Appeals will hear challenges to these rules later this month.

But… The FCC claims that the IIJA gave it expansive power to prevent and eliminate discrimination across all aspects of broadband access, and that its rules are needed to address both intentional discrimination and practices that have discriminatory effects, even if unintentional.

However… Critics contend that the commission has overstepped its authority, that its rules will hinder broadband deployment and adoption, and that they will stifle innovative business models. This puts the rules at odds with the IIJA’s stated goal of fostering greater broadband access and adoption.

KEY TAKEAWAYS

THE LOPER BRIGHT DECISION

The Supreme Court’s decision in Loper Bright Enterprises v. Raimondo overturned the longstanding principle of “Chevron deference,” which previously gave agencies like the FCC significant leeway in interpreting ambiguous statutes. Under the new precedent, courts must use their own independent judgment to determine if an agency has acted within its statutory authority, rather than deferring to the agency’s interpretation.

This could make it harder for the FCC to justify its broad interpretation of the IIJA’s mandate. The FCC’s rules go beyond merely preventing intentional discrimination in broadband deployment, extending to practices that might have unintended discriminatory effects across various aspects of broadband service. 

Without Chevron deference, the court will need to be convinced that this broad interpretation aligns with congressional intent in writing the IIJA.

THE FISCHER DECISION

The Court’s holding in Fischer v. United States emphasized the importance of reading statutory provisions in context. The FCC’s expansive interpretation of its authority under the IIJA may not align with Fischer’s principle that, when interpreting a law, all parts of the statute should be considered together, not in isolation.

When read as a whole, the IIJA appears to be focused primarily on deployment discrimination by broadband providers. The FCC’s decision to extend its rules to cover pricing, marketing, and other practices by a wide range of entities might be seen as overreaching the authority granted by the law.

THE JARKESY DECISION

SEC v. Jarkesy established that defendants have a right to a jury trial in an Article III court when facing civil monetary penalties for alleged wrongdoing. This decision could have significant consequences for the FCC’s planned enforcement process for digital discrimination.

The current FCC process positions the commission as both prosecutor and judge in potential discrimination cases. The agency would determine if a violation occurred, whether the entity had a valid justification, and what penalties to impose, including “monetary forfeitures.” 

Under Jarkesy, this process will likely need to be revised to include a jury trial in federal court, at least in cases where the FCC seeks to impose monetary penalties.

THE OHIO V EPA DECISION

Ohio v. EPA reinforced the principle that agency actions must be “reasonable and reasonably explained.” Agencies must offer a satisfactory explanation for their actions, including a rational connection between the facts found and the choices made.

The FCC may face challenges on this front. Critics contend that the commission’s rules will discourage broadband investment and adoption, contrary to the IIJA’s goals of expanding broadband access. 

If the court finds that the FCC hasn’t adequately addressed these potential negative consequences, it could rule that the agency’s action was arbitrary and capricious.

THE CORNER POST DECISION

Corner Post Inc. v. Federal Reserve clarified that the statute of limitations for challenging agency actions doesn’t begin until an entity is directly affected by the action. This could have far-reaching implications for the FCC’s digital-discrimination rules.

Given the broad and somewhat vague definition of entities, policies, and practices that the rules cover, some organizations might not realize they’re subject to the regulations until the FCC takes action against them. Corner Post would allow these entities to challenge the rules at that point, even if it’s years after the rules were initially adopted.

For more on this issue, see “ICLE Ex Parte on Digital Discrimination“ and “Brief of ICLE and ITIF to 8th Circuit in Minnesota Telecom Alliance v FCC.”

Executive Summary

We appreciate the opportunity to respond to this request for information on corporate consolidation through serial acquisitions and roll-up strategies issued by the U.S. Justice Department (DOJ) and the Federal Trade Commission (FTC) (collectively, “the agencies”). We agree that robust competition across markets is critical to consumer welfare and the U.S. economy. The agencies have important roles to play in protecting competition and consumers against anticompetitive conduct—including, specifically, those mergers and acquisitions that have harmed, or are likely to harm, competition and consumers. We do not gainsay the importance of effective and efficient enforcement of the federal antitrust laws—including, but not limited to, the Clayton Act. Moreover, we recognize the substantial contributions that agency staff have made to such enforcement through systematic economic research and other policy studies.

While serial acquisitions and roll-up strategies merit further study, there is no apparent basis—in either the economic literature or the agencies’ enforcement experience—for any general changes to the procedures or substantive standards by which serial acquisitions are scrutinized. We also have concerns about the RFI’s approach and framing. The inquiry appears to presume that serial acquisitions tend to be harmful, without adequately acknowledging or requesting information on the potential procompetitive effects or efficiencies of such acquisitions. In addition, the broad scope of the inquiry, covering numerous industries and transaction types, may not yield sufficiently focused insights to inform policy initiatives going forward.

Our response comprises, at the highest level of generality, one observation and one recommendation. The observation is that, while serial acquisitions may constitute an important domain of merger scrutiny, neither enforcement experience nor the economic literature support any fundamental changes in procedural or substantive antitrust law or regulation for these types of transactions. Competition policy is not, and should not be, static. At the same time, sound policy reform is a difficult and iterative process, and one that requires a firm foundation in both research and enforcement experience, along with attention to established precedent.

Correspondingly, our overarching recommendation is that the agencies build on the substantial body of research regarding mergers and acquisitions that has been conducted over the course of several decades by agency staff and others. More specifically, we recommend that economic and policy staff at the agencies synthesize the existing body of research at their disposal. To be sure, market developments and developments in research methods and available data might suggest new avenues of research, as well as those in need of significant updates. But a serious, critical synthesis of the available literature will help to sharpen the agencies’ sense of new research demands, just as it will provide a basis from which to contemplate new enforcement initiatives.

This research-focused approach is particularly crucial for understanding serial acquisitions and roll-up strategies, given their complexity and likely variation across industries. In pharmaceuticals, for example, serial acquisitions may be a vital mechanism to bring innovations to market, while in other sectors they might raise more significant or more frequent competitive concerns. Only through rigorous, sector-specific research can the agencies develop a nuanced understanding of these strategies’ competitive effects.

Furthermore, we encourage the agencies to:

1. Develop more focused and productive requests for information on critically important issues in serial acquisitions going forward.
2. Recognize that sound policy reform requires a firm foundation in both research and enforcement experience, along with attention to established precedent.
3. Be cautious about drawing general conclusions about whole industries, business models, or methods of acquisition, and more cautious still in condemning them.
4. Consider the tradeoffs inherent in any new reporting requirements or regulatory approaches, carefully weighing potential benefits against the burdens imposed on businesses and agency resources.

The following sections elaborate on these points and provide specific responses to the RFI questions. Our goal is to encourage a more balanced, evidence-based approach to addressing potential competitive concerns arising from serial acquisitions and roll-up strategies, grounded in rigorous research and analysis.

I. Framing and Conceptual Problems

The role of an RFI should be to collect diverse objective data and perspectives that can help the agencies to better understand complex market dynamics within and across industries. Unfortunately, the current RFI appears to deviate from this purpose. The questions appear designed to elicit responses supporting a predetermined negative view of serial acquisitions, rather than to gather balanced information. It thus creates the unfortunate appearance that the agencies are primarily interested in collecting negative perspectives about serial acquisitions, rather than seeking to learn about their potential—and potentially varied—benefits and harms to competition and consumers.

A. The General Framing Suggests an Answer

The framing of this RFI seems, in many ways, conclusory. It does not suggest a neutral inquiry but, rather, raises significant concerns about bias in the agencies’ approach to serial acquisitions.

The RFI’s language often presumes negative outcomes from serial acquisitions. For instance, the document states that “some companies use serial acquisitions of small firms as a business strategy that can harm competition to the detriment of consumers, workers, and innovation in an industry or business sector without detection by the Agencies.”[1] This statement, presented in the background section, sets a tone that appears to prejudge the issue. It suggests that serial acquisitions are inherently harmful, without acknowledging potential procompetitive effects or efficiencies. After all, only 2% of all mergers subject to premerger notification receive second requests; moreover, a second request is not a complaint, much less a final decision that a proposed merger would be unlawful.[2] Do the agencies believe that the proportion of harmful mergers is higher for small acquisitions or multiple acquisitions by a given firm, whether within or across markets? If not, the presumption of harm seems misguided.

Similarly, several of the RFI’s questions are framed in ways that appear to seek confirmation of harm, rather than objective information. Question 2(c), for example, asks whether respondents have “witnessed any actual or attempted coordination or collusion between competitors that you did not notice prior to the serial acquisitions”[3]—a highly unscientific approach, both as a means of sampling and in that it assumes or suggests causality. Question 3 lists nine specific business practices, all of which are framed negatively, such as “[s]elling products or services below cost with the goal or effect of rivals or driving them out of the market.”[4]

Meanwhile, the agencies’ public outreach around the RFI betrays the same presumption of harm, as well as an apparent quest for enforcement targets, rather than better understanding. The subheading on both the FTC’s and DOJ’s press releases announcing the inquiry explicitly asks for information on harmful acquisitions: “Agencies launch public inquiry to identify serial acquisitions, roll-ups that have harmed competition, consumers, workers, and innovation.”[5] Both press releases also include similarly biased quotes from FTC Chair Khan, referring to “stealth consolidation schemes” and highlighting that “[f]irms can use serial acquisitions to roll up markets, consolidate power and undermine fair competition, all while jacking up prices and degrading quality.”[6] Of course, serial acquisitions, like any acquisition, can have these deleterious effects. But the plain implication in the agencies’ communications is that serial acquisitions (perhaps unlike other acquisitions) inherently do have these harmful characteristics.

As noted by former antitrust enforcers in their response to this RFI, “[a]lthough several questions take a neutral approach, many of them solicit negative information about acquisitions, and not one asks about any benefits.”[7] This imbalance is particularly striking, given the inquiry’s breadth and the fact that, as the former enforcers likewise highlight, the agencies have stated that “the vast majority of mergers are either procompetitive and enhance consumer welfare or are competitively benign.”[8]

This apparent bias in the RFI is disconcertingly reinforced by the agencies’ recent statements to the Organisation for Economic Co-operation and Development (OECD).[9] In those comments, the FTC stated that “[s]erial acquisition strategies which aim to achieve—or actually achieve—high market share, exclusion of competitors, suppression of wages, reduction in innovation, or pricing power may violate Section 2.”[10] The commission also noted that “[t]he Agencies are focused on enforcement against serial acquisition strategies.”[11]

If “may violate,” in this context, suggests mere possibility, it is uninformative: Any given acquisition may or may not run afoul of Section 2, as there is no recognized class of mergers that is per se lawful. But “may violate” here would appear to be more naturally read as a suggestion of suspicion, or a predisposition on the part of the federal enforcement agencies toward viewing serial acquisitions as inherently or likely problematic.

These statements, along with examples of enforcement actions, suggest that the FTC has already formed conclusions about the competitive effects of serial acquisitions and determined its enforcement approach. This predetermined stance would contradict the ostensibly open nature of the current RFI, which purports to seek objective information to inform policy, not to vindicate prior policy suspicions.

As enforcement agencies, the FTC and DOJ have a responsibility to be vigilant and duly skeptical when examining potential anticompetitive practices. Their mandate to protect competition and consumers necessitates scrutiny of business practices that could harm market dynamics. This skepticism is not only appropriate, but essential to their role. The RFI’s framing, however, appears to go beyond reasonable skepticism to approach presumption, if not prejudgment. While it is legitimate for the agencies to investigate practices they believe may be harmful, an RFI is a tool for gathering objective information to inform policy and to help establish enforcement priorities.

Careful information gathering, like systematic research, also requires a degree of skepticism about its own priors, including working hypotheses. The current framing lacks such balance or care, ahead of any specific findings or even preliminary analysis of data. For instance, while it is appropriate to ask about potential harms from serial acquisitions, a balanced inquiry would also—at a bare minimum—seek information about potential benefits or efficiency gains associated with or caused by such acquisitions. As noted, the current RFI’s lack of such balance suggests a predisposition toward finding harm, rather than an open-ended exploration of the complex competitive dynamics surrounding serial acquisitions.

This approach not only risks biasing the information received, but it may also undermine the credibility of the agencies’ ultimate enforcement efforts. Effective enforcement requires not just identifying anticompetitive practices but also understanding the full context in which business strategies operate. By appearing to presume harm from serial acquisitions, the agencies may miss important nuances that could inform more targeted and effective enforcement actions.

These concerns are especially troubling when viewed alongside, for example, the 2024 interagency RFI regarding consolidation in health-care markets[12] and the 2023 proposed changes to Hart-Scott-Rodino Act (HSR) rules.[13] As we pointed out in our comments to those inquiries, the workshop that accompanied the healthcare RFI “seems to have been conclusory by design.”[14] Similarly, the HSR rules would dramatically increase the burden of premerger notification, requiring extensive new information on labor markets, nonhorizontal relationships, and other areas—again, without clear evidence that such information is necessary or appropriate for effective merger screening.[15] Both of these initiatives, like the current RFI, suggest presumptions that mergers are unlawful far more often than they have proven to be. Both also suggest substantially increasing scrutiny and reporting burdens on merging parties without clear evidence that such measures are likely to identify anticompetitive harms.

The leading nature of this RFI’s questions, especially when considered alongside other recent agency actions, raises concerns about the objectivity of the agencies’ approach to serial acquisitions. We urge the agencies to adopt a more balanced, evidence-based approach that considers both the potential harms and benefits of these business strategies. Such an approach would be more consistent with established antitrust principles and more likely to yield insights that truly enhance the agencies’ ability to protect competition and consumer welfare.

B. Conceptual Issues

The RFI also suffers from conceptual issues—mostly concerning the question of what constitutes a “serial acquisition”—that would make subsequent enforcement based on these issues inherently arbitrary.

The RFI’s definition of serial acquisitions as “the same firm consolidating a fragmented market through a number of acquisitions, typically of many relatively small companies” lacks precision and raises several questions.[16] What constitutes a “number” of acquisitions? Over what period should these acquisitions be considered “serial”? How small must the acquired companies be to fit this definition?

Precise numerical answers to these questions are not required at this stage of the agencies’ inquiry, nor are fixed thresholds. Still, without clearer parameters—without some way to cabin the agencies’ inquiry—there is a risk of disparate and unfocused responses to the RFI. Further, a signal of overly broad agency scrutiny could chill legitimate business activity. For instance, two acquisitions over 14 years might be considered “serial” under a broad definition, yet such a pattern hardly suggests a systematic strategy to consolidate a market. Indeed, even within a given industry or sector, a considerably larger number of acquisitions and a narrower time frame may commonly be procompetitive or benign if they range across product or geographic markets.

Moreover, the difficulty in precisely defining “serial acquisitions” raises concerns about potential arbitrary enforcement. While wide-ranging requests for diverse inputs may be suitable for preliminary information gathering, clear, objective criteria are essential for rigorous study and more focused investigations and enforcement. Such criteria are also likely important to develop guidance for businesses to understand their obligations, and for consistent application of the law.

II. Do Serial Acquisitions Present Unique Competition Issues?

Many of the activities described as “serial acquisitions” are indistinguishable from normal patterns of business growth and consolidation in maturing industries. As a general matter, it is not clear why a company growing through multiple small acquisitions should be viewed differently than one growing “organically” or through fewer, larger acquisitions. This raises important questions about the agencies’ underlying theory of harm. If the concern is market concentration, this can manifest through various means, not just serial acquisitions. If the concern is the specific process of multiple small acquisitions, it’s unclear why this would be inherently more problematic than other forms of growth.

Recent research by Jonathan Cohn, Edith Hotchkiss, and Erin Tower sheds light on the motivations behind roll-up strategies in private-equity buyouts of private firms.[17] Their study suggests that these strategies are often driven by two primary motives: unlocking growth potential in capital-constrained firms and improving operational performance in underperforming firms. They find that acquired firms often experience significant increases in sales growth and moderate improvements in profitability post-acquisition, supporting the view that these strategies can create value through both growth and operational improvements. These findings suggest that properly executed roll-up strategies can serve legitimate business purposes beyond mere market consolidation.

Given the legitimate business reasons for acquisitions (serial or not), we are aware of no theoretical or empirical grounds on which to suppose that multiple acquisitions are typically anticompetitive. At the same time, there is no reason to suppose that the organic growth of a firm precludes anticompetitive conduct. The competitive effects of growth, whether through acquisition or internal expansion, depend on a variety of factors—including market structure, barriers to entry, and the specific capabilities and assets being acquired or developed. For example, in some cases, serial acquisitions might allow a firm to quickly assemble complementary assets and capabilities, leading to increased innovation and more robust competition. In other instances, organic growth might allow a firm to build market power in ways that are difficult for competitors to challenge.

By identifying such a broad range of serial acquisitions for special scrutiny, the agencies risk deterring transactions that may be procompetitive or competitively neutral. This is particularly concerning given that, as noted in our HSR comments, “the vast majority of mergers are either procompetitive and enhance consumer welfare or are competitively benign.”[18] The costs of deterring beneficial transactions could be substantial, including reduced incentives for startup formation and innovation, decreased liquidity in capital markets, lost efficiencies from beneficial consolidation, and reduced competitive pressure on incumbent firms.

A. Distinctions Between Organic Growth and Acquisitions Are Baseless

The agencies’ RFI appears animated by a fundamental presumption that it is better for firms to “build” new capabilities than to “buy” them.

While eschewing the more contentious form of the statement in their draft guidelines,[19] the agencies assert in the 2023 Merger Guidelines that “[i]n general, expansion into a concentrated market via internal growth rather than via acquisition benefits competition.”[20] The FTC’s challenge of the Meta/Within transaction centers on the agency’s claim that Meta’s decision to buy its way into a new market obviated its role as a potential entrant through organic growth, thereby reducing competition.[21] As John Newman, then-deputy director of the FTC Bureau of Competition, said in criticizing Meta’s acquisition strategy:

Instead of competing on the merits, Meta is trying to buy its way to the top. Meta already owns a best-selling virtual reality fitness app, and it had the capabilities to compete even more closely with Within’s popular Supernatural app. But Meta chose to buy market position instead of earning it on the merits. This is an illegal acquisition, and we will pursue all appropriate relief.[22]

The FTC’s view has been supported by commentators who contend that, despite their ability to grow organically, many incumbent firms have “opportunities to ‘roll up’ (willing) startups to ‘get there faster,’ ‘buying’ instead of expending effort in rival innovation.”[23] Their conclusion is that “[f]oregoing such effort is never good for consumers and society as a whole.”[24] Indeed, these commentators contend that “the general conclusion of the academic literature is that consumers and society are better off when innovative firms are not permitted to merge.”[25]

But this blunt distinction between “competition on the merits” and growth and entry through acquisition is unwarranted, and not well-supported by the literature. As Jay Ezrielev has observed: “There is, however, no basis for concluding that build-or-buy acquisitions harm innovation. Despite providing ‘build’ incentives, restrictions on buying may lead to less building and less innovation.”[26]

Not only do many firms pursue “hybrid strategies,” as mentioned above, but acquisition by itself can constitute competition on the merits for several reasons. First, acquisitions can lead to significant efficiency gains. These may include cost savings, better resource allocation, and enhanced innovation capabilities, which can ultimately benefit consumers. Second, acquisitions may be the most efficient and effective way to increase competition, by facilitating incumbents’ entry into other markets. As Ezrielev notes: “Buying rather than building frees up resources that the acquirer may use for other innovations. Buying also allows firms to expand into new markets faster and with more certainty.”[27] Third, the potential for takeovers can also drive dynamic competition by encouraging managers to perform sufficiently well to avoid becoming targets. This competitive pressure can also lead to overall improvements in industry performance.[28]

Acquisitions may also serve to reallocate and recombine resources in ways that spur innovation.[29] It may be more efficient to acquire a firm with specialized technologies than to develop those technologies internally. Conversely, acquired firms that excel at developing technology may themselves require managerial and other resources of the acquired firm to develop and commercialize their innovations.[30]

Potential acquisition is also a key exit strategy that makes financing startups more attractive. As Joanna Shepherd similarly notes: “Forcing inefficient builders to build would raise costs and discourage more efficient builders from undertaking building projects as these projects would have fewer potential buyers.”[31] As a result, the existence of a robust merger market offers incentives to create new firms in the first place.

As Jay Ezrielev explains:

The main purpose of build-or-buy acquisitions is to expand into a new market (or to expand a firm’s capacity in a market it already serves) in the most cost-effective way. These acquisitions do not weaken incentives to continue ongoing innovations. On the contrary, many acquirers may be looking to accelerate the development of the target’s innovation. These acquirers may be attracted to the target’s technology precisely because it is promising. Acquirers may be uniquely qualified to recognize the true potential of the target’s innovation because of their related expertise.[32]

By the same token, with respect to vertical acquisitions, there is no basis for the oft-repeated claim that anything a firm can accomplish by merger, it can accomplish by contract, nor that the latter is preferable because it avoids increased consolidation and foreclosure incentives.[33] In reality, just like the decision to buy rather than build, the decision to buy rather than contract is often motivated by important distinctions that make the one preferable to the other.[34]

It is true that there are “limits to vertical integration. In particular, there may be limits to his [sic] entrepreneur’s resources or abilities, such that adding an additional transaction within the firm would generate decreasing returns.”[35] But acquisition can also be distinctly preferable to contracting. It can often be more efficient to organize production within a firm: “[F]irms have a fundamentally different production function from separate, additive market-based production—and [] cooperative team-based production could be much more efficient.”[36]

At the same time, at various stages of the industry life cycle, firms exhibit increasing returns to scale, making acquisition relatively more efficient than contracting, depending on the maturity and other attributes of the relevant industry.[37] Meanwhile, contracting becomes less attractive as asset specificity (i.e., specialized investments or assets that are extremely costly to relocate or redeploy) increases.[38] By the same token, as the costs of contracting increase due to complexity and lack of information (and the risks of asset specificity), the costs of the resulting incomplete contracts increase, again making acquisition relatively more efficient.[39]

Of course, the distinction between vertical and horizontal transactions will also often be blurred, and this, too, provides further reason to be skeptical of claims that perceived horizontal “buy-rather-than-build” acquisitions are harmful. In the case of Meta/Within, for example, while the FTC focused on the acquisition’s purported horizontal effects in the virtual-reality (VR) fitness-app market, it is undeniable that the deal simultaneously allowed Meta to rapidly expand its portfolio of VR-app offerings.[40] In this way, it also signaled a commitment to Meta’s development of the broader ecosystem on which all VR-app markets depend. Meta’s primary interest in the transaction (and the real benefit of the deal for consumers and other app developers) may have been for its broader contribution to the development of Meta’s platform, rather than its narrow contribution to Meta’s place in any particular VR-app market.

In much the same way, the distinction between “organic” growth and growth through acquisition is not always clearcut. This is particularly true at a time when antitrust enforcers around the world are increasingly tempted to treat minority shareholdings and key personnel hires as notifiable mergers.[41] Indeed, the concept of the so-called “acqui-hire” provides a compelling argument against a rigid preference for internal growth over mergers. An “acqui-hire” is an acquisition primarily aimed at acquiring the talent of a company, rather than its products, services, or other assets.

