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

March 10, 1995 UNITED STATES COURT OF APPEALS
UNITED STATES COURT OF APPEALS

FOR THE FIRST CIRCUIT
FOR THE FIRST CIRCUIT

No. 94-1621

EQUAL EMPLOYMENT OPPORTUNITY COMMISSION,
Plaintiff, Appellee,

v.

STEAMSHIP CLERKS UNION, LOCAL 1066,
Defendant, Appellant.

No. 94-1656

EQUAL EMPLOYMENT OPPORTUNITY COMMISSION,
Plaintiff, Appellant,

v.

STEAMSHIP CLERKS UNION, LOCAL 1066,
Defendant, Appellee.

ERRATA SHEET
ERRATA SHEET

The opinion of the court issued on February 28, 1995, is
corrected as follows:

Cover page, next-to-last line replace "Bladewood" with
"Blackwood"

On page 16, line 2 replace "Judge Coffin" with "it"

On page 26, line 17 delete "written" after "submit"

UNITED STATES COURT OF APPEALS
UNITED STATES COURT OF APPEALS
FOR THE FIRST CIRCUIT
FOR THE FIRST CIRCUIT

No. 94-1621

EQUAL EMPLOYMENT OPPORTUNITY COMMISSION,
Plaintiff, Appellee,

v.

STEAMSHIP CLERKS UNION, LOCAL 1066,
Defendant, Appellant.

No. 94-1656

EQUAL EMPLOYMENT OPPORTUNITY COMMISSION,
Plaintiff, Appellant,

v.

STEAMSHIP CLERKS UNION, LOCAL 1066,
Defendant, Appellee.

APPEALS FROM THE UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF MASSACHUSETTS

[Hon. Richard G. Stearns, U.S. District Judge]

Before

Selya, Boudin and Stahl, Circuit Judges.

Christopher N. Souris, with whom Thomas F. Birmingham and

Feinberg, Charnas & Birmingham were on brief, for Local 1066.

Paul D. Ramshaw, Attorney, with whom James R. Neely, Jr.,

Deputy General Counsel, Gwendolyn Young Reams, Associate General

Counsel, Vincent J. Blackwood, Assistant General Counsel, and

Lamont N. White, Attorney, were on brief, for EEOC.

February 28, 1995

SELYA, Circuit Judge. Labor unions have historically
SELYA, Circuit Judge.

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 calumnizes 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
I. BACKGROUND

The relevant facts are not disputed. The Union is "a

labor organization engaged in an industry affecting commerce," 42

3

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

1The 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

4

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 provided an opportunity for cultivation; he advised the

from the entry of judgment below. Consequently, we treat the
case as if the Union were the sole defendant.

2The 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)(1) (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.

3The 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).

5

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

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

6

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
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 (1971),

and more subtle forms of discrimination, known as disparate

impact discrimination, arising from "the consequences of

employment practices, not simply the motivation." Id. at 432.

In this instance, we limit our inquiry to whether the court below

supportably determined that the MSP resulted in race-based

disparate impact discrimination during the years 1980 through

1986.

A. The Disparate Impact Approach.
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

7

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. 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 (1977); accord Watson v. Fort Worth Bank &

Trust, 487 U.S. 977, 987 (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

5The 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. III 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 boggard of retroactive application
hovers. See Rivers v. Roadway Express, Inc., 114 S. Ct. 1510,

1519-20 (1994) (holding that 101 of the Act is nonretroactive);
Landsgraf v. USI Film Prods., 114 S. Ct. 1483, 1508 (1994)

8

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 (1975); McDonnell Douglas

Corp. v. Green, 411 U.S. 792, 802 (1973); Johnson v. Allyn &

Bacon, Inc., 731 F.2d 64, 69 (1st Cir.), cert. denied, 469 U.S.

1018 (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

(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.

Third, the plaintiff must demonstrate a causal relationship

between the identified practice and the disparate impact. See

(holding that 102 is nonretroactive); 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, 113 S. Ct. 207 (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.

6While 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., 114 S. Ct. 1483, 1489 (1994), our

reliance on Wards Cove is limited to portions of the opinion not

affected by this legislative backlash.

