League of United Latin American Citizens (LULAC), Council #4434 v. Judger Sharolyn Wood and Judge F. Harold Entz Supplemental Brief on Rehearing En Banc of Plaintiff-Intervenor-Appellees

    IN THE
UNITED STATES COURT OF APPEALS 

FOR THE FIFTH CIRCUIT

No. 90-8014

LEAGUE OF UNITED LATIN AMERICAN 
CITIZENS (LULAC), COUNCIL #4434, 
et al.,

Plain ti ffs-Appell ees,

HOUSTON LAWYERS’ ASSOCIATION, 
et al.,

Plaintiff-Intervenor-
Appellees,

V.
DAN MORALES, et al.,

State Defendants- 
Appellants,

JUDGE SHAROLYN WOOD AND 
JUDGE F. HAROLD ENTZ,

Defendant-Intervenor-
Appellant.

Appeal from the United States District Court for the Western 
District of Texas, Midland-Odessa Division

SUPPLEMENTAL BRIEF ON REHEARING EN BANC 
OF PLAINTIFF-INTERVENOR-APPELLEES 
HOUSTON LAWYERS’ ASSOCIATION, etal.

ELAINE R. JONES 
CHARLES STEPHEN RALSTON 
SHERRILYN A. IFILL

99 Hudson Street, 16th Floor 
New York, New York 10013 
(212) 219-1900

GABRIELLE K. McDONALD, ESQ. 
Walker & Satterthwaite 
7800 N. Mopac, Suite 215 
Austin, Texas 78759 
(512) 346-6801



TABLE OF CONTENTS
Introductory S ta te m e n t ....................................................................................................................... 1

Summary of Argument ....................................................................................................................... 2

A rg u m e n t................................................................................................................................................. 3

I. The Standards of Review Advanced by the Dissent are Incompatible with 
the Guidelines for Assessing Vote Dilution Claims Set O ut by Congress
and the Supreme Court, and Followed by this C i r c u i t ............................................. 3

A. Plaintiffs in Section 2 Cases Need Not Show that W hite 
Voters are Racist in O rder to Prove that Voting is Racially
Polarized ..................................................................................................................  4

B. Gingles Does Not Require that Plaintiffs Prove the Cause of
Racially Polarized Voting ...................................................................................  8

C. No O ther Court Has Required that Plaintiffs Prove the Cause of
Racially Polarized Voting ................................................................................. 10

II. The Legitimacy of the State’s Interests is Not a Defense to Proven Vote
D ilu t io n ...............................................................................................................................  11

III. The State’s Interest Should be W eighed in the Totality of Circumstances . . .  13

IV. By Any Standard, Plaintiffs Proved Their Case in Harris C o u n ty .........................  18

A. None of the Purportedly Non-Racial Rationale Offered by the 
Defendants In Fact Explain Racially Polarized Voting in H arris
C o u n ty .....................................................................................................................  18

B. The Houston Lawyers’ Association Proved the Existence of Vote
Dilution in Harris C o u n ty ................................................................................. 20

C. The State H as Never, In Fact, Proved its Interest in "Linkage" ........... 23

Conclusion 26



TABLE OF AUTHORITIES

Cases: Pages:

Bazemore v. Friday,
478 U.S. 385 (1986) ..........................................................................................................  9

Bradley v. Swearingen,
525 S.W.2d 280 (Tex. Civ. App. 1975) ......................................................................  25

Campos v. City of Baytown,
840 F.2d 1240 (5th Cir. 1988) .........................................................................  10, 11, 21

Citizens for a B etter G retna v. City of Gretna,
636 F. Supp. 1113 (E.D. La. 1986), affd , 834
F.2d 496 (5th Cir. 1987) ............................................................................................. 11

Dillard v. Crenshaw County,
831 F.2d 246 (11th Cir. 1986) ......................................................................................  24

Gingles v. Edmisten,
590 F. Supp. 345 (E.D.N.C. 1984) .........................................................................passim

Growe v. Emison,
61 U.S.L.W. 4163 (February 23, 1993) ......................................................................  18

HLA v. A ttorney General of Texas,
115 L. Ed. 2d 379 (1 9 9 1 ) .........................................................................................passim

Jeffers v. Clinton,
730 F. Supp. 196 (E.D. Ark. 1989) affd  mem.,
112 L. Ed. 2d 656 (1 9 9 1 ) ................................................................................................  10

Kirksey v. Bd. of Sup’rs Hinds County, Miss.,
554 F.2d 139 (5th Cir. 1977) ........................................................................................  15

Kirksey v. City of Jackson, Miss.,
663 F.2d 659 (5th Cir. 1981) reh’g and reh’g en
banc denied, 669 F.2d 316 (5th Cir. 1982) .................................................................  6

LULAC v. Clements,
914 F.2d 620 (5th Cir. 1990) ...................................................................................  12, 23

li



Pages

LULAC v. Clements,
914 F.2d 620 (5th Cir. 1990) ...................................................................................  12, 23

LULAC v. Midland ISD,
812 F.2d 1494 (5th Cir. 1987), vacated on other
grounds, 829 F.2d 546 (5th Cir. 1987) ................................................................. passim

Mallory v. Eyrich,
707 F. Supp. 947 (S.D. Ohio 1989) ...........................................................................  10

M obile v. City of Bolden,
446 U.S. 55 (1980) .............................................................................................................  4

Overton v. City of Austin,
871 F.2d 529 (5th Cir. 1989) ........................................................................................  10

Sanchez v. Bond,
875 F.2d 1488 (10th Cir. 1 9 8 9 )...................................................................................... 10

Solomon v. Liberty County, Florida,
899 F.2d 1012 (11th Cir. 1 9 9 0 )...................................................................................... 10

Sweatt v. Painter,
339 U.S. 629 (1950) ................................................................................................... 22, 23

Thornburg v. Gingles,
478 U.S. 30 (1986) ....................................................................................................... passim

Westwego Citizens for a Better Gov’t v. City of 
Westwego, 872 F.2d 1201 (5th Cir. 1989), on remand, 946

F.2d 1109 (5th Cir. 1991) ........................................................................................  21, 24

Statutes: Pages:

Senate R eport No. 97-417, 97th Cong., 2nd Sess. ( 1 9 8 2 ) .......................................... passim

42 U.S.C. § 1973, as amended ............................................................................................... passim

iii



Introductory Statem ent

The interested persons are identical to those listed in the Houston Lawyers’ 

Association’s October 1991 Brief on Remand.

Plaintiff-intervenors incorporate by reference the Statem ent of Jurisdiction, 

Statem ent of Facts, and Statem ent of the Case contained in their O ctober 1991 Brief on 

Remand.

1



Summary o f Argument

This supplem ental brief addresses three core arguments advanced by the defendant- 

intervenors and by Judge Higginbotham in his dissent from the recently vacated panel 

opinion1 as bases for denying relief to plaintiffs in this case2: (1) that plaintiffs must show 

that the cause of racially polarized voting patterns is racism on the part of white voters; (2) 

that the existence of some non-racial state interest may, standing alone, rebut a showing 

of vote dilution, and (3) that the state of Texas has proven that it has a "substantial" 

interest in linking the electoral and jurisdictional base for judges. All three arguments rest 

on fundam ental misreadings of the Voting Rights Act, its legislative history, and relevant 

Supreme Court precedent. But even if this Court were to adopt these unsupportable

'On January 27,1993, in LULAC v. Attorney General of Texas, a panel of this court affirmed the 
district court’s finding of vote dilution in eight of the nine counties at issue in this case. As stated 
by a majority of the panel, the record in this case supports the inescapable factual reality "that in 
Texas district court elections minority voters have less opportunity than white voters to participate 
in the political process and to elect representatives of their choice." Slip Op. at 8. The panel 
majority ("the majority") in a carefully reasoned, thorough analysis, describes the nature of the vote 
dilution which it found to be present in the eight relevant counties. Plaintiff-intervenors endorse 
the reasoning, analysis, and interpretation of the law articulated in that opinion.

References to the majority panel opinion of January 27,1993 will be to "Panel Op. a t____."
References to Judge Higginbotham’s dissent and proposed opinion will be to "Dissent a t ___."

2The Houston Lawyers’ Assocation has filed three briefs on appeal in this case. The first 
Brief on Appeal filed on February 27, 1990, we discussed the factual evidence in the record which 
supports the district’s court’s finding that district judge elections in Harris County violate Section
2 of the Voting Rights Act, as amended. In our Supplemental Brief to this court en banc, filed on 
June 5,1990, we argued that 59 trial judges elected in Harris County are not single-person officers, 
that all elections -  including judicial elections ~ are covered by the Voting Rights Act. After the 
Supreme Court’s decision in HLA v. Attorney General of Texas, 115 L.Ed.2d 379 (1991), we filed a 
third brief in October 1991 responding to specific questions posed by this Court. In that brief, we 
set out the appropriate standards for assessing the state’s interest in a vote dilution analysis.

Copies of these briefs have been furnished to the court.

2



positions, it should nonetheless affirm the district court’s holding that plaintiff-intervenors 

proved a violation of section 2 in H arris County.

ARGUMENT

I. The Standards of Review Advanced by the Dissent are Incompatible with the 
Guidelines for Assessing Vote Dilution Claims Set Out by Congress and the 
Supreme Court, and Followed by this Circuit

D efendant-intervenors have advanced two novel and eccentric interpretations of how 

vote dilution claims should be analyzed. First, defendant-intervenor Wood claims, in direct 

contradiction to Congress’ express and repeated directives, that plaintiffs must show that 

racial bloc voting is caused by the racism of the white electorate. Second, the defendant- 

intervenors and Judge Higginbotham seek to change the "totality of circumstances" test into 

an inquiry that gives dispositive weight to the state’s interest in maintaining its system.

These proposed standards conflict with three fundam ental legal principles which 

must govern this case. First, the central purpose of the 1982 am endm ents to the Voting 

Rights Act, 42 U.S.C §1973 as amended, was to eliminate any purpose or intent 

requirem ent from plaintiffs’ proof. Second, the "totality of circumstances" test set out by 

Congress and upheld by the Supreme Court in Thornburg v. Gingles, 478 U.S. 30 (1986), 

m andates that no one circumstance or factor dispose of section 2 claims. Third, HLA v. 

Attorney General o f Texas, 115 L.Ed. 3d 379 (1991), squarely holds that section 2 of the 

Voting Rights Act and the "totality of circumstances" test applies to the election of judges.

3



A. Plaintiffs in Section 2 Cases Need Not Show that W hite Voters are 
Racist in O rder to Prove that Voting is Racially Polarized

The very purpose of the amendments to section 2 was to eliminate any purpose or 

intent requirem ent from section 2 cases. U nder amended section 2, courts must find 

violations whenever plaintiffs show that a challenged election practice results in denying 

minority voters an equal opportunity to participate in the political process and elect their 

candidates of choice, regardless of why the system was adopted or m aintained. Requiring 

plaintiffs to prove that the refusal of white voters to support particular African American 

candidates is motivated by racism, as the defendant-intervenors advocate, requires asking 

about the presence of a discriminatory motivation and thus re-introduces "intent" as a 

critical elem ent of proof under section 2.

Such a requirem ent is entirely incompatible the 1982 am endm ents.3 The Senate 

R eport that accompanied the amendm ents contains the most compelling and unequivocal 

evidence of Congress’ commitment to eliminating proof of "intent" from section 2 vote 

dilution analysis:

[T]he specific intent of this am endm ent [to Section 2] is that 
the plaintiffs may choose to establish discriminatory results 
without proving any kind of discriminatory purpose.

Senate R eport No. 97-417, 97th Cong., 2nd Sess. 28 (1982) (hereafter "S. Rep. a t ___")

3It is undisputed that Section 2 of the Voting Rights Act was amended in 1982 to overturn 
the Supreme Court’s decision in Mobile v. City of Bolden, 446 U.S. 55 (1980) and to remove any 
requirement that minority voters prove racist intent on the part of whites in order to prevail in a 
Section 2 vote dilution claim. See Thornburg v. Gingles, 478 U.S. at 35 ("Congress substantially 
revised § 2 to make clear that a violation could be proved by showing discriminatory effects alone").

4



(emphasis added).4 In the Senate R eport Congress articulated three principal reasons for 

rejecting any "intent" requirement, all of which compel the rejection of defendant- 

intervenors’ position.

First, Congress determined that the motivation behind the adoption of a particular 

election practice is irrelevant, so long as that practice has the current effect of excluding 

minorities from meaningful participation in the political process. See S. Rep. at 36. Thus, 

the relevant question is not why white voters have cast their ballots in a particular way, but 

whether their voting pattern, in concert with other factors, denies African American voters 

the opportunity to elect their candidates of choice.

Second, Congress’ review of prior cases showed that the "intent" inquiry proved 

unnecessarily divisive to communities, and undermined "existing racial progress" in local 

jurisdictions. S. Rep. at 36. Congress wanted to eliminate the need to "brand individuals 

as racist in order to obtain judicial relief." Id. Defendant-intervenors’ position would 

ultimately require plaintiffs to brand entire communities as racist. Only by engaging in the 

most cramped reading of the legislative history of the am endm ents to section 2 can one 

conclude that Congress’ policy concerns apply only to proof "that legislators intended to 

discriminate when they enacted or maintained a challenged electoral system" as Judge 

Higginbotham recently contended. Dissent at 71 (emphasis in original). Indeed, Congress 

expressly rejected the intent test because it required leveling charges of racism against

4The Supreme Court has specifically recognized the Senate Report as the "authoritative source 
for legislative intent" in interpreting amended Section 2. Gingles, 478 U.S. at 43 n.7.

5



"individual officials or entire communities" S. Rep. at 36 (emphasis added).5 If anything, 

requiring plaintiffs to prove the racism of individual white citizens to show polarized voting 

would be even m ore divisive than focusing solely on the behavior of a few, often long-gone 

state legislators.

Third, Congress was concerned that an "intent" requirem ent imposed too high a 

burden of proof on the plaintiffs. Defendant-intevenors’ position would in fact impose an 

impossible burden. This Court has already, and correctly, held that "[t]he m otivation^] of 

... individual voters may not be subjected to ... searching judicial inquiry" by plaintiffs who 

ask them to reveal individually how they cast their votes. Kirksey v. City o f Jackson, Miss., 

663 F.2d 659., 662 (5th Cir. 1981) reh’g and reh’g en banc denied, 669 F.2d 316 (5th Cir. 

1982). Thus, plaintiffs are barred from seeking the most direct, relevant evidence of 

causation — the motives of individual white voters who voted against African American 

candidates. Surely, plaintiffs cannot be required to prove by indirect evidence something 

they are barred from showing directly.