But in this regard, the very aim of an aqui-hire is ultimately to spur organic growth. And, indeed, many firms pursue hybrid strategies, combining internal development with strategic acquisitions. Mergers of this sort can be essential for internal growth, as they bring in critical human resources that drive innovation and competitiveness. The agencies should be cautious about drawing sharp distinctions between these growth strategies without robust empirical evidence of their differential competitive effects.

B. Issues Particular to Small, Serial Acquisitions

Certainly, it is possible for a series or sequence of acquisitions to prove anticompetitive, whether those acquisitions fall under or above HSR reporting thresholds. Indeed, even a single acquisition may prove anticompetitive. At the same time, the costs of increased scrutiny must be weighed against the likely benefits. As noted in comments we previously submitted to the FTC,[42] the 2023 proposed changes to the premerger notification rules could lead to between $350 million and $2.23 billion in additional annual compliance costs,[43] with other additional costs imposed on agency staff. Extending similar requirements to smaller transactions would necessarily add to those costs, potentially imposing disproportionately high burdens on relatively small filers. Moreover, reviewing a larger number of small transactions would significantly increase the agencies’ workload, potentially diverting resources from more critical enforcement priorities.

The potential for errors in judgment is particularly high when dealing with small acquisitions and evolving markets. False positives could stifle innovation and efficiency, while false negatives could allow anticompetitive harm. Given that most mergers are procompetitive or benign, an approach that scrutinizes all serial acquisitions risks an unacceptably high false-positive rate.

While we acknowledge that a pattern of acquisitions could potentially raise competitive concerns in specific circumstances, we are skeptical that “serial acquisitions” represent a distinct phenomenon that requires special treatment. We urge the agencies to focus on case-specific analysis rather than creating new, potentially overbroad categories of acquisitions for heightened scrutiny. In the alternative, the agencies should use the information informally gathered in response to the RFI—together with further economic research and enforcement experience—to sharpen our understanding of serial acquisitions and the circumstances under which greater scrutiny of such transactions may be cost-justified. Any changes to the current approach should be based on robust empirical evidence of harm and a careful consideration of costs and benefits.

Private-equity investment can bring substantial benefits. Even the DOJ has previously recognized the potential advantages of private-equity buyers in certain merger-remedy scenarios, though this view has since shifted.[44] Research has shown that private-equity investments can lead to improved operational efficiency and total factor productivity, increased innovation, and enhanced competitiveness through entry and exit in the industry.[45]

For example, Steven Davis and his co-authors find that private-equity buyouts lead to a 2.1 log point increase in total factor productivity at target manufacturing firms over two years post-buyout, compared to controls.[46] The authors also find that buyouts catalyze creative destruction, increasing job creation and destruction rates by 14 percentage points over two years.[47] Relatedly, contrary to concerns about short-term benefits but long-term harms, Lerner et al. found that leveraged buyouts do not sacrifice long-term investments in innovation.[48]

It’s important to note, however, that the effects of private-equity investment can vary significantly depending on the specific industry, firm, and investment strategy involved.[49] The shift in the agencies’ stance towards private equity, particularly in merger remedies, should be carefully evaluated against this nuanced body of evidence to ensure that potentially beneficial investments are not unduly deterred.

III. Is Industry Variance in Serial Acquisitions Evidence of Harm?

The competitive implications of multiple acquisitions can also vary significantly both across and within industries, underscoring the importance of context-specific analysis, rather than broad generalizations or simple numerical thresholds. Just as we cannot look at concentration measures across industries as a uniform indicator of competitive harm, we similarly cannot view the number of acquisitions in isolation as a reliable signal of market harm.

Different industries have distinct characteristics that shape their competitive dynamics, innovation cycles, and efficient scales of operation.[50] These factors critically influence the role and effect of acquisitions within each sector. For instance, in pharmaceuticals, serial acquisitions may be a vital mechanism to bring innovations to market, especially given regulatory hurdles and the substantial costs of adequate clinical trials above and beyond those of primary pre-clinical research.[51] Small pharmaceutical or biotech firms, while excelling in early-stage research, frequently lack the resources for large-scale clinical trials, regulatory approval, and commercialization.[52] Acquisitions by larger entities can bridge this gap, facilitating the entry and delivery of new drug therapies to the benefit of both competition and consumers (namely, patients).

Thus, while some pharmaceutical and bio-tech acquisitions might prove anticompetitive, multiple or serial acquisitions could often, or on average, prove pro-competitive, signaling a vibrant innovation ecosystem rather than anti-competitive behavior. Indeed, as Joanna Shepherd has noted: “In industries in which most innovation originates externally…, analyses should be less concerned with mergers’ impacts on internal innovation, and more focused on whether consolidation will increase demand for externally-sourced innovation and, ultimately, increase aggregate drug innovation.”[53] This perspective underscores the importance of understanding the specific dynamics and needs of different sectors in order to accurately assess the implications of serial acquisitions.

Similarly, in technology sectors, acquisitions of small, innovative firms can accelerate the development and deployment of new technologies.[54] Many tech startups are founded with the intent and expectation of being acquired, and the viability of this type of exit strategy in turn creates incentives for investment and innovation. The fast-paced nature of technological change means that the competitive landscape can shift rapidly, and what might appear to be market consolidation through acquisitions might instead be a series of attempts to keep pace with evolving consumer demands and technological possibilities.

In contrast, in more mature industries with stable technologies, a pattern of acquisitions might warrant closer scrutiny. Even here, however, the specific market conditions matter. In fragmented industries like home services or local retail, roll-up strategies can create efficiencies of scale that benefit consumers through lower prices or improved service quality.[55]

Moreover, the relevant geographic market can vary dramatically across industries, further complicating any attempt to apply uniform standards. A series of local acquisitions in a national market might have minimal competitive impact, while similar acquisitions in a local or regional market could be more significant. This variability in geographic-market definition is well-documented in antitrust literature and case law. For instance, in the health-care industry, hospital markets are often defined locally or regionally, as evidenced by the FTC’s lawsuit against U.S. Anesthesia Partners Inc., which focused on the Houston and Dallas markets.[56]

The dynamic nature of many industries also means that market boundaries and competitive threats can change rapidly. Today’s competitors might be tomorrow’s complements, or vice versa. This fluidity makes it particularly challenging to assess the long-term competitive impact of a series of acquisitions based solely on their number or frequency.

IV. Conclusion

In light of these complexities, we urge the agencies to resist the temptation to rely on simple metrics or across-the-board presumptions about serial acquisitions. Instead, a nuanced, industry-specific approach is necessary to accurately assess the competitive implications of multiple acquisitions. This approach should consider such factors as the pace of technological change, the role of innovation, typical firm lifecycles, efficient scale of operations, and the specific competitive dynamics of each industry.

For example, in many smaller markets, independent providers of hospital-based services (such as anesthesiology) may be highly concentrated on any standard for “highly concentrated” markets. Further research might aid the agencies in examining highly concentrated provider markets to develop filters or screens for provider acquisitions below the HSR filing threshold, so that the agencies might identify and investigate those subthreshold filings that are the most likely candidates for investigations and, depending on the results of those investigations, enforcement actions. Efficient matter-selection tools will be critical to that effort, less the agencies commit scarce resources to small, unpromising investigations and impose undue costs on health-care providers.

By recognizing and accounting for these industry-specific factors, the agencies can more accurately identify truly problematic acquisition patterns, while avoiding undue interference in benign or beneficial market evolution. This nuanced approach is crucial to maintain a competitive economy that fosters innovation and efficiency across all sectors.

[1] Request for Information for Public Comment on Corporate Consolidation Through Serial Acquisitions and Roll-Up Strategies, Fed. Trade Comm’n & U.S. Dep’t of Justice, Docket No. FTC-2024-0028 (May 22, 2024), at 1, https://www.regulations.gov/docket/FTC-2024-0028 [hereinafter “RFI”].

[2] Lina Khan & Jonathan Kanter, Hart-Scott-Rodino Annual Report, Fiscal Year 2022, Fed. Trade Comm’n & U.S. Dep’t of Justice (Jan. 4, 2024), at 23, available at https://www.ftc.gov/system/files/ftc_gov/pdf/FY2022HSRReport.pdf.

[3] RFI at 5.

[4] RFI at 6.

[5] Press Release, FTC and DOJ Seek Info on Serial Acquisitions, Roll-Up Strategies Across U.S. Economy, Fed. Trade Comm’n (May. 23, 2024), https://www.ftc.gov/news-events/news/press-releases/2024/05/ftc-doj-seek-info-serial-acquisitions-roll-strategies-across-us-economy; Press Release, Justice Department and Federal Trade Commission Seek Information on Serial Acquisitions, Roll-Up Strategies Across U.S. Economy, U.S. Dep’t of Justice (May 23, 2024).

[6] Id.

[7] Asheesh Agarwal et al., Former Enforcers Comment on Request for Information on Corporate Consolidation Through Serial Acquisitions and Roll-Up Strategies (Jun. 26, 2024), at 2, available at https://www.regulations.gov/comment/FTC-2024-0028-0214.

[8] Id., at 3 (citing Statement of Ass’t Att’y Gen. Christine Varney, Merger Guidelines Workshops, Third Annual Georgetown Law Global Antitrust Enforcement Symposium (Sep. 22, 2009)).

[9] See Serial Acquisitions and Industry Roll-ups—Note by the United States, OECD DAF/COMP/WD(2023)99 (Dec. 6, 2023), available at https://one.oecd.org/document/DAF/COMP/WD(2023)99/en/pdf.

[10] Id., at 6.

[11] Id., at 7.

[12] Request for Information on Consolidation in Health Care Markets, Docket No. ATR-102, Dep’t Justice, Dep’t Health & Human Servs., & Fed. Trade Comm’n (Mar. 5, 2024), https://www.regulations.gov/docket/FTC-2024-0022; Daniel J. Gilman & Geoffrey A. Manne, ICLE Comments Re: Request for Information on Consolidation in Health Care Markets, Int. Ctr. Law Econ. (Jun. 5, 2024), https://laweconcenter.org/resources/icle-comments-re-request-for-information-on-consolidation-in-health-care-markets.

[13] Premerger Notification Rules, 88 Fed. Reg. 42178, (Jun. 29, 2023), (to be codified at 16 C.F.R. Parts 801 and 803); Brian Albrecht, Dirk Auer, Daniel J. Gilman, Gus Hurwitz, & Geoffrey A. Manne, Comments of the International Center for Law & Economics on Proposed Changes to the Premerger Notification Rules, Int. Ctr. Law Econ. (Sep. 27, 2023), https://laweconcenter.org/resources/comments-of-the-international-center-for-law-economics-on-proposed-changes-to-the-premerger-notification-rules.

[14] Gilman & Manne, supra note 12, at 6.

[15] Albrecht et al., supra note 13, at 7-9.

[16] RFI at 1.

[17] Jonathan B. Cohn, Edith S. Hotchkiss, & Erin M. Towery, Sources of Value Creation in Private Equity Buyouts of Private Firms, 26 Rev. Fin. 257 (2022).

[18] Albrecht et al., supra note 13, at 5.

[19] See Draft Merger Guidelines, U.S. Dep’t of Justice & Fed. Trade Comm’n (2023), available at https://www.ftc.gov/system/files/ftc_gov/pdf/p859910draftmergerguidelines2023.pdf (The assertion in the draft guidelines was: “The antitrust laws reflect a preference for internal growth over acquisition.”)

[20] 2023 Merger Guidelines, U.S. Dep’t of Justice & Fed. Trade Comm’n (2023), available at https://www.ftc.gov/system/files/ftc_gov/pdf/2023_merger_guidelines_final_12.18.2023.pdf.

[21] See Press Release, FTC Seeks to Block Virtual Reality Giant Meta’s Acquisition of Popular App Creator Within, Fed. Trade Comm’n (Jul. 27, 2022), https://www.ftc.gov/news-events/news/press-releases/2022/07/ftc-seeks-block-virtual-reality-giant-metas-acquisition-popular-app-creator-within (“Meta is a potential entrant…. But instead of entering, it chose to try buying Supernatural. Meta’s independent entry would increase consumer choice, increase innovation, spur additional competition to attract the best employees, and yield other competitive benefits. Meta’s acquisition of Within, on the other hand, would eliminate the prospect of such entry, dampening future innovation and competitive rivalry.”).

[22] Id.

[23] Cristina Caffarra, Gregory S. Crawford, & Tommaso Valletti, “How Tech Rolls”: Potential Competition and “Reverse” Killer Acquisitions, CPI Antitrust Chron. (May 2020), at 2, available at https://www.competitionpolicyinternational.com/wpcontent/uploads/2020/05/CPI-Caffarra-Crawford-Valletti.pdf.

[24] Id.

[25] Id., at 4.

[26] Jay Ezrielev, Antitrust Policy and Legal Standards for Build-or-Buy Decisions, CPI Columns US & Canada (Mar. 2024), at 2, https://www.pymnts.com/cpi-posts/antitrust-policy-and-legal-standards-for-build-or-buy-decisions.

[27] Id.

[28] Henry G. Manne, Mergers and the Market for Corporate Control, 73 J. Pol. Econ. 110 (1965).

[29] See generally Carl Shapiro, Competition and Innovation: Did Arrow Hit the Bull’s Eye?, in The Rate and Direction of Inventive Activity Revisited (Josh Lerner and Scott Stern, eds., 2012), at 361–404. Shapiro calls this the “Synergies Principle: ‘Combining complementary assets enhances innovation capabilities and thus spurs innovation.’” Id., at 365. As he explains: “The Synergies principle emphasizes that ?rms typically cannot innovate in isolation. The quest for synergies is especially important in industries where value is created by systems that incorporate multiple components, as in the information and communications technology sector. The Synergies principle is directly relevant for competition policy since procompetitive mergers and business practices allow for the more e?cient combination of complementary assets.” Id.

[30] See, e.g., Joanna Shepherd, Consolidation and Innovation in the Pharmaceutical Industry: The Role of Mergers and Acquisitions in the Current Innovation Ecosystem, 21 J. Health Care L. & Pol’y 1 (2018).

[31] Id.

[32] Ezrielev, supra note 26, at 3.

[33] See, e.g., Steven C. Salop, The AT&T/Time Warner Merger: How Judge Leon Garbled Professor Nash, 6 J. Antitrust Enforcement 459, 468 (2018).

[34] See generally Geoffrey A. Manne, Kristian Stout, & Eric Fruits, The Fatal Economic Flaws of the Contemporary Campaign Against Vertical Integration, 68 Kansas L. Rev. 923 (2020).

[35] Id., at 939 (citing Ronald Coase, The Nature of the Firm, 4 Economica 386, 394 (1937)).

[36] Id., at 940.

[37] See George Stigler, The Division of Labor Is Limited by the Extent of the Market, 59 J. Pol. Econ. 185, 188 (1951); Manne, Stout, & Fruits, supra note 32, at 941 (“[I]n Stigler’s model, the degree of vertical integration is a dynamic interaction among the firm’s economies of scale as well as the extent of the firm’s market and the market for its inputs.”).

[38] See Oliver E. Williamson, The Economic Institutions of Capitalism 85-102 (1985); see also Paul L. Joskow, Vertical Integration, 55 Antitrust Bull. 543, 563 (2010).

[39] See Benjamin Klein, Robert G. Crawford, & Armen A. Alchian, Vertical Integration, Appropriable Rents, and the Competitive Contracting Process, 21 J.L. Econ. 297, 306-09 (1978).

[40] FTC v. Meta Platforms Inc., U.S. Dist. LEXIS 29832 (2023);

[41] See, e.g., Ryan Browne, Microsoft’s Hiring of Staff from AI startup Inflection Referred for UK Merger Probe, CNBC (Jul. 16, 2024), https://www.cnbc.com/2024/07/16/microsoft-hiring-of-ai-inflection-staff-referred-for-uk-merger-probe.html; Brandon Vigliarolo, UK Antitrust Cops Thrust Probe into Microsoft, Inflection AI Merger, The Register (Jul. 16, 2024), https://www.theregister.com/2024/07/16/uk_competition_officials_kick_off.

[42] Albrecht et al., supra note 13.

[43] Id., at 7-8.

[44] See Bernard A. Nigro Jr. & Harrisson C. Kummer, Unraveling the Roll-Up: Private Equity’s Misunderstood Investment Strategy, CPI Antitrust Chron. (Jun. 2024). The authors discuss the benefits of private-equity investment, citing the DOJ’s 2020 Merger Remedies manual (now withdrawn), which acknowledged that “in some cases a private equity purchaser may be [the] preferred” remedies buyer. See Merger Remedies Manual 24, Dep’t of Justice Antitrust Division (Sep. 2020, withdrawn Apr. 2022), https://www.justice.gov/atr/page/file/1312416/dl. Nigro & Kummer also reference a recent court decision that rejected DOJ arguments against a private-equity divestiture buyer, finding the buyer “well-positioned to maintain, and perhaps even improve upon” competition. United States v. UnitedHealth Group Inc., 630 F. Supp. 3d. 118, 137 (D.D.C. 2022).

[45] See Shai Bernstein et al., Private Equity and Industry Performance, 63 Mgmt. Sci. 1198 (2017).

[46] Steven J. Davis et al., Private Equity, Jobs, and Productivity, 104 Am. Econ. Rev. 3956 (2014).

[47] Id.

[48] Josh Lerner et al., Private Equity and Long-Run Investment: The Case of Innovation, 66 J. Fin. 445 (2011).

[49] Steven J. Davis, John C. Haltiwanger, Kyle Handley, Ben Lipsius, Josh Lerner, & Javier Miranda, The (Heterogeneous) Economic Effects of Private Equity Buyouts (Nat’l Bureau of Econ. Research, No. 26371, 2019), https://doi.org/10.3386/w26371.

[50] See generally Stigler, supra note 35.

[51] See Shepherd, Consolidation and Innovation in the Pharmaceutical Industry, supra note 30.

[52] See Jeffery M. Drazen et al., Drug-Development Challenges for Small Biopharmaceutical Companies 376 New Engl. J. Med. 469 (2017), (“Small biopharmaceutical companies often encounter important challenges in designing and implementing clinical development programs. In a context in which only approximately 10% of clinical programs result in drugs that achieve regulatory approval, small-company clinical programs may have an even lower rate of success than that of large companies owing to limited internal experience in clinical development and limited infrastructure, which may also affect manufacturing and clinical supply. However, these challenges are largely overshadowed by limited resources and funding, which in turn fuel demand for short timelines owing to the need to demonstrate progress to investors. As such, these companies must focus their resources on small, less-costly development programs for very specific targets and often must spearhead new approaches to testing new products in order to survive.”).

[53] Joanna Shepherd, Understanding Innovation Markets in Antitrust Analysis, Truth on the Market (Mar. 30, 2017), https://truthonthemarket.com/2017/03/30/understanding-innovation-markets-inantitrust-analysis-ag-biotech-symposium.

[54] See discussion of the Meta/Within merger, infra Section III.A.

[55] See Stigler, supra note 35, at 190.

[56] Fed. Trade Comm’n v. U.S. Anesthesia Partners, Inc., et al., No. 2010031, Complaint Filed (S.D. Tex. Sep. 21, 2023).

I. Introduction

On behalf of the International Center for Law & Economics (ICLE), we thank the National Telecommunications and Information Administration (NTIA) for the opportunity to respond to NTIA’s Proposed Broadband Equity, Access, and Deployment (BEAD) Alternative Broadband Technology Guidance.[1]

The BEAD program provides $42.45 billion to U.S. states, territories, and the District of Columbia to help with broadband planning, deployment, mapping, equity, and adoption. To fulfill BEAD’s promise, the NTIA’s guidance should encourage participation from as many types of broadband providers as possible. Congress required that subgrantees provide broadband service with at least 100 Mbps speeds for downloads and 20 Mbps for uploads, with low latency and low risk of network outages, be made available to every customer served by the project. See 47 U.S.C. §1702(f)(4)(A). In other words, the statutory language does not stipulate any preference for a particular technology, so long as those metrics are met. Currently, the proposed guidance strongly favors fiber projects by placing burdens on other providers that simply do not exist for fiber. This will discourage unlicensed fixed wireless and low-earth-orbit (LEO) satellite services from participating, to the detriment of many areas where those technologies are the best—and perhaps only—available option.

While the NTIA may have reasons to prefer fiber, it must acknowledge the tradeoffs that accompany giving it such strong preferences. Fiber is much more expensive to build out, especially in areas with low population density. Alternative technologies like fixed wireless and LEO satellite services can reach those areas in a much more cost-efficient manner. If grantees fail to maximize the value of BEAD funding by adopting less efficient means to reach unserved locations, it will mean less money available to connect underserved locations and anchor institutions, as well as depleting the funds available for other equity programs. The NTIA should adopt a technology-neutral approach that promotes the most efficient use of BEAD funds.

A. Summary of Proposed Guidance

The NTIA’s proposed guidance on technologies eligible for broadband deployment under the BEAD program establishes a hierarchy for awarding projects:

1. Priority Broadband Projects (end-to-end fiber);
2. Other Reliable Broadband Service projects; and
3. Alternative Technology projects (only where cost exceeds the Extremely High Cost Per Location Threshold).

Alternative technology projects are defined as those employing any broadband-access technology that does not qualify as reliable broadband service but that meets BEAD’s minimum technical requirements (100/20 Mbps speeds, ≤100ms latency). The proposed guidance identifies unlicensed fixed wireless and LEO satellite services as examples of alternative technologies.

Before funding new deployments, eligible entities (i.e., states, territories, and the District of Columbia) must determine whether (1) there are enforceable commitments to deploy alternative technologies with ongoing performance monitoring, or (2) an existing alternative-technology provider can demonstrate that it currently meets the BEAD program requirements. NTIA’s proposed guidance requires the alternative-technology provider to demonstrate that it has the technical and operational capacity, as well as the financial and managerial capacity, to deliver service meeting BEAD’s technical requirements to all broadband serviceable locations (BSL) in the project area.

For LEO satellite projects, the guidance allows funding for the reservation of network capacity. This, however, comes with additional requirements, including a 10-year performance period and reimbursement based only on actual subscribers served. This differs from how other types of providers are treated, as they receive reimbursement based on areas of coverage regardless of the number of actual subscribers.

B. Summary of Our Recommendations

These comments advocate for a technology-neutral approach in the BEAD program guidance. The proposed hierarchy favoring fiber-optic deployment may inadvertently hinder the program’s goal of achieving universal broadband access efficiently and expeditiously. By embracing a technology-neutral stance that includes LEO satellite broadband as a viable alternative, the NTIA can foster competition, drive innovation, and maximize the impact of BEAD funding.