9

id. at 656-57; Watson, 487 U.S. at 994.

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 plaintiff's 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 (1977), asserting, say, that no identifiable

policy exists, or that the policy's implementation produces no

disparate impact, or that the plaintiff's 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; see also Albemarle Paper, 422 U.S. at 425, 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-11; see also 1

Charles A. Sullivan et al., Employment Discrimination 4.5-4.8

(2d ed. 1988). In all events, however, a defendant's good faith

is not a defense to a disparate impact claim. See Griggs, 401

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

10

U.S. at 432 (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

plaintiff's prima facie case, then the factfinder is entitled

though not necessarily compelled, cf. St. Mary's Honor Ctr. v.

Hicks, 113 S. Ct. 2742, 2748-50 (1993) to enter judgment for

the plaintiff. See, e.g., Cabrera v. Jakabovitz, 24 F.3d 372,

381 (2d Cir.), cert. denied, 115 S. Ct. 205 (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; Johnson, 731

F.2d at 69-70; see also McDonnell Douglas, 411 U.S. at 804. 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;

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 (quoting McDonnell Douglas, 411 U.S. at

804-05).

B. Standards of Review.
B. Standards of Review.

11

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,

F.3d , (1st Cir. 1995) [No. 94-1176, slip op. at 5].

Hence, "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, F.3d at [slip op. at 5]. 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 [slip op.

at 5-6] (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

12

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

8Of 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).

13

ready for decision on the merits." Id. Consequently, the

standard for appellate oversight shifts from de novo review to

clear-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.
C. Application of the Law.

In this case, the district court adroitly applied the

substantive law and concluded that the Union's sponsorship-based

9Our 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 U.S. 1, 17 (1st Cir.) (explaining that

theories which are not briefed or argued are waived), cert.

denied, 494 U.S. 1082 (1990).

14

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
1. The Prima Facie Case.

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 elaboration.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.
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

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

15

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 percent

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, a defendant who asserts that a plaintiff's 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 plaintiff's

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 plaintiff's case by bringing forth

sufficient evidence to create a question of fact on that

element") (emphasis supplied). In this case, the Union proffered

16

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. In this instance, the sample, though small, is telling.

Given the unique factual mosaic from which the statistical

11Furthermore, 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.

17

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 (1971); accord

United States v. United Bhd. of Carpenters & Joiners, 457 F.2d

210, 214 (7th Cir.), cert. denied, 409 U.S. 851 (1972).

b.
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

18

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 (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

19

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 al., 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

12None 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, 114 S. Ct. 1641 (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.

20

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

(citing Griggs, 401 U.S. at 430). 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

21

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),

22

cert. denied, 464 U.S. 937 (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 case, 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
2. The Union's Response.

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

13We 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.

23

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 case 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 adjuncts 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

14The 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.").

24

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 (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, nondiscriminatory justification for its membership

policy, we uphold the grant of partial summary judgment in the

EEOC's favor.

III. RELIEF
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 motion,

entered a judgment that awarded several items of permanent

equitable relief. See supra p. 5. 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.

15The 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.

25

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.

16Of 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.

26

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., F.3d , (1st Cir. 1995) [No. 94-

1498, slip op. at 21] (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, F.3d (1st Cir. 1994) [No. 94-1653]. 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

27

affording [the adversely affected party] an opportunity to submit

argument explaining his objections to the Secretary's

determination . . . ." Id. at [slip op. at 2]. 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 [slip op. at 6] (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

17We 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 (1968) (recognizing that "[t]here is a place in our
jurisprudence for ex parte issuance, without notice, of temporary

restraining orders of short duration").

28

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

(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

Caribe, 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

29

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

18Indeed, 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.

30

issues, and no opportunity to make its case to the decisionmaker.

Viewed in that light, the remedial rulings cannot endure.19

IV. CONCLUSION
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, nondiscriminatory 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

19In 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 input, 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 meritless. 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.

31

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
Affirmed in part, vacated in part, and remanded. Costs

in favor of plaintiff.
in favor of plaintiff.

20Notwithstanding the foregoing, the unappealed judgment in
the Union's favor on the record-keeping count, see supra note 4

and accompanying text, endures.

32