Judge Higginbotham thus properly rejects the defendant-intervenors’ expansive 

purpose requirem ent because "a rule conditioning relief under Section 2 upon proof of 

racial animus in the electorate ... would impose far too great a burden on plaintiffs." 

Dissent at 36-37. But he then wrongly attem pts to rescue the "causation" argum ent by 

recasting it and proposing that the plaintiffs’ showing be "limit[ed]" only to "whether

5The strength of Congress’ commitment to avoiding any kind of "intent" inquiry is reflected 
in its list of the factors most likely to prove the existence of dilution (hereafter "Senate Factors"). 
Neither evidence of motivation or causation for vote dilution were included among the most 
probative factors set out by Congress.

6



divergent voting patterns are caused by partisan differences." Dissent at 37.6 But even 

a limited version of the "causation" requirem ent cannot withstand practical or legal scrutiny.

Judge Higginbotham’s position only highlights the peculiarity of the new standard. 

Causation is an evidentiary Pandora’s box. Once evidence regarding the salience of 

partisan voting patterns becomes an essential elem ent of the inquiry into racial bloc voting, 

the court can provide no reasoned justification for limiting review to only this one causative 

factor. Defendant-intevenor W ood’s somewhat breathless list of hypothetical, and 

unsupported explanations for the consistent defeat of African American candidates — 

perhaps it was partisan voting,7 or perhaps the endorsem ent of the Houston Bar 

Association, or perhaps the endorsem ent of the Houston Post or the Houston Chronicle; 

maybe, sometimes, unflattering press accounts, or candidates’ Anglo-Saxon names, or the 

support of the Gay Political Caucus explain the outcomes — shows the untethered nature 

of such a standardless inquiry: the recitation of excuses differs from election to election and 

from candidate to candidate. But even if the evidence were organized scientifically, and 

applied uniformly to all elections, a multivariate analysis inevitably would provide more

6Judge Higginbotham does contend however, that proof that white bloc voting cannot be 
explained by partisan voting establishes only "an inference . .  . that race is at work." Dissent at 26. 
It is not readily apparent whether Judge Higginbotham contemplates limiting the causation 
requirement only to cases involving judicial elections, or whether this standard would be imposed 
upon all vote dilution cases.

’Moreover, the defendant-intervenors’ insistence that judicial elections are controlled by partisan 
voting patterns conflicts with their passionately argued view, see, e.g., April 2, 1992, Brief of 
Defendant-Intervenor Entz at 19-22, shared by Judge Higginbotham, see Dissent at 47-48, that 
judicial elections should be assessed under different standards than those used to analyze elections 
for non-judicial offices, because the function of judges is different than other elected offices.

Defendant-intervenors must choose between their characterization of elected judges as 
independent, impartial officers above the political fray, or as political candidates whose election is 
beholden to the rough and tumble of Texas partisan politics.

7



heat than light, since the analysis itself would say nothing about whether the allegedly non- 

racial factors raised by defendants were themselves the products of intentional 

discrimination.8 The same barrage of excuses to explain the racially divergent voting 

patterns in North Carolina was offered by the State in Gingles. As discussed below, the 

district court rejected these offerings by the defendant, and determ ined that voting was 

racially polarized based on the statistical patterns in the relevant elections. See Gingles v. 

Edmisten, 590 F.Supp 345, 368 n.32 (E.D.N.C. 1984) (three-judge court). The district 

court’s finding of racially polarized was affirmed by the Supreme Court.

B. Gingles Does Not Require that Plaintiffs Prove the Cause of 
Racially Polarized Voting

The theory that Gingles even permits, let alone requires, that plaintiffs prove that 

racism lies behind white voters’ refusal to support the African American community’s 

candidates of choice, rests on a fundamental flawed reading of the case. Justice Brennan’s 

plurality opinion squarely rejects such a claim: "the reasons black and white voters vote 

differently have no relevance to the central inquiry of § 2." Gingles, 478 U.S. at 63. Justice 

W hite’s concurring opinion did not dispute this central point. Rather, he thought that the

8For example, defendant-intervenor’s position would potentially permit such far-afield 
inquiries as whether the refusal of newspapers like the Houston Post and Chronicle to endorse 
minority judges running for a particular office was motivated by racism and that the Houston Bar 
Association’s failure to endorse African American candidates in particular elections was motivated 
by racism within the Bar Association.

This would impose an inordinately high burden on plaintiffs, and would require the most 
divisive kind of investigation into the racial motivations of the electorate and local community 
institutions. It is difficult to imagine, given Congress’ emphasis in amending Section 2 as set out 
above, that Congress would have condoned an interpretation of section 2 which required this kind 
of proof.

8



electoral success of minority candidates, regardless of whether they w ere supported by the 

African American community, might be used by defendants as rebuttal evidence.9 (That 

difference of opinion is clearly irrelevant to Harris County, since white voters refused to 

vote for virtually any African American candidates, thus foreclosing the claim.) Nor does 

Justice O ’Connor’s concurrence provide any support. W hite she suggested that she might 

find "[ejvidence that a candidate preferred by the minority group in a particular election was 

rejected for reasons other than those which made that candidate the preferred choice of 

the minority group," id. at 100 (emphasis added), she never intim ated that defendants could 

use a bevy of excuses to justify maintaining a system in which "bloc voting by white voters 

will consistently defeat minority candidates." Id. (emphasis added). She rem ained firmly 

focused on the question whether minority voters could elect their preferred candidates.

In any case, nothing in either Justice W hite’s or Justice O ’C onnor’s respective 

concurrences urges imposing the burden on plaintiffs to prove that no non-racial factors 

explain the statistically proven racially divergent voting patterns. In short, Gingles does not 

authorize or require the kind of causation analysis prom oted by the defendants or Judge 

Higginbotham.10 Cf. Bazemore v. Friday, 478 U.S. 385, 404 (1986) (rejecting the claim that 

plaintiffs must rebut all possible multivariate explanations as part of their case in chief).

9This view is also the view expressed by Judge Higginbotham in his Dissent in LULAC. See e.g., 
Dissent at 10. No other justice on the Supreme Court in Gingles has supported giving weight to 
elections in which minority candidates who were not the choice of minority voters were elected.

10Indeed, the Supreme Court in Gingles affirmed the district court’s finding of racially 
polarized voting, based on statistical evidence, despite the State’s insistence that partisan voting and 
other ostensibly non-racial factors explained the racially divergent voting patterns found in that case. 
See Gingles v. Edmisten, 590 F.Supp. at 368.

9



C. No O ther Court Has Required that Plaintiffs Prove the Cause of 
Racially Polarized Voting

Courts have been uniform in the view that the most salient elections for determining 

w hether voting is racially polarized in a jurisdiction are elections involving African 

American and white candidates. Jeffers v. Clinton, 730 F.Supp. 196, 208-209 (E.D. Ark. 

1989) (three-judge-court) aff'd mem., 112 L.Ed.2d 656 (1991); Mallory v. Eyrich, 707 

F.Supp. 947, 951-952 (S.D. Ohio 1989); Sanchez v. Bond, 875 F.2d 1488 (10th Cir. 1989). 

None of these courts have required proof of causation in order for plaintiffs to prove 

racially polarized voting.11

The obvious limitations of an attem pt to ascertain the "cause" of racially divergent 

voting patterns, as described above, have persuaded this Court to reject any requirem ent 

that plaintiffs prove racially polarized voting through "multivariate" statistical analysis.12

Every decision in this circuit which has addressed the question of the role of causation in 

an analysis of polarized voting, has concluded that a court need not engage in an inquiry 

into the motives of white voters in rejecting African American candidates. See Overton v. 

City o f Austin, 871 F.2d 529,538 (5th Cir. 1989); Campos v. City o f Baytown, 840 F.2d

"T he Eleventh Circuit did not endorse the view proferred by Judge Tjoflat and cited by the 
Dissent in LULAC, that plaintiffs must prove that the cause of white bloc voting is racism in the 
white community. See Solomon v. Liberty County, Florida, 899 F.2d 1012 (11th Cir. 1990). The 10th 
Circuit has expressly adopted the view that the "cause" of racial differences in candidate preferences 
is "unimportant" to the determination of bloc voting. Sanchez v. Bond, 875 F.2d 1488, 1493 (10th 
Cir. 1989).

12In Overton v. City of Austin, for example, this Court held that the analysis used by the 
district court in LULAC, which focuses on the results of bivariate regression and homogenous 
precinct analysis and supporting lay testimony, rather than extrinsic factors such as political party, 
is an appropriate method of determining the existence of legally significant racial bloc voting. 871 
F.2d 529, 538 (5th Cir. 1989). See also LULAC v. Midland Ind. Sch. Dist., 812 F.2d 1494 (5th Cir. 
1987), vacated on other grounds, 829 F.2d 546 (5th Cir. 1987).

10



1240,1243 (5th Cir. 1988); Citizens for a Better Gretna v. City o f Gretna, 636 F.Supp. 1113, 

1130 (E.D. La. 1986), affd, 834 F.2d 496 (5th Cir. 1987).13 These decisions have 

consistently affirmed the view that the statistical bivariate regression analysis approved by 

the Supreme Court in Gingles is sufficient to prove the existence of racially polarized 

voting.

II. The Legitimacy of the State’s Interests is Not a Defense to Proven Vote Dilution

Plaintiff-intervenors have never argued that the state’s interests are entitled to no 

consideration in a vote dilution analysis. Thus there are only two questions before this 

court with regard to the state’s interest: (1) at what point in the "totality of circumstances" 

test is it appropriate to consider particular state interests, and; (2) how much weight should 

be afforded those interests in the calculus of dilution.

Judge Higginbotham has interpreted HLA as "directing] that on rem and [the state’s] 

in te re s t... be weighed in determining whether a violation of the Act has occurred." Dissent 

at 44. H e then proposes that in making this determ ination this court must decide "whether 

the state’s substantial interest [in the at-large election of district judges] outweighs plaintiffs’ 

proof of dilution." Dissent at 50. U nder such a standard, the state’s interest is given 

controlling weight in the dilution analysis. The legitimacy or "substantiality" of the state’s 

interest in using a particular election structure is balanced against the plaintiffs’ proof of 

dilution. If the state’s interest is "substantial," according to the dissent, even proven vote

I3This Circuit has consistently affirmed findings of racially polarized voting in the lower court 
based on a statistical review of elections involving African American and white candidate. See 
Campos v. City of Baytown, 840 F.2d 1240,1243 (5th Cir. 1988); LULAC v. Midland ISD, 812 F.2d 
1494,1501 n.14 (5th Cir. 1987), vacated on other grounds, 829 F.2d 546 (5th Cir. 1987).

11



dilution is outweighed. This view is simply in direct conflict with the "results" test.

Much like the defendants’ proposed imposition of a causation showing for racially 

polarized voting, the view that a "substantial" state interest in at-large elections can trump 

plaintiffs’ proof of dilution re-imports an "intent" inquiry into the section 2 "results" 

determination. By removing any "intent" requirem ent from section 2, Congress deliberately 

shifted the focus of the dilution inquiry away from the purposes behind the state’s use of 

a challenged electoral practice. Giving the state’s interest dispositive weight in the dilution 

determ ination shifts the focus of the inquiry back to the intent underlying the enactm ent 

or m aintenance of the challenged election practice, in direct conflict with Congress’ 

intention in amending Section 2. This court may not "supplant the stated aims of Congress 

with its own policy preferences." LULAC  v. Clements, 914 F.2d at 654 (Johnson, J., 

dissenting).

Congress’ views with regard to the significance of the state’s interest in the liability 

determ ination under amended section 2 is amply documented in the legislative history. In 

fact, Congress responded to precisely the position advanced by the defendants and Judge 

Higginbotham. It unequivocally declared that "if an electoral system operates today to 

exclude blacks or Hispanics from a fair chance to participate," then legitimate reasons for 

enacting the challenged law 'TOO years ago [are] of the most limited relevance." S. Rep. 

at 36.

Congress specifically defined the limitation of the relevance of the state’s interest 

through the Senate R eport factors. In keeping with the "results-oriented" inquiry, Congress 

deliberately assigned limited probative value to the state’s interest, citing the tenuousness

12



of the state’s interests as one factor perhaps to be considered in some cases. But it rejected 

the idea that the legitimacy of the challenged electoral practice somehow rebuts plaintiffs’ 

proof of dilution.14 See Additonal Views of Senator R obert Dole, S. Rep. at 195 

(rejecting suggestion that "defendants be perm itted to rebut a showing of discriminatory 

results by a showing of some nondiscriminatory purpose behind the challenged voting 

practice or structure"). M oreover, Congress specifically cautioned against giving too much 

weight to the state’s interest, advising that "even a consistently applied practice premised 

on a racially neutral policy would not negate a p laintiffs showing through other factors that 

the challenged practice denies minorities fair access to the process." S. Rep. at 29 n.117.15

III. The State’s Interest Should be Weighed in the Totality of Circumstances

Since this case was first appealed, the parties have presented differing views on when 

and how the state’s interests should be assessed in a vote dilution inquiry. The defendants 

have consistently advanced the view that the state’s interest must be weighed at the liability 

stage. R ather than just weighing the state’s interest with other relevant factors in the 

dilution analysis, however, the defendants and recently Judge Higginbotham, have argued 

that in this case the state’s asserted interest in linking the electoral and jurisdictional base 

of its trial judges must be weighed against the proof of vote dilution proferred by minority 

voters.

14Neither the presence nor absence of any single factor can constitute "rebuttal evidence of non­
dilution." S. Rep. at 29 n. 118.

15Texas has failed to even demonstrate that its purported interest in the "linkage" of electoral 
and jurisdictional base for trial judges is a "consistently applied practice."

13



In our view, the state’s interest, like other factors identified by Congress, is one 

factor which can be considered at the liability stage. But the state’s interest in entitled to 

no greater weight than other factors, such as the lack of minority electoral success, Texas’ 

long history of race discrimination, or the pervasiveness of racial bloc voting.

During the liability phase of a section 2 case, the state’s interest in retaining its 

system should be accorded substantial weight only if tenuousness forms a major part of the 

plaintiff's case, that is, if plaintiffs are challenging a particular practice precisely because of 

its reflection of the state’s policy. For example, if minority voters were to challenge a 

state’s decision to have a single chief executive, then the state’s interest in electing one 

Governor for the entire state would necessarily be a central feature of the liability phase. 