Economic analysis demonstrates that technological neutrality promotes market efficiency, allowing the most cost-effective solutions to emerge for diverse geographic and demographic contexts. To implement a technology-neutral framework, we recommend evaluating all proposals based on performance metrics, rather than which specific technologies are employed. Special restrictions on LEO satellite projects should be removed or revised, allowing for equitable treatment in funding structures and performance periods. Additionally, hybrid approaches that combine multiple technologies could yield optimal solutions for challenging deployment scenarios. While concerns about long-term viability and service quality are valid, these can be addressed through performance-based requirements and ongoing monitoring, rather than technology-specific requirements that effectively restrict the use of alternative technologies.

By adopting these recommendations, the NTIA could ensure that the BEAD program leverages the full spectrum of broadband technologies to achieve its goal of rapid and universal broadband access in the most efficient and effective manner possible.

II. Economic Analysis Demonstrates that Technological Neutrality Promotes Market Efficiency

In a technology-neutral environment, different broadband solutions (e.g., fiber, fixed wireless, LEO satellite) compete on their merits. This competition drives providers to innovate and improve their offerings, leading to better service and lower prices for consumers. An ICLE white paper on broadband competition released earlier this year finds that, relative to 2018, more U.S. households are now connected to the internet; broadband speeds have increased, while prices have fallen; more households are served by multiple providers; and new technologies like satellite and 5G have expanded internet access and intermodal competition among providers.[2] For example:

• Starlink’s LEO service was launched a little more than five years ago. Today, the service is available to all locations in the United States with speeds of between 25/5 Mbps and 220/25 Mbps.[3] Project Kuiper has successfully launched its first test satellites, with commercial service expected to begin in 2025.[4]
• In 2018, 5G fixed wireless first launched in the United States.[5] The technology now accounts for roughly 6% of U.S. internet connections and download speeds have nearly tripled.[6]

The fact that providers have invested billions of dollars in LEO satellite and 5G technologies, and that millions of consumers have adopted them, demonstrates that so-called “alternative technologies” already play an important role in a competitive broadband marketplace.

The cost structures and performance characteristics of these various technologies differ in ways that may make them more or less suitable for differing geographies, demographics, and use cases. For instance, fiber might be cost-effective in densely populated areas, while LEO satellite could be more efficient in remote, sparsely populated regions.[7] While 5G fixed wireless is often considered a potential solution for rural areas difficult to serve by fiber, 5G fixed wireless also accounts for 6% of urban connections.[8] That’s because the technology can be used to improve connectivity in public areas, such as parks, shopping malls, and transportation centers.[9] By allowing all technologies to compete, resources are more likely to be allocated to the most efficient solution for each specific context.

Technological neutrality encourages ongoing innovation. When policies are technology-specific, they can create lock-in effects that reduce the incentives to develop new technologies. In contrast, a neutral approach would maintain incentives for continuous improvement and disruptive innovations. While the NTIA claims that fiber is “future proof,”[10] other technologies make the same claim, such as Ericsson’s claim that 5G is future proof.[11] Regardless, in a little more than five years, the United States went from having no LEO satellite or 5G fixed-wireless broadband that was broadly available to having a vibrant broadband market in which both technologies are viable and competitive alternatives. Technological neutrality allows the market to adapt more quickly to changing economic conditions, technological advancements, and consumer preferences. This flexibility can lead to more resilient and sustainable broadband ecosystems.

Different technologies also may have differing economies of scale or scope. For example, LEO satellite networks might achieve greater economies of scale over large geographic areas, while fiber might have economies of scope in providing multiple services over the same infrastructure. A technology-neutral approach allows these efficiencies to be realized where appropriate.

By allowing these market mechanisms to operate, technological neutrality promotes an environment where the most cost-effective solutions can emerge to serve diverse consumers, technological constraints, and cost factors. A technology-neutral approach not only optimizes current-resource allocation, but also maintains the incentives for future innovations in broadband technologies, which can thereby lead to greater long-term economic benefits.

III. Case 2’s Unintended Consequences

The proposed guidance provides a three-case hierarchical approach to determining the need for BEAD funding for alternative technologies, identified as Case 1, Case 2, and Case 3. The case hierarchy is intended to ensure that BEAD funds are used efficiently and only where necessary. In concept, this approach should align with the economic principle of avoiding redundant investments and maximizing the effectiveness of public funds.

• Case 1 avoids unnecessary new investments where investment plans are already in place;
• Case 2 attempts to ensure that providers have the technical, operational, financial, and managerial capacity to deliver service, while also recognizing that some providers may already have the capability to meet BEAD requirements without additional funding; and
• Case 3 serves as a fallback option to ensure that all areas can be served, even if they do not fall under the first two cases.

On its face, these guidelines appear to be a reasonable attempt to ensure deployment, minimize redundancies, and avoid the problems the Federal Communications Commission (FCC) encountered with the Rural Digital Opportunity Fund, when it awarded funds to ISPs that subsequently defaulted on their obligations.[12] The proposed structure under Case 2 might, however, inadvertently create significant obstacles to achieving BEAD’s universal coverage goals. The proposed criteria—particularly the stringent requirements for alternative-technology providers to demonstrate their capacity—risk narrowing the competitive field and undermining the diversity of technological solutions that the program needs to fulfill its mission. By refining these criteria and reducing the uncertainties that providers face, the BEAD program can better support a competitive and innovative broadband landscape.

For example, the “requirement”[13] to demonstrate a network capacity of at least 5 Mbps or 2 TBs of monthly usage per BSL is particularly wrong-headed, as it may not align with the operational realities of technologies like LEO satellites and fixed wireless, which have different capacity dynamics than traditional fiber-optic networks. LEO satellites and fixed-wireless networks often operate on shared capacity models, where total network capacity is distributed among users dynamically. As such, these technologies may not guarantee a fixed capacity per-user in the same way that fiber-optic networks can.

This rigid requirement does not account for the varying performance characteristics across different broadband technologies. Alternative-technology providers, who have the most in-depth knowledge of their own networks and operational capacities, should be afforded the flexibility to determine what level of capacity is sufficient to comply with BEAD service standards and ongoing performance-measurement testing. Moreover, the capacity requirements are likely less relevant to users than cost or performance (e.g., speed and latency).[14]

While such a requirement may make sense for fiber technologies, it makes less sense for some alternative technologies. In areas with lower population density or less intense usage patterns, this high-capacity standard could impose unnecessary burdens, driving up costs and potentially making it financially unfeasible for certain providers to participate in the BEAD program. This could inadvertently reduce competition and stifle innovation, as only the largest and most resource-rich providers would be able to meet such demanding requirements.

By considering a lower capacity requirement that more accurately reflects the operational realities of different technologies, the BEAD program could encourage broader participation. This would foster greater innovation and competition, allowing for a more diverse array of solutions tailored to the specific needs of various geographic and demographic areas, ultimately leading to more effective and efficient broadband deployment across the country. By narrowing the range of available solutions, the proposed capacity requirement could make it difficult for states to consider alternative technologies where such options would be most effective.

While the NTIA understandably seeks to minimize redundancies, the capacity requirements likely will lead to overprovisioning, as not all users consume maximum capacity simultaneously.[15] By mandating capacity that may be idle much of the time, the proposed guidelines could result in inefficient uses of scarce resources. A better approach would be to require providers to demonstrate that they can dynamically allocate adequate capacity during peak usage periods.

The uncertainty and high barriers introduced by Case 2 not only might deter participation, but could also leave some regions without adequate broadband service. In areas where fiber deployment is impractical or exceedingly costly, alternative technologies like fixed wireless or LEO satellites offer potentially effective solutions. The combination of strenuous capacity requirements and a process that heavily favors fiber could, however, prevent these technologies from being implemented, leading to a scenario where certain geographic areas remain unserved or underserved despite the availability of viable alternatives. This outcome would directly conflict with the BEAD program’s mandate to ensure that all unserved and underserved locations receive broadband access.

To mitigate these risks, it is crucial that states establish objective and consistent criteria to evaluate alternative-technology providers under Case 2 that are based on technological and operational realities, rather than unrealistic ideals. This would require potentially revisiting the capacity requirements to better reflect the capabilities of various technologies. Such criteria would allow for a more transparent assessment process, ensuring that all technologies capable of meeting the overarching BEAD requirements reflected in the Infrastructure Investment and Jobs Act (IIJA) are given adequate consideration. By adopting the performance-based approach envisioned in the statute, states could better align the BEAD program with its goal of universal coverage, enabling the deployment of the most effective solutions tailored to the specific needs of different regions.

IV. Mismatch Between LEO Satellite Technical Requirements and Proposed Reimbursement Policy

The guidance’s proposed reimbursement structure would introduce significant challenges for LEO satellite providers, potentially jeopardizing their participation in the BEAD program. NTIA’s guidance instructs states to reimburse providers based only on the number of actual subscribers within a project area during a specific period, rather than on the mandated capacity reserved for potential customers. Technologies like LEO satellites, however, operate very differently from terrestrial technologies like fiber, and these differences must be recognized in the reimbursement model.

A subscriber-based reimbursement requirement for fiber does not impose additional costs beyond the initial capital expenditure of laying dark fiber. Once a customer decides to subscribe, the provider has a reasonable amount of time to establish the last-mile connection. This process involves predictable operational expenses within a defined and constrained footprint.

In contrast, the operational requirements for a LEO constellation are fundamentally different. Demanding a subscriber-based reimbursement model for LEO providers is akin to requiring fiber providers to wire every house in their project footprint, provide functional Wi-Fi to those homes, but merely withholding the password to access the internet until a subscription is confirmed. Just as such an approach would drastically increase the costs of providing fiber, the recommended guidance similarly inflates the costs for LEO providers, who must maintain constant capacity across their entire service area regardless of the actual number of subscribers. The infrastructure and capacity needs for LEO are inherently more expansive and less predictable, given the wide area that satellites cover and the dynamic nature of space-based networks.

The NTIA’s proposed approach would impose an undue burden on LEO-satellite providers, who must make substantial upfront capital investments to ensure sufficient capacity is available, even if subscription levels are unpredictable and unstable in the early stages. By limiting cost recovery to actual subscribers, the guidance not only conflates broadband adoption with access—potentially leaving areas deemed “unserved” until a subscription occurs—but also risks harming LEO providers’ ability to serve other customers by requiring them to reserve unused capacity for extended periods. The proposed reimbursement policy might discourage providers from serving areas with initially low adoption rates, even if these areas have strong potential for future growth.

Moreover, discouraging LEO participation could have long-term consequences, particularly in hard-to-reach areas where fiber is infeasible, which would run counter to the BEAD program’s goal of universal broadband access. The NTIA’s one-size-fits-all reimbursement model may inadvertently exclude certain technologies de facto, raising concerns about whether the BEAD program is truly technology-neutral. Ensuring that the program supports a broad array of technologies is crucial to achieve its stated goals and avoid unintended inequities in broadband deployment.

V. Recommendations

The recommendations we offer in these comments are grounded in the economic principles of efficient resource allocation and technological neutrality. They aim to support a framework that promotes competition and innovation, allows for tailored solutions to diverse geographic and demographic contexts, and maximizes the effectiveness of BEAD funding by focusing on outcomes, rather than means. In adopting these recommendations, the NTIA could create a more flexible, efficient, and effective framework to achieve BEAD’s goal of universal broadband access.

A. Eliminate the Proposed Technological Hierarchy

The NTIA’s proposed guidance establishes a clear hierarchy that favors fiber-optic deployments, followed by other “reliable broadband service” technologies, with alternative technologies like LEO satellites and unlicensed fixed wireless only considered when other options exceed certain cost thresholds. While well-intentioned, this proposed hierarchy inappropriately presumes that one technology (i.e., fiber) is universally superior, which may not be true in all contexts. As such, the proposed guidance potentially excludes innovative solutions that could be more efficient in certain scenarios. Following the guidance might therefore lead to greater overall costs by encouraging the use of more expensive technologies in areas where alternatives could provide equivalent service at lower cost.

Instead, we recommend a technology-neutral approach that evaluates all proposals based on performance metrics. These metrics could include speed (download and upload); latency; reliability (uptime, consistency of service); scalability (ability to increase capacity over time); time to deployment; and cost-effectiveness (both initial deployment and ongoing operation). By focusing on these performance metrics, rather than specific technologies, the NTIA could ensure that the most effective solution is chosen for each unique deployment scenario, maximizing the effect of BEAD funding.

B. Remove Special Restrictions on LEO Satellite Projects

To resolve the mismatch between the LEO satellite technical requirements and proposed reimbursement policy, we recommend adjustments that more closely align the two policies.

We advise NTIA to replace the fixed-capacity requirements with performance-based metrics (e.g., speed, latency, reliability) that are more relevant to user experience and can be tailored to the capabilities of LEO technologies. If, however, NTIA insists on mandating capacity requirements, we advise that, instead of mandating 5 Mbps or 2 TBs per BSL, the agency should adopt a more flexible approach that considers the shared-capacity model of LEO networks. One way would be to require providers to demonstrate sufficient overall network capacity and ability to dynamically meet peak demand in the service area, rather than mandating a fixed per-BSL minimum allocation. These revisions would better reflect the technical realities of LEO satellite broadband vis-à-vis fiber.

Below, we offer several suggestions to improve the reimbursement policy. Any one of these suggestions would better align incentives and encourage LEO satellite participation in the BEAD program:

• Allow LEO satellite providers to recover costs, or at least a portion of them, for locations where they can connect a customer within 10 business days of a service request. This would ensure that the BEAD program fosters a competitive landscape that includes diverse technological solutions, rather than unduly favoring certain technologies over others.
• Implement a “ramp-up” or “grace” period at the beginning of the project where providers are reimbursed at a higher rate to account for lower initial adoption, while maintaining necessary infrastructure.
• Instead of reimbursing only for actual subscribers, consider a model that compensates for a percentage of total BSLs in the project area, gradually increasing as adoption rates grow.
• Use the reimbursement policy as an incentive to satisfy the Case 2 capacity objectives. For example, providers meeting the capacity objectives would be reimbursed based on all BSLs in the project area, but those that do not meet the objectives would be reimbursed based on actual subscribers. This approach would provide higher payments to providers who invest in adding capacity and provide a financial incentive to encourage providers to meet the NTIA’s capacity objective. Such an approach would better align the reimbursement model with the operational realities of different broadband solutions.

C. Consider Hybrid Approaches

In many challenging deployment scenarios, a combination of technologies may provide the most effective solution. For example, LEO satellites could provide rapid initial coverage to an area, while fiber is gradually deployed for long-term capacity. In addition, a mix of fiber backhaul and fixed-wireless last-mile connections might be optimal in areas with dispersed populations.[16] Such hybrid approaches, integrating multiple technologies, could provide redundancy and improved reliability. By explicitly encouraging consideration of hybrid approaches, the NTIA could promote creative solutions that leverage the strengths of different technologies to overcome deployment challenges.

VI. Conclusion

We applaud the NTIA for seeking to issue guidance that would allow more participation for alternative technologies in the BEAD program. Simply put, building out broadband to hard-to-reach areas with fiber alone is not possible with BEAD funding.[17] The NTIA should embrace a technology-neutral approach to maximize the value of these limited resources.

[1] Proposed BEAD Alternative Broadband Technology Guidance, NTIA (Aug. 26, 2024), available at https://www.ntia.gov/sites/default/files/publications/bead-alternative-broadband-technology-policy-notice-for-public-comment-final.pdf.

[2] Eric Fruits, Geoffrey A. Manne, Ben Sperry, & Kristian Stout, Dynamic Competition in Broadband Markets: A 2024 Update (ICLE White Paper 2024-06-04), available at https://laweconcenter.org/wp-content/uploads/2024/06/Broadband-Competition-2024-Update.pdf.

[3] Dan Heming, Starlink No Longer Has a Waitlist for Standard Service, and 10 MPH Speed Enforcement Update, Mobile Internet Resource Center (Oct. 3, 2023), https://www.rvmobileinternet.com/starlink-no-longer-has-a-waitlist-for-standard-service-and-10-mph-speed-enforcement-update; Starlink Specifications, Starlink, https://www.starlink.com/legal/documents/DOC-1400-28829-70 (last accessed Sep. 5, 2024).

[4] Everything You Need to Know About Project Kuiper, Amazon’s Satellite Broadband Network, About Amazon (Jun. 17, 2024), https://www.aboutamazon.com/news/innovation-at-amazon/what-is-amazon-project-kuiper.

[5] Robert Wyrzykowski, 5G Fixed Wireless Access (FWA) Success in the US: A Roadmap for Broadband Success Elsewhere?, OpenSignal (Jun. 6, 2024), https://www.opensignal.com/2024/06/06/5g-fixed-wireless-access-fwa-success-in-the-us-a-roadmap-for-broadband-success-elsewhere.

[6] Id.

[7] See Colby Leigh Rachfal, Low Earth Orbit Satellites: Potential to Address the Broadband Digital Divide, Cong. Research Serv. R46896 (Aug. 31, 2021), https://crsreports.congress.gov/product/pdf/R/R46896.

[8] Id.

[9] 5G Fixed Wireless Access Market, Global Market Insights (Feb. 29, 2024), https://www.gminsights.com/industry-analysis/5g-fixed-wireless-access-market.

[10] Evan Feinman, Choosing the Right Mix of Technologies to Achieve Internet for All, NTIA (Aug. 26, 2024), https://www.ntia.gov/blog/2024/choosing-right-mix-technologies-achieve-internet-all?source=email.

[11] Ericsson, Insight 6 of 6: 5G Offers a Future-Proof Platform for FWA Growth (Dec. 2023), available at https://www.ericsson.com/4ade15/assets/local/reports-papers/further-insights/doc/fwa_insights_6_offers_extracted.pdf.

[12] See Julia King, FCC Faces Pressure to Pardon RDOF Defaults, Fierce Network (Mar. 6, 2024), https://www.fierce-network.com/broadband/fcc-faces-pressure-pardon-rdof-defaults.

[13] It is worth noting here that this “requirement” is contained nowhere in the enabling statute for BEAD. See 47 U.S.C. §1702.

[14] See, Broadband Basics: How it Works, Why It’s Important, and What Comes Next, Pew Trusts (Aug. 18, 2023), https://www.pewtrusts.org/en/research-and-analysis/fact-sheets/2023/08/broadband-basics-how-it-works-why-its-important-and-what-comes-next (reporting that the main reason people do not subscribe to broadband is because the monthly cost is too expensive).

[15] See Jacob B. Malone, Aviv Nevo, & Jonathan W. Williams, The Tragedy of the Last Mile: Economic Solutions to Congestion in Broadband Networks (May 30, 2021), available at https://jonwms.web.unc.edu/wp-content/uploads/sites/10989/2021/06/Congestion_WP-2021.pdf (showing data usage is lowest around 4 a.m., increases throughout the day, and peaks around 10 p.m.).

[16] Backhaul is also known as the “middle mile,” and is the cost of transmitting information from a provider’s aggregation points to the internet “backbone.” These costs are higher for rural areas and can adversely affect the deployment of broadband to these areas. See generally U.S. Gov’t Accountability Off., Telecommunications: Broadband Deployment Is Extensive Throughout the United States, but It Is Difficult to Assess the Extent of Deployment Gaps in Rural Areas, GAO-06-426 (May 2006), available at https://www.gao.gov/assets/gao-06-426.pdf.

[17] See Broadband Funding Optimization Tool, Vernonburg Group (last accessed Sep. 5, 2024), https://www.vernonburggroup.com/broadband-funding-optimization-tool.

I. Introduction

We thank the National Institute of Standards and Technology (NIST) for the opportunity to comment on the initial public draft of AI 800-1, “Managing Misuse Risk for Dual-Use Foundation Models” (“Draft”). NIST’s efforts to address the potential risks associated with advanced artificial-intelligence (AI) systems are commendable, but several aspects of the proposed approach raise significant areas of concern.

Those concerns are outlined in these comments, which focus on what we believe to be the draft’s overly restrictive risk-assessment model and problematic dual-use classification, as well as the potential for negative effects on open-source development. If implemented as proposed, the framework could stifle innovation, limit the economic benefits of AI, and potentially undermine U.S. leadership in AI development.

NIST should instead pursue a harm-focused approach based on the concept of “regulation as a discovery process,” which would more effectively balance the need for responsible AI development with the imperative to foster innovation. Our recommendations in this area are designed to contribute to developing a more nuanced, flexible, and effective regulatory framework for AI that aligns with the rapidly evolving nature of these transformative technologies.

II. Regulation of ‘Dual-Use Foundation Models’

A. Overly Restrictive Risk Assessment

One critical issue with the draft is the approach it takes to risk and liability assessment. The document appears to recommend a broad liability framework for foundation-model developers without adequate consideration of the nuances that differentiate various use cases, or developers’ ability to control downstream usage.[1] When considering liability for AI systems, it is useful to draw parallels with existing legal frameworks, such as the “least-cost avoider” principle in common law.[2]

The law & economics concept of the “least-cost avoider” refers to the party in a conflict who can most efficiently reduce the probability of a costly interaction.[3] In the context of AI, this principle suggests that liability should be assigned to the entity or entities that can most effectively and economically mitigate potential harms.

Applying this principle judiciously to AI systems will encourage developers and deployers to take reasonable steps to protect both their own users and potential non-users who might be affected by the AI models in question. It should be emphasized that this approach doesn’t necessarily mean holding any particular AI developer liable for all harmful outcomes of their models or products. Rather, it militates for imposing a duty of care, as appropriately dictated by the circumstances, that would require a developer to take those reasonable measures necessary to mitigate foreseeable risks, given the degree of control they have over users of their products.

Any liability framework should be sufficiently flexible to account for the diverse instantiations of AI technologies. It should also consider the specific context of each use case, the degree of control each party has over the AI system, and the practical ability that different relevant parties possess to prevent harm.[4]

While there may be instances where it’s appropriate for a product developer to bear presumptive liability, policymakers should remain skeptical of the wisdom of imposing such liability as an ex ante rule to all foundation models. For regulators to assign liability in this way would run the risk of creating a strict liability regime, which appears particularly ill-suited for vast swathes of the AI ecosystem.

Ideally, courts should be allowed to parse the specifics of particular cases, and thereby to shape the contours of liability and relevant duties of care. This would allow for more balanced consideration of the social benefits that AI platforms provide (e.g., innovation, expression, and commerce) as weighed against the need to limit the social costs of unlawful or harmful AI conduct. Short of that, regulatory intervention in the assignment of liability should follow the same principles as do courts at common law: proceed case-by-case and remain sensitive to the full context of particular developers, deployers, products, and consumers.

Furthermore, the draft fails to adequately account for the diverse nature of AI technologies and the various commercialization models employed by the emerging AI industry. As noted by Jason Potts, “AI is not a single technology, or even industry, and is composed of extremely complex and varied ownership and governance at each composite layer.”[5] This complexity makes it challenging to craft a one-size-fits-all regulatory approach, much less to assume that a broad category of developers is always best-positioned to assume liability for AI development.

