SINGH v. UBER TECHNOLOGIES, INC.

*FOR PUBLICATON*

UNITED STATES DISTRICT COURT

DISTRICT OF NEW JERSEY

_______________________________________

JASWINDER SINGH, on behalf of himself

and all those similarly situated,

Plaintiff,

No. 16-3044 (FLW)

v.

OPINION

UBER TECHNOLOGIES, INC.,

Defendant.

JAMES CALABRESE, GREGORY

CABANILLAS, and MATTHEW

MECHANIC, individually and on behalf of

all those similarly situated,

Plaintiffs,

No. 19-18371 (FLW)

v.

UBER TECHNOLOGIES, INC., and

RASIER, LLC,

Defendants.

WOLFSON, Chief Judge:

Jaswinder Singh, James Calabrese, Gregory Cabanillas, and Matthew Mechanic

(collectively, “Plaintiffs”) were drivers with the rideshare company Uber Technologies, Inc., who

allege individually, and on behalf of a class of similarly situated New Jersey drivers,1 that Uber

1 A motion for class certification is also pending. Case No. 16-3044, ECF No. 61. It appears that

Plaintiffs filed this motion based on the misunderstanding that certain language in the Federal Arbitration

Act requires it. However, the FAA uses the term “class” to refer to a category or group of workers, not in

the sense of class action litigation or in connection with Fed. R. Civ. P. 23. In any event, since I compel

arbitration in this Opinion and accompanying Order, the motion is moot.

misclassified them as independent contractors, thereby depriving them of overtime pay and other

benefits afforded to employees. Uber moves to compel arbitration under the Federal Arbitration

Act (“FAA”) pursuant to a clause in Plaintiffs’ contracts. 9 U.S.C. § 1, et. seq. Plaintiffs argue that

arbitration is inappropriate because they fall within an exemption to the FAA as transportation

workers who move riders across state lines. Uber responds that Plaintiffs do not belong to such a

class of workers because interstate rides constitute a small fraction of all rides, and in any event, I

should order arbitration under the New Jersey Arbitration Act (“NJAA”), which embodies the

same pro-arbitration policy as the FAA without the exemptions. For the following reasons, I

GRANT Uber’s motions, COMPEL arbitration under the FAA, and DENY Plaintiffs’ motion for

class certification as moot.

I. FACTUAL BACKGROUND AND PROCEDURAL HISTORY

Uber is a billion-dollar technology company whose ridesharing app enables drivers to

connect with riders, based on location, at the click of a button. Def. Statement of Material Facts I

(“SUMF”), ¶¶ 1, 9. Plaintiffs are gig-economy workers who used the Uber app to provide rides

between 2014 and 2020. They allege that Uber must reimburse certain business expenses (e.g., the

cost of maintaining cars, gas, insurance, and phone/data expenses), comply with guaranteed

minimum wage laws, and pay overtime, as state law requires for employees. The present dispute

centers on the validity of an arbitration provision in their contracts.

A. The Arbitration Provision

Drivers who sign up with Uber must accept the company’s Technology Services

Agreement (“TSA”) before completing any rides. Id. ¶¶ 2, 28, 30, 41. Uber presents the TSA to

drivers as soon as they login to the app by populating a “TERMS AND CONDITIONS” screen

with a hyperlink. Id. ¶¶ 3-4. Clicking the hyperlink opens the TSA. Id. ¶ 5. After drivers scroll

through the document for as long as they need to review it, id. ¶ 6, the app prompts them to click

“YES, I AGREE.” Id. ¶ 8. As this screen makes clear, “[b]y clicking below, you represent that you

have reviewed all the documents above and that you agree to all the contracts above.” Id. ¶ 9. Once

a driver indicates agreement, the app generates another screen, which reads: “PLEASE CONFIRM

THAT YOU HAVE REVIEWED ALL THE DOCUMENTS AND AGREE TO ALL THE NEW

CONTRACTS.” Id. ¶ 11. At this point, drivers may select buttons reading “NO” or “YES, I

AGREE.” Id. ¶ 12. If drivers select yes, Uber stores the executed TSA in an online portal,

reviewable to this day. Id. ¶ 13. Singh joined Uber on June 21, 2014. Id. ¶¶ 16, 25. Mechanic joined

on December 11, 2015. Def. SUMF II, ¶ 61. Calabrese joined on June 8, 2017. Id. ¶ 55. Cabanillas

joined on August 18, 2017. Id. ¶ 56. Each driver accepted the TSA as a condition of signing up.

Def. SUMF I, ¶¶ 14-15; Def. SUMF II, ¶¶ 55, 57-59, 61.

The applicable version of the TSA contains an arbitration provision visible on the first

page, which provides:

IMPORTANT: PLEASE NOTE THAT TO USE THE UBER SERVICES,

YOU MUST AGREE TO THE TERMS AND CONDITIONS SET FORTH

BELOW. PLEASE REVIEW THE ARBITRATION PROVISION SET

FORTH BELOW CAREFULLY, AS IT WILL REQUIRE YOU TO

RESOLVE DISPUTES WITH THE COMPANY ON AN INDIVIDUAL

BASIS THROUGH FINAL AND BINDING ARBITRATION UNLESS YOU

CHOOSE TO OPT OUT OF THE ARBITRATION PROVISION…. IF YOU

DO NOT WISH TO BE SUBJECT TO ARBITRATION, YOU MAY OPT

OUT OF THE ARBITRATION PROVISION BY FOLLOWING THE

INSTRUCTIONS PROVIDED IN THE ARBITRATION PROVISION

BELOW.

WHETHER TO AGREE TO ARBITRATION IS AN IMPORTANT

BUSINESS DECISION. IT IS YOUR DECISION TO MAKE, AND YOU

SHOULD NOT RELY SOLELY UPON THE INFORMATION PROVIDED

IN THIS AGREEMENT AS IT IS NOT INTENDED TO CONTAIN A

COMPLETE EXPLANATION OF THE CONSEQUENCES OF

ARBITRATION. YOU SHOULD TAKE REASONABLE STEPS TO

CONDUCT FURTHER RESEARCH AND TO CONSULT WITH

OTHERS—INCLUDING BUT NOT LIMITED TO AN ATTORNEY—

REGARDING THE CONSEQUENCES OF YOUR DECISION, JUST AS

YOU WOULD WHEN MAKING ANY OTHER IMPORTANT BUSINESS

OR LIFE DECISION.

Def. SUMF I, ¶ 18. The arbitration provision specifies the FAA as the governing law and contains

a class action waiver:

This Arbitration Provision is governed by the Federal Arbitration Act, 9 U.S.C. § 1

et seq. (the “FAA”) and evidences a transaction involving commerce. This

Arbitration Provision applies to any dispute arising out of or related to this

Agreement or termination of the Agreement and survives after the Agreement

terminates . . . .

Except as it otherwise provides, this Arbitration Provision is intended to apply

to the resolution of disputes that otherwise would be resolved in a court of law

or before a forum other than arbitration. This Arbitration Provision requires

all such disputes to be resolved only by an arbitrator through final and binding

on an individual basis only and not by way of court or jury trial, or by way of

class, collective, or representative action.

Id. ¶ 19. The arbitration provision also contains a delegation clause, which encompasses a wide

range of potential disputes between drivers and Uber, including threshold questions such as

whether a particular dispute is arbitrable:

Such disputes include without limitation disputes arising out of or relating to

interpretation or application of this Arbitration Provision, including the

enforceability, revocability or validity of the Arbitration Provision. All such

matters shall be decided by an Arbitrator and not by a court or judge.

Id. ¶ 20. At the same time, the TSA offers an opt-out provision, which drivers may exercise for up

to 30 days after accepting the TSA by emailing Uber. It states:

Your Right To Opt Out Of Arbitration.

Arbitration is not a mandatory condition of your contractual relationship with

the Company. If you do not want to be subject to this Arbitration Provision,

you may opt out of this Arbitration Provision by notifying the Company in

writing of your desire to opt out of this Arbitration Provision . . . .

Should you not opt out of this Arbitration Provision within the 30-day period,

you and the Company shall be bound by the terms of this Arbitration

Provision. You have the right to consult with counsel of your choice concerning

this Arbitration Provision. You understand that you will not be subject to

retaliation if you exercise your right to assert claims or opt-out of coverage

under this Arbitration Provision.

Id. ¶ 21. Though thousands of drivers exercised their opt-out rights, Plaintiffs did not.2 Id. ¶ 23-

24; Def. SUMF II, ¶¶ 56, 60, 62.

B. The FAA

While state law forms the substantive basis for most of Plaintiffs’ claims, the present

arbitration dispute arises under the FAA. Congress enacted the FAA in 1925 “in response to a

perception that courts were unduly hostile to arbitration.” Epic Sys. Corp. v. Lewis, 138 S. Ct.