Similarly, if plaintiffs in this case challenged the requirem ent that district judges in Texas 

have legal training in order to qualify for office, the state’s interest in having a judiciary 

familiar with the law would be central to the liability question.

But where minority voters do not claim that section 2 renders the state’s interest 

illegitimate, it is m ore appropriate to defer consideration of the state’s interest to the 

remedy phase. In this case, plaintiff-intervenors are not claiming that the state’s purported 

interest in linkage denies them an ability to participate equally in the political process or 

to elect the candidates of their choice. African American and Hispanic voters merely 

challenge the dilutive nature of the current system. Dilution in this case, does not turn on 

the presence or absence of the state’s interest in linkage. This is illustrated by the fact that 

non-dilutive electoral alternatives exist that would accomodate the state’s asserted interest

14



in linkage.16 At the remedy phase, where the State has the first opportunity to fashion 

an appropriate remedy, its interest in maintaining the use of the non-dilutive features of 

the challenged election system can be afforded appropriate weight.17

The Supreme Court’s decision in HLA supports our view. HLA held that "the 

State’s interest in maintaining an at-large, district-wide electoral scheme for single-member 

offices is merely one factor to be considered in evaluating the ‘totality of circumstances’." 

115 L.Ed. 2d at 387. In accordance with that view, the Supreme Court reaffirm ed the role 

of the state’s interest as a factor for consideration in the "totality of circumstances" vote 

dilution analysis. Id. The "totality of circumstances" test informs both the liability and 

remedy phase of the litigation. The Court did not direct this court to consider the state’s 

purported interest in "linkage" at the liability phase. To the contrary, the Court expressly 

contem plated that the state’s interest could be properly considered at the remedy phase:

we believe [the state’s interests concerns] are m atters that are 
relevant either to an analysis of the totality of circumstances 
that must be considered in an amplication of the results test 
embodied in Section 2, . . or to a consideration of possible 
rem edies in the event a violation is proved.

16Moreover, where the state, as in this case, has failed to prove that it in fact has a state interest 
in "linkage" for trial court judges, consideration of that interest cannot be at the core of the liability 
determination.

"Nevertheless, this Court specifically has cautioned against accomodating the state’s interests 
"at the expense of effective black minority participation in democracy" when determining the 
appropriate remedy for proven dilution. Kirksey v. Bd. of Sup ’rs Hinds County, Miss., 554 F.2d 139, 
151 (5th Cir. 1977).

15



HLA, 115 L.Ed. 2d at 387 (emphasis added).18 The interests asserted by Texas in this case 

are most appropriately considered at the remedy phase, since rem edies exist which could 

cure the dilutive nature of the current system, while preserving the State’s asserted interests 

in "linkage." If "linkage" is such an im portant interest, the State may propose rem edies such 

as cumulative or limited voting, which will cure the proven dilution, yet m aintain the 

linkage of electoral and jurisdictional base of judges.

The relevance of these alternative remedies to the "state’s interest" question cannot 

be dismissed as "red herringfs]." See Dissent at 51. The Supreme Court in HLA expressly 

recognized the availability of such remedies and cited their possible im plem entation as part 

of its rationale for rejecting the "single-person office exemption" argum ent for judicial 

elections from the scope of Section 2. 115 L.Ed. 2d at 388. See also discussion infra at 

. The availability of cumulative and limited voting rem edies dramatized to the Supreme 

Court, in fact, the danger of immunizing elections for "single-person offices" from Section 

2 coverage. The Supreme Court in HLA specifically rejected the rationale of the 

concurring judges in LULAC  who proposed that trial judges should be exempt from the

18The Court’s subsequent discussion of the state’s interest culminates in its statement, often 
quoted by the defendants out of context, that "a state interest is a factor to be considered by the 
court in evaluating whether the evidence in a particular case supports a finding of a vote dilution 
violation in an election for a single-member office." That discussion cannot be cited as a directive 
by Congress to courts to so consider the state’s interest.

Instead, any fair reading of this section of HLA reveals that the Court’s statement is merely 
its response to a hypothetical conclusion which the Court never reached in HLA: "assuming], 
arguendo, that the State’s interest in electing judges on a district-wide basis may preclude a remedy 
that involves redrawing boundaries." HLA, 115 L.Ed. 2d at 387 (emphasis in original). The Court 
in HLA never held that Texas’ asserted interest in "linkage" in fact precludes a districting remedy; 
nor did the Court hold that the current election system is necessary to protect Texas’ asserted 
interests. In fact, the Court in HLA expressly refrained from addressing the merits of the specific 
interests raised by Texas and Judge Higginbotham in LULAC. See HLA, 115 L.Ed. 2d at 386.

16



strictures of Section 2, because of the strength of the state’s interest in "linking" the 

jurisdictional and electoral base of "single-person officers." In refusing to adopt the 

LULAC  concurrence’s reasoning, the Court specifically identified the possibility that the 

"impairment of a minority group’s voting strength could be rem edied without signficantly 

impairing the State’s interest in electing judges on a district-wide basis." HLA, 115 L.Ed. 

2d at 388.

Similarly, the availability of cumulative and limited voting options dram atize the 

danger of affording the state’s interest controlling weight at the liability stage of Section 

2 vote dilution analysis. Contrary to Judge Higginbotham’s contention,19 cumulative and 

limited voting are remedies which address the very features of the election system 

challenged by the plaintiff-intervenors. Plaintiff-intervenors specifically challenged the 

features of the at-large election system that cumulative and limited voting cure: "the 

exclusionary at-large, winner-take-all, num bered place" m ethod of electing district judges. 

See Complaint of Houston Lawyers’ Association at paragraph 42.

Judge Higginbotham also mistakenly concludes that the "Gingles e lem en ts . . .  do not 

establish whether other features of an electoral scheme, such as anti-single shot rules or 

majority runoffs cause dilution." Dissent at 51. First, the test set out in Gingles is a 

"flexible, fact-intensive" test which can accomodate an analysis of many different kinds of

19Judge Higginbotham states that

"[cjumulative or limited voting are election mechanisms which preserve district-wide 
elections. Thus, they are not even remedies for the particular structural problem 
which the plaintiffs have chosen to attack."

Dissent at 51.

17



vote dilution problems. See e.g., Growe v. Emison, 61 U.S.L.W. 4163 (February 23, 1993) 

Slip Op. at 14 (holding that the Gingles elements are applicable to claims challenging 

minority fragmentation in single-member districts). Second, both the Senate R eport and 

Gingles expressly direct that courts examine the very electoral features cited by the Dissent 

as incompatible with a Gingles analysis. See S. Rep. at 29 (identifying anti-single shot 

provisions as procedures which "may enhance the opportunity for discrimination against the 

minority group"); Gingles, 478 U.S. at 38 (citing Senate Report). These enhancing features 

together create the exclusionary nature of the at-large system at issue in LULAC, which 

plaintiff-intervenors have contended, since the very outset, could be cured through the use 

of alternative non-exclusionary at-large election systems.

IV. By Any Standard, Plaintiffs Proved Their Case in Harris County

A. None of the Purportedly Non-Racial Rationale Offered by the 
Defendants In Fact Explain Racially Polarized Voting in H arris 
County

Even if the defendants’ and Judge Higginbotham’s proposed standards for assessing 

racially polarized voting were adopted by this Court, plaintiff-intervenors would prevail in 

H arris County.20 The uncontroverted evidence in the record shows that partisan voting 

does not, in fact, explain racially divergent voting patterns or lack of minority electoral 

success in district judge elections in H arris County, Texas.

20This brief addresses only the claims of the Harris County plaintiff-intevenors. We therefore 
do not address the claims of plaintiffs in other counties, and nothing in this brief should be taken 
to suggest that they are not entitled to relief.

18



Judge Higginbotham concludes based on the defendant-intervenors’s exhibits, that "of 

22 black Dem ocrats only three were elected, about 13.64%. By contrast, white Democrats 

won four out of five, 80% of their races."21 Dissent at 112. Judge Higginbotham also 

found that "four out of six black Democratic candidates, Berry, Fitch, Fisher, and Lee, lost 

in 1986, a year in which all other Democrats won." Id. The Plaintiff-intervenors’ exhibits 

correctly indicate that in addition to those African American candidates, incum bent African 

American judge M atthew Plum mer also ran for election in 1986 and lost. See P-I Exhibit 

1. Plummer, Berry, Fitch, and Williams were all Democratic incumbents who had been 

appointed to fill unexpired terms, but lost their election bids in 1986. It should be noted 

that-the record showed that the Dem ocrat at the top of the ticket that year, Governor 

M ark White, won a majority of the votes in Harris County. The district court found that 

Governor W hite at the top of the ticket "swept" into office every incum bent white 

Democratic district judge, but failed to "sweep" in the African Am erican Democratic 

incumbents. LULAC v. Clements, RE at 28. The evidence is overwhelmingly clear: 

partisan voting does not, in fact, explain the loss by African American candidates in district 

judge elections in H arris County.

Nor does partisan voting explain the differences in candidate preferences between 

white and African American voters found by the experts. Plaintiff-intervenors’ expert, Dr. 

Engstrom, testified that the overwhelming majority of African American district judge

21Dr. Engstrom, the expert for the plaintiff-intervenors in Harris County testified that between 
1980 and 1988 only 12.5% of African American Democratic district judge candidates were elected, 
while during the same period 52% of white Democratic district judge candidates were elected. TR. 
at 3-134-135. By both the plaintiff-intervenors and the defendant-intervenors’ evidence, partisan 
voting does not explain lack of minority electoral success or racially polarized voting in Harris 
County.

19



candidates who ran in general elections between 1980 and 1988 in H arris County received 

a num ber of votes which consistently put African American candidates in the bottom half 

of all candidate vote getters in the County. TR. at 3-134-136. See Gingles v. Edmisten, 590 

F.Supp. at 368 (white voters in heavily Democratic areas consistently ranked African 

American candidates next to last among all candidates).

The H arris County evidence withstands even the racial polarization standards 

advanced by the defendant-intervenors, Judge Higginbotham, and the concurring justices 

in Gingles. The record is replete with evidence that the non-racial explanations offered by 

the defendants do not explain the consistent defeat of African American judicial candidates 

in H arris County. In 1986 when all of the white Democratic judges were swept into the 

office by the D em ocrat at the top of the ticket, all of the African American Democratic 

district judge lost their election bids. Several of the defeated African American candidates 

were incumbents (Judges Berry, Fitch and Plummer had been appointed and thus ran as 

incumbents). Judge Francis Williams was endorsed by the Houston Bar Poll and at least 

one of the major Houston newspapers.

B. The Houston Lawyers’ Association Proved the Existence of Vote 
Dilution in Harris County

The Houston Lawyers’ Association also proved the existence of other factors which 

support its claim that the current method of electing district judges in H arris County, 

dilutes the voting strength of politically cohesive African American voters.

It was undisputed at trial that African Americans could constitute a majority of the

20



voting age population in a t least nine single-member judicial districts. Plaintiff-intervenors 

profferred illustrative electoral sub-districts in order to illustrate to the court, in accordance 

with the first prong of the three-part Gingles test, the nature of the dilution challenged. 

It was also conceded by all parties that African American voters in H arris County are 

politically cohesive. In 16 of 17 election contests involving white and African American 

candidates examined by the plaintiff-intervenors’ expert, African American voters 

consistently gave at least 97% of their vote to the African American candidate. R E at 26.

The record also attests to the lack of African American electoral success in Harris 

County. Only three African Americans successfully ran in an opposed general election for 

district judge in H arris County during the 1980s.22 One of those judges, John Peavy, was 

initially appinted to the bench and has run unopposed since 1978. The other sitting African

22The suggestion that the success of African American candidates who were not the candidate 
of choice of African American voters should be given weight in the polarized voting analysis has 
been rejected by this court. Campos v. City of Baytown, 840 F.2d 1240,1245 (5th Cr. 1988)(a "viable 
candidate is one which the minority group sponsors"). Similarly, the suggestion that the election 
of white candidates who were supported by African American voters in any way changes plaintiff- 
intervenors’ proof of racially polarized voting ignores the common-sense conclusion of this court 
that "[ejvidence of black support for white candidates in an all-white field . . .  tells us nothing about 
the tendency of white bloc voting to defeat black candidates." Westwego Citizens for a Better Gov’t 
v. City of Westwego, 872 F.2d 1201, 1208 n.7 (5th Cir. 1989), on remand, 946 F.2d 1109, 1119 n. 15 
(5th Cir. 1991).

Finally, Judge Higginbotham’s suggestion that elections involving Hispanic candidates must 
be considered in the racial bloc voting analysis in Harris County is entirely insupportable. Plaintiffs 
have advanced no claim on behalf of Hispanics in Harris County. Plaintiff-intervenors advanced 
a claim on behalf of African American voters only. African American voters in Harris County 
cannot be forced to include Hispanics in their vote dilution analysis simply because entirely different 
parties claim that African Americans and Hispanics are politically cohesive in an entirely different 
county in the state. The vote dilution inquiry under Section 2 is "an intensely local appraisal' 
Gingles, 478 U.S. at 79 (emphasis added). The presence of political cohesiveness between African 
Americans and Hispanics in Midland County, therefore, does not "compel the conclusion that there 
is also black-Hispanic cohesion in Harris" County as Judge Higginbotham contends. Dissent at 40.

Moreover, racial minorities protected under the Voting Rights Act are not interchangeable. 
White voters support of Hispanic candidates does not tell us anything about the willingness of white 
voters to support African American candidates.

21



American district judge, Carl Walker, won in 1986, winning a judicial seat that had been 

held by a white judge also named Walker. The late Thomas Routt, the third African 

American judge who ran in an opposed district judge general election and won, testified 

that minority candidates enjoy greater success if they do not disclose their racial identity 

when they run for judicial office in Harris County. TR. at 3-206. Then-incum bent Judge 

R outt barely defeated his white opponent, a virtual unknown in 1982, by receiving 51%, a 

bare majority of the total vote.23 See Gingles, 478 U.S. at 60-61 (appropriateness of closely 

examining circumstances of election of African Americans).

In addition to proof of racially polarized voting and lack of minority electoral 

success,24 plaintiff-intervenors proved the presence of the other particularly relevant 

Senate R eport factors as well. The District Court took judicial notice of the history of 

discrimination which has touched the right of African Americans to vote and otherwise 

participate in the political process. R E at 69-70. Plaintiff-intervenor W eldon Berry, a 

m em ber of the Houston Lawyers’ Association, attested to the history of discrimination in 

education in H arris County which precipitated his attendance at Texas State University for 

Negroes in 1950 in order to receive a legal education in the state of Texas. Texas State 

University for Negroes was erected by the State to avoid permitting African Americans to 

attend the University of Texas Law School. See Sweatt v. Painter, 339 U.S. 629 (1950).

