Arthur Diamond v. Pennsylvania State Education A

PRECEDENTIAL

UNITED STATES COURT OF APPEALS

FOR THE THIRD CIRCUIT

___________

Nos. 19-2812 and 19-3906

ARTHUR DIAMOND, on behalf of himself and others

similarly situated;

JEFFREY SCHAWARTZ; SANDRA H. ZIEGLER, on

behalf of themselves

others similar situated; MATTHEW SHIVELY; MATTHEW

SIMKINS;

DOUGLAS R. KASE; JUSTIN BARRY,

Appellants in case no. 19-2812

v.

PENNSYLVANIA STATE EDUCATION ASSOCIATION;

CHESTNUT RIDGE EDUCATION ASSOCIATION,

as representative of the class of all chapters and

affiliates of the Pennsylvania State Education Association;

NATIONAL EDUCATION ASSOCIATION; JOSH

SHAPIRO,

in his official capacity as Attorney General of Pennsylvania;

JAMES M. DARBY; ALBERT MEZZAROBA; ROBERT H.

SHOOP, JR.,

in their official capacities as chairman and members of the

Pennsylvania

Labor Relations Board; LESLEY CHILDER-POTTS, in her

official capacity

as district attorney of Bedford County, and as representative

of the class

of all district attorneys in Pennsylvania with the authority to

prosecute violations

of 71 Pa. Stat. 575

JANINE WENZIG and CATHERINE KIOUSSIS,

Appellants in case no. 19-3906

v.

SERVICE EMPLOYEES INTERNATIONAL UNION

LOCAL 668

On Appeal from the United States District Court

for the Western District of Pennsylvania and the Middle

District of Pennsylvania

(District Court Nos.: 3-18-cv-00128 and 1-19-cv-01367)

District Judges: Honorable Kim Gibson and Honorable

Malachy E. Mannion

Argued April 24, 2020

(Opinion Filed: August 28, 2020)

Before: PHIPPS, RENDELL, and FISHER, Circuit Judges

2

Joseph F. Canamucio, Esq.

Pennsylvania State Education Association

400 North Third Street

Harrisburg, PA 17101

Leon Dayan, Esq. [ARGUED]

Bredhoff & Kaiser

805 15th Street, N.W.

Suite 1000

Washington, DC 20005

Jacob Karabell, Esq.

Bredhoff & Kaiser

805 15th Street, N.W.

Suite 1000

Washington, DC 20005

John M. West, Esq.

Bredhoff & Kaiser

805 15th Street, N.W.

Suite 1000

Washington, DC 20005

Counsel for Appellees Pennsylvania State

Education Association, et al.

Daniel B. Mullen, Esq.

Office of Attorney General of Pennsylvania

1251 Waterfront Place

Mezzanine Level

Pittsburgh, PA 15222

3

Counsel for Appellees Attorney General Joshua

D. Shapiro, et al.

Jonathan F. Mitchell, Esq. [ARGUED]

Direct: 512-686-3940

Email: jonathan@mitchell.law

Fax: 512-686-3941

[COR NTC Retained]

Suite 400

111 Congress Avenue

Austin, TX 78701

Counsel for Appellants Arthur Diamond, et al.

Meredith Johnson, Esq.

Altshuler Berzon

177 Post Street

Suite 300

San Francisco, CA 94108

Scott A. Kronland, Esq.

Altshuler Berzon

177 Post Street

Suite 300

San Francisco, CA 94108

P. Casey Pitts, Esq. [ARGUED]

Altshuler Berzon

177 Post Street

Suite 300

San Francisco, CA 94108

4

Counsel for Appellee Service Employees

International Union Local 668

Charles O. Beckley, II, Esq.

Beckley & Madden

212 North Third Street

Suite 301

Harrisburg, PA 17108

Brian Kelsey, Esq. [ARGUED]

Liberty Justice Center

190 South LaSalle Street

Suite 1500

Chicago, IL 60603

William L. Messenger, Esq.

National Right to Work Legal Defense Foundation

8001 Braddock Road

Suite 600

Springfield, VA 22151

Counsel for Appellants Janine Wenzig and

Catherine Kioussis

O P I N I O N

RENDELL, Circuit Judge:

In reliance on a Pennsylvania statute and the Supreme

Court’s decision in Abood v. Detroit Bd. of Educ., 431 U.S. 209

(1977), Appellee Unions, the Service Employees International

5

Union Local 668 and the Pennsylvania State Education

Association, collected “fair-share fees” from Appellants over

Appellants’ objections. But the Supreme Court overruled

Abood in Janus v. AFSCME Council 31, holding that state

legislation condoning public-sector fair-share fees was

unconstitutional. 138 S. Ct. 2448 (2018) (“Janus I”). Now,

Appellants bring these § 1983 lawsuits seeking reimbursement

of the sums they were required to pay. The District Courts,

joining a consensus of federal courts across the country,

dismissed Appellants’ claims for monetary relief, ruling that

because the Unions collected the fair-share fees in good faith

reliance on a governing state statute and Supreme Court

precedent, they are entitled to, and have successfully made out,

a good faith defense to monetary liability under § 1983. We

will affirm.

I

A. Legal background

Labor laws in the United States have long authorized

employers and labor organizations to bargain for an “agency

shop,” an arrangement in which one union is allowed to

exclusively represent an entity’s employees on the condition

that the union represent all the entity’s employees—even those

who do not join the union. See, e.g., Janus I, 138 S. Ct. at 2460;

45 U.S.C. § 152 (Railway Labor Act); 29 U.S.C. § 159

(National Labor Relations Act). Agency shop arrangements

are intended to promote uniform bargaining, streamlined

administration, and other interests, but they also create an

incentive for employees to decline to join their union (and

therefore avoid paying dues) while still accruing the benefits

of union representation. See, e.g., Janus I, 138 S. Ct. at 2465-

69 (describing the intended purpose of agency shops to create

6

“labor peace” and describing the hypothetical potential for

“free rider” problems in agency shop arrangements). To

address this incentive, Congress often allowed unions and

employers who opt for an agency shop arrangement to require

all employees either to join the union and pay dues or, if an

employee does not join the union, to nonetheless contribute to

the costs of representation, bargaining, and administration of

bargaining agreements. This requirement that non-members

pay some form of union dues is often referred to as a “fair-

share” fee, and is present in various pieces of federal

legislation, including, for instance, the Railway Labor Act, 45

U.S.C. § 152, and the National Labor Relations Act, 29 U.S.C.

§§ 157, 158(a)(3).

The Supreme Court has upheld the constitutionality of

these agency shop arrangements, including fair-share fees. For

instance, in Railway Employees’ Dep’t v. Hanson, the Supreme

Court ruled that the Railway Labor Act’s provisions allowing

agency shop arrangements and fair-share fees did not violate

the First Amendment. 351 U.S. 225, 236-38 (1956). Although

the employees in that case argued that the agency shop

“agreement forces men into ide[o]logical and political

associations which violate their right to freedom of conscience,

freedom of association, and freedom of thought protected by

the Bill of Rights,” id. at 236, the Court “h[e]ld that the

requirement for financial support of the collective-bargaining

agency by all who receive the benefits of its work . . . does not

violate” the First Amendment, id. at 238. The Supreme Court

later reaffirmed this ruling. See Int’l Ass’n of Machinists v.

Street, 367 U.S. 740, 749 (1961) (affirming the

constitutionality of the Railway Labor Act’s agency shop and

fair-share provisions).

7

Eventually, state legislatures across the country passed

laws authorizing public-sector unions to collect fair-share fees

and bargain for agency shop arrangements with state

government employers. In Abood, the Supreme Court affirmed

the constitutionality of one such law, a Michigan statute

permitting state employers to negotiate for agency shop

arrangements and fair-share fees with the public-sector unions

that represented their employees. 431 U.S. at 224-26. The

Abood Court ruled that the important government interests in

creating functional and peaceful labor relations and preventing

the free rider problem “support the impingement upon

associational freedom created by the agency shop.” Id. at 225.

Although the Court recognized that the “government may not

require an individual to relinquish rights guaranteed [] by the

First Amendment as a condition of public employment,” id.

at 234, the Court held that there was no reason to distinguish

Abood from cases like Hanson that had upheld agency shop

arrangements in the private sector, id. at 232 (holding that the

“differences between public- and private-sector collective

bargaining simply do not translate into differences in First

Amendment rights”).

But the Abood Court also ruled that—as in the private

sector—non-members’ fair-share fees could only be used to

pay for union activities that were “germane to [the union’s]

duties as collective-bargaining representative,” but not the

union’s political or other work. Id. at 235. In the Abood

Court’s view, this limitation struck an appropriate balance

between the non-members’ speech rights under the First

Amendment and the government’s interests in regulating labor

relations. Id. at 237 (describing the Court’s ruling as

“preventing compulsory subsidization of ideological activity

by employees who object . . . without restricting the [u]nion’s

8

ability to require every employee to contribute to the cost of

collective-bargaining activities”). Over the course of the

following four decades, the Supreme Court affirmed its

holding in Abood against similar challenges to the

constitutionality of state laws allowing for agency shop

arrangements between public-sector employers and public-

sector unions. See, e.g., Lehnert v. Ferris Faculty Ass’n, 500

U.S. 507 (1991); Locke v. Karass, 555 U.S. 207 (2009);

Friedrichs v. Cal. Tchrs. Ass’n, 136 S. Ct. 1083 (2016) (per

curiam) (equally divided Court affirming without opinion).

In light of Abood, Pennsylvania enacted a law allowing

public-sector agency shop arrangements and authorizing

unions that serve as exclusive representatives to collect fair-

share fees. See 71 Pa. Stat. and Cons. Stat. Ann. § 575 (West

2020). Under section 575(b), “[i]f the provisions of a

collective bargaining agreement so provide, each nonmember

of a collective bargaining unit shall be required to pay to the

exclusive representative a [fair-share] fee.” Fair-share fees

could consist of normal dues minus “the cost for the previous

fiscal year of [the union’s] activities or undertakings which

were not reasonably employed to implement or effectuate the

duties of the employe[e] organization as exclusive

representative.” Id. § 575(a). The law also set forth the

procedure by which fair-share fees would be deducted from

non-member employees’ paychecks, see id. § 575(c), and a

procedure through which non-member employees could obtain

information about how their fees were used, see § 575(d). If

this information reflected any improper uses, non-members

could challenge the fair-share fees. See id. § 575(e).

In 2018, the Supreme Court “overruled” Abood. Janus

I, 138 S. Ct. at 2460. Holding that Abood “was poorly

reasoned” and led to “practical problems and abuse,” the Court

9

ruled that Abood was “inconsistent with other First

Amendment cases” and was not entitled to continued

precedential status. Id. The Janus I Court held that Abood had

mischaracterized the government’s interests in promoting

“labor peace” and preventing “free-riders.” Id. at 2465-70.

