DAVID BISHOP and ERIC LISH v. Air Line Pilots Association , International

In the
United States Court of Appeals
For the Seventh Circuit
____________________
No. 17-1438
DAVID BISHOP and ERIC LISH ,
Plaintiffs-Appellants,
v.
A IR LINE PILOTS A SSOCIATION ,
I NTERNATIONAL,
Defendant-Appellee.
____________________
Appeal from the United States District Court for the
Northern District of Illinois, Eastern Division.
No. 1:13-cv-06243 — Gary Feinerman, Judge.
____________________
A RGUED S EPTEMBER 27, 2017 — DECIDED A UGUST 13, 2018
____________________
Before R IPPLE, S YKES, and HAMILTON , Circuit Judges.
R IPPLE, Circuit Judge. United Airlines pilot instructors
David Bishop and Eric Lish brought this action against their
union, the Air Line Pilots Association (“ALPA”). They al-
leged that ALPA had breached its duty of fair representation
in its allocation of a retroactive pay settlement among differ-
ent groups of pilots. ALPA moved for judgment on the
pleadings; it contended that the plaintiffs had not alleged

-- 1 of 30 --

2 No. 17-1438
adequately that ALPA acted arbitrarily, discriminatorily, or
in bad faith. The district court granted the motion and dis-
missed the case. Mr. Bishop and Mr. Lish timely appealed
and seek reversal of the district court’s dismissal of their
claims. We hold that the district court prematurely dis-
missed the plaintiffs’ well-pleaded allegations. We therefore
reverse the judgment of the district court and remand the
case for further proceedings consistent with this opinion.
I
BACKGROUND
A.
In 2003, when United Airlines was in the throes of bank-
ruptcy, United and ALPA negotiated a concessionary collec-
tive bargaining agreement (“CBA”) that resulted in wage
and benefit cuts.1 This 2003 CBA became amendable on Jan-
uary 1, 2010, and, as required under the Railway Labor Act,
45 U.S.C. § 151 et seq., the pilots continued to work under the
2003 CBA until ALPA and United could negotiate a new
one. See id. § 156; see also Detroit & Toledo Shore Line R.R. v.
United Transp. Union, 396 U.S. 142, 148–49 (1969).
It took nearly three years for ALPA and United to settle
on a new CBA (the “2012 United Pilot Agreement” or the
1 Because the district court dismissed the plaintiffs’ claims on ALPA’s
motion under Federal Rule of Civil Procedure 12(c), we recite the plain-
tiffs’ version of the facts. In doing so, “we draw all reasonable inferences
and facts in favor of the [plaintiffs], but need not accept as true any legal
assertions.” Wagner v. Teva Pharm. USA, Inc., 840 F.3d 355, 358 (7th Cir.
2016).

-- 2 of 30 --

No. 17-1438 3
“2012 UPA”). Throughout this extended negotiation process,
United pilots and ALPA assumed that the new CBA would
include retroactive compensation for the years that the pilots
had worked under the amendable CBA. United had merged
with Continental Airlines while the 2003 CBA still was in ef-
fect. The parties therefore covered both legacy United and
legacy Continental pilots in the 2012 UPA.
The vast majority of United’s pilots are “line pilots.”2
Their “work consists exclusively of flying customers from
one location to another—referred to as ‘flying the line.’”3 Mr.
Bishop and Mr. Lish are in a second classification. They are
pilot instructors. The present dispute centers on how ALPA
chose to allocate the negotiated retroactive pay settlement
between line pilots and pilot instructors.4
Line pilot pay is determined by two factors: (1) the pilot’s
“fleet, seat, and longevity” combination, and (2) an hourly
rate multiplied by the number of hours worked. A pilot’s
fleet, seat, and longevity combination is determined by the
type of aircraft a pilot flies (“fleet”); the rank the pilot occu-
pies in that aircraft (“seat”); and the length of time since the
pilot was hired (“longevity”). The CBA then sets an hourly
rate for each possible fleet, seat, and longevity combination.
Line pilots are paid per number of hours actually worked
based on this rate.
2 R.29 at 6.
3 Id.
4 There is also a third group of pilots, management pilots. They were
once parties to this litigation but have settled their claims.

-- 3 of 30 --

4 No. 17-1438
Pilot instructors, by contrast, are salaried pilots. To arrive
at the appropriate salary, the CBA adopts a fleet, seat, and
longevity combination that applies equally to all pilot in-
structors.5 Then, United takes the hourly rate that would be
assigned to a line pilot with that same fleet, seat, and longev-
ity combination and multiplies it by a predetermined num-
ber of hours. This number of hours has no relation to the
number of hours actually worked by each pilot instructor.
Under the 2003 concessionary CBA, all pilot instructors were
capped at a fleet, seat, and longevity combination of “a
767/757 First Officer with six (6) years longevity.”6 They
were given a salary equivalent to what a line pilot with that
combination would earn for eighty-nine flight hours in a
month.
The line pilots gained wage increases in the 2012 UPA
through increased hourly rates and some redefinition of how
the fleet, seat, and longevity combination is calculated. The
pilot instructors’ pay was affected by a change in their pre-
determined fleet, seat, and longevity cap, as well as the
number of credit hours per month they were given. Under
the 2012 UPA, pilot instructors’ fleet, seat, and longevity
combination was increased to “a First Officer with nine (9)
5 The predetermined fleet, seat, and longevity combination for pilot in-
structors is written like a salary cap. If a pilot instructor’s actual fleet,
seat, and longevity combination were lower than the predetermined cap,
the pilot instructor’s pay would be calculated based on his actual combi-
nation. Here, though, during the relevant time period, all pilot instruc-
tors had actual fleet, seat, and longevity combinations that exceeded the
predetermined cap.
6 R.29 at 12.

