Equal Employment Opportunity Commission v. Steamship Clerks Union, Local 1066

48 F.3d 594

• Court: Court of Appeals for the First Circuit
• Decided: February 28, 1995
• Docket: 94-1621, 94-1656
• Status: Published

Opinion

SELYA, Circuit Judge.

Labor unions have historically been instruments of solidarity, forged in an ostensible effort to counterbalance the weight of concentrated industrial power. It is, therefore, ironic — but not unprecedentedly so, inasmuch as “irony is no stranger to the law,” Amanullah v. Nelson, 811 F.2d 1, 17 (1st Cir.1987) — that unions themselves sometimes engage in exclusionary membership practices. The court below detected such an elitist strain in the operation of the Steamship Clerks Union, Local 1066 (the Union), determining that the Union’s policy requiring prospective members to be “sponsored” by existing members — all of whom, from time immemorial, have been white — constituted race-based discrimination. See EEOC v. Costello, 850 F.Supp. 74, 77 (D.Mass.1994).

In this venue, the Union ealumnizes both the district court’s evaluation of the sponsorship practice and the court’s remedial rulings. The Equal Employment Opportunity Commission (the EEOC), plaintiff below, cross-appeals, likewise voicing dissatisfaction with the court’s remedial rulings (albeit for very different reasons). Though we uphold the finding of disparate impact discrimination, we conclude that the lower court acted too rashly in fashioning remedies without pausing to solicit the parties’ views. Hence, we affirm in part, vacate in part, and remand for further proceedings.

I. BACKGROUND

The relevant facts are not, disputed. The Union is “a labor organization engaged in an industry affecting commerce,” 42 U.S.C. § 2000e(d)-(e) (1988). It has approximately 124 members, 80 of. whom are classified as active. The members serve as steamship clerks who, during the loading and unloading of vessels in the port of Boston, check cargo against inventory lists provided by shippers and consignees. The work is not taxing; it requires little in the way of particular skills.

On October 1, 1980, the Union formally adopted the membership sponsorship policy (the MSP) around which this suit revolves. The MSP provided that any applicant for membership in the Union (other than an injured longshoreman) had to be sponsored by an existing member in order for his application to be considered. The record reveals, without contradiction, that (1) the Union had no African-American or Hispanic members when it adopted the MSP; (2) blacks and Hispanics constituted from 8% to 27% of the relevant labor pool in the Boston area; (3) the Union welcomed at least 30 new members between 1980 and 1986, and then closed the membership rolls; (4) all the “sponsored” applicants during this period and, hence, all the new members, were Caucasian; and (5) every recruit was related to — usually the son or brother of — a Union member.

After conducting an investigation and instituting administrative proceedings, the EEOC brought suit on June 7, 1991, alleging that the Union had discriminated against African-Americans and Hispanics by means of the MSP. 1 The EEOC accused the Union of discrimination in violation of 42 U.S.C. § 2000e-2(c). 2 In addition, the EEOC charged that the Union had neglected to keep records (including so-called EEO-3 reports) in the manner required by law. 3

■After ample discovery, the EEOC moved for partial summary judgment, limiting its motion to the liability issues. The Union followed suit. On February 7, 1994, Judge Stearns held a hearing, reserved decision on the cross-motions, and extolled the virtues of settlement. Having planted the seed, the judge then proyided an opportunity for cultivation; he advised the parties that he would take no action for the time being and instructed them that, should no settlement eventuate within 30 days, he would thereafter render his decision. A month later, the Union informed Judge Stearns that settlement discussions had stalled. The EEOC, however, remained in a negotiating mode. On March 24, 1994, it mailed a letter to the court and the Union describing relief that it proposed for potential “inclusion in a consent decree.”

On the very same date, the district judge, presumably unaware of the EEOC’s letter, issued his decision. Judge Stearns granted the EEOC’s motion for partial summary judgment, holding that the. MSP evinced unlawful discrimination on the basis of race. See Costello, 850 F.Supp. at 77-78. He also granted the Union’s cross-motion for summary judgment on the record-keeping count. 4 See id.

Nothing significant occurred until April 10, 1994, when the court, without awaiting further motions or soliciting any input from the parties, entered final judgment. Among other things, it ordered the Union to (1) scrap the MSP; (2) open its membership “to enable admission of at least one new member for each listed member who, since the books were closed in 1986, has died, retired or [become inactive]”; (3) submit a plan for publicizing membership opportunities, taking special cognizance of the need to recruit minority applicants; (4) periodically submit membership information to the EEOC; and (5) comply with the EEOC’s record-keeping requirements, including the filing of EEO-3 reports. These appeals followed.

II. LIABILITY

We begin with the liability issue. The EEOC’s allegations against the Union find their genesis in Title VII of the Civil Rights Act of 1964, 42 U.S.C. §§ 2000e et seq. (1988). Broadly speaking, Title VII outlaws discrimination based on race, color, religion, gender, or national origin. In so doing, the law forbids both “overt discrimination” in the form of disparate treatment, Griggs v. Duke Power Co., 401 U.S. 424, 431, 91 S.Ct. 849, 853, 28 L.Ed.2d 158 (1971), and more subtle forms of discrimination, known as disparate impact discrimination, arising from “the consequences of employment practices, pot simply the motivation.” Id. at 432, 91 S.Ct. at 854. In this instance, we limit our inquiry to whether the court below support-ably determined that the MSP resulted in race-based disparate impact discrimination during the years 1980 through 1986.

