Pleading Sets the Case: Summary Judgment Evidence Must Match the Disparate-Treatment Theory Plausibly Alleged in Union Referral Disputes

1. Introduction

Ronald King, Anthony Robinson, Chris Samuel, Nolan Jones, Jr., and Brian Struggs—five Black members of UA Local 91—sued the United Association of Journeymen and Apprentices of the Plumbing and Pipe Fitting Industry of the United States and Canada (UA) and Local 91 alleging race discrimination in union referral and leadership-placement practices. They asserted: (1) disparate-impact claims under Title VII; (2) disparate-treatment claims under Title VII and 42 U.S.C. § 1981 (including tangible employment actions and hostile environment); and (3) proposed class allegations on behalf of similarly situated Black workers.

The district court (Northern District of Alabama) dismissed class allegations at various points, granted judgment on the pleadings on certain theories, and granted summary judgment on the remaining claims. The Eleventh Circuit affirmed across the board, largely on the ground that the plaintiffs’ evidence and theories outstripped what they had actually (and plausibly) pleaded.

2. Summary of the Opinion

The Eleventh Circuit affirmed all challenged rulings. Its core holdings were:

  • Scope-of-pleadings constraint: At summary judgment, plaintiffs cannot rely on evidence supporting unpleaded or implausibly pleaded theories; the district court properly limited analysis to the specific disparate-treatment theory that was plausibly alleged.
  • No “new element” added: The district court did not modify McDonnell Douglas; it merely applied the plaintiffs’ pleaded theory (failure to obtain leadership roles on projects where plaintiffs were referred and worked, while whites were referred to leadership).
  • Disparate impact failed on causation/practice identification: Although statistics showed disparity, plaintiffs did not adequately identify and prove the specific neutral practice(s) causing it, as required by Supreme Court and circuit precedent.
  • Call-by-name referrals: Even where a “call-by-name” practice existed, the union’s role was ministerial—honoring employer requests—so the causal responsibility for any disparity lay with employers’ choices, not the union’s neutral procedure.
  • Hostile environment evidence insufficient: Confederate-flag exposure plus isolated or vague racial remarks did not meet the “severe or pervasive” standard.
  • Class claims affirmed under abandonment doctrine: Plaintiffs failed to challenge all independent grounds for dismissal on appeal, requiring affirmance.

3. Analysis

3.1 Precedents Cited

Pleading boundaries and litigation-stage discipline

  • Gilmour v. Gates, McDonald & Co.
    The court treated Gilmour as a hard boundary: a plaintiff who fails to plausibly plead allegations cannot effectively “add” them at summary judgment. This principle drove the outcome on disparate-treatment claims because much of plaintiffs’ evidence supported broader referral theories than the complaint plausibly alleged.
  • Ashcroft v. Iqbal and Bell Atl. Corp. v. Twombly
    These supplied the plausibility standard. The court used them not to newly tighten pleading burdens, but to justify why the complaint’s factual framing constrained what theories were “in the case” later.
  • Sears v. Roberts
    Cited for the summary judgment standard (de novo review; evidence viewed favorably to the nonmovant). Importantly, even under that favorable lens, only evidence relevant to a properly pleaded claim matters.

Title VII and § 1981 alignment; methods of proving intent

  • Berry v. Crestwood Healthcare LP
    Provided the analytic linkage: Title VII and § 1981 disparate-treatment claims “require the same proof and analytical framework,” letting the court analyze them together.
  • Tynes v. Fla. Dep't of Juv. Just.
    Used for the burden-shifting description and the centrality of intentional discrimination to disparate treatment. It also supported the court’s rejection of the argument that the district court improperly demanded statistical proof of intent—statistics were plaintiffs’ strategic choice.
  • McDonnell Douglas Corp. v. Green
    The opinion reemphasized that McDonnell Douglas is a flexible framework; what matters is pleading facts that reasonably infer intentional discrimination.
  • Lewis v. City of Union City
    Used to explain the “convincing mosaic” approach as another way to survive summary judgment—functionally a restatement of the general requirement to produce enough circumstantial evidence for a triable issue of discriminatory intent.

Union failure-to-refer doctrine (prima facie contours)

  • Mills v. Int'l Bhd. of Teamsters and Barber v. Int'l Bhd. of Boilermakers
    These anchored the court’s view that, in union referral contexts, a plaintiff generally must show he requested referral (often via being properly placed on the out-of-work list) and was treated less favorably than comparable white members. The court used them to explain why the complaint’s omission of clear “requested referral/out-of-work list” allegations would “normally” doom broader failure-to-refer claims—while still crediting a narrow inference for jobs the plaintiffs actually worked.

