Ronald King v. UA Local 91

USCA11 Case: 24-13658 Document: 70-1 Date Filed: 08/10/2026 Page: 1 of 19

NOT FOR PUBLICATION In the

United States Court of Appeals

For the Eleventh Circuit ____________________ No. 24-13658 ____________________

RONALD KING, ANTHONY ROBINSON, CHRIS SAMUEL, NOLAN JONES, JR., BRIAN STRUGGS,

Plaintiffs-Appellants, versus UA LOCAL 91, United Association of Plumbers, Steamfitters, Welders and HVAC Techni- cians, UNITED ASSOCIATION OF JOURNEYMEN AND APPRENTICES OF THE PLUMBING AND PIPEFITTING INDUSTRY OF THE UNITED STATES AND CANADA, UNION OF PLUMBERS, FITTERS, WELDERS AND SERVICE TECHNICIANS,

Defendants-Appellees, DAY & ZIMMERMANN NPS, INC., et al.,

Defendants.

____________________

Appeal from the United States District Court for the Northern District of Alabama D.C. Docket No. 2:19-cv-01115-ACA ____________________

Before JORDAN, JILL PRYOR, and NEWSOM, Circuit Judges.

PER CURIAM:

This appeal involves race-discrimination claims brought by five black union members—Ronald King, Anthony Robinson, Chris Samuel, Nolan Jones, Jr., and Brian Struggs—against the United Association of Journeymen and Apprentices of the Plumb- ing and Pipe Fitting Industry of the United States and Canada (UA), and Local 91, an affiliate in Alabama. The plaintiffs brought dispar- ate-impact claims under Title VII of the Civil Rights Act of 1964, as well as disparate-treatment claims (related to both tangible em- ployment actions and a hostile work environment) under Title VII and 42 U.S.C. § 1981, and they sought to certify a class of similarly situated black workers. 1 At various points, the district court dis- missed the class allegations, ordered judgment on the pleadings in favor of the unions as to some claims, and granted the unions’ mo- tions for summary judgment on the remaining claims.

On appeal, the plaintiffs argue that each decision was incor- rect. After careful review and with the benefit of oral argument, we reach the opposite conclusion and affirm the district court.

I

The plaintiffs first argue that the district court erred in grant- ing summary judgment on the tangible-employment-action dispar- ate-treatment claims because it failed to consider certain evidence of racial discrimination. 2 For reasons we’ll explain, that evidence was beyond the scope of the claims that the plaintiffs had plausibly pleaded. The district court didn’t err by limiting its analysis to the claims properly before it.

A

If a plaintiff fails to plausibly plead allegations in his com- plaint, he may not raise those allegations at the summary judgment stage. See Gilmour v. Gates, McDonald & Co., 382 F.3d 1312, 1315 (11th Cir. 2004). The plausibility-pleading standard is familiar: “[A] complaint must contain sufficient factual matter, accepted as true, to ‘state a claim to relief that is plausible on its face.’” Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009) (quoting Bell Atl. Corp. v. Twombly, 550 U.S. 544, 570 (2007)). “A claim has facial plausibility when the plain- tiff pleads factual content that allows the court to draw the reason- able inference that the defendant is liable for the misconduct al- leged.” Id.

Title VII of the Civil Rights Act of 1964 prohibits employ- ment discrimination based on race, among other protected charac- teristics. As relevant here, the law prohibits labor organizations from failing or refusing to refer someone for employment based on his race. 42 U.S.C. § 2000e-2(c)(2). Relatedly, § 1981 bans race- based discrimination in contracting. Id. § 1981(a). Title VII and § 1981 “require the same proof and analytical framework,” so we analyze claims brought under the two statutes together. Berry v. Crestwood Healthcare LP, 84 F.4th 1300, 1307 (11th Cir. 2023). “To prove a claim under either statute, a plaintiff can use direct evi- dence, circumstantial evidence, or both.” Tynes v. Fla. Dep’t of Juv. Just., 88 F.4th 939, 944 (11th Cir. 2023). Here, the plaintiffs’ plead- ings were based on circumstantial evidence.