A. The Disparate Impact Approach.

It has long been understood that discrimination, whether measured quantitatively or qualitatively, is not always a function of a pernicious motive or malign intent. Discrimination may also result from otherwise neutral policies and practices that, when actuated in real-life settings, operate to the distinct disadvantage of certain classes of individuals. See, e.g., John Hart Ely, Democracy and Distrust 84 (1980) (observing that technical enfranchisement, under certain conditions, has often fallen far short of actual enfranchisement). Within the world of Title VII, this understanding is reflected in the concept of disparate impact discrimination — a concept born of a perceived need to ensure that Title VII’s proscriptive sweep encompasses “not only overt discrimination but also practices that are fair in form, but discriminatory in operation.” Griggs, 401 U.S. at 431, 91 S.Ct. at 853. Thus, the disparate impact approach roots out “employment policies that are facially neutral in their treatment of different groups but that in fact fall more harshly on one group than another and cannot be justified by business necessity.” International Bhd. of Teamsters v. United States, 431 U.S. 324, 335 n. 15, 97 S.Ct. 1843, 1854 n. 15, 52 L.Ed.2d 396 (1977); accord Watson v. Fort Worth Bank & Trust, 487 U.S. 977, 987, 108 S.Ct. 2777, 2785, 101 L.Ed.2d 827 (1988) (explaining that “the necessary premise of the disparate impact approach is that some employment practices, adopted without a deliberately discriminatory motive, may in operation be functionally equivalent to intentional discrimination”). Beyond this abecedarian premise, however, the nature and allocation of the relevant burdens of proof must be clearly understood. 5

Under the legal framework that applies in this case, see supra note 5, it is incumbent upon the plaintiff to demonstrate a prima facie case of discrimination. See Albemarle Paper Co. v. Moody, 422 U.S. 405, 425, 95 S.Ct. 2362, 2375, 45 L.Ed.2d 280 (1975); McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802, 93 S.Ct. 1817, 1824, 36 L.Ed.2d 668 (1973); Johnson v. Allyn & Bacon, Inc., 731 F.2d 64, 69 (1st Cir.), cert. denied, 469 U.S. 1018, 105 S.Ct. 433, 83 L.Ed.2d 359 (1984). In the disparate impact milieu, the prima facie case consists of three elements: identification, impact, and causation. First, the plaintiff must identify the challenged employment practice or policy, and pinpoint the defendant’s use of it. See Wards Cove Packing Co. v. Atonio, 490 U.S. 642, 656, 109 S.Ct. 2115, 2124, 104 L.Ed.2d 733 (1989). 6 Second, the plaintiff must demonstrate a disparate impact on a group characteristic, such as race, that falls within the protective ambit of Title VII. See generally id. at 650-55, 109 S.Ct. at 2121-24. Third, the plaintiff must demonstrate a causal relationship between the identified practice and the disparate impact. See id. at 656-57, 109 S.Ct. at 2124-25; Watson, 487 U.S. at 994, 108 S.Ct. at 2788.

When the plaintiff rests, declaring herself satisfied that she has established a prima facie case of disparate impact discrimination, the ball bounces into the defendant’s court. At that point, the defendant has several options. First, it may attack the plaintiffs proof head-on, debunking its sufficiency or attempting to rebut it by adducing countervailing evidence addressed to one or more of the three constituent strands from which the prima facie case is woven, see Dothard v. Rawlinson, 433 U.S. 321, 331, 97 S.Ct. 2720, 2728, 53 L.Ed.2d 786 (1977), asserting, say, that no identifiable policy exists, or that the policy’s implementation produces no disparate impact, or that the plaintiffs empirical claims — such as the claim of causation — -are insupportable.

Alternatively, the defendant may confess and avoid, acknowledging the legal sufficiency of the prima facie case but endeavoring to show either that the challenged practice is job-related and consistent with business necessity, see Griggs, 401 U.S. at 431, 91 S.Ct. at 853; see also Albemarle Paper, 422 U.S. at 425, 95 S.Ct. at 2375, or that it fits within one or more of the explicit statutory exceptions covering bona fide seniority systems, veterans’ preferences, and the like. 7 See 42 U.S.C. §§ 2000e-2(h), 2000e-ll; see also 1 Charles A. Sullivan et al., Employment Discrimination. §§ 4.5-18 (2d ed. 1988). In all events, however, a defendant’s good faith is not a defense to a disparate impact claim. See Griggs, 401 U.S. at 432, 91 S.Ct. at 854 (holding that “good intent or absence of discriminatory intent does not redeem employment procedures or testing mechanisms that operate as ‘built-in headwinds’ for minority groups and are unrelated' to measuring job capability”).

If the defendant fails in its efforts to counter the plaintiffs prima facie case, then the factfinder is entitled — though not necessarily compelled, cf. St. Mary’s Honor Ctr. v. Hicks, — U.S. -, -, 113 S.Ct. 2742, 2748-50, 125 L.Ed.2d 407 (1993) — to enter judgment for the plaintiff. See, e.g., Cabrera v. Jakabovitz, 24 F.3d 372, 381 (2d Cir.), cert. denied, — U.S. -, 115 S.Ct. 205, 130 L.Ed.2d 135 (1994). On the other hand, even if the defendant stalemates the prima facie case by elucidating a legitimate, nondiscriminatory rationale for utilizing the challenged practice, the plaintiff may still prevail if she is able to establish that the professed rationale is pretextual. See Wards Cove, 490 U.S. at 658-59, 109 S.Ct. at 2125-26; Johnson, 731 F.2d at 69-70; see also McDonnell Douglas, 411 U.S. at 804, 93 S.Ct. at 1825. The plaintiff might demonstrate, for example, that some other practice, without a similarly undesirable side effect, was available and would have served the defendant’s legitimate interest equally well. See Wards Cove, 490 U.S. at 660-61, 109 S.Ct. at 2126; Johnson, 731 F.2d at 69-71. Such an exhibition constitutes competent evidence that the defendant was using the interdicted practice “merely as a ‘pretext’ for discrimination.” Albemarle Paper, 422 U.S. at 425, 95 S.Ct. at 2375 (quoting McDonnell Douglas, 411 U.S. at 804-05, 93 S.Ct. at 1825-26).

