Stephanie Turner v. McDonald's USA LLC

In the
United States Court of Appeals
For the Seventh Circuit
____________________
Nos. 22-2333 & 22-2334
LEINANI DESLANDES and STEPHANIE TURNER,
Plaintiffs-Appellants,
v.
MCDONALD’S USA, LLC, and MCDONALD’S CORPORATION,
Defendants-Appellees.
____________________
Appeals from the United States District Court for the
Northern District of Illinois, Eastern Division.
Nos. 17 C 4857 & 19 C 5524 — Jorge L. Alonso, Judge.
____________________
ARGUED MARCH 31, 2023 — DECIDED AUGUST 25, 2023
____________________
Before EASTERBROOK, RIPPLE, and WOOD, Circuit Judges.
EASTERBROOK, Circuit Judge. Until recently, every McDon-
ald’s franchise agreement contained an anti-poach clause.
Each franchise operator promised not to hire any person em-
ployed by a different franchise, or by McDonald’s itself, until
six months after the last date that person had worked for
McDonald’s or another franchise. A related clause barred one
franchise from soliciting another’s employee. We use “anti-
2 Nos. 22-2333 & 22-2334
poach clause” or “no-poach clause” to refer to these collec-
tively.
Plaintiffs in this suit under §1 of the Sherman Act, 15
U.S.C. §1, worked for McDonald’s franchises while these
clauses were in force and were unable to take higher-paying
offers at other franchises. They contend that the no-poach
clause violates the antitrust laws. If this clause holds down the
price of labor by reducing competition for fast-food workers,
that could benefit owners—and conceivably consumers too.
But the antitrust laws prohibit monopsonies, just as they pro-
hibit monopolies. See NCAA v. Alston, 141 S. Ct. 2141 (2021).
Claims under §1 fall into two principal categories: naked
restraints, akin to cartels, are unlawful per se, while other re-
straints are evaluated under the Rule of Reason. (The quick-
look approach, see NCAA v. University of Oklahoma, 468 U.S.
85 (1984), is a subset of analysis under the Rule of Reason.)
The district court rejected plaintiffs’ per se theory after stating
that the anti-poach clause is not a naked restraint but is ancil-
lary to each franchise agreement—and, as every new restau-
rant expands output, the restraint is justified. 2018 U.S. Dist.
LEXIS 105260 (N.D. Ill. June 25, 2018).
The court deemed the complaint deficient under the Rule
of Reason because it does not allege that McDonald’s and its
franchises collectively have power in the market for restau-
rant workers’ labor. Market power is essential to any claim
under the Rule of Reason. See Ohio v. American Express Co., 138
S. Ct. 2274, 2284 (2018); Leegin Creative Leather Products, Inc. v.
PSKS, Inc., 551 U.S. 877, 885–86 (2007); Ball Memorial Hospital,
Inc. v. Mutual Hospital Insurance, Inc., 784 F.2d 1325, 1334–35
(7th Cir. 1986). The absence of such an allegation rendered the
claim implausible, the court held. See Bell Atlantic Corp. v.
Nos. 22-2333 & 22-2334 3
Twombly, 550 U.S. 544 (2007) (establishing the plausibility re-
quirement for antitrust complaints). The judge invited plain-
tiffs to file an amended complaint alleging market power. Af-
ter they declined to do so, the judge dismissed the complaint
with prejudice, ending the suit. 2022 U.S. Dist. LEXIS 113524(N.D. Ill. June 28, 2022).
On appeal plaintiffs assert that they didn’t “really” waive
or forfeit their opportunity to allege market power, but the
district court’s contrary conclusion is not an abuse of discre-
tion. Plaintiffs also contend that the existence of market power
is too obvious to need allegations and proof, but that line of
argument depends on treating “workers at McDonald’s” as
an economic market. That’s not sound. People who work at
McDonald’s one week can work at Wendy’s the next, and the
reverse. People entering the labor market can choose where to
go—and fast-food restaurants are only one of many options.
If wages are too low at one chain, people can choose other em-
ployers. The mobility of workers—both from one employer to
another and from one neighborhood to another—makes it im-
possible to treat employees at a single chain as a market.
The district judge found it undisputed that within three
miles of Deslandes’s home there are between 42 and 50 quick-
service restaurants as well as two McDonald’s franchises, and
that within ten miles of her home there are 517 quick-service
restaurants. This is not a situation in which a court can treat
employment for a single enterprise as a market all its own. See
also, e.g., Elliott v. United Center, 126 F.3d 1003 (7th Cir. 1997)
(peanut sales in or near a sports arena is not a meaningful
market); Menasha Corp. v. News America Marketing In-Store,
Inc., 354 F.3d 661 (7th Cir. 2004) (store coupons, ice cream fla-
vors, and diet soda are not meaningful markets). So the Rule
4 Nos. 22-2333 & 22-2334
of Reason is out of this suit, and, as quick-look analysis is part
of the Rule of Reason, it is out too.
But the district judge jettisoned the per se rule too early.