1612, 1621 (2018). The statute provides that “agreements to arbitrate [are] ‘valid, irrevocable, and

enforceable, save upon such grounds as exist at law or in equity for the revocation of any

contract.’” AT&T Mobility LLC v. Concepcion, 563 U.S. 333, 336 (2011) (quoting 9 U.S.C. § 2).

In this sense, the FAA places arbitration agreements on equal footing with all other contracts and

requires courts to enforce them according to their terms. Epic Sys., 138 S. Ct. at 1621 (describing

“a liberal federal policy favoring arbitration agreements”) (quoting Moses H. Cone Mem’l Hosp.

v. Mercury Constr. Corp., 460 U.S. 1, 24 (1983)); Mitsubishi Motors Corp. v. Soler Chrysler-

Plymouth, Inc., 473 U.S. 614, 631 (1985) (describing the policy as “emphatic”). Not only must

“questions of arbitrability [ ] be addressed with a healthy regard for the federal policy favoring

arbitration,” but “any doubts concerning the scope of arbitrable issues should be resolved in favor

of arbitration.” Moses, 460 U.S. at 24-25. The FAA “establishes procedures by which federal

courts implement” this “substantive rule.” Rent-A-Ctr., W., Inc. v. Jackson, 561 U.S. 63, 68 (2010).

2 Uber required Plaintiffs to agree to its TSA in 2015, 2019, and 2020. Each TSA is materially

identical with respect to the arbitration provision, Def. SUMF II, ¶¶ 41, 52-54, except that the 2019 and

2020 TSAs clarify that drivers could not opt out of arbitration then if they had not done so initially. Id. ¶

40. As one court explained, “this section provides that, despite any opt out of the 2019 TSA’s arbitration

provision, any existing agreement to arbitrate disputes concerning use of the Uber App remains binding.”

Nicholas v. Uber Techs., Inc., No. 19-08228, 2020 WL 4039382, at *6 (N.D. Cal. July 17, 2020).

Specifically, § 4 permits litigants to seek a court order “directing the parties to proceed to

arbitration in accordance with the terms of the agreement,” while § 3 requires courts to stay

litigation “until such arbitration has been had in accordance with the terms of the agreement” if

the court concludes that the action involves “any issue referable to arbitration.” Id.

The FAA does not “extend to all private contracts, no matter how emphatically they may

express a preference for arbitration.” New Prime Inc. v. Oliveira, 139 S. Ct. 532, 537 (2019). As

such, “a court’s authority . . . to compel arbitration . . . isn’t unconditional.” Id. For example,

“[a]rbitration is strictly a matter of consent, and thus is a way to resolve those disputes—but only

those disputes—that the parties have agreed to submit to arbitration.” Granite Rock Co. v. Int’l

B’hd of Teamsters, 561 U.S. 287, 299 (2010) (emphasis in original). This typically includes

“‘gateway’ questions of ‘arbitrability,’ such as whether the parties have agreed to arbitrate or

whether their agreement covers a particular controversy.” Rent-A-Center, 561 U.S. at 68-69.

Likewise, certain contracts are exempt from the FAA’s coverage entirely. Chief among these are

“contracts of employment of seamen, railroad employees, or any other class of workers engaged

in foreign or interstate commerce.” 9 U.S.C. § 1; Circuit City Stores, Inc. v. Adams, 532 U.S. 105,

109 (2001). This is commonly called the “residual clause,” Waithaka v. Amazon.com, Inc., 966

F.3d 10, 16 (1st Cir. 2020), and the Supreme Court has limited its scope to transportation workers

whose jobs are akin to seamen and railroad workers. Circuit City, 532 U.S. at 118-20 (explaining

that the history of § 1 is “quite sparse” yet very “particular,” that the exemption is reserved for

“transportation workers and their necessary role in the free flow of goods,” and that it must

otherwise be “afforded a narrow construction” to further the FAA’s “purpose of overcoming

judicial hostility to arbitration”). The Supreme Court has also cautioned that “[t]he plain meaning

of the words ‘engaged in commerce’ is narrower than the more open-ended formulations ‘affecting

commerce’ and ‘involving commerce.’” Id. at 118-19. “[A] court should decide for itself whether §

1’s ‘contracts of employment’ exclusion applies before ordering arbitration,” even if an otherwise

valid delegation clause in the arbitration agreement gives the arbitrator the authority to decide that

threshold question. New Prime, 139 S. Ct. at 537-38.

C. Litigation History

i. The Court’s Prior Decision

Notwithstanding the arbitration provision in the TSA and the federal policy in the FAA,

Singh filed suit in this Court alleging that Uber misclassified him as an independent contractor (a

status conferring flexibility but little security) and owes, inter alia, overtime pay plus

reimbursement for business expenses under the Fair Labor Standards Act (“FLSA”), the New

Jersey Hour and Wage Law, and the New York Labor Law. Case No. 16-3044, ECF No. 7. Uber

moved to dismiss and to compel arbitration pursuant to the TSA. Id., ECF No. 5. Singh opposed,

arguing that he fell under the § 1 exemption described above for transportation workers on par

with seaman and railroad employees. I granted Uber’s motion. Id., ECF No. 15, at 8-9. In doing

so, I found the TSA’s arbitration provision to be valid and enforceable. Id. at 8-14, 23-26. I also

found that the TSA’s class waiver provision is permissible under the National Labor Relations Act

(“NLRA”) because Singh could opt out within thirty days without consequence, and that the

delegation clause is lawful under Rent-A-Center. Further, I construed the FAA to exempt only

transportation workers engaged in moving goods across state lines, not people, and ordered the

parties to arbitrate. Id. at 14-15. I did not address any state law questions.

ii. The Third Circuit Decision

On appeal, the Third Circuit vacated and remanded. Singh v. Uber Techs. Inc., 939 F.3d

210 (3d Cir. 2019). Fundamentally, the court held that the FAA’s exemption for transportation

workers is not limited to those engaged in moving interstate goods, but may encompass those who

move interstate passengers, like Uber drivers, or whose work is “so closely related [to interstate

commerce] as to be in practical effect part of it.” Id. at 214. The court then instructed me to

consider, upon remand, whether Singh belongs to such a class of workers and directed the parties

to engage in “limited discovery” on that issue. Id. at 219. If not, the court explained, then the FAA

applies and all remaining issues are reserved for the arbitrator, pursuant to the TSA’s delegation

clause. Id. at 219, 228 (“[T]he District Court shall . . . decide only this aspect of the § 1 residual

clause inquiry, which will be dispositive as to whether the FAA applies.”). The court also identified

certain guideposts to inform the inquiry, “including, but not limited to and in no particular order,

the contents of the parties’ agreement, information regarding the industry in which the class of

workers is engaged, information regarding the work performed by those workers, and various

texts—i.e., other laws, dictionaries, and documents—that discuss the parties and the work.” Id. at

227-28. The court upheld my determination that the arbitration provision (including the delegation

clause) is valid and enforceable, and that the class waiver is permissible, while declining to reach

other issues raised by the parties, such as state law arbitrability questions, to the extent that they

“are contingent on the FAA’s applicability.” Id. at 219.

Calabrese, Cabanillas, and Mechanic subsequently filed a separate suit seeking unpaid

wages and unreimbursed expenses under the FLSA, the New Jersey Wage and Hour Law, and the

New York Labor Law.3 Case No. 19-18471, ECF No. 1; Def. SUMF II, ¶¶ 66-81. When Uber

moved to dismiss, I noted identical issues to Singh’s case, terminated the motion, and ordered

limited discovery on the question whether Plaintiffs are exempt under the FAA. Case No. 19-

18371, ECF No. 33, at 2. Discovery has closed, and Uber has renewed dismissal motions in both

3 This case was originally assigned to the Hon. Madeline Cox Arleo, U.S.D.J., but was transferred

to this Court to be decided with Singh.

cases. As before, Plaintiffs seek to avoid arbitration under the FAA based on the residual clause,

compel Uber to comply with state labor laws, and classify them as employees not independent

contractors.

iii. Other Third Circuit Decisions

While this matter was on remand and in discovery, the Third Circuit decided another FAA

case, Harper v. Amazon.com Services, Inc., 12 F.4th 287 (3d Cir. 2021). There, a divided panel

held that a court facing an arbitrability dispute in a case such as this one must proceed as follows:

(1) consider, based on the face of the complaint and related documents alone, whether the FAA

governs or whether the residual clause applies; (2) if the answers to those questions are “murky,”

consider whether the dispute must be arbitrated under applicable state law; and (3) if state law does

not compel arbitration, go back to the FAA for limited discovery. Id. at 296. The stated logic for

resolving state law arbitrability before turning to “questions of fact and discovery” under the FAA

is that doing so “honors the principles of federalism” while preventing “delays, costs, and

uncertainty,” since “the parties might still have an enforceable agreement to arbitrate under state

law.” Id. at 294. The majority also explained that FAA fact-finding “can always come later.” Id.

at 296. This conclusion appears to center on doubt regarding “the judicially created presumptions

atop both §§ 1 and 2 of the FAA,” as well as a desire to place “everyone back to the starting line

in the text of the law.” Id. at 297-98 (Matey, J., concurring).