-- 4 of 30 --

No. 17-1438 5
years longevity at the second highest-rated aircraft rate (i.e.,
A350, 747, 777, 787 rates).”7 They were compensated at an
amount equivalent to ninety hours worked each month for a
pilot with that fleet, seat, and longevity combination. Under
this 2012 UPA, pilot instructors received the largest pay in-
crease.
When it settled on the 2012 UPA with ALPA, United
agreed to give ALPA a $400 million lump-sum settlement to
compensate pilots for the nearly three-year delay in reaching
a new CBA. An intra-union arbitration designated $225 mil-
lion for legacy United pilots and $175 million for legacy Con-
tinental pilots. The $225 million allocated for legacy United
pilots was not enough to compensate the United pilots fully
for the pay they should have been receiving while the 2012
UPA was being negotiated. United left ALPA to allocate the
settlement among its various groups of members.
ALPA generated a formula to calculate each pilot’s share
of retroactive pay relative to the $225 million designated for
legacy United pilots. In crafting this formula, ALPA em-
ployed a Delta Airlines CBA as a comparator to ascertain
what each pilot should have been earning during the three
years of negotiation. ALPA reasoned that because Delta was
a peer competitor, its pay rates during that time period
would be similar to what the United pilots should have been
earning during the same period.
To calculate the line pilots’ retroactive pay, ALPA ap-
plied Delta’s hourly rate to each line pilot’s actual fleet, seat,
7 Id. at 13.

-- 5 of 30 --

6 No. 17-1438
and longevity combination during the negotiation period
and then considered the number of hours the United line pi-
lot actually worked during the negotiation period. ALPA
then took the difference between this fictional wage and the
line pilot’s actual wages for that time period to calculate
what it called a “Delta differential” for each line pilot.8 Final-
ly, ALPA used the Delta differential to calculate each pilot’s
pro rata share of the $225 million settlement payment.
ALPA employed a different approach for pilot instruc-
tors. Instead of basing the United pilot instructors’ retroac-
tive pay on what Delta pilot instructors were paid under the
comparator Delta CBA, ALPA applied its line pilot formula
equally to pilot instructors. It calculated what a Delta line pi-
lot would have earned for the pilot instructors’ predeter-
mined fleet, seat, and longevity combination and predeter-
mined number of hours under the 2003 CBA and based each
pilot instructor’s “Delta differential” on that calculation.9
Because the $225 million was not enough to compensate
each pilot fully for the delay, no pilot received 100 percent of
his retroactive pay, but line pilots received 38 percent of
their retroactive pay under ALPA’s formula. This stands in
contrast to the pilot instructors, who received only 15 per-
cent of theirs. Although ALPA applied the same formula
equally to both groups of pilots, the same formula resulted
8 Appellant’s Br. 4
9 Delta did not use a fleet, seat, and longevity cap in its calculation of pay
for its pilot instructors. Instead, Delta based its pilot instructors’ pay on
their actual fleet, seat, and longevity combinations.

-- 6 of 30 --

No. 17-1438 7
in artificially deflated retroactive pay calculations for pilot
instructors because of the significant differences in how their
pay is calculated. Furthermore, because there was a fixed
amount of money in the lump-sum settlement, any ad-
vantage the line pilots gained in the application of this for-
mula came at the expense of the pilot instructors.
The difference comes from how line pilots were compen-
sated for the three-year delay in negotiations. Under the Del-
ta CBA, the hourly rate for each fleet, seat, and longevity
combination automatically increased each year for which
ALPA used the Delta CBA to calculate retroactive pay.10 The
2003 United CBA similarly provided for automatic increases
in the hourly rate for each fleet, seat, and longevity combina-
tion for each year contemplated by the 2003 CBA,11 but dur-
ing the negotiations period, the pilots were locked in at the
2009 rates, the last year for which the 2003 United CBA set
rates.12 By using the Delta CBA rates as a comparator, ALPA
gave the line pilots the advantage of a built-in increase in the
most significant part of their pay formula—the hourly rate—
for each year that ALPA was in negotiations with United.
However, the most significant variable in the calculation of
pilot instructors’ pay is the predetermined fleet, seat, and
longevity combination and number of monthly hours. Be-
cause ALPA did not take into account the increased fleet,
seat, and longevity cap in calculating the pilot instructors’
10 R.134-2 at 10–12.
11 R.134-3 at 4–8.
12 R.134-1 at 41.

-- 7 of 30 --

8 No. 17-1438
retroactive pay, the pilot instructors were locked in at their
previous combination (set by the 2003 concessionary CBA)
for the entire time that the 2012 UPA was being negotiated.
Put another way, even though ALPA did not use the terms
of the 2012 CBA in calculating any pilot’s retroactive pay, us-
ing the Delta CBA as a comparator had the effect of giving
the line pilots a competitive hourly rate for each year that
ALPA was in negotiations with United. ALPA also consid-
ered each line pilot’s actual fleet, seat, and longevity combi-
nation for the negotiation period. The pilot instructors also
got the benefit of Delta’s hourly rates but not the benefit of
the increase to their fleet, seat, and longevity cap. Under this
formula, pilot instructors were hurt more by the three-year
delay than the line pilots were.
In sum, although pilot instructors also received increased
hourly rates based on the Delta CBA, they did not receive
any increases in the most significant factor for their pay cal-
culation—their fleet, seat, and longevity combination. Ra-
ther, their fleet, seat, and longevity combinations, as well as
their set number of monthly hours, remained locked in at the
2009 amounts set by the 2003 CBA.
B.
Mr. Bishop and Mr. Lish believed that ALPA favored the
line pilots over pilot instructors for political reasons because
the line pilots make up such a large percentage of ALPA’s
membership and, therefore, violated its duty of fair repre-
sentation. They allege that the pilot instructors received the
largest pay increase under the 2012 UPA but the smallest
percentage of their retroactive pay for the negotiations peri-
od.