The document’s narrow focus on initial developers overlooks the reality that many components within the AI stack are often developed via open-source processes or in a modular fashion, and are subsequently assembled by firms using business models that may vary along the spectrum of relatively open to relatively closed. For instance, while some aspects of AI development—such as model training in advanced data centers—may tend to be centralized due to efficiency considerations, other components are developed in a more distributed manner, such as low-level AI tools[6] or products that incorporate trained models.[7]

By placing undue burden on foundation-model developers and, as discussed below, potentially restricting open-source development, the draft risks stifling innovation and limiting the broader economic benefits of AI in the United States. The lack of a vibrant U.S.-based foundation-model ecosystem could undermine U.S. global tech leadership, as developers worldwide turn to foundation models developed in other countries—including those that lack some of the United States’ basic legal norms.

B. The Problem with ‘Dual-Use’ Classification

The draft relies heavily on the concept of “dual-use foundation models.”[8] However, this classification raises significant issues that could potentially lead to stifled innovation in the AI field without generating concomitant consumer benefits from regulation.

Even applying the term “dual use” to AI technologies raises significant issues. Virtually any technological tool can be considered “dual use,” in the sense that it has both benign and potentially harmful applications. This renders the term nearly meaningless as a basis for targeted regulation. Further, the draft’s definition of dual-use foundation models creates, at least implicitly, a presumption of harmfulness. This presumption is not only unwarranted, but could also lead to unnecessary restrictions on beneficial AI development and deployment.

Moreover, the draft is unclear about how to prioritize potential risks from “dual-use” technologies, which could push developers to make binary determinations (either “harmful” or “not harmful”) when considering new products and improvements. The draft offers examples of risks that it posits may be representative, such as “the risks that models will facilitate the development of chemical, biological, radiological, or nuclear weapons; enable offensive cyber attacks; aid deception and obfuscation; and generate child sexual abuse material (CSAM) and non-consensual intimate imagery (NCII) of real individuals.”[9] But it remains unclear whether this list is meant to be comprehensive or if foundation-model developers are also responsible for analyzing other, as-yet undefined risks. Further, the risk of “deception” can mean very different things in different contexts. Some may call political parody “deceptive” misinformation, while others may regard it as merely entertaining. The draft should be amended to clarify these issues.

More broadly, the draft’s dual-use classification opens what is potentially a giant definitional loophole that threatens to subsume virtually any useful AI system. The document employs expansive definitions to describe a generalized threat—specifically that dual-use models could “permit the evasion of human control or oversight through means of deception or obfuscation.”[10] But such language could be said to apply to a broad array of general-purpose AI technologies. All software—particularly highly automated software—may in some circumstances pose serious risks to any of a host of other concerns that we wish to protect. Encryption and other privacy-protecting tools certainly fit this definition.[11] While it is crucial to mitigate harms associated with the misuse of AI technologies, the blanket treatment of all technologies under this category raises the risk of stifling overregulation.

Rather than rely on broad and potentially misleading classifications assigned to potential dual-use capabilities, a more effective approach would be to focus on specific and demonstrable harms. This would involve developing a more nuanced framework to assess AI technologies—one that considers the specific application domain and context of use.[12] It’s crucial to clarify and prioritize specific risks that require regulatory attention, rather than using catch-all categories. Many technologies pose the potential for misuse, but this potential alone should not be the sole basis for restrictive regulation.

C. Impacts on Open-Source Development

The safeguards and restrictions proposed by the draft could also significantly undermine the open-source AI ecosystem. While the draft’s recommendations are intended to mitigate risk, they fail to adequately consider the unintended consequences that these measures could have on open-source innovation.

Open-source software development is a vital component of the broader technological ecosystem, and this is especially true in the realm of AI. Many components of the AI stack are developed through open-source processes or in a modular fashion, and are assembled by firms using business models that vary in the degree to which they are relatively open or relatively closed.[13] This underscores the difficulty of clearly assigning liability to “foundation models” or to any of the various elements of the complex, multi-layered AI ecosystem.

At the foundation of that ecosystem is the hardware layer, which includes semiconductors and raw computing power, alongside “XaaS” (everything-as-a-service) providers that offer virtual access to storage, processing, and software resources.[14] The next layer is data, which may be structured (organized in databases) or unstructured (text, images, videos).[15] The quality and quantity of these data are fundamental to an AI system’s performance. The data layer involves processes for collection, preparation (cleaning, transforming, and labeling), and curation. Next is the model-training layer, where various techniques—such as supervised learning, unsupervised learning, reinforcement learning, and transfer learning—are employed to create AI models that can perform specific tasks.[16]

The deployment layer represents the final stage, where trained models are put into operation. This can occur in cloud environments, for scalability; on edge devices, for reduced latency; or on-premises, for enhanced data control. Each deployment method serves different user needs, and is often managed by specialized firms with distinct business models. This layered structure of the AI stack underscores the diversity among AI technologies and the challenges that inevitably attend defining broad product markets around these heterogeneous components.

In short, while the draft focuses primarily on foundation models, in reality, there are many distinct pieces that constitute the creation and training of AI models. Moreover, this diversity and flexibility in development approaches has been instrumental in driving rapid advancements in AI capabilities and applications.

The draft’s recommended safeguards could, however, severely restrict this open-development model. For instance, the document suggests implementing measures to prevent model theft and limit access to model weights.[17] While these measures might be appropriate in some contexts, their blanket application could effectively shut down many open-source AI projects, where such requirements don’t make sense given the distributed nature of their development. In particular, these measures presume a proprietary model of development that can be secured at a centralized source. This runs completely contrary to the nature of many (if not all) open-source development projects, which rely on decentralized control, widely distributed code, and data sharing. The ability to freely access, modify, and distribute model weights is often fundamental to open-source AI development. Restricting this ability could stifle innovation and collaboration across the field.

Furthermore, such restrictions would pose a direct and fundamental threat to the development of open-source foundation models. By making it more difficult or risky to develop and share open-source models, the draft’s recommendations could inadvertently push the field toward more closed and proprietary models. This shift would not only limit the diversity of approaches in AI development, but could also concentrate power in the hands of a few large companies with the resources to develop and maintain proprietary models.

Another significant concern is the potential that such restrictions could limit the broader economic benefits of AI in the United States.[18] Open-source technologies have been a driving force behind the rapid adoption and integration of AI across various sectors of the economy. By potentially restricting open-source development, the draft could slow this adoption process, limiting the economic gains and productivity improvements that generative AI promises.[19]

Moreover, the document’s approach appears to overlook the self-regulatory mechanisms that often emerge within open-source communities. These communities frequently develop their own best practices and ethical guidelines, which can be more adaptive and nuanced than top-down regulatory approaches.[20] By potentially sidelining these community-driven efforts, following the draft’s recommendations might actually serve to reduce overall safety and responsibility in AI development, rather than enhance it.

The draft’s proposed restrictions could also have far-reaching implications for AI research and education. Open-source models and tools are vital resources for academic researchers and students, providing accessible platforms for learning, experimentation, and innovation.[21] Limiting access to these resources could create a significant barrier to entry for new talent in the field, potentially slowing the pace of AI advancement and reducing the diversity of voices contributing to its development.

III. The Need for Context-Specific Regulation

As we note below, rather than the draft’s current approach, what is needed is a more flexible and adaptive governance framework that would allow industry best practices to evolve organically in response to technological change and consumer demands.[22] Policymakers should embrace a dynamic and context-specific framework that focuses on addressing tangible harms while leaving room for experimentation and innovation.

Crucial to this approach is that AI regulation must not disproportionately focus on mitigating risks without giving sufficient consideration to the immense benefits that AI technologies might also yield. As many of AI’s potential risks remain largely hypothetical, regulatory proposals can quickly become unmoored from the real world and how the technology is actually used. This problem is exacerbated by the nature of risk perception. As Aaron and Adam Wildavsky have observed, familiarity with known hazards does not necessarily determine whether or to what degree an individual or organization will perceive a given technology as safe or dangerous.[23] This holds true not only for laypeople, but also for experts in risk assessment.

Ideally, AI regulation should be focused on empirically observed harms, but a risk-based framework that is limited in scope to better match anticipated real-world harms would the next-best option. The approach recommended by  the National Telecommunications and Information Administration (NTIA) in its recent report “Dual Use Foundation Models with Widely Available Model Weights” offers such a framework to analyze marginal risks.[24]

Following extensive stakeholder outreach,[25] NTIA developed its framework to provide a more nuanced and balanced assessment of the potential impact of open-foundation models. NTIA defines marginal risks and benefits as “the additional risks and benefits that widely available model weights introduce compared to those that come from non-open foundation models or from other technologies more generally.”[26] This framework was adopted to avoid targeting dual-use foundation models with restrictions that are unduly stricter than alternative systems that pose a similar balance of benefits and risks.[27] By focusing on the marginal differences between open and closed foundation models (and between AI and non-AI products), NTIA advocates for rooting analysis of potential risks in empirical reality, as opposed to speculation.

The NTIA report outlines three key conditions to assess marginal risks and benefits:

1. There must be a difference in the magnitude of risks/benefits between open and closed models;
2. The risks/benefits must be greater for dual-use foundation models than for non-AI technologies or other AI models; and
3. The risks/benefits must arise from future models over and above those already generally released.[28]

The benefits of this approach are that it focuses on the measurable differences among both AI and non-AI technologies, as well as the relevant differences between AI technologies with open weights and those with closed weights. That is, the NTIA framework focuses on the empirically observable facts that actually affect adoption, use, and potential harms arising from AI use, rather than developing ex-ante regulatory frameworks based on speculation. It also acknowledges the challenges inherent in assessing such risks, noting that “[r]isks and benefits that satisfy all three conditions are difficult to assess based on current evidence,”[29] and thus acknowledging the need to proceed cautiously.

Indeed, this approach is eminently sensible, insofar as we don’t yet even have a clear taxonomy of what it means to regulate “AI.” The diversity of AI technologies has profound implications for regulation and development. For example, leaving aside the glitz around newly popular LLMs and their ability to plausibly pass the Turing test, there remain big questions about the nature of the harms expected from such a technology. Is it the system’s autonomous behavior that is the primary concern, as was the case for high-frequency trading and the flash crash?[30] Or is it the possibility that bad actors will have new tools that help them to break existing criminal laws? Or is it some mixture of the two?

This isn’t merely an academic question. The application domain for practical uses of AI technology requires nuance, as different regulatory considerations are relevant for AI systems designed for autonomous vehicles versus those used in financial algorithms, creative content generation, autonomous weapons systems, or predictive policing. An approach focused on assessing marginal risks is broadly compatible with the diverse nature of AI technologies and products in a way that an ex-ante risk framework is not.

Ideally, NIST should advocate for a harm-based approach that roots regulation in empirically observable harms that actually exist. However, short of that, NIST should consider a marginal-risk framework along the lines of what NTIA has adopted. This would be valuable not merely because such an approach preserves space for innovation in AI, while also focusing on actual harms. It also would help regulatory agencies focus on their actual missions. A harm-based or marginal-risk framework would enable more targeted and efficient allocations of resources in the face of a technology that is incredibly diverse both in its core technologies, its current product offerings, and the innumerable use cases that individuals will adopt.

IV. Regulation as a Discovery Process

In the rapidly evolving field of AI, traditional methods of regulation are likely to prove ineffective. Instead, regulators should adapt their mission and view regulation as a “discovery” process, an approach that is particularly well-suited to the dynamic nature of AI technologies. This perspective offers a more flexible and adaptive framework for governing AI development and deployment.

As proposed by Geoffrey Manne and Gus Hurwitz, “regulation as a discovery process” counsels for treating regulation not merely as a mechanism to decree and enforce rules, but as a process to discover information that can inform and improve regulatory approaches over time.[31] This perspective is particularly pertinent to AI, where the pace of innovation and the complexity of technologies often outstrip regulators’ understanding and ability to predict future developments.

At its core, this approach asks regulators to consider that they might be wrong—that they might be asking the wrong questions, collecting the wrong information, or analyzing it incorrectly.[32] It acknowledges that there is no amount of information collection or analysis that is guaranteed to be “enough,” especially when dealing with complex, dynamic industries. This epistemic humility is crucial in a field where our understanding of the technology’s capabilities and implications is constantly evolving.

Adopting regulation as a discovery process means shifting how regulatory agencies conceptualize their mission. Instead of focusing solely on writing and enforcing regulations, agencies should see themselves as assisting lawmakers to assemble, filter, and focus on the most relevant and pressing information needed to understand the market’s changing dynamics. This requires setting up ongoing mechanisms to gather and report data, such as regulatory sandboxes,[33] and directing the process toward specifications for how information should be used prior to its collection (i.e., how it feeds back into the regulatory process).

This approach stands in contrast to the draft’s treatment of transparency and reporting requirements, as outlined in Objective 7.[34] While NIST recognizes the importance of transparency, its approach seems to view transparency primarily as a means of ensuring compliance with predetermined rules. Regulation as a discovery process, on the other hand, sees transparency and reporting as part of ongoing learning, where the information gathered actively shapes and refines the regulatory approach.

Embracing regulation as a discovery process is particularly crucial in the context of AI, given the limits of our collective knowledge about the technology’s potential risks and benefits. It underscores why regulators should prioritize generating and using new information through regulatory experiments, iterative rulemaking, and feedback loops. A more adaptive regulatory framework could respond to new developments and insights in AI technologies, thereby ensuring that regulations remain relevant and effective without stifling innovation.

Moreover, this approach emphasizes the importance of considering regulation as an information–producing activity. In the context of AI regulation, this extends beyond mere data collection. It suggests that any regulatory mechanism should incorporate processes to ensure that information is not only generated as part of the regulatory process, but also fed back into it. This cyclical approach to information flow is crucial to maintain relevance and effectiveness in the rapidly evolving field of AI.

Adopting regulation as a discovery process would offer policymakers several avenues to create a more dynamic and informed regulatory environment for AI.

First, it would mitigate the issue of overly restrictive risk assessment by embracing a more flexible and adaptive framework. Instead of imposing rigid, predetermined rules based on speculative risks, this approach allows for ongoing learning and adjustment as real-world evidence emerges. This aligns with the NTIA’s marginal-risk framework, which focuses on actual and measurable differences in risks and benefits, rather than hypothetical scenarios.

Second, it addresses the problems associated with the broad dual-use classification. By viewing regulation as an information-gathering and learning process, we can move away from blanket categorizations and toward a more nuanced understanding of AI technologies and their applications. This approach also allows for context-specific assessments, recognizing that the risks and benefits of AI systems can vary greatly depending on their specific use cases and deployment contexts.

Third, the discovery-process approach is more conducive to fostering open-source development. By emphasizing ongoing dialogue and collaboration among regulators, developers, and other stakeholders, this approach can help to balance safety concerns with the need for innovation. It also allows for the development of more adaptive and nuanced guidelines that account for the idiosyncratic nature of different approaches to software development and deployment, while still addressing legitimate safety concerns.

Moreover, this approach’s emphasis on information production and feedback loops can help regulators stay abreast of rapid technological developments. This is crucial to address the concern that rigid regulations might stifle innovation or quickly become obsolete in the fast-moving field of AI.

By adopting regulation as a discovery process, we can create a regulatory environment that is more responsive to the actual state of AI technology, more attuned to real-world impacts, and more supportive of beneficial innovation. This approach provides a framework to develop policies that are empirically grounded, adaptable, and better equipped to balance the immense potential of AI with responsible risk management.

V. Conclusion and Recommendations

The draft’s focus on “dual-use” models, its implicit presumption of harm, and its recommended restrictions that would make open-source development difficult or impossible all would likely yield more harms than benefits. We advise that NIST pursue an alternate approach to regulating foundation models that aims to strike a more appropriate balance between managing risks and fostering innovation.

The preferred regulatory framework would be adaptable, context-specific, and would focus on addressing tangible harms, rather than speculative risks. Moreover, it would not presume that any given type of model is harmful (as is implicitly the case when regulating models as “dual use”).  The rapidly evolving nature of AI technology necessitates a dynamic approach to regulation—one that can keep pace with technological advancements and respond to emerging challenges in real time.

A key element of this proposed approach is to focus on tangible and demonstrable harms, rather than potential risks. Ideally, this would be a harm-based approach, grounded in legal traditions that appreciate the nuances of relative risk, with calculations of actual harms derived from longstanding legal principles and empirical realities. By prioritizing the mitigation of real, observed harms over speculative risks, regulators can ensure that their efforts are targeted and effective, without unduly constraining beneficial innovation.

Short of that, NIST would do well to incorporate a marginal-risk framework into its proposed regulatory approach. Such a framework would assess the additional risks and benefits that widely available model weights introduce, relative to those from non-open foundation models or other non-AI technologies. By adopting this disposition, regulators would avoid imposing unnecessarily strict restrictions on open-weight models that may not pose significantly greater risks than their closed counterparts or existing technologies.

This alternate approach would also place a strong emphasis on preserving and promoting open-source AI development. Open-source models have played a crucial role in democratizing AI technology, fostering innovation, and accelerating research. Any regulatory framework should recognize this value and avoid measures that would unduly restrict or discourage open-source development. Instead, regulators should work collaboratively with the open-source community to develop guidelines and best practices that address legitimate safety concerns, while preserving the benefits of openness.

Whatever risk- or harm-based framework is adopted, NIST should encourage the use of discovery processes in AI regulation. In practical terms, this approach would involve a more flexible and iterative regulatory process. Rather than imposing rigid rules ex ante, regulators would engage in ongoing monitoring and assessment of AI developments, adjusting their approach as new information becomes available. Moreover, this alternative approach would emphasize the importance of multi-stakeholder collaboration in developing and implementing AI governance frameworks. This would involve ongoing dialogue among regulators, industry experts, researchers, and civil-society organizations to ensure that regulatory measures are informed by diverse perspectives and the latest technological insights.

By adopting this more balanced, dynamic, and context-specific approach to AI regulation, policymakers can better ensure an environment that fosters responsible innovation while effectively managing genuine risks. This approach would not only be more conducive to technological progress, but would also be more resilient in the face of rapid advancements in AI technology. It would allow the United States to maintain its leadership in AI development, while ensuring that these powerful technologies are deployed in ways that align with social values and priorities.

Ultimately, the goal of this alternate approach is to create a regulatory framework that is as innovative and adaptive as the technology it seeks to oversee. By focusing on tangible harms, embracing context specificity, and recognizing the value of open-source development, this approach can help to unlock the immense potential of AI while responsibly managing its risks.

[1] See, Managing Misuse Risk for Dual-Use Foundation Models, Nat’l Inst. of Standards and Tech., available at https://nvlpubs.nist.gov/nistpubs/ai/NIST.AI.800-1.ipd.pdf (last visited Sep. 5, 2024) (“This document focuses on misuse risk from dual-use foundation models… [which] includes foundation models that exhibit, or could be easily modified to exhibit, high levels of performance at tasks that can pose a serious risk to security, economic security, public health or safety, or any combination of those.”) [hereinafter “Draft”].

[2] See Geoffrey A. Manne et al., Who Moderates the Moderators?: A Law & Economics Approach to Holding Online Platforms Accountable Without Destroying the Internet, Int’l Cent. L. & Econ. (Nov. 9, 2021), https://laweconcenter.org/resources/who-moderates-the-moderators-a-law-economics-approach-to-holding-online-platforms-accountable-without-destroying-the-internet.

[3] See Harold Demsetz, When Does the Rule of Liability Matter?, 1 J. Leg. Stud. 13, 28 (1972); see generally Ronald Coase, The Problem of Social Cost, 3 J. L. & Econ. 1 (1960).

[4] See e.g., Manne et al., supra note 2, at 36-47.

[5] Jason Potts, Jason Potts: “Sources of Innovation in Generative AI”, Nat’l L. Rev. (Feb. 21, 2024), https://www.networklawreview.org/jason-potts-generative-ai.

[6] See, e.g., Learn the Basics, PyTorch, https://pytorch.org/tutorials/beginner/basics/intro.html (last visited Sep. 6, 2024); Tim Mucci, Five Open-source AI Tools to Know, Int’l. Bus. Mach. Corp. (Dec. 15, 2023), https://www.ibm.com/blog/five-open-source-ai-tools-to-know.

[7] See, e.g., Qualcomm Enables Meta Llama 3 to Run on Devices Powered by Snapdragon, Qualcomm (Apr. 18, 2024), https://www.qualcomm.com/news/releases/2024/04/qualcomm-enables-meta-llama-3-to-run-on-devices-powered-by-snapd; Clip Studio Paint to Include Experimental Image Generator Palette in Winter Update, Clip Studio Paint,  https://www.clipstudio.net/en/news/202211/29_01 (last visited Dec. 02, 2022).

[8] Id. at 1.  

 

 

[9] See, Managing Misuse Risk for Dual-Use Foundation Models, supra note 1, at 1.

[10] See Id. at 18.

[11] Encryption and the “Going Dark” Debate, Cong. Rsch. Serv., https://crsreports.congress.gov/product/pdf/R/R44481 (last updated Jan. 25, 2017).

[12] See discussion, infra, at nn. 29-32 and accompanying text.

[13] Alex Engler, How Open-Source Software Shapes AI Policy, Brookings Inst. (Aug. 10, 2021), https://www.brookings.edu/articles/how-open-source-software-shapes-ai-policy.

[14] Romit Dey & George Korizis, How Anything-As-A-Service (XaaS) Can Help Reinvent Business Models and Transform Outcomes Across Industries, Price Waterhouse and Coopers & Lybrand, https://www.pwc.com/us/en/services/consulting/business-transformation/library/use-xaas-to-reinvent-business-models.html (last visited Sep. 6, 2024).

[15] See, e.g., Structured vs Unstructured Data, IBM (Jun. 29, 2021), https://www.ibm.com/think/topics/structured-vs-unstructured-data; Dongdong Zhang et al., Combining Structured and Unstructured Data for Predictive Models: A Deep Learning Approach, BMC Med. Informatics Dec. Making 280 (2020), https://link.springer.com/article/10.1186/s12911-020-01297-6 (describing generally the use of both structured and unstructured data in predictive models for health care).

[16] Anil Ananthaswamy, Why Machines Learn: The Elegant Math Behind Modern Ai 12-13 (2024).

[17] Draft at 8-9.

[18] Id. at 7. 

[19] See, e.g., Michael Chui, et al., The Economic Potential Of Generative AI: The Next Productivity Frontier, McKinsey Digital (2023), https://www.mckinsey.com/capabilities/mckinsey-digital/our-insights/the-economic-potential-of-generative-ai-the-next-productivity-frontier (“Generative AI’s impact on productivity could add trillions of dollars in value to the global economy. Our latest research estimates that generative AI could add the equivalent of $2.6 trillion to $4.4 trillion annually across the 63 use cases we analyzed—by comparison, the United Kingdom’s entire GDP in 2021 was $3.1 trillion. This would increase the impact of all artificial intelligence by 15 to 40 percent. This estimate would roughly double if we include the impact of embedding generative AI into software that is currently used for other tasks beyond those use cases”); Generative AI Could Raise Global GDP by 7%, Goldman Sachs (Apr. 5, 2023), https://www.goldmansachs.com/insights/articles/generative-ai-could-raise-global-gdp-by-7-percent (“As tools using advances in natural language processing work their way into businesses and society, they could drive a 7% (or almost $7 trillion) increase in global GDP and lift productivity growth by 1.5 percentage points over a 10-year period”).