It is difficult to reconcile Harper with the sequence of decisions in Singh, where the Third

Circuit advised that discovery should be ordered before “leav[ing] it to the District Court to

address” state-law arguments as to arbitrability, if necessary. 939 F.3d at 228. As the dissent noted,

this is the usual approach in both the Third Circuit and other circuits. Notwithstanding federalism

principles, it respects the parties’ “chosen law,” effectuates the “plain language” of their agreement

to specify the FAA as the rule of decision, and avoids potentially “tricky” state issues. Harper, 12

F.4th at 303-06 (Shwartz, J., dissenting). Nevertheless, since the Third Circuit ordered discovery

in Singh before deciding Harper, and discovery is now complete, the Harper three-step analysis

is not applicable here. That is, I need not set aside the evidentiary record the parties have developed

and proceed to state law questions before adjudicating fact issues. Following Singh, at least one

other court in this district ordered the parties to engage in discovery to determine whether the

FAA’s residual clause is triggered, notwithstanding pending state law arbitrability questions.

Gonzalez v. Lyft, Inc., No. 19-20569, 2021 WL 303024, at *5-6 (D.N.J. Jan. 29, 2021). I thus

continue with the FAA, and find that it applies, though I nonetheless conclude that state law would

yield the same result—arbitration is compelled.

II. LEGAL STANDARD

On remand, I apply the standard for summary judgment. Singh, 939 F.3d at 219 (“If Uber

chooses to reassert its motion after this discovery is completed, the District Court shall apply the

summary judgment standard under Federal Rule of Civil Procedure 56.”); Aliron Int’l, Inc. v.

Cherokee Nation Indus., Inc., 531 F.3d 863, 865 (D.C. Cir. 2008) (same). Summary judgment is

appropriate where “the pleadings, depositions, answers to interrogatories, and admissions on file,

together with the affidavits if any, . . . demonstrate the absence of a genuine issue of material fact”

and that the moving party is entitled to a judgment as a matter of law. Celotex Corp. v. Catrett,

477 U.S. 317, 322-23 (1986) (quotations omitted). An issue is “genuine” when “a reasonable jury

could return a verdict for the non-moving party.” Anderson v. Liberty Lobby, Inc., 477 U.S. 242,

248 (1986). A fact is “material” when it “might affect the outcome of the suit under the governing

law.” Id. I construe all facts in the light most favorable to the nonmoving party, Boyle v. Cty. of

Allegheny Pa., 139 F.3d 386, 393 (3d Cir. 1998), whose evidence “is to be believed,” and I make

“all justifiable inferences . . . in [its] favor.” Marino v. Indus. Crating Co., 358 F.3d 241, 247 (3d

Cir. 2004); Wishkin v. Potter, 476 F.3d 180, 184 (3d Cir. 2007).

The moving party “always bears the initial responsibility of informing the district court of

the basis for its motion.” Celotex, 477 U.S. at 323. That party may discharge its burden by

“showing — that is, pointing out to the district court — that there is an absence of evidence to

support the nonmoving party’s case when the nonmoving party bears the ultimate burden of

proof.” Singletary v. Pa. Dep’t of Corr., 266 F.3d 186, 192 n.2 (3d Cir. 2001) (quotations and

citations omitted). The nonmoving party must then identify, by affidavits or otherwise, specific

facts showing that there is a triable issue. Celotex, 477 U.S. at 324. To do so, the nonmoving party

“may not rest upon the mere allegations or denials of the . . . pleading[s].” Saldana v. Kmart Corp.,

260 F.3d 228, 232 (3d Cir. 2001) (quotations omitted). Instead, “[it] must make a showing

sufficient to establish the existence of [every] element essential to [its] case, and on which [it] will

bear the burden of proof at trial.” Cooper v. Sniezek, 418 Fed. App’x. 56, 58 (3d Cir. 2011)

(quotations and citations omitted). “While the evidence that the non-moving party presents may

be either direct or circumstantial, and need not be as great as a preponderance, [it] must be more

than a scintilla,” Hugh v. Butler Cnty. Family YMCA, 418 F.3d 265, 267 (3d Cir. 2005), and

conclusory declarations, even if made in sworn statements, will not suffice. Lujan v. Nat’l Wildlife

Fed’n, 497 U.S. 871, 888 (1990).

III. DISCUSSION

A. The Scope of the Dispute

At the outset, it is critical to clarify what is not in dispute, as I do not write on a blank slate.

First, the FAA governs the TSA. This is so because the TSA is a contractual provision evincing an

intention to settle certain (if not all) disputes by arbitration, Mohamed v. Uber Techs., Inc., 848

F.3d 1201, 1208-09 (9th Cir. 2016), and because the TSA indicates that the FAA governs it, i.e.,

stipulates to the statute’s application. Uhl v. Komatsu Forklift Co., Ltd., 512 F.3d 294, 302-03 (6th

Cir. 2008). There is also no remaining doubt as to the validity or enforceability of the TSA’s

arbitration provision (or its class waiver or delegation clause), which I upheld in my prior decision

and the Third Circuit affirmed. To the extent that Plaintiffs continue to argue to the contrary, I do

not consider it. See, e.g., Pl. Rep. Br. I, at 6-10.

Next, as many courts have held, I define the relevant class of transportation workers as

Uber drivers nationwide, not in any particular state, region, or locality.4 See, e.g., Davarci v. Uber

Techs., Inc., No. 20-9224, 2021 WL 3721374, at *8 (S.D.N.Y. Aug. 20, 2021); Osvatics v. Lyft,

Inc., No. 20-1426, 2021 WL 1601114, at *11 (D.D.C. Apr. 22, 2021) (Ketanji, J.) (holding that

contrary position “would undermine the underlying purposes of the FAA”); Islam v. Lyft, Inc., 524

F. Supp. 3d 338, 350 (S.D.N.Y. 2021) (“[I]t would be illogical if Lyft drivers performing the same

work for the same company in different cities were to have completely different rights and

obligations.”); Sienkaniec v. Uber Techs., Inc., 401 F. Supp. 3d 870, 872 (D. Minn. 2019) (“Uber

drivers [nationwide] perform the same job for the same company pursuant to the same

agreement.”); Aleksanian v. Uber Techs. Inc., 524 F. Supp. 3d 251, 261 (S.D.N.Y. 2021) (rejecting

argument attempting to “frame the class of workers as ‘New York City Uber drivers’”); Buckeye

Check Cashing, Inc. v. Cardegna, 546 U.S. 440, 443 (2006) (“The FAA embodies a

“national policy favoring arbitration.”). Both parties accept a class definition at this level of

4 At one point, Singh confusingly states that Uber drivers belong to a nationwide class that includes

“all other transportation workers employed in the transportation industry who work . . . [on] public interstate

highways and roads[],” such as “bus line workers.” Pl. Rep. Br. I, at 3-4. Singh provides no legal authority

for this proposition, and if it were true, then seemingly every transportation worker (except perhaps airline

workers) would be exempt under § 1—an exception that would swallow the rule and contravene the plain

text of the residual clause requiring courts to take each “class” of transportation workers on its own terms

and to exempt only those on par with seamen and railroad employees.

generality. Pl. Br. I, at 9 (“Viewing Uber rideshare drivers nationwide is an appropriately bounded

class to determine application of the FAA [ ] exemption.”); Def. Br. I, at 13-14 (same). Importantly,

along these lines, “the inquiry regarding § 1’s residual clause asks a court to look to classes of

workers rather than particular workers,” so Plaintiffs’ individual driving histories are not relevant,

except to the extent that they reflect the work of the class on the whole. Singh, 939 F.3d at 227;

Capriole v. Uber Techs., Inc., 460 F. Supp. 3d 919, 929 (N.D. Cal. 2020) (“[T]he relevant inquiry

is not whether an individual driver has crossed state lines, but whether the class of drivers crosses

state lines.”); Rogers v. Lyft, Inc., 452 F. Supp. 3d 904, 915-16 (N.D. Cal. Apr. 7, 2020) (“The

plaintiffs’ personal exploits are relevant only to the extent they indicate the activities performed

by the overall class.”). In other words, a member of a class of transportation workers who never

personally crosses a border may still be “engaged in interstate commerce,” so long as the class on

the whole is defined by such engagement, and conversely, a class of transportation workers may

not qualify for the residual clause exemption even if its members occasionally or often perform

that kind of work. The focus is always the overall class.