-- 8 of 30 --

No. 17-1438 9
The pilot instructors first raised their complaint through
the dispute resolution mechanism set forth in Letter of
Agreement 24, where ALPA formalized its allocation meth-
odology. That document provided for the resolution of all
disputes about the retroactive pay allocation through an ex-
pedited arbitration before the ALPA Executive Council. The
pilot instructors invoked their right to arbitration and also
challenged the arbitration procedures themselves. They re-
quested that “(1) the parties jointly select the arbitrator; (2)
the arbitrator be instructed to make an independent decision
without giving deference to either side; (3) the parties be al-
lowed to take some discovery; and (4) the arbitration pro-
ceed on behalf of a class of affected individuals.”13 ALPA re-
jected these proposals and conducted the arbitration accord-
ing to the established procedures.
During arbitration, an ALPA representative testified that
Mark Arellano, the pilot instructors’ representative on the
council that set the allocation methodology, was the indi-
vidual who proposed the retroactive pay formula about
which the pilot instructors now complain. Specifically, the
representative testified that the committee “relied heavily on
[Arellano’s] advice,” that “[Arellano] was an integral part of
the committee,” and that Arellano had recommended the
formula because it “treat[ed] [the pilot instructors] just like
they get treated by everybody else.”14
13 R.84 at 4.
14 R.149-2 at 24–25.

-- 9 of 30 --

10 No. 17-1438
The record contains, however, material that tells a differ-
ent story. According to those documents, Arellano had told
the pilot instructors that he did not agree with the retroactive
pay formula as it applied to them because he knew that it
did not accurately reflect the retroactive pay to which they
were entitled.15 Further, Arellano had expressed this opinion
“on numerous occasions” to the committee charged with set-
ting the formula.16 When the pilot instructors inquired fur-
ther, Arellano told them that the committee wanted “some-
thing as simple as possible” and that the decision to treat pi-
lot instructors with the same formula was “just politics.”17
Arellano also warned the pilot instructors that if they chal-
lenged the allocation, there would be “repercussions” and
that the committee would try to take away other benefits
from the pilot instructors.18
C.
The arbitrator confirmed ALPA’s allocation methodology
in May 2013. The pilot instructors filed the operative com-
plaint on November 5, 2013, in the Northern District of Illi-
nois. They were joined as plaintiffs with another group of
pilots, management pilots, who alleged that they were
harmed by the retroactive pay allocation in different ways
that are not relevant to the merits of this appeal. The pilot
15 Id. at 6.
16 Id.
17 Id. at 3.
18 Id. at 4.

-- 10 of 30 --

No. 17-1438 11
instructors alleged that ALPA breached its duty of fair rep-
resentation under the Railway Labor Act, 45 U.S.C. § 151 et
seq. ALPA first moved for summary judgment as to both
groups of pilots on the basis of the arbitration. The district
court denied ALPA’s motion for summary judgment and
held that the plaintiffs were entitled to discovery on the in-
tegrity of the arbitration.19 ALPA then moved for judgment
on the pleadings against both groups of plaintiffs under
Federal Rule of Civil Procedure 12(c), urging that regardless
of the integrity of the arbitration, the complaint did not al-
lege adequately that it had breached its duty of fair represen-
tation.
The district court found that, as to the pilot instructors,
ALPA had not breached its duty of fair representation be-
cause the limited pool of funds required ALPA to compro-
mise, and it was not irrational or arbitrary for ALPA to ap-
ply the same pay formula to all of the pilots. Further, the dis-
trict court concluded, the pilot instructors had received the
largest pay increase under the 2012 UPA, and it was rational
for ALPA to conclude that they should not also receive “a
larger share of the $225 million pie.”20 Finally, although the
pilot instructors alleged that ALPA had acted in bad faith to
benefit the line pilots and harm the pilot instructors (a mi-
nority segment of ALPA’s membership), the district court
19 Although the pilot instructors have briefed their contentions relating
to the arbitration, these issues were never decided definitively by the
district court and therefore are not ripe for decision on this appeal.
20 R.189 at 10.

-- 11 of 30 --

12 No. 17-1438
concluded that ALPA’s motive was “irrelevant” because the
pay allocation decision was rational and not arbitrary.21
The district court therefore granted ALPA’s Rule 12(c)
motion as it pertained to the pilot instructors, although it al-
lowed the management pilots’ claims to proceed for unrelat-
ed reasons. Because the management pilots remained in the
litigation as plaintiffs after Mr. Bishop and Mr. Lish were
dismissed, the district court did not enter a final judgment
against the pilot instructors until February 9, 2017, after the
management pilots settled with ALPA just before trial.
II
The pilot instructors now challenge the district court’s
dismissal of their claims on ALPA’s Rule 12(c) motion. For
the reasons stated below, we reverse the district court’s dis-
missal of the plaintiffs’ claims.
DISCUSSION
A.
The law governing the procedural context of this case is
well established. We review de novo a district court’s grant
of a Rule 12(c) motion for judgment on the pleadings. Mil-
waukee Police Ass’n v. Flynn, 863 F.3d 636, 640 (7th Cir. 2017).
To survive a motion for judgment on the pleadings, “a com-
plaint must state a claim to relief that is plausible on its
face.” Id. (quoting Wagner v. Teva Pharm. USA, Inc., 840 F.3d
355, 358 (7th Cir. 2016)). “When assessing the facial plausibil-
ity of a claim, ‘we draw all reasonable inferences and facts in
21 Id. at 12.