[20] See, e.g., Hugging Face, GPT-4chan, https://huggingface.co/ykilcher/gpt-4chan (last visited Sep. 5, 2024).

[21] Miguel A. Cardona et al., Artificial Intelligence and the Future of Teaching and Learning, U.S. Dept. of Educ., https://www2.ed.gov/documents/ai-report/ai-report.pdf (last visited Sep. 5, 2024).

[22] See, infra, at nn. 29-32 and accompanying text.

[23] See Aaron Wildavsky & Adam Wildavsky, Risk and Safety, Econlib, https://www.econlib.org/library/Enc/RiskandSafety.html (last visited Sep. 5, 2024).

[24]See, Dual-Use Foundation Models with Widely Available Model Weights, Nat’l Telecomm. Info. Admin., available at https://www.ntia.doc.gov/sites/default/files/publications/ntia-ai-open-model-report.pdf, at 2 (last visited Sep. 5, 2024).

[25] Id. at 3.

[26] Id. at 10.

[27] Id. at 10.

[28] Id. at 10-11.

[29] Id. at 11.

[30] See Tom C.W. Lin, The New Investor, 60 UCLA L. Rev. 678 (2013).

[31] Justin (Gus) Hurwitz & Geoffrey A. Manne, Pigou’s Plumber: Regulation as a Discovery Process, SSRN (2024), https://laweconcenter.org/resources/pigous-plumber.

[32] Id. at 32.

[33] See Thomas A. Hemphill, Technology Entrepreneurship and Innovation Hubs: Perspectives on the Universal Regulatory Sandbox, 50 Sci. & Pub. Pol’y 350, 352 (2022), https://doi.org/10.1093/scipol/scac072; Sofia Ranchordds, Experimental Regulations and Regulatory Sandboxes – Law Without Order?, 2021 Law & Method 2-3, available at https://www.lawandmethod.nl/tijdschrift/lawandmethod/2021/12/lawandmethod-D-21-00012.pdf.

[34] See Draft, supra note 1, at 16-17.

Abstract

Consumer welfare has been a north star of the Federal Trade Commission (FTC), providing an organizing principle for diverse issues under the Commission’s dual competition and consumer protection missions and, specifically, a uniform ground on which to examine the law and economics of privacy matters and the tradeoffs that privacy policies entail. This paper provides the first contemporary literature synthesis by former FTC staff that brings together the legal and economics literatures on privacy. Our observations are the following: (a) privacy is a complex subject, not a simple attribute of goods and services or a simple state of affairs; (b) privacy policies entail complex tradeoffs for and across individuals; (c) the economic literature finds diverse effects, both intended and unintended, of privacy policies, including on competition and innovation; (d) while there is diverse and growing evidence of the costs of privacy policies, countervailing benefits have been understudied and, as of yet, empirical evidence of such benefits remains slight; and (e) observed costs associated with omnibus policies suggest caution regarding one-size-fits-all regulation.

Abstract

AI has taken the world by storm, and competition law is no exception. Policymakers, academics, and commentators are struggling to make sense of how to apply competition law principles to burgeoning AI markets. The question is spurred by an impending sense that inaction is likely to lead to monopolistic outcomes that will later be impossible to revert. What is feared is that AI will become dominated by a few large technology companies and, more spuriously, that these will be the same companies that already control vast swathes of the so-called digital sphere. In other words: it will make big tech even bigger.

One difficulty with this narrative, however, is that, strictly speaking, there is no such thing as an “AI market”because AI is not a unitary, monolithic technology. In this chapter, we argue that the first step to ensuring that antitrust law stays relevant in the age of AI is developing a principled approach to defining AI relevant markets; one that is legally, economically, and technologically sound. Relevant market definition is central to antitrust law because it is the starting point of most, if not all antitrust law cases.A relevant product market is typically comprised of all those products which consumers view as substitutable and which can therefore be said to compete against each other. By delineating the boundaries of competition between firms, relevant product market definition alerts of the presence of market power and, by extension, of the likelihood of anticompetitive effects.Despite its limitations and despite not being an end in itself,  relevant product market definition is, and likely will remain, the main tool for thinking about the contours of competition between firms for the foreseeable future.

We suggest how this can be done by grasping the internal heterogeneity of AI and by understanding what makes AI similar to HI and HI-powered tasks, thus eschewing simplistic narratives about AI’s supposed ubiquity and uniqueness that are bound to impede antitrust law from discharging its social role, which is to protect competition for the ultimate benefit of consumers.

 

Abstract

It is often assumed that the creation and protection of competitive markets, whenever possible, is a central tenet of classical liberalism and that antitrust laws are an essential feature of a classical liberal political economy. The logic behind this proposition is that monopolization and cartelization produce private gains for their participants while resulting in “deadweight loss”: lower output, higher prices, and a general decline in overall wealth and productivity. But that cannot be the end of the story. Remedies to guard against these risks are both costly and (often) erroneous, thus deterring beneficial conduct, as well. This unavoidable tradeoff makes it unclear what concrete legal and policy prescriptions classical liberalism prescribes to maintain, protect, and promote competition—if any. As a result, classical liberal ideals have been invoked to justify deeply opposed policies: from antitrust abstentionism to aggressive trust-busting. This chapter aims to elucidate the perennial, albeit elusive question of whether classical liberalism supports the enactment of general antitrust laws and, if so, under what principles.

Executive Summary

Privately run digital-payments systems are helping to bring informal market transactions and unbanked people into the formal economies of many countries—in the process, improving livelihoods and fostering innovation and economic growth.

The success these private systems experienced in expanding financial inclusion has led some governments to create their own systems for digital payments. “India Stack,” for example, combines a national ID and associated database (Aadhaar), an interoperability framework for real-time payments (UPI), and a set of standards for sharing financial data (DEPA). Brazil, meanwhile, has focused on the middle part of that stack, with the nation’s central bank, BCB, creating its own real-time payment system (Pix).

The results of these interventions have been mixed. Where they have supported natural market developments (as is largely the case with Aadhaar), they have had a positive impact. But where they have competed directly with market actors, they have created distortions that may inhibit more effective market solutions. For example:

• Restrictions on interchange fees imposed by UPI have favored market participants who are able to monetize the operation of payment networks in other ways, such as Google (which monetizes primarily through search advertising) and Walmart (which monetizes through the sale of goods).
• Similarly, subsidies from the BCB and Pix’s restrictions on interchange fees crowd out private-sector competition and undermine innovations that might lead to increased financial inclusion.

In some cases, governmental attempts to force a shift toward digital money have impeded financial inclusion. India’s “demonetization” in 2016 harmed M-Pesa, contributing to greater financial exclusion. Arbitrary governmental regulations, including restrictions on the number of WhatsApp Pay users, have also likely inhibited both competition and financial inclusion.

The key lesson is that the private sector is well-placed to identify opportunities for digital payments and to implement solutions appropriate to local consumer culture and other contingent facts. Meanwhile, governments should focus on delivering genuine public goods, such as identity verification, and avoid the temptation to participate in digital-payments infrastructure.

I. Introduction

Digital-payment systems are expanding access to financial services around the world. In many countries, they are helping to bring the previously unbanked into the formal economy. In so doing, they undoubtedly improve livelihoods and foster innovation and economic growth. But not all digital-payment systems are alike.

Early digital-payment systems were largely private-sector initiatives. Notable among these are: M-Pesa in Kenya; Venmo in the United States; AliPay and WeChat Pay in China; and MobiKwik and Paytm in India. These emerged as bottom-up responses to the felt needs of consumers and merchants.

The success these private systems experienced in expanding financial inclusion has led some governments to create their own systems for digital payments. “India Stack,” for example, combines a national ID and associated database (Aadhaar), an interoperability framework for real-time payments (UPI), and a set of standards for sharing financial data (DEPA). Brazil, meanwhile, has focused on the middle part of that stack, with the nation’s central bank, BCB, creating its own real-time payment system (Pix).

To better understand the costs and benefits of government interventions, this brief investigates the economics of mobile digital payments, their effects on financial inclusion, and the consequences of specific government interventions. Section II begins with a discussion of the core problems that payment systems must address, providing a law & economics framing. Section III describes the development of digital-payment platforms in Kenya, China, India, and Brazil. Section IV puts the evidence presented in section III into broader context and draws conclusions for policymakers.

II. The Economics of Mobile Digital Payments and Financial Inclusion

This section briefly considers the economics of payments, focusing on the two core problems that all payments systems must address: counterparty risk and the two-sided nature of the payments markets.

A. Trust and Counterparty Risk

In all transactions, there is a risk that a counterparty will not perform the promised action. For example, in the sale of any good, the seller may fail to deliver that good; the good may not function as described; or the buyer may fail to make payment. This is particularly problematic in the context of transactions among parties who are unknown to one another, or who are separated by time and space, as is the case in many online transactions.

One way to address counterparty risk is through the law of contract. In common-law jurisdictions, buyers may bring breach-of-contract actions against sellers if the goods purchased are damaged, defective, or otherwise not fit for purpose (or some equivalent term). Meanwhile, sellers may bring breach-of-contract actions against buyers who fail to make payment as agreed in a timely manner.

Often, however, the cost of legal action is large relative to the value of a transaction. As such, while contract law may function as a backstop, it is usually more efficient to resolve such disputes through alternative means. In the online marketplace, entrepreneurs have developed several important technologies that serve to facilitate trust and otherwise address counterparty risk; these include user ratings, security protocols and fraud-detection systems, payment gateways, and online escrow systems. The digital-payment systems discussed in Section III typically rely on some or all of these technologies. (The Appendix includes a more detailed discussion of the various ways counterparty risk has been addressed in the context of online transactions.)

B. Two-Sided Markets

Payment systems are an example of what economists call “two-sided markets”: buyers are on one side, sellers on the other, and the payment system itself acts as the platform that facilitates interactions between them.[1] The primary challenge faced by any two-sided market is to ensure that there are sufficient participants on both sides for it to be self-sustaining. Conceptually, this is known as a cross-side network effect.

In the case of payment systems, if too few sellers accept a particular form of payment, buyers will have little reason to adopt it. Likewise, if too few buyers hold a particular form of payment, sellers will have little reason to accept it. This applies equally to other two-sided markets, such as shopping malls. For customers to want to go to a mall, there must be stores that they want to visit. And for stores to want to locate in a mall, they need to know that there will be customers who will visit.

Two-sided markets also must cover the operational costs. In the case of shopping malls, this means the costs of building and operating the mall, including ongoing maintenance, lighting and heating/cooling of common areas, and various services, usually including security. In the case of payment networks, it includes the development and maintenance of technology needed to facilitate secure transactions, verify identity, and deter fraud.

Platforms typically address these problems simultaneously by charging differential fees and offering incentives. Thus, malls typically charge store owners (one side of the market) rent and service fees sufficient to cover both capital and operational costs. Those store owners thus effectively subsidize access to the mall for customers (the other side of the market), who are generally free to walk around enjoying the mall’s climate and security even if they do not ultimately make any purchases. This highlights a feature common to most two-sided markets: it is usually efficient for one side to subsidize the other, thereby maximizing the value to all participants. Typically, the side that provides the subsidy has a lower price elasticity of demand (i.e., it is less sensitive to changes in the cost of platform services).[2]

Meanwhile, to encourage popular retailers to locate in the mall in the first place, mall owners may offer discounted rent to early occupants and/or specific retailers who are especially likely to attract customers (such as hotels and restaurants with a loyal following).[3] This highlights another feature of two-sided markets: it may also be efficient for some participants on one side to subsidize other participants on the same side, again in order to maximize the value to all participants.

Like malls, payment networks typically facilitate cross-network subsidies. Merchants typically place a higher value on the ability to accept a specific payment type than consumers place on the ability to use that payment type. In other words, merchants have lower price elasticity of demand for the payment network. Because of this, the subsidy typically flows from the merchant to the consumer.

In some cases, payment networks also exhibit own-side subsidies. For example, smaller-ticket items may be subject to proportionally lower fees—a merchant-side subsidy. Meanwhile, some consumers may receive an initial rewards bonus and/or pay a lower fee for the first year they use a certain payment type (such as a credit card)—a consumer-side subsidy.

III. Mobile Digital-Payment Platforms

While Venmo and Zelle may be better known in the United States, the first cellphone-based payment networks first emerged informally in Tanzania and some other sub-Saharan African countries in the early 2000s, allowing users to send cellphone airtime minutes from one phone to another.[4] These informal networks inspired the development of more formal systems, such as M-Pesa. In China, meanwhile, mobile payments emerged from two different sources: an online payment system (Alipay) and a cellphone-messaging service (WeChat). India’s early leader in mobile payments, Paytm, grew out of a cellphone-charging platform.

A. M-Pesa and the Mobile-Payments Revolution in Africa

In 2006, Kenya had 41 retail banks with 400 branches and 600 automated-teller machines (ATMs) for a population of 36 million.[5] By comparison, the United States at the time had 7,523 banks with 72,958 branches and 400,000 ATMs for a population of 296 million. On a per-capita bases, that’s nearly 20 times the number of banks and branches, and 80 times as many ATMs.[6]

Unsurprisingly, Kenyans—especially those in rural areas—were not major users of the banking system. A survey undertaken in 2006 found that just 18.9% of Kenyan adults had access to a bank account or other formal financial-services provider, such as insurance, while a further 7.5% had accounts with semi-formal savings and credit organizations and/or a microfinance institution. Thus, no more than about a quarter of the adult population had access to a formal or even semiformal financial-services provider.[7] Of the remainder, slightly less than half of the population (35.2% of those surveyed) had access to informal financial services, while the rest (38.4%) were considered “financially excluded.” The proportion of adults with access to electronic payments was even smaller, with less than 10% having a credit card.[8]

Three years before that survey was undertaken, Nick Hughes, then head of social enterprises at Vodafone, learned of the use of air-time credits as an informal currency in Tanzania, and believed it should be possible to create a more formal system to facilitate microcredit and microinsurance—services that were popular among development groups at the time. [9] Hughes put together a proposal to do just that and submitted it to the UK government’s Department for International Development (DFID). The DFID awarded Vodafone £1 million for a pilot program, which Vodafone matched with a combination of cash and personnel time.

For the pilot project, which launched in October 2005, Hughes selected Commercial Bank of Africa (CBA), which would hold customer funds in a trust account; Faulu Kenya, a microfinance institution that would provide loans; and Vodafone’s 40% owned subsidiary Safaricom.[10]

Hughes called the system M-Pesa: “pesa” is Swahili for money and M stands for mobile, so M-Pesa is “mobile money.” The pilot project proved to be a wild hit, though not primarily for microfinance loans. It was mainly used to send money from one person to another, including people who were not part of the pilot. For example, it was used for business transactions, for remittances, and to store money overnight.[11]

In 2007, Safaricom formally launched M-Pesa at-scale. That meant onboarding thousands of agents, who receive funds deposits from accountholders, and whose identity they verify using a national ID card or passport. The agents add those funds to the system, as well as pay-out funds that are withdrawn. By 2012, there were more than 30,000 M-Pesa agents. As of October 2023, there were 160,000 M-Pesa agents in Kenya alone, and 500,000 across all the markets in which M-Pesa operates.[12]

M-Pesa began as a text-based system with limited functionality. As smartphones became more prevalent, it expanded to include a wider range of financial services. In 2011, Safaricom partnered with CBA to enable users to create a linked bank account, called M-Shwari, that enables M-Pesa users to save and borrow.[13] By 2015, more than 10 million Kenyans had opened M-Shwari bank accounts; for many, this represented their first such formal bank account.[14] CBA also launched similar products in Tanzania, Uganda, Rwanda, and Cote d’Ivoire.[15]

Using a mobile app or web interface, it is also now possible to pay money into M-Pesa electronically by linking to a bank account.[16] Since 2022, M-Pesa has had a partnership with Visa, which enables M-Pesa accountholders to establish a virtual account that can then be used to make international purchases online.[17]

According to the most recent FinAccess report, M-Pesa helped to expand financial inclusion in Kenya from 26% of the population in 2006 to 84% in 2021.[18] It has also facilitated economic growth and, according to a study published in Science by MIT economists Tavneet Suri and William Jack, has helped to lift hundreds of thousands out of poverty.[19]

M-Pesa has been so successful that it has expanded to the Democratic Republic of Congo (DRC), Egypt, Ghana, Lesotho, Mozambique, and Tanzania. It now boasts more than 51 million users and $314 billion in annual transactions.[20] Other mobile operators have also followed suit, both in Kenya and across Africa, with Orange providing its “Orange Money” service to more than 80 million customers in 17 countries in Africa and the Middle East.[21]

B. China

The success of mobile-payments systems in Africa has been much vaunted. But by some metrics (e.g., rate of adoption, value of transactions, and broad economic effects), it pales in comparison with the success of the two primary mobile-payments systems that have developed in China: Alipay and WeChat Pay.

1.  AliPay

In 2003, there were only 3 million credit cards issued in China, implying that less than 5% of adults had a card.[22] Meanwhile, China had around 20 million internet-connected computers and 60 million internet accounts.[23] The vast majority of Chinese internet users thus lacked a reliable means to make online purchases. This posed significant problems for online auctions and other consumer-to-consumer marketplaces, where there is an inherent trust deficit between buyer and seller—and associated counterparty risk.[24]

To address this problem, in 2003, online retailer Alibaba introduced an escrow service, called Alipay, for its consumer-to-consumer subsidiary Taobao.[25] When making a payment, funds are initially deposited in Alipay’s escrow account and only released to the seller once both parties are satisfied that the transaction is complete.[26] Initially, users would fund their Alipay personal digital wallet by linking it to their bank account.[27] Over time, Alipay has expanded the ways in which wallets can be funded to include debit cards, credit cards, bank transfers, and other digital wallets.[28]

Alipay’s mobile wallet, launched in 2008, had grown to roughly 100 million users by 2013. Today, it has more than 900 million users in China and about 1.3 billion users around the world.[29]

2. WeChat Pay

WeChat Pay (officially “Weixin Pay”) is a mobile wallet and payments system launched in 2014 with the express intention of enabling users of the WeChat messaging app to send money to one another.[30] To drive adoption, WeChat introduced the ability to share virtual “red envelopes” with family and friends over the Lunar New Year holiday—replicating an ancient tradition.[31] This proved enormously popular, with 20 million red envelopes sent in the first year, and 3.2 billion sent in the following year. Over time, WeChat’s owner Tencent has built an ecosystem around WeChat Pay, including mini-programs that enable the user to hail rides or pay for movie tickets.[32] As of 2022, about 60% of WeChat users use these mini-programs.[33]

Unlike AliPay, WeChat Pay does not provide escrow services, and is thus less well-suited to certain kinds of online transactions. It works well, however, both for in-person payments to merchants and transfers to other individuals (including gift giving). WeChat Pay has approximately 1.1 billion users worldwide.[34]

3. Discussion

AliPay and WeChat Pay’s precursor (TenPay) both began life as mechanisms to enable payments within existing ecosystems of products and services. AliPay’s pivot to mobile payments proved an enormous success, motivating Tencent to create its own mobile-payments system. Over time, both systems have gradually expanded to provide payment services outside their own ecosystems. Payments using either platform can be made at an increasingly large number of merchants using a range of technologies, including quick-response (QR) codes for in-store or peer-to-peer (P2P) payments.[35]

For nearly a decade, AliPay and WeChat Pay have been engaged in fierce competition. As access to credit cards in China has grown, this competition has expanded to include China Union Pay. As of 2022, AliPay and WeChat Pay were by far the most used means of making online payments (see Figure 1).

FIGURE 1: Most Used Online-Payment Services in China (2022)

SOURCE: Daxue Consulting[36]

In 2015, Tencent and Alibaba established banks associated with their payment networks. WeBank (Tencent) and MyBank (Alibaba, now part of Ant Financial) have since grown rapidly. Ant Financial reports that, as of June 2022, MyBank had provided loans to more than 49 million small and medium-sized enterprise (SMEs), of which 80% had not previously had a bank loan.[37] WeBank, meanwhile, reports having more than 370 million individual clients and more than 4 million micro-small-and-medium-sized-enterprise (MSME) clients.[38]

In July 2023, the Chinese authorities permitted AliPay and WeChat Pay to integrate payments made using Mastercard, Visa, American Express, and Discover.[39] This, combined with increasing adoption of China Union Pay cards domestically, has led to an increase in the use of AliPay and WeChat Pay as mobile-payment gateways, in contrast to their role as digital wallets.

The emergence and expansion of digital payments in China has undoubtedly resulted in improvements in access to financial services, both directly (since payments are a financial service) and indirectly, by enabling tens of millions of businesses and hundreds of millions of individuals to hold bank accounts and thereby access related services, including loans.

C. India

India’s mobile-payment system has grown rapidly over the past decade. The market’s key players are MobiKwik, Paytm, PhonePe, and Google Pay. This section discusses those services, as well as WhatsApp Pay and M-Pesa, which have been less successful in India.

A major factor affecting the development of mobile payments in India has been the national digital ID (Aadhaar) and the Unified Payment Interface (UPI), both part of what is now referred to as “India Stack.”[40] I plan to look at India Stack in more detail in a future piece, but it is worth noting the following:

1. Aadhaar was developed by the Unique Identification Authority of India (UIDAI), a government authority established in 2009, in order to enable every resident of India to have a unique digital identity that could be instantaneously verified. The goal was to thereby reduce identity theft and related fraud, and facilitate more rapid onboarding to banks and other financial entities that require proof of identity for know-your-customer (KYC) purposes.[41] UIDAI began enrolling Indian residents in Aadhaar in 2010 and, by the end of 2015, had enrolled more than 800 million people.[42] As of February 2023, 1.38 billion Indian residents are enrolled—about 96% of the population.[43]
2. UPI was developed in 2016 by the National Payments Corporation of India (NPCI), a public-private partnership, as an open-source interoperable application programming interface (API) that facilitates real-time transfers between individuals with accounts at participating banks who have integrated the API into their smartphone apps.[44]

1. MobiKwik

Among the first mobile-payments systems in India, MobiKwik was initially established in 2009 by husband-and-wife team Bipin Preet Singh and Upasana Taku.[45] In 2012, it launched a digital wallet feature and, in 2013, the Reserve Bank of India awarded it a license for that wallet.[46] MobiKwik grew organically, aided by significant investments from Bajaj Capital and Sequoia, and signed up 1.5 million merchants and 55 million registered users by 2015.[47] It has continued to grow in the nearly decade since, albeit at a slower rate, recording around 4 million merchants, 140 million registered users, and 35 million active monthly users in 2023.[48]

MobiKwik generates revenue from commissions and advertisements from its Zaak payment-gateway franchise subsidiary,[49] as well as loans—including short-term credit, buy-now-pay-later, and personal loans—and investment advice.[50]

2. Paytm

Paytm began life in 2010 as a prepaid mobile-charging platform, but pivoted to payments in 2014 with the introduction of a mobile wallet. Paytm now generates revenue from three main sources: payment services, financial services, and “commerce and cloud.” Payment services—the core of its business, contributing 58% of its revenue in Q3 2023[51]—arise from users making payments from mobile wallets, debit cards, and credit cards, for which Paytm charges merchants a fee (merchant-discount rate) that ranges from 0.4% to 2.99% of the transaction amount, depending on the payment type (for small to medium-size businesses).[52] Financial services—including loans, investments, and insurance—accounted for 22% of Paytm’s revenue in the most recent quarter, while “commerce and cloud” services—Paytm’s marketplace that features more than 100,000 merchants—accounted for 20% of revenue.