Whereas the Abood Court had decided that those interests

justified the fair-share fee laws’ impingement on the union

non-members’ speech rights, the Court in Janus I stated that,

instead, “‘labor peace’ can readily be achieved through means

significantly less restrictive of associational freedoms,” and

that “avoiding free riders is not a compelling interest.” Id. at

2466 (internal quotation marks and citations omitted).

Accordingly, “the First Amendment does not permit the

government to compel a person to pay for another party’s

speech just because the government thinks that the speech

furthers the interests of the person who does not want to pay.”

Id. at 2467. State legislation allowing public-sector employers

and public-sector unions to collect fair-share fees

unconstitutionally forced non-members “to subsidize a union,

even if they choose not to join and strongly object to positions

the union takes in collective bargaining and related activities,”

and thereby compelled non-members “to subsidize private

speech on matters of substantial public concern.” Id. at 2459-

60. On this basis, the Court ruled that “[s]tates and public-

sector unions may no longer extract agency fees from

nonconsenting employees.” Id. at 2486. Therefore, under

Janus I, Pennsylvania’s public sector agency shop law was no

longer constitutional.1

1

We assume without deciding that the right announced by the

Supreme Court in Janus I is retroactive. Janus v. AFSCME,

Council 31, 942 F.3d 352, 360 (7th Cir. 2019) (“Janus II”)

10

B. Factual background

1. Diamond facts

Plaintiff Arthur Diamond and his six co-plaintiffs (the

“Diamond Plaintiffs”) are current or former teachers in

Pennsylvania public schools. They were not members of the

Pennsylvania State Education Association (“PSEA”), the

union that exclusively represented their bargaining unit. But

PSEA’s collective bargaining agreement contained a fair-share

clause that required they pay fair-share fees to either the union

or to a union-approved nonreligious charity. See Diamond

Appellants’ Br. at 5 (citing D.A. 73-74). Only Diamond paid

his fair-share fee to PSEA. Id. at 6. The other six Plaintiffs

directed their fees to be diverted to nonreligious charities,

though Sandra H. Ziegler did not identify a charity. Id. at 5-6.

The fair-share fees were no longer collected after June 27,

(“Rather than wrestle the retroactivity question to the ground,

we think it prudent to assume for the sake of argument that the

right recognized” by the Supreme Court in Janus I is

retroactive.); Danielson v. Inslee, 945 F.3d 1096, 1099 (9th

Cir. 2019) (“[W]e will assume that the right delineated in

[Janus I] applies retroactively and proceed to a review of

available remedies.”); Lee v. Oh. Educ. Ass’n, 951 F.3d 386,

389 (6th Cir. 2020) (“[T]he most prudent course of action is to

assume without deciding that the right recognized in [Janus I]

has retroactive application.”). Even if Janus I is retroactive,

the good faith defense may constitute a “previously existing,

independent legal basis” for denying the Appellants’ claims.

See Reynoldsville Casket Co. v. Hyde, 514 U.S. 749, 759

(1995).

11

2018, the date that the Supreme Court issued its decision in

Janus I. A. 74, 93-96.

The Diamond Plaintiffs originally sued PSEA on the

same theory as the plaintiffs in Janus I, but once the Supreme

Court ruled in that case, the Diamond Plaintiffs amended their

Complaint to seek repayment of the fair-share fees they had

previously paid to their union. See Diamond Appellants’ Br.

at 6. PSEA moved to dismiss the amended complaint, arguing

that because it had collected the fees in good faith reliance on

a Pennsylvania statute and pre-Janus I Supreme Court

precedent authorizing fair-share fees, they could not be held

liable for monetary damages. Id. at 7. The District Court

granted the motion to dismiss, ruling that because PSEA had

relied on a prevailing state statute and federal caselaw, they

were entitled to a good faith defense to § 1983 liability that

barred the Diamond Plaintiffs’ claims. D.A. 50-51. The

Diamond Plaintiffs timely appealed. D.A. 1.

2. Wenzig facts

Janine Wenzig and Catherine Kioussis (the “Wenzig

Plaintiffs”) work for the Commonwealth of Pennsylvania.

W.A. 8. Like the Diamond Plaintiffs, they were forced to pay

fair-share fees to their union, the Service Employees

International Union Local 668, without their consent. Id.

Their bargaining unit’s CBA contained the following

provision:

The Employer further agrees to deduct a [fair-

share] fee from all compensation paid to all

employees in the bargaining unit who are not

members of the Union. Authorization from non-

members to deduct [fair-share] fees shall not be

12

required. The amounts to be deducted shall be

certified to the Employer by the Union and the

aggregate deductions of all employees shall be

remitted together with an itemized statement to

the Union by the last day of the succeeding

month after such deductions are made.

Wenzig App. 42.

More than a year after Janus I was issued, the Wenzig

Plaintiffs filed suit on behalf of themselves and a putative class

of similarly situated employees to recover damages under

§ 1983 for the fair-share fees that they had paid to their union.

See Wenzig Appellants’ Br. at 3. They sought a declaratory

judgment that the union’s pre-Janus I collection of fair-share

fees violated the First Amendment and repayment of all fair-

share fees that were collected. W.S.A. 9.

The SEIU filed a motion to dismiss their claims, which

the District Court granted. The District Court ruled the good

faith defense shielded the union from monetary liability for

collecting fair-share fees in good faith reliance on then-

prevailing Supreme Court precedent. W.A. 16. The Wenzig

Plaintiffs timely appealed, and their case was consolidated for

argument and opinion with the Diamond Plaintiffs’ case.

W.A.1.

II

The District Courts had jurisdiction pursuant to 28

U.S.C. §§ 1331 and 1343. We have jurisdiction under 28

U.S.C. § 1291. We review the District Courts’ judgments

granting the Defendants’ motions to dismiss de novo. See, e.g.,

13

Foglia v. Renal Ventures Mgmt., LLC, 754 F.3d 153, 154 n.1

(3d Cir. 2014).

III

We are not the first court of appeals to rule on this

question, and we join a growing consensus of our sister circuits

who, in virtually identical cases, have held that because the

unions collected the fair-share fees in good faith reliance on a

governing state statute and Supreme Court precedent, they are

entitled to a good faith defense that bars Appellants’ claims for

monetary liability under § 1983. See Janus v. AFSCME,

Council 31, 942 F.3d 352 (7th Cir. 2019) (“Janus II”); Mooney

v. Ill. Educ. Ass’n, 942 F.3d 368 (7th Cir. 2019); Danielson v.

Inslee, 945 F.3d 1096 (9th Cir. 2019); Lee v. Oh. Educ. Ass’n,

951 F.3d 386 (6th Cir. 2020); Ogle v. Ohio Civil Serv. Emps.

Ass’n, AFSCME Local 11, 951 F.3d 794 (6th Cir. 2020);

Wholean v. CSEA SEIU Local 2001, 955 F.3d 332 (2d Cir.

2020).

A. Private parties may assert a good faith defense to

§ 1983 liability.

42 U.S.C. § 1983 creates a cause of action for plaintiffs

who are injured by a person who, acting “under color of any

statute . . . of any State,” causes the plaintiff to suffer “the

deprivation of any rights, privileges, or immunities secured by

the Constitution.” Appellants assert that the Unions—acting

under color of a Pennsylvania statute—caused them to be

deprived of their First Amendment rights when the Unions

collected fair-share fees from Appellants’ paychecks.

In Lugar v. Edmondson Oil Co., the Supreme Court held

that § 1983 allows suits against private parties acting under

14

color of state law. 457 U.S. 922, 941 (1982). Under Lugar, a

private party may be liable under § 1983 when the private-

party defendant deprived the plaintiff of a constitutional right

by exercising “a right or privilege having its source in state

authority” and where the private-party defendant may be

“appropriately characterized as [a] ‘state actor[].’” Id. at 939.2

But while the Lugar Court confirmed that private-party

defendants may be subject to suit under § 1983, the Court also

recognized a “concern” that its ruling could unfairly subject

these private entities to liability even though the private parties

had “innocently [made] use of seemingly valid state laws.” Id.

at 942 n.23.

Despite voicing this “concern,” the Court in Lugar left

open the question of whether private parties may avail

themselves of immunity to suit. Id. In Wyatt v. Cole, the

Supreme Court answered this question, ruling that immunity is

reserved for governmental entities, not private parties subject

to suit under § 1983. 504 U.S. 158, 168 (1992). The Court

nonetheless noted—without explicitly ruling—that “principles

of equality and fairness may suggest . . . that private citizens

who rely unsuspectingly on state laws they did not create and

may have no reason to believe are invalid should have some

protection from liability.” Id. But the Court left the question

of whether private-party defendants are entitled to a “defense

based on good faith” for “another day.” Id. at 169. Later, the

2

Under Lugar, a private party may be appropriately

characterized as a state actor where the private party “is a state

official, . . . has acted together with or has obtained significant

aid from state officials, or [where its] conduct is otherwise

chargeable to the State.” Lugar, 457 U.S. at 937. Appellants

do not challenge the Unions’ statuses as state actors.

15

Supreme Court again alluded to, without adopting, this good

faith defense. See Richardson v. McKnight, 521 U.S. 399, 414

(1997) (“Like the Court in Wyatt, . . . we do not express a view

on [the good faith defense].”).

We addressed this open question shortly after Wyatt was

issued. In Jordan v. Fox, Rothschild, O’Brien & Frankel, we

held that a “good faith defense is available” to private parties

who act under color of state law and are sued for monetary

liability under § 1983. 20 F.3d 1250, 1277 (3d Cir. 1994). We

stated our “basic agreement” that “private defendants should

not be held liable under § 1983 absent a showing of malice and

evidence that they either knew or should have known of the

statute’s constitutional infirmity.” Id. at 1276 (citations

omitted). We noted that good faith gives private defendants “a

defense that depends on their subjective state of mind, rather

than the more demanding objective standard of reasonable

belief that governs qualified immunity.” Id. at 1277.3

3

In his concurrence, JUDGE FISHER suggests that a historical

approach to the issue of good faith requires a complex analysis

based on common law. He asserts that the various opinions in

Wyatt imply “that any limitation on private-party liability must

be grounded in the common-law approach.” Fisher Op. at I.C.

JUDGE PHIPPS similarly urges that the good faith defense

should be available if and only if a “deeply rooted common-

law tradition exists” to support it. See Phipps Op.