-- 12 of 30 --

No. 17-1438 13
favor of the non-movant, but need not accept as true any le-
gal assertions.’” Id. (quoting Wagner, 840 F.3d at 358). At the
Rule 12 motion stage, a plaintiff must support an allegation
of bad faith with “subsidiary facts,” not just “[b]are asser-
tions of the state of mind.” Yeftich v. Navistar, Inc., 722 F.3d
911, 916 (7th Cir. 2013). Plaintiffs fail to state a claim of bad
faith when their complaint contains nothing but “conclusory
labels,” such as that union officials had acted “invidiously”
or “intentional[ly], willful[ly], wanton[ly], and mali-
cious[ly].” Id. We have noted that a plaintiff’s complaint
might pass muster if it “offer[s] facts that suggest a motive
for the union’s alleged” bad-faith conduct. Id.
The Railway Labor Act and the Supreme Court decisions
interpreting the collective bargaining obligations of parties
under that statute provide the doctrinal anchor for our sub-
stantive analysis. Under the statute, when a union serves as
the exclusive bargaining agent for a group of employees, it
assumes a duty of fair representation. The Supreme Court
has described this duty as an implied “statutory obligation
to serve the interests of all members without hostility or dis-
crimination toward any, to exercise its discretion with com-
plete good faith and honesty, and to avoid arbitrary con-
duct.” Vaca v. Sipes, 386 U.S. 171, 177 (1967).22 This duty is
grounded in the practical reality that employees give up
22 The duty of fair representation was first “developed … in a series of
cases involving … the Railway Labor Act” and, later, was “extended to
unions certified under the N.L.R.A.” Vaca v. Sipes, 386 U.S. 171, 177
(1967). Therefore, duty-of-fair-representation cases apply interchangea-
bly to claims arising under either statutory scheme. See, e.g., Air Line Pi-
lots Ass’n, Int’l v. O’Neill, 499 U.S. 65, 75–77 (1991).

-- 13 of 30 --

14 No. 17-1438
their rights to bargain individually with their employers
when they designate a union as their exclusive bargaining
representative. Therefore, the statute imposes on the exclu-
sive bargaining agent the duty of fair representation. The
Supreme Court has likened this duty “to the duty owed by
other fiduciaries to their beneficiaries.” Air Line Pilots Ass’n,
Int’l v. O’Neill, 499 U.S. 65, 74 (1991). It has further said that
this duty applies equally to all aspects of union activity, not
merely to a union’s negotiations with employers. Id. at 77.
A union breaches the duty of fair representation if its ac-
tions are (1) arbitrary, (2) discriminatory, or (3) made in bad
faith. Id. at 67. A plaintiff is successful in pleading a breach
of the duty of fair representation if he plausibly pleads a
breach under any of the three prongs. See Rupcich v. United
Food & Commercial Workers Int’l Union, 833 F.3d 847, 854 (7th
Cir. 2016). Therefore, each prong “must be considered sepa-
rately in determining whether or not a breach has been
[pleaded].” Neal v. Newspaper Holdings, Inc., 349 F.3d 363, 369
(7th Cir. 2003).
The appropriate inquiry under each of these prongs is
somewhat different. In determining whether a union acted
arbitrarily, we employ “an objective inquiry,” Rupcich, 833
F.3d at 854 (quoting Neal, 349 F.3d at 369), and must remain
constantly mindful of the Supreme Court’s admonition that
“[a]ny substantive examination of a union’s performance …
must be highly deferential, recognizing the wide latitude
that negotiators need for the effective performance of their
bargaining responsibilities,” O’Neill, 499 U.S. at 78. There-
fore, under the “arbitrary” prong of the fair representation
analysis, “the final product of the bargaining process may
constitute evidence of a breach of duty only if it can be fairly

-- 14 of 30 --

No. 17-1438 15
characterized as so far outside a ‘wide range of reasonable-
ness’ that it is wholly ‘irrational’ or ‘arbitrary.’” Id. (citation
omitted) (quoting Ford Motor Co. v. Huffman, 345 U.S. 330,
338 (1953)).
The other two prongs of the fair representation analysis
necessarily require a different approach. When we ask
whether “a union’s actions are discriminatory or in bad
faith,” there must be “proof that the union acted (or failed to
act) due to an improper motive.” Neal, 349 F.3d at 369. Such
an inquiry calls for a subjective inquiry into the union’s mo-
tives. Id. The case law recognizes that unions often must
make decisions that distinguish among different categories
of employees. In that situation, unions must represent “all
the employees as a unit acting by majority vote,” even
though “each individual employee in the unit is a benefi-
ciary of this collective action.” Garcia v. Zenith Elecs. Corp., 58
F.3d 1171, 1175 (7th Cir. 1995). This tension inherent in this
obligation requires us to recognize when assessing claims of
discrimination or bad faith that “[t]he interests of individual
employees sometimes may be compromised for the sake of
the larger bargaining collective.” Id. at 1176.
Our obligation to recognize the tension caused by the un-
ion’s concurrent obligations to its collective membership and
to the individual members means that a claim of discrimina-
tion or bad faith must rest on more than a showing that a un-
ion’s actions treat different groups of employees differently.
A union member’s claim must be based on more than the
discriminatory impact of the union’s otherwise rational deci-
sion to compromise. See O’Neill, 499 U.S. at 81 (noting that
“some form of allocation was inevitable”). Instead, a claim of
discrimination “requires proof that the union acted (or failed