B. Standards of Review.

In general, summary judgment is proper only if, in the context of the motion and any opposition to it, no genuine issue of material fact exists and the movant has demonstrated its entitlement to judgment as a matter of law. See Fed.R.Civ.P. 56(c); see also National Amusements, Inc. v. Town of Dedham, 43 F.3d 731, 735 (1st Cir.1995). Hene.e, “a party seeking summary judgment [must] make a preliminary showing that no genuine issue of material fact exists. Once the movant has made this showing, the nonmovant must contradict the showing by pointing to specific facts demonstrating that there is, indeed, a trialworthy issue.” National Amusements, 43 F.3d at 735. An issue is “genuine” when the evidence relevant to it, “viewed in the light most flattering to the party opposing the motion, [is] sufficiently open-ended to permit a rational factfinder to resolve the issue in favor of either side.” Id. at 735 (citation omitted). Since the summary judgment standard requires the trial court to make a legal determination rather than to engage in differential factfinding, appellate review is plenary. See Garside v. Osco Drug, Inc., 895 F.2d 46, 48 (1st Cir. 1990).

Having recited the norm, we place it to one side, for certain unique aspects of the instant case dictate that we depart from the customary standard. The record discloses that, at the time the parties cross-moved for summary judgment, the Union voiced no disagreement with the facts on which the EEOC had constructed its case. 8 It gave no indication either that it intended to introduce any additional evidence or that any such evidence existed. To the exact contrary, the Union’s contentions centered entirely around the ultimate legal significance to be accorded to conceded facts. In effect, then, the parties submitted their dispute to the district court as a case stated.

Circuit precedent teaches that in such a situation — where, in a nonjury case, “the basic dispute between the parties concerns the factual inferences ... that one might draw from the more basic facts to which the parties have drawn the court’s attention,” where “[t]here are no significant disagreements about those basic facts,” and where neither party has “sought to introduce additional factual evidence or asked to present witnesses” — the district court is freed from the usual constraints that attend the adjudication of summary judgment motions. Federacion de Empleados del Tribunal Gen. de Justicia v. Torres, 747 F.2d 35, 36 (1st Cir.1984) (Breyer, J.). The court may then engage in a certain amount of differential factfinding, including the sifting of inferences. By the same token, the court of appeals may assume that “the parties considered the matter to have been submitted.below as a case ready for decision on the merits.” Id. Consequently, the standard for appellate oversight shifts from de novo review to dear-error review.- See id., (“Under these circumstances ... we should set aside the district court’s factual inferences only if they are ‘clearly erroneous.’ ”); see also United States v. Ven-Fuel, Inc., 758 F.2d 741, 744 n. 1 (1st Cir.1985) (stating in connection with a motion for summary judgment that when there are “no significant disagreements about the underlying facts,” and no indications that “any further factual evidence” might be available, the district court’s factual inferences should be set aside “only if they are clearly erroneous”) (citing other cases).

Based on these precedents, we are constrained to apply the more deferential clear-error standard when scrutinizing the inferences drawn by the court below. 9 Nonetheless, the court’s legal conclusions engender plenary review. See McCarthy v. Azure, 22 F.3d 351, 354 (1st Cir.1994).

C. Application of the Law.

In this case, the district court adroitly applied the substantive law and' concluded that the Union’s sponsorship-based membership policy constituted disparate impact discrimination. See Costello, 850 F.Supp. at 77. We descry no error.

1. The Prima Facie Case. We agree with the district court, see id. at 76-77, that the EEOC carried its burden of producing facts' sufficient to limn the three elements essential to its prima facie case. The first element — identification—requires no elabora tion. 10 We start, therefore, with the element of disparate impact and then move to causation. In both instances, the relevant facts are not disputed.

a.

Population statistics for the Boston area, proffered by the EEOC and unchallenged by the Union, show that in the relevant time frame African-Americans comprised 21%, and Hispanics 6%, of the available labor force. Although there are no known statistics on the racial composition of the steamship clerk industry — if such an “industry” exists — “Census Bureau statistics that merge the transportation industry’s employment statistics with similar statistics for public utilities ... show that blacks and Hispanics participate in the labor force as clerical/clerks at a rate of 7% and 1% of the total, respectively.” Id. at 77 n. 6. Despite the fact that the combined pool of potential black and Hispanic applicants for union membership ranged between 8% and 27% of the overall pool of potential applicants, no African-American or Hispanic was granted Union membership. Finally, during the MSP’s heyday — the six-year period from 1980 through 1986 — the Union admitted 30 new members. Based on a comparison of these figures with the profile of the newly minted Union members — 0 of 30, or zero pereént— the district court found that the EEOC adequately demonstrated a race-based disparate impact.

The Union is of a more skeptical mind. Although it does not challenge either the accuracy or the relevance of the underlying data, it contends that the small sample size renders the figures statistically insignificant, thus undercutting the EEOC’s attempt to establish a disparate impact. This contention is doubly flawed.

