LET ME RIDE: NO SHORT-CUTS IN THE ANTITRUST ANALYSIS OF RIDE HAILING

By Daniel Bitton, David Pearl and Patrick Shaw¹

The ride-hailing industry, and gig economy more generally, is under pressure. On the one hand, there has been a push in the California legislative and judiciary branches to treat drivers and other gig economy participants as employees of gig economy companies. On the other hand, ride-hailing companies like Uber have faced antitrust lawsuits, including claims that their centralized pricing algorithms amount to Section 1 Sherman Act price fixing among drivers.

Some have suggested that ride-hailing companies like Uber and Lyft face a catch-22: if their drivers are classified as employees they would have to comply with costly labor regulations, but if their drivers are deemed to be independent contractors Uber and Lyft could be considered to be illegally fixing prices because the companies determine how much drivers charge riders on their platforms.² In the authors’ view, this is a false analytical choice. It is certainly true that if drivers are employees of a ride-hailing company, the company could not be liable for fixing prices with them under Section 1. But, it does not logically follow that ride-hailing companies incur Section 1 liability for setting prices of rides purchased through their app if drivers are, instead, viewed as independent contractors. To the contrary, there is a legitimate question whether ride hailing pricing amounts to unilateral conduct, a vertical restraint or a horizontal restraint—all of which differ analytically for purposes of determining potential antitrust liability under Section 1.

Moreover, ride-hailing apps and pricing algorithms do not fit the types of historical practices that always or nearly always restrict competition or decrease output—conduct for which courts have deemed per se antitrust analysis appropriate. Given their relatively recent advent, it is the view of these authors that the rule of reason should be applied to analyze whether pricing algorithms like the ones Uber and Lyft and other gig economy companies use provide procompetitive market benefits.

Ride-hailing apps and pricing algorithms have injected new competition and innovation into what were historically monopolistic markets insulated from competition and innovation by government licensing rules. Additionally, there are usually multiple ride-hailing options in each city, in addition to cab service, and output and efficiency of transportation has increased, in part because ride hailing pricing algorithms efficiently match supply and demand for rides. Ride-hailing companies thus have brought about greater competition and innovation in transportation services, as well as new, flexible, freelance work opportunities for gig-economy workers.

[Page 20]

The rule of reason offers a more searching³ and far more appropriate analysis in the ride hailing context because it takes into account the beneficial effects of the pricing algorithms that Uber, Lyft and other gig economy companies use. A recent court decision concluding the opposite on a motion to dismiss seems to miss the mark and ignores Supreme Court precedent.

I. EMPLOYEES OR INDEPENDENT CONTRACTORS?

This past April, the National Labor Relations Board ("NLRB") published an advisory memorandum providing support for Uber and Lyft’s position that their drivers are independent contractors, and thus not subject to the protections offered by the National Labor Relations Act.⁴ However, in 2018, the California Supreme Court in the case Dynamex Operations West, Inc. v. Superior Court⁵ adopted the "ABC" test for determining whether delivery drivers were independent contractors or employees. The test strongly favors a determination that a driver is an employee. On the heels of the Dynamex decision, the California State Legislature passed Assembly Bill 5, a bill that codified the "ABC" test in determining whether workers are employees or contractors.⁶

In an Op-Ed in the San Francisco Chronicle, Uber’s CEO and Lyft’s founders have explained that "a change to the employment classification of ride-share drivers would pose a risk to [their] business."⁷ Moreover, as they also explain in their Op-Ed, reclassification would ultimately be harmful to drivers too because many prefer the freedom and flexibility that the current system allows rather than the more rigid schedules associated with traditional employment. Additionally, the Op-Ed mentions that many drivers are often driving as a means to supplement income, on top of another job or venture. Whether drivers would actually qualify as employees of ride-hailing companies like Uber or Lyft under the new California state legislation is still an open question.

Some have suggested that gig economy companies like Uber and Lyft face an even greater risk if they do not classify their drivers as employees: Section 1 antitrust liability.⁸ For example, in Meyer v. Kalanick⁹, plaintiff Spencer Meyer filed a class action against the co-founder and then-CEO of Uber, Travis Kalanick, alleging that Kalanick had facilitated an illegal price-fixing conspiracy among Uber and its drivers by requiring drivers to use Uber’s pricing algorithm to set the prices that the drivers would charge riders, in violation of Section 1 of the Sherman Act.¹⁰ The plaintiff’s complaint relied, in part, on Uber’s statement that it "is not a transportation company and does not employ drivers."¹¹ Although the case was ultimately settled through binding arbitration,¹² the plaintiff’s claims did survive Kalanick’s motion to dismiss.¹³

[Page 21]

It is certainly the case that Uber or Lyft cannot be illegally fixing prices if their drivers are employees: the antitrust laws recognize an exemption for intra-firm agreements. This exemption is based on the idea that firms are a single entity, and you cannot have an illegal agreement when there is only one entity involved. As the United States Supreme Court wrote in the case Copperweld Corp. v. Independence Tube Corp.,¹⁴ "officers or employees of the same firm do not provide the plurality of actors imperative for a Section 1 conspiracy."