While Plaintiffs further dispute which party has the burden of proof, there is no question

that Plaintiffs bear it, because they are challenging the applicability of the TSA. See, e.g., Singh,

939 F.3d at 231-32 (Porter, J., concurring in part and concurring in the judgment) (“Singh bears

the burden on remand to show why the District Court should not compel arbitration under the

FAA.”); Green Tree Fin. Corp.-Ala. v. Randolph, 531 U.S. 79, 91 (2000) (“[T]he party resisting

arbitration bears the burden of proving that the claims at issue are unsuitable for arbitration.”);

Capriole, 460 F. Supp. 3d at 928 (“As the party opposing arbitration, Plaintiffs have the burden of

proving that the exemption applies.”); Osvatics, 2021 WL 1601114, at *6. Under the FAA, once a

party seeking to enforce an arbitration agreement carries its initial burden as to the agreement’s

validity, as Uber has done, the burden shifts to the opposing party to show that § 1 applies. See,

e.g., Bean v. ES Partners, Inc., No. 20-62047, 2021 WL 1239899, at *2 (S.D. Fla. Apr. 4, 2021)

(“The party resisting arbitration bears the burden of showing that [the residual clause] exemption

applies.”); Smith v. AllState Power Vac, Inc., 482 F. Supp. 3d 40, 45 (E.D.N.Y. 2020) (“A plaintiff

opposing arbitration under the FAA has the burden of demonstrating the exemption.”) (citation

omitted).

Finally, I reject Uber’s contention that, because the TSA is merely a software license, the

arbitration provision does not appear in a “contract of employment” in the sense of the FAA. In

Singh, the Third Circuit stated that “New Prime eliminated Uber’s ‘contract of employment’

argument, so we are left with its transportation-of-goods and ‘engaged in interstate commerce’

arguments.” 939 F.3d at 217; New Prime, 139 S. Ct. at 541 (“Congress used the term ‘contracts of

employment’ in a broad sense to capture any contract for the performance of work by workers.”)

(emphasis in original). The Third Circuit also stated that whether Plaintiffs are engaged in

interstate commerce or “sufficiently related work” is “dispositive as to whether the FAA applies,”

an implicit acknowledgement that the FAA’s “contract of employment” criteria is satisfied here.

Singh, 939 F.3d at 219. And, “[b]y now, it is axiomatic that on remand for further proceedings

after decision by an appellate court, the trial court must proceed in accordance with the mandate

and the law of the case as established on appeal.” Bankers Trust Co. v. Bethlehem Steel Corp., 761

F.2d 943, 949 (3d Cir. 1985) (quotations omitted). Lyft, another rideshare company, has conceded

this point in similar litigation, notwithstanding the underlying dispute about whether drivers are

employees or independent contractors. Osvatics, 2021 WL 1601114, at *8 n.5. Further, Uber has

failed to identify any court that has found its contrary position availing. Davarci, 2021 WL

3721374, at *9 n.20 (concluding same). As the purported software license provides, Plaintiffs may

use it “solely for the purposes of providing transportation services.” Def. SUMF I, ¶ 19; Def.

SUMF II, ¶¶ 24-25. That is, for “the performance of work.” New Prime, 139 S. Ct. at 541 (emphasis

omitted). Having resolved these issues, I turn to the question before me: what it means to be

“engaged” in interstate commerce under the FAA, which I answer based on the record the parties

developed in discovery.

B. The Legal Landscape

Since my decision in 2019, many district courts have decided similar FAA cases based on

varying levels of factual development. The majority view is that rideshare drivers nationwide do

not engage in interstate commerce and are not covered by the residual clause in the FAA. See, e.g.,

Davarci, 2021 WL 3721374; Osvatics, 2021 WL 1601114; Capriole, 460 F. Supp. 3d 919;

Aleksanian, 2021 WL 860127; Rogers, 452 F. Supp. 3d 904; Sienkaniec, 401 F. Supp. 3d 870;

Hinson v. Lyft, Inc., No. 20-2209, 2021 WL 838411 (N.D. Ga. Feb. 26, 2021); Heller v. Rasier,

LLC, No. 17-8545, 2020 WL 413243 (C.D. Cal. Jan. 7, 2020); Tyler v. Uber Techs., Inc., No. 19-

3492, 2020 WL 5569948 (D.D.C. Sept. 17, 2020); see also In re Grice, 974 F.3d 950, 954 (9th

Cir. 2020) (denying plaintiff’s petition for a writ of mandamus and holding that district court’s

decision that Uber drivers did not fall within residual clause of Section 1 exemption was not

“clearly erroneous as a matter of law”). These courts reason that nationwide rideshare drivers

complete a small percentage of cross-border rides (e.g., just over 2%), and their overall driving

activities demonstrate that they serve a fundamentally local transportation function.

The majority view culminated in a recent Ninth Circuit decision, where a court of appeals

directly addressed the issue for the first time, found in favor of Uber, and compelled arbitration.

Capriole v. Uber Techs., Inc., 7 F.4th 854 (9th Cir. 2021). Specifically, the Ninth Circuit found

Uber drivers nationwide not to be “engaged in interstate commerce” in the sense of the FAA

because “Uber trips are often short and local” as well as “primarily . . . intrastate in nature,”

crossing state lines just a fraction of the time largely as a result of geography, they “only

infrequently involve . . . a trip to a transportation hub” such as an airport, and interstate movement

“cannot be said to be” central to what drivers do. Id. at 864-65. A few months later, the First Circuit

rejected the same twin arguments as the Ninth Circuit, finding that drivers do not “fit within

the section 1 exemption [just] because some of them occasionally transport passengers across state

lines” and because they occasionally transport passengers to Logan International Airport.

Cunningham v. Lyft, Inc., Nos. 20-1373, 20-1379, 20-1544, 20-1549, 20-1567, 2021 WL 5149039,

at *3 (1st Cir. Nov. 5, 2021). According to that court, drivers are “among a class of workers

engaged primarily in local intrastate transportation, some of whom infrequently find themselves

crossing state lines, and are thus fundamentally unlike seamen and railroad employees when it

comes to their engagement in interstate commerce.” Id. *6-7.

Two district courts have bucked this trend, holding that rideshare drivers are transportation

workers engaged in interstate commerce under the FAA. Islam, 524 F. Supp. 3d 338; Haider v.

Lyft, Inc., No. 20-2997, 2021 WL 1226442 (S.D.N.Y. Mar. 31, 2021). What matters to these courts

is that rideshare drivers nationwide complete tens of millions of interstate trips per year, regardless

of whether that equals just 2% of all rides, and frequently pick up and drop off passengers at

airports.

C. Residual Clause Analysis

Notwithstanding the limited momentum (in the Southern District of New York) for

Plaintiffs’ theory, I agree with the majority view that nationwide rideshare drivers are not a class

of transportation workers engaged in interstate commerce.

i. The Standard

The Third Circuit has construed the FAA’s residual clause to cover workers “actually

engaged in the movement of interstate or foreign commerce or in work so closely related thereto

as to be in practical effect part of it.” Tenney Eng’g, Inc. v. United Elec. Radio & Mach. Workers

of Am., (U.E.) Local 437, 207 F.2d 450, 452 (3d Cir. 1953) (en banc) (citing Shanks v. Del.,

Lackawanna & W.R. Co., 239 U.S. 556, 558 (1916)). The Third Circuit declared Tenney good law

in Great W. Mortgage Corp. v. Peacock, 110 F.3d 222, 226-27 (3d Cir. 1997), and again in Palcko

v. Airborne Express, Inc., 372 F.3d 588, 593 (3d Cir. 2004). Singh followed suit, describing Tenney

as requiring at a minimum active engagement in work that is “sufficiently related” to interstate

commerce. 939 F.3d at 219. Singh also set forth multiple factors calculated to determine whether

a class of transportation workers satisfies Tenney. Id. at 227. The factors include “the contents of

the parties’ agreement, information regarding the industry in which the class of workers is

engaged, information regarding the work performed by those workers, and various texts—i.e.,

other laws, dictionaries, and documents—that discuss the parties and the work.” Id. (“The District

Court must be equipped with a wide variety of sources.”). The factors are co-equal and non-

exhaustive. Id. (“Nor must its analysis hinge on any one particular factor, such as the local nature

of the work.”).

Singh squares with the inquiry in the First and Ninth Circuits. See, e.g., Waithaka, 966 F.3d

at 22 (analyzing the “geographic footprint and nature of the business for which [delivery drivers]

work,” among other things, because “workers’ activities are not pursued for their own sake” but

to “carry out the objectives of a business,” and because seamen and railroad employees are

“defined by the nature of the business for which they work”); Rittmann v. Amazon.com, Inc., 971

F.3d 904, 917-18 (9th Cir. 2020) (focusing on same); Grice, 974 F.3d at 956 (same). The Seventh

Circuit has summed up the inquiry in its own words: “actual engagement in interstate commerce”

means that the interstate work must be a “central part” of the overall work of the class. Wallace v.