First, the contention misperceives the facts. While we appreciate that “small sample size may ... detract from the value of [statistical] evidence,” Teamsters, 431 U.S. at 339 n. 20, 97 S.Ct. at 1857 n. 20, a defendant who asserts that a plaintiffs prima facie case is insufficient must point out real deficiencies, not simply hurl epithets from behind gauzy generalizations. In particular, where, as here, a plaintiff has made out a colorable prima facie showing of discrimination, a challenger must do more than trumpet conclusory averments concerning the validity of the plaintiffs statistical foundation. See 1 Sullivan et al., supra, § 4.3.1, at 184 (explaining that a defendant must “attempt to undermine at least one element of the plaintiffs ease by bringing forth sufficient evidence to create a question of fact on that element”) (emphasis supplied). In this case, the Union proffered no such evidence.

Second, the Union’s contention misperceives the law. The cornerstone of its legal argument is our opinion in Fudge v. City of Prov. Fire Dep’t, 766 F.2d 650 (1st Cir.1985), and, yet, its point-by-point reliance on Fudge leaves much to be desired. While the Fudge court cautioned against the use of “an intuitive judicial judgment” as the sole basis for discerning a disparate impact, it carefully confined this admonition to cases “involving a claim that a screening test for admission to employment imposes a disparate and adverse impact” on a protected group. Id. at 657. Indeed, in a later case, not involving a screening test, we cited Fudge for the proposition that, in weighing the probative value of statistical evidence, “[e]ven small samples are not per se unacceptable.” Freeman v. Package Mach. Co., 865 F.2d 1331, 1342 n. 5 (1st Cir.1988). So it is here: because the EEOC’s claim does not involve an examination or other screening test, and because it nestles in a singularly compelling factual context, the Union’s repeated references to Fudge shed far more heat than light. 11

The utility of statistical evidence “depends on all of the surrounding facts and circumstances.” Teamsters, 431 U.S. at 340, 97 S.Ct. at 1857. In this instance, the sample, though small, is telling. Given the unique factual mosaic from which the statistical scaffolding hangs, and the logical force of the conclusion that the numbers suggest, it would blink reality to .conclude that a serious “sample size” problem lurks here. In our judgment, the lower court did not err in considering the available statistical evidence, and drawing founded inferences from it, en route to a disparate impact determination. See, e.g., United States v. Ironworkers Local 86, 443 F.2d 544, 551 (9th Cir.) (“On the basis that a showing of an absence or a small black union membership in a demographic area containing a substantial number of black workers raises an inference that the racial imbalance is the result of discrimination, the burden of going forward ... is shifted to the accused, for such a showing is enough to establish a prima facie case.”), cert. denied, 404 U.S. 984, 92 S.Ct. 447, 30 L.Ed.2d 367 (1971); accord United States v. United Bhd. of Carpenters & Joiners, 457 F.2d 210, 214. (7th Cir.), cert. denied 409 U.S. 851, 93 S.Ct. 63, 34 L.Ed.2d 94 (1972).

b.

Reluctant to raise a white flag, the Union further contends that, even if the EEOC established a significant racial disparity, its prima facie case misfired on the element of causation. The district court rejected this analysis. After reviewing the MSP and the evidence of disparate racial impact, it concluded that the former had caused the latter. See Costello, 850 F.Supp. at 77 (“Chance is not a likely explanation for this result.”). The pertinent question on review is whether the court erred in finding causation. We think not.

On this issue, the Union suggests three reasons why the court blundered, asseverating that the EEOC (1) did not identify particular African-Americans or Hispanics who unsuccessfully sought Union membership; (2) confused nepotism with race-based discrimination; and (3) failed to offer a suitably sophisticated statistical analysis, beyond a mere presentation of accumulated data. In the argot of the port, none of these arguments holds water.

As for the absence of identifiable minority applicants, the Union would have us rule that causation may be proven only by demonstrating that a flesh-and-blood African-American or Hispanic, who applied and was turned away, would have been admitted as a member but for the MSP. This isthmian view is a product of tunnel vision. The concept of causation under Title VII, like the larger concept of discrimination itself, is sometimes only discernible and inferable when viewed in context. See, e.g., Julia C. Lamber et al., The Relevance of Statistics to Prove Discrimination: A Typology, 34 Hastings L.J. 553, 553 (1983) (“Discrimination is difficult to define, observe, and prove.... [I]t may have no intrinsic meaning at all; rather, it acquires meaning in the context of a larger whole.”); see, also Teamsters, 431 U.S. at 340, 97 S.Ct. at 1856-57 (explaining that the value of statistical data depends on the totality of the surrounding circumstances). Here, the unvarnished reality of the situation — a sponsorship-based membership policy, enacted by an all-white union, and a- six-year track record of zero minority members despite 30 new white members, all of whom had family ties to existing members — renders the district court’s conclusion irresistible notwithstanding the lack of a specific unsuccessful minority applicant. 12

If bolstering is needed — and we do not believe that it is — we would add only that the cases and the commentators teach that evidence involving the rejection of actual applicants is not always necessary to prove causation. See, e.g., United States v. Sheet Metal Workers Int’l Ass’n, Local Union No. 36, 416 F.2d 123, 127 (8th Cir.1969) (holding, in an intentional discrimination case, that it was not “necessary for the government to prove that the Locals have refused membership” to actual black applicants); 1 Sullivan et ah, supra, §-4.3.1, at 186 (enunciating similar view). The Court’s opinion in Dothard bears stout witness to this principle. There, the plaintiff alleged that Alabama’s height and weight requirements for correctional counselors had a disparate impact on female applicants. In upholding the challenge, the Justices explicitly disavowed any rule “that a statistical showing of disproportionate impact must always be based on analysis of the characteristics of actual applicants.’-’ Dothard, 433 U.S. at 330, 97 S.Ct. at 2727 (citing Griggs, 401 U.S. at 430, 91 S.Ct. at 853). In. a passage that has marked relevance to the instant case, the Court reasoned that “[t]he application process itself might not adequately reflect the actual potential applicant pool, since otherwise qualified people might be discouraged from applying because of a self-recognized inability to meet the very standards challenged as being discriminatory.” Id. In this case, as in Dothard, we think that the court below could have inferred causation, despite the dearth of actual applicants, in part because the MSP would itself naturally have discouraged potential minority candidates.