I can find no such implication, let alone any directive to that

effect. Indeed, the point—the very narrow ruling—of the

majority in Wyatt is that qualified immunity is uniquely a

creature of common law to which private parties are not

entitled. And the Wyatt concurrence’s statement (which Judge

Fisher quotes as the basis for this implication), that “[w]e may

16

not transform what existed at common law based on our

notions of policy or efficiency,” 504 U.S. at 171-72, did no

more than provide support for the majority’s reasoning

rejecting an expansion of the concept of qualified immunity,

and speaks not at all to the issue of the good faith defense or its

contours.

JUDGE FISHER also suggests that my reading of Jordan is

“expansive[],” Fisher Op. at II.B., and JUDGE PHIPPS “does not

see a valid basis for recognizing such a defense,” Phipps Op.,

but urges that, instead, our adoption of the good faith defense

in Jordan was a “misnomer,” id.

I disagree. In Jordan, we embraced the good faith defense

and opined on the contours of its relatively modest

requirements. 20 F.3d at 1275-77. We concluded that good

faith gives private actors a defense that depends on their

“subjective state of mind,” id. at 1277, and looked to whether

the private party acted with “malice” or “either knew or should

have known of the statute’s constitutional infirmity,” id. at

1276. And I note that, importantly, in Jordan, we made no

mention of the common-law approach. Jordan is controlling

precedent as to the legal standard that we apply in this case.

And let us be clear: we are not talking about an across-the-

board good faith defense to a § 1983 action that is inconsistent

with the common law. Instead, we are talking about

prohibiting monetary liability when a private-party defendant

acted in good faith reliance on a statute enacted in accordance

with binding Supreme Court precedent in a situation that has

no exact analogue at common law. Doesn’t the analogy to

abuse of process in note 4 below—or, in its own way, JUDGE

FISHER’s intensive historical analysis—make that very point?

See also, e.g., Janus II, 942 F.3d at 365 (noting that no common

law tort “is a perfect fit”).

17

B. Appellants’ § 1983 claims are barred by the Unions’

good faith defense.

Jordan therefore established that the good faith defense

is available to a private-party defendant in a § 1983 case if,

after considering the defendant’s “subjective state of mind,” id.

at 1277, the court finds no “malice” and no “evidence that [the

defendant] either knew or should have known of the statute’s

constitutional infirmity,” id. at 1276.

There was no such finding of malice or knowledge in Jordan,

and, similarly here, Appellants have not asserted that either of

these disqualifying factors is implicated. Indeed, as noted

above, the Unions’ collection of fair-share fees was authorized

by over four decades of Supreme Court precedent and a

Pennsylvania statute, 71 Pa. Stat. and Cons. Stat. Ann. § 575

(West 2020), that explicitly authorized fair-share fees for

public-sector unions like the Unions. Accordingly, in this case,

Appellants cannot possibly make any “showing of malice” or

demonstrate that the Unions “either knew or should have

known of [§ 575]’s constitutional infirmity.” Jordan, 20 F.3d

at 1276 (citation omitted). The Unions are therefore entitled to

the good faith defense under Jordan.

This is not the huge jurisprudential leap that my colleagues

urge. This is a reasonable way to afford private parties some

of the protection that government actors are afforded when

they act in a situation in which the existing state and federal

law explicitly condoned their behavior. Do we need to chart a

complex path to ensure that this underlying principle is

recognized? We did not in Jordan, and we do not need to do

so here.

18

Moreover, “principles of equality and fairness,” Wyatt,

504 U.S. at 168, independently weigh in favor of the Unions

being protected from suit. It is fair—and crucial to the

principle of rule of law more generally—that private parties

like the Unions should be able to rely on statutory and judicial

authorization of their actions without hesitation or fear of

future monetary liability. Janus II, 942 F.3d at 366 (“The Rule

of Law requires that parties abide by, and be able to rely on,

what the law is . . . .”); Danielson, 945 F.3d at 1105 (finding

that the defendant unions did “exactly what we expect of

private parties: adhering to the governing law of its state and

deferring to the Supreme Court’s interpretations of the

Constitution”); Wholean, 955 F.3d at 336 (noting that unions

“cannot reasonably be deemed to have forecasted whether,

when, and how Abood might be overruled” and holding that

they “were entitled to rely on directly controlling Supreme

Court precedent”).

Appellants present numerous arguments that the good

faith defense should not bar their claims against the Unions.

First, Appellants urge us to rule that the good faith defense only

applies to § 1983 suits that allege theories of liability for which

the most analogous common law tort requires malice or

probable cause. We decline to do so for several reasons. First,

Wyatt applied this most analogous tort concept in considering

the way courts have analyzed immunity from suit under

§ 1983. The Wyatt Court did not mention this concept in

relation to the good faith defense and there is no reason to think

that it would apply a historical immunity analysis to what it

obviously considered to be a distinct good faith analysis. See

Wyatt, 504 U.S. at 168. Other courts have concurred in this

view. See Danielson, 945 F.3d at 1101 (observing that Wyatt’s

discussion of the most closely analogous common law tort

19

“applies only to . . . qualified immunity” and not to the good

faith defense); Janus II, 942 F.3d at 365 (“[T]he Supreme

Court in Wyatt [] embarked on the search for the most

analogous tort only for immunity purposes—the Court never

said that the same methodology should be used for the good-

faith defense.”); Lee, 951 F.3d at 392. In any event, because

the legal basis for § 1983 immunity is distinct from the legal

basis for the good faith defense, we see no independent reason

to adopt the most analogous common law tort inquiry here. See

Danielson, 945 F.3d at 1101 (“The rationales behind

[immunity and the good faith defense], and their limitations,

are not interchangeable.”). Instead, as noted above, our

decision is based on the “principles of equality and fairness”

identified in Wyatt. 504 U.S. at 168.4

4

We note that the Appellants did not urge (or even suggest)

that we delve into the historical “common-law approach” with

the level of historical detail and specificity that JUDGE

FISHER’s concurrence would require, so we need not consider

it. Our sister circuits have construed what JUDGE FISHER refers

to broadly as the “common-law approach” as a narrower most

analogous common law tort approach, and, although they

ultimately reject the idea that this approach should be

incorporated into our analysis, they have uniformly determined

that, even if we were to adopt this mode of analysis, abuse of

process is the most analogous common law tort on these facts.

See Janus II, 942 F.3d at 365; Danielson, 945 F.3d at 1102;

Lee, 951 F.3d at 392 n.2; cf. Ogle, 951 F.3d at 797. Abuse of

process, which provides a “cause[] of action against private

defendants for unjustified harm arising out of the misuse of

governmental processes,” Wyatt, 504 U.S. at 164, corresponds

to the Unions’ use of a Pennsylvania statute to collect fair-share

20

Next, Appellants cite numerous cases in which

defendants who have taken money or property in violation of a

plaintiff’s constitutional rights have been required to disgorge

or return the money or property. First, most of these cases

involved government defendants, not private parties. But in

addition, one of the main considerations in Abood was the

benefit conferred on plaintiffs by the union activities. This has

no role in the various cases cited by Appellants. But it does

play a role when we are considering fairness because

Appellants benefitted from the fair-share fees they paid. Thus,

we are not disputing that a cause of action for return of money

or property exists for Appellants. We are merely saying that

principles of fairness make this situation different.

Third, Appellants urge that the good faith defense does

not apply to claims for restitution, which they allegedly seek.

But contrary to their urging, Appellants’ claims do not

constitute claims for restitution. “[R]estitution in equity

typically involved enforcement of a constructive trust or an

equitable lien, where money or property identified as

belonging in good conscience to the plaintiff could clearly be

traced to particular funds or property in the defendant’s

possession.” Montanile v. Bd. of Trustees of Nat’l Elevator

Indus. Health Benefit Plan, 136 S. Ct. 651, 657 (2016)

fees through government employer payroll withholding.

Abuse of process also requires a showing of malice and

probable cause, which would support the availability of the

good faith defense here. Id.; see also Jordan, 20 F.3d at 1275-

77. So, although JUDGE FISHER’s opinion goes well beyond an

analogy to abuse of process in its “common-law approach,” see

Fisher Op. at II.B.-III.B., I would not go so far, even if I were

to look to the common law for guidance on this issue.

21

(quotation marks and citation omitted). In contrast, where a

plaintiff pursues a “personal claim against the defendant’s

general assets,” then that plaintiff is seeking “a legal remedy,

not an equitable one.” Id. at 658. Appellants have not

demonstrated that their lawsuit seeks recovery from anything

more specific than the Unions’ general assets, and therefore

they fail to persuade us that they are suing for restitution. See

also Mooney, 942 F.3d at 371 (finding that the plaintiff’s claim

was “[i]n substance . . . one for damages”); Danielson, 945

F.3d at 1102-03; Lee, 951 F.3d at 391.

Appellants next theorize that the Unions can only avoid

liability—even if there is a good faith defense—if they acted

appropriately to benefit Appellants as Abood reasoned. Thus,

they urge that the District Courts should not have dismissed

their claims without allowing discovery as to whether the

Unions’ conduct was consistent with what Abood required.

But because Appellants have pled an entitlement to return of

their money based on Janus I, not on the Unions’ conduct, this

argument falls flat. See Danielson, 945 F.3d at 1105 (noting

that because plaintiffs’ “claims arise from the [u]nion’s

reliance on Abood, not allegations that the [u]nion flouted that

authority, the [u]nion need not show compliance with Abood’s

strictures to assert successfully a good faith defense”); Lee, 951

F.3d at 392 (“[I]f Defendants improperly spent the fair-share

fees, Plaintiff would have an independent Abood claim but it

would not render the exaction of the fee an act in bad faith.”

(citation omitted)).

Finally, Appellants argue that an “entity”—as opposed

to an “individual”—cannot invoke the good faith defense. But

this argument is plainly contradicted by our ruling in Jordan,

which made the good faith defense available to a law firm.

Jordan, 20 F.3d at 1277; see also Danielson, 945 F.3d at 1100

22

(rejecting argument that only individuals may invoke the good

faith defense). Appellants’ argument that the good faith

defense is incompatible with the text of § 1983 falls flat for the

same reason: Jordan involved a § 1983 cause of action.

Jordan, 20 F.3d at 1277.

IV

As Judge Wood noted in Janus II, the good faith defense

to section 1983 liability is “narrow” and “only rarely will a

party successfully claim to have relied substantially and in

good faith on both a state statute and unambiguous Supreme

Court precedent validating that statute.” 942 F.3d at 367. In

this unique circumstance, the good faith defense applies here

to protect the Unions from monetary liability under § 1983.

Accordingly, we will affirm the District Courts’ judgments.

23

Diamond v. Pa. State Educ. Ass’n, No. 19-2812

Wenzig v. Serv. Emps. Int’l, No. 19-3906

FISHER, Circuit Judge, concurring in the judgment.