Grubhub Holdings, Inc., 970 F.3d 798, 801 (7th Cir. 2020) (Barrett, J.) (“That is the inquiry that

Circuit City demands.”). Though somewhat different in form, the centrality test in Wallace is not

different in substance from the approach in other circuits. The Ninth and First Circuits relied on

Wallace when they found that rideshare drivers are subject to arbitration under the FAA. Capriole,

7 F.4th at 865; Cunningham, 2021 WL 5149039, at *6. The same goes for virtually every district

court to consider the issue, including those holding the minority view. Davarci, 2021 WL 3721374,

at *11 (focusing on “whether a central feature of class members’ jobs involves interstate

commerce”); Osvatics, 2021 WL 1601114, at *12 (“[W]hether a worker is ‘engaged in . . .

interstate commerce’ turns on whether ‘interstate movement’ is ‘a central part of the class

members’ job description.’”) (citation omitted); Aleksanian, 524 F. Supp. 3d at 260-61 (holding

“cases out of the First, Third, and Ninth Circuit,” including Singh, “support rather than contradict”

the Seventh Circuit’s holding as to the centrality of the interstate work); Islam, 524 F. Supp. 3d at

344 (“Put another way, as the Seventh Circuit did in Wallace, the relevant inquiry is whether the

‘interstate movement of goods [or people] is a central part of the class members’ job description.’”)

(citation omitted).

The centrality of the interstate work is likewise implicit in the Singh factors, particularly

Singh’s focus on “the information regarding the [rideshare] industry,” “the work performed by

[rideshare] workers,” and the “extent to which [drivers’] activities constitute engagement in

interstate commerce.” 939 F.3d at 227 & n.11 (emphasis added). It also inheres in Tenney, whose

benchmark is work “so closely related” or “sufficiently related” to interstate travel. If driving for

Uber does not, at its core, involve transporting riders across state lines, then I would be hard-

pressed to find that drivers could “actually engage[] in the movement of interstate or foreign

commerce” or that their work could be “sufficiently related” to it. Tenney, 207 F.2d at 453. Most

importantly, comparing a class of workers to seamen and railroad employees necessarily

implicates the centrality of the interstate work, for a core feature of those jobs is interstate

movement. Davarci, 2021 WL 3721374, at *9 (“The transportation of goods or passengers in the

flow of interstate commerce must be a definitional feature of the workers’ job duties, such that the

work of the class can be deemed analogous to that of seamen and railroad employees, whose

occupations center on the transportation of goods or persons in interstate commerce.”); Osvatics,

2021 WL 1601114, at *12 (same); Eastus v. ISS Facility Servs., Inc., 960 F.3d 207, 210 (5th Cir.

2020) (examining whether a class of workers is “engage[d] in the movement of goods in interstate

commerce in the same way that seamen and railroad workers are”) (quotations and citation

omitted). Or, as the concurrence put it in Harper, the inquiry turns on whether the interstate work

forms an “ordinary and regular part of the class of work.” 12 F.4th at 302.

The case law teaches that one other principle bears on the residual clause analysis: crossing

state lines is not the “touchstone of the exemption’s test.” Waithaka, 966 F.3d at 25. Indeed,

crossing state lines is not a necessary condition because, for instance, “workers moving goods or

people destined for, or coming from, other states” may be “engaged in interstate commerce” even

if the workers are “responsible only for an intrastate leg of that interstate journey.” Id. at

22; Rittmann, 971 F.3d at 915; Osvatics, 2021 WL 1601114, at *12. By the same token, crossing

state lines by itself may not be sufficient to trigger the residual clause because, as the Eleventh

Circuit has described, “a pizza delivery person who deliver[s] pizza across a state line to a customer

in a neighboring town” cannot claim the FAA’s exemption solely for that reason. Hill v. Rent-A-

Ctr., Inc., 398 F.3d 1286, 1289-90 (11th Cir. 2005). All told, what matters is “[t]he nature of the

business for which a class of workers perform[s] their activities” and other industry-related factors

under Singh, rather than whether the workers actually “cross[] state lines.” Grice, 974 F.3d at 956

(quotations and citation omitted). With these principles in hand, and having defined the contours

of the analysis, I turn to the Singh factors.

ii. The Factors

1. The TSA

I begin with the TSA. While both parties point to it as evidence of the type of work in

which Uber drivers engage, it hardly settles the issue. The TSA provides for “Transportation

Services,” which, Uber points out, are defined as the “provision of [peer to peer] passenger

transportation [ ] to Users via the Uber Services in the Territory using the Vehicle.” Def. SUMF I,

¶ 27. “Territory” refers to “the city or metro area in the United States in which [drivers] are enabled

by the Driver App to provide [rides].” Id. ¶ 28. Uber’s “operations are organized on a city-by-city

basis” and its business is “very city-oriented” to comply with the disparate patchwork of

ridesharing regulations governing drivers. Id. ¶¶ 33, 37. Uber adds that, when drivers sign up to

use its app, they input “the city in which they want that account associated to,” which becomes the

“home city” and whose regulations control where they may pick up passengers. Id. ¶¶ 34-35. Thus,

in some territories, drivers may not make pick-ups in states other than the state in which their home

city is located. Id. ¶ 36.

At the same time, Plaintiffs note, Uber has created many multi-state territories. Pl. Br. II,

at 31. This is true in parts of 34 states, and 47% of Uber’s 217 territories span borders. Pl. SUMF

II, ¶¶ 28, 32. Likewise, many purportedly single state territories come with exceptions. For

instance, in Pennsylvania, there are 11 territories in which drivers are limited to pick-ups within

the state, but in the Philadelphia territory, some drivers may pick up riders in New Jersey and

Delaware, while in the Lehigh Valley territory, some drivers can pick up riders in New Jersey, and

in the New York territory, some drivers may pick up riders in New Jersey, but not vice versa. Pl.

Br. II, Ex. D. Plaintiffs also reasonably point out that Uber’s territory system did not exist before

2016, during which time drivers could begin trips anywhere, and at present it only limits where

drivers can receive ride requests. Drivers may complete rides to any place a car can go. As well,

Uber designs its service to facilitate interstate travel; its fare schedule contemplates interstate travel

in that it is based in part on time and distance; its deactivation policy penalizes drivers for declining

interstate pick-ups in multi-state territories; and it markets its business as “transportation as reliable

as running water everywhere for everyone.” Because this evidence weighs equally on both sides,

and because I cannot substitute general notions about Uber’s business for actual data in the record,

the TSA alone does not answer the question whether rideshare drivers are engaged in interstate

commerce.

2. The Rideshare Industry

What matters are driving statistics for Uber.5 The parties rely on this information, indeed

some of the same data points, to reach opposite conclusions about the nature of Uber’s business

and whether the residual clause applies. Uber argues that its drivers do not engage in interstate

commerce because fewer than 2.5% of all trips cross state lines, and those that do cover on average

only 13.5 miles, last on average 30 minutes, and occur largely by the happenstance of geography,

such as when a rider needs to travel from Arlington, Virginia, to Washington, D.C., or Hoboken,

New Jersey, to Manhattan in New York City. Def. SUMF I, ¶¶ 39-42, 104. In this sense, Uber

maintains, “the way drivers use [the app] is very, very localized.” Id. ¶ 38.

While in Uber’s estimation the proportion of interstate rides is relatively rare, according to

Plaintiffs, such rides are nevertheless numerically many: 2% of all Uber rides nationwide nets to

5 All statistics run through May 2020. Def. SUMF I, ¶ 43 nn.1-2.

140 million interstate rides since 2010, equal to one interstate ride every second for the last three

years and an average of 32 million interstate rides per year from 2017 to 2019. Pl. SUMF I, ¶¶ 12-

14. Plaintiffs also respond that Uber skews the data, and elides the true character of the class work,

by emphasizing the ratio of interstate to intrastate rides. If any ratio matters, Plaintiffs propose, it

is that in any given year approximately 16% of drivers cross state lines to complete a ride, which

equals 500,000 drivers per year from 2017 to 2019. According to Plaintiffs, this is a much more

revealing datapoint than Uber’s “2% of all rides” statistic. Pl. Br. II, Ex. A.

Plaintiffs further rely on the activity of drivers who complete the most rides, drive the most

miles, and spend the most time “on the clock,” or drivers with 50 or more rides per year, which

they argue offer the best picture of what rideshare drivers nationwide actually do. Pl. Br. I, at 22.