This combination of offerings has allowed Paytm to become a highly effective multisided market. A year after launching its mobile wallet, Paytm had 100 million registered accounts, making it the early leader in the field in India.[53] The user base grew by 50% in 2016,[54] likely in part as a result of the government’s decision to cancel the majority of its banknotes.[55] It now has approximately 100 million active users.[56] While Paytm’s wallet remains the most popular digital wallet in India, its share of the mobile-payments market has declined as a result of competition from PhonePe and Google Pay.

In addition to offering payments, Paytm provides a range of financial services, including personal and business loans. It also operates its own ecommerce platform.

3. PhonePe

Created in 2015 by former Flipkart executives Sameer Nagim and Rahul Chari, PhonePe is India’s market-leading mobile-payments platform and was acquired by Flipkart in 2016 prior to its launch.[57] Flipkart itself—founded in Bengaluru in 2007—was purchased by Walmart in 2018.[58] In December 2022, PhonePe was separated from Flipkart, though both remain majority owned by Walmart.[59]

By launching as a wholly owned subsidiary of Flipkart, India’s largest online marketplace, PhonePe was able to leverage the marketplace’s then 100 million users (it has since grown to 400 million),[60] enabling it rapidly to acquire customers. PhonePe also differentiated itself from Paytm by primarily targeting middle-class consumers and associated merchants in second-tier cities.[61] PhonePe was also an early adopter of the Unified Payments Interface (UPI), a real-time payments solution that facilitates account-to-account transfers at no cost to either party.

Since separating from Flipkart, PhonePe’s offerings are now remarkably similar to those of Paytm. It does, however, have a larger user base (approximately 200 million monthly active users, double Paytm’s 100 million) its historic connection to Flipkart enables it more effectively profit from low-margin, high-volume activities.

4. Google Pay

Google Pay, which launched in India in 2017 (originally under the brand Tez), is now the second most popular mobile-payments system in India.[62] Google Pay does not act as a wallet in India, but is essentially a payment gateway that works exclusively with UPI.[63]  Google Pay has built a large Indian customer base due largely to brand recognition. Google is able to monetize Google Pay through advertising and its local online marketplace.[64] Google has also been granted a license to operate as a payment aggregator.[65]

5. WhatsApp Pay

Meta launched a pilot program for WhatsApp Pay in India in 2017, but did not fully launch until 2020 and has been subject to limits on its number of users.[66] Despite operating the most popular messaging service in the country, with more than 400 million users, WhatsApp Pay has not taken off. Indeed, its peak market share was about 0.4%, and it has since fallen to about 0.1%.[67]

WhatsApp Pay’s failure in India is likely due in no small part to regulatory delays in launching and the NPCI-imposed limits on the number of users—initially 40 million, but increased to 100 million in 2022.[68] The delay in launching meant that PhonePe and Google Pay were able to gain the upper hand on WhatsApp, becoming the predominant UPI-based wallet and payment gateways, respectively. The limit on signups, meanwhile, has forced WhatsApp to impose arbitrary restrictions on which customers can adopt its payment service, which effectively undermines the self-organization that underpins P2P-transaction networks.

6. M-Pesa

M-Pesa entered the Indian market in 2011, hoping to repeat its success in Kenya. Despite attracting more than 400 million Indian subscribers, many of them in rural areas, Vodafone was only able to onboard 8.4 million subscribers and eventually shut down the operation in 2019.[69] In an interview in 2020, the former head of M-Pesa, Michael Joseph, identified four factors that explained the failure:[70]

• Difficulty recruiting agents: In Kenya, Safaricom had invested heavily in building a large network of M-Pesa agents across the country, including in remote villages. This allowed easy cash deposits and withdrawals. But in India, Vodafone struggled to find shops and merchants to serve as last-mile M-Pesa agents in rural areas.
• Lack of understanding among consumers: In contrast to Kenya, poor and unbanked Indian consumers needed more handholding and awareness-building to adopt M-Pesa, which required significant time, money, and labor from Vodafone.
• Competition from well-funded startups targeting more profitable market segments: Meanwhile, mobile-payment startups such as Paytm targeted urban middle-class and affluent consumers. They could link to bank accounts, unlike M-Pesa, which dealt in cash for the unbanked. Paytm also had significant investor backing.
• Demonetization: The 2016 Indian demonetization benefited digital payments that were linked to bank accounts, but it harmed M-Pesa, which focused primarily on the unbanked.

7. Discussion

The veritable explosion in mobile payments in India over the past decade has undoubtedly contributed to improvements in financial inclusion. India’s primary mobile-payments apps—MobiKwik, Paytm, PhonePe, and Google Pay—have sought to leverage their platforms to maximize value for both sides of the market (merchants and consumers) in different ways:

• MobiKwik (Zaakpay), Paytm, PhonePe, and Google Pay provide payment-gateway and/or aggregator services.[71]
• MobiKwik, Paytm, and PhonePe offer users wallet features and seek to supply additional financial services, such as insurance and investment.
• Paytm, PhonePe, and Google Pay generate revenue through online marketplaces.
• All the apps (but especially Google Pay) generate revenue from advertising.

As noted in the introduction to this section and throughout, UPI likely contributed to the rapid adoption of mobile payments, especially for PhonePe and Google Pay, which were built around it. The fact that UPI requires parties to have a bank account has likely created incentives to open such account, thereby increasing financial inclusion. The provision of welfare support through e-RUPI vouchers that are delivered over UPI also likely increases financial inclusion.[72]

Some aspects of UPI, however, may also have hindered more widespread adoption of mobile digital payments and inhibited financial inclusion. For example, NPCI currently prohibits card and app operators from charging merchants for most transactions made using UPI (fees of up to 1% are permitted only for prepaid debit cards and pre-funded mobile wallets). Those card issuers and app operators who lack existing means to monetize their relationships with customers are therefore likely to find it difficult to break into the market using UPI, as they cannot easily offer rewards and other inducements to customers to on-board and use their card or app.

The Payments Council of India estimates that its members lose 55 billion rupees ($660 million) annually as a result of the zero MDR on UPI and RuPay transactions.[73] This is effectively a transfer from banks to the companies whose apps monetize UPI transactions. India’s government partly offsets this loss through a subsidy to UPI participants of between 15 and 25 billion rupees.[74] But experience with other systems that impose restrictions on payment-transaction fees suggests that banks will seek to recover these losses from other fees.[75] To the extent that such additional fees fall on lower-income accountholders, the effect on financial inclusion is likely negative. (I plan to explore this and related issues in more detail in a future brief on India Stack.)

D. Brazil

The first mobile-payments system in Brazil appears to have been Oi Paggo, a closed-loop short-messaging-service (SMS) “credit card” system founded in 2004 and formally launched in 2006 as an app on the Oi mobile-phone service (it was subsequently purchased by Oi).[76] In 2008, Oi Paggo reportedly had around 1 million users, but does not appear to have achieved mass adoption.[77]

Over the course of the past dozen years, several other entrants have built successful digital mobile-payments systems. According to a recent Survey by Statista, the top five most-used mobile-payments systems in Brazil at the end of 2023, based on proportion of respondents who had used the payment online or in stores in the previous year, were: PicPay (50%), PayPal (46%), Mercado Pago (44%), PagBank (31%), and Google Pay (25%).[78]

Before discussing these mobile-payment solutions, it is worth briefly noting that, in 2021, the Central Bank of Brazil (BCB) introduced its own real-time payment system, Pix, and required all of Brazil’s larger banks to implement it for all accountholders. As a result, Pix has had a significant effect on mobile digital payments in Brazil.

1. PicPay

PicPay was founded in 2012 as a digital wallet and initially launched in the city of Vitória, where it has gained significant traction, with approximately 75% of the population using it.[79] It subsequently launched in Rio de Janeiro and São Paulo. In 2022, PicPay—owned by closely held J&F Investments—obtained a banking license via J&F’s bank, Banco Original, enabling it to expand its offerings and lower its funding cost.[80] PicPay has more than 30 million active users and is accepted at more than 10 million merchants.[81] It generates revenue from a range of product lines, including loans, credit cards, investments, insurance, and a crypto-currency exchange.[82]

2. PayPal

Originally conceived as a payment system in its own right, PayPal functions both as a payment gateway—processing card payments—and as a wallet. In Brazil, PayPal also has a Pix integration, which allows users to send and receive payments to their PayPal account using their Pix keys.[83]

3. Mercado Pago

A subsidiary of the online marketplace Mercado Libre, Mercado Pago is able to leverage the customer base of that market, offering a full range of payments and banking services, including checking accounts, deposit accounts, loans, debit cards, and credit cards.[84] Mercado Pago operates in several countries and reported having nearly 65 million accountholders in 2023.[85]

4. PagBank and PagSeguro

PagSeguro was founded in 2006 as an online payment system by Universo Online (UOL), Brazil’s largest internet-content and services provider.[86] In 2013, PagSeguro diversified into point-of-sale (POS) payments and then into banking, offering a range of services. PagBank now has around 15 million active Brazilian users and 7.7 million merchants.[87]

5. Google Pay

In Brazil, Google Pay operates primarily as a payment gateway, facilitating transactions made using debit and credit cards whose (tokenized) information is stored on a user’s phone and transmitted to POS machines using radio-frequency identification (RFID).[88]

6. Discussion

The above-mentioned survey suggests that there is considerable multi-homing for mobile digital payments in Brazil, with the same consumers using several different mobile-payments apps. It also seems plausible that these apps have contributed to financial inclusion. As Figure 2 shows, in the five years from 2017 to 2022, there was a dramatic shift away from the use of cash toward a combination of card and digital mobile payments.

FIGURE 2: POS Payments in Brazil by Method

SOURCE: Statista

Meanwhile, the proportion of Brazilians with accounts at a financial institution rose from 70% in 2017 to 84% in 2021. Some have suggested that this was the result of Pix’s introduction, but that appears unlikely, as Pix was only introduced at the end of 2020. Two other factors seem better candidates: first, the growing use of digital payments and associated onboarding by financial-technology firms (fintechs), such as PagBank; second, some government subsidies offered in response to the COVID-19 pandemic required Brazilians to open bank accounts.

A study by Americas Market Intelligence (commissioned by Mastercard) found that, during the COVID-19 pandemic: “Brazil reduced its unbanked population by an astounding 73%.”[89] The study was based on research conducted between June and August 2020 and was published in October 2020, the month before Pix launched. It described the implementation of state and federal programs that Brazil launched in response to the pandemic:

• The “Coronavoucher” program distributed emergency funds to low-income informal workers exclusively via the state-owned bank Caixa Econômica Federal (CEF). Applications for funds could only be made via CEF’s Caixa Tem smartphone app, and funds were distributed using the same app. As of Aug. 5, 2020, 66 million people had received Coronavouchers via the Caixa Tem app. Of those, 36 million were previously unbanked.
• Merenda em Casa (“snack at home”), a program run by state governments, distributed funds to low-income families with children at public schools to help pay for food while schools were closed due to COVID-19. The program distributed funds via PicPay and PagBank’s PagSeguro, both private-sector payment apps.[90]

Following the launch of Pix, the BCB-run RTP program was made available to clients of Caixa Tem, PicPay, and PagBank.[91] As a result, previously unbanked individuals who had become banked because of the Coronavoucher and Merenda em Casa programs were able to obtain and use Pix keys to send and receive payments.

IV. Conclusions and Policy Implications

At one level, every successful payment system solves the same set of problems—in essence: creating a self-sustaining network that facilitates the transfer of funds, while limiting counterparty risk. But as this study shows, each system does so in ways that are economically, legally, and culturally contingent.

In Kenya—and, subsequently, many other countries—M-Pesa provided a means for millions of people without ready access to banks to send and receive money over long distances. To do so, it created a network of tens of thousands of agents providing on- and off-ramps for cash. Over time, this has gradually transformed into a broader ecosystem that includes banking and other financial services.

In China, AliPay provided a solution to counterparty risk for internet-based transactions in an economy where alternative solutions, such as credit cards, were not widely available. AliPay was able to leverage its reputation for reliable and secure payments to facilitate transactions using increasingly widely available smartphones. Meanwhile, WeChat Pay was able to leverage its network of hundreds of millions of chat users to establish an effective mobile-payments ecosystem and associated markets for goods and services.

In India, MobiKwik and Paytm both offered digital wallets as an alternative to card payments for middle-class consumers in an economy where credit cards were tightly regulated. But over time, and especially with the introduction of UPI, all of the mobile-payment systems’ business models have converged around offering a range of financial products and markets.

In Brazil, an early attempt to introduce a mobile digital platform predicated on the credit-card model seems to have foundered, but later entrants—notably PicPay, PayPal, Mercado Pago, PagSeguro and other players—have been able to build and monetize payment networks by offering payment-gateway services and a range of adjacent products.

Public policy has played a substantial role in the development of all these mobile-payments ecosystems. In some cases, the effects have likely been positive. For example, India’s Aadhaar digital ID has dramatically reduced the cost of proving the identity of hundreds of millions of people, thereby facilitating quicker and less expensive access to bank accounts. The distribution of welfare payments via payment apps in India and Brazil also likely increased adoption at the margin.

On the other hand, restrictions on transaction fees in India (zero MDR on UPI and RuPay) and Brazil (extremely low interchange fees permitted on Pix) have almost certainly made it more difficult to reach some potential users. In the absence of adequate merchant-transaction fees, the only way to monetize mobile-payments systems is through advertising or by selling other products. Operators of mobile-payments systems might encourage some participants to use their platform by offering discounts on certain products and services. For example, in India, Google offered cashback incentives for use on apps within its own (Android) ecosystem.[92] While there is nothing wrong with this, it is less likely to attract marginal users who are not significant consumers of Android’s app-based services than would simply offering unencumbered cashback rewards.

It is also worth noting that limiting MDR fees on UPI and interchange fees on Pix does not necessarily grant merchants a free ride. They will often have to pay hefty per-transaction fees to participate in online markets, with those fees paid to the parent company (e.g., Google) rather than the payment-service provider (in this example, Google Pay). Moreover, since mobile-payments-system operators have incentives to keep customers within their ecosystem through internal rewards programs, brick-and-mortar merchants may find that they lose customers.

At base, these policy failures stem from the fact that payments markets do not cease being two-sided simply because governments choose to intervene in them. The operators of digital mobile-payments systems are highly aware of this fact. Consider this note from PayPal:

PayPal began with a vision of an online world where consumers could safely, securely, and easily use digital payments to buy what they need and want from businesses. Over the past 24 years, our two-sided network has grown significantly, seamlessly connecting shoppers and retailers across the globe.

Our platform now connects businesses, who look to us to help them navigate the digital economy, with shoppers, individuals who increasingly value flexibility and ease in where and how they shop and manage their daily financial lives. [93]

Meanwhile, PicPay notes:

We now have a more comprehensive portfolio of products and services, further enhancing our value proposition beyond the digital wallet and day to day payments services, becoming a much broader two-sided financial ecosystem.[94]

Interventions in two-sided markets change the participants’ incentives. For example, when MDR or IF rates are subject to price controls, system participants look for other ways to monetize their products (or exit). Fintech companies that operate mobile digital payments have found ways to monetize consumers through advertising and/or online marketplaces. But for banks that don’t also operate mobile digital-payments systems, participating in payment apps for which they cannot charge fees on most transactions, such as UPI and Pix, is often both unavoidable and unfortunate. It is unavoidable because accountholders now expect to be able to use these apps for payments. It is unfortunate because it adds cost, while bringing in minimal IF revenue at best. Banks therefore look for other ways to monetize those accounts, which can include increasing account-maintenance fees or hiking borrower interest rates and/or other fees. Such actions typically harm the poorest consumers, who are least able to afford additional account fees.

Following enactment in the United States of the so-called “Durbin amendment,” which imposed price controls on debit-card IFs, covered banks raised account fees and increased the minimum balance required to maintain an account. In response, it appears that many lower-income consumers closed their bank accounts.[95] In other words, in this and likely other cases, price controls on payment-network fees appear to have had negative effects on financial inclusion.

Furthermore, it is unclear whether the systems established by Brazil and India are sustainable in their current form. One problem is that both rely on government subsidies. In the case of UPI, these come from the central government and must be financed by taxpayers. In the case of Pix, they come from the central bank (BCB), and are thus presumably paid, in part, out of fees from the banks and other organizations that are licensed by BCB. Because such subsidies are not directly related to the system’s performance, the system operator’s incentives are not necessarily aligned with users’ incentives (banks, fintechs, consumers). This may result in underinvestment in security (as arguably has been the case with Pix) and in other necessary ongoing upgrades.

Finally, potential conflicts of interest arise when a payment system’s operator is also the regulator of other payment systems, especially when there are no apparent mechanisms to address this conflict. For example, BCB both operates Pix and regulates payment-service providers, but does not appear to even acknowledge that this may be a conflict.[96] Such concerns were evident Brazil when WhatsApp proposed to introduce a payment app in the country, but the BCB prevented it from doing so until after it had launched Pix.[97]

Looking at changes in the proportion of adults with an account at a financial institution from 2011 to 2021 (Figure 3), the clear winners out of the four countries studied here are China and Kenya. While India and Brazil have also seen significant improvements, the growth has been less dramatic. In part, this is likely because both Brazil and India started from a higher base, but policy also likely played a role. It is probably too soon to see an effect from Pix in the data (Brazil only launch that system at the end of 2020) but UPI and demonetization appear to have had a detrimental effect on financial inclusion in India, with the proportion of adults having an account at a financial institution falling slightly after 2016

FIGURE 3: Share of Adults with an Account at a Financial Institution (2011 to 2021)

SOURCE: World Bank

Governments should not distort the market in these ways. Quite the opposite: they should be as neutral as possible. And they should limit themselves to the production of genuine public goods, such as courts and identity registers. Doing so will enable participation, competition, and innovation, which will drive financial inclusion.

Appendix: Trust and Counterparty Risk in Online-Payment Systems

As noted in Section I, entrepreneurs have developed several important technologies to improve trust and otherwise address counterparty risk in the online world, including user ratings, security protocols and fraud-detection systems, payment gateways, and online escrow systems. The following subsections discuss these technologies.

A. User Ratings as a Means to Enhance Trust

In February 1996, eBay founder Pierre Omidyar introduced the “feedback forum” on the site in order to encourage buyers and sellers to provide constructive feedback to one another. Accompanying the forum’s launch, he included the following note:

Most people are honest. And they mean well. Some people go out of their way to make things right. I’ve heard great stories about the honesty of people here. But some people are dishonest. Or deceptive. This is true here, in the newsgroups, in the classifieds, and right next door. It’s a fact of life. But here, those people can’t hide. We’ll drive them away. Protect others from them. This grand hope depends on your active participation. Become a registered user. Use our feedback forum. Give praise where it is due; make complaints where appropriate. [98]

The forum became extremely popular and Omidyar credits it with much of eBay’s subsequent success, because it enabled users to generate trust with one another.[99]

Such ratings systems have now become a common feature of e-commerce sites; indeed, they are a more-or-less ubiquitous and expected feature of larger sites. Moreover, the granularity of such ratings arguably makes them superior to other forms of regulation.[100]

B. SSL and Credit-Card-Security Protocols

From the perspective of merchants, one of the advantages of accepting credit-card payments is the credible commitment made by card issuers (embedded in the agreement those issuers make with the networks) that they will pay the merchant (via the acquiring bank, in the case of bank-issued cards) as long as the merchant has undertaken the necessary authentication.

Early internet-based transactions, however, posed some novel problems for card networks. In particular, there were security concerns on both sides of the market: issuers and cardholders were concerned about the risk that card details would be stolen and used fraudulently, while merchants and acquirers were concerned that they would be presented with stolen card details that, if used, would result in liability through chargebacks. For online payments to work required the development of solutions that would enhance security and trust on both sides.

Part of the solution came with the development in 1994 of the “secure sockets layer” (SSL) by Marc Andreessen and his team at Mosaic/Netscape.[101] SSL allowed for encrypted information to be sent between a web browser and the host server, preventing that information from being stolen. It was the precursor to modern online browsers, which use transport layer security (TLS) to transmit encrypted information. The incorporation of a padlock sign in a browser’s uniform resource locator (URL) helped to reassure users that their connection was secure, and information could be transmitted without fear of theft.

Another part of the solution came with card networks’ introduction of additional security protocols. Initially, individual payment networks developed their own standards, but these were superseded in 2004 with the development and implementation of an initiative known as the Payment Card Industry Data Security Standard (PCI DSS), which since 2006 has been an independent entity that includes all the major card networks.[102]

C. Credit-Card Guarantees and Chargebacks

Another important mechanism by which credit cards increase trust is card issuers’ commitment (as part of their agreement with networks, in the case of four-party cards) to guarantee payment to merchants on the condition that the merchants themselves meet certain criteria, including undertaking authentication checks. Meanwhile, card issuers may initiate a chargeback against a merchant on behalf of a cardholder in cases of fraud or theft (e.g., the cardholder’s information was stolen) or if the merchant failed to meet a cardholder’s legitimate expectations (e.g., it failed to provide goods or services in a timely fashion; the goods were damaged, defective, or missing parts; or the cardholder was charged an incorrect amount).[103]

D. Online-Payment Gateways as Trusted Intermediaries

Another important piece of the two-sided-market puzzle for online markets was solved in 1999 with the launch of PayPal. Originally conceived as a payment system in its own right, PayPal’s most important function (at least, until recently) has been as a payment gateway—a trusted intermediary that processes card payments.[104] PayPal is itself also a two-sided market, with merchants on one side and consumers on the other.[105] But it’s more complicated: in 2009, it partially opened up its APIs to third parties, thereby enabling others to build interoperable payments systems that would permit true peer-to-peer payments.[106]

E. Escrow and ‘Holds’

Parties to transactions frequently use trusted intermediaries to hold funds until specific conditions have been met, as documented in an escrow agreement. In the United States, escrow had traditionally been used mainly for transactions involving high-value idiosyncratic items, such as real-estate properties and paintings. This also appears to be the case thus far with online escrow services, such as those offered by escrow.com. The likely explanation is that escrow adds complexity and delays to a transaction; where financial remedies are adequate in the event of nonperformance, it is not an efficient solution.