In April 1871, Congress passed, and President Grant

signed, an extraordinary act, variously called the Ku Klux Klan

Act, Third Force Act, or Civil Rights Act of 1871. On its face,

the first section of that act—what we now know as 42 U.S.C.

§ 1983—provided its violators no immunities from or defenses

to liability. See Act of Apr. 20, 1871, ch. 22, § 1, 17 Stat. 13,

13. Of course, the Supreme Court has since read immunities

and defenses into § 1983, but it has done so principally on the

conceit that they were available at common law in 1871, and

implicitly incorporated into the statute. While this approach

certainly limits the scope of liability, it also constrains judges

from straying too far from the statutory text. In only one

context has the Court invented a freestanding defense: the

qualified immunity of certain state officials. Whatever might

be said for that doctrine—and it is increasingly under

scrutiny—I believe that the precedent of neither the Supreme

Court nor our own Court warrants another divergence from the

common-law approach in the present context. And however

strongly considerations of equality and fairness might

recommend such action, it is beyond our remit to invent

defenses to § 1983 liability based on our views of sound policy.

I must, therefore, respectfully disagree with the reasoning of

JUDGE RENDELL’s opinion announcing the Court’s judgment.

Nevertheless, I concur in the affirmance of the District

Courts’ orders. There was available in 1871, in both law and

equity, a well-established defense to liability substantially

similar to the liability the unions face here. Courts consistently

1

held that judicial decisions invalidating a statute or overruling

a prior decision did not generate retroactive civil liability with

regard to financial transactions or agreements conducted,

without duress or fraud, in reliance on the invalidated statute

or overruled decision. Because this defense comports with the

history and purposes of § 1983, I conclude that it is available

to the unions here and supports the dismissal of the plaintiffs’

complaints.

I

A

Section 1983 “cannot be understood in a historical

vacuum.” City of Newport v. Fact Concerts, Inc., 453 U.S. 247,

258 (1981). Despite the statute’s “general language,” Tenney v.

Brandhove, 341 U.S. 367, 376 (1951), creating a form of

liability in law and equity that seemingly “admits no

immunities,” Tower v. Glover, 467 U.S. 914, 920 (1984), the

Supreme Court has consistently construed § 1983 “in the light

of common-law principles that were well settled at the time of

its enactment,” Kalina v. Fletcher, 522 U.S. 118, 123 (1997).

Those principles “provide the appropriate starting point” for

“defining the elements of damages [under § 1983] and the

prerequisites for their recovery,” Carey v. Piphus, 435 U.S.

247, 257-58 (1978), including any available immunities and

defenses, see Pulliam v. Allen, 466 U.S. 522, 529 (1984).

The paradigm application of this common-law approach

has been the absolute immunity of legislators, judges, and

certain other state officials. Congress, the Supreme Court has

said, gave “no clear indication” in passing § 1983 that it “meant

to abolish wholesale all common-law immunities.” Pierson v.

Ray, 386 U.S. 547, 554 (1967); see also Bauers v. Heisel, 361

F.2d 581, 587-88 (3d Cir. 1966) (en banc). As a result, when an

official asserts absolute immunity, the Court has demanded “a

2

considered inquiry into the immunity historically accorded the

relevant official at common law and the interests behind it.”

Imbler v. Pachtman, 424 U.S. 409, 421 (1976). This inquiry

involves “consult[ing] the common law to identify those

governmental functions that were historically viewed as so

important and vulnerable to interference by means of litigation

that some form of absolute immunity from civil liability was

needed.” Rehberg v. Paulk, 566 U.S. 356, 363 (2012); see also

Burns v. Reed, 500 U.S. 478, 484-86 (1991); Imbler, 424 U.S.

at 422-24; Pierson, 386 U.S. at 553-54; Tenney, 341 U.S. at

376. While the scope of immunity at common law in 1871 does

not exclusively define its scope under § 1983—the statute is

not “simply a federalized amalgamation of pre-existing

common-law claims,” Rehberg, 566 U.S. at 366—the inquiry

nevertheless remains grounded in historical analogy. Judges

“do not have a license to create immunities based solely on

[their] view of sound policy.” Id. at 363.

Even when absolute immunity does not apply, the Court

has still employed the common law approach. To “defin[e] the

contours and prerequisites of a § 1983 claim,” Manuel v. City

of Joliet, 137 S. Ct. 911, 920 (2017), it has read the statute

“against the background of tort liability that makes a man

responsible for the natural consequences of his actions.”

Monroe v. Pape, 365 U.S. 167, 187 (1961); see also Memphis

Cmty. Sch. Dist. v. Stachura, 477 U.S. 299, 305-06 (1986). In

particular, the Court has looked to “[t]he common-law cause of

action . . . [that] provides the closest analogy to claims of the

type considered” pursuant to § 1983. Heck v. Humphrey, 512

U.S. 477, 484 (1994); see also Nieves v. Bartlett, 139 S. Ct.

1715, 1726 (2019). Yet here too, the elements and limitations

of a § 1983 claim will not necessarily be co-extensive with the

most analogous common-law cause of action. “Common-law

principles are meant to guide rather than to control the

3

definition of § 1983 claims,” and so “[i]n applying, selecting

among, or adjusting common-law approaches, courts must

closely attend to the values and purposes of the constitutional

right at issue.” Manuel, 137 S. Ct. at 921.

B

The singular exception to this practice is the doctrine of

qualified immunity. Early on, the Court did refer to the

common law. In Pierson, which concerned common-law and §

1983 claims against police officers, the Court held that because

“the defense of good faith and probable cause” was “[p]art of

the background of tort liability[] in the case of police officers

making an arrest,” it was available to the officers in the § 1983

action as well as the common-law action. 386 U.S. at 556-57

(citing Monroe, 365 U.S. at 187). Soon, however, as it

confronted cases involving other executive officials, the Court

generalized this defense without regard to its common-law

moorings. “[T]he relevant question” became “whether [the

official] ‘knew or reasonably should have known that the

action he took within his sphere of official responsibility would

violate the constitutional rights of [the plaintiff], or if he took

the action with the malicious intention to cause a deprivation

of constitutional rights or other injury to [the plaintiff].’”

O’Connor v. Donaldson, 422 U.S. 563, 577 (1975) (quoting

Wood v. Strickland, 420 U.S. 308, 322 (1975)); see also

Procunier v. Navarette, 434 U.S. 555, 561-62 (1978); Scheuer

v. Rhodes, 416 U.S. 232, 247 (1974).

This drift culminated in Harlow v. Fitzgerald, 457 U.S.

800 (1982), where “the Court completely reformulated

qualified immunity along principles not at all embodied in the

common law,” Anderson v. Creighton, 483 U.S. 635, 645

4

(1987).1 The Court abandoned any reference to a subjective

good-faith standard, noting that such “[i]nquiries . . . can be

peculiarly disruptive of effective government.” Harlow, 457

U.S. at 817. Instead, the question was now purely one of

objective reasonableness, and it would apply “across the

board,” id. at 821 (Brennan, J., concurring) (citation omitted),

to all “government officials performing discretionary

functions,” id. at 818 (majority opinion).

Yet even as it departed from the common-law model,

the Court indicated its unwillingness to extend Harlow’s

policy-based rationale to other contexts. “We reemphasize,” it

said in 1986, “that our role is to interpret the intent of Congress

in enacting § 1983, not to make a freewheeling policy choice,

and that we are guided in interpreting Congress’ intent by the

common-law tradition.” Malley v. Briggs, 475 U.S. 335, 342

(1986); see also Filarsky v. Delia, 566 U.S. 377, 389 (2012)

(“Nothing about the reasons we have given for recognizing

immunity under § 1983 counsels against carrying forward the

common law rule.”). Outside of qualified immunity, the

“general approach” remained the same: a court first determines

“whether an official claiming immunity under § 1983 can point

to a common-law counterpart to the privilege he asserts”; if a

sufficiently analogous counterpart exists, the court is then to

“consider[] whether § 1983’s history or purposes nonetheless

counsel against recognizing the same immunity in § 1983

actions.” Malley, 475 U.S. at 339-40 (citation omitted).

1

Although Harlow arose under the cause of action created in

Bivens v. Six Unknown Named Agents of Federal Bureau of

Narcotics, 403 U.S. 388 (1971), the Court saw no reason to

distinguish between that context and § 1983, see Harlow, 457

U.S. at 818 n.30.

5

C

This background informs the context we confront in

these cases—the far less developed area of private-party

liability under § 1983. Any limitation on such liability should,

as with official liability, “be dealt with . . . by establishing an

affirmative defense.” Lugar v. Edmondson Oil Co., 457 U.S.

922, 942 n.23 (1982); see also Adickes v. S.H. Kress & Co.,

398 U.S. 144, 174 n.44 (1970) (citing Pierson, 386 U.S. 547).

The Supreme Court has not, however, definitively stated what

such a defense might be. Rather, in Wyatt v. Cole, 504 U.S. 158

(1992), it refused to apply Harlow-style qualified immunity to

private parties sued under § 1983 for invoking a state replevin

statute later declared unconstitutional. And that is where the

doctrine remains. JUDGE RENDELL’s opinion suggests that in

rejecting the application of qualified immunity, Wyatt opened

the door to another freestanding, judge-made defense. In my

view, however, Wyatt stands for the proposition that the

common-law approach must guide any limitation on private-

party liability under § 1983.

The Wyatt defendants were private parties who invoked

a Mississippi statutory procedure that obliged state officials,

solely upon the declaration of the applicant, “to issue a writ of

replevin for the seizure of the property described in [the]

declaration.” Wyatt v. Cole, 710 F. Supp. 180, 182 (S.D. Miss.

1989). The plaintiff, whose property had been seized, filed an

action under § 1983 seeking damages and a declaratory

judgment on the statute’s constitutionality. The district court

declared the statute unconstitutional but declined to hold the

private defendants monetarily liable. Id. at 183. The Fifth

Circuit affirmed, finding the defendants entitled to qualified

immunity. Wyatt v. Cole, 928 F.2d 718, 721-22 (5th Cir. 1991)

(per curiam).

6

In reversing, the Supreme Court distinguished between

post-Harlow qualified immunity and a good-faith defense. The

basic approach, the Court said, is the one grounded in the

common law: whether the “parties seeking immunity were

shielded from tort liability when Congress enacted the Civil

Rights Act of 1871”; and, if so, whether “§ 1983’s history or

purpose counsel against applying [the immunity] in §

1983 actions.” Wyatt, 504 U.S. at 164. The defendants in fact

argued along these lines, claiming a defense under Pierson

because they acted without malice and with probable cause. Id.

at 165. The Court’s response was telling: “Even if there were

sufficient common law support to conclude that [the

defendants] . . . should be entitled to a good faith defense, that

would still not entitle them to what they sought and obtained

in the courts below: the qualified immunity from suit accorded

government officials under Harlow.” Id. As to that issue, the

Court concluded that the “special policy concerns,” articulated

in Harlow, that “mandat[e] qualified immunity for public

officials are not applicable to private parties.” Id. at 167.