Judgment on the pleadings

  • Horsley v. Feldt
    Provided the de novo review framework for judgment on the pleadings and reinforced that claims beyond the plausibly pleaded theory were properly eliminated at the pleadings stage.

Disparate impact: identifying the practice and proving causation

  • EEOC v. Joe's Stone Crab, Inc.
    Supplied the circuit’s three-part disparate-impact test: statistical disparity, a neutral practice, and a causal relationship between the practice and the disparity. The court accepted plaintiffs’ showing on disparity but found failures on practice identification and causation.
  • Watson v. Fort Worth Bank & Trust
    The key constraint: even strong statistics do not relieve plaintiffs of the duty to “isolate and identify the specific employment practices” responsible for the disparity. The court used Watson to reject plaintiffs’ effort to let statistics stand in for proof that alleged practices (nepotism, ignoring posting, racial hostility, deviations from first-in/first-out, etc.) actually existed and caused the outcomes.
  • Tex. Dep't of Hous. & Cmty. Affs. v. Inclusive Communities Project, Inc.
    Cited for the “robust causality requirement” concept (in Fair Housing Act context) to underscore that disparate impact should not impose liability for disparities a defendant did not create—supporting the conclusion that “call-by-name” outcomes were attributable to employer requests, not the union’s ministerial honoring of them.

Hostile environment severity/pervasiveness

  • Adams v. Austal, U.S.A., L.L.C.
    The court used Adams as the comparator: examples where claims survived (daily slurs/graffiti, repeated “boy,” pervasive Confederate imagery) versus where they did not. It held plaintiffs’ evidence here did not meet that benchmark.
  • Yelling v. St. Vincent's Health Sys.
    Supported the proposition that vague, nonspecific testimony about racist comments is insufficient to show frequency/pervasiveness.
  • Melton v. I-10 Truck Ctr. Inc
    Used to reinforce that a single racist comment generally will not establish a hostile environment unless exceptionally severe.

Appellate abandonment and class allegations

  • Sapuppo v. Allstate Floridian Ins. Co.
    Controlled the class-claim affirmance: where a judgment rests on multiple independent grounds, the appellant must challenge each; failure to address one ground constitutes abandonment and requires affirmance.

3.2 Legal Reasoning

  1. The complaint defined the permissible disparate-treatment theory.
    The panel carefully parsed the Third Amended Complaint and concluded it plausibly pleaded disparate treatment only in a narrow scenario: plaintiffs were referred to and worked projects, but whites were referred for leadership positions on those same projects. Two pleading features were decisive: (a) the complaint did not clearly allege plaintiffs were on the out-of-work lists/requested referral for the disputed jobs, and (b) the complaint repeatedly framed the discriminatory act as whites being referred “for all [leadership] opportunities.”
  2. “No new McDonnell Douglas element”—just fidelity to pleadings.
    Plaintiffs argued the district court invented a requirement that a white foreman referral had to occur on a project where at least one plaintiff worked. The panel rejected this: that linkage was not a new legal element, but the factual theory the complaint itself advanced.
  3. Statistics alone did not establish disparate impact without a proven practice and causal nexus.
    The court credited plaintiffs’ statistical disparity evidence but found the record lacked proof that many alleged neutral practices existed as practices (or that they drove outcomes). The panel stressed that Watson requires identifying the specific practice responsible; broad allegations such as nepotism, failure to post foreman jobs, racial hostility, or “first-in/first-out” noncompliance cannot be inferred merely from disparity.
  4. Call-by-name: ministerial union action breaks the causation chain for union liability.
    Even assuming “call-by-name” existed, the union’s “shall honor” obligation made its role non-discretionary. The panel treated employer preference as the operative decision, echoing Inclusive Communities’ causality concerns—disparate impact should not impose liability on an actor who did not create the disparity.
  5. Hostile environment: Confederate-flag exposure plus isolated/vague incidents did not meet the objective standard.
    The panel applied the “severe or pervasive” test and found plaintiffs’ specific evidence thin: an isolated comment from 2001, “kind of racial jokes” without details, one “black ass” remark, and generalized claims of “punitive action” without specifics. Under Adams, Yelling, and Melton, this did not establish an objectively hostile environment. Notably, the court again limited itself to the legal sufficiency of the hostile-environment theory and expressly declined to opine on whether certain conduct might be actionable under other theories (e.g., retaliation).
  6. Class allegations: affirmance by abandonment.
    Applying Sapuppo, the panel affirmed dismissal because plaintiffs did not meaningfully challenge one independent ground: they lacked permission to reassert previously struck class allegations. A passing sentence about amendment was not enough to preserve the issue.