A plaintiff can establish a disparate-treatment violation based on circumstantial evidence in either of two ways. First, he can rely on the burden-shifting framework from the Supreme Court’s decision in McDonnell Douglas Corp. v. Green, 411 U.S. 792 (1973). Under McDonnell Douglas, the plaintiff must first make out a prima facie case of discrimination. Tynes, 88 F.4th at 944. What’s required to establish a prima facie case depends on the facts of the case. McDonnell Douglas, 411 U.S. at 802 n.13. In failure-to-refer cases, a plaintiff can state a prima facie case of discrimination if he shows that:

(1) [he] was a member of a group protected by Title VII; (2) [he] was qualified and requested referral for a job for which the union was referring applicants; (3) despite [his] qualifications, [he] was not referred; and (4) after [his] nonreferral, the union continued to refer applicants with plaintiff’s qualifications for available positions.

Mills v. Int’l Bhd. of Teamsters, 634 F.2d 282, 285 (5th Cir. Unit B 1981); see also Barber v. Int’l Bhd. of Boilermakers, 778 F.2d 750, 756 (11th Cir. 1985) (requiring a plaintiff to show “that he is the mem- ber of a minority, that he was properly on the union’s out-of-work list, and that he was not referred at the same wage rate as were white trainees”). As Mills and Barber show, even within the failure- to-refer context, the McDonnell Douglas test varies based on the facts of the case. But the gist is that a plaintiff must show that he is a member of a protected group, was qualified for and requested a job, and was treated less favorably than comparable non-minority union members.

Once a plaintiff makes out a prima facie case, the burden shifts to the defendant to “rebut[] that presumption . . . by offering evidence of a valid, non-discriminatory justification for the adverse employment action.” Tynes, 88 F.4th at 944. If the defendant can muster a justification, “the presumption of discrimination falls away” and the plaintiff must “persuad[e] the factfinder that [he] has been the victim of intentional discrimination.” Id. (citation modi- fied).

The McDonnell Douglas framework isn’t the only way to show disparate treatment. A plaintiff can also prevail if he “pre- sents a convincing mosaic of circumstantial evidence” that shows intentional discrimination. Lewis v. City of Union City, 934 F.3d 1169, 1185 (11th Cir. 2019) (citation modified). This “convincing mosaic” approach is really just an artful way to describe the normal summary judgment standard: A “plaintiff will always survive sum- mary judgment if he . . . presents circumstantial evidence that cre- ates a triable issue concerning the employer’s discriminatory in- tent.” Id.

In sum, to avoid dismissal on their disparate-treatment claims, the plaintiffs were required to plead facts supporting a rea- sonable inference that the union intentionally discriminated against them. They could do this either by alleging a prima facie case un- der McDonnell Douglas or by asserting facts that, when viewed as a “mosaic,” suggest discrimination.

B

Here, we agree with the district court that the plaintiffs plau- sibly pleaded disparate-treatment claims only in instances where Local 91 referred the plaintiffs for jobs and the plaintiffs worked those jobs, but the union referred white members for the leader- ship positions. This limitation on the claims stems from two ele- ments of the complaint.

First, the plaintiffs didn’t plead that they were on the out-of- work lists for the jobs at issue. The complaint only generally avers that the plaintiffs were eligible to act as foremen:

[Plaintiffs] were qualified for those positions, and that they were not referred and/or appointed to such po- sitions despite their qualifications even though Local 91 continued to seek candidates to fill vacancies in those positions that needed to be filled. . . . A prima facie inference of racial discrimination is established by such facts.

Third Am. Compl. ¶ 82, Dkt. No. 89.