However, that does not mean that the centralized pricing algorithms that Uber and Lyft deploy to set ride fares for drivers amount to unlawful price fixing if drivers are independent contractors. Ride-hailing pricing does not easily fall into a traditional antitrust paradigm. We therefore consider whether pricing algorithms like the ones Uber and Lyft deploy for their ride-hailing services should be classified as unilateral single firm pricing, or as a horizontal or vertical restraint, as that can have a significant impact on the mode of antitrust scrutiny: rule of reason or per se. Ultimately, even if one classifies ride hailing pricing as horizontal price restraints, in these authors’ view the Supreme Court’s decision in Broadcast Music Inc. v. Columbia Broadcasting System, Inc.¹⁵ ("BMI") suggests that the rule of reason antitrust analysis should apply. And, as discussed below, there are a number of reasons to believe that a rule of reason analysis would lead to a finding that the pricing algorithms of ride hailing apps have had net procompetitive effects.

II. SHOULD CENTRALIZED RIDE-HAILING PRICING BE TREATED AS UNILATERAL PRICING?

Ride-hailing companies like Uber and Lyft offer riders an app through which they can hail rides when and where they need it, based on a fare the ride-hailing company determines. Uber’s algorithms set the fare when a user hails a ride on Uber’s platform,¹⁶ and Lyft’s algorithm sets the fare when a user hails a ride on Lyft’s platform. The algorithm sets the fare based on a number of different factors such as the length and distance of the trip, as well as rider demand at the time the user hails the ride.¹⁷ Uber charges a user the fare for the ride, retains a percentage of the fare, and then remits the remainder to the driver.¹⁸ Lyft’s approach and pricing works largely the same way.¹⁹

[Page 22]

In a market or antitrust analysis, drivers could be viewed as input suppliers to Uber’s and Lyft’s transportation platform services to consumers. Alternatively, Uber and Lyft could be viewed as two-sided platforms,²⁰ if one considers Uber and Lyft as providing both a service to riders and a service to drivers (i.e. to find riders, drive them, and receive payment).

When viewed that way, the pricing algorithms Uber and Lyft use could be treated as plain unilateral pricing of their transportation platform services to their customers: (a) the ride-hailing fare for consumers, (b) a fee for the rider access service for the drivers (the percentage of the fare that Uber and Lyft take vs. pass onto the drivers), or (c) both. In an analogous context, limousine-for-hire companies set the fare for their limo services centrally, even though in many cases their drivers are independent contractors.²¹ We are not aware of allegations that these arrangements are hub and spoke conspiracies between limousine-for-hire companies and their drivers.

[Page 23]

III. SHOULD RIDE-HAILING PRICING BE TREATED AS A SERIES OF VERTICAL RESTRAINTS?

Uber and Lyft’s ride-hailing service and pricing arrangement have also been characterized as a series of vertical agreements between Uber or Lyft and drivers.²² Under a vertical agreement framework, drivers provide services to riders, and Uber and Lyft supply an input service to the drivers. When people sign up to drive for Uber or Lyft, they do so to connect to riders using Uber’s platform or Lyft’s platform. Drivers pay Uber and Lyft a portion of the fare for each ride they provide using the ride-hailing platform. In return, Uber and Lyft connect drivers to riders and provide drivers an easy way to bill for the rides.

Under the federal antitrust laws, such vertical pricing arrangements between a supplier and customer of services, even if they impose minimum resale prices, typically receive rule of reason analysis, not a per se analysis. In contrast, some states, including California, still treat minimum resale price maintenance arrangements as per se unlawful.²³

However, as even the court in Kalanick recognized, the pricing algorithms of ride-hailing companies like Uber and Lyft arguably do not really qualify as "resale" pricing, since no resale occurs in the transaction.²⁴ In the vertical framework outlined above, drivers consume the platform services offered to them by ride-hailing apps like Uber and Lyft as an input into the service the drivers provide to riders. The drivers do not resell the platform service to riders. So even under the laws of states that have per se prohibitions against minimum resale price maintenance, such as California, one question would be whether ride hailing app pricing can really be characterized as such.

IV. SHOULD RIDE-HAILING PRICING BE TREATED AS A "HUB-AND-SPOKE" CONSPIRACY?

In Meyer v. Kalanick, the court declined to accept as a matter of law that Uber’s many agreements with drivers were purely vertical, holding that the plaintiff had sufficiently pled a "hub-and-spoke" conspiracy.²⁵ A "hub and spoke" conspiracy can be thought of as a hybrid vertical and horizontal agreement. Ultimately, though, the construct is meant to infer a horizontal agreement among competitors. As the court explained in Meyer v. Kalanick:

[C]ourts have long recognized the existence of "hub-and-spoke" conspiracies in which an entity at one level of the market, the "hub," coordinates an agreement among competitors at a different level, the "spokes." These arrangements consist of both vertical agreements between the hub and each spoke and a horizontal agreement among the spokes to adhere to the [hub’s] terms, often because the spokes would not have gone along with the vertical agreements except on the understanding that the other spokes were agreeing to the same thing.²⁶