The complaint alleges a horizontal restraint, and market
power is not essential to antitrust claims involving naked
agreements among competitors. See, e.g., Palmer v. BRG of
Georgia, Inc., 498 U.S. 46 (1990).
An agreement among competitors is not naked if it is an-
cillary to the success of a cooperative venture. See, e.g., Polk
Bros., Inc. v. Forest City Enterprises, Inc., 776 F.2d 185 (7th Cir.
1985); Rothery Storage & Van Co. v. Atlas Van Lines, Inc., 792 F.2d
210, 229 (D.C. Cir. 1986). Consider a partnership to practice
law. The partners devote their time to the law firm and pool
their revenues; that’s a horizontal agreement. The partners
also promise not to compete with the law firm by taking their
own clients. That agreement is lawful because the promise to
devote all legal time to the firm’s business helps each law firm
compete against its rivals; in antitrust jargon, the no-compete
pledge is ancillary to the venture in the sense that it makes the
partnership more effective when competing in the market for
legal services. See Broadcast Music, Inc. v. Columbia Broadcast-
ing System, Inc., 441 U.S. 1, 9 (1979).
The complaint alleges that McDonald’s operates many
restaurants itself or through a subsidiary, and that it enforced
the no-poach clause at those restaurants. This made the ar-
rangement horizontal: workers at franchised outlets could not
move to corporate outlets, or the reverse. See Klor’s, Inc. v.
Broadway-Hale Stores, Inc., 359 U.S. 207, 212–13 (1959); Inter-
state Circuit, Inc. v. United States, 306 U.S. 208 (1939).
Nos. 22-2333 & 22-2334 5
Still, the district court thought that the anti-poach clause is
justified as an ancillary restraint. The court deemed the re-
straint ancillary because it appeared in franchise agree-
ments—and each agreement expands the output of burgers
and fries. (We need not consider the possibility that new fran-
chises replace old ones, so that “new franchise” need not im-
ply “more output,” though this may need attention later.)
One problem with this approach is that it treats benefits to
consumers (increased output) as justifying detriments to
workers (monopsony pricing). That’s not right; it is equiva-
lent to saying that antitrust law is unconcerned with compe-
tition in the markets for inputs, and Alston establishes other-
wise.
Another problem with using the appearance of a clause in
a contract that, on the whole, increases output, is that the
clause may have nothing to do with the output. A “restraint
does not qualify as ‘ancillary’ merely because it accompanies
some other agreement that is itself lawful.” Phillip E. Areeda
& Herbert Hovenkamp, Antitrust Law ¶1908b (4th ed. 2022).
Is there some reason to think that a no-poach clause promotes
the production of restaurant food? See Polk Bros., 776 F.2d at
189. Maybe it just takes advantage of workers’ sunk costs and
helps each business’s bottom line, without adding to output.
What we mean is this: People who choose to work at
McDonald’s or one of its franchises acquire business-specific
(or location-specific) skills. Employees may choose to work
for less than their marginal product in order to compensate
the employer for the training. In a competitive market, work-
ers recover these investments as their wages rise over time, in
response to their greater productivity. But if McDonald’s
specifies a limited number of classifications of workers
6 Nos. 22-2333 & 22-2334
(something the complaint also alleges), that may delay pro-
motion and frustrate workers’ ability to recoup their invest-
ments in training. One way to obtain a higher salary, after
paying for one’s own training through lower wages, is to seek
employment at another similar business where the skills can
be put to use at the market wage. Deslandes alleges that this
is what she tried to do, only to be blocked by the no-poach
clause. And if this is what the no-poach agreement does—if it
prevents workers from reaping the gains from skills they
learned by agreeing to work at lower wages at the outset of
their employment—then it does not promote output. It pro-
motes profits, to be sure, as franchises capitalize on workers’
sunk costs. But it does not promote output and so cannot be
called “ancillary” in the sense antitrust law uses that term.
Common training and job classifications could in principle
justify restraints on poaching. Suppose Franchise A hires
workers and pays for necessary training, rather than requir-
ing the workers to cover their own training costs through
lower wages. During training in this approach, the wage ex-
ceeds the worker’s productivity, but after training the worker
produces enough value to pay back the costs of training and
allow A to recoup the “excess” wage during training time. A
needs to keep the worker for this to pay off. If Franchise B
offers no training but a higher wage, this will be attractive to
the worker who was trained at A, and B can make a profit
from free riding on A’s investment. B can do this because the
restaurants have the same layout, tasks, and so on. In these
circumstances a ban on poaching could allow A to recover its
training costs and thus make training worthwhile to both
franchise and worker. It would not imply monopsony. But
eventually the cost of training will have been amortized, and
Nos. 22-2333 & 22-2334 7
a ban on transfer to another restaurant after that threshold
could be understood as an antitrust problem.
So what was the no-poach clause doing? Was it protecting
franchises’ investments in training, or was it allowing them to
appropriate the value of workers’ own investments? That
question can’t be answered by observing that any given fran-
chise contract, viewed by itself, expands the output of food.