Normally, this would have doomed the plaintiffs’ claims: Our precedent says that under McDonnell Douglas a prima facie case of failure-to-refer discrimination requires that the plaintiff re- quested referral. Mills, 634 F.2d at 285; Barber, 778 F.2d at 756. And the plaintiffs didn’t plead a mosaic of facts suggesting that the un- ion discriminated by failing to refer them when they didn’t request work. The district court, however—giving the plaintiffs the benefit of the doubt—inferred that the plaintiffs were on the out-of-work list for the jobs for which they were referred and worked. We think that this is a fair assumption—signing the out-of-work list, after all, is generally the first step of the referral process.

Second, the complaint’s allegations of referral-based dis- crimination were limited to instances in which white members were nominated for leadership positions. For example, Paragraph 25 stated that “[Local 91] referred . . . Caucasian union members for all [leadership] opportunities.” Third Am. Compl. ¶ 25 (empha- sis added). Similarly, Paragraph 44 asserted that “[e]ach [leader- ship] position and opportunity the Plaintiffs were interested in was given instead to a non-African American employee or Union mem- ber by Local 91.” Id. ¶ 44. There’s only one way to read these par- agraphs: In every instance of alleged discrimination, a white per- son was referred for a leadership position.

Putting these two elements of the complaint together, we agree with the district court that the plaintiffs plausibly pleaded dis- parate-treatment claims only based on scenarios where they were referred for and worked a job, but the union referred white mem- bers for the leadership positions.

The plaintiffs fault the district court for considering only ev- idence relevant to these claims. But the district court appropriately limited its summary judgment analysis, because the plaintiffs hadn’t plausibly pleaded other theories of liability. See Gilmour, 382 F.3d at 1315. The court didn’t err by considering only evidence rel- evant to the claims before it.

II

The plaintiffs next argue that the district court erred by re- quiring them to show, as part of their McDonnell Douglas prima facie case, that a white member was referred for a leadership position for a project on which at least one plaintiff was referred. The plaintiffs contend that the district court improperly “adopt[ed] a new McDon- nell Douglas element.” Br. of Appellants at 23. But this argument misunderstands the court’s decision. In its summary judgment or- der, the court reiterated that the plaintiffs plausibly pleaded dispar- ate-treatment claims only in situations where a white member, ra- ther than a plaintiff, was referred for a leadership position on a pro- ject for which the plaintiff was referred. The court didn’t add to McDonnell Douglas; it limited the case based on the plaintiffs’ plead- ings. Doing so was appropriate.

III Next, the plaintiffs argue that when all their evidence is con- sidered, genuine factual issues preclude summary judgment. But for reasons already discussed, much of the evidence offered at sum- mary judgment was beyond the scope of the claims that the plain- tiffs had plausibly pleaded. So that evidence could not create gen- uine factual issues relevant to the claims at issue. Perhaps that evi- dence would preclude summary judgment on other, unpleaded claims, but that question is irrelevant here.3 3 Regarding the district court’s summary judgment order, the plaintiffs also argue that the court erred by applying the law governing disparate-impact claims to their disparate-treatment claims. Not so. The district court stated simply that the plaintiffs’ statistical evidence—which didn’t establish a causal link between the unions’ practices and a disparate outcome—failed to prove that the unions intentionally discriminated against the plaintiffs. Intentional discrimination is the core of a disparate-treatment claim. See Tynes, 88 F.4th at 944. The district court also did not err by “requiring statistical proof of dis- parate treatment.” See Br. of Appellants at 40. Statistical analysis of the de- fendants’ employment practices was a strategic decision made by the plaintiffs, not one imposed by the district court.

IV

The plaintiffs next turn to the district court’s grant of judg- ment on the pleadings in favor of the unions. 4 In its order, the court addressed claims predicated on four scenarios:

(1) referrals where Local 91 did not designate anyone as its nominee for foreman; (2) any contractors’ selec- tions of someone other than Plaintiffs as a foreman where the employee who made the hiring decision was a member of Local 91; (3) any contractors’ selec- tion of foremen without seeking a referral for the foreman position from Local 91; and (4) referrals to job locations not covered by the [Southern Company Maintenance and Modification Agreement].