[Page 24]

The court concluded that the plaintiff plausibly alleged that drivers agreed to Uber’s pricing terms with "the clear understanding that all other Uber drivers are agreeing to charge the same fares."²⁷ The court found that the plaintiff’s allegations were bolstered by the occasional events Uber holds for its drivers, such as picnics, which would theoretically give drivers an opportunity to organize, as well as an instance in which drivers in New York City once collectively negotiated higher fares from Uber.²⁸ Defendant Kalanick argued that a conspiracy among drivers is "wildly implausible" because it would involve an agreement "among hundreds of thousands of independent transportation providers all across the United States."²⁹ The court gave short shrift to that argument reasoning that it was Uber’s own "genius" in utilizing the "magic of smartphone technology" that could enable Uber to orchestrate such a large conspiracy.³⁰

This reasoning is troubling. First, the allegation that the countless Uber drivers had reached a horizontal agreement was not plausible, nor supported with substantial factual allegations.³¹ Illustrative is the precedent upon which the Kalanick court relied. That precedent involved far fewer spokes to the conspiracy than in the case of Uber or Lyft, which have thousands of drivers, and it was shown that those spokes extensively communicated with each other. For example, in Interstate Circuit Inc. v. United States, there were only eight film distributors making up the rim of the hub-and-spoke agreement,³² while in United States v. Apple Inc., et al., there were only five publishing companies making up the rim.³³ This is a far cry from the hundreds of thousands of Uber drivers all over the world.

Moreover, Interstate Circuit involved express communications, in the form of letters between the hub (a film exhibitor) and the spokes (film distributors).³⁴ In Apple, the court found that the spokes (book publishers) expressly colluded with each other and that the hub (Apple) consciously played a role in organizing their collusion.³⁵ Specifically, the district court in Apple found that the collusion among the publishers was accomplished through regular meetings, writing, "[o]n a fairly regular basis, roughly once a quarter, the CEOs of the Publishers held dinners in the private dining rooms of New York restaurants, without counsel or assistants present, in order to discuss the common challenges they faced, including most prominently, Amazon’s pricing policies."³⁶ Additionally, Apple also participated in these direct conversations and meetings.³⁷ While the court in Kalanick correctly recognized that technology may be able to facilitate hub-and-spoke conspiracies, Uber and Lyft drivers do not have nearly the same level of direct conversation around pricing with each other or with Uber and Lyft, either through the app or otherwise, as was present in the Apple case.

[Page 25]

Second, the allegations that the drivers (the alleged spokes) agreed with Uber (the alleged hub) on the condition or assurance by the hub that other spokes agreed to those same terms were not well-supported in the complaint. It is true that ride-hailing companies offer a platform service that applies the same pricing algorithm to each ride, no matter who drives or who rides. Drivers for the most part probably know this when they sign up. But that does not mean they agreed to sign up to the Lyft or Uber platform on the condition or assurance that other drivers would agree to or be subject to the same Uber or Lyft pricing algorithm.

Inference of a horizontal (hub-and-spoke) conspiracy requires more than allegations of mere knowledge. Other hub-and-spoke cases illustrate this point. For example, in Toys "R" Us, Inc. v. FTC,³⁸ the court found that the spoke toy manufacturers accepted restrictions on their ability to sell toys to certain Toys "R" Us competitors on the express condition that the other largest toy manufacturers agreed to do the same.³⁹ The court found these assurances important because the manufacturers were worried that other manufacturers would cheat by breaking the terms of the agreement.⁴⁰ The FTC explained, "[t]hese manufacturers were in effect being asked by TRU to reduce their output . . . and as is classically true in such cartels, they were willing to do so only if TRU could protect them against cheaters."⁴¹

Furthermore, the anticompetitive agreement in Toys "R" Us was directly targeted at harming low-cost competitors, with the FTC writing, "TRU sought to eliminate the competitive threat the [low-cost competitor toy] clubs posed by denying them merchandise, forcing clubs’ customers to buy products they did not want, and frustrating customers’ ability to make direct price comparisons of club prices and TRU prices."⁴² There is no factual support for the idea that all Uber drivers, or even a significant portion of them, entered into the Uber driver agreement with an explicit assurance that other drivers would be using the same pricing algorithm. Many of the other factors present in Toys R Us—including efforts to drive out low-cost competitors, forcing consumers to buy rides they do not want, or frustrating consumers’ ability to price other ride-for-hire services—are not alleged to exist in the hail-riding context either.

[Page 26]

Uber’s and Lyft’s pricing algorithms do not exist to provide drivers assurances that competition on price will be eliminated between them or to prevent competition from a lower priced competitor. Rather, this pricing mechanism attempts to ensure the greatest degree of supply and demand efficiency as discussed below. Furthermore, the allegations in Kalanick were devoid of a suggestion that drivers could not sign up for multiple ride-hailing apps. Those ride-hailing platforms compete with each other on price and what pay they offer drivers, which in turn leads to additional competition among drivers. One would expect a true anticompetitive pricing agreement to prohibit drivers from breaking ranks with the conspiracy by driving for another ride-hailing company.