Another reason escrow has not been used more widely for online purchases is that payment networks’ dual-message systems perform a similar function at much lower cost, enabling payments to be put on hold by the merchant until they are satisfied that the client will not issue a chargeback.[107] Some online payment gateways have similar “hold” systems that are very similar to escrow.[108]

But escrow has emerged as an important means to facilitate online commerce in China, where simple financial remedies facilitated by credit-card companies were not available to the majority of consumers. Within that context, online marketplace Alibaba established Alipay in 2003, which was initially an escrow-based system in which a purchaser’s funds were held by Alibaba until the purchase was delivered. As explored in Section 3, this proved to be a phenomenally successful strategy.

[1] Todd J. Zywicki, The Economics of Payment Card Interchange Fees and the Limits of Regulation, Int’l. Ctr. L. & Econ. (Jun. 2, 2010), available at https://laweconcenter.org/images/articles/zywicki_interchange.pdf.

[2] Marc Rysman, The Economics of Two-Sided Markets, 23  J. ECON. PERSP. 125, 130 (2009).

[3] See, e.g., Philip Moore, The State of the American Mall: Competitive, Attractive and Here To Stay, CORESIGHT RESEARCH (Jun. 27, 2023), https://coresight.com/research/the-state-of-the-american-mall-competitive-attractive-and-here-to-stay (noting that a new mall in Atlanta features a Nobu hotel and restaurant).

[4] Gunnar Camner, Emil Sjöblom, & Caroline Pulver, What Makes a Successful Mobile Money Implementation? Learnings from M-PESA in Kenya and Tanzania, GSMA (2009), available at https://www.gsma.com/mobilefordevelopment/wp-content/uploads/2012/06/mpesa_case_study9983.pdf (noting that airtime sharing was not observed in Kenya).

[5] See TONNY K. OMWANSA & NICHOLAS P. SULLIVAN, MONEY REAL QUICK: THE STORY OF M-PESA (Guardian Books, 2012); See also Enabling Mobile Money Transfer: The Central Bank of Kenya’s Treatment of M-Pesa, Alliance for Financial Inclusion (2010), available at afi_casestudy_mpesa_en.pdf.

[6] Banks Find Suite: Find Annual Historical Bank Data, FEDERAL DEPOSIT INSURANCE CORPORATION, https://banks.data.fdic.gov/explore/historical?displayFields=STNAME%2CTOTAL%2CBRANCHES%2CNew_Char&selectedEndDate=2022&selectedReport=CBS&selectedStartDate=1934&selectedStates=0&sortField=YEAR&sortOrder=desc (last visited Jun. 19, 2024); Geographical Outreach: Number of Automated Teller Machines (ATMs), Country Wide for United States, FED. RES. BANK ST. LOUIS (2016), https://fred.stlouisfed.org/series/USAFCACNUM (retrieved from FRED).

[7] Financial Access in Kenya: Results of the 2006 National Survey, Nairobi, Kenya: The Steadman Group Res. Div. (2007).

[8] Id.

[9] Omwansa & Sullivan, supra note 6.

[10] Id.

[11] Id.

[12] Experience M-PESA, Safaricom, https://www.safaricom.co.ke/personal/m-pesa/getting-started/experience-m-pesa (last visited Jun. 13, 2024); Driven by Purpose: 15 Years of M?Pesa’s Evolution, McKinsey & Co. (Jun. 29, 2022), https://www.mckinsey.com/industries/financial-services/our-insights/driven-by-purpose-15-years-of-m-pesas-evolution.

[13] M-Shwari FAQs, SAFARICOM, https://www.safaricom.co.ke/personal/m-pesa/credit-and-savings/m-shwari (last visited June 13, 2024)

[14] Tavneet Suri, Mobile Money, 9 Ann. Rev. Econ. 497, 515 (2017).

[15] Id.

[16] M-Pesa and Your Bank, SAFARICOM, https://www.safaricom.co.ke/personal/m-pesa/do-more-with-m-pesa/m-pesa-and-your-bank (last visited Jun. 13, 2024)

[17] M-Pesa Partners with Visa for Virtual Card Payments in Africa, PYMNTS (Jun. 2, 2022), https://www.pymnts.com/digital-payments/2022/m-pesa-partners-with-visa-for-virtual-card-payments-in-africa; About the M-PESA Globalpay Virtual VISA Card, Safaricon, https://www.safaricom.co.ke/mpesaglobalpay/about (last visited Jun. 13, 2024).

[18] 2021 FinAccess Household Survey, Cent. Bank of Kenya, Kenya Bureau of Nat’l Stat., & FSD Kenya (Dec. 2021), available at https://www.centralbank.go.ke/wp-content/uploads/2022/08/2021-Finaccesss-Survey-Report.pdf.

[19] Suri Tavneet & Jack William, The Long-Run Poverty and Gender Impacts of Mobile Money, 354 SCI. 1288 (2016),  https://www.science.org/doi/10.1126/science.aah5309.

[20] M-Pesa, VODAFONE, https://www.vodafone.com/about-vodafone/what-we-do/consumer-products-and-services/m-pesa (last visited Jun. 13, 2024).

[21] Improving People’s Everyday Financial Experience, Orange (Aug. 28, 2020), https://www.orange.com/en/groupe/nous-connaitre/linnovation-utile-et-source-de-progres-pour-tous/improving-peoples-everyday.

[22] Don Lee, China Charges into Credit Cards, L.A. Times (Oct. 22, 2008), https://www.latimes.com/archives/la-xpm-2008-oct-22-fi-chinacredit22-story.html; The card penetration rate was considerably higher in cities—around 18%, according to a contemporaneous Nielsen survey (rising to 22% the following year). See Survey: China Steps into Credit Card Era, CHINA DAILY (Aug. 9, 2004), https://www.chinadaily.com.cn/english/doc/2004-08/09/content_363531.htm.

[23] The Internet Timeline of China 1986~2003, CHINA INTERNET NETWORK INFO. CTR. (Jun. 28, 2012), https://www.cnnic.com.cn/IDR/hlwfzdsj/201306/t20130628_40563.htm#:~:text=On%20January%2016%2C%202003%2C%20China,to%20the%20Internet%20in%20China.

[24] Infra Section II.A and Appendix

[25] China: A Digital Payments Revolution, Consultative Group to Assist the Poor (Sep. 2019), https://www.cgap.org/research/publication/china-digital-payments-revolution.

[26] What Is Alipay?, CHECKOUT.COM (May 18, 2023), https://www.checkout.com/blog/what-is-alipay.

[27] Yichen Zhu & Sarah Hui Li, A Hangzhou Story: The Development of China’s Mobile Payment Revolution (Lee Kuan Yew Sch. Pub. Pol’y, Nat’l Univ. Sing., 2018), available at a-hangzhou-story.pdf.

[28] Payment Methods, ANTOM DOCS (Apr. 24, 2024), https://global.alipay.com/docs/ac/cashierpay/payment_method.

[29] Barry Elad, Alipay Statistics 2023 – Market Share, Facts and Marketing Trends, ENTERPRISE APPS TODAY (last updated Oct. 10, 2023), https://www.enterpriseappstoday.com/stats/alipay-statistics.html.

[30] WeChat Now Supports Payments Between Users and One-Click Payments, Finance Magnates (Jun. 24, 2014), https://www.financemagnates.com/fintech/payments/wechat-now-supports-payments-between-users-and-one-click-payments (Tencent, the owner of WeChat, established a payment system called TenPay in 2005. That failed to take off, but it remains the official licensed payment provider underpinning Weixin Pay).

[31] Eveline Chao, How WeChat Became China’s App for Everything, Fast Company (Feb. 1, 2017), https://www.fastcompany.com/3065255/china-wechat-tencent-red-envelopes-and-social-money.

[32] Matthew Fulco, The WeChat Economy, From Messaging to Payments and More, CKGSB Knowledge (Aug. 28, 2017), https://english.ckgsb.edu.cn/knowledge/article/the-wechat-economy-from-messaging-to-payments-and-more/#:~:text=Matthew%20Fulco%20Authors,August%2028%2C%202017.

[33] All You Need to Know About WeChat Mini-Programs, GOCLICK CHINA (Jun. 28, 2022), https://www.goclickchina.com/blog/wechat-mini-programs-all-you-need-know.

[34] Shubham Singh, 18 WeChat Statistics — Users & Revenue Data, DEMANDSAGE (May 27, 2024), https://www.demandsage.com/wechat-statistics.

[35]  Multiple Payment Methods, WECHAT PAY, https://pay.weixin.qq.com/wechatpay_guide/intro_method.shtml (last visited Jun. 19, 2024); In-Store Payment, ANTOM DOCS, https://global.alipay.com/docs/instorepayment (last visited Jun. 19, 2024).

[36] Payment Methods in China: How China Became a Mobile-First Nation, Daxue Consulting (Jan. 29, 2024), https://daxueconsulting.com/payment-methods-in-china.

[37] ANTGROUP, https://www.antgroup.com/en/business-development/digital-finance-tab-details/mybank (last visited Jun. 18, 2024).

[38] WEBANK, https://www.webank.com/en/characteristic (last visited Jun. 18, 2024).

[39] Laura He, Visa and Mastercard Can Now be Used on China’s Biggest Payment Apps, CNN (Jul. 21, 2023), https://edition.cnn.com/2023/07/21/tech/china-alipay-wechat-pay-international-credit-cards-intl-hnk/index.html.

[40] INDIA STACK, www.indiastack.org (last visited Jun. 18, 2024)

[41] Vision and Mission, UNIQUE IDENTIFICATION AUTH. INDIA, https://uidai.gov.in/en/about-uidai/unique-identification-authority-of-india/vision-mission.html (last visited Jun. 18, 2024).

[42] Aadhaar Verification API: Innovation in Identity Verification, PERFIOS (Dec. 5, 2024), https://www.perfios.com/post/aadhaar-verification-api-innovation-in-identity-verification; India Population 1950–2024, MACROTRENDS, https://www.macrotrends.net/countries/IND/india/population (last visited June 18, 2024).

[43] Enrolment Dashboard, UNIQUE IDENTIFICATION AUTH. INDIA, https://uidai.gov.in/aadhaar_dashboard/india.php (last visited Jun. 18, 2024).

[44] Unified Payment Interface (UPI), NPCI, https://www.npci.org.in/what-we-do/upi/product-overview (last visited Jun. 18, 2024); UPI Live Members, NPCI, https://www.npci.org.in/what-we-do/upi/live-members (last visited Jun. 18, 2024).

[45] MOBIKWIK, https://www.mobikwik.com/about (last visited Jun. 18, 2024).

[46] Mahesh Sharma, Payments Startup MobiKwik Launches Mobile Wallet As India’s Central Bank Acts To End Country’s Cash Dependence, TechCrunch (Sep. 27, 2013), https://techcrunch.com/2013/09/27/payments-startup-mobikwik-launches-mobile-wallet-as-indias-central-bank-acts-to-end-countrys-cash-dependence.

[47] Jon Russell, Indian Payments Startup MobiKwik Nabs $25M From Tree Line, Cisco, AmEx and Sequoia, TechCrunch (Apr. 7, 2015), https://techcrunch.com/2015/04/07/mobikwik-series-b.

[48] MobiKwik Continues Profitable Streak for Second Quarter in a Row, ECONOMIC TIMES (Oct. 5, 2023), https://economictimes.indiatimes.com/tech/technology/mobikwik-continues-profitable-streak-for-second-quarter-in-a-row/articleshow/104183594.cms?from=mdr.

[49] MobiKwik Consolidated Financial Statement, MOBIKWIK (2023), available at https://documents.mobikwik.com/files/investor-relations/statements/zaakpay/zaakpay-financials-sept2023.pdf; financial statements of other subsidiaries available at https://www.mobikwik.com/ir/subsidiary-financials; RBI notice of “in-principle authorization” available at https://www.rbi.org.in/Scripts/bs_viewcontent.aspx?Id=4236; Report on the Audit of Special Purpose Interim Financial Statements, TATTVAM & CO. (Dec. 31, 2023), available at https://documents.mobikwik.com/files/investor-relations/statements/zaakpay/zaakpay-financials-sept2023.pdf; Subsidiary Financials, MOBIKWIK,  https://www.mobikwik.com/ir/subsidiary-financials (last visited Jun. 19, 2024); Status of Applications Received from Online Payment Aggregators (PAs) under Payment and Settlement Systems Act, 2007, RES. BANK OF INDIA, https://www.rbi.org.in/Scripts/bs_viewcontent.aspx?Id=4236 (last updated Jun. 16, 2024).

[50] Id., Res. Bank of India.

[51] Press Release, Paytm’s Earning’s Release for Quarter and Year Ending March 2024, Paytm (May 22, 2024), available at https://paytm.com/document/ir/financial-results/Paytm_Earnings-Release_INR_Q4_FY24.pdf.

[52] Paytm’s Pricing, Paytm, https://business.paytm.com/pricing (last visited Jun. 19, 2024).

[53] Paytm Reaches 100 Million Users, Business World (Aug. 11, 2015), https://www.businessworld.in/article/Paytm-Reaches-100-Million-Users-/11-08-2015-84698.

[54] Patrick Jenkins, Paytm and Transmit Security Win FT Fintech Awards, FINANCIAL TIMES (Nov. 16, 2016), https://www.ft.com/content/1a92d762-ac1b-11e6-ba7d-76378e4fef24.

[55] Justin Rowlatt, Why India Wiped Out 86% Of Its Cash Overnight, BBC NEWS (Nov. 14, 2016), https://www.bbc.co.uk/news/world-asia-india-37974423.

[56] Paytm Surpasses 100 Million Monthly Transacting Users for the First Time in Q3 FY24, Livemint (Jan. 22, 2024), https://www.livemint.com/companies/news/paytm-surpasses-100-million-monthly-transacting-users-for-the-first-time-in-q3-fy24-11705932856486.html.

[57] PHONEPE, https://www.phonepe.com/about-us (last visited Jun. 19, 2024); Charlie Graham, PhonePe, CONTRARY RESEARCH (last updated Dec. 1, 2023), https://research.contrary.com/reports/phonepe.

[58] Press Release, Walmart to Invest in Flipkart Group, India’s Innovative eCommerce Company, Walmart (May 9, 2018), https://corporate.walmart.com/news/2018/05/09/walmart-to-invest-in-flipkart-group-indias-innovative-ecommerce-company.

[59] Sri Deepti, Flipkart Completes Separation from PhonePe, TECH IN ASIA (Dec. 23, 2022), https://www.techinasia.com/flipkart-completes-separation-phonepe.

[60] Alnoor Peermohamed, Flipkart Grows User Base to 100 million, Business Standard (Sep. 22, 2016), https://www.business-standard.com/article/companies/flipkart-grows-user-base-to-100-million-116092100216_1.html; FLIPKART CATAPULT, https://brands.flipkart.com/catapult-about (last visited Jun. 19, 2024).

[61] Graham, supra note 58.

[62] Ingrid Lunden, Google Debuts Tez, A Mobile Payments App for India that Uses Audio QR to Transfer Money, TECHCRUNCH (Sep. 17, 2017), https://techcrunch.com/2017/09/17/google-debuts-tez-a-mobile-wallet-and-payments-app-for-india.

[63] Google Pay API for India & Google Pay & Wallet Console, GOOGLE, https://support.google.com/google-pay-and-wallet-console/answer/10945206?hl=en#:~:text=Google%20Pay%20API%20for%20India%20operates%20on%20a%20unique%20India,used%20as%20your%20UPI%20ID (last visited Jun. 19, 2024). The version of Google Pay implemented in India is thus quite different to the version implemented in the United States and other markets.

[64] Manish Singh, Google Paves Way to Monetize Pay Users’ Data in India, TECHCRUNCH (Mar. 11, 2021), https://techcrunch.com/2021/03/11/google-pay-paves-way-to-tap-pay-users-data-in-india.

[65] Certificates of Authorisation Issued by the Reserve Bank of India Under the Payment and Settlement Systems Act, 2007 for Setting Up and Operating Payment System in India, RES. BANK OF INDIA, https://rbi.org.in/Scripts/PublicationsView.aspx?id=12043 (last updated Jun. 18, 2024).

[66] Manish Singh, WhatsApp Permitted to Extend Payments Service to 100 Million Users in India, TECHCRUNCH, (Apr. 13, 2022), https://techcrunch.com/2022/04/13/whatsapp-permitted-to-extend-payments-service-to-100-million-users-in-india.

[67] Id.

[68] Id.

[69] Press Release, Vodafone India Launches M-Pesa Pay for Merchants and Retailers, Vodafone (Jan. 5, 2017), https://www.vodafone.com/news/inclusion/m-pesa-pay-india; Aman Rawat, Vodafone Winds Up M-Pesa in India After Huge Losses, INC 42 (Jan. 22, 2020) https://inc42.com/buzz/vodafone-winds-up-m-pesa-in-india-after-huge-losses.

[70] Jackson Lott & Mona Sinha, M-Pesa’s Failure in India: Why Couldn’t Vodafone Replicate its Kenyan Success? An International Marketing Case Study, 6 KENNESAW J. UNDERGRADUATE RES. 1, 12-13 (2019), available at https://digitalcommons.kennesaw.edu/cgi/viewcontent.cgi?article=1160&context=kjur (“In Kenya, Safaricom had invested considerable effort and money over the years to build a critical mass of M-Pesa agents. These were essentially small shop owners across the country, including in villages, who could also serve as M-Pesa agents, available to sell mobile airtime as well as enable M-Pesa wallet top-ups and withdrawals, as needed. However, when Vodafone went to villages in India where most migrant workers came from, they did not find much economic activity. Unable to find small shop keepers who could serve as agents, they could not create as good a network of last-mile agents crucial to the service. Moreover, unlike Kenya, poor, unbanked, or underbanked consumers struggled to adopt self-service technologies and needed assistance. Creating awareness and driving behavior change amongst this segment of the population required tremendous resources in terms of time, money, and human capital that Vodafone would have to divert from its core business in India, i.e., cell phone service. Meanwhile, a slew of mobile payment upstarts entered the Indian market. Unlike Vodafone that aimed to service the unbanked and underbanked, these new entrants, like Paytm, reached out to middle class and affluent consumers who wanted the convenience and lived in urban and semiurban areas where agents could be easily appointed. Given the income profile of their target consumers, their technologies could be based on linking their app to the customers’ bank accounts. This demographic was markedly different from the unbanked/underbanked consumers that M-Pesa serviced who used feature/basic phones and dealt in cash to top-up or withdraw cash from their M-Pesa wallets. Moreover, Paytm had the backing of large investors like Soft Bank and Alibaba. One key trigger event for spurring adoption of digital payments in India was the demonetization announcement by the Prime Minister of India. However, a shortage of cash at that time due to high-value currencies being declared defunct meant that a cash dependent system like M-Pesa did not benefit from the surge of mobile payment adopters.”)

[71] Zaakpay and PhonePe have “in principle approval,” see supra note 47; Google has full approval, see supra note 63; Paytm has been asked to revise and resubmit its application, see Press Release, Update on PA License: Paytm Payments Services Receives Extension From RBI for Resubmission of Application & Remains Hopeful of Getting Necessary Approvals, Paytm (Mar. 27, 2023), https://paytm.com/blog/investor-relations/update-on-pa-license-paytm-payments-services-receives-extension-from-rbi-for-resubmission-of-application-remains-hopeful-of-getting-necessary-approvals.

[72] Hiral Thanawala, e-Rupi Vouchers Get a Boost from RBI Monetary Policy, MONEY CONTROL (Jun. 8, 2023), https://www.moneycontrol.com/news/business/personal-finance/e-rupi-vouchers-get-a-boost-from-rbi-monetary-policy-10766151.html.

[73] Roll Back Zero Merchant Discount Rate on UPI, RuPay Debit Card Payments, Industry Body Payments Council of India Writes to Finance Ministry, INDIAN EXPRESS (Jan. 23, 2022),  https://indianexpress.com/article/business/banking-and-finance/merchant-discount-rate-rollback-on-upi-rupay-debit-cards-7737229.

[74] Pratik Bhakta, Fintechs Await Government Word on MDR Subsidy Allocation, ECONOMIC TIMES (Feb. 22, 2024), https://economictimes.indiatimes.com/tech/technology/fintechs-await-government-support-for-promoting-digital-payments-for-current-fiscal/articleshow/107891943.cms?from=mdr.

[75] Julian Morris, Todd J. Zywicki, & Geoffrey A. Manne, The Effects of Price Controls on Payment-Card Interchange Fees: A Review and Update, Int’l. Ctr. L. & Econ. (Mar. 3, 2022), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4063914.

 

[76] Oi Paggo, LINKEDIN, https://www.linkedin.com/company/paggo/about (last visited Jun. 19, 2024); Paggo, WIKIPEDIA, https://pt.wikipedia.org/wiki/Paggo (last visited Jun. 19, 2024), (information unverified because the links are broken).

[77] Oi Paggo: A Disruptive Brasileiro Credit Play, STL PARTNERS (Jun. 2008), https://stlpartners.com/insights/oi_paggo_a_disruptive_brasilei.

[78] Umair Bashir, Most Used Mobile Payments by Brand in Brazil 2023, Statista (Feb 13, 2024), https://www.statista.com/forecasts/1226566/most-used-mobile-payments-by-brand-in-brazil#:~:text=We%20asked%20Brazilian%20consumers%20about,in%20our%20Consumer%20Insights%20tool.

[79] Crescimento do PicPay: 34 Milhões Usam o App No Dia a Dia, PICPAY (Dec. 14, 2021), https://blog.picpay.com/picpay-crescimento/?utm_source=iupana&utm_medium=web&utm_campaign=010822EN.

[80] Fabiane Z. Menezes, PicPay, Latin America‘s Largest Digital Wallet, Is Now A Bank, THE BRAZILIAN REP. (Jul. 13, 2022), https://brazilian.report/liveblog/2022/07/13/picpay-digital-wallet-bank; PicPay 1H23 Earnings Release, PICPAY (Aug. 2023), at 11, https://api.mziq.com/mzfilemanager/v2/d/f4269298-97ef-427d-ac29-04f9d6a6054a/765c070b-b1ca-b07b-d699-b6a7eb7be191?origin=1.

[81] PicPay, supra note 79, at 2.

[82] Id., at 4.

[83] PayPal User Agreement, PAYPAL, https://www.paypalobjects.com/marketing/ua/pdf/BR/en/ua-032122.pdf?locale.x=en_BR (last visited Jun. 19, 2024).

[84] Mercado Libre SEC Proxy Statement (2023), at 6, https://api.mziq.com/mzfilemanager/v2/d/098a2d95-0ea8-4ed5-a340-d9ef6a2b0053/bfba3d21-84f6-7263-8abc-f32f54dc122d?origin=1.

[85] Id., at 7.