-- 15 of 30 --

16 No. 17-1438
to act) due to an improper motive.” Neal, 349 F.3d at 369 (em-
phasis added).23 If such an improper motive can be shown, it
can taint a union’s decision that otherwise would not consti-
tute actionable discrimination or be considered arbitrary. We
have said that “[a] union would act in bad faith if, for exam-
ple, it disfavored members who supported a losing candi-
date for union office.” Cunningham v. Air Line Pilots Ass’n,
Int’l, 769 F.3d 539, 542 (7th Cir. 2014). “Deceptive actions” or
“fraud” can be relevant to whether a union acted in bad
faith. Yeftich, 722 F.3d at 916 (quoting Humphrey v. Moore, 375
U.S. 335, 348 (1964)).24
B.
Having set forth the principles governing this appeal, we
now apply those principles to the case before us.
We begin with the allegations of the plaintiffs’ operative
complaint. There, the plaintiffs allege that pilot instructors
“represent[] only a small minority” of ALPA’s membership
and that “ALPA designed an allocation formula … designed
to ensure that the members of … minority groups would re-
23 See also Schwartz v. Bhd. of Maint. of Way Employes, 264 F.3d 1181, 1186
(10th Cir. 2001) (requiring plaintiffs to show not just that they “were
treated differently than other employees” but “why [they] were treated
differently—the crucial question for a duty of fair representation dis-
crimination claim” (emphasis in original)).
24 Another example of bad-faith conduct would be collusion between a
union and an employer, “such as attempts by the union and manage-
ment to conceal an agreement to not pursue [an employee’s] arbitration.”
Rupcich v. United Food & Commercial Workers Int’l Union, 833 F.3d 847, 859
(7th Cir. 2016).

-- 16 of 30 --

No. 17-1438 17
ceive, at most, a fraction of their deserved retro pay.”25 They
allege that the allocation was “irrational and discriminatory”
and imposed for “[n]o good faith reason.”26 Finally, they al-
lege that in applying an allocation formula that benefitted
the line pilots at the expense of the pilot instructors, ALPA
breached its duty of fair representation “by discriminating
against, expressing hostility and acting with animosity to-
wards [the pilot instructors] and arbitrarily choosing to dis-
regard their interests in favor of the interests of the stronger,
more politically favored majority.”27
Later, in opposition to ALPA’s Rule 12(c) motion, the
plaintiffs submitted, and the district court considered, sever-
al exhibits that are consistent with the allegations in the
complaint.28 These exhibits included evidence that ALPA
lied about having the support of the pilot instructors’ union
25 R.29 at 6.
26 Id. at 12, 14.
27 Id. at 21–22.
28 It was proper for the district court to consider this evidence. Although
a party moving for dismissal ordinarily converts a Rule 12 motion into a
Rule 56 motion for summary judgment by attaching extrinsic materials
to its motion, a party opposing a Rule 12 motion “has much more flexibil-
ity,” and we (and the district court) may consider the additional materi-
als as long as they are “consistent with the pleadings.” Geinosky v. City of
Chicago, 675 F.3d 743, 745 n.1 (7th Cir. 2012). In fact, under the more
stringent pleading standards of Ashcroft v. Iqbal, 556 U.S. 662 (2009), and
Bell Atlantic Corp. v. Twombly, 550 U.S. 544 (2007), “a plaintiff who is op-
posing a Rule 12(b)(6) or Rule 12(c) motion and who can provide such
illustration may find it prudent to do so.” Geinosky, 675 F.3d at 745 n.1.

-- 17 of 30 --

18 No. 17-1438
representative when it devised the formula and evidence
that ALPA warned the pilot instructors not to complain
about the formula lest they risk “repercussions” and “suf-
fer[ing] the consequences.”29
Here, we are confident that Mr. Bishop and Mr. Lish
have sufficiently and plausibly pleaded that ALPA acted in
bad faith in its allocation of retroactive pay between the line
pilots and pilot instructors. A union may not make decisions
“solely for the benefit of a stronger, more politically favored
group over a minority group.” Barton Brands, Ltd. v. NLRB,
529 F.2d 793, 798–99 (7th Cir. 1976) (emphasis in original). A
union therefore breaches its duty of fair representation if it
makes decisions “for no apparent reason other than political
expediency.” Id. at 800. The plaintiffs have alleged that pilot
instructors make up a minority of ALPA’s membership and
that ALPA acted with the intent to appease its majority
membership, the line pilots, after a lengthy and contentious
CBA negotiation. Under our case law, this is an improper
motive.30
29 R.149-2 at 4.
30 See, e.g., Cunningham v. Air Line Pilots Ass’n, Int’l, 769 F.3d 539, 542 (7th
Cir. 2014) (“A union would act in bad faith if, for example, it disfavored
members who supported a losing candidate for union office.”); Barton
Brands, Ltd. v. NLRB, 529 F.2d 793, 799–800 (7th Cir. 1976) (noting that a
union may not make “decisions … solely for the benefit of a stronger,
more politically favored group over a minority group” (emphasis in
original)); see also Addington v. US Airline Pilots Ass’n, 791 F.3d 967, 984–
85 (9th Cir. 2015) (“Decisions benefitting a majority of the group may not
be made merely because the ‘losers ha[d] too few votes to affect the out-
come of an intra-union election.’” (alteration in original) (quoting
(continued … )