This cohort must be the focal point, Plaintiffs contend, because 60% of drivers leave within 6

months of signing up for Uber, 59% average less than 2 hours of driving prior to ending their

employment, 50% work less than 10 hours per week, and 5% perform 45% of all rides. Pl. SUMF

I, ¶¶ 18-21. Viewed through the lens of Plaintiffs’ “most active drivers” measure, 27% of drivers

with 50 or more annual trips completed an interstate ride between 2010 and 2020, comprising

99.3% of all interstate rides. Id. ¶ 19; Pl. Br. II, at 23. Finally, according to Plaintiffs, Uber drivers

frequently pick up/drop off passengers at airports who are heading to/returning from interstate

travel, which is sufficiently related to interstate commerce. For instance, 10% of all Uber rides

start or end at an airport, and 15 percent of Uber’s gross bookings are airport trips, which nets to

more than 500 million airport rides from 2016 to 2020. Pl. Br. II, Ex. C.

a. Driver/Ride Data

Plaintiffs’ argument that 140 million interstate rides is sufficient to qualify all Uber drivers

as “actually engaged in interstate commerce” has a certain surface-level appeal, especially when

Plaintiffs characterize these rides as happening every second of every day. Similar arguments have

persuaded a minority of district courts in the Southern District of New York. Islam, 2021 WL

871417, at *8 (finding that rideshare drivers conduct tens of millions of interstate rides each year,

meaning rideshare drivers “perform sufficient numbers of interstate rides, with sufficient

regularity, to make them ‘engaged in’ interstate commerce,” “even if interstate transportation is

not the predominant daily service provided by rideshare drivers”); Haider, 2021 WL 1226442, at

*3-4 (finding that “the sheer number of interstate trips rideshare drivers make places them ‘within

the flow of interstate commerce’”) (quoting Circuit City, 532 U.S. at 118). And, admittedly, the

circumstances of this case fall into a grey area on the interstate commerce spectrum. Courts have

long recognized that “there is no[] clear definition or consensus of what constitutes a

‘transportation worker’ who is ‘engaged in interstate commerce,’” Heller, 2020 WL 413243, at

*6, and there is “a gap in the case law . . . between cases in which no member of a class transported

goods or services across a state line and cases in which all members of a class did so.” Sienkaniec,

401 F. Supp. 3d at 872 (emphasis in original).

I nevertheless agree with the majority view (espoused in both the First and Ninth Circuits)

that focusing solely on the number of cross-border trips produces a simple but incorrect answer.

Plaintiffs first falter by assigning too much weight to this metric. The lynchpin is not “the

frequency vel non with which a type of worker traverses state boundaries,” Davarci, 2021 WL

3721374, at *12, because crossing state lines is neither necessary nor sufficient to trigger § 1.

Osvatics, 2021 WL 1601114, at *12; Rogers, 452 F. Supp. 3d at 915 (“[T]he fact that some workers

cross state lines in the course of their duties does not mean that the class of workers as a whole is

engaged in interstate commerce.”); Aleksanian, 524 F. Supp. 3d at 262 (“[J]ust because Uber is set

up to handle the occasional interstate trip does not mean that ‘interstate movement of goods is a

central part of the job description of the class of workers to which [Plaintiffs] belong.’”) (quoting

Wallace, 970 F.3d at 800). In fact, insofar as Plaintiffs rely on cases holding that traversing borders

is not the sine qua non of the residual clause, they tacitly concede that that the total tally of

interstate trips cannot be dispositive and should not bear the most significance. Davarci, 2021 WL

3721374, at *12 (“[T]hose cases, in this Court’s opinion, directly call into question the minority

view’s reliance on the raw number of trips rideshare drivers purportedly take across state

borders.”). “The raw number of cross-border trips conducted by Uber drivers is [thus] irrelevant

to the ultimate inquiry.”6 Id. at *11.

Even accepting Plaintiffs’ argument on its terms, it still runs headlong into the disparity at

the heart of this case: interstate trips, albeit numerically many, do not constitute a central part of

what Uber does when placed in the context of its drivers’ overall work activities. Capriole, 7 F.4th

at 865-66. For one thing, tens of billions of rides are local, never crossing a border, equal to dozens

of rides per second, as almost any Uber rider would attest. Davarci, 2021 WL 3721374, at *9

(“The vast majority [97.5%] of Uber drivers’ trips are purely intrastate.”); Osvatics, 2021 WL

1601114, at *13 (holding that rideshare drivers “offer services that are primarily local and

intrastate in nature”); Capriole, 460 F. Supp. 3d at 932 (“Uber drivers do not perform an integral

role in a chain of interstate transportation.”); Rogers, 452 F. Supp. 3d at 916 (“[Drivers’] work

predominantly entails intrastate trips, an activity that undoubtedly affects interstate commerce but

6 Two examples helpfully illustrate this point. As the Davarci Court wrote, a bartender who spends

all of his or her time giving advice to patrons is not “engaged in” therapy or counseling as the Supreme

Court has construed that term. 2021 WL 3721374, at *12 & n.21. Conversely, as the Islam Court observed,

“any federal district judge would answer in the affirmative if asked whether or not she was ‘engaged in’

conducting criminal trials, notwithstanding that most federal judges’ dockets consist primarily of civil

actions and the majority of criminal prosecutions are resolved by a guilty plea. Overseeing criminal trials

is unquestionably a ‘central part of [a federal district judge’s] job description,’ even if it is not something

that she does every day or even every month.” 524 F. Supp. 3d at 351 (citation omitted). These affirm that

the focus of the residual clause inquiry must remain on the nature of the business and the centrality of the

interstate work, not numerical tallies.

is not interstate commerce itself.”). The fact that drivers completed 140 million interstate rides

over the last 10 years is more likely a function of Uber’s popularity or the rapidly growing “scope

of its operations,” Harper, 12 F.4th at 302, than the nature of its business and the centrality of

interstate rides to the work of the class.

Also problematic for Plaintiffs: their total rides tally is “likely influenced by the fact that

many interstate trips are performed by drivers (or for riders) who live close to state borders,

especially on the East Coast,” thereby eliding the nationwide character of the class. Capriole, 7

F.4th at 864. Similarly, interstate trips largely resemble intrastate trips in time and distance,

suggesting that when they do happen, they result from a combination of geography plus routine

travel patterns, buttressing the fundamentally local transportation function provided by drivers and

the nominal or incidental character of the interstate trips, no matter how numerous they are. See,

e.g., Def. SUMF I, ¶ 39 (finding that interstate trips average 13.5 miles and 30 minutes, while

intrastate trips average 6.1 miles and 16.6 minutes); Davarci, 2021 WL 3721374, at *9 (relying on

these statistics and holding that “[o]nce the Court concludes that interstate trips are merely

incidental to Uber drivers’ local transportation function, there is no raw number of interstate trips

that can transform them into workers who are engaged in interstate commerce”); Rogers, 452 F.

Supp. 3d at 916 (same). In other words, “Uber drivers, even when crossing state lines . . . , are

‘merely convey[ing] interstate . . . passengers between their homes and [their destination] in the

normal course of their independent local service.’” Capriole, 7 F.4th at 865 (citation omitted).

Contrast this with seamen and railroad workers, for whom the interstate movement of people and

things over long distances and across state lines is intrinsic to the type of work they

perform. Wallace, 970 F.3d at 803.

Plaintiffs’ alternative interpretation of the data only underscores this point: though they

make much of the fact that 16% of all drivers gave at least one interstate ride from 2010 to 2020,

the flip side is that 84% of drivers never did so, and New Jersey and New York drivers gave almost

8 million interstate trips in 2019, or nearly one-fourth of all such trips that year. Insofar as Plaintiffs

eschew these statistical nuances, their position is at once overinclusive of “Uber’s service in

metropolitan markets” nearby other states, yet underinclusive of its service in major markets far

from other state borders—for example, San Francisco, which is 200 miles from Nevada, or Austin,

which is 125 miles from Oklahoma, or even Massachusetts, where 99.7% of percent of rides begin

and end in-state despite its proximity to other states in the dense New England region. Capriole,

460 F. Supp. 3d at 929. In any event, accepting arguendo Plaintiffs’ focus on 16% of all drivers,

interstate trips still constitute a fraction of their workload, not a central or definitional feature. For

half of this cohort, not more than 4.2% of rides crossed state lines between 2015 and 2020, while

for 80%, it was fewer than one-fifth, and 5% completed 80% of all trips. Def. SUMF I, ¶¶ 43-44,

46.

Plaintiffs further “inexplicably” limit their analysis to the most active drivers, or those with

50+ trips per year. Capriole, 7 F.4th at 866; cf. Haider, 2021 WL 1226442, at *3 (focusing on

“full-time” divers). Their basis for doing so is that most Uber drivers work for a short period of

time, then quit. Assuming that is true, it still does not permit the inference that the work patterns

of the most active drivers transform the entire class into one engaged in interstate commerce. For

one thing, the crux of the inquiry is not whether a certain proportion of a specific kind of driver’s

work is out-of-state, but whether the entire category of workers to which the driver belongs is a

part of the stream of interstate commerce based on its overall work. See, e.g., Hinson, 522 F. Supp.

3d at 1261. To glean the nature of Uber’s business and the core of all ridesharing jobs from a

narrow subset of drivers, and to assume their experience is universal to the class, risks undermining

the well-established rule that “the idiosyncratic patterns of [ ] drivers . . . are largely irrelevant,”

Davarci, 2021 WL 3721374, at *9 (collecting cases), and “personal exploits” matter only insofar

as “they indicate the activities performed” by workers overall. Rogers, 452 F. Supp. 3d at 915.