The Union’s second asseveration need not detain us. Although the district court did not find a formal policy of nepotism, it recognized, as any thinking person must, that the MSP appeared to operate nepotistically. See Costello, 850 F.Supp. at 76 n. 4. The Union claims that this recognition betokens a confusion of two separate concepts: nepotism and discrimination. We do not agree. The history of the MSP’s actual implementation — an archive which reveals that every new member has been a relative of an existing member — is competent evidence on the element of causation. See Thomas v. Washington County Sch. Bd., 915 F.2d 922, 925 (4th Cir.1990) (explaining that “when the work force is predominantly white, nepotism and similar practices which operate to exclude outsiders may discriminate against minorities as effectively as any intentionally discriminatory policy”).

In mounting its third asseveration, the Union once again eschews any challenge to the EEOC’s basic data — the percentages of blacks and Hispanics in the relevant labor populations, as compared with the percentage of blacks and Hispanics on the Union’s membership roster — but, rather, impugns the EEOC’s failure to subject these proportionality data to some kind of formal statistical analysis. Although the Union’s frustration is understandable, its position that a prima facie case of disparate impact discrimination must invariably include a formal statistical analysis is untenable.

We say that the Union’s frustration is understandable because it would almost certainly have been helpful to the parties and to the court if the EEOC had processed its data in a slightly more sophisticated manner. Moreover, given its resources and institutional experience, the EEOC has no easily ascertainable excuse for neglecting this avenue. Nonetheless, though one would normally expect sound statistical analyses to assist a plaintiff in making out a prima facie case, see Lamber et al., supra, at 584-95, the absence of such analyses, by itself, does not automatically doom the plaintiff’s efforts. See, e.g., Ingram v. Madison Square Garden Ctr., Inc., 709 F.2d 807, 810-11 (2d Cir.) (affirming determination of Union’s liability under Title VII despite weak statistical evidence), cert. denied, 464 U.S. 937, 104 S.Ct. 346, 78 L.Ed.2d 313 (1983). To hold otherwise would effectively subordinate the whole of Title VII, in every last disparate impact case, to the sometimes vagarious sway of statistical proof.

In sum, it was not error for the lower court to conclude,, on the idiosyncratic facts of this ease, that the MSP, though neutral on its face, proximately caused the exclusion of minorities between 1980 and 1986. 13

2. The Union’s Response. Once the EEOC demonstrated a prima facie case of discrimination, the burden of production shifted. In the absence of any applicable statutory exemption, see supra note 7, it became incumbent upon the Union either to mount a satisfactory empirical rebuttal or to show that the challenged practice was job-related and consistent with business necessity. For all intents and purposes, the Union travels only the second path. Its sojourn is unavailing.

The Union suggests that the MSP is job-related and consistent with business necessity because it represents an important vehicle for continuing family traditions. Most of the 30 new members, according to the Union, “joined simply because their fathers had been members and because they wanted to maintain a family tradition_” We approach the task of evaluating this rationale mindful that the meaning and scope of the “business necessity” concept.are blurred at the edges. 14 • See 1 Sullivan et al., supra, § 4.3.2. In the-ease at bar, however, such potential-indeterminacy is of no consequence, for the Union’s “family tradition” thesis falls hopelessly short of limning a business necessity, and, thus, does not require us to explore terra incognita.

We will not tarry. Here, the Union has not shown even the glimmerings of a business necessity defense. Instead, it asks us to undertake a leap of faith. It makes absolutely no effort to explain, logically, why family tradition, and, thus,- the MSP, are necessary adjunets to carrying on the business of steamship clerks; and we, like the district court, can discern no essential connection. See Costello, 850 F.Supp. at 77 (concluding that the Union’s justification “does not explain, much less justify, the nexus between family tradition and the job of steamship clerk,” but “is merely an illumination of the motives of those who have had its advantage”). If courts were to accept an employer’s arbitrary ipse dixit as a satisfactory justification for retaining a policy that produces an invidiously discriminatory impact, Title VII would be reduced to no more than a toothless tiger. A policy that is neutral on its face, but that discriminates in fact, cannot elude the proscriptions of the law merely because its sponsor prefers to retain it. 15 See Wards Cove, 490 U.S. at 659,109 S.Ct. at 2126 (warning that courts must not “permit discrimination to be practiced through the use of spurious, seemingly neutral employment practices”).

The finish line looms. Because the Union neither rebutted the EEOC’s prima facie case nor articulated a legitimate, nondiserim-inatory justification for its membership policy, we uphold the grant of partial summary judgment in the EEOC’s favor.

III. RELIEF

The remedial rulings rest on a less even keel. Although the EEOC restricted its Rule 56 motion to the issue of liability, the district court, shortly after granting the mo tion, entered a judgment that awarded several items of permanent equitable relief. See supra p. 600. The court acted entirely on its own initiative, without convening a hearing and without affording the litigants any warning that it intended to resolve the matter of remediation.