Order Granting Mot. J. Pleadings 4, Dkt. No. 220. We agree with the district court that these claims weren’t properly pleaded.

As an initial matter, the plaintiffs contend that the allegations discussed in the order are not claims, but mere “factual scenarios.” Br. of Appellants at 28. But whether the allegations are separate claims or merely factual scenarios makes no difference because, as already stated, the plaintiffs adequately pleaded only claims based on the unions’ failure to refer them for leadership positions when 4 “We review de novo a district court’s entry of judgment on the pleadings.” Horsley v. Feldt, 304 F.3d 1125, 1131 (11th Cir. 2002). We apply the same ana- lytical framework as when reviewing a district court’s dismissal of claims, ac- cepting the facts in the complaint as true and granting judgment to the defend- ant “only when the plaintiff can prove no set of facts in support of his claim which would entitle him to relief.” Id. (citation modified). they were referred for work on a project and a white member was instead referred for a leadership role. The allegations at issue in the judgment-on-the-pleadings order were beyond that category of claims. So they weren’t plausibly pleaded, and judgment on the pleadings was appropriate.

V

Next, the plaintiffs argue that the district court erred in granting summary judgment to the unions on their disparate-im- pact claims. Title VII prohibits not only intentional discrimination; it also bars “neutral employment practices which, while non-dis- criminatory on their face, visit an adverse, disproportionate impact on a statutorily-protected group.” EEOC v. Joe’s Stone Crab, Inc., 220 F.3d 1263, 1274 (11th Cir. 2000). To succeed on a disparate-impact claim, a plaintiff must show three things: (1) a significant statistical disparity, (2) a facially neutral employment practice that allegedly causes the disparity, and (3) a causal relationship “between the spe- cific employment practice identified and the statistical disparity shown.” Id.

The plaintiffs allege that numerous practices cause a dispar- ate impact. They point to “the consideration of experience and nepotism when making referrals, failing to announce and post fore- man positions, and racial hostility,” Br. of Appellants at 32 (quoting Mem. Op. Granting Defs.’ Mot. Summ. J. 10, Dkt. No. 362), as well as use of “a ‘call-by-name’ criterion,” id. at 36, overrepresenting black union members in out-of-jurisdiction positions, id. at 37, and “Local 91’s non-compliance with the mandatory ‘first-in/first-out’ criterion,” id. at 38.

We conclude that although the plaintiffs have ticked the first disparate-impact box—they’ve produced statistics showing a signif- icant racial disparity in foreman referral rates—they’ve failed to identify evidence suggesting that the alleged discriminatory prac- tices existed or caused the disparity. The plaintiffs rely almost ex- clusively on statistics to substantiate their disparate-impact claim. And, relying on Watson v. Fort Worth Bank & Trust, 487 U.S. 977 (1988), they contend that statistical disparities, if large enough, can support an inference of causation. The problem, however, is that under Watson plaintiffs remain “responsible for isolating and iden- tifying the specific employment practices that are allegedly respon- sible for any observed statistical disparities.” Id. at 994. The plain- tiffs haven’t presented evidence that Local 91 maintains practices of considering prior experience, engaging in nepotism, failing to an- nounce and post foreman positions, acting with racial hostility, sys- tematically selecting employees for out-of-jurisdiction roles, and failing to comply with the “first-in/first-out” rule. So their dispar- ate-impact claims related to these “practices” fail. 5 5 One of the plaintiffs’ attempts to show a practice’s existence is worth discuss- ing in greater detail, because the plaintiffs produce evidence that at first glance seems probative. To substantiate their allegation that Local 91 maintained a practice of ignoring the first-in/first-out rule, the plaintiffs point to the decla- ration of Karen Allen, which identifies instances where people who either were not on the out-of-work list or occupied a lower position on the list were referred instead of higher-ranked plaintiffs. But the plaintiffs don’t address the