Similarly, the evidence regarding the lingering effects of discrimination in H arris

^In 1991, African American Judge John Kyles was appointed to fill the seat vacated upon the 
death of Judge Routt. In his election bid to retain his seat, Judge Kyles was defeated by a white 
opponent in the November, 1992 general election.

24Proof of racially polarized voting and lack of minority electoral success are the two most 
important factors to consider in the vote dilution analysis. Gingles, 478 U.S. at 48 n.15.

22



County which affects the socioeconomic condition of African Americans, see Plaintiffs 

Exhibit H-08, was essentially undisputed.

C. The State H as Never, In Fact, Proved its Interest in "Linkage"

A t trial the State of Texas offered 3 reasons to support its interest in using the 

challenged election district judge election system:

"(1) judges elected from smaller districts would be m ore 
susceptible to undue influence by organized crime; (2) changes 
in the current system would result n costly administrative 
changes for District C lerk’s offices; and (3) the system of 
specialized courts in some counties would disenfranchise all 
voters rights to elect judges with jurisdiction over some 
m atters.25"

R E at 76. The District correctly found that these interests were not "compelling," in part 

because the State’s concerns could be accomodated by several rem edial options. Id.

In reliance on the analysis set out in LULAC II, 914 F.2d 620 (5th Cir. 1990), the 

defendants have now discovered a "new" primary interest in preserving the current district 

judge election system as an interest in "linking" the electoral and jurisdictional base of 

district judges. But Texas’ rhetorical assertion of its interest on appeal cannot translate into 

proof that such an "interest" exists, nor that such an interest is sufficiently strong to 

outweigh the consistent denial of equal voting opportunity to African American voters.

The interests actually asserted by Texas at trial are de minimis, and obviously 

insufficient to warrant great weight in the vote dilution analysis. . . The district court’s

25It was in the context of specialized courts that the State raised the "linkage" argument at
trial.

23



findings as to the state’s interest actually asserted at trial is undoubtedly not clearly 

erroneous.

The state’s asserted fear that judges elected from small districts would be subject to 

the influence of organized crime is undermined by the existence of current judicial districts 

smaller than the hypothetical sub-districts for H arris County proferred by plaintiff- 

intervenors. The state’s administrative concerns regarding specialized courts and the cost 

of reorganization are irrelevant to the question w hether the current system discriminates 

against minority voters. This Court has held that administrative considerations should not 

"play a role in determining whether there has been a violation of section 2." Westwego, 872 

F.2d at 1211. The d is ru p tio n  of the specialized functions of elected officials resulting from 

a proposed change in the electoral system "is not a sufficient ground for maintaining an 

otherwise flawed system." Id., citing Dillard v. Crenshaw County, 831 F.2d 246 252 (11th Cir. 

1986). Clearly the interest asserted by the State during trial — the appropriate forum for 

advancing such claims -  are not "compelling" or even "substantial."

Even the "linkage" interest Texas asserts on appeal is not persuasive. First, Texas’ 

interest is underm ined by its own Constitution, which provides that counties may choose 

to elect their judges from sub-districts. Art. 5 Section 7(a)(i), Texas Constitution of 1876, 

as amended. The fact that since 1985, when this provision was enacted in the Constitution 

no county has exercised the option to elect district judges from sub-districts does not, as 

Judge Higginbotham overstates, constitute "254 distinct affirmations of [the state’s] policy." 

Dissent at 50. Texas’ policy interest is indicated more directly by its decision to include 

such an option in its Constitution.

24



Texas’ claim that the "linkage" of the jurisdictional and electoral base for judges is 

essential to m aintaining judicial independence and accountability is simply not supported 

by Texas’ practices. First, Texas does not require "linkage" for its Justices of the Peace, 

who are elected from sub-districts within the county, but exercise countywide jurisdiction. 

See Bradley v. Swearingen, 525 S.W. 2d 280,282 (Tex. Civ. App. 1975). M oreover, Texas 

does not insure that litigants will appear before judges over whom they have electoral 

control — a central argum ent advanced by the proponents of the "linkage" argument. See 

Higginbotham Dissent at 47. Chief Justice Phillips, chair of the Texas Judicial Districts 

Board, testified at trial that judges in Texas are often called to hear cases in counties 

outside of the county from which they were elected to assist with docket control. TR. 5- 

120. As a result, litigants in Texas may often appear before judgs over whom they have no 

electoral control.

This court may not simply assume the legitimacy of the "linkage" interest. Many 

states, such as North Carolina, New York, New Mexico, Louisiana, Mississippi and 

Arkansas, elect some of their trial judges from districts which are not co-terminous with the 

judge’s area of primary jurisdiction. These states are presumably also concerned with 

having a qualified, independent judiciary. Nothing in the record, Texas law, or the 

imaginative hypothesizing of inventive defendants can justify concluding that Texas’ 

eleventh-hour invocation of linkage outweighs the national policy of giving minority voters 

an equal opportunity to participate and elect the candidates of their choice.

25



CONCLUSION

For the foregoing reasons, plaintiff-intervenors respectfully request that this Court 
affirm the findings of the district court as to Harris County, Texas.

Respectfully submitted,

Gabrielle K.McDonald, Esq. 
W alker & Satterthwaite 
7800 N. M opac 
Suite 215 
Austin, TX 78759 
(512) 346-6801

Elaine R/ Jobes 
Charles Stephen Ralston 
Sherrilyn A. Ifill 
99 Hudson Street 
16th Floor
New York, NY 10013 
(212) 219-1900

Council for H ouston Lawyers’ Association

April 19, 1993



CERTIFICA TE O F SERV ICE

I hereby certify that on this 19th day of April, 1993 a true and correct copy of an 

Appellee-Intervenors’ Supplemental Brief on Rehearing En Banc was mailed to counsel of 

record in this case by first class United States mail, postage pre-paid, as follows:

William L. G arrett 
Brenda Hall Thompson 
G arrett, Thom pson & Chang 
8300 Douglas, Suite 800 
Dallas, TX 75225

Rolando L. Rios 
The Law Office of Rolando 

L. Rios
Milam Building, Suite 1024 
115 E. Travis Street 
San Antonio, TX  78205

Edward B. Cloutman, III 
Cloutman, Albright & Bower 
3301 Elm Street 
Dallas, TX 75226-1637

J. Eugene Clements 
John E. O ’Neill 
Evelyn V. Keyes 
Porter & Clements 
700 Louisiana, Suite 3500 
Houston, TX 7002-2730

M ichael J. W ood 
Attorney a t Law 
440 Louisiana, Suite 200 
Houston, TX  77002

John L. Hill, Jr.
Liddell, Sapp, Zivley,

Hill & LaBoon 
3300 Texas Commerce Tower 
Houston, TX  77002

David R. Richards 
Special Counsel 
600 W est 7th Street 
Austin, TX 78701

R obert H. Mow, Jr.
Hughes & Luce 
2800 M omentum Place 
1717 M ain Street 
Dallas, TX 75201

Gabrielle K. M cDonald 
W alker & Satterthwaite 
7800 N. M opac 
Suite 215 
Austin, TX 78750



D an M orales 
Will Pryor 
Mary F. Keller 
R enea Hicks 
Javier Guajardo 
A ttorney G eneral’s Office 
Price Daniel Sr.

Office Building 
209 W. 14th Street 
Austin, TX  78701-2548

Seagal V. W heatley 
Donald R. Philbin, Jr. 
Oppenheim er, Rosenberg, 

Kelleher & Wheatley, Inc.
711 Navarro, Sixth Floor 
San Antonio, TX  78205

E. Brice Cunningham
777 South R.L. Thornton Freeway
Suite 121
Dallas, TX  75203

D arrell Smith
10999 Interstate Highway 10 
Suite 905
San Antonio, TX 78230

W alter L. Irvin 
5787 South Ham pton Road 
Suite 210, Lock Box 122 
Dallas, TX 75232-2255

Ken Oden
Travis County Attorney 
P.O. Box 1748 
Austin, TX 78767

Tom Rugg 
Jefferson County 

Courthouse 
Beaumont, TX 77701

Jessica Dunsay Silver, Esq. 
M ark Gross, Esq. 
c/o A ttorney General of the 

United States 
United States D epartm ent 

of Justice
M ain Justice Building 
10th & Pennsylvania 

Avenue, N.W.
Washington, D.C. 20530



Blacks were ever promoted, with the differences in promotion 
rates being at the statistically significant level at grades 7,
9, and 13 in the period as a whole, and significant at the 
GS-11 level in the period before January 1, 1977. (PX 5; PX 63; 
R. Vol. X 63-65.)

e. The Defendants' Statistical Evidence
To the extent the statistics presented by the defendants 

addressed plaintiff's case, they supported the conclusions drawn 
by Dr. Levin. The sole source for defendants statistical case 
was the testimony and report presented by Dr. James Beckett, II. 
Dr. Beckett studied essentially the same data that was the basis 
of Dr. Levin's study. However, instead of examining time in 
grade he looked at levels of progression over a time period 
using for comparison purposes Blacks and non- Blacks who entered 
in particular grade levels in particular years. (DX 201).

Dr. Beckett's methodology fragmented the numbers down to 
such small segments that in most instances there were only one 
to three, and at most four, Blacks in each cohort group studied. 
(DX 201; R. Vol. XV, 222-23.) Not surprisingly, the use of 
very small samples tend to produce results which lack statistical 
significance. (R. Vol X, 79-81.) This is becaue, as Dr. Beckett 
himself admitted, the level of statistical signficance is deter­
mined both by the number of examples studied as well as the dif­
ferences between the groups studied. (R. Vol. XVI, 12-13; 21).

Dr. Beckett made no attempt to combine his results.
(R. Vol. XV, 230-31; XVI, 21-22) Dr. Levin, on the other hand, 
testified that accumulation was possible through accepted

21



methods and that when Dr. Beckett's figures were added together 
they produced generally statistically significant differences 
in the progression rates between Blacks and whites. (R. Vol. X, 
81-87.)

Further, even though the number of Blacks in each group 
studied by Dr. Beckett was very small he found statistical signi­
ficance in a number of instances. (DX 201; R. Vol. XV, 222-23.)
Indeed, if one uses the level of statistical significance, .05 or

11/5 chances out of 100, more than 5% of the instances were at 
the level of statistical significance. (DX 201; R. Vol. XVI, 
28-31.) This result in and of itself was statistically signifi­
cant and demonstrated that the differences in progression rates 
between Blacks and non-Blacks was at a level from which an 
inference of discrimination could be drawn.

Moreover, on cross-examination Dr. Beckett revealed that 
his approach and method of structuring his data obscured the 
actual pattern of advancement of Blacks and non-Blacks. For 
example, he removed from his study a group of Blacks with long 
times to promotion (R. Vol. XV, 187-89; XVI, 61-65; 66-73.). 
Further, his actual data shows a consistent pattern of differ­
ences in distribution of Blacks between grade levels to their 

±8/disfavor. Thus, in many instances the overall distribution
disfavored Blacks, but the median grades were the same. The

17/ Dr. Beckett used the probability level of .02275 as 
establishing statistical significance. R. Vol. XVI, 22-27.
18/ For example, if one examines the table in DX 201 for Field 
Examiners in the 1976 entry level group, the GS-7 critical 
grade for the year 1978, the actual distribution of Blacks and 
non-Blacks was as follows:

22



The same general pattern was consistently present. (R. Vol.
XV, 219; XVI, 7-11; 15-18.)

In sum, Dr. Beckett acknowledged that the findings of his 
study were consistent with those of Dr. Levin. The persons who 
lagged behind in statistically significant numbers tended to be 
Black, and more Blacks lagged behind than one would expect by 
the random operation of chance. (R. Vol. XV, 223-24; XVI, 
13-15; 30-32; 74-75.) On the other hand, virtually all whites 
moved ahead rapidly and on time.

Finally, defendants in no way challenged or rebutted plain­
tiff's showing with regard to the low number of Blacks in super­
visory or management positions above the journeyman level. Al­
though defendants attempted to suggest that one explanation was 
that Blacks left the agency without waiting for opportunities to 
develop, it was not shown that Blacks left at a higher rate than 
Whites. Moreover, the testimony of many Black class members 
established that Blacks left because they became discouraged

Contd.
Blacks Non-Blacks

GS-9 1 0
GS-11 3 35
GS-12 0 1

Vol. XV, 214-15.)
1980, the distribution was:

Blacks Non-Blacks
GS-11 1 0
GS-12 2 16
GS-13 0 12

(Id. at 216-17).

23



from the patterns they observed and concluded that their oppor­
tunities to advance to higher level positions were much less than 
those of their White counterparts.

5. The Expert Analytical Proof
Dr. James Outtz, a reputable industrial psychologist who has 

been qualified as an expert in employment discrimination cases 
against the federal government, was retained by the plaintiffs 
for the purpose of analyzing the promotion schemes used at NRLB. 
Industrial psychology is the field of psychology which applies 
the basic principles of human behavior to the world of work. (R. 
Vol. XIII 217) In preparing his analysis and report Dr. Outtz 
analyzed the various components of the promotion scheme.

Dr. Outtz testified and his expert report confirms that the 
processes involved in the NLRB's promotion scheme are standard­
less and non-objective. There are no official or written or oral 
standards for supervisors which guide them in the preparation of 
the appraisal. (R. Vol. XV 91-93; XVI 202; 203.) Prior to 1980 
the agency did not provide any written or verbal guidanace 
regarding how to conduct appraisals. (R. Vol. XVI 202.) There 
is no requirement that the individual submitting the appraisal 
have substantive knowledge of the employee's work product or
performance to any specific degree or for any fixed period of

1 9 /
time. Thus, the favorable evaluation of a first line supervisor

19/ For example, the Regional Director in St. Louis admitted 
during the trial that he did not work directly with Ms. Cathy 
Carey, an attorney in the region. (R. Vol. XIII 7, 34, 36). Nor 
did he ever discuss her written work product prior to evaluating 
her. (R. Vol. 13-14, 36-37.) Yet he was the "seminal force" in 
denying her a promotion to the GS-12 level. (R. Vol. XVIII, 112; 
see also, XVII 73; XII 116.)