[86] Our History, PAGBANK, https://investors.pagbank.com/about-us/our-history (last visited Jun. 19, 2024).

[87] Our History, PAGSEGURO, https://international.pagseguro.com/about-us (last visited Jun. 19, 2024).

[88] Google Wallet Help, GOOGLE, https://support.google.com/wallet/answer/12059326?hl=en-GB&co=GENIE.CountryCode%3DBR (last visited Jun. 19, 2024).

[89] The Acceleration of Financial Inclusion During the COVID-19 Pandemic: Bringing Hidden Opportunities to Light, MasterCard (Oct. 12, 2020), available at https://www.mastercard.us/content/dam/public/mastercardcom/na/us/en/banks-and-credit-unions/other/mastercard-financial-inclusion-during-covid-whitepaper-20201012.pdf.

[90] Vinicius Colares, Bolsa Merenda PagBank 2021: CADASTRO, VALOR e como RECEBER o BENEFÍCIO…, PRONATEC (Apr. 11, 2021), https://pronatec.pro.br/bolsa-merenda-pagbank-2021-cadastro.

[91] Eduarda Andrade, CAIXA Tem Libera Transferência do Auxílio e Bolsa Família por PIX e TED, FDR (Aug. 18, 2021), https://fdr.com.br/2021/08/18/caixa-tem-libera-transferencia-do-auxilio-e-bolsa-familia-por-pix-e-ted; PICPAY, https://picpay.com/pix (last visited Jun. 19, 2024); Pix, PAGBANK, https://faq.pagseguro.uol.com.br/pix/408#rmclgBank (last visited Jun. 19, 2024).

[92] Manish Singh, Google‘s New Plan to Push Google Pay in India: Cashback Incentives in Android Apps, TECHCRUNCH (May 16, 2019), https://techcrunch.com/2019/05/16/google-pay-india-android-cashback.

[93] A Peek Inside PayPal’s Two-Sided Network of Consumers and Businesses, PayPal (Oct. 27, 2022), https://newsroom.paypal-corp.com/2022-10-27-A-Peek-Inside-PayPals-Two-Sided-Network-of-Consumers-and-Businesses.

[94] PicPay, supra note 79, at 3.

[95] Vladimir Mukharlyamov & Natasha Sarin, Price Regulation in Two-Sided Markets: Empirical Evidence from Debit Cards, SSRN (Nov. 24, 2022), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3328579.

[96] Julian Morris, Central Banks and Real-Time Payments: Lessons from Brazil’s Pix, Int‘l. Ctr. L. & Econ. (Jun. 1, 2022), available at https://laweconcenter.org/wp-content/uploads/2022/06/Lessons-from-Brazils-Pix.pdf.

[97] Julian Morris, Pixtopia Is Not Real, TRUTH ON THE MARKET (Jun. 2, 2022), https://truthonthemarket.com/2022/06/02/pixtopia-is-not-real.

[98] Pierre Omidyar, Founder’s Letter, eBay (Feb. 26, 1996), https://pages.ebay.co.uk/services/forum/feedback-foundersnote.html.

[99] Pierre Omidyar – On Innovation, The Henry Ford (Apr. 29, 2010), https://www.youtube.com/watch?v=RKVmsifohgM&t=295s.

[100] Julian Morris, Consumer Protection in the 21st Century, Int‘l. Ctr. L. & Econ. (Feb. 24, 2023), https://laweconcenter.org/resources/consumer-protection-in-the-21st-century.

[101] The original version of SSL was easily hacked, and thus was not launched publicly. Version 2.0, which was more secure, was included in releases of Netscape in late 1994. See Huzaifa Sidhpurwala, Evolution of the SSL and TLS protocols, RED HAT (Nov. 16, 2016), https://access.redhat.com/blogs/766093/posts/2758801; Phillip Hallam-Baker, Crypto Standards vs. Engineering Habits – Was: NIST About to Weaken SHA3?, ianG.org (Oct. 4, 2013), https://www.metzdowd.com/pipermail/cryptography/2013-October/018041.html.

[102] PCI SECURITY STANDARDS COUNCIL, https://www.pcisecuritystandards.org (last visited Jun. 19, 2024).

[103] Troy Segal, What Is a Chargeback? Definition, How to Dispute, and Example, Investopedia (Dec. 15, 2023), https://www.investopedia.com/terms/c/chargeback.asp.

[104] There are now many other online-payment gateways, including Worldpay, Stripe, and Square/Block.

[105] The company even talks about itself as a “two-sided network,” see PayPal, supra note 93.

[106] Sebastian Rupley, PayPal’s (Partially) Open Platform to Usher in New Payment Models & Apps, GIGAOM (Nov. 3, 2009),  https://web.archive.org/web/20110102142203/http://gigaom.com/2009/11/03/paypals-partially-open-platform-to-usher-in-new-payment-models-apps.

[107] Credit/Debit Authorisation Holds Explained, GOCARDLESS (last updated Mar. 2022), https://gocardless.com/guides/posts/credit-debit-authorisation-holds-explained.

[108] How to Resolve Payment on Hold or Unavailable Funds, PAYPAL (Oct. 10, 2023),  https://www.paypal.com/uk/brc/article/funds-availability.

ICLE Chief Economist Brian Albrecht was a guest on Marketplace to discuss how market definition is determined in antitrust cases like the Federal Trade Commission’s (FTC) challenge to the Kroger-Albertsons merger. Audio of the full clip is embedded below.

ICLE Editor-in-Chief R.J. Lehmann was a guest on the Insurance Hour to discuss California’s Proposition 103, its impact on insurance regulation, and the controversial intervener system. Video of the full episode is embedded below.

ICLE President Geoffrey A. Manne was a guest on the Tech Policy Podcast to discuss the flaws in the Federal Trade Commission’s antitrust lawsuit against Meta. Audio of the full episode is embedded below.

WCPO – ICLE Senior Scholar Eric Fruits was quoted by Cincinnati television station WCPO in a story about a petition from several state attorneys general asking the Federal Trade Commission to challenge to the merger of Kroger and Albertsons. You can read full piece here.

Kroger has a 75% chance of completing the proposed merger, but it might have to fight the FTC in court to get the deal done, said Eric Fruits, an antitrust expert and senior scholar for the International Center for Law & Economics in Portland Oregon.

“For the past 25-30 years, almost every single grocery merger has been allowed to go through with these spinoffs, or what are known as divestitures,” said Fruits, a Sycamore High School graduate who recently co-authored a detailed analysis of the likely legal arguments in the case. “My guess is that’s what Kroger and Albertsons are going to offer. The real question is whether or not the Federal Trade Commission will take that offer. And if they don’t, they could go to court and the judge could say, ‘This seems like a reasonable remedy. We’ll let you spin off the stores.’”

When Kroger announced the planned merger last October, it said it hoped to complete the deal by early next year. Fruits doesn’t think that can happen if the case goes to court.

“I would think that there’s bigger fish to fry than the Kroger-Albertsons merger,” said Fruits. “But you just don’t know what sort of political pressure these federal trade commissioners are under to try to show that they’re doing something big. And this is something big that they can try to block.”

ICLE President Geoffrey Manne joined the Free the Economy podcast to discuss the recent actors and writers’ strikes, competition in Hollywood, and the future of streaming entertainment. Audio of the full episode is embedded below.

ICLE Editor-in-Chief R.J. Lehmann was cited by Jalopnik in a story about former President Donald Trump’s proposal to cut auto insurance rates. You can read the full piece here.

Insurance is a private industry — a regulated one, sure, but not one where prices can be decided by fiat. Editor-in-chief at the International Center for Law & Economics, Ray Lehmann, said as much to Insurance News Net, though with a slightly more optimistic view towards Trump’s comments than Hartwig…

ICLE Senior Scholar Eric Fruits was quoted by Supermarket News about the prospects for the Federal Trade Commission in its lawsuit to block the merger of Kroger and Albertsons. You can read the full piece here.

Supermarket News reached out to Eric Fruits, senior scholar at the International Center for Law & Economics, and David Balto, an antitrust lawyer who served as the policy director of the Bureau of Competition for the FTC between 1998 and 2001, and both believe an injunction on the $24.6 billion merger is likely to happen.

ICLE Editor-in-Chief R.J. Lehmann was quoted by Insurance News Net about former President Donald Trump’s pledge to slash auto insurance rates. You can read the full piece here.

InsuranceNewsNet talked to Ray Lehmann, editor-in-chief and senior fellow at the International Center for Law & Economics, to find out how much impact a president and his administration could have on auto insurance costs. Premiums increased 18.6% from July 2023 to July 2024, according to Consumer Price Index, with the average cost of car insurance at $2,348, according to research from Bankrate.

“It’s interesting,” said Lehmann, “You look at the last decade going into the pandemic –  about 2014 to 2017 – you were seeing average increases of about 7% a year. And then 2018 to 2021 – which includes the pandemic – was basically flat, maybe a 1% increase a year, but generally none. And then we’ve had several years in a row of these double-digit increases. If you normalize that over time, if we didn’t have the flat period, 7% a year would actually look normal.”

Rate hikes ‘more like catch-up growth’

Lehman said the recent rate jumps are ”more like catch-up growth.” During the earlier period going into the pandemic, said Lehmann, the industry was running combined ratios in the mid-nineties: 93%, 94%, 98%. (The combined ratio – also called “the combined ratio after policyholder dividends ratio” – is a measure of profitability used by insurance companies. A number below 100% means the company is making an underwriting profit. A number above 100% means the company is taking a loss, paying out more in claims than it’s taking in from premiums.)

“In the pandemic year of 2020, that dropped like a rock, because nobody was driving,” said Lehmann. [Insurers] had a combined ratio of about 89%. And then we’ve been over 100% the last three years with 2022 being 113%.

“So, 2022 was a really extraordinarily high year for combined ratio,” said Lehman. “It went down a lot in 2023, but was still over 100%. I think it’s probably like 102%. We’re seeing some normalization in the cost trends, but the overall drivers of things we’ve known is that you have social inflation and the cost of medical care for casualty claims.”

ICLE Editor-in-Chief R.J. Lehmann was quoted by Insurance Journal in a story about former President Donald Trump’s proposal to slash auto insurance rates. You can read the full piece here.

Whether Trump intended for the post to elicit the types of comments it has received, many pointed to the illegal immigrant population as a factor in high auto insurance rates. Hartwig said he was unaware of any study related to immigrants as a driver of high auto insurance costs, or being associated with issues related to uninsured motorist coverage. R.J. Lehmann, senior fellow at the International Center for Law & Economics, said the same.

“I wouldn’t jump to that conclusion,” Lehmann said an interview. “Uninsured or underinsured drivers may be a factor, but [in terms of illegal immigrants] I don’t know of a study. It doesn’t come up. An uninsured driver could be anybody. Rates have gone up due to claims affecting insurers’ results.”

Lehmann explained that rates were going up around 7% per year until the pandemic, when the volume of vehicles on the road dropped significantly and insurers weren’t increasing rates at all. But since vehicles returned to the roads, studies have shown driving behavior has taken a downturn and motor vehicle deaths have increased. Insurers are “catching up,” said Lehmann, and recent reports suggest they may have. S&P Global Market Intelligence forecasted a “dramatic return to underwriting profitability” with a personal auto 2024 combined ratio of 98.4, down from 104.9 in 2023 and 112.2 in 2022.

…Lehmann said he believes “Trump isn’t thinking about federal regulation,” and there is little that he or Congress can do with the system of state regulation firmly in place. However, a president could fund road infrastructure improvements or campaigns with agencies like the National Highway Traffic Safety Administration to promote safety.

ICLE Nonresident Scholar Roger Bate was cited by The Business Post in a story about his recent issue brief looking at the problem of unapproved vaping products on the market. You can read the full piece here.

A new paper by Roger Bate, a nonresident scholar at the International Center for Law and Economics (ICLE), has raised the alarm over the safety and quality of unapproved tobacco harm reduction (THR) products in the US market.

Published by ICLE, the paper highlights a troubling gap between regulatory policies and market realities, which could put consumers at risk, read a press release.

The ICLE is a nonprofit, nonpartisan research center that promotes the use of law and economics to inform public policy. Its goal is to advance data-driven solutions for efficient policies that promote consumer welfare and economic growth.

Bate’s research, conducted at 14 gas stations in the Philadelphia suburbs, found that over 99% of the THR products on sale were either illegal or operating in a legal grey area.

ICLE Director of Law & Economics Programs Gus Hurwitz was quoted by the New York Post in a story about TikTok’s legal challenge to a law that would force its Chinese parent ByteDance to divest the company. You can read the full piece here.

“I had expected TikTok to face an uphill battle at this hearing, but the questioning they faced was much more critical than anticipated,” said Gus Hurwitz, a senior fellow at the University of Pennsylvania Carey Law School. “The judges seemed quite skeptical that the law merits strict scrutiny, or even intermediate scrutiny.”

“It’s hard to make predictions about these things, but after today’s argument I would say the smart money is on a unanimous and very unambiguous loss for TikTok,” added Hurwitz, who noted the judges “seemed to take the national security arguments very seriously.”

ICLE Director of Competition Policy Dirk Auer and Director of Law & Economics Programs Gus Hurwitz were quoted by The Drum in a story about Google’s failed appeal of the European Commission’s shopping inquiry. You can read the full piece here.

And in the telling of Dirk Auer, the director of competition policy at the International Center for Law & Economics, for these tech giants, “the DMA is is way stronger than competition law has much tougher provisions.“

…“It is remarkable that [this CJEU] opinion, demonstrating European hostility to innovation and backwards attitude towards the global, digital economy, would come out only one day after that report was released,” says Gus Hurwitz, senior fellow and academic director at University of Pennsylvania’s Center for Technology, Innovation, and Competition.

…Whether the CJEU’s verdict today will have any bearing on these other appeals remains to be seen, but Auer, for one, is skeptical. ”It’s not impossible, but it’s unlikely that on a subsequent Google Android appeal, [for example] the higher court, the Court of Justice, would set aside the lower court’s assessment.”

…Similar investigations are also underway in the UK, underscoring the global nature of regulatory efforts to rein in big tech’s market power. In Auer’s telling, ”there is a huge appetite, both in the US and the EU, to go after tech firms.”

ICLE Chief Economist Brian Albrecht was cited by Forbes in a piece about the decision in the Google adtech antitrust case. You can read the full piece here.

Brian Albrecht is an applied economic theorist in competition and information. Albrecht is also Chief Economist at the International Center for Law & Economics, earned his PhD in economics from the University of Minnesota in 2020, and was Assistant Professor at Kennesaw State University.

ICLE Senior Scholar Julian Morris was quoted by the Center Square in a piece about a Georgia House of Representatives study on the impact of credit-card fees on merchants and consumers. You can read the full piece here.

“Over the course of the past 20 years, [there] has been a significant shift away from both cash and checks and towards using card payments,” Julian Morris, senior scholar with the International Center for Law and Economics, told committee members.’

 

ICLE President Geoffrey A. Manne was cited by the Washington Examiner in a piece about Google’s loss in the U.S. Justice Department’s antitrust trial concerning Google search. You can read the full piece here.

“But for Microsoft’s lighter causation standard to apply, it must also involve a nascent threat and conduct proven sufficient to prevent rivals from achieving minimum efficient scale. Arguably, neither is true in Google Search,” wrote Geoffrey A. Manne of the International Center for Law and Economics.

ICLE Chief Economist Brian Albrecht was cited by the Northwoods River News in a column about Vice President Kamala Harris’ price controls proposals. You can read the full piece here.

Writing on Substack after Harris’s policy pronouncements, Brian Albrecht, the chief economist at the International Center for Law & Economics, said her policy was just that — price controls.

“Some might argue that calling Harris’s proposal ‘price controls’ is unfair or hyperbolic,” Albrecht wrote. “After all, she’s not directly setting prices, right? Any policy that gives the government the power to decide what price increases are ‘fair’ or ‘unfair’ is effectively a price control system. It doesn’t matter if you call it ‘anti-gouging,’ ‘fair pricing,’ or ‘consumer protection’ — the effect is the same.”

When bureaucrats, not markets, determine acceptable prices, we’re dealing with price controls, Albrecht wrote.

Albrecht listed several reasons why that was the case. First, vague definitions lead to broad interpretations.

“Without a clear, objective definition of ‘price gouging,’ regulators would have wide latitude to decide what constitutes an ‘unfair’ price increase,” he wrote. “It typically refers to unfairly high prices during emergencies. This isn’t about emergencies. This ambiguity could easily lead to de facto price controls across a wide range of grocery items.”

Second, Albrecht wrote, enforcement requires price monitoring.

“To enforce a price gouging ban, the government would need to monitor price changes,” he wrote. “This creates a system where prices are effectively controlled by bureaucrats rather than market forces.”

Finally, Albrecht asserted, companies will err on the side of caution.

“Faced with potential penalties for ‘gouging,’ grocery retailers and suppliers are likely to be overly cautious about any price increases, even when justified by rising costs or changes in supply and demand,” he wrote. “This chilling effect is tantamount to informal price controls.”

That chilling effect rivals that of explicit price ceilings, even if not true ceilings, Albrecht asserted.

“If companies face severe penalties for ‘excessive’ price increases (however that’s defined), they’ll err on the side of caution and keep prices artificially low,” he wrote. “This informal price control can be just as damaging to market efficiency as a government-mandated price ceiling.”

ICLE Senior Scholars Julian Morris and Ben Sperry were cited in a blog post at Conversable Economist about their recent paper on the costs and benefits of various payment methods. You can read the full piece here.

There was a time, not all that long ago, when the primary options for making everyday payments were cash and checks. Those times are past. Julian Morris and Ben Sperry discuss the tradeoffs and consequences in “The Cost of Payments: A Review,” (International Center for Law & Economics Working Paper, August 28, 2024). They begin with this anecdote…

ICLE Chief Economist Brian Albrecht was quoted by GZero in a story about California’s efforts to regulate artificial intelligence. You can read the full piece here.

Brian Albrecht, the chief economist at the International Center for Law & Economics, was not surprised by the companies’ willingness to share their models with the government. “This is a very standard response to expected regulation,” Albrecht said. “And it’s always tough to know how voluntary any of this is.”

…On Aug. 28, the state’s legislature passed the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act, or SB 1047, which aims to establish “common sense safety standards” for powerful AI models. Written by California State Sen. Scott Wiener, and supported by AI pioneers like Geoffrey Hinton and Yoshua Bengio, the bill has divided Silicon Valley companies. Albrecht said that what’s been proposed by California is much closer to the European model of AI regulation — the EU’s AI Act passed in March — while Washington hasn’t yet adopted a unified view on how the technology should be regulated.

Threads from ICLE scholars on trending issues for the month of August 2024.

An updated version of my @LawEconCenter wp on “The #Privacy/#Antitrust Curse: Insights from #GDPR Application in Competition Law Proceedings” has been published in the Antitrust Bulletin https://t.co/bkCFaaPkRz. 1/4

— Giuseppe Colangelo (@GiuColangelo) September 30, 2024

#Antitrust #AIPartnerships

The @CMAgovUK announced today that Amazon’s partnership with Anthropic “does not qualify for investigation under the merger provisions of the Enterprise Act 2002”. A ?https://t.co/TYEt7PHcaO

— Mario Zúñiga (@MZunigaP) September 27, 2024

#Antitrust #AIPartnerships

The @CMAgovUK announced today that Amazon’s partnership with Anthropic “does not qualify for investigation under the merger provisions of the Enterprise Act 2002”. A ?https://t.co/TYEt7PHcaO

— Mario Zúñiga (@MZunigaP) September 27, 2024

ANNOUNCEMENT: Alden Abbott and I have co-edited a book on Artificial Intelligence and Competition Policy for @Concurrences. The volume features 20 contributions, organized into three sections:
? “Market Dynamics, Mergers, and Partnerships in AI”
? “AI Challenges for Competition… pic.twitter.com/QJVKPHI7gG

— Thibault Schrepel (@ProfSchrepel) September 23, 2024

New @LawEconCenter working paper with Alba Ribera Martinez on “The Secret (#Metrics) of DMA’s Success.”
While much has been written about the #DMA, one key question remains underexplored: How do we determine its #success? 1/2 https://t.co/Q9i1rz6vAP

— Giuseppe Colangelo (@GiuColangelo) September 18, 2024

The BEAD Program is supposed to deploy broadband to millions of unserved homes. But so far, not a single site has been connected under the $42.5 billion program.

The NTIA's "fiber over all" approach is one reason why.

A ? … pic.twitter.com/QGUmk0Cv26

— Eric Fruits, Ph.D. (@ericfruits) September 11, 2024

What do we want? Technology neutrality! When do we want it? Now! @ericfruits and my latest @TOTMblog on our comments on the NTIA's guidance on "alternative technologies." (a ?)https://t.co/XLxXOLMNNR

— Ben Sperry (@RBenSperry) September 11, 2024

The DOJ's antitrust trial against Google's adtech business just kicked off.

The market is technically complicated, which makes the case complicated. What do you need to know?

Luckily, I have a new primer to get you up to speed! https://t.co/g5QsjF233b pic.twitter.com/6qhqy7IkO3

— Brian Albrecht (@BrianCAlbrecht) September 10, 2024

Draghi: EU’s regulatory stance towards tech co’s hinders innovation. EU now has over 100 tech-focused laws & over 270 regulators active in the digital space. Many laws take a precautionary approach, dictating business practices ex ante. We’ve been saying this for years. A ? pic.twitter.com/Ddk6oJRZop

— Lazar Radic (@laz_radic) September 10, 2024

As arguments begin today in Alexandria, Virginia in @JusticeATR's bench trial alleging that Google has maintained a monopoly in advertising technology, we wanted to share some @LawEconCenter resources that might be useful for those watching the case.

— Int'l Ctr Law & Econ (@LawEconCenter) September 9, 2024

1/14 ENEL, the Italian electricity giant and former state monopoly, complained to @antitrust_it that Google was limiting access of its app on Android Auto. In the opinion delivered today in Case C 233/23, AG Medina has sided with the Italian antitrust watchdog. A?

— Lazar Radic (@laz_radic) September 5, 2024

#Antitrust #AI #AIPartnerships

Last week, @CadeGovBr, ?? competition agency, opened investigations into Amazon, Google and Microsofts deals with AI startups.

Is this warranted or just a political move? A ?: pic.twitter.com/0ZRcP25vQe

— Mario Zúñiga (@MZunigaP) September 3, 2024

Huge defeat for the Commission. It tried to push the limits of the law. Now the Court is rebuking it in no uncertain terms.

The big question is how much this will impact merger enforcement in Europe (as well as GRAIL & Illumina)?

My guess is very little.

? https://t.co/s58uXQ7OEM

— Dirk Auer (@AuerDirk) September 3, 2024

My new @LawEconCenter wp on “Trendy #antitrust for digital markets: are market #investigations the new black?” is forthcoming @JECLAP https://t.co/K7iplZGMCH. 1/3

— Giuseppe Colangelo (@GiuColangelo) September 2, 2024