We think that there’s one practice that the plaintiffs have es- tablished: referring members whom D&Z “called by name”—i.e., specifically requested. Br. of Appellants at 36 (citation modified). Under Local 91’s hiring hall procedures, “[r]equests . . . for key men to act as supervisors, general foremen or foremen, shall be honored without regard to the requested man’s place on the out-of-work list.” Local 91’s Mot. Summ. J. Ex. 4, at 2, Dkt. No. 275-4. The problem for the plaintiffs, however, is that Local 91’s role in call-by- name referrals is purely ministerial; the union “shall” honor em- ployers’ requests for specific foremen. Id. (emphasis added). To the extent that call-by-name referrals produce racially disparate out- comes, the culprit is employers’ decisions to request certain people, not the mere existence of the call-by-name procedure. Cf. Tex. Dep’t of Hous. & Cmty. Affs. v. Inclusive Communities Project, Inc., 576 U.S. 519, 542 (2015) (emphasizing—albeit in the Fair Housing Act context—that disparate impact’s “robust causality requirement . . . protects defendants from being held liable for racial disparities they did not create.”). fact that the first-in/first-out rule is subject to multiple exceptions. Nor does the Allen Declaration. So the Allen Declaration doesn’t create a genuine fac- tual issue regarding Local 91’s compliance with the first-in/first-out proce- dure.

VI

The plaintiffs next assert that the district court erred by granting summary judgment for the unions on the plaintiffs’ hos- tile-union-environment claims.6 We agree with the district court that the plaintiffs failed to present evidence suggesting that they faced a hostile environment.

To succeed on a hostile-environment claim, a plaintiff must establish that his environment is “permeated with discriminatory intimidation, ridicule, and insult, that is [] severe or pervasive.” Ad- ams v. Austal, U.S.A., L.L.C., 754 F.3d 1240, 1248 (11th Cir. 2014) (ci- tation modified). The plaintiff’s environment must be “both sub- jectively and objectively hostile.” Id. at 1249.

Adams provides useful guideposts for determining whether racial harassment creates a hostile environment. There, we held that a hostile-environment claim survived summary judgment based on the following facts: A black employee heard racial slurs and saw racist graffiti in the bathroom every day, he periodically saw racist graffiti elsewhere, his white supervisor called him “boy” several times, his coworkers wore clothes with Confederate flags every day, and he heard that a white coworker called a black coworker a “black bitch.” Id. at 1252. By contrast, a hostile-envi- 6 Our circuit has not decided whether Title VII provides a cause of action based on a hostile union environment. For the sake of argument, we assume that hostile-union-environment claims are cognizable and subject to the standard we apply to hostile-work-environment claims. ronment claim didn’t survive summary judgment based on the fol- lowing record: A black employee regularly saw coworkers wear Confederate flag apparel and saw racist graffiti in the bathroom daily, heard people say the N-word a few times over two years, and heard about a noose in the breakroom. Id. at 1254.

Here, none of the plaintiffs have pointed to evidence creat- ing a genuine factual issue as to whether they faced an objectively hostile union environment. Two facts are common to all plaintiffs: They often saw the Confederate flag at the union hall, and the un- ion refused to take down the flag even after people complained about it. We address the plaintiff-specific facts next.

Robinson

In 2001, after a black union member expressed concerns about the flag, Robinson heard a white member tell the black mem- ber that “he would be removed before the flag would be removed.” However objectionable, exposure to the Confederate flag, plus an isolated, decades-old racist comment, falls short of the Adams standard for a hostile environment.

Strus

Struggs sometimes heard union members make “kind of ra- cial jokes.” He, however, couldn’t remember the content of the jokes or who made them. Isolated, unspecified, “kind of racial” comments—again, however objectionable—don’t constitute se- vere and pervasive harassment. See Yelling v. St. Vincent’s Health Sys., 82 F.4th 1329, 1335 (11th Cir. 2023) (holding that a plaintiff’s testi- mony that her workplace “became ‘kind of heated’ with racist comments” and that “her coworkers generally made racist com- ments multiple times” wasn’t specific enough to show frequent harassment). And regularly seeing the Confederate flag doesn’t push the environment over the line. See Adams, 754 F.3d at 1254.