-- 18 of 30 --

No. 17-1438 19
Further, the plaintiffs have alleged that ALPA ignored
Arellano’s warnings about the effect of the allocation on pi-
lot instructor pay and later lied about Arellano’s having
supported the allocation in an attempt to justify the pilot in-
structors’ retroactive pay settlement. “Deceptive actions can
be evidence of bad faith,” Yeftich, 722 F.3d at 916, and the
plaintiffs have plausibly pleaded that ALPA acted decep-
tively in its post-decision justifications for the allocation.
By plausibly pleading that ALPA acted with the improp-
er motive of appeasing the majority-membership line pilots
at the expense of the minority pilot instructors, the plaintiffs
have met their burden of pleading not only that they “were
treated differently than other employees” but “why [they]
were treated differently—the crucial question for a duty of
fair representation discrimination claim.” Schwartz, 264 F.3d
at 1186 (emphasis in original).
Because the plaintiffs have plausibly pleaded allegations
of bad faith and discrimination, they are not required to ne-
gate ALPA’s claim that it acted rationally and, therefore, not
arbitrarily. First, we reiterate that in order to succeed on
their duty of fair representation claim, the plaintiffs need on-
ly prove a breach under one of the prongs of the tripartite
( … continued)
Rakestraw v. United Airlines, Inc., 981 F.2d 1524, 1530 (7th Cir. 1992));
Ramey v. Dist. 141, Int’l Ass’n of Machinists & Aerospace Workers, 378 F.3d
269, 277 (2d Cir. 2004) (noting that a union may not attempt “to punish a
disfavored group,” “prefer [one group of] workers based solely on [their
loyalty] to the guild,” or “punish a minority group within the union”
(quoting Teamsters Local Union No. 42 v. NLRB, 825 F.2d 608, 612 (1st Cir.
1987) (alterations in original)).

-- 19 of 30 --

20 No. 17-1438
analysis. See Rupcich, 833 F.3d at 854. Second, at this stage in
the litigation, we are required to “take all well-pleaded alle-
gations in the plaintiffs’ pleadings to be true, and we view
the facts and inferences to be drawn from those allegations
in the light most favorable to the plaintiffs.” Alexander v. City
of Chicago, 994 F.2d 333, 336 (7th Cir. 1993). ALPA might well
be able to prove later in the litigation that, as a factual mat-
ter, it did not act with the bad-faith motive that the plaintiffs
have pleaded at this stage.
The district court disregarded the plaintiffs’ allegations of
bad faith and discrimination because it concluded that “the
allocation was rational and not arbitrary, rendering irrele-
vant ALPA’s subjective motivation.”31 In so holding, the dis-
trict court relied on our decisions in Rakestraw v. United Air-
lines, Inc., 981 F.2d 1524 (7th Cir. 1992), and Cunningham, 769
F.3d 539. ALPA similarly relies on Rakestraw and Cunning-
ham to contend that “ALPA’s motive for … equality of
treatment is irrelevant” because ALPA’s allocation decision
was not arbitrary.32 ALPA appears to read Rakestraw and
Cunningham as establishing that, as long as ALPA can put
forth a nonarbitrary, rational basis for its allocation method-
ology, the plaintiffs never can succeed on their claims based
on bad faith and discrimination.
This broad reading of these two cases has several infirmi-
ties. First, in the case of Rakestraw, it ignores the significant
difference in the procedural postures of the cases. Rakestraw
31 R.189 at 12.
32 Appellee’s Br. 18–19.

-- 20 of 30 --

No. 17-1438 21
had a substantially more developed record and did not turn
on the adequacy of the pleadings.33 In Cunningham, the issue
of bad faith apparently was never actually at issue.34 More
fundamentally, in the years since these two cases, our court
has acknowledged consistently the three independent in-
quiries required by the Supreme Court analysis under
O’Neill. E.g., Rupcich, 833 F.3d at 854; Yeftich, 722 F.3d at 916;
Neal, 349 F.3d at 369. The law of this circuit follows O’Neill
and is compatible with the law of the other circuits. See, e.g.,
Addington v. US Airline Pilots Ass’n, 791 F.3d 967, 985 (9th
Cir. 2015); Carter v. United Food & Commercial Workers, Local
No. 789, 963 F.2d 1078, 1082 (8th Cir. 1992). Therefore, even if
Rakestraw and Cunningham deviated from the national con-
sensus, we found our way home long ago.
Conclusion
The district court acted prematurely in crediting ALPA’s
allegation of rationality over the plaintiffs’ well-pleaded al-
legations of bad faith and discrimination. In doing so, the
district court deprived the plaintiffs of the opportunity to
conduct discovery and develop a factual record to support
their claims. Because the plaintiffs plausibly pleaded that
ALPA acted with an improper motive in allocating the
lump-sum retroactive pay settlement in a way that dispar-
33 Rakestraw involved two consolidated cases. One was decided on a full
summary judgment record; the other was decided after a full trial on the
merits. Rakestraw v. United Airlines, Inc., 981 F.2d 1524, 1527, 1529 (7th
Cir. 1992).
34 See Cunningham, 769 F.3d at 542 (“Plaintiffs do not accuse the Union”
of “act[ing] in bad faith … .”).

-- 21 of 30 --

22 No. 17-1438
ately affected the pilot instructors, we reverse the judgment
of the district court and remand for further proceedings con-
sistent with this opinion.
REVERSED and REMANDED

-- 22 of 30 --

No. 17-1438 23
HAMILTON , Circuit Judge, concurring in the judgment. I
agree with much of Judge Ripple’s thoughtful opinion, but re-
gretfully cannot join it in all respects. I agree with my col-
leagues that the dismissal of this case on the pleadings was
premature. In my view, though, later dismissal on summary
judgment or at trial seems likely once we focus on just how
small a target plaintiffs must hit to prove a breach of the duty
of fair representation.
We all agree that plaintiffs have not alleged a viable claim
for “arbitrary” action by the union. See ante at 19. In Air Line
Pilots Ass’n v. O’Neill, 499 U.S. 65, 67 (1991), the Supreme
Court restated the established formula that a union may
breach its duty of fair representation by taking action that is
either (1) arbitrary, (2) discriminatory, or (3) in bad faith. On
the arbitrary prong of the standard, O’Neill emphasized, as
have many other cases, that when a union must decide how
to balance competing interests of different members, the un-
ion may operate within a “wide range of reasonableness.” Id.
In this case, the retroactive payment of $225 million for
United legacy pilots presented a zero-sum negotiation within
the union, at least if the allocation of that sum is viewed in
isolation. But it would be a mistake for us to view this negoti-
ation in isolation (and to let a jury view it in isolation). Per-
haps the most striking feature of this case is that the plaintiffs
themselves received the largest prospective raises under the
new 2012 collective bargaining agreement. But plaintiffs also
contend that they received “one of the biggest pay cuts” un-
der the 2003 concessionary agreement in the United bank-
ruptcy. Dkt. 149-2, at 2 (Casciano Aff.). Their complaint in this
case is that they did not receive enough of the retroactive pay
adjustment. We can be confident that every union officer and