Empirical evidence of an actual price-fixing conspiracy among Uber drivers in Washington, D.C. suggests that ride-hailing apps, on their own, would be insufficient vehicles for facilitating such a conspiracy. In May of this year, reports emerged that Uber drivers at Reagan National Airport near Washington, D.C. were conspiring to game the Uber algorithm and artificially inflate prices.⁴³ Groups of fifty or so drivers apparently met near the airport and all agreed to turn off their apps right before multiple flights were scheduled to land. Because lack of supply tends to influence the Uber algorithm to increase prices, prices for rides from Reagan National would begin to creep up. Two drivers who kept their apps on would monitor the pricing and signal to the others to turn their apps back on once prices reached a certain threshold. Fixing prices among 50 drivers in a very limited vicinity near Reagan National Airport required real-time and in-person communications. If Uber or Lyft already provided a mechanism for drivers to conspire to charge supracompetitive fares, then surely this elaborate scheme would not have been necessary.

V. NO MATTER HOW YOU CLASSIFY RIDE-HAILING PRICING, THE RULE OF REASON SHOULD APPLY: BMI

In Broadcast Music Inc. v. Columbia Broadcasting System, Inc.⁴⁴ ("BMI") the Supreme Court held that while the pricing restraint at issue was horizontal price fixing "in the literal sense,"⁴⁵ it nonetheless had to be analyzed under the "rule of reason" not the per se rule.⁴⁶ The defendants in BMI, ASCAP and BMI, were licensing agencies whose members were music composers.⁴⁷ ASCAP and BMI were given the authority by their composer-members to provide a non-exclusive blanket license to use the members’ copyrighted music to organizations such as radio stations or television networks. Under the blanket license, a single fee gave the licensee access to any of the songs in the ASCAP or BMI repertoires, respectively.⁴⁸ The blanket license did not preclude each member from negotiating individual licenses to their music. CBS sued ASCAP and BMI, claiming that the single-fee blanket license amounted to per se illegal price fixing in violation of Section 1 of the Sherman Act.⁴⁹

[Page 27]

The Supreme Court disagreed, holding that although facially "price fixing," ASCAP and BMI’s blanket licenses should be analyzed under the rule of reason framework.⁵⁰ The Court did so for three main reasons, each of which applies equally to ride-hailing pricing. We discuss them below.⁵¹

A. Lack of Experience with Ride-Hailing Platforms and Pricing

Under antitrust law, practices are deemed per se illegal when "the practice facially appears to be one that would always or almost always tend to restrict competition and decrease output."⁵² The BMI Court pointed out that courts only classify a business practice as per se illegal when they have considerable experience with it, noting that it had never examined a practice like the blanket licenses before.⁵³

That reasoning applies in ride-hailing pricing as well. Courts do not have "considerable experience" with Uber and Lyft’s pricing policies: these companies and others with similar pricing models have only existed for about a decade,⁵⁴ and there has been minimal antitrust inquiry into whether their centralized pricing algorithms constitute price fixing.⁵⁵

B. Pricing Algorithm Ancillary to Ride-Hailing

The BMI Court pointed out that horizontal restraints are not properly classified as per se illegal when they are not "naked restraints of trade," but ancillary to other procompetitive agreements.⁵⁶ The Court suggested that the blanket license program at issue in BMI had procompetitive effects because it was necessary to enable thousands of different entities to negotiate licenses for the many different copyrighted pieces of music, as well as to enforce the copyrights on these works; a nearly impossible task without blanket licensing. The setting of a fixed price was therefore a "necessary consequence" of offering a blanket license,⁵⁷ such that a rule of reason analysis applied.

[Page 28]

Here, the pricing algorithms at issue are similarly integral to the ride-hailing service. There are intuitive reasons to believe that a centralized pricing algorithm is key to the functioning and output enhancing virtues of ride-hailing. Historically, taxi prices are regulated, and the rationale for these regulations was, in part, to ensure predictability and avoid price gouging.⁵⁸ Regulated prices, however, are not actually based on supply and demand, and thus highly inefficient.⁵⁹ The Uber and Lyft pricing algorithms, however, adjust for supply and demand in a particular locale, optimizing market efficiency, while preserving the benefits of centralized pricing that customers have come to expect. Thus, as discussed below, this increases the probability that there will be rides available when and where people need them. Further, a centralized pricing algorithm contributes to greater transparency and thus a reduction of transaction costs. Because of the consumer benefit of these centralized pricing algorithms, the algorithms are integral to the product just as a blanket license in BMI was necessary to effectuate the copyright protections of the license.