Both parties appeal from this aspect of the judgment. The Union attacks on two fronts, assailing the district court for proceeding too fast and for venturing too far. In the first place, the Union asserts that the court flouted due process by vaulting to the remedial stage' without first putting the litigants on notice of its intentions and giving them an opportunity to be heard. In the second place, the Union denounces certain components of the injunction, especially the court’s command that the membership rolls be reopened. The EEOC, for its part, castigates the court for not proceeding far enough; it says'that backpay and mandated preferences to encourage minority membership should have been included in the compendium of relief.

Judicial dispensation of equitable remedies usually is reviewed for abuse of discretion. See Rosario-Torres v. Hernandez-Colon, 889 F,2d 314, 323 (1st Cir.1989) (en banc). Here, however, we need not consider the propriety of the remedies bestowed or withheld, for the district court’s failure to provide notice taints its remedial rulings and necessitates vacating virtually the entire relief-related portion of the judgment. 16

The question of whether notice is required is a question of law and is, therefore, subject to plenary review. See McCarthy, 22 F.3d at 354. We are in full agreement with the Second Circuit that “[n]o principle is more fundamental to our system of judicial administration than that a person is entitled to notice before adverse judicial action is taken against him.” Lugo v. Keane, 15 F.3d 29, 30 (2d Cir.1994). Examples abound. We, ourselves, have had occasion to address issues involving notice and its faithful companion, the opportunity to be heard, in a variety of contexts. See, e.g., Foster-Miller, Inc. v. Babcock & Wilcox Can., 46 F.3d 138, 148-49 (1st Cir.1995) (cautioning that, preparatory to deciding important issues, judges should strive to see that parties are given adequate notice and meaningful opportunities to be heard). We offer two illustrations.

First, while we have acknowledged that district courts possess the raw power to enter summary judgment sua sponte, we have repeatedly cautioned that this power must be “tempered by the need to ensure that the parties are given adequate notice to bring forward their evidence.” Stella v. Town of Tewksbury, 4 F.3d 53, 55 (1st Cir.1993); accord Jardines Bacata, Ltd. v. Diaz-Marquez, 878 F.2d 1555, 1560-61 (1st Cir.1989); Bonilla v. Nazario, 843 F.2d 34, 37 (1st Cir.1988). A second, very recent, example of our adherence to this principle can be found in Banks v. Shalala, 43 F.3d 11 (1st Cir.1994). There, we vacated the district court’s denial of Social Security disability benefits, not on the merits but because “the district court issued its affirmance [of the Secretary’s decision] before affording [the adversely affected party] an opportunity to submit argument explaining his objections to the Secretary’s de-termination_” Id. at 12. In taking that tack, we relied upon, and expressed our agreement with, the Fifth Circuit’s statement “that ‘district courts reviewing disability determinations should not conclude their review without an appropriate opportunity for the presentation of the parties’ contentions.’” Id. at 13 (quoting Flores v. Heckler, 755 F.2d 401, 403 (5th Cir.1985)).

The same principles also apply to and inform the dispensing of most types of equitable remedies. 17 Thus, absent exigent or other extraordinary circumstances- — and there are none reflected in this record — a court generally may not award equitable relief without first providing all affected parties actual notice that it is contemplating remedial action and affording them a meaningful chance to be heard. Nor does the fact that the judge enjoys broad discretion in shaping solutions relieve him from the obligation to afford procedural due process to all parties in interest. The rights of due process are constitutional and inviolable; hence, once a district court chooses to exercise its discretion, its conduct must comport with the promise of the Constitution. Discretion ensures the judge’s right to choose rather freely among plausible remedial options; it does not insulate him from listening to — or, at least, reading — the parties’ importunings.

We hold, therefore, that under ordinary circumstances litigants must be accorded fair opportunities to submit proposals for the judge’s consideration and to offer arguments in support of their positions before an award of equitable relief is made. We caution, however, that due process does not necessarily require any particular kind of hearing. See, e.g., In re Nineteen Appeals, 982 F.2d 603, 611 (1st Cir.1992) (noting that “in many, if not most, instances, due process does not require a full-scale trial, or even a hearing strictly conforming to the rules of evidence”); Domegan v. Fair, 859 F.2d 1059, 1065 (1st Cir.1988) (discussing district courts’ discretion to bypass oral argument); see generally Morrissey v. Brewer, 408 U.S. 471, 481, 92 S.Ct. 2593, 2600, 33 L.Ed.2d 484 (1972) (explaining that due process is a malleable concept, calling “for such procedural protections as the particular situation demands”). Accordingly, many matters can lawfully — and satisfactorily — be heard on the papers. See Aoude v. Mobil Oil Corp., 862 F.2d 890, 894 (1st Cir.1988); Cia. Petrolera Carite, Inc. v. Arco Caribbean; Inc., 754 F.2d 404, 411 (1st Cir.1985).

In the last analysis, whether any particular proceeding within any specific case warrants live arguments before the judge, as opposed to some other approach, is simply a function of the characteristics of the situation. “The test should be substantive: given the nature and circumstances of the case, did the parties have a fair opportunity to present relevant facts and arguments to the court, and to counter the opponent’s submissions?” Aoude, 862 F.2d at 894. In connection with this inquiry, one must bear in mind that litigants have no absolute right to present their arguments in whatever way they may prefer, or to expostulate for as long as they may choose. The inmates do not run the asylum. Thus, 'the trial judge has -broad authority to place reasonable limits on the parties’ presentation of their positions. See, e.g., United States v. Gleeson, 411 F.2d 1091, 1096 (10th Cir.1969).