-- 23 of 30 --

24 No. 17-1438
virtually every union member was aware of that larger pic-
ture, and of the even longer history of United pilots’ compen-
sation since long before the Continental merger, the 2012 col-
lective bargaining agreement, and the allocation of the United
pilots’ share of the retroactive payment. Different groups of
pilots can use different baselines to gauge fairness and to
reach diametrically opposite, yet all reasonable, conclusions.
The fairness of one transaction or allocation decision
simply cannot be evaluated in isolation. In this situation, vir-
tually any group will be able to identify past actions that “vic-
timized” them. Federal labor law has long avoided trying to
dictate the outcome of collective bargaining. That reluctance
helps explain the highly deferential approach to union deci-
sions when plaintiffs claim a decision was so arbitrary as to
breach the duty of fair representation. See O’Neill, 499 U.S. at
74, 78 (affirming summary judgment for union); Ford Motor
Co. v. Huffman, 345 U.S. 330, 338 (1953) (same); Yeftich v. Navis-
tar, Inc., 722 F.3d 911, 916 (7th Cir. 2013) (affirming dismissal).
Against virtually any allocation decision a union might make,
someone could make the same arguments plaintiffs make
here. See ante at 16–17 (summarizing plaintiffs’ arguments).
The district court correctly dismissed this theory at the plead-
ing stage.
We are remanding for further proceedings on plaintiffs’
allegations that in allocating the retroactive pay sum among
its members, the union acted for a discriminatory motive or
in bad faith. I do not agree with my colleagues’ statement of
the applicable pleading standard regarding those states of
mind. See ante at 13. Federal Rule of Civil Procedure 9(b), af-
ter requiring that fraud or mistake be alleged “with particu-
larity,” provides further: “Malice, intent, knowledge, and

-- 24 of 30 --

No. 17-1438 25
other conditions of a person’s mind may be alleged gener-
ally.” The majority opinion says that such an allegation of bad
faith must be supported with allegations of subsidiary facts,
and that “bare assertions of the state of mind” are not enough.
Ante at 13, quoting Yeftich, 722 F.3d at 916. In saying that, the
majority opinion contributes to our circuit’s decisions relying
on Ashcroft v. Iqbal, 556 U.S. 662, 686–87 (2009), which have us
drifting further and further away from Rule 9(b)’s express au-
thorization that, apart from fraud or mistake, a defendant’s
state of mind may be alleged “generally.” To my knowledge,
we have not yet seen the efficiencies and cost savings prom-
ised, or at least hoped for, in Iqbal. See id. at 685. The principal
effect of Iqbal seems to be only an extra layer of motions prac-
tice (and attorney fees) in vast numbers of civil cases. While I
do not join the language of the majority opinion in this re-
spect, I still agree that at the pleading stage, these plaintiffs
have alleged sufficiently the subjective states of mind—dis-
criminatory and acting in bad faith.
The more difficult problem in this case, and in the field of
unions’ duty of fair representation generally, is what should
count as bad faith or discrimination? The answer is that not
much should count, and a jury should be instructed accord-
ingly. The principle of majority rule is at the core of American
labor law. We should not be surprised if union leaders opt for
the greatest good for the greatest number, and we should not
step in to protect union minorities from majority rule except
in narrow classes of cases.
The district court looked at the Supreme Court’s decision
in O’Neill and our precedents, including Rakestraw v. United
Airlines, Inc., 981 F.2d 1524 (7th Cir. 1992), and Cunningham v.
Air Line Pilots Ass’n, 769 F.3d 539 (7th Cir. 2014), which all held

-- 25 of 30 --

26 No. 17-1438
that this same defendant union did not breach its duty of fair
representation in similarly painful zero-sum allocations
among union members. Based on those opinions, Judge
Feinerman reached the understandable conclusion that sub-
jective intentions did not matter, at least as long as the union’s
decision on allocation of the retroactive pay was not arbitrary.
We explained in Rakestraw that the duty of fair representa-
tion prevents a union from punishing political opponents of
union leaders or another minority “for no reason other than
that the losers have too few votes to affect the outcome of an
intra-union election,” or based on race, sex, or other legally
forbidden grounds. 981 F.2d at 1530. But in Rakestraw we also
warned of the ease with which difficult allocation decisions
can be described as arbitrary, discriminatory, or in bad faith,
at least if those terms are understood broadly:
The same acts may be characterized as “bad
faith” efforts to take a perquisite away from one
group, or as normal efforts to provide a level
playing field for all workers. Slogans can be
mustered on all sides.
So we are back to the beginning. Bargaining
has winners and losers. A better retirement plan
and other deferred compensation helps older
workers at the expense of younger ones, who
prefer cash to cover current expenses. Accepting
lower fringe benefits (less health insurance,
fewer days off) in exchange for higher cash
wages helps the majority at the expense of the
sick and those whose religious faith requires
them to take extra days off. Higher wages assist
those who remain with the firm at the expense