C. Ride-Hailing Apps and Price Algorithms Are a New Product

The BMI Court explained that the blanket license also required rule of reason analysis rather than per se treatment because it actually created a new product:⁶⁰ an aggregating, one-stop shop service that allowed users of the compositions to have a lot of flexibility in selecting which songs they would like to use. Since the blanket license was a new product, and each composer member remained free to license and price its own work, the uniform price of the blanket license was not a price restraint that limited competition; the individual composers did not participate in the market for the blanket license product.⁶¹ When they were first launched, ride-hailing apps were a completely new product, as were their pricing algorithms, which for the first time presented a mechanism for customers to find a truly on-demand ride. Because of the pricing algorithm, cars efficiently met this demand in a way that taxi cabs and other ride services had not. Just as the individual composers in BMI could not offer a blanket license for a single fee covering the repertoires of multiple composers, individual drivers could not, pre-Uber and Lyft, offer a service that showed all (or multiple) drivers available in an area, much less determine the most efficient ride fares based on the availability of drivers. Meanwhile, as in BMI, neither Uber nor Lyft precludes drivers from individually selling rides or signing up with other ride-hailing services. Indeed, many drivers multi-home Uber, Lyft, and other ride-hailing apps.⁶²

[Page 29]

The emergence of ride-hailing apps and centralized pricing algorithms thus does not appear to have eliminated competition; to the contrary, as discussed below, all signs points to it having created a lot of new competition in transportation services. As a consequence, driver agreements on pricing attendant to ride-hailing services should be analyzed under the rule of reason.⁶³

[Page 30]

VI. RULE OF REASON ANALYSIS SUGGESTS RIDE-HAILING PRICING IS PROCOMPETITIVE

Under a rule of reason analysis, a court likely would find ride hailing app pricing procompetitive. The Ninth Circuit in Tanaka v. University of Southern California,⁶⁴ laid out the rule of reason test as follows, and many courts have followed a similar approach:

The plaintiff bears the initial burden of showing that the restraint produces ‘significant anticompetitive effects’ within a ‘relevant market’. If the plaintiff meets this burden, the defendant must come forward with evidence of the restraint’s procompetitive effects. The plaintiff must then show that ‘any legitimate objectives can be achieved in a substantially less restrictive manner.⁶⁵

While we have not seen evidence that Uber’s or Lyft’s centralized pricing algorithms have resulted in "significant anticompetitive effects," there is ample evidence that these algorithms have had significant procompetitive effects.

First, output has grown significantly since ride-hailing apps emerged. An analysis conducted by Statista shows that in 2009, before the introduction of ride-hailing apps, there were 170.9 million pickups in New York City.⁶⁶ In 2017 ride-hailing apps were responsible for 159.9 million pickups in New York City, while taxis were still responsible for 125.5 million, for a total of 285.4 million pickups. This suggests the introduction of ride-hailing services is correlated with an increase of 114.5 million pickups in New York City.

Second, the quality and efficiency of hailing rides has also improved substantially since ride-hailing apps emerged. Many taxi cab sers innovation by rivices were forced to adopt apps for ride-hailing and payment in direct response to thide-hailing platforms like Uber. ⁶⁷ Studies suggest that since Uber’s introduction, taxis have also responded by improving the quality of their service, as reflected by a decreased number of passenger complaints to taxi regulators about things such as broken heating and air conditioning or rude driver behavior. ⁶⁸

Finally, initial data suggests that the centralized pricing algorithms of apps like Uber and Lyft have created a highly efficient mechanism to optimize supply and demand and maximize the availability of nearby drivers when there is local rider demand.⁶⁹ When there is high demand, Uber’s and Lyft’s pricing algorithms increase the fares to encourage drivers who would not otherwise plan to be working to come out and drive. This has been confirmed empirically in at least one study.⁷⁰ In The Effects of Uber’s Surge Pricing: A Case Study, the authors analyzed a natural experiment, comparing two nights in which there was present an unusually high demand for Uber rides in New York City: the night of a sold-out Ariana Grande concert in Madison Square Garden and New Year’s Eve. On the latter night, an error in Uber’s system meant that surge pricing was not in effect for 26 minutes. The first night saw the amount of drivers in the area near the venue double and wait times remained constant at 2.6 minutes. During the latter night, average wait times spiked during the outage periods to 8 minutes. Additionally, the number of ride requests that were fulfilled dropped precipitously, meaning that many would-be riders could not get a ride.

[Page 31]

Thus, there is significant evidence of pro-competitive benefits arising from ride-hailing apps, and the pricing algorithms themselves.

There also are other indicia suggesting that these ride-hailing apps are not stifling competition. Uber, as the first mover, and Lyft, as second mover, presumably both have substantial positions in ride-hailing or transportation services. One third party report contains estimates that Uber accounts for 71 percent and Lyft for 27 percent of app-based ride-hailing sales in the US, respectively.⁷¹

Of course, it is not clear how meaningful these numbers are for antitrust purposes, since one would first have to determine in what relevant market ride-hailing companies operate. A discussion of the relevant antitrust market is beyond the scope of this article, but obvious questions are whether to include taxi cabs and other limousine-for-hire services, among others. Public transit may also be a substitute, depending on where someone is and needs to go. In any event, at a minimum, Uber and Lyft face competition not only from each other, but also from other ride-hailing apps, such as Juno, Via, or Flywheel, as well as taxi cabs, most of which now also offer their services on hailing and payment apps, such as Curb. Riders and drivers both can and regularly do multi-home ride-hailing platforms.⁷² There thus is little cost in switching between providers. Meanwhile, Uber and Lyft reportedly make little profit on their ride-hailing services. This all suggests that barriers to entry, expansion, and switching in ride-hailing are low, and competition is vigorous. Such robust competition leaves little room for accumulation or exercise of market power.