In response, Plaintiffs contend that they are doing just that: identifying certain structural

features of Uber’s business exemplified by the most active drivers, not conveniently limiting the

class to that group. Plaintiffs’ position fails to persuade because high attrition and low productivity

are also defining class characteristics, certainly as much as the fact that a small number of drivers

do a large percentage of the work. This is the gig economy: on-demand jobs, supplemental income,

tradeoffs between stability and flexibility, and a market where many drivers use multiple apps at

once, such as Uber and Lyft, or multiple platforms, such as Uber Rides and Uber Eats, to earn the

equivalent of full-time pay. See, e.g., Christopher Mims, In a Tight Labor Market, Gig Workers

Get Harder to Please: Companies like Uber, Lyft, Postmates, and Instacart Could Run Out of Man

Power As High Turnover Plagues the Side-Hustle Economy, WALL ST. J. (May 4, 2019)

(describing turnover as high as 500% per year, workers who use as many as eight different apps

to supplement income, and incentive programs to encourage drivers to stay); Sarah Kessler,

GIGGED: THE END OF THE JOB AND THE FUTURE OF WORK (2018) (describing a structural shift away

from permanent employment and to delivery services coordinated by apps). It is inappropriate to

exclude this type of work and these kinds of drivers from the nature of Uber’s business or the core

activities of the class.

Finally, precedent teaches that “the residual clause must be interpreted in light of the

specifically enumerated categories of workers that directly precede it, consistent with the ejusdem

generis canon of statutory construction.” Waithaka, 966 F.3d at 17 (citing Circuit City, 532 U.S.

at 118); Wallace, 970 F.3d at 801-02 (holding that interstate transportation work under the residual

clause must be on par with seamen and railroad employees); Asplundh Tree Expert Co. v. Bates,

71 F.3d 592, 600-01 (6th Cir. 1995) (holding that § 1 “should be narrowly construed to apply to

employment contracts of seamen, railroad workers, and any other class of workers actually

engaged in the movement of goods in interstate commerce in the same way that seamen and

railroad workers are”); Harper, 12 F.4th at 293. For seamen and railroad workers, interstate travel

is a central, intrinsic, practically unavoidable task. See, e.g., CSX Transp., Inc. v. Healey, 861 F.3d

276, 278 (1st Cir. 2017) (referring to the railroad industry as a “quintessentially interstate

business”); Baker v. United Transp. Union, AFL-CIO, 455 F.2d 149, 153-54 (3d Cir. 1971) (stating

that the railroads “remain the backbone of much of our interstate transportation system” and

describing the railroads as a “vital link in our nation’s commerce”). “Railroads are interstate at

their core, regardless of the fact that some rail lines are entirely intrastate.” Davarci, 2021 WL

3721374, at *12. The same cannot be said for rideshare drivers, who close to 100% of the time

give short, local rides, and whose interstate trips are generally of the same character as their

intrastate trips. Capriole, 7 F.4th at 865 (contrasting Uber drivers with seamen and railroad

workers in this respect, for whom “the interstate movement of goods and passengers over long

distances and across national or state lines is an indelible and ‘central part of the job description’”)

(quoting Wallace, 970 F.3d at 803).

b. Airport Data

What remains is data on Uber’s airport trips. Plaintiffs’ theory here is that, because Uber

drivers occasionally transport riders to and from airports, where interstate travel frequently occurs,

drivers themselves are “within the flow of interstate commerce” under Circuit City or at least they

engage in work “sufficiently related” to it pursuant to Tenney. The text of the residual clause places

a heavy thumb on the scale against Plaintiffs’ argument from the start. Section 1 uses the phrase

“engaged in commerce,” while § 2 uses the phrase “involving” commerce. Section 2’s open-ended

phrasing “signals an intent to exercise Congress’ commerce power to the full.” Allied-Bruce

Terminix Cos. v. Dobson, 513 U.S. 265, 277 (1995). By contrast, § 1 is written to capture a much

narrower set of activities, Circuit City, 532 U.S. at 118, has a “more limited reach,” id. at 115, and

evinces “[n]o [ ] concern for the impact of intrastate conduct on interstate commerce.” United

States v. American Building Maintenance Industries, 422 U.S. 271, 278 (1975); Capriole, 460 F.

Supp. 3d at 929 (“The plain meaning of the words ‘engaged in commerce’ includes not everyone

whose work might generally affect commerce.”) (quoting Rogers, 452 F. Supp. 3d at 915)

(emphasis in original). To the extent that Uber drivers engage in fundamentally local conduct that

is only tangentially or incidentally related to interstate movement, they are not within the flow of

interstate commerce as courts have interpreted that term. Stated differently, as the Ninth Circuit

held, “‘the residual exemption is . . . about what the worker does,’ not just ‘where the goods [or

people] have been,’” which is the main focus of Plaintiffs’ “airport trips” argument. Capriole, 7

F.4th at 865 (quoting Grice, 974 F.3d at 958 (omission and alteration in original) (quoting Wallace,

970 F.3d at 802)).

The Supreme Court also foreclosed an argument like Plaintiffs’ eighty years ago in United

States v. Yellow Cab Co., 332 U.S. 218 (1947), overruled on other grounds by Copperweld Corp.

v. Indep. Tube Corp., 467 U.S. 752 (1984). The reasoning there is persuasive. In interpreting the

language in the Sherman Act, which is broader than the language in the FAA,7 Circuit City, 532

U.S. at 118, the Court held that local taxicab operators in Chicago “merely convey interstate train

7 Because the Sherman Act is broadly construed, whereas the FAA is narrowly construed, conduct

that does not affect interstate commerce under the Sherman Act a fortiori is not conduct “engaged in

interstate commerce” under the residual clause. Cunningham, 2021 WL 5149039, at *5.

passengers between their homes and the railroad station in the normal course of their independent

local service,” which is “not an integral part of interstate transportation” but only “causal and

incidental” to it. Id. at 233. At the same time, the Court held that a transportation service designed

to transfer passengers between rail stations two miles apart is part of the stream of interstate

commerce. Id. at 228-29. The Court found various factors important, including that the none of

the cab companies “serve[d] only railroad passengers, all of them being required to serve ‘every

person’ within the limits of Chicago,” there was “no contractual or other arrangement with the

interstate railroads,” cab “fares [were not] paid or collected as part of the railroad fares,” and

passengers “contracted for” the cab rides “independently of the railroad journey.” Id. at 231-32.

Multiple courts have held that Uber rides are technologically advanced “local taxicab

transport,” and have rejected Plaintiffs’ “airport trips” argument to this extent. See, e.g.,

Aleksanian, 2021 WL 860127, at *8 (explaining that Yellow Cab’s “reasoning is just as applicable”

to rideshare drivers); Rogers, 452 F. Supp. 3d at 916-17 (summarizing Yellow Cab and concluding

that the same analysis applies to “these modern-day taxi drivers”); Hinson, 2021 WL 838411, at

*6 (analogizing Lyft drivers to taxi drivers, who “have been found to have an ‘only casual and

incidental’ relationship to interstate transit”). Included in this consensus are the Ninth and First

circuits, the latter of which explicitly reversed the district court’s finding on this point. Capriole,

12 F.4th at 865; Cunningham, 2021 WL 5149039, at *4-6. I agree as well.

At bottom, Uber drivers do not serve many—let alone only—airport passengers, anyone

can hail an Uber from anywhere to just about everywhere, and passengers neither order nor pay

for Ubers through their airline or as part of their plane tickets. Accord Osvatics, 2021 WL 1601114,

at *13-14 (“[T]his is especially so given the apparently undisputed fact in the instant case that

passengers seeking a ride to or from an airport or railroad station use the Lyft application

unilaterally to hail a driver, and there is no evidence that Lyft has a ‘contractual or other

arrangement’ with airlines or railways for Lyft drivers to transport passengers who have taken trips

with those companies.”). Likewise, “even when transporting passengers to and from transportation

hubs as part of a larger foreign or interstate trip, Uber drivers are unaffiliated, independent

participants,” called separately and on-demand. Capriole, 7 F.4th at 867; Cunningham, 2021 WL

5149039, at *4 (“The Lyft driver contracts with the passenger as part of the driver’s normal local

service to take the passenger to the start (or from the finish) of the passenger’s interstate journey.”).

For at least these reasons, Uber drivers do not “participate in a single, unbroken stream of interstate

commerce” in completing airport rides, Capriole, 7 F.4th at 863-65, and they lack the requisite

“practical, economic continuity” with interstate air travel to satisfy the residual clause, Gulf Oil

Corp. v. Copp. Paving Co., Inc., 419 U.S. 186, 195 (1974), even under Tenney’s (seemingly) more

permissive “sufficiently related” or “so closely related” standard. Uber drivers are “but one,

segmented part” of a rider’s overall journey to, through, and beyond transportation hubs such as

airports. Davarci, 2021 WL 3721374, at *13-14.

Analogies buttress this conclusion. For instance, “[o]ne would not reasonably say that

plaintiffs are engaged in interstate trucking merely because they sometimes give truck drivers rides

to and from their garages.” Cunningham, 2021 WL 5149039, at *5. On the other hand, airport

shuttle drivers do engage in interstate commerce because shuttle bus companies have “practical

continuity of movement” with overall interstate journeys based on agreements with travel

companies and the fact that customers typically buy travel packages that include shuttle service.