For present purposes, this holding has two relevant

implications. First, contrary to what some of our sister circuits

have said, the Court in Wyatt made no suggestion that the

common-law approach applies only in the context of immunity

and not in the context of a good-faith defense. See Janus v. Am.

Fed’n of State, Cnty. & Mun. Emps., Council 31, 942 F.3d 352,

365-66 (7th Cir. 2019) (Janus II); Danielson v. Inslee, 945 F.3d

1096, 1101 (9th Cir. 2019); Lee v. Ohio Educ. Ass’n, 951 F.3d

386, 391-92 (6th Cir. 2020). In fact, the implication was

precisely the opposite: “we do not foreclose the possibility,”

the Court wrote, “that private defendants . . . could be entitled

to an affirmative defense based on good faith and/or probable

cause.” Wyatt, 504 U.S. at 169. That is the same defense

Pierson recognized, explicitly deriving it by analogy from the

7

common law. It was also the argument that the defendants in

Wyatt made before the Court, but which was “of no avail”

because it was neither sought nor ruled upon in the lower

courts. Id. at 165. And, accordingly, it was the basis of the Fifth

Circuit’s recognition of a good-faith defense on remand. See

Wyatt v. Cole, 994 F.2d 1113, 1120 (5th Cir. 1993) (Wyatt II).2

Second, in declining to extend qualified immunity to

private-party defendants, the Court did not imply, as today’s

opinion announcing our judgment holds, see Rendell Op. at

III.B, that alternative policy grounds might supply an

affirmative defense.

Although principles of equality and fairness may

suggest . . . that private citizens who rely

unsuspectingly on state laws they did not create

and may have no reason to believe are invalid

should have some protection from liability, as do

their government counterparts, such interests are

not sufficiently similar to the traditional

purposes of qualified immunity to justify such an

expansion.

Wyatt, 504 U.S. at 168. Rather than open the door to an

independent defense based on “principles of equality and

fairness,” this statement asserts that, at least in the context of

private-party § 1983 defendants, equality and fairness

2

Moreover, the distinction between immunities and defenses

is potentially misleading because qualified immunity is itself

“an affirmative defense that must be pleaded by a defendant

official.” Harlow, 457 U.S. at 815 (citing Gomez v. Toledo, 446

U.S. 635 (1980)). As I note above, the relevant distinction in

Wyatt is between Harlow-style qualified immunity and a good-

faith defense based on the common-law approach.

8

considerations are not significant enough in themselves to

warrant divergence from the common-law model in the manner

of Harlow. Those concerns “may be well founded,” but courts

“do not have a license to establish immunities from § 1983

actions in the interests of what [they] judge to be sound public

policy.” Tower, 467 U.S. at 922-23.

Justice Kennedy’s concurrence in Wyatt, joined by

Justice Scalia, underlines both of these points. “Our immunity

doctrine,” he wrote, “is rooted in historical analogy, based on

the existence of common-law rules in 1871, rather than in

‘freewheeling policy choices.’” Wyatt, 504 U.S. at 170

(Kennedy, J., concurring) (alteration omitted) (quoting Malley,

475 U.S. at 342). Although Harlow “depart[ed] from history in

the name of public policy,” Justice Kennedy joined the Court’s

opinion in resisting “exten[sion] [of] that approach to other

contexts.” Id. at 171. “[W]e may not transform what existed at

common law based on our notions of policy or efficiency.” Id.

at 171-72. The implication is that any limitation on private-

party liability must be grounded in the common-law approach.

Justice Kennedy then went further than the Court in

laying out what such an inquiry, at least on the Wyatt facts,

should look like. All of the Justices, including those in dissent,

accepted that at common law in 1871 the tort actions “most

closely analogous” to the Wyatt action were “malicious

prosecution and abuse of process.” Id. at 164 (majority

opinion); see id. at 172 (Kennedy, J., concurring); id. at 176

(Rehnquist, C.J., dissenting). Both torts required the plaintiff

to prove that the defendant acted with malice and without

probable cause. Id. at 166 n.2 (majority opinion); id. at 172

(Kennedy, J., concurring); id. at 176 n.1 (Rehnquist, C.J.,

dissenting). For Justice Kennedy, proof of “subjective bad faith

on the part of the defendant”—rather than an objective

standard—went “far towards proving” both elements. Id. at

9

173 (Kennedy, J., concurring). “[T]here is support in the

common law,” he observed, “for the proposition that a private

individual’s reliance on a statute, prior to a judicial

determination of unconstitutionality, is considered reasonable

as a matter of law; and therefore under the circumstances of

this case, lack of probable cause can only be shown through

proof of subjective bad faith.” Id. at 174 (citing Birdsall v.

Smith, 122 N.W. 626, 627 (Mich. 1909)). Further, five Justices

agreed that a “good-faith defense” in this context represented

both the plaintiff’s burden to prove the elements of the offense

and, relatedly, the defendant’s opportunity to avoid liability by

showing good faith. See id. at 175; id. at 176 n.1 (Rehnquist,

C.J., dissenting).

II

Under Wyatt, then, any defense to private-party liability

under § 1983 must derive from the common-law approach and

may not rest on freestanding policy grounds. The next question

is whether the defense suggested there—whether the defendant

acted with malice and without probable cause—is context

dependent or applies categorically to all cases involving

private-party defendants. Only the former view is faithful to

the common-law approach; the latter, like the Supreme Court’s

qualified-immunity standard in cases such as Procunier,

O’Connor, and Wood, generalizes a subjective good-faith

defense, unmooring it from its common-law origins. JUDGE

RENDELL’s opinion, in addition to its policy-based holding,

takes this latter view, relying upon our decision in Jordan v.

Fox, Rothschild, O’Brien & Frankel, 20 F.3d 1250 (3d Cir.

1994). See Rendell Op. at III.A-B. On my reading, however,

Jordan did not announce a categorical rule, and so we must

conduct an independent inquiry based on the common-law

approach. And on that score, I think that instead of determining

whether a pre-1871 tort is sufficiently analogous, resolution on

10

an alternative ground, also based in the common-law approach,

is preferable.

A

Lugar and Wyatt both concerned “private defendants

charged with 42 U.S.C. § 1983 liability for invoking state

replevin, garnishment, and attachment statutes later declared

unconstitutional.” Wyatt, 504 U.S. at 159. So too did Jordan.

Pursuant to a cognovit clause in a commercial real estate lease,

the defendants obtained and executed a confessed judgment

against the plaintiffs in state court. Jordan, 20 F.3d at 1258.

Along with their complaint, the defendants invoked a

Pennsylvania procedure that required the prothonotary of the

court to issue a writ ordering the court’s sheriff to garnish the

plaintiffs’ bank account. Jordan v. Fox, Rothschild, O’Brien &

Frankel, 787 F. Supp. 471, 473-74 (E.D. Pa. 1992) (Fox

Rothschild). The law required neither pre-deprivation notice

nor issuance of a writ of service, and indeed the plaintiffs

received notice only after the seizure. Id. Unsurprisingly

aggrieved, the plaintiffs thereafter sought, among other things,

a declaratory judgment that the Pennsylvania procedure was

unconstitutional and damages under § 1983.

The district court held that the post-judgment

garnishment phase of the procedure violated due process, id. at

477-78, but it dismissed the § 1983 action, determining that the

defendants were entitled to qualified immunity, id. at 479-80.

While the case was pending on appeal, however, the Supreme

Court decided Wyatt. Our question, then, was whether the

defendants were entitled to a good-faith defense. Jordan, 20

F.3d at 1276. We held that they were, declaring ourselves “in

basic agreement” with the Fifth Circuit’s holding on remand in

Wyatt that “[p]rivate defendants should not be held liable under

§ 1983 absent a showing of malice and evidence that they

11

either knew or should have known of the statute’s

constitutional infirmity.” Id. (quoting Wyatt II, 994 F.2d at

1120).

In my view, Jordan’s holding is best read as limited to

the context before it. Immediately after announcing our

agreement with the Fifth Circuit, we clarified that by “malice”

we had in mind “a creditor’s subjective appreciation that its act

deprives the debtor of his constitutional right to due process.”

Id. To support this standard, we cited Justice Kennedy’s

reference, in his Wyatt concurrence, to Birdsall v. Smith. Id. at

1276 n.30. That case concerned a malicious-prosecution action

brought by a milk vendor who had been charged, solely on the

basis of a report filed with state officials, under a state statute

later declared unconstitutional. See 122 N.W. at 626-27. We

also referred to “Pennsylvania cases that place state law

limitations on the use of judgment by confession” because we

thought they may “sometimes be relevant on the good faith

issue.” Jordan, 20 F.3d at 1277. This all suggests that we had

in mind the factual circumstances of the immediate case—

circumstances essentially similar to those of Lugar and Wyatt.

B

Because Jordan cannot be read as expansively as JUDGE

RENDELL’s opinion suggests, the proper question is whether

the abuse-of-process and malicious-prosecution torts, from

which the Wyatt defense is derived, are sufficiently analogous

to the present action, such that our recognition of that defense

in Jordan is applicable here. For their part, our sister circuits

that have confronted the question have so far uniformly

concluded that those torts do provide the best analogy. See,

e.g., Janus II, 942 F.3d at 365; Danielson, 945 F.3d at 1102;

Lee, 951 F.3d at 392 n.2. I think that view is worth questioning,

12

at least to the extent that it supplies the unions a good-faith

defense here.

In both Wyatt and Jordan, the private-party defendants

invoked a generally available state procedure. Upon the

defendants’ independent initiative, state officials were

compelled to seize or garnish property of the plaintiffs. That

mandate was what rendered the state laws unconstitutional in

each case. See Fox Rothschild, 787 F. Supp. at 477-78; Cole,

710 F. Supp. at 183. Here, Pennsylvania law required the public

employer to deduct the fair-share fee from the nonmembers’

paychecks, if the collective-bargaining agreement so provided.

Yet (and this is the key difference) the agreements triggering

collection of the fees were not the fruit of the unions’

independent initiative—the relevant public employer was a

party to them and necessarily had to agree to them. See 71 Pa.