3.3 Impact

  • Litigation discipline in discrimination cases (especially union referral cases): The decision underscores that plaintiffs must align discovery, evidence, and summary judgment arguments with what was plausibly pleaded. Evidence suggesting broader discrimination will not salvage a narrow pleaded theory.
  • Pleading strategy for failure-to-refer claims: Plaintiffs should plead facts showing they requested referral (often via out-of-work lists), the union’s referral role, and clear comparators or a coherent “mosaic.” Vague assertions of eligibility and generalized discrimination allegations risk later being deemed outside the case.
  • Disparate impact claims require operational proof of “the practice”: Statistical disparity is necessary but not sufficient. Plaintiffs must identify and prove the existence of the neutral practice and show it caused the disparity. This is especially consequential where union procedures are constrained by collective bargaining or mandatory hiring-hall rules.
  • Ministerial compliance defenses: Where a union’s action is contractually mandatory (e.g., honoring employer “call-by-name” requests), plaintiffs may face a heightened causation challenge in assigning disparate-impact liability to the union rather than the employer.
  • Hostile environment threshold remains high and fact-specific: Confederate symbols and isolated slurs may be evidentially relevant, but plaintiffs must show severity or pervasiveness with concrete, specific facts.
  • Appellate briefing rigor for class issues: The class portion is a cautionary tale: failure to challenge every independent ground for dismissal results in automatic affirmance, regardless of merits.

4. Complex Concepts Simplified

  • Disparate treatment vs. disparate impact: Disparate treatment is intentional discrimination (someone is treated worse because of race). Disparate impact targets neutral rules that disproportionately harm a protected group, even without proven intent—so long as a specific neutral practice causes the disparity.
  • McDonnell Douglas framework: A step-by-step method for proving intent with circumstantial evidence: (1) plaintiff shows basic facts suggesting discrimination (prima facie case), (2) defendant offers a legitimate reason, (3) plaintiff shows that reason is pretext.
  • “Convincing mosaic”: Not a separate legal test, but a way of saying: does the total circumstantial evidence allow a reasonable jury to infer discriminatory intent?
  • Out-of-work list / first-in-first-out: In hiring-hall systems, workers often sign an out-of-work list and are referred in order. Exceptions may apply. If a plaintiff claims the union “skipped” him, he must account for exceptions and show the skip was discriminatory or violated the neutral rule as applied.
  • Call-by-name: An employer requests a specific worker by name. If the union must honor such requests, the union has little discretion—making it harder to prove the union caused the disparity.
  • Judgment on the pleadings vs. summary judgment: Judgment on the pleadings tests whether the complaint states a claim. Summary judgment tests whether evidence creates a genuine dispute for trial. This opinion highlights that summary judgment evidence cannot expand the pleaded claims.
  • Hostile environment (“severe or pervasive”): The law targets environments saturated with intimidation/insult. Isolated incidents usually are not enough unless extremely severe; vague recollections are typically insufficient.
  • Abandonment on appeal: If the trial court gave multiple independent reasons for a ruling, the appellant must challenge each one. Missing one means automatic affirmance.

5. Conclusion

Ronald King v. UA Local 91 is a procedural-and-substantive reminder that employment discrimination litigation is constrained by what is plausibly pleaded, not by the full universe of evidence a plaintiff might later develop. The Eleventh Circuit affirmed because: (1) the district court correctly limited disparate-treatment analysis to the narrowly pleaded “leadership referral on projects worked” theory; (2) the disparate-impact claims failed for lack of identified, proven practices and causation notwithstanding statistical disparities; (3) the hostile-environment record did not meet the “severe or pervasive” standard; and (4) class allegations were properly dismissed and, in any event, affirmance was compelled by appellate abandonment under Sapuppo v. Allstate Floridian Ins. Co..

The decision’s practical significance lies in its insistence on theory-to-pleading-to-proof alignment—particularly in union hiring-hall settings, where causation and discretion are often the pivotal fault lines.