This case, however, is about complete deprivation rather than the reasonableness of limits. The Union- received no notice that the court had begun to mull proposed remedial rulings. And all available indications were to the contrary: the EEOC’s motion for partial summary judgment, by its own terms, was “confined to questions of liability only and [did] not address relief’; the district court had not hinted at the hearing on the cross-motions for summary judgment that it intended to exceed the scope of the EEOC’s motion; 18 and the district court’s March 24 rescript sounded no warning bells. The Union, then, had no reason to marshal its arguments on relief-related issues, and no opportunity to make its case-to the decisionmaker.' Viewed in that light, the remedial rulings cannot endure. 19

IV. CONCLUSION

Our voyage is nearly complete. Having navigated the waters of Title VII, we now steer this case into the port of judgment and unload the cargo we have hauled. We affirm the district court’s grant of partial summary judgment in favor of the EEOC on its claim of disparate impact discrimination. The Union adopted a membership policy which, by its very nature, created a strong likelihood that no non-white face would ever appear in the Union’s ranks. Based on the evidence we have recounted, the EEOC established a prima facie case of discrimination. Because the Union failed either to rebut that case or to offer a legitimate, nondiseriminatory justification for maintaining the membership policy, the district court did not err in finding for the EEOC in respect to liability.

The court’s remedial rulings float in more turbulent seas. We agree with the Union that the district court’s gadarene rush to judgment deprived it of any meaningful opportunity to propose appropriate remedies or otherwise to participate in the formulation of a decree. Hence, we vacate the remedial rulings (save only for the exception previously mentioned, see supra note 16) and remand for further proceedings. 20 In so doing, we take no view as to what forms of relief, apart from barring continued use of the MSP itself, would, or would not, appear proper; and we specifically decline to address the parties’ substantive concerns as to the remedies granted and withheld.

Affirmed in part, vacated in part, and remanded. Costs in favor of plaintiff.

1 . The EEOC joined Bernard S. Costello, Inc. (Costello), a firm that regularly employed steamship clerks, as a codefendant. Costello is reportedly defunct, and, in any event, did not appeal from the entry of judgment below. Consequently, we treat the case as if the Union were the sole defendant.

2 .The statute provides in pertinent part:

It shall be an unlawful employment practice for a labor organization ... to exclude or eXpel from its membership, or otherwise discriminate against, any individual because of his race, color, religion, sex, or national origin.

42 U.S.C. § 2000e-2(c)(l) (1988). The district court found the Union to have practiced disparate impact discrimination in violation of this provision, and, therefore, did not consider the EEOC's parallel charge of intentional discrimination. See Costello, 850 F.Supp. at 76 n. 5. We emulate the district court's example.

3 . The operative statute obligates covered labor organizations, inter alia, to:

(1) make and keep such records relevant to the determinations of whether unlawful employment practices have been or are being committed, (2) preserve such records ..., and (3) make such records therefrom as the Commission shall prescribe by regulation or order....

42 U.S.C. § 2000e-8(c).

4 . The EEOC has not appealed from this portion of the judgment.

5 . The Civil Rights Act of 1991, Pub.L’. No. 102-166, 105 Stat. 1071 (1991), altered these burdens in some respects. See id. § 105 (codified at 42 U.S.C. § 2000e-2(k) (Supp. Ill 1991)) (specifically addressing the allocation and nature of burdens in disparate impact cases); see generally Rosemary Alito, Disparate Impact Discrimination Under the 1991 Civil Rights Act, 45 Rutgers L.Rev. 1011 (1993). Here, however, because the EEOC sued before the Act became law, the bog-gard of retroactive application hovers. See Rivers v. Roadway Express, Inc., — U.S. — , —, 114 S.Ct. 1510, 1519-20, 128 L.Ed.2d 274 (1994) (holding that § 101 of the Act is nonretro-active); Landsgraf v. USI Film Prods., — U.S. -, -, 114 S.Ct. 1483, 1508, 128 L.Ed.2d 229 (1994) (holding that § 102 is nonretroac-tive); see also Mozee v. American Commercial Marine Serv. Co., 963 F.2d 929, 932 (7th Cir.) (holding that the 1991 Act does not apply retroactively to a disparate impact claim), cert. denied, — U.S. -, 113 S.Ct. 207, 121 L.Ed.2d 148 (1992). We need not probe this point, for, although the Union alluded to the 1991 Act in its appellate brief, neither party sought to invoke it either in the court below or on appeal. It is, therefore, not properly before us. See United States v. Slade, 980 F.2d 27, 30 (1st Cir.1992); Clauson v. Smith, 823 F.2d 660, 666 (1st Cir. 1987) (collecting cases). Thus, our ensuing discussion reflects the legal framework as it existed without regard to the 1991 Act.

6 . While Congress passed the 1991 Act partly in an effort to nullify certain aspects of the Court's opinion in Wards Cove, see Landsgraf v. USI Film Prods., — U.S. -, -, 114 S.Ct. 1483, 1489, 128 L.Ed.2d 229 (1994), our reliance on Wards Cove is limited to portions of the opinion not affected by this legislative backlash.

7 . Because the Union has never suggested that the MSP comes within any such exception, we do not pursue this alternative.

8 . Of course, the mere fact that all parties move simultaneously for summary judgment neither unties the district court's hands nor renders the customary standard of review obsolete. Barring special circumstances, the nisi prius court must consider each motion separately, drawing inferences against each movant in turn, and the court of appeals must engage in de novo review. See El Dia, Inc. v. Hernandez Colon, 963 F.2d 488, 492 n. 4 (1st Cir. 1992); Griggs-Ryan v. Smith, 904 F.2d 112, 115 (1st Cir. 1990).