[Page 32]

VII. CONCLUSION

Ride-hailing apps like Uber and Lyft have revolutionized the transportation market. They have unseated taxi monopolies, hastened the rise of the gig economy, and provided a new, convenient way to get around. Centralized pricing algorithms play a key part in facilitating these benefits. Courts and regulators should be careful to consider these benefits, and appropriately apply rule of reason analysis to allegations of anticompetitive conduct in the ride-hailing industry. Anyone who remembers what it was like trying to find a cab in San Francisco or Los Angeles before 2009 surely will agree.

[Page 33]

——–

Notes:

^1. Daniel Bitton is a partner in Axinn, Veltrop & Harkrider LLP’s San Francisco office and a member of their antitrust group. David Pearl is an antitrust associate in Axinn’s San Francisco office and Patrick Shaw is an antitrust associate in Axinn’s Washington D.C. office.

^2. Aaron Gordon, The Legal Argument That Could Destroy Uber, Jalopnik (May 20, 2019), available at https://jalopnik.com/the-legal-argument-that-could-destroy-uber-1834790506("Gordon").

^3. See, e.g., Rebecca Haw Allensworth, The Commensurability Myth in Antitrust, 69 Vand. L. Rev. 1, 5 n.7 (2016) ("The Rule of Reason . . . [is] a standard that balances pro- with anticompetitive effects. . . .")

^4. U.S. National Labor Relations Board, Office of General Counsel, Advice Memorandum, Cases 13-CA-163062, 14-CA0158833, 29-CA-177483 (Apr. 16, 2019) available at https://www.laborrelationsupdate.com/files/2019/05/NLRB-Uber-memo.pdf.

^5. 4 Cal.5th 903 (2018).

^6. Cal. Assem. Bill 5 (2019-2020, Reg. Sess.); see Kate Conger & Noam Scheiber, California Bill Makes App-Based Companies Treat Workers as Employees, The New York Times (Sept. 11, 2019), available at https://www.nytimes.com/2019/09/11/technology/california-gig-economy-bill. html?smid=nytcore-ios-share.

^7. Dara Khosrowshahi, Logan Green, & John Zimmer, Open Forum: Uber, Lyft ready to do our part for drivers, San Francisco Chronicle (June 12, 2019), available at https://www.sfchronicle.com/opinion/openforum/article/Open-Forum-Uber-Lyft-ready-to-do-our-part-for-13969843. php?psid=k9i1h.

^8. See, Hal Singer, Uber Under The Antitrust Microscope: Is There A "Firm Exemption" To Antitrust?, Forbes (Feb. 25, 2019), available at https://www.forbes.com/sites/washingtonbytes/2019/02/25/uber-under-the-antitrust-microscope-is-there-a-firm-exemption-to-antitrust/#145b3e092a47; Gordon, supra, note 2.

^9. 174 F.Supp.3d 817 (S.D.N.Y. 2016).

^10. Id. at 820.

^11. Id.

^12. See Meyer v. Kalanick, 291 F. Supp. 3d 526 (S.D.N.Y. 2018).

^13. 174 F. Supp. 3d at 829.

^14. 467 U.S. 752, 769 (1984).

^15. 441 U.S. 1 (1979).

^16. According to former Uber CEO Travis Kalanick’s brief in Meyer v. Kalanick, Uber’s terms of service with its drivers at that time allowed for downward departure in the price charged from the recommended price. Memorandum of Law In Support of Defendant Travis Kalanick’s Motion to Dismiss, Meyer v. Kalanick, 174 F.Supp.3d 817, at *10 (S.D.N.Y. 2016). We are not aware of the details of the current driver arrangement, but such a provision would make it even more unlikely that Uber’s pricing would be considered price-fixing because there would have to be additional evidence that Uber and its drivers had an agreement not to depart from the recommended fare despite being able to do so.

^17. How Much Can Drivers Make With Uber, https://www.uber.com/us/en/drive/how-much-drivers-make/ (last visited Sept. 11, 2019).

^18. Uber Service Fee: How Can Pricing Serve Drivers and Riders?, https://marketplace.uber.com/pricing/service-fee (last visited Sept. 11, 2019).

^19. How and When Driver Pay is Calculated, https://help.lyft.com/hc/en-us/articles/115013080008-How-and-when-driver-pay-is-calculated#calculations (last visited Sept. 11, 2019); The Service Fee, https://help.lyft.com/hc/en-us/articles/115013081048 (last visited Sept. 11, 2019).