Abel v. So. Shuttle Servs., Inc., 631 F.3d 1210, 1216-17 (11th Cir. 2011); Rogers, 452 F. Sup. 3d

at 916. Similarly, “last leg” drivers are an integral part of an unbroken stream of interstate

commerce: the deliveries they make are coordinated and controlled by the shipping company for

which they work from origin to destination, and they are hired specifically to complete that task—

circumstances not present here. Osvatics, 2021 WL 1601114, at *15; Hinson 2021 WL 838411, at

*6 (“Lyft drivers are more like taxi drivers than last-mile delivery drivers of Amazon products.”)

(alterations omitted); Baltimore & Ohio Southwestern Railroad Co. v. Burtch, 263 U.S. 540, 544

(1924) (discussing “the loading or unloading of an interstate shipment”); Philadelphia & Reading

Railroad Co. v. Hancock, 253 U.S. 284, 286 (1920) (discussing first intrastate leg of interstate coal

route); Rittmann, 971 F.3d at 916-18 (holding that last intrastate leg of interstate package delivery

is intrinsic to interstate commerce); Waithaka, 966 F.3d at 26 & n.11 (same).

iii. Conclusion

In sum, I find that nationwide Uber drivers are not exempt from the FAA, consistent with

holdings in virtually every other court to address the issue, including two circuits.8 In doing so, I

reject both pillars of Plaintiffs’ argument: that they are engaged in interstate commerce because

drivers have crossed state lines 140 million times in 10 years and because 10% of trips begin or

end at an airport. This data (though certainly true) is not dispositive when viewed against

uncontroverted evidence that such rides constitute just 2% of all rides, resemble in character the

other 98% of rides, and likely occur due to the happenstance of geography; and that airport trips

are unaffiliated with and independent from the interstate commerce in which passengers partake

8 In fact, in the two cases reaching a different conclusion, each court has specifically noted that

further evidence may compel the majority view. Haider, 2021 WL 1226442, at *4 (describing Lyft’s claim

as resting on the “apparent instinct that [] trips across state lines must be vanishingly rare,” while stating

that “[n]othing . . . shall prejudice Lyft renewing its motion if a more developed factual record shows that

[plaintiff] is not among a class of workers engaged in interstate commerce,” as here); Islam, 2021 WL

871417, at *8 (holding that Lyft’s claim “presupposes that Lyft and Uber rides are necessary short, local

trips . . . and Lyft has not produced evidence to support that presupposition,” but with such evidence, “[t]he

interstate nature of a trip [from Philadelphia to Camden to, for example, meet a friend for lunch] might

indeed be considered an incidental byproduct of geography”). Thus, with a more complete record, I am not

convinced that the courts in Haider and Islam would not reach the same conclusion as the majority of

courts.

once at airports. Uber drivers nationwide are in the “general business of giving people [local] rides,

not the particular business of offering interstate transportation to passengers,” unlike railroad

workers and seamen, whose jobs revolve around interstate travel/movement. Rogers, 452 F. Supp.

3d at 916. The FAA therefore applies and the parties must arbitrate pursuant to the TSA.

D. State Law Issues

Because I compel arbitration under the FAA and TSA, I need not decide the state law

issues. Cunningham, 2021 WL 5149039, at *7 (“Because we find that the FAA applies, we need

not examine the role of the Massachusetts Uniform Arbitration Act.”); Smith Barney, Inc. v.

Critical Health Sys. of N.C., Inc. of Raleigh, N.C., 212 F.3d 858, 860-61 (4th Cir. 2000) (“Once a

dispute is covered by the [FAA], federal law applies to all questions of interpretation, construction,

validity, revocability, and enforceability.”) (alteration in original); In re Salomon Inc. S’holders’

Derivative Litig., 68 F.3d 554, 559 (2d Cir. 1995) (same).

In any case, state law would furnish an alternative basis to arbitrate. Plaintiffs begin by

reading a nonexistent requirement into the TSA: it forecloses application of the arbitration

provision (in the event that the agreement is taken outside the context of the FAA) because there

is no express state law safe harbor. As the Third Circuit explained in Harper, however, “[f]inding

the § 1 exemption applies does not mean all state law about arbitration vanishes,” regardless of

whether an arbitration provision mentions only the FAA. 12 F.4th at 295; see also Palcko, 372

F.3d at 595 (“There is no language in the FAA that explicitly preempts the enforcement of state

arbitration issues.”); Diaz v. Michigan Logistics Inc., 167 F. Supp. 3d 375, 381 (E.D.N.Y. 2016)

(“Plaintiffs argue that given the parties’ explicit choice to apply the FAA, the FAA is the only law

the Court should consider in determining whether to compel arbitration, effectively rendering the

arbitration provision unenforceable . . . . the inapplicability of the FAA does not render the parties’

arbitration provision unenforceable . . . . assuming that the FAA does not apply, state arbitration

law governs.”) (emphasis in original); Islam, 2021 WL 871417, at *14 (collecting cases holding

same).

The question is what state law governs. Plaintiffs propose California. This follows from §

15.1 of the TSA, they claim, which provides that the “interpretation of [the TSA] shall be governed

by California law.” Id. But the TSA is not that straightforward. In the very same sentence, § 15.1

states that “[t[he choice of law provisions contained [herein] do not apply to the arbitration clause,”

which is subject to the FAA instead. Id. Although Uber certainly “could have specified more

clearly what law applies,” if any, when the FAA does not, Waithaka, 966 F.3d at 27 & n.13, I do

not construe the arbitration provision as subject to the choice of law clause. See, e.g., Rimel v. Uber

Techs., Inc., 246 F. Supp. 3d 1317, 1324 (M.D. Fla. 2017) (concluding that “the Service

Agreement’s California choice of law provision has no effect on . . . the Arbitration Provision”

because “the Arbitration Provision is severable from the Service Agreement” that contains the

choice of law provision); Carey v. Uber Techs., Inc., No. 16-1058, 2017 WL 1133936, at *6 (N.D.

Ohio Mar. 27, 2017) (“The Agreement generally provides that California law applies, but

challenges to the validity of an arbitration provision are considered independently from the rest of

the Agreement . . . . Neither the arbitration provision nor the delegation provision contain[s] a

choice-of-law clause. ‘Absent an effective choice of law provision, Ohio courts apply the law of

the state with the most significant relationship to the contract.’”) (citations omitted). The TSA

treats the arbitration provision differently from the rest of the agreement in this regard, and I would

respect that distinction. In fact, it appears that the parties specifically contracted that “California

law should not apply to the arbitration [provision].” Islam, 2021 WL 87147, at *14 (applying New

York law to Lyft drivers under “standard choice of law principles”) (emphasis added). To

nevertheless apply California law would run the risk of “rewrit[ing] the contract under the guise

of [choice of law principles],” and doing so in a way contrary to the parties’ intentions. Rittman,

971 F.3d at 920 (declining to apply Washington law in the context of a similar arbitration provision

for Amazon delivery drivers).

Notwithstanding the TSA’s choice of law clause, a district court sitting in diversity (such

as this Court, assuming Plaintiffs are exempt from the FAA) applies the choice of law rules of the

forum state (here, New Jersey). Collins v. Mary Kay, Inc., 874 F.3d 176, 183 (3d Cir. 2017). Under

New Jersey choice of law rules, “the law of the state which has ‘the most significant relationship’

with the transaction would apply.” Polarome Mfg. Co. v. Commerce & Indus. Ins. Co., 310 N.J.

Super. 168, 172 (App. Div. 1998). Again, that is New Jersey, where Plaintiffs live and work and

where the bulk of the allegations arise. Islam, 2021 WL 871417, at *14. And under New Jersey

law, which does not contain a residual clause exemption, the arbitration provision in the TSA is

lawful and controlling. N.J.S.A. § 2A:23B-6(a) (“An agreement contained in a record to submit to

arbitration any existing or subsequent controversy arising between the parties to the agreement is

valid [and] enforceable.”); Angrisani v. Fin. Tech. Ventures, L.P., 402 N.J. Super. 138, 148 (App.

Div. 2008) (“[T]here is no material difference between the approach to the interpretation of

arbitration agreements mandated by the FAA and the approach our courts have taken as a matter

of State law even when the FAA does not apply.”). Accordingly, while I need not reach the issue,

even if Plaintiffs’ contract with Uber fell outside the FAA, I would apply New Jersey law and

compel arbitration all the same.

IV. CONCLUSION

For the foregoing reasons, I GRANT Uber’s motions and COMPEL arbitration under the

FAA. In accordance with the Third Circuit’s instruction in Singh, all other issues are reserved for

the arbitrator. Plaintiffs’ motion for class certification is DENIED as moot.

DATED: November 23, 2021

/s/ Freda L. Wolfson

Hon. Freda L. Wolfson

U.S. Chief District Judge