Why did the clause have a national scope, preventing a res-
taurant in North Dakota from hiring a worker in North Caro-
lina, when the market for restaurant jobs is local? Why did the
restriction last as long as the employment (plus six months),
rather than be linked to any estimate of the time a franchise
would need to recover its investments in training? If the an-
swer to some of these questions depends (as McDonald’s as-
serts) on the fact that the system as a whole advertises for
workers and wants to prevent some outlets from free riding
on the contributions of others, how do the terms of the no-
poach clause reflect this objective?
These are all potentially complex questions, which cannot
be answered by looking at the language of the complaint.
They require careful economic analysis. More than that: the
classification of a restraint as ancillary is a defense, and com-
plaints need not anticipate and plead around defenses. Gomez
v. Toledo, 446 U.S. 635, 640–41 (1980); Craftwood II, Inc. v. Gen-
erac Power Systems, Inc., 920 F.3d 479, 482 (7th Cir. 2019); Fed.
R. Civ. P. 8(c).
Some language in the district court’s opinions suggests
that a complaint must contain enough to win, but that is not
so. It suffices, Twombly holds, to make out a plausible claim,
and this complaint does so. Nor need a complaint plead law
or match facts to elements of legal theories. See Johnson v.
8 Nos. 22-2333 & 22-2334
Shelby, 574 U.S. 10 (2014); Swierkiewicz v. Sorema N.A., 534 U.S.
506 (2002). Once a complaint has identified a plausible anti-
trust claim, further development requires discovery, eco-
nomic analysis, and potentially a trial.
Plaintiffs sought class certification, and the district court
said no. The court may think it wise to reconsider in light of
the need for a remand and the analysis in this opinion.
The judgment is vacated and the case is remanded for fur-
ther proceedings.
Nos. 22-2333 & 22-2334 9
RIPPLE, Circuit Judge, concurring. I join the opinion and the
judgment of the court. The issue presented by this case is an
important and timely one. I therefore write separately to
make clear my understanding of what we decide, and do not
decide, today.
Our opinion sends the ancillary restraint defense back to
the district court for further analysis. It makes clear that, in
further proceedings before the district court, the defendants
bear the burden of establishing that the no-poaching clause in
the franchise agreement qualifies as an ancillary restraint. It
further suggests the sort of inquiry that the district court
should undertake in considering this question. Our opinion’s
discussion of these perspectives hopefully will be helpful to
the district court and to the parties. However, I do not under-
stand the court’s opinion to assess in any definitive way the
merits of any of these suggested avenues of further economic
analysis, nor do I understand the court to preclude other ap-
proaches that the parties believe pertinent and that the district
court believes relevant.
Nor do I read the court’s discussion as addressing the rel-
ative usefulness of the various considerations that it dis-
cusses. As I understand the court’s opinion, it leaves the dis-
trict court, with the assistance of the parties, to determine the
relative importance of these considerations and to identify
those issues worthy of its prime attention. For instance, the
district court might determine that the scope and duration of
the restriction in question reduces substantially the need for
extended economic analysis of other “potentially complex
questions.” Op. 7. If the restriction cannot be justified because
of its scope and duration, it is difficult to see how it can be
reasonably necessary to the achievement of the
10 Nos. 22-2333 & 22-2334
procompetitive objectives of the franchise agreement. See, e.g.,
Blackburn v. Sweeney, 53 F.3d 825, 828–29 (7th Cir. 1995) (con-
cluding that the “infinite duration” of the restraint meant it
had no “necessary relation” to the procompetitive arrange-
ment and so was not ancillary); Schering-Plough Corp. v. FTC,
402 F.3d 1056, 1073 (11th Cir. 2005) (“[T]he restraint imposed
must relate to the ultimate objective, and cannot be so broad
that some of the restraint extinguishes competition without
creating efficiency.”). If we are to retain the benefits of apply-
ing a per se analysis to horizontal agreements, we need to en-
sure that our adjudication of possible defenses is a focused
one. Cf. Broadcast Music, Inc. v. Columbia Broadcasting System,
Inc., 441 U.S. 1, 19 n.33 (1979) (“BMI”) (cautioning against al-
lowing the threshold inquiry to “subsume the burdensome
analysis required under the rule of reason,” which would ef-
fectively amount to “apply[ing] the rule of reason from the
start”).
Perhaps most importantly, I do not understand the court
to question the continued vitality of the rule that the ancillary
restraint defense requires that the defendants establish both
that the restriction in question be “subordinate and collat-
eral,” Rothery Storage, 792 F.2d at 224, to a “legitimate business
collaboration” among the defendants, Texaco Inc. v. Dagher,
547 U.S. 1, 7 (2006), and be reasonably necessary to achieve a
procompetitive objective of the franchise agreement. See
Blackburn, 53 F.3d at 828. This rule is well-established, and I
do not understand this opinion to weaken surreptitiously a
principle upon which the bench and bar rely.