^20. Two—sided platforms enable two distinct types of participants to interact more readily and realize gains from trade or other interaction. They provide each customer group with access to the other customer group.

^21. See, e.g., Groundlink.com, FAQs, Rates & Fees, https://www.groundlink.com/faq/ and Groundlink. com, Independent Contractors Agreement Terms and Conditions, https://driver.groundlink.com/terms-and-conditions. Note that, similar to surge pricing from ride-hailing services, Groundlink features "demand-based pricing," which raises prices during times of high demand in order to ensure there is sufficient driver supply.

^22. See 174 F. Supp. 3d at 823.

^23. For example, minimum resale price maintenance is still per se illegal under California’s Cartwright Act. See, e.g., Alan Darush MD APC v. Revision LP, No. CV 12-10296 GAF (AGRx), 2013 WL 1749539 (C.D. Cal. Apr. 10, 2013).

^24. See 174 F. Supp. 3d at 826 ("Here, unlike in Leegin, Uber is not selling anything to drivers that is then resold to riders.").

^25. 174 F. Supp. 3d at 823-24.

^26. Id. at 824 (internal alterations and emphasis in original) (quoting United States v. Apple, Inc., 791 F.3d 290, 314 (2d Cir. 2015)).

^27. Id.

^28. Id. at 821, 825.

^29. Id. at 825.

^30. Id.

^31. Id. at 824 (internal alteration omitted).

^32. 306 U.S. 208, 214 (1939).

^33. 791 F.3d at 296.

^34. 306 U.S. at 216-17.

^35. 791 F.3d at 316.

^36. United States v. Apple, 952 F. Supp. 2d 638, 651 (S.D.N.Y. 2013).

^37. Id. at 657-58.

^38. 221 F.3d 928 (7th Cir. 2000).

^39. Id. at 932.

^40. Id. at 936.

^41. Id.

^42. Id. at 932.

^43. Minda Zetlin, Here’s Why Uber and Lyft Drivers Are Artificially Creating Surge Prices, Inc.com (May 23, 2019), available at https://www.inc.com/minda-zetlin/uber-lyft-drivers-artificial-surge-pricing-reagan-national-washington-arlington-drive-united.html.

^44. 441 U.S. 1.

^45. 441 U.S. at 8.

^46. Id. at 24-25.

^47. Id. at 5.

^48. Id. at 5-6.

^49. Id. at 6.

^50. Id. at 16-24.

^51. Note that BMI arguably involved a much clearer example of a horizontal restraint than does ride-hailing, since BMI and ASCAP are licensing agencies representing competing performing artists, who were their members, while it is at a minimum debatable whether Uber and Lyft represent drivers who sign up for their service.

^52. 441 U.S. at 19-20.

^53. See also United States v. Topco Associates, Inc., 405 U.S. 596, 608 (1972) ("It is only after considerable experience with certain business relationships that courts classify them as per se violations of the Sherman Act."); Northrop Corp. v. McDonnell Douglas Corp., 705 F.2d 1030, 1051-52 (9th Cir. 1983) (finding agreements under which Northrop would limit its marketing of aircraft to those suitable for land-based operation and McDonnell Douglas limited its marketing to aircraft suitable for aircraft-carrier operation governed by rule of reason analysis where, among other factors, the courts lacked significant experience with either the business practice or industry involved).

^54. Uber was founded in 2009 and the current incarnation of Lyft began in 2012. The History of Uber, https://www.uber.com/newsroom/history/ (last visited Aug. 31, 2019); Andrew Graeiner, Matt McFarland, Ivory Sherman, & Jen Tse, A History of Lyft, From Fuzzy Pink Mustaches to Global Ride Share Giant, CNN (Apr. 2, 2019) available at https://www.cnn.com/interactive/2019/03/business/lyft-history/index.html.

^55. In addition to Meyer v. Kalanick, Uber’s pricing system has been analyzed under a theory of price-fixing under the Maryland Antitrust Law. See The Yellow Cab Co. v. Uber Technologies, Inc., Case No. 14-CV-2764-RDB, 2015 WL 4987653 (D. Md. Aug. 19, 2015).

^56. 441 U.S. at 19-20.

^57. Id. at 21.

^58. Transportation Research Board, Between Public and Private Mobility: Examining the Rise of Technology-Enabled Transportation Services, Committee for Review of Innovative Urban Mobility Services Special Report 319, p. 40 (2016) available at https://www.nap.edu/read/21875/chapter/5 ("[Taxicab] fare regulation is designed to ensure predictability in the amount customers will be charged, to eliminate price gouging, and to ensure a reasonable return for owners and drivers.").

^59. See Mark W. Frankena & Paul A. Pautler, An Economic Analysis of Taxicab Regulation, Fed. Trade Comm’n, Bureau of Econ. Staff Report, 6-7 (1984) available at https://www.ftc.gov/sites/default/files/documents/reports/economic-analysis-taxicab-regulation/233832.pdf ("It appears that taxi regulations have often been designed to protect public transit systems and existing taxi firms from competition. . . . Some of the more obvious ways in which the allocation of resources under existing regulations is inefficient include: (a) the number of taxi rides taken is inefficiently low, because of regulation that raise fares, restrict the amount of service, and increase waiting times; (b) the cost of producing taxi trips is unnecessarily high, because of regu lations that prevent ride sharing and increase deadheading and waiting in taxi lines; and (c) there are shortages of certain types of services because of the incentives provided by the structure of fares.").