9 . Our conclusion concerning the applicable standard of review is reinforced by the Union's brief on appeal. In it, the Union neither promotes the conventions of Rule 56 nor asserts that the district court should have left the matter for trial, but, rather, argues that the court entered judgment for the wrong party because the EEOC failed to present a prima facie case; and, alternatively, that even if a prima facie case emerged, the Union successfully rebutted it. This scenario not only is consistent with the submission of the matter as a case stated but also amounts to a waiver of any contrary contention. See United States v. Zannino, 895 F.2d 1, 17 (1st Cir.) (explaining that theories which are not briefed or argued are waived), cert. denied, 494 U.S. 1082, 110 S.Ct. 1814, 108 L.Ed.2d 944 (1990).

10 . It is transparently clear that the EEOC singled out the MSP, identified it as the challenged employment practice, and linked it to the Union.

11 . Furthermore, even if we were to overlook these important distinctions and apply Fudge wholesale to the case at hand, we would endorse the trial court’s meticulous explanation of why a finding of disparate impact discrimination would still be appropriate. See Costello, 850 F.Supp. at 77 n. 7.

12 . None of the three cases brandished by the Union is to the contrary. Two of them — Johnson v. Uncle Ben’s, Inc., 965 F.2d 1363 (5th Cir. 1992), cert. denied, — U.S. -, 114 S.Ct. 1641, 128 L.Ed.2d 362 (1994), and Walls v. City of Petersburg, 895 F.2d 188 (4th Cir.1990) — are reminiscent of Fudge inasmuch as both involved challenges to an employment-related testing practice,, such as an examination, that had no inherently obvious tendency to discriminate against protected classes of individuals. Here, by contrast, we are dealing with a union consisting exclusively of whites which only accepts applicants who have already been sponsored by a member. Under such highly suggestive circumstances, far less additional proof is necessary to establish causation. Similarly, in EEOC v. Chicago Miniature Lamp Works, 947 F.2d 292 (7th Cir.1991), the court rejected the EEOC’s theory of causation because its statistical analysis totally omitted several key explanatory variables, thereby potentially skewing the results. See id. at 301. Here, by contrast, there is nothing to indicate any omitted variable or to cast doubt upon the apparent outcome.

13 . We add one further note. Though it is perhaps true; as the Union claims, that no court has ever invalidated a facially valid sponsorship-based membership policy under Title VII, it seems equally true that no such policy has ever been upheld. It would be a peculiar' rule of construction if a statute could not be applied in a certain manner unless it had already been applied in that manner in a previous case.

14 . The 1991 Act did little to sharpen the focus. See Note, The Civil Rights Act of 1991: The Business Necessity Standard, 106 Harv.L.Rev. 896, 903-06 (1993) ("On the issue of business necessity, the Act merely returns the courts to where they were just prior to Wards Cove, and appears to provide little guidance as to what direction they should take from there. The courts are saddled, instead, with a rich but uncertain legislative history arising from two years of complicated political maneuvering.”).

15 .The Union attempts to profit from the "family tradition" gambit in another way as well. Though offering no empirical rebuttal to the EEOC’s prima facie case, the Union posits that no African-Americans or Hispanics joined between 1980 and 1986 because of "the stark economic reality” of membership dues and the lack of any guaranteed employment. It then seeks to explain the 30 new recruits on the basis of family tradition. Although this twist, if believed, might conceivably furnish an' alternative theory of causation, it is unsupported by any cogent evidence, and, in all events, did not foreclose the district court from making a contrary, inference-based determination of causation.

16 . Of course, the district court plainly possessed the authority, without further proceedings, to order the Union to cease using the MSP. This portion of the decree may stand because it flows ineluctably from the court's finding of disparate impact discrimination. Hence, our comments and our instructions for vacatur are confined to the remainder of the equitable relief ordered sua sponte by the trial court.

17 . We exempt from this discussion provisional remedies, such as temporary restraining orders and ex parte attachments, which may from time to time be justified to preserve the status quo in a given case despite the absence of either notice or a predeprivation hearing. See, e.g., Fed.R.Civ.P. 65(b); Carroll v. President & Comm’rs of Princess Anne, 393 U.S. 175, 180, 89 S.Ct. 347, 351, 21 L.Ed.2d 325 (1968) (recognizing that "[t]here is a place in our jurisprudence for ex parte issuance, without notice, of temporary restraining orders of short duration").

18 . Indeed, the trial court stated then that the case was "not yet at the remedy stage.” In its rescript of March 24, 1994, the court noted explicitly that “[t]he EEOC is seeking a determination as to liability. It has not as yet suggested an appropriate remedy.” Costello, 850 .F.Supp. at 75 n. 2.

19 . In a vain attempt to salvage the court’s remedial rulings, the EEOC speculates that the Union was on constructive notice because the EEOC had requested injunctive relief in its complaint. Relatedly, it theorizes that a hearing was unnecessary because, no matter what the Union’s in put, injunctive relief was warranted. These arguments are jejune, and we reject them out of hand. The EEOC's insistence that its transmittal of March 24 placed the Union on notice that the court was pondering remediation is equally mer-itless. There is nothing about the EEOC's discussion of possible anodynes in the context of a proposed consent decree that would have alerted the most vigilant litigant to array its relief-related arguments, or risk preclusion.

20 . Notwithstanding the foregoing, the unap-pealed judgment in the Union's favor on the record-keeping count, see supra note 4 and accompanying text, endures.