Jones Jones heard about the 2001 incident where the white mem- ber told the black member that he would be removed before the Confederate flag, but he didn’t personally observe it. Jones also saw pictures of the flag on the union’s Facebook page. For reasons al- ready explained, exposure to the flag, plus hearing secondhand about an isolated racist incident, doesn’t support a hostile-environ- ment claim.

King

Taking the facts in the light most favorable to King, he was suspended from apprentice school in 2012 because he complained about the Confederate flag. King also saw a picture of the flag on the union’s Facebook page. These two isolated incidents, plus ex- posure to the flag at the union hall, don’t constitute the kind of “permeat[ing]” harassment that can support a hostile-environment claim. See Adams, 754 F.3d at 1248 (citation modified).7 7 We take no position on whether the allegedly retaliatory suspension could be actionable under some other theory of liability.

Samuel

On one occasion, Samuel was told by a white union leader that he was being insubordinate and that his “black ass” didn’t get to tell the leader how to run the union hall. Samuel also states that he faced racially derogatory remarks while an apprentice and suf- fered “punitive action . . . for expressing his concerns relating to the apprentice school,” although he fails to provide details about either. Ultimately, Samuel points only to one concrete example of racial harassment at the union hall: the “black ass” remark. Under our precedent, though, “one instance of a racist comment does not it- self render a workplace hostile unless the comment is sufficiently severe to make up for the absence of pervasiveness.” Melton v. I-10 Truck Ctr. Inc, 166 F.4th 905, 918 (11th Cir. 2026). The union leader’s offensive remark doesn’t rise to that level. And Samuel’s vague al- lusion to other derogatory comments and “punitive action” doesn’t substantiate his hostile-environment claim. Cf. Yelling, 82 F.4th at 1335 (discounting a plaintiff’s vague testimony about the harass- ment she faced).

* * *

In sum, no plaintiff has established a genuine issue of mate- rial fact as to whether he faced an objectively hostile union envi- ronment, so the district court properly granted summary judg- ment for the unions.

VII

Lastly, the plaintiffs argue that the district court erred by dis- missing their class claims. The plaintiffs, however, failed to chal- lenge all the district court’s grounds for dismissing the claims, so we affirm the court’s dismissal.

“To obtain reversal of a district court judgment that is based on multiple, independent grounds, an appellant must convince us that every stated ground for the judgment against him is incorrect.” Sapuppo v. Allstate Floridian Ins. Co., 739 F.3d 678, 680 (11th Cir. 2014). “When an appellant fails to challenge properly on appeal one of the grounds on which the district court based its judgment, he is deemed to have abandoned any challenge of that ground, and it follows that the judgment is due to be affirmed.” Id.

That’s what happened here. In its order, the district court laid out two reasons for dismissing the class allegations. First, the court stated that the plaintiffs didn’t have permission to reallege the class claims that had previously been struck. Second, the court con- cluded that the class claims failed on the merits because they were substantively the same as the claims the court previously dismissed. In their initial brief, the plaintiffs don’t address the court’s first ra- tionale for dismissing their claims: that they lacked permission to reallege them. See Br. of Appellants at 45–52. The portion of their brief discussing the class allegations has only one sentence that could be relevant to this issue: The plaintiffs state that “[t]he dis- trict court erred in not allowing at least one opportunity to amend the class certification allegations.” Id. at 46. This “passing refer- ence[]” to the issue is insufficient to raise it on appeal. See Sapuppo, 739 F.3d at 681. So the plaintiffs have abandoned any challenge to the first rationale, and we therefore must affirm the district court’s dismissal order.

* * *

In sum, the district court’s various dismissal, judgment-on- the-pleadings, and summary judgment orders were correct.

AFFIRMED.