Stat. § 575(b)-(c); see also 43 Pa. Stat. § 1101.901 (the

collective-bargaining agreement is “between the

representatives of the public employes and the public

employer”). And the collection of the fees—the compelled

subsidization of speech—was the constitutional violation. See

Janus v. Am. Fed’n of State, Cnty. & Mun. Emps., Council 31,

138 S. Ct. 2448, 2464, 2478 (2018).

Thus, the relevant state action in our cases stems not

merely from the involvement of state officials in

unconstitutional conduct, see Lugar, 457 U.S. at 941, but also,

to some extent, from the command or express authorization of

the state to engage in that conduct, see Blum v. Yaretsky, 457

U.S. 991, 1004 (1982). From this perspective, the torts of abuse

of process and malicious prosecution provide at best attenuated

analogies. It seems apparent that we are not dealing here

simply with a civil “process . . . willfully made use of for a

purpose not justified by the law,” Thomas M. Cooley, A

Treatise on the Law of Torts 189 (1876), let alone “the

13

malicious institution of a civil suit,” id. at 187. Insofar as the

state establishes a law’s justified purposes, we confront the use

of a procedure for a purpose that the state in part set.3

It may be, as the Seventh Circuit observed in Janus II,

that abuse of process and malicious prosecution are the most

analogous torts, however imperfect the analogy. See 942 F.3d

at 365. But it does not necessarily follow that they therefore

supply the basis of a defense. By that logic, a defense is

potentially always available, no matter how attenuated the

connection between the common-law cause of action and the

injury alleged. We must remember that “[c]ommon-law

principles are meant to guide rather than to control the

definition of § 1983 claims.” Manuel, 137 S. Ct. at 921. True

commitment to the common-law approach may eventually

require deciding where to draw the line between analogous and

non-analogous causes of action. But at least in this case, I find

it unnecessary to do so.

In what follows, I describe an alternative basis for a

defense, well established at both common law and equity in

1871, and providing a closer similarity to the facts that we

confront. Resolving these cases on this ground would both

avoid the knotty problems raised by a most-analogous-tort test

and preserve the notion, accepted by six Justices in Wyatt, that

Harlow was an exception that should not swallow the

common-law rule. Indeed, in my view, that latter benefit is

especially compelling, given the recent cogent critiques of

3

It follows from this argument that the parties’ other proposed

torts—conversion, defamation, tortious interference with

contract, and intentional infliction of emotional distress—are

also insufficiently analogous. Their elements are even further

afield than those of abuse of process and malicious

prosecution.

14

qualified immunity as incongruent with the principles of

statutory interpretation. See, e.g., Ziglar v. Abbasi, 137 S. Ct.

1843, 1871-72 (2017) (Thomas, J., concurring in part and

concurring in the judgment); Baxter v. Bracey, 140 S. Ct. 1862,

1864 (2020) (Thomas, J., dissenting from the denial of

certiorari); William Baude, Is Qualified Immunity Unlawful?,

106 Calif. L. Rev. 45 (2018).

III

“An unconstitutional act is not a law; . . . it is, in legal

contemplation, as inoperative as though it had never been

passed.” Norton v. Shelby County, 118 U.S. 425, 442 (1886).

Derived from the common law, see Robinson v. Neil, 409 U.S.

505, 507 (1973), this principle from the late nineteenth century

was premised on the then-prevalent legal theory that judges

“find” or “declare” rather than “make” law, see Linkletter v.

Walker, 381 U.S. 618, 622-23 (1965); Kuhn v. Fairmont Coal

Co., 215 U.S. 349, 370 (1910) (Holmes, J., dissenting). That

theory fell out of fashion in the early twentieth century, but the

Norton principle nevertheless proved remarkably influential.

See, e.g., Ex Parte Young, 209 U.S. 123, 159 (1908). Most

notably, it underlies the Supreme Court’s more recent

retroactivity jurisprudence—and thus the plaintiffs’ theory of

liability in the present cases. See Harper v. Va. Dep’t of Tax’n,

509 U.S. 86, 95-97 (1993); James B. Beam Distilling Co. v.

Georgia, 501 U.S. 529, 540 (1991) (opinion of Souter, J.);

Griffith v. Kentucky, 479 U.S. 314, 326-29 (1987).

Yet there was a contemporaneous exception to this

general view, in which a judicial decision either voiding a

statute or overruling a prior decision does not generate

retroactive civil liability with regard to financial transactions

or agreements conducted, without duress or fraud, in reliance

on the invalidated statute or overruled decision. See, e.g.,

15

Benjamin N. Cardozo, The Nature of the Judicial Process 146-

47 (1921); Oliver P. Field, The Effect of an Unconstitutional

Statute 221-28 (1935); Note, The Effect of Overruled and

Overruling Decisions on Intervening Transactions, 47 Harv. L.

Rev. 1403 (1934). An assessment of the cases applying this

exception demonstrates its applicability in the present context.

The exception appears to have developed as a sort of

corollary to originally English legal and equitable doctrines.

One such doctrine is that voluntary payments made upon an

illegal demand are not recoverable except where the payments

were made under an immediate and urgent necessity. See, e.g.,

Valpy v. Manley (1845), 135 Eng. Rep. 673, 677; 1 C. B. 594,

602-03 (Tindal, C.J.) (citing and quoting Fulham v. Down

(1798), 170 Eng. Rep. 820 n.; 6 Esp. 26 n. (Kenyon, C.J.));

Brisbane v. Dacres (1813), 128 Eng. Rep. 641, 645; 5 Taunt.

143, 152 (Gibbs, J.). Another is that money paid pursuant to a

contract may not be recovered if the contract was formed under

a mutual mistake of law. See, e.g., Bilbie v. Lumley (1802) 102

Eng. Rep. 448, 449-50; 2 East 469, 472. Although nineteenth-

century American courts straightforwardly applied these

doctrines in the contexts in which they originated, see, e.g.,

Bank of U.S. v. Daniel, 37 U.S. (12 Pet.) 32, 55-56 (1838); Hunt

v. Rhodes, 26 U.S. (1 Pet.) 1, 15 (1828); Sprague v. Birdsall, 2

Cow. 419, 421 (N.Y. Sup. Ct. 1823), they also invoked them

when confronting the effects of the practice of judicial review.

Two lines of cases—one at law, the other in equity—are

especially notable.

A

At common law, money extracted illegally by taxes or

fees could be recovered through an action of assumpsit. See,

e.g., 3 William Blackstone, Commentaries on the Laws of

England *158-59 (describing as a form of assumpsit an action

16

to recover tax or fee payments to a government or other body

of which one is a member).4 As noted, in Janus the

unconstitutional act was the compelled subsidization of speech

through the payment of the fair-share fees. The plaintiffs here

seek a repayment of the fees they paid prior to Janus and whose

extraction only became illegal as a result of that decision.

Several pre-1871 state cases address a similar situation, where

repayment of a tax, fee, or other expenditure is sought when

the law or court decision under which it was made is declared

unconstitutional or overruled. The courts in these cases

developed a limitation on such liability, uniformly barring

repayment where the initial expenditure was made voluntarily

and without duress.

The most succinct formulation of this doctrine came in

an 1846 decision of the Maryland high court:

It is now established, by an unbroken series of

adjudications in the English and American

4

Although the Supreme Court has often referred specifically to

tort law when enunciating the common-law approach to § 1983

immunities and defenses, it has never suggested that

application of that approach is limited to tort, rather than

contract, law where the latter is most applicable. Moreover, the

assumpsit action was in fact a form of the writ of trespass on

the case—the fountainhead of modern tort law—that officially

came to supplant actions in debt due to the institutional rivalry

of the Courts of Common Pleas and King’s Bench. See David

Ibbetson, Sixteenth Century Contract Law: Slade’s Case in

Context, 4 Oxford J. Legal Stud. 295 (1984). Assumpsit treats

misperformance or nonperformance of an implied agreement

as a tort-like wrong. See John H. Langbein et al., History of the

Common Law: The Development of Anglo-American Legal

Institutions 252 (2009).

17

courts, that where money is voluntarily and fairly

paid, with a full knowledge of the facts and

circumstances under which it is demanded, it

cannot be recovered back in a court of law, upon

the ground, that the payment was made under a

misapprehension of the legal rights and

obligations of the party.

City of Baltimore v. Lefferman, 4 Gill 425, 431 (Md. 1846). The

operative legal fiction—consistent with the Supreme Court’s

later statement in Norton—is that a statute or ordinance

subsequently declared unconstitutional is void even at the time

the money is transacted pursuant to it, thus creating the

“misapprehension.” The burden, however, is on the payor to

establish more than mere reliance on the law’s presumptive

validity. As the California Supreme Court put it: “The illegality

of the demand paid constitutes of itself no ground for relief.

There must be in addition some compulsion or coercion

attending its assertion, which controls the conduct of the party

making the payment.” Brumagim v. Tillinghast, 18 Cal. 265,

266 (1861). The payment, according to the Ohio Supreme

Court, “can only be considered involuntary when it is made to

procure the release of the person or property of the party from

detention, or when the other party is armed with apparent

authority to seize upon either, and the payment is made to

prevent it.” Mays v. City of Cincinnati, 1 Ohio St. 268, 278

(1853). Simply because the law was assumed valid at the time

of the payment, and therefore that non-payment might result in

legal enforcement proceedings, was not enough. See Town

Council of Cahaba v. Burnett, 34 Ala. 400, 404 (1859); see also

Town of Ligonier v. Ackerman, 46 Ind. 552, 559 (1874),

overruled in part on other grounds by Jennings v. Fisher, 2

N.E. 285, 288 (Ind. 1885).

18

The Pennsylvania Supreme Court at midcentury also

adopted this general doctrine. See Taylor v. Phila. Bd. of

Health, 31 Pa. 73, 75 (1855); Borough of Allentown v. Saeger,

20 Pa. 421 (1853). In Saeger, the Court stated in dictum that

“[i]f [the money] had been paid under protest, that is, with

notice that [the payor] would claim it back, this would repel the

implication of an assent, and give rise to the right of

reclamation.” 20 Pa. at 421. It is unclear, however, if this

standard required the payor actually to bring the threatened

legal action. Other courts were more explicit in imposing this

requirement. See, e.g., Burnett, 34 Ala. at 405 (“[T]he case is

not altered by the fact, that the party so paying protests that he

is not answerable, and gives a notice that he shall bring an

action to recover the money back. He has an opportunity in the

first instance to contest th[e] claim at law.” (quoting Benson v.

Monroe, 61 Mass. (7 Cush.) 125, 131 (1851))).5

Finally, although the United States Supreme Court did

not, during this period, have a factually similar case, it did

approvingly recite this doctrine in analogous situations. For

example, in an 1877 case involving payments to Confederate

5

The Alabama Supreme Court’s adoption of Benson’s

language is significant. Benson, also an assumpsit action, more

nearly approximates abuse of process because the plaintiffs,

who were ship owners, only paid after their vessel was

attached. Nevertheless, the Massachusetts Supreme Judicial

Court still denied recovery. The plaintiffs had the choice of

either paying or litigating. Benson, 61 Mass. at 131. Burnett’s

importation of Benson’s standard suggests the similarity

between the sort of cases described here and abuse-of-process

situations (though still litigated in assumpsit). It suggests the

closeness of this rule to the one Wyatt suggested and our Court

adopted in Jordan.