^60. 441 U.S. at 22.

^61. A number of cases since BMI have developed the exception further. In Arizona v. Maricopa Cnty Med. Soc., 457 U.S. 332, 340-41 (1982), the Supreme Court, among other things, emphasized that the price restraints had to be necessary to achieve any claimed efficiencies, and that it is important that participants in a restraint don’t compete much outside of the restraint anyways. Id. at 353-57. In Nat. Collegiate Athletic Assoc. v. Board of Regents of the Univ. of Oklahoma, 468 U.S. 85 (1984), the Supreme Court analyzed restrictions that the National Collegiate Athletic Association ("NCAA") placed on its member schools. The court determined that this practice was to be analyzed under a rule of reason framework because the horizontal restraints are essential if the "college football product" is to be available at all. Id. at 101-02. See also O’Bannon v. Nat. Collegiate Athletic Ass’n, 802 F.3d 1049 (9th Cir. 2015). Ultimately, however, the Court concluded the rules were unlawful under a rule of reason analysis because they resulted in increased prices and decreased output. 468 U.S. at 113, 120.

^62. It is well known that some agents may multi-home in two-sided markets (e.g., a consumer may carry and a merchant may accept multiple credit cards). In the case of ride-hailing, a driver may drive for both Uber and Lyft, and a rider may use both companies’ apps and request a ride from the company that has a driver closer by.

^63. These market realities suggest that even under California law, the rule of reason should apply. California courts have been more reticent to move away from the per se rule than federal courts. For example, courts have recognized that even vertical pricing restraints still receive per se treatment in California. See, e.g., Alan Darush MD APC v. Revision LP, Case No. 12-CV-10296-GAF (AGRx), 2013 WL 1749539, *6 (C.D. Cal. Apr. 10, 2013). Nonetheless, there is still some precedent that can be found suggesting California’s Supreme Court would, at least, consider applying rule of reason. The Fisher v. City of Berkeley Court has recognized the BMI decision and noted that the "price-fixing illegal per se rule . . . has . . . suffered steady and growing criticism as an often arbitrary, mechanical, and inconsistently applied rule that ignores the realities of market power and net economic effects." 37 Cal. 3d 644, 666-67 & n.14 (1984). It went on to consider whether applying the per se rule to municipal defendants concerning rent control ordinances would be "economically reliable" or "overinclusive[]." Id. at 668-671. That court concluded it would be, and declined to apply the per se rule. Id. at 671.

^64. 252 F.3d 1059 (9th Cir. 2001).

^65. Id. at 1063. Other circuits have established similar burden shifting frameworks. See e.g., FTC v. H.J. Heinz Co., 246 F.3d 708, 715 (D.C. Cir. 2001).

^66. Patrick Wagner, Ride-Hailing Apps Surpass Regular Taxis in NYC, Statista (Apr. 10, 2018), available at https://www.statista.com/chart/13480/ride-hailing-apps-surpass-regular-taxis-in-nyc/.

^67. See, e.g., Luz Lazo, The D.C. Taxi App Has Arrived, The Washington Post, (Feb. 11, 2016), available at https://www.washingtonpost.com/news/dr-gridlock/wp/2016/02/11/the-d-c-taxi-app-has-arrived/.

^68. Scott Wallsten, The Competitive Effects of the Sharing Economy: How is Uber Changing Taxis? (June 1, 2015), available at https://www.ftc.gov/system/files/documents/public_comments/2015/06/01912-96334.pdf.

^69. How Surge Pricing Works, https://www.uber.com/us/en/drive/partner-app/how-surge-works/(last visited Sept. 9, 2019).

^70. Jonathan Hall, Cory Kendrick & Chris Nosko, The Effects of Uber’s Surge Pricing: A Case Study, working paper (2016) available at http://economicsforlife.ca/wp-content/uploads/2015/10/effects_of_ubers_surge_pricing.pdf.

^71. Kathryn Gessner, Uber vs. Lyft: Who’s Tops in The Battle of U.S. Rideshare Companies, Second Measure (Aug. 21, 2019) available at https://secondmeasure.com/datapoints/rideshare-industry-overview/.

^72. Melissa Berry, How Many Uber Drivers Are There?, The Ride Share Guy (June 1, 2019), available at https://therideshareguy.com/how-many-uber-drivers-are-there/; Eric. A. Morris et al., Assessing the Experience of Providers and Users of Transportation Network Company Ridesharing Services, Center for Connected Multimodal Mobility Final Report, 58 (May, 2019) available at https://cecas.clemson.edu/C2M2/assessing-the-experience-of-providers-and-users-of-transportation-network-company-ridesharing-services/.