-- 26 of 30 --

No. 17-1438 27
of workers who may be laid off when the em-
ployer loses business to its competitors. Reduc-
ing the difference between the wages of skilled
and unskilled workers assists the less skilled,
who also tend to be more numerous. Unless its
leaders are unfathomably dense, the union
knows who wins and who loses. The losers al-
ways may say that the union “intended” them to
lose. As all of society is an assembly of minori-
ties, the losers can point to some respect (age, sex,
religion, politics, skill, health, membership in the un-
ion, status during the most recent strike) in which
they are in a minority, and insist that the distinction
must have been part of a “bad faith” plot to “penal-
ize” them on account of that status. Taken to its
limits, the approach prevents the union from re-
solving differences internally and representing
the interests of workers as a group. Yet one of
the premises of the Railway Labor Act (like the
National Labor Relations Act), and a require-
ment of the Labor-Management Reporting and
Disclosure Act, is that unions act democratically
to reach a collective decision—the majority is
entitled to prevail.
981 F.2d at 1530–31 (citations omitted, emphasis added).
Drawing on O’Neill’s comparison to judicial review of stat-
utes, we added in Rakestraw: “Slapping the label ‘bad faith’ or
‘discrimination’ on a classification that is rationally related to
a legitimate objective does not alter the analysis. A discrimi-
natory motive without a discriminatory rule does not con-
demn a statute.” Id. at 1532. And the same applies to union

-- 27 of 30 --

28 No. 17-1438
allocation decisions: “For the reasons developed in Part II of
this opinion, however, a ‘bad’ motive does not spoil a collec-
tive bargaining agreement that rationally serves the interests
of workers as a whole . . . .” Id. at 1535.
Similarly, we pointed out in Cunningham that the kind of
discrimination alleged there (favoring pilots of one merging
airline over those of another) was not the sort of discrimina-
tion courts are worried about. 769 F.3d at 542. Cunningham
noted that “some form of allocation [between competing
groups of pilots] was inevitable. A rational compromise on
the initial allocation was not invidious ‘discrimination’ of the
kind prohibited by the duty of fair representation.” Id. at 542,
quoting O’Neill, 499 U.S. at 81.
My colleagues portray Rakestraw and Cunningham as out
of step with O’Neill and other Seventh Circuit cases, but that
view is not persuasive. Both Rakestraw and Cunningham leave
a little room to consider subjective motives, such as in dis-
crimination cases involving race, sex, or other legally prohib-
ited reasons, and in bad faith cases where plaintiffs can show
the union leaders’ sole motive was to punish their political op-
ponents. 981 F.2d at 1530; 769 F.3d at 542. But expanding the
realm of discriminatory or bad faith actions much beyond
those narrow categories runs into O’Neill’s broad deference to
union leaders in solving intractable problems in distributive
justice.
The majority views Rakestraw and Cunningham as contrary
to O’Neill, but I do not think that is correct, at least if we pay
close attention to what actually happened in the relevant
cases, and not just to what the courts’ opinions said. The ma-
jority cites several cases to show that we have “found our way
home” from Rakestraw and Cunningham, but the cited cases do

-- 28 of 30 --

No. 17-1438 29
not really show that. See ante at 21. In each of the cited cases,
the union won on the claims of bad faith and discrimination.
See Rupicich v. United Food & Commercial Workers Int’l Union,
833 F.3d, 847, 854–59 (7th Cir. 2016) (union arbitrarily refused
to pursue arbitration of plaintiff’s grievance, but union did
not act in bad faith); Yeftich, 722 F.3d at 916 (affirming dismis-
sal on pleadings); Neal v. Newspaper Holdings, Inc., 349 F.3d
363, 369 (7th Cir. 2003) (holding that union did not act in bad
faith or with discriminatory motive). Opinions explaining
why unions won should not be read as expanding unions’ po-
tential liability.
My colleagues write that it would be an improper motive
for the union to have acted “with the intent to appease its ma-
jority membership, the line pilots, after a lengthy and conten-
tious CBA negotiation.” Ante at 18. Taken by itself, that one
sentence does not capture the applicable law. The cases cited
to support it show that plaintiffs in fact must hit a much
smaller target here. In Barton Brands v. N.L.R.B., we wrote that
a union breaches its duty of fair representation if it makes de-
cisions “for no apparent reason other than political expedi-
ency,” and that a union may not make decisions “solely for the
benefit of a stronger, more politically favored group over a
minority group.” 529 F.2d 793, 799–800 (7th Cir. 1976) (first
emphasis added; second in original).
The cited decisions from other circuits similarly required
plaintiffs to show that a union acted solely for an improper
purpose. Addington v. US Airline Pilots Ass’n, 791 F.3d 967,
984–85 (9th Cir. 2015) (“Decisions benefitting a majority of the
group may not be made merely because the ‘losers ha[d] too
few votes to affect the outcome of an intra-union election.’”),
quoting Rakestraw, 981 F.2d at 1530; Ramey v. District 141, Int’l

-- 29 of 30 --

30 No. 17-1438
Ass’n of Machinists & Aerospace Workers, 378 F.3d 269, 277 (2d
Cir. 2004) (union may not attempt “to punish a disfavored
group,” “prefer [one group of] workers based solely on [their
loyalty to the guild],” or “punish a minority group within the
union”) (emphasis added). I trust the district judge will apply
this high standard at summary judgment or trial.
In the end, this case will need to be decided on the basis of
facts rather than pleadings. Though plaintiffs have barely
cleared the pleading standard here, they will have a difficult
time proving what they need to prove on the merits. The evi-
dence of bad faith proffered so far—different accounts of the
role of the instructor pilots’ representative in the allocation
negotiations—looks a lot less like bad faith and more like dif-
ferences in “spin” about a difficult decision bound to disap-
point everyone in varying degrees.

-- 30 of 30 --