19

officials for the right to export cotton, the Court said that to

“justify an action against [the payees], either for the return of

the money paid . . . or for damages of any kind,” “the doctrine

established by the authorities is[] that ‘a payment is not to be

regarded as compulsory, unless made to emancipate the person

or property from an actual and existing duress imposed upon it

by the party to whom the money is paid.’” Radich v. Hutchins,

95 U.S. 210, 212-13 (1877) (quoting Lefferman, 4 Gill. at 436,

and citing Brumagim, 18 Cal. at 265; and Mays, 1 Ohio St. at

268); see also Elliott v. Swartwout, 35 U.S. (10 Pet.) 137, 153-

55 (1836). This voluntariness rule remains the applicable

standard. See McKesson Corp. v. Div. of Alcoholic Beverages

& Tobacco, 496 U.S. 18, 38 n.21 (1990).

B

The doctrine was also applied in equitable actions,

usually involving not the payment of a tax or fee, but rather a

financial transaction between private parties. Its most well-

known enunciation was by Chancellor Kent in 1815: “A

subsequent decision of a higher Court, in a different case,

giving a different exposition of a point of law from the one

declared and known when a settlement between parties takes

place, cannot have a retrospective effect, and overturn such

settlement.” Lyon v. Richmond, 2 Johns. Ch. 51, 60 (N.Y. Ch.

1815), rev’d on other grounds, Lyon v. Tallmadge, 14 Johns.

501 (N.Y. 1817). In addition to general policy grounds, the key

principle was, again, that parties may not be relieved of “acts

and deeds fairly done on a full knowledge of facts, though

under a mistake of the law.” Id.; see also Shotwell v. Murray, 1

Johns. Ch. 512, 515-16 (N.Y. Ch. 1815). Later state equity

courts adopted or followed this doctrine, see, e.g., Doll v.

Earle, 59 N.Y. 638, 638 (1874); Hardigree v. Mitchum, 51 Ala.

151, 155-56 (1874); Harris v. Jex, 55 N.Y. 421, 424 (1874);

Kenyon v. Welty, 20 Cal. 637, 642 (1862), as did at least one

20

federal court, see In re Dunham, 8 F. Cas. 37, 38-39 (D.N.J.

1872).

***

When Congress in 1871 enacted the law that became §

1983, it was well established at both law and equity that court

decisions that invalidated a statute or overruled a prior

decision, and thereby affected transactional relationships—

between private parties and government officials or

representatives, or between private parties alone—established

in reliance on that statute or decision, did not generate civil

liability for repayment except where duress or fraud was

present. Whatever the nature of the state action in the present

cases—whether the state “act[ed] jointly with” the unions or

“compel[led] the [unions] to” collect the fees, Manhattan

Cmty. Access Corp. v. Halleck, 139 S. Ct. 1921, 1928 (2019)—

the factual circumstances underlying this doctrine bear a

substantial similarity to those we confront here. Therefore, in

my view the doctrine constitutes “a previously existing,

independent legal basis” sufficient to limit the unions’ liability

under § 1983. Reynoldsville Casket Co. v. Hyde, 514 U.S. 749,

759 (1995).6 I know of no authority on “§ 1983’s history or

purposes” that might “counsel against” recognition of this

defense, Tower, 467 U.S. at 920, and the consistency of its

application in law and equity safely permits the conclusion that

Congress did not wish to “impinge” on it “by covert inclusion

6

The Diamond appellants argue strenuously that this is a case

of restitution. Even if it is, every case upon which they rely can

be explained according this doctrine. Moreover, they cite cases

only from the mid-twentieth century or later. There is no

suggestion that the principle they claim was established in

1871. The reverse, in fact, seems to be the case.

21

in the general language” of § 1983, Tenney, 341 U.S. at 376.

IV

It may be tempting, in cases like the present, to read

precedent broadly, or appeal to freestanding principles such as

the rule of law and basic notions of fairness. But we must

interpret and apply § 1983 as we would any other statute,

always prepared for the faithful execution of that duty to result

in a seemingly extreme outcome. For even when that does not

occur, there is value in adhering to the well-established

principles of interpretation.

Because the plaintiffs in these cases have not pleaded

any facts, suggesting that their payments were either

sufficiently involuntary or exacted on a fraudulent basis,7 to

permit a reasonable person to infer that the unions might be

liable, I concur in the affirmance of the orders granting the

unions’ motions to dismiss.

7

JUDGE PHIPPS asserts that, even accepting the standard I adopt

here, the plaintiffs’ payments were not voluntary. I think it

apparent that none of the plaintiffs have pleaded anything

approaching the kind of involuntariness or duress articulated in

the cases I discuss.

22

Diamond v. Pa. State Educ. Ass’n, No. 19-2812

Wenzig v. Serv. Emps. Int’l, No. 19-3906

PHIPPS, Circuit Judge, dissenting.

The central question presented in these consolidated

cases, which seek recovery of agency fees garnished from the

wages of non-union members, is whether a good faith

affirmative defense exists to a First Amendment compelled

speech claim under 42 U.S.C. § 1983. I do not see a valid basis

for recognizing such a defense. A good faith affirmative

defense was not firmly rooted in the common law in 1871 when

§ 1983 was enacted, and nothing else compels recognition of

such a defense today. For that reason, I would reverse the

orders dismissing these cases and remand them for further

proceedings.

My colleagues see it differently. Judge Rendell

recognizes such a defense from precedent and out of

consideration of “principles of equality and fairness.” Rendell

Op. at III.B. In concurring in the judgment only, Judge Fisher

does not rely on a good faith defense. Instead, from an

examination of pre-1871 common law, he identifies another

limitation on the § 1983 cause of action: it may not be used to

collect voluntary payments. See Fisher Op. at III.A. I disagree

with these perspectives and respectfully dissent.

The Supreme Court has articulated standards for

supplementing the plain text of § 1983, which itself identifies

no immunities or defenses. Such supplementation requires a

tradition “so firmly rooted in the common law and . . .

supported by such strong policy reasons that ‘Congress would

have specifically so provided had it wished to abolish the

doctrine.’” Owen v. City of Independence, 445 U.S. 622, 637

(1980) (quoting Pierson v. Ray, 386 U.S. 547, 555 (1967)).

Even if such a deeply rooted common-law tradition exists, that

will still not permit supplementation of § 1983 in a manner

inconsistent with the statute’s history or purpose. See Wyatt v.

Cole, 504 U.S. 158, 164 (1992) (“[I]rrespective of the common

law support, we will not recognize an immunity available at

common law if § 1983’s history or purpose counsel against

applying it in § 1983 actions.”).

I. A GOOD FAITH DEFENSE WAS NOT FIRMLY ROOTED IN

THE COMMON LAW IN 1871 WHEN CONGRESS ENACTED

§ 1983.

The specific inquiry here focuses on whether a good

faith defense was firmly rooted in the common law in 1871.

But as an initial point of reference, the good faith affirmative

defense is not firmly rooted in the common law today – either

generally or for any specific cause of action.

In articulating 18 affirmative defenses that must be

raised in a responsive pleading, Rule 8(c) of the Federal Rules

of Civil Procedure does not include good faith. See Fed. R.

Civ. P. 8(c). The rule’s listing is not exhaustive, and leading

treatises supplement those 18 listed defenses, but those

treatises do not identify a common-law good faith affirmative

defense. See, e.g., Arthur R. Miller et al., Federal Practice and

Procedure § 1271 (3d ed., Apr. 2020 Update) (recognizing no

common-law good faith affirmative defense); 2 Jeffrey A.

Parness, Moore’s Federal Practice § 8.08 (3d ed. 2020) (listing

affirmative defenses, such as immunities, but not including

good faith). If a good faith affirmative defense were deeply

rooted in the common law, such as defenses like statute of

2

limitations, laches, or accord and satisfaction, then one would

expect to find it listed in Rule 8(c) – or at least to make a

showing in a leading treatise.

Similarly, a review of other statutory causes of action

reveals that Congress has not understood good faith to be so

deeply rooted as to go unspoken. Rather, when Congress

wants to include good faith as an affirmative defense, it does

so expressly.1 And that begs the question: if the good faith

defense were so well established that it could be assumed “that

Congress [in enacting § 1983] would have specifically so

provided had it wished to abolish the doctrine,” then why did

Congress find the need to expressly provide for the defense in

many other statutes but not in § 1983? Pierson, 386 U.S. at

555.

In sum, the absence of a good faith affirmative defense

from Rule 8(c) along with its presence as a defense in other

federal statutes suggests that today the good faith affirmative

defense is not firmly rooted in the common law.

1

See, e.g., 15 U.S.C. § 78r (providing a good faith defense to

securities fraud); 15 U.S.C. § 1115(b) (providing a good faith

defense to trademark infringement); 15 U.S.C. §§ 1640,

1691e(e), 1692k(e), 1693m(d) (providing a good faith defense

to claims related to consumer credit protection); 16 U.S.C

§ 1540(a)(3), (c)(3) (providing a good faith defense to certain

claims under the Endangered Species Act); 29 U.S.C. § 259(a)

(providing a good faith defense to certain claims under the Fair

Labors Standards Act); 29 U.S.C. § 2617(a)(1)(iii) (providing

a good faith defense to a liquidated damages claim under the

Family Medical Leave Act).

3

That conclusion, of course, is not dispositive – it could

be that a good faith affirmative defense was deeply entrenched

in the common law in 1871 but has lost traction over time. But

cf. Fed. R. Civ. P. 8(c) (continuing to identify the virtually

obsolete affirmative defense of injury to fellow servant). To

make such a showing would require proof similar to that

adduced in Tenney v. Brandhove, 341 U.S. 367 (1951),

wherein the Supreme Court determined that legislative

immunity applied to § 1983 claims. See id. at 377-78. In

reaching that conclusion, the Supreme Court relied on evidence

of that immunity dating back to sixteenth and seventeenth

century English law, provisions of the Articles of

Confederation and the Constitution, as well as protections

specifically articulated in 41 of the then 48 admitted States.

See id. at 372-76.