24



could be discarded and replaced by a negative recommendation of
a manager who had less direct contact with the employee. (R.
Vol. XVIII, 53-54, 63; Vol. XVII, 97-100; 102-104; XIX, 6-9, 38;
XIX 122; DX 256.) Supervisors are not required to document the
basis of their appraisals. (Trunkes Dep. 63; R.Vol. XVIII 63.)

The determinants a supervisor uses to develop the appraisal
are undefined, arbitrary and in large part based merely on what
the individual supervisor believes is important to success on the
job. (R. Vol. XIV 6.) Supervisors do not receive comprehensive
training which would enable them to adequately and objectively
evaluate an employee. (Trunkes Dep. 63; Herman Dep. 19-20; R.

20/
Vol. XVI 181-182.) Many supervisors did not receive supervisory 
training until several years after they had supervised and eval­
uated employees. (R. Vol. XVIII 4; Vol. XV 67-68; Vol. XVI 
198-199, 202; Vol. XIX 18-19, 42; XIX 88; Herman Dep. 19;
Trunkes Dep. 44.)

Dr. Outtz' testimony was based upon the professional lit­
erature and his experience in the field. PX 70; R. Vol. XIV 14.) 
He testified that the NLRB's appraisal system is subjective in 
that it is not a barometer which measures actual performance 
nor does it evalute observable behavioral patterns. (R. Vol. XIV 
15, 25, 26-28.)

Dr. Outtz testified that there are controls which could be 
effectively used to reduce rater bias. (Id., 30, 33.) In con-

20/ The training which the agency does offer to supervisors is 
de minimus, (id.) and the sessions which are a component in a 
broader annual program, did not last more than two hours (id.; DX 239).

25



elusion, Dr. Outtz established that the appraisal system is not 
related to job performance nor does it preedict an employee's 
ability to successfully perform the job. The NLRB has not con­
ducted any studies to ascerain the validity of their appraisal

21/or promotion processes. (R. Vol. XVIII 72, XVI 152, 160.)
Thus, Dr. Outtz testified that the appraisal system "of its 
own force, or by its very nature, operates in a negative manner 
against Blacks." (R. Vol. XIV 7.) The defendants failed to call 
their expert witness to rebut the testimony of Dr. Outtz which 
recendered the evidenced presented by Dr. Outtz conclusive.

6. The Named Plaintiff and Class Members 
a. The Claim of Donald C. Lewis 

Mr. Lewis, who is presently a GS-13 field examiner, is 
employed in Region 23, Houston, Texas. He was initially hired 
by the agency in 1971 as a field examiner with a rating of 
GS-9. In October of 1972 he was promoted to GS-11 and in 1973 
to GS-12. He was not promoted for five years, or until May of 
1978, when he became a GS-13. In January, 1975, Mr. Lewis was

21/ The Uniform Guidelines and the Federal Personnel Manual 
apply the principles of adverse impact and validation to 
all "selection procedures" 5 C.F.R. § 1607.2-C. "Selection 
Procedures" are defined as "any measure, combination of meas­
ures or procedures used as a basis for any employment decision," 
including the "full range of assessment techniques" such as 
"educational and work experience requirements through ... 
unscored application forms." § 1607.16. Similarly, the offi­
cial Questions and Answers on the Uniform Guidelines state 
that they apply "to all selection procedures ... including ... 
review of experience or education from application forms."
See 401 FEP Cases 2304 (1979); § 1607.16. Chapter 335 of the 
Federal Personnel Manual, at 335-1-4, provides that "methods 
of evalution for promotion and placement "must be consistent 
with instructions" (FPM Supl. 335-1) that refer to validation 
and the Uniform Guidelines.

26



rated highly qualified for promotion to a GS-13 position. (PX 
34). That rating was essential for his promotion beyond the 
journeyman level. After he had been so rated, he inquired of the 
regional director, Mr. Baldovin, about opportunities to obtain a 
GS-13 position. For the first time he was advised that the 
agency maintained a 50% quota on the number of GS-13 promotions 
which he was told blocked any advancement. (R Vol. XI, 21-22.)

After futile attempts to obtain written confirmation of
this rule, Mr. Lewis exercised his right to file an EEO complaint
claiming that the quota discriminatorily blocked the promotional
opportunities of Blacks. (R Vol. X, 22-25; 32-33.) Prior to the
filing of the EEO complaint, Mr. Baldovin had informed Mr. Lewis
that in lieu of a promotion he would obtain a quality within-grade
increase. (R. Vol. X, 31-32.) After his complaint was filed,
however, the increase was not given to Mr. Lewis. It was only
after Mr. Lewis complained to the agency's EEO director and the
director contacted Mr. Baldovin that Baldovin did the paper work
necessary for Lewis to obtain the increase, which he did not
receive until May or June. (R. Vol. X, 32-34; R. Vol. XIX,
192-198.) Mr. Baldovin testified that it was his ordinary
practice to put through such increases once per year and he
had simply "forgotten" about Mr. Lewis. However, the employee
record cards for Region 23 reflect a substantial number of
employees receiving within-grade increases at various times

22/during the year. (PX 75).

22/ Thus, Mr. Lewis' employee record card (PX 32), reflects that 
in other years he received within-grade increases in April, 1977 
and April, 1978. Theodore Arter received one in September, 1976;

27



A further act of retaliation occured when Mr. Lewis was
assigned to a new team, into which were placed all the Black 
professionals in the office, and Mr. Baldovin attempted to 
enveigle the senior Black in the office, Mr. Robert Penrice, 
to lay the ground work for further action against Mr. Lewis.
(R. Vol. XIII 153-58; PX 75) In addition, Mr. Lewis was 
inundated with a barrage of "challenge and objection" cases 
which are extremely difficult and time consuming to process.
(R. Vol. XI 19-22) He had never received these case assignments 
prior to filing his complaint, nor was any other agent ever 
assigned as many of these cases, which usually were handled by 
persons well above Mr. Lewis' grade level. (PX. 80)

Only after Mr. Lewis filed the EEO complaint did negative 
comments begin appearing in his file (R. Vol. XI 6-8;
11-19.) Indeed, he had received a quality increase for his 
work in 1974. (R. Vol. XI 31.) In January 1976, the next rating
period after the filing of the EEO complaint, Mr. Baldovin 
overruled the favorable recommendation of Mr. Lewis' immediate 
supervisor and diminished his rating to "not well-qualified" for 
a GS-13 position. (PX 34; R. Vol. XI 25-27; XIX 202-05.) In 
February, 1976 such a position became vacant and was given to a 
White field examiner, Mr. Robert Markey. Mr. Baldovin's attempted 
explanation of this action is also squarely contradicted by the

22/ Contd.
Robert Crain in June, 1974, November, 1975, and December, 1978; 
Robert Levy in August, 1977; Richard Linton in June, 1975 
and November, 1976; Michael McReynolds in January, 1974 and 
April, 1978; Guadalupe Ruize in August, 1978; Arthur Safos in 
July, 1975; and Wilton Waldrop in May 1979. (PX 75). Of these 
nine persons only Mr. Lewis is black.

28



the unrefutable documentary evidence.
Baldovin testified that Mr. Markey had been rated well

qualified for a GS-13 position before Mr. Lewis' initial rating
in January of 1975. Therefore, he claimed, even if Mr. Lewis
had still been rated well qualified in January 1976, he would
not have received the position that opened in February, because
Markey had seniority in terms of the well qualified rating. (R. 

23/
Vol. XIX 200) However, Markey's employee record card
evidences that he was first rated well qualified for the GS-13
position in August of 1975, some eight months after Mr. Lewis
was rated as being well qualified in January, 1975. (PX 75; PX 24/
32). Therefore, if Mr. Lewis had retained his rating of 
well-qualified in January of 1976, he would have been entitled 
the promotion in February of 1976. As it was, he did not

23/ Under an arbitrator's decision, promotions to GS-13 field 
examiner, non supervisory positions were non-competitive and were 
to be given in order of seniority based on the time at which the 
eligible persons were rated well-qualified. (PX 24).
24/ Mr. Lewis' employee record card has the following 
sequence of entries:

1- 22-75 —  13CO: WQ; 13S: WQ
2- 9-76 —  13CO: NR; 13S: NR 
1-13-77 —  13CO: WQ; 13S: WQ
1- 31-78 —  13CO: WQ; 13S: WE 
5-21-78 —  Promotion to GS-13.

(DX 416).
Mr. Markey's card reflects the following:

8-16-74 —  13CO: NWQ [Not Well Qualified]
8-26-75 —  13CO: WQ; 13S: NR
2- 29-76 —  Promotion to GS-13.

(PX 75).

29



• 25/receive such a promotion until two years later in 1978.— '

In contrast, two white field agents in the region took a sub­
stantially shorter time to reach the GS-13 level than did Mr. 
Lewis. (R. Vol. XVI 32-33.)

Although Mr. Lewis has been consistently rated "well 
qualified" for supervisory and management positions since 1978, 
he has not received any further promotions to date. (R. Vol. X, 
29-30.) Throughout this period he has continued to file excep­
tions to and has challenged as retaliatory negative comments in 
the appraisals forwarded to Washington for review. (PX 34.)

b. Testimony of class members.
During the trial eleven members of the class, both field 

examiners and attorneys, from nine regional offices presented 
illuminating testimony. In addition, depositions were introduced 
on behalf of three other class members, and documentary evidence 
was admitted into the record itself with respect to eight others. 
In sum, plaintiffs introduced evidence on behalf of twenty-two 
class members from eleven regional offices. Because of space 
limitations, plaintiffs can not present a comprehensive summary 
of their testimony. We refer the Court to plaintiffs' proposed 
Findings of Fact, pp. 47-87, and to the discussions at pp. 24-26 
supra and 55-57 infra.

25/ The action against Mr. Lewis was part of a consistent pat­
tern of retaliation by Mr. Baldovin against persons who filed 
EEO complaints or who were otherwise active in EEO matters.
Thus, Mr. Penrice testified, without contradiction, that Mr. 
Baldovin had told him (Mr. Penrice) that if a Hispanic employee 
did not drop her EEO complaint, "she would never get promoted." 
(R. Vol. XIII 163-65.) And Baldovin reprised against Penrice

30



Summary of Argument
I .

The mechanical adoption of the defendant's proposed findings 
of fact renders the court's decision vulnerable under Rule 53(a). 
Therefore, this Court is free to review them in light of the 
entire record herein.

II.
Plaintiffs presented a clear prima facie case of racial 

discrimination in promotions through statistical evidence. The 
defendants failed to meet their burden of rebutting it either 
by counter statistical evidence or by demonstrating non-dis- 
criminatory reasons for the observed disparities in promotions.

III.
The district court failed to assess all of the relevant 

evidence before it. It therefore ignored much proof of discri­
minatory practices responsible for the statistical disparities 
that were shown.

IV.
The district court's concentration on irrelevant evidence

25/ Contd.himself for his refusal to act as a hatchet man against the only 
other Black professionals in the region, Mr. Lewis and a Black 
attorney, Mr. James Palmer. (R. Vol. XIII 171-76).

Eventually, Mr. Baldovin fired the Black attorney and as 
recently as April, 1982, verbally abused a Black supervisory 
attorney, Mr. Bernard Middleton. Despite this obvious pattern 
of harassment, abuse, and discrimination against virtually every 
Black professional in Region 23 since he became director, the 
agency has taken no action against Mr. Baldovin. (R. Vol. XV,
3-11? PX 32).

31



introduced by defendants relating to hiring, initial assignments, 
and programs for non-professional employees distored its analysis 
of the issues and the proof before it.

V.
The named plaintiff established a prima facie case of discrim­

ination and reprisal against him as an individual. The defendants' 
attempted rebuttal failed to provide legitimate and nondiscrimina- 
tory reasons for the failure to promote plaintiff. Moreover, the 
uncontradictable documentary evidence establishes that the reasons 
were pretextual.

VI.
Under a Department of Justice directive issued in 1978 and 

not rescinded until April, 1983, a defendant federal agency may not 
seek costs in a Title VII action except in cases where the require­
ments established in Christianburg Garment Co. v. EEOC are met.
The present case clearly was not baseless, meritless, or brought 
for vexatious reasons.

ARGUMENT

I. THE DISTRICT COURT'S VERBATIM ADOPTION OF THE DEFENDANT'S 
FINDINGS OF FACT AND CONCLUSIONS OF LAW UNDERMINED THE 
PURPOSE OF RULE 52(a) OF THE FEDERAL RULES OF CIVIL PRO­
CEDURE AND THEREFORE ITS OPINION IS CLEARLY ERRONEOUS.
In the instant action the trial judge adopted almost verba­

tim and in toto the proposed findings of fact submitted by the 
defendants, and did not independently formulate even one. The 
differences in the two documents are minor involving either the

32



addition or deletion of a single word or a few words, or the
27/

re-numbering of the paragraphs. Notably, the trial court
followed the defendant's findings so closely, that the court even

28/adopted the mistakes made by the defendant.
The Supreme Court has criticized a trial judge's mechanical

adoption of the findings of the prevailing party. In United States
v. El Paso Gas Co., 376 U.S. 651 (1964) the court quoted at
length from a statement of Judge J. Skelly Wright:

I suggest to you strongly that you avoid as far as you 
possibly can simply signing what some lawyer puts under 
your nose. These lawyers, and properly so, in their zeal 
and advocacy and their enthusiasm are going to state the 
case for their side in these findings as strongly as they 
possibly can. When these findings get to the court of 
appeal, they won't be worth the paper they are written on 
as far as assisting the court of appeals in determining 
why the judge decided the case. Seminars for Newly 
Appointed United States District Judges (1963), p. 166.

376 U.S. at 656 n. 4.

27/ As is detailed in the attached Appendix "A", which is a 
paragraph by paragraph comparison of the Court's opinion 
with the defendant's proposed findings of fact, the district 
court's findings of fact were identical or substantially the 
same as the findings proposed by the defendants.
28/ The circumstances surrounding the submission of defendants' 
proposed findings of fact and conclusions of law reveal the 
suspect nature of the court's total adoption of defendants' 
brief. At the conclusion of the trial on July 9, 1982 the Court 
directed both of the parties to simultaneously file their pro­
posed findings of fact and conclusions of law within three weeks. 
(R. Vol. XX, 38-40.) Accordingly, the brief was due on August 
6, 1983. Plaintiffs' counsel, obtained extensions of time for 
filing the brief, and their pleading was filed on August 23,
1983. Contrariwise, defendants who never sought an extension of 
time for filing their brief; nor advised plaintiffs that their 
submission would be substantially late did not file their findings 
of fact and conclusions of law until September 25th. Thus, they 
had the opportunity to review plaintiffs brief in excess of one 
month prior to submitting their findings of fact. Defendants' 
brief was accepted by the Court in the absence of any explanation 
and was the basis of the Court's opinion of November 23, 1982.

33



This Court has also denounced the "unfortunate practice"
of trial judges adopting the proposed findings of one of the 

30/
parties and it has been disapproved to varying degrees by

31/the majority of the courts of appeal. The procedure has been 
rejected because it does not assure that the trial judge ana­
lyzed the facts with the care necessary to insure that they 
support his decision (see, e ,g., United States v. Forness, 125 
F.2d 928, 942 (2nd Cir., cert, denied, 316 U.S. 694 (1942)) and 
the proposed findings formulated by one of the parties may not 
adequately reflect the reasoning process used by the judge in 
reaching his decision (see, e.g., Roberts v. Ross, 344 F.2d 747 
(3rd Cir. 1965)). El Paso, 376 U.S. at 656; Industrial Build­
ings Materials, Inc, v. International Corp., 437 F.2d 1336,

32/1340 (9th Cir. 1970).

2 9 /

29/ Volkswagen of America v. Jahre, 472 F.2d 557, 559 (5th 
Cir. 1973).
30/ See, e.g., FMC Corp. v. Varco International, Inc. 677 F.2d 
500, 501-502 n.2 (5th Cir. 1982), Amstar Corporation v. Dominic*s 
Pizza, Inc, , 615 F.2d 252, 258 (5th Cir. 1980); Kaspar Wire Works, 
Inc, v. Leco Engineering, 575 F.2d 530, 543 (5th Cir. 1978),_ 
Kinnett v. Dairies Inc, v. Farrow, 508 F.2d 1268 (5th Cir.
1978); James v. Stockham Valves & Fitting Co. 559 F.2d 310, 314 
n.1 (5th Cir. 1977). Louis Dreyfus and Cie v. Panama Canal Co.,
298 F.2d 733, 737 (5th Cir. 1962).
31/ International Controls Corp. v. Vesco, 490 F.2d 1334, 1341 
n.6 (2nd Cir. 1974); In re Feara Mir Candy Corporation, 432 
F.2d 1060, 1062 n.2 (2nd Cir. 1970). Schlensky v. Dorsey, 574 
F.2d 131, 148-49 (3rd Cir. 1978). Askew v. United States, 680 
F.2d 1206, 1207-08 (8th Cir. 1982); Bradley v. Maryland Casulty 
Co., 382 F.2d 415, 422-23 (8th Cir. 1967). Ramey Construction 
Co. v. Apache Tribe, 616 F.2d 464, 466 (9th Cir. 1980); Kelson v. 
United States, 503 F.2d 1291, 1294 (9th Cir. 1974).
32/ The notes of the Advisory Committee on Rule 52(a) Fed. Rules 
Civ. Proc. 28 U.S.C. further establish that "these findings 
should represent the judge's own determination and not the often 
argumentative statements of successful counsel."

34



The courts have primarily considered two criteria in deter­
mining whether the a court has mechanically adopted a party's 
proposed findings of the fact: first, whether the wording of the
judge's opinion follows counsels' findings verbatim, or with only 
a few substantially insignificant changes; and second, the rela­
tive number of findings so adopted. The greater the number of such 
findings, the less willing the appellate courts have been to con­
clude that the trial judge independently considered the issues in 
writing the opinion. U.S. v. Marin, 651 F.2d 28, n.10 (1st Cir. 
1981). The courts have become particularly suspicious when por­
tions of the findings which obviously should have been modified
upon adoption by the court were not. See, e.g., Ramey Construc-

33/tion Co. v. Apache Tribe, 616 F.2d 464,467-68 (9th Cir. 1980).
There are three basic reasons why Rule 52(a) requires the 

trier to find facts specially:
1. ... to aid the appellate court by affording it a clear 

understanding of the ground or basis of the decision 
of the trial court. Wright and Miller, Federal 
Practice and Civil Procedure, S. 2571 at 679 (1971);

2. ... to make definite just what is decided by the case 
in order to apply the doctrines of estoppel and res 
judicata in future cases, id. and

33/ For example, in Amstar Corp. v. Domino's Pizza, Inc., 615 
F.2d 252, 258 (5th Cir. 1980), where the appellate court found 
clear error in one of the trial court's findings, it held:

While the "clearly erroneous" rule of Fed. R. Civ. P.52(a) 
applies to a trial judge's findings of fact whether he 
prepared them or they were developed by one of the parties 
and mechanically adopted by the judge, "we can take into 
account the District Court's lack of personal attention to 
factual findings in applying the clearly erroneous rule."

See also In Re Los Colinas, 426 F.2d 1005 (1st Cir. 1981).

35



3. ... to evoke care on the part of the trial judge in
ascertaining the facts. Id. at 680.

The district court's failure to independently analyze the
facts and prepare findings of fact undermines all three of these
purposes. See, e.g., U.S. v. El Paso National Gas. Co., 376 U.S.
at 656-7 n.4. More importantly, although a court does not have
an obligation to explicitly address every point raised by the
parties, there exists an "undeniable right of the losing counsel
to be assured that his position has been thoroughly considered."
In re Los Colinas, 426 F.2d 1005 at 1008 (1st Cir. 1981).

In summary, in James v. Stockham Valves & Fitting Co., 559
F.2d 310, 314 n.1 (5th Cir. 1977), this Court reviewed a Title VII
action in which the trial judge relied heavily on the prevailing
party's Findings of Fact and found them to be "clearly erroneous"
in many respects. The Court stated:

[T]he appellate court can feel slightly more confident 
in concluding that important evidence has been overlooked 
or inadequately considered when factual findings were 
not the product of personal analysis and determination 
by the trial judge.
In this action, the trial judge's mechanical adoption of the 

prevailing defendant's proposed findings of fact evidences a 
blatant lack of independent consideration of the issues. Thus, 
it is appropriate for this court to disregard the findings. See, 
e.g., FMC Corp. v. Varco Intern., Inc., 577 F.2d 500, 501-502 n.2 
(5th Cir. 1982) citing James. Plaintiffs submit that despite the 
weakness of the trial court's findings of fact the appellate court 
should conduct a thorough review of the entire record in this 
case and dispose of the appeal on its merits. See, e.g. , B.F.

36



Goodrich v. Rubber Products/ Inc., 400 F.2d 401 (6th Cir. 1968).

II. THE STATISTICAL EVIDENCE INTRODUCED BELOW ESTABLISHES 
A PRIMA FACIE CASE OF RACIAL DISCRIMINATION THAT WAS 
UNREBUTTED BY THE DEFENDANTS

A. Introduction
In order to put the discussion of the statistical evidence 

that will follow into proper perspective, it is first necessary 
to recapitulate briefly the specific issues this case presents.

First, this case concerns promotions and promotional oppor­
tunities of professional employees only. The issues of hiring, 
initial assignment, and programs related to non-professional 
personnel are simply not present, despite the discussion of those
issues in the decision of the trial court (or, more accurately,

34/
in the proposed findings of fact filed by the defendants). Thus, 
the statistics presented by the defendants and accepted by the 
court that purport to show a lack of discrimination in hiring 
of Blacks were simply irrelevant to the questions before the court.

Second, virtually all professional employees of the N.L.R.B. 
fall in the categories of Field Agents and Field Attorneys. The 
two jobs are closely related and to a significant extent encompass 
the same duties and subject matter. Therefore, the statistical 
evidence relates to the opportunities of both categories.

Third, the statistical and anecdotal evidence relate to 
two types of promotions. Career ladder advancement from initial

34/ The decision below devoted almost equal space to the discus­
sion of these irrelevant issues and to the relevant statistics. 
Compare Record Excerpts pp. 17-20 with pp. 30, 26-29.

37



hire up to journeymen level, is supposed to take place virtually 
automatically on a regular basis unless N.L.R.B. supervisory and 
administrative personnel intervene. After an employee has reached 
the top rung of the professional career ladder, future advancement 
comes about either through the accretion of further duties, which 
leads to higher grade levels, or competitive promotions into 
higher level supervisory and administrative positions.

Fourth, the net result of both a lower rate of career ladder 
advancement and the lack of any realistic opportunity to advance 
to higher level positions in the agency has resulted in Black 
employees, including ones who have received high ratings, becoming 
discouraged and deciding to move on to other employment where 
opportunities for advancement are available. The combination of 
actual refusals to promote and discouragement of Black employees 
has resulted in the almost total absence of any Black employees 
being in higher level positions in N.L.R.B. regional offices, 
despite significant numbers of Black professional employees form­
ing a potential pool from which such employees may be selected.

B. The Governing Principles
When Congress passed the Equal Employment Opportunity Act of 

1972 it recognized that the issue of employment discrimination 
was more complex, far reaching, and entrenched than had been 
perceived in 1964:

In 1964, employment discrimination tended to be viewed 
as a series of isolated and distinguishable events, for the 
most part due to ill-will on the part of some identifiable 
individual or organization. . . Experience has shown this view to be false.

38



Employment discrimination as viewed today is a far 
more complex and pervasive phenomenon.

S. Rep. No. 92-415 (92nd Cong., 1st Sess., 1971) p. 5.
With regard to agencies of the federal government Congress

found in the concentration of Blacks in the lower grade levels
evidence both of employment discrimination and of the failure
of existing programs to bring about equal employment opportun-

35/ity. The present case presents the same pattern that led 
Congress to extend Title VII to federal agencies; Blacks are 
largely relegated to lower positions, regardless of their qualifi­
cations and capabilities.

Title VII, of course, is based on the fundamental proposition 
that in an employment system that is fair and neutral with regard 
to race, one would expect to see persons receiving employment 
benefits on an equal basis irrespective of their race. Thus, if 
the issue is hiring, one would expect to see a workforce reflec­
tive of the workforce from which employees are hired. Teamsters

35/ The House Report stated:
Statistical evidence shows that minorities and women 

continue to be excluded from large numbers of government 
jobs, particularly at the higher government levels ....
This disproportionate distribution of minorities and women 

throughout the Federal bureaucracy and their exclusion from 
higher level policy-making and supervisory positions 
indicates the government's failure to pursue its policy of 
equal opportunity.

H. Rep. No. 92-238 (92nd Cong., 1st Sess., 1971) p. 23. The 
Senate report also included statistics which showed the 
concentration of minorities in the lower grade levels, and 
concluded that this indicated that their ability to advance to 
the higher grade levels had been restricted. S. Rep. No. 92-415 
(92nd Cong., 1st Sess.) pp. 13-14.

39



v. United States, 431 U.S. 324, 339 n. 20 (1977). If the issue
is internal promotions one would expect over a period of time to 
see Blacks distributed fairly throughout the workforce. Davis 
v. Califano, 613 F.2d 957, 963-64 (D.C. Cir. 1980). Indeed, it 
was this expectation and its disappointment that led Congress 
to conclude that minority federal employees suffered from employ-

36/ment discrimination and that corrective action was needed.
The burden on plaintiffs in a Title VII action is not, and

never has been, to disprove every conceivable explanation for
a maldistribution of Blacks in the workforce, but to show patterns
which demonstrate that the underlying presumptions of Title VII
are not met. Blacks are disproportionately in lower grades; they
advance at slower rates; fewer are promoted and they are underrep-

37/resented in supervisory and managerial positions. Upon such
a showing, the burden shifts to the defendant employer to come
forward with legally sufficient reasons for these disparities,
maldistributions, and inequities. An employer cannot simply sit
back and demand that the plaintiffs counter every speculative

38/explanation that may be invented by a fertile mind.

36/ See S. Rep. No. 92-415, supra, pp. 5-6.
37/ See Hazelwood School District v. United States, 433 U.S. 
299, 307 (1977); Teamsters v. United States, 431 U.S. 324, 
336-338 (1977); Barnett v. W. T. Grant Co., 518 F.2d 543,
549 (4th Cir. 1975); Segar v. Civiletti, 508 F. Supp. 690 
(D.D.C. 1981); Grant v. Bethlehem Steel Corp., 635 F.2d 1007, 1018 (2nd Cir. 1980).
38/ As one Court has put it:

When a plaintiff submits accurate statistical data, 
and a defendant alleges that relevant variables are

40



In sum, the thrust of Title VII is to provide an effective 
remedy to correct the historical denial to Blacks of equal 
opportunity and a fair share of employment benefits. It is a 
remedial statute and must be construed and applied in light of 
the problems it was passed to address and correct. As we will 
now show, the statistical evidence in this case establishes a 
consistent pattern of disparate treatment of Black professional 
employees at the N.L.R.B. which requires the conclusion that 
Title VII has been violated.

C. The Undisputed Evidence Establishes Discrimination in
Promotions to Supervisory Positions
As set out in the Statement of Facts above, it is clear

that virtually all promotions to supervisory and higher level
administrative and managerial positions have come from within the
agency. It is well established that the relevant labor force,
and therefore comparison group, is the internal labor force of 

39/
the employer. The undisputed evidence taken from the defendants

38/ continued
excluded defendant may not rely on hypothesis to lessen the 
probative value of plaintiffis statistical proof. Rather, 
defendant in his rebuttal presentation, must either rework 
plaintiffis statistics incorporating the omitted factors or 
present other proof undermining plaintiffis claims.

Segar v. Civiletti, 508 F. Supp. 690, 712 (D.D.C. 1981), citing 
Davis v. Califano, 613 F.2d at 964. See Hazelwood School 
District v. United States, supra; Teamsters v. United States, 
supra; Grant v. Bethlehem Steel Corp., 635 F.2d 1007, 1015 
(2nd Cir. 1980); Pegues v. Mississippi State Employment 
Service, 699 F. 2d 760, 769 (5th Cir. 1983); EEOC v. American 
National Bank, 652 F.2d 1176, 1186-89 (4th Cir. 1981).
39/ Payne v. Travenol Laboratories, 673 F.2d 798 (5th Cir. 
1982); Johnson v. Uncle Bents, Inc. 628 F.2d 419, 425 (5th Cir. 
1980); Davis v. Califano, 613 F.2d 957 (D.C. Cir. 1979).

41



own affirmative action plans demonstrates that as of 1979
approximately 10.5% of the professional employees at the N.L.R.B.

40/
regions were Black. It is also clear that promotions to super-

41/
visory positions come from both Field Attorneys and Field Agents.

The same affirmative action plans demonstrate that Blacks
have comprised at best only 4.4% of those in supervisory posi- 

42/tions. Under the general rule set out in Hazlewood School
District v. United States, 433 U.S. 299 (1977), this disparity
establishes a prima facie case of discrimination since the actual
number of Black supervisors deviate from the expected number of

43/
Black supervisors by 3.72 standard deviations. Put in another 
way, the chances are less than 2 in ten thousand that such a 
disparity could be the result of chance alone. Under Hazelwood, 
the inference of discrimination must be drawn and the burden is 
shifted to the employer to come forward with a satisfactory 
explanation for the observed disproportion.

When one looks to promotions by year similar disparities 
exist. Set out in the table on p. 19, n.14, which is also 
derived from defendants' own affirmative action plan (PX 9, pp. 
64-65; PX 10, Table 1), is the number of promotions and number 
and percentage of Black professionals promoted each year from

±0/ PX 8, pp. 35-36; PX 10, Table 3.
41/ DX 217, pp.24-25.
42/ PX 8; PX 10, Table 3.
43/ See supra, p. 18 for the Hazlewood computation. The 
result is the same using the formula approved by this Court 
in a number of cases, most recently Pegues v. Mississippi 
State Employment Serv., 699 F.2d 760, 768, n. 9 (1983). The

42



1973 through 1980. In 1974 and 1975, the two years prior to this
proceeding, no Blacks were promoted to supervisory positons despite

_44/
there being 117 positions available. Even after the lawsuit the
proportion of Blacks promoted remained substantially below their

45/representation in the eligible work force.
With regard to higher level managerial positions in the 

N.L.R.B., those of regional director, assistant regional director

43/ continued 
formula is:

NP - X
Standard deviations = \/ NPQ

Here the figures are: N = Total number of supervisory posi­
tions; P = percentage of Black eligible professionals; X = 
number of Black supervisors; Q = percentage of white eligi­ble professionals.

317 x 10.5% - 13 = 20.285 = 3.72
SD = v/317 x 10.5% x 89.5% 5,46

44/ Using the Pegues formula, the disparity between expected 
Black and actual promotions is at the level of 3.61 standard 
deviations. The figures used are N = 117 supervisory postions;
P = 10% Black available labor force; X = 0 Blacks selected; Q = 90% white labor force.

SD = NP - X 117 x 10% - 0 11.7
\/ NPX~ = s/ 117 x 10% x 90% = 3,24 = 3,61

45/ For the period 1976-1979, there were 379 promotions to higher 
level positions, 19 of which went to Blacks (n. , supra). Usinq
the Pegues formula, N = 379, P = 10%, X = 19, and Q = 90%.

SD = 379 x 10% - 19 = 18.9 = 3.24 standard deviations 
n/379 x 10% x 90% 5.84

For the entire period 1974-79, 
X = 19% and Q = 90%. SD =496 x 10!
deviations. J496 x 10%

N = 496 promotions 
- 19 = 30.6 =
x 90% 6.68

P = 10%,
4.58 standard

43



regional attorney and assistant regional attorney, the statistics 
are even more stark. In the entire history of the National Labor 
Relations Board since 1933 only three Blacks have every held any 
of those positions, and never more than one at a single time. 
Therefore, out of the approximately 100 such positions available 
in the varying regional offices at any one time at most only one 
has been held by a Black.

In response to these statistics, many of which are at or 
near the famous "inexorable zero", the defendants have offered 
no evidence in explanation. Rather, they have resorted to 
precisely the type of speculation and "violent presumptions" that 
as long ago as 1881 the Supreme Court held could not overcome a 
prima facie case of racial discrimination. See Neal v. Delaware, 
103 U.S. 370, 397 (1881).

Thus, defendants have suggested, without offering any evi­
dence in support, that Blacks were not as "well qualified" for 
supervisory positions (see Record Excerpts 30-31; 112-113), 
that positions might have been filled from different areas of
consideration (ibid), or that Blacks were not as interested

46/in accepting supervisory positions (ibid). The court below 
essentially held that plaintiffs had not demonstrated that 
some other reason besides race could have explained the

46/ Such speculations are remarkably similar to those offered 
on numerous occasions in attempts to explain away low levels 
of representation of Blacks on jury roles. See, e.g. , Neal v. 
Delaware, supra at 397 (Blacks not qualified for jury duty); 
Patton v. Mississippi, 332 U.S. 463, 468 (1947)(Blacks not 
qualified or more likely to have criminal records); Hill v.
Texas, 316 U.S. 400, 405-06 (1942); Castaneda v. Partida,
430 U.S. 482, 498-99 (1977)

44



disparities seen in the selection and placement of persons in 
managerial and supervisory positions; but as Hazlewood makes 
clear such was not the plaintiffs* burden.

In Hazlewood the Supreme Court simply held that the relevant 
labor market to examine in determining whether or not Blacks were 
unrepresented in teachers* positions was persons trained as 
teachers. There is no suggestion in Hazlewood that the plaintiff 
had the burden of proving that Blacks who were teachers had equal 
qualifications as did Whites, had similar length of experience 
as teachers, or to demonstrate in another way an exact correspond­
ence between the qualifications of Blacks and Whites. To the 
contrary, the Court clearly cast on the employer the burden of 
"offering evidence concerning wage differentials, commuting 
problems, ... the relative advantages [of the jobs involved]", or 
other reasons that could explain the observed disparities. 433 
U.S. at 317.

Similarly, in Teamsters v. United States, 431 U.S. 324
(1977), the Court looked to the general population, which had the
basic qualifications to be truck drivers, as the relevant compar- 

47/
ison group. Other courts in dealing with internal promotions 
have also held that the relevant comparison group consists of 
those persons that have the minimum qualifications for promotion 
to the position. Thus, in Davis v. Califano, 613 F.2d 957 (D.C. 
Cir. 1979), the court held that the plaintiff had made out a prima 
facie case of discrimination through statistical evidence by demon-

47/ Teamsters was distinguished in Hazlewood only because 
Hazlewood involved a profession that required certain basic min­imum qualifications beyond those held by the general population.

45



strating a disparity between women employees having the minimal 
qualifications for higher level positions and the number of 
women employees holding such positions.

In the present case plaintiffs-appellants have made precisely 
such a demonstration, since the minimum qualification for promo­
tion to higher level supervisory and managerial positions at the
N.L.R.B. is the holding of positions in the lower levels as field

48/
examiners and field attorneys. This case, therefore, is on 
all fours with Payne v. Travenol Laboratories, 673 F.2d 798, 826-27 
(5th Cir. 1982), where this Court held that work force statistics 
should be used to establish a prima facie case of discrimination 
in promotions to higher level supervisory positions. The burden 
was then on the employer to provide an adequate explanation for 
the disparity. Carroll v. Sears Roebuck & Co., 32 FEP Cases 286 
(5th Cir. 1983).

The employer's burden is to come forward with evidence which 
will rebut a prima facie case either by showing that the statisti­
cal evidence introduced by the plaintiffs can be explained or by
demonstrating non-discriminatory reasons for the observed re- 

49/suits. The reasons for this shifting of the burden of proof 
are clear. In an employment case the employer is uniquely able 
to come forward with evidence to demonstrate the reasons for 
particular selections since it is in possession of the records,

48/ See DX 217, pp.15-17; 18.
49/ See Hazlewood School District v. United States, 433 
U.S. at 317; Alexander v. Louisiana, 405 U.S. 625, 631-32 
(1972); Teamsters v. United States, 431 U.S. at 342, n. 24. 
Castaneda v . Partida, 430 U.S. 482, 498-99, 500 (1977); Boykin 
v. Georgia-Pacific, 32 FEP Cases 25, 31 (5th Cir. 1983).

46



is in control of the promotion process, and has the ability to
50/explain it.

Therefore, not only did the district court err in casting 
the burden on plaintiffs to disprove any possible explanation for 
the observed disparities, but the record itself is barren of any 
counter explanation other than racial discrimination. To the con­
trary, one of the key aspects in promotion to higher level positions 
were the ratings received by field examiners and field agents as 
they progressed up the career ladder. The uncontradicted expert 
testimony presented by plaintiffs by its industrial psychologist 
demonstrated that the rating systems used were highly subjective

51/and capable of discrimination. The final selection process 
itself was also highly subjective and not controlled by objective 
standards. Thus, selections for supervisory and higher level 
positions were made at the regions or at N.L.R.B. headquarters 
by White managerial personnel based on a variety of unarticu­
lated standards. See Payne v. Travenol Laboratories, 673 F.2d 
at 824, 827; Phillips v. Joint Legislative Committee, 637 F.2d 
1014, 1026 (5th Cir. 1981); Fisher v. Proctor & Gamble Manuf. Co.,

50/ As the Supreme court cogently noted in a jury discrimination 
case in which the state attempted to rely on a variety of hypo­
thetical explanations for a proven underrepresentation of Blacks:

. . . had there been evidence to contradict the inference 
to be drawn from this testimony, the state would not have 
refrained from introducing it . . .  .

Hill v. Texas, 316 U.S. 400, 405 (1942); Patton v. Mississippi,332 U.S. 463, 468 (1977).
1 1 / R. Vol. XIII 224-25; XIV, 6-7; 10-14; 16-18; PX 6 (Record Excerpts 179-86).

47



613 F.2d 527, 544 (5th Cir. 1980); James v. Stockham Valves &
Fittings Co., 559 F.2d 310, 330 (5th Cir. 1977).

Defendants could have met the prima facie case by a counter
statistical showing. For example, they could have, if the data
were available, demonstrated that, although the pool of persons
eligible for promotion was 10.5% Black, those actually expressing
an interest in accepting such positions was nearer the proportion
of Blacks selected. Again, this type of information, if it
existed, was uniquely within the possession of the defendants and
it was their burden to come forward with it. They did not do so
but simply resorted to unfounded and unsupported speculation as

52/
to why the disparity existed. Carroll v. Sears Roebuck & Co., 
supra.

For all of the above reasons, therefore, the district court 
erred as a matter of law in failing to hold that the plaintiffs 
had demonstrated a prima facie case of discrimination in the se­
lection of persons for higher level supervisory and managerial 
positions at the National Labor Relations Board. Since the evi­
dence with regard to this issue is clear and undisputed, judgment 
should be entered for the plaintiffs with regard to this issue.

52/ The district court also clearly erred in attempting to 
explain the disparity by holding that Blacks who were quali­
fied for these positions left the agency for higher paid jobs.
The testimony in this case is clear that Blacks left because 
they had become convinced that they could not advance further in 
the N.L.R.B. The fact that they sought higher paid positions 
elsewhere does not in any way disprove the fact that they did so because they had no expectation of receiving higher paid higher 
level positions at the agency. That expectation was fully 
supported by the statistical evidence which demonstrated that the 
number of Blacks holdings such positions was far below their 
representation in the work force of those eligible to receive 
such positions.

48



D. The Evidence Demonstrates A Racial Disparity in The Rates 
of Promotion of Professional Employees
Initial promotions of Field Agents and Field Attorneys are 

noncompetitive through a career ladder up to a certain grade 
level in each job category. The highest career ladder levels 
were different with Field Agents than Field Attorneys; there­
fore, both parties analyzed the statistical evidence relating to 
rates of progression for the two groups separately. Even though 
the statisticians used different methods there was substantial 
agreement that there were differences in rates of progression 
between Blacks and Whites, and that in a number of instances 
those differences were statistically significant.

Plaintiffs# expert, Dr. Bruce Levin, concluded that
there were disparities large enough to indicate discrimination53/
based on race. His analysis looked at the two job
categories separately, examining the rate at which Blacks and

54/Whites moved from each grade level to the one higher. For 
example, since field agents were hired in at the GS-5 level, 
plaintiffgs expert looked at the beginning dates of blacks and 
Whites entering GS-5, the date at which Blacks and whites were 
promoted from GS-5 to GS-7, and then determined whether the

53/ The expert used a method originally developed in the biolog­
ical sciences called a "survival analysis." Survival analysis is 
particularly suited to examine differences in lengths of time to 
promotion from one level to another since it accounts for persons 
who leave a position before they are promoted and persons who 
reach a certain level and are never promoted beyond it. See R. 
Vol. X 47; 51-53; 55-58; 90-91, for a detailed explanation of 
survival analysis.
54/ R. Vol. X 57. Thus, the methodology used by Dr. Levin 
avoided the flaw found by this court in Pouncy v. Prudential 
Ins. Co. of America, 668 F.2d 795, 803-4 (5th Cir. 1982), since 
it examined the time it took Black and White employees to 
progress from one level to the next, level by level.

49



difference in time if any between the two groups to promotion was
statistically significant. The expert found statistical signifi-

55/
cance in a number of grade levels in both professional series.

Dr. Levin also compared the percentages of Blacks and Whites
ever promoted out of each grade level and found a consistent pat-

56/
tern of relatively fewer Blacks being promoted. The overall 
result has been that Blacks advance at a lower rate and have

57/
lower average grade levels than comparable White professionals.

The defendants' expert substantially agreed with the con­
clusions of the plaintiff, although his methodology was differ- 

58/
ent. Thus, he testified that indeed the overall statistics 
showed statistically significant differences in the rates of pro­
gression in career ladder when Black and White field agents and

59/
field attorneys were compared. He sought to explain away

55/ For field agents, Blacks remained in grades longer than 
whites in GS-levels 5, 7, 11, and 12, with statistically 
significant differences in levels 5 and 11 at a probability 
level of less than 1 in 100 (PX 5, Table 1). Moreover, the 
differences overall were greater in the period before 1977 
(PX 5, Table 2). For field attorneys, a similar pattern is 
present with a statistically significant difference at Grade 
12. (PX 5, Table IV.)
56/ For field examiners, the difference in percentage promoted 
was statistically significant at GS-5 and 11 at a probability 
of less than 1 in 100 (PX 5, Table 1). For field attorneys the 
differences at GS levels 7, 9, and 13 were significant at a prob­
ability of less than 5 in 100 (the 2 standard deviation level).
57/ On the average, Black field agents and field attorneys 
were one grade level below Whites. See p. 16-17, supra.
58/ Defendants' expert looked at groups of employees at a par­
ticular GS level in a specific year. He then compared the grade 
levels of Blacks and Whites in a series of later years. His study 
was, in essence, one based on cohorts. See Segar v. Civiletti,
508 F.Supp. 690, 697-98 (D.D.C. 1981), for a cohort study disap­
proved because it tended to fragment the data into small groups.
The studies here and in Segar were prepared by the same statistical 
consulting firm, Criterion, Inc. (R. XV, 191.)
5 9 /  R. V o l .  XV, 2 2 3 - 2 4 ;  XVI,  1 3 - 1 5 ;  3 0 - 3 2 ;  7 4 - 7 5 .

50



these differences by fragmenting the data down into such small
6CI/groups so that the statistical significance would disappear, 

and by making artificial dividing lines between groups of agents 
and attorneys.

For example, he asserted that if one looked at most of the
field agents and field attorneys, Black or White, they progressed
at approximately the same rate. The differences in the overall
progression of Blacks and Whites, he testified, were due to a
group of Blacks progressing substantially slower than others.
Therefore, he separated out this group of Blacks and came to the
conclusion that there was no discrimination because with this

61/group excluded the differences in rates disappeared.
Since the district court accepted this line of reasoning, 

the issue here is whether, as a matter of law, differences in 
rates of progression between two groups which are statistically 
significant to the point of raising an inference of discrimina­
tion may be explained away by pointing out that the differences 
were the result of the treatment of one part of the protected 
group, rather than differences in the treatment of all its members. 
The defense, if accepted, would allow an employer to discriminate 
against part of its Black employees without hindrance as long as 
the employer treated other Blacks the same as the employer treated 
its White employees. Such a result of course, is inconsistent 
with the purpose of Title VII, which is intended to protect all

60/ Thus, there were at most four Blacks in any one cohort, 
and many had 1, 2, or 3 Blacks. Nevertheless, there was 
statistical significance at the .05 level in a substantial number of instances. (DX 201, R. Vol. XV, 222-23.)
61 /  R. V o l .  XVI,  6 1 - 6 5 ;  6 6 - 7 5 .

51



and also confuses whether dis-employees from discrimination,
crimination has been proven against a group with whether relief
will be available to all members of the group at the second stage
in the proceedings. See also Boykin v. Georgia-Pacific Corp.,
32 FEP Cases 25, 31 (5th Cir. 1983).

The defense is also basically inconsistent with Connecticut
v. Teal, ____U .S. ___, 73 L.Ed.2d 130 (1982), which holds that the
fact that there was no discrimination against some part of a
group does not excuse discrimination at another stage in the
process or against another segment of a protected group. In
short then, the attempted defense of the defendants is no defense
at all, because it amounts to excusing differential treatment of
one part of the workforce by a showing that others are treated on
an equal basis with the favored group.

Further, as a factual matter the defendants' showing was
fatally flawed. It was clear from the testimony of defendants'
expert that the dividing line between the group asserted to be
treated the same as White employees and the group from which the

63/
difference overall arose was arbitrarily drawn. Moreover, 
the data was so fragmented as to render many of the expert's 
conclusions suspect. If Defendants' Exhibit 201 is examined, 
it is not surprising that a lack of statistical significance 
was found in most instances, since most involved three or fewer 
Blacks. It is a commonplace that the use of extremely small 
groups reduces the chances of showing statistical significance,

62/ Furnco Construction Co. v. Waters, 438 U.S. 567 (1978).

6 2 /

6_3/ R. V o l .  XVI, 6 1 - 6 5 ;  6 6 - 7 3 .

52



and for this reason, the courts are in agreement that sufficiently 
large enough sample sizes must be used in drawing conclusions 
from statistical evidence. Wheeler v. City of Columbus, 686 
F.2d 1144, 1151 (5th Cir. 1982); Pegues v. Miss. State Employment 
Services, 699 F.2d 760, 769 (5th Cir. 1983); Trout v. Lehman,
702 F.2d 1094, 1106 (D.C. Cir. 1983).

On cross-examination it became clear that if complete data 
had been set out even for the small groups looked at by defend­
ants' expert a consistent pattern emerged of Black employees
with the same anniversary date being at lower grade levels than

6 4 /their corresponding White colleagues. Further, defendants' 
expert seemed to concede that if a summation of all of the small 
groups set out had been made, the result could have been a con­
sistent pattern of a lower rate of advancement of Black employees
as compared to White employees which would have in many instances

65/reached a level of statistical significance. Indeed, plain­
tiff's expert testified to the results of his summation of de­
fendants' data, which revealed statistically significant differ-

66/ences in the advancement of Blacks in many instances.
In short, the expert testimony on both sides demonstrated a 

consistent pattern of Black field examiners and field attorneys 
advancing at a lower rate than did Whites. Moreover, it is clear 
from the testimony of plaintiffs' expert psychologist that the 
primary factor in determining the rate of advancement was the

64/ R. Vol. XV, 214-17; 219; XVI
65/ R. Vol. XVI, 21-22; 28-31.
66/ R. Vol. •

i-~001i—00X

53



subjective and nonvalidated performance appraisal system. Of
course, the decision as to whether a particular employee would
advance on his or her anniversary date was in the hands of the
virtually all-White supervisory and managerial force at N.L.R.B.
The combination of these circumstances thus creates a classic
case of discrimination since, as this court has held on numerous
occasions, a lower rate of promotion of Blacks where there is a
White supervisory force unguided by clear and objective standards

67/
necessarily gives rise to an inference of discrimination.

III. CONTRARY TO THE MANDATES OF THE SUPREME COURT THE DIS­
TRICT COURT FAILED TO ASSESS THE ENTIRE BODY OF EVIDENCE

The lower court's opinion violated the basic mandate of the 
Supreme Court set forth in International Brotherhood of Teamsters 
v. United States, 431 U.S. 324 (1977), and Hazelwood School Dis­
trict v. United States, 433 U.S. 299 (1977) which require that 
the existence vel non of discrimination be assesed from the 
entire record when viewed in a light most favorable to plaintiffs. 
Contrary to these principles , the court below did not so much 
make a cursory analysis of much of the evidence which is pivotal 
to the case.

The parties jointly introduced over 200 documents; of these 
over 100 were presented by the plaintiffs. Yet, the Court's 
Findings of Fact and Conclusions of Law fail to even mention 
much of this probative data; the preponderance of the plain­
tiffs' documentary evidence was virtually undiscussed.

67/ See Payne v. Travenol Laboratories, Inc., 672 F.2d at 824 
and 827, and cases cited therein.

54



One of the most egregious omissions made by the lower court 
was its failure to consider the expert study as well as testimony 
of Dr. James Outtz, who presented a clear, comprehensive and un­
contradicted analysis of the promotion schemes utilized within the 
N.L.R.B. (R. Vol XIII 216-227; R. Vol XIV 2-80; Record Excerpts
179-186) Although the defendants made their expert witness, Dr.

68/Edward J. O'Conner available for a deposition and had him 
present in the courtroom during the trial, neither he nor any of 
the defendants' other witnesses offered one iota of testimony in 
rebuttal to Dr. Outtz. The trial court's findings are a carbon 
copy of the defendants' proposed findings, which mimic the defend­
ants' approach to this litigation. Consequently, the court's
determinations are barren of any discussion of the expert

69/analytical proof presented by Dr. Outtz.
The court's second overt omission of paramount concern

pertains to the extensive documentary evidence presented on
70/behalf of members of the class. (PX 244PL - 413PL) Plain­

tiffs introduced personnel records on behalf of eight class 
members representing nine regional offices and presented a compre­
hensive analysis of this proof in their findings of facts (R. 136). 
Blindly following the lead of the defendants, the lower court

68/ Record Excerpts 10; R. Vol. X 29.
69/ The Findings cite Dr. Outtz in one instance at Finding 
Number 81 (Record Excerpts 81). However, this statement cannot 
be genuinely or reasonably attributed to Dr. Outtz. (R. Vol. Ill 
216-227; R. Vol XIV 2-80). It should be noted that the Court had 
access to a transcript of the testimony of all of the expert 
witnesses. (Record Excerpt 10.)
70/ These documents which were submitted and later withdrawn by 
the defendants were subsequently adopted and admitted by the plaintiffs.

55



completely neglected to discuss, analyze or even mention this 
revealing proof in its findings. This evidence, which is based
solely on the defendants' records, is consistent with the attesta-

11/tions of class members who appeared during the trial. Finally,
the court neglected to consider the depositions of either of the

72/
parties' witnesses who did not appear at the trial.

In failing to assess the record in its entirety the trial 
court failed to take account of numerous practices which are 
detrimental to the advancement of Blacks. As the ignored evidence

71/ a. Blacks upon reaching the journeyman level are uniformly 
denied promotions to expert and supervisory positions. (R. 135
р . 61, PX 289; R. 135 p. 66-67; PX 408; R. 135 p. 72-74, PX 413;
R. 135 P. 80, PX 284, 285; R. 135 p. 68-70; PX 263.)
b. The N.L.R.B., which has never in the history of the agency 
appointed a Black to the position of Assistant Regional Director 
or Regional Attorney, has rejected qualified Blacks for these 
positions. (R. 135 pp.61-62; PX 289.)
с. The appraisal system is infected with arbitrariness in that 
Black employees who receive favorably appraisals from the immedi­
ate or "first line supervisors" are, subsequently rated "not 
ready" by Whites who have never had any direct supervisory 
contact with the employee. (R. 135, p. 66 PX 408; R. 135, p. 
70-71; PX 272; R. 135, p.74; PX 413B; R. 135, p.77, PX 283A.)
d. The criteria used by the predominantly White supervisory 
workforce, in developing appraisals is undefined, subjective and 
non-job related, results in Blacks receiving unfavorable evalu­
ations which preclude them from attaining timely step increases 
and promotions. (R. 135, p. 61-62, PX 289; R. 135, p. 67-68, PX 
32, PX 279; R. 135, p. 72-74, PX 413A and C; R. 135, p. 75-76,
PX 287A, B, C.)
e. Black agents do not receive assignments or guidance which 
would improve their skills and enhance their opportunities to 
advance. (R. 135, p.78, PX 275; R. 135, p. 77, PX 283; R. 135, p. 
78, PX 275; R. 135, p. 79, PX 285.)
72/ Pursuant to Rule 32(a)(3)(1) of the Federal Rules of Civil 
Procedure the depositions of the defendants' witnesses and 
the plaintiffs' were properly introduced. In this regard, al­
though defendants' newly appointed counsel in this case submitted 
the deposition of Dr. James Beckett on March 4, 1983, Dr. Beckett 
did testify and it is appropriate to exclude his deposition from 
the record.

56



indicates, Black employees are evaluated by a predominantly White
73/supervisory staff based on vague and vacillating factors. 

Nevertheless, despite the existence of a multitude of practices 
of this nature, the trial court failed to address their impact.

IV. THE DISTRICT COURT’S ANALYSIS OF THE PROOF IS FATALLY FLAWED

As noted above, a great deal of the opinion is devoted
to a discussion of procedures, programs and policies which are

74/not at issue in this litigation. The Court's analysis of the
N.L.R.B.'s affirmative action program is superficial. Although
Findings of Fact 11, 12, 13 and 15 all relate to the defendant's
alleged affirmative action efforts, even upon a cursory review
of the documents and testimony upon which they were based it
becomes evident that they do not so much as mention the issues
raised in the lawsuit, viz., promotions, assignments and training

75/affecting professionals already employed by the agency.

73/ Practices of this sort have repeatedly been condemned by 
this Court because of their tendency to perpetuate racially-dis- 
criminatory promotion practices. See, e.g., Robbins v. White- 
Wilson Medical Clinic, 642 F.2d 1531 (5th Cir. 1981); Johnson v. 
Uncle Ben's Inc., 628 F.2d 419 (5th Cir. 1980); Payne v. Travenol 
Laboratories, Inc., 565 F.2d 895 (5th Cir. 1978); Rowe v. General 
Motors Corp., 457 F.2d 348 (5th Cir. 1975).
74/ For example, Findings of Fact Numbers 4, 5, 6, 7, 11 and 14, 
which all pertain to hiring policies or procedures are clearly 
not in dispute in this litigation. Findings of Fact Numbers 8 
and 9, which relate solely to the programs, that only affect 
non-professional employees are not at all germane to the action. 
Similarly, Finding of Fact Number 10 extolling the virtures of 
the Cooperative Education Program, which was specifically ex­
cluded from consideration by the trial judge himself in the 
court's class action certification, obviously has no bearing on this case.
75/ Clearly, the only explanation for this void in contrast to 
the plethora of affirmative action other efforts, is that no such

57



This Court has consistently held that a defendant's efforts
to rectify its discriminatory actions subsequent to the institution
of a Title VII suit do not constitue a persuasive defense. See
Rowe v. General Motors Corporation, 457 F.2d 348, 349, (5th Cir.
1972), quoting from Cypress v. Newport News General & Nonsectarian
Hosp. Assoc., 375 F.2d 648 (1967). Moreover, in Findings of Fact
121, 155 and 187 the trial court exonerates the defendants on the
basis of formal and informal union or in-house determinations
contrary to the Supreme Court's decision in Alexander v. Gardner

76/
Denver Co., 415 U.S. 36, 49 ( 1974). See, e_^., Kremer v .
Chemical Construction Corp., 456 U.S. 461, 477 (1982); Chandler

77/
v. Roudebush, 425 U.S. 630 (1976).

V. THE NAMED PLAINTIFF, DONALD R. LEWIS, ESTABLISHED A PRIMA 
FACIE CASE OF DISCRIMINATION WHICH WAS UNREBUTTED BY THE 
DEFENDANTS
Preliminarily, we must emphasize that the trial court 

erred in rejecting plaintiff's individual claim because it erred 
in finding no discrimination in promotions against Blacks as a

75/ Contd.activities have taken place; hence there are no reports reflecting 
such. (R. XIV 109-112; PX 289; DX 228, 229, 230, 231, 232, 233, 
234, 235 and 236.
76/ Similarly, appellate courts which have decided the issue 
have concluded that an arbitration decision did not preclude a 
plaintiff from seeking remedies under a Title VII action. See, 
e.g., Allen v. General Felt Industries, 661 F.2d 135, 136 (9th 
Cir. 1981).
77/ In any event, the court's ruling with respect to this issue 
is defective in that the Rule 408 of the Federal Rules of Evi­
dence specifically prohibits the court from considerating 
evidence as to settlement offers and rejections. See generally 
J. Weinstein & M. Berger, Weinstein's Evidence, at 408(01]- 
[02](1981).

58



class. Of course, once a finding of class discrimination has
been made, the burden is on the employer to prove by clear and
convincing evidence that individual decisions were made free

78/of any taint of bias. For the reasons discussed below, defen­
dants made no such showing with regard to the named plaintiff. 
Therefore, if this Court holds that class discrimination has been 
proven, plaintiff also must prevail on his individual claim. See 
Donaldson v. Pillsbury Co., 554 F.2d 825, 833 (8th Cir. 1977).

Even