By contrast no such evidence is present here. No party

identifies a pre-1871 case recognizing a common-law good

faith affirmative defense – either as a general matter or in the

context of any particular cause of action. Judge Rendell’s

opinion does not identify any common-law basis for such a

defense. Nor do any of the other courts applying a good faith

defense to agency fee cases identify any grounding in common

law for such an affirmative defense.2

2

See Wholean v. CSEA SEIU Local 2001, 955 F.3d 332, 334-

36 (2d Cir. 2020); Lee v. Ohio Educ. Ass’n, 951 F.3d 386, 392

n.2 (6th Cir. 2020); Ogle v. Ohio Civil Serv. Emps. Ass’n,

AFSCME Local 11, 951 F.3d 794, 797 (6th Cir. 2020) (per

curiam); Danielson v. Inslee, 945 F.3d 1096, 1102 (9th Cir.

2019); Janus v. Am. Fed. of State, Cty. & Mun. Emps., Council

31, 942 F.3d 352, 364 (7th Cir. 2019) (finding “no common-

4

The strongest case for such a defense comes from Chief

Justice Rehnquist’s dissenting opinion in Wyatt v. Cole. There,

he viewed the good faith defense as “something of a

misnomer” because it actually referred to elements of the

common-law torts of malicious prosecution and abuse of

process. 504 U.S. 158, 176 & n.1. That perspective is telling.

Chief Justice Rehnquist identified no authority for the

proposition that good faith functions as a transsubstantive

affirmative defense – applicable across a broad class of claims,

such as the defenses of accord and satisfaction, laches, and res

judicata. See id. at 175-80. Nor did his dissenting opinion

recognize good faith as a claim-specific affirmative defense,

such as the defenses of assumption of risk, contributory

negligence, or duress. See id. At most, Chief Justice Rehnquist

determined that the elements of two common-law tort claims

could be defeated by proof of subjective good faith. See id. at

176 & n.1.

Judge Fisher picks up on that theme. From an

examination of the common law, he concludes that in 1871 no

cause of action allowed for later recovery of voluntary

payments. See Fisher Op. at III.A. But unlike the cases he

relies upon, the agency fee payments at issue here were not

voluntary – they were wage garnishments that were paid to

unions.3 More fundamentally, Judge Fisher’s approach is

law history before 1871 of private parties enjoying a good-faith

defense to constitutional claims”).

3

See 71 Pa. Stat. and Cons. Stat. Ann. § 575(c) (West 1988)

(requiring employers to garnish wages for fair-share agency

fees for transmittal to unions); see also Wenzig Compl. ¶¶ 9-

10 (Wenzig App. 42) (alleging that non-union members were

5

analogous to the one that the Supreme Court did not adopt in

Wyatt – which prompted Chief Justice Rehnquist’s dissent.

Section 1983 created a new statutory cause of action, not one

pre-defined by the common law. Thus, it is immaterial that no

pre-1871 cause of action permitted recovery for voluntary

payments that were subsequently declared unconstitutional:

the Civil Rights Act of 1871 established a new cause of action

in part to provide “a remedy where state law was inadequate.”

Monroe v. Pape, 365 U.S. 167, 173 (1961), overruled on other

grounds by Monell v. Dep’t of Soc. Servs. of N.Y., 436 U.S. 658

(1977).

For these reasons, I do not see the common law as

limiting the scope of a § 1983 claim for compelled speech –

either through a good faith affirmative defense or through a

separate limitation on the statutory cause of action.

II. BOTH THE HISTORY AND THE PURPOSE OF § 1983

COUNSEL AGAINST RECOGNITION OF A GOOD FAITH

AFFIRMATIVE DEFENSE.

For completeness, even supposing that the common law

did recognize good faith as an affirmative defense in 1871,

more would be required. Before a deeply rooted affirmative

“forced to pay” fair-share agency fees and that those fees were

deducted from nonmembers’ wages “without their consent”);

Diamond Second Am. Compl. ¶ 24 (Diamond App. 74)

(alleging that the class representatives were “compelled . . . to

pay a financial penalty for exercising their constitutional right

to not join a union”), ¶ 39 (Diamond App. 77) (defining the

putative class as persons who were “compelled to pay money .

. . as a condition of employment”).

6

defense can apply to a § 1983 action, it must also be “supported

by such strong policy reasons that Congress would have

specifically so provided had it wished to abolish the doctrine.”

Owen, 445 U.S. at 637 (internal quotation marks omitted). Put

differently, a common-law defense will not be read into § 1983

when it is inconsistent with the history or the purpose of

§ 1983. See Wyatt, 504 U.S. at 164. And neither the history

nor the purpose of § 1983 supports the recognition of good

faith as an affirmative defense for violations of every

constitutional right.

A good faith defense is inconsistent with the history of

the Civil Rights Act of 1871. As the Supreme Court has

explained, that statute is predicated on the understanding that

“Congress has the power to enforce provisions of the

Fourteenth Amendment against those who carry a badge of

authority of a State and represent it in some capacity, whether

they act in accordance with their authority or misuse it.”

Monroe, 365 U.S. at 171-72 (emphasis added). As this

statement makes clear, the history behind the Civil Rights Act,

which Congress enacted pursuant to the Enabling Clause of the

Fourteenth Amendment,4 demonstrates the need to remedy

actions taken in accordance with state law. And thus a good

faith affirmative defense – that a state actor was merely

following state law – is an especially bad fit as an atextual

addition to § 1983.

4

See Civil Rights Act of 1871, Pub. L. 42-22, 17 Stat. 13, 13

(Apr. 20, 1871) (entitling the legislation as “[a]n Act to enforce

the [p]rovisions of the Fourteenth Amendment . . . and for other

[p]urposes”).

7

Nor can a good faith affirmative defense be reconciled

with the purpose of the Civil Rights Act of 1871. The Supreme

Court has identified “three main aims” for § 1983. Monroe,

365 U.S. at 173. Those were (i) “to override certain kinds of

state laws”; (ii) to provide “a remedy where state law was

inadequate”; and (iii) “to provide a federal remedy where the

state remedy, though adequate in theory, was not available in

practice.” Id. at 173-74. Each of those purposes reflects a

dissatisfaction with the redress provided by state law for

constitutional violations. It would seem, then, that state law

would be the last place to look for limitations on the redress

§ 1983 allows – the whole point of the statute was to overcome

the limitations of state law. Thus, absent some foundation in

federal law, incorporating a defense rooted only in state

common law into § 1983 is inconsistent with the purpose of

that statute.

The later enactment of § 1988 also supports this

conclusion. There, Congress allowed for consideration of state

common law, but only to supplement “deficienc[ies] in the

provisions necessary to furnish suitable remedies and punish

offenses against law.” 42 U.S.C. § 1988. That is quite

different than looking to state common law to limit the

remedies permitted by § 1983.

Thus, even if it were firmly entrenched in the common

law, a good faith affirmative defense should not be grafted onto

the text of § 1983 – either as a transsubstantive defense (such

as accord and satisfaction or res judicata) or a cause-of-action

specific defense (such as assumption of the risk or duress).

8

III. THE ROLE OF GOOD FAITH IN § 1983 LITIGATION DOES

NOT RISE TO THE LEVEL OF AN AFFIRMATIVE DEFENSE.

Although good faith does not operate as an affirmative

defense, it still may have a role in § 1983 litigation. As this

Circuit recognized, proof of good faith may negate an element

of a § 1983 claim. See Jordan v. Fox, Rothschild, O’Brien &

Frankel, 20 F.3d 1250, 1277-78 (3d Cir. 1994). Specifically,

the gross negligence mental state element required for a

procedural due process claim can be rebutted by a showing of

subjective good faith through adherence to then-existing law.

See id. at 1278. That holding was context specific, and it

recognized good faith as a means to disprove a mental state

requirement. See id. at 1277-78. Consistent with Chief Justice

Rehnquist’s observation, the Jordan decision used the term

‘good faith defense’ as a misnomer – it was actually applying

good faith to negate a specific element of a cause of action, as

opposed to asserting it as an affirmative defense. See id.; see

generally Affirmative Defense, Black’s Law Dictionary (11th

ed. 2019) ( “A defendant’s assertion of facts and arguments

that, if true, will defeat the plaintiff’s or prosecution’s claim,

even if all the allegations in the complaint are true.”). Thus, I

do not read our precedent as recognizing good faith as an

across-the-board affirmative defense, or even as cause-of-

action specific affirmative defense. At most, a showing of

good faith can negate a mental state element of a claim – such

as gross negligence required for a procedural due process

claim. See Jordan, 20 F.3d at 1277-78. But that is of no

moment here because a claim for compelled speech does not

have a mens rea requirement. See Janus v. Am. Fed’n of State,

Cty. & Mun. Emps., Council 31, 138 S. Ct. 2448, 2464 (2018)

(“[T]he compelled subsidization of private speech seriously

9

impinges on First Amendment rights[.]”); see also United

States v. United Foods, Inc., 533 U.S. 405, 408, 416 (2001);

Wooley v Maynard, 430 U.S. 705, 717 (1977); W. Va. State Bd.

of Educ. v. Barnette, 319 U.S. 624, 642 (1943).

Beyond Jordan, Judge Rendell relies on “principles of

equality and fairness” to justify a good faith defense. Rendell

Op. at III.B. But in full context, the Supreme Court made clear

that “principles of equality and fairness” were insufficient to

establish immunity:

Although principles of equality and fairness may

suggest . . . that private citizens who rely

unsuspectingly on state laws they did not create

and may have no reason to believe are invalid

should have some protection from liability, as do

their government counterparts, such interests are

not sufficiently similar to the traditional

purposes of qualified immunity to justify such an

expansion.

Wyatt, 504 U.S. at 168. Nothing about that quotation validates

“principles of equality and fairness” as standards for evaluating

potential affirmative defenses. As explained above, the

appropriate inquiry looks instead to the common law.

But even still, principles of equality and fairness would

not carry the day here. Neither equality nor fairness

overwhelmingly favors the reliance interests of the unions in

pre-existing law over the free speech rights of non-members

who were compelled to support the unions. The Supreme

Court in Janus already accounted for those reliance interests in

overturning Abood. See Janus, 138 S. Ct. at 2484-86; see also

10

Abood v. Detroit Bd. of Educ., 431 U.S. 209 (1977). Those

considerations need not be double-counted under the guise of

a good faith affirmative defense. And that is to say nothing of

the text, history, and purpose § 1983, which make it

particularly ill-suited to a construction that elevates reliance

interests over the vindication of constitutional rights.

***

Good faith was not firmly rooted as an affirmative

defense in the common law in 1871, and treating it as one is

inconsistent with the history and the purpose of § 1983. Nor

does our precedent or even principles of equality and fairness

favor recognition of good faith as an affirmative defense to a

compelled speech claim for wage garnishments. I respectfully

dissent and vote to reverse the orders dismissing the complaints

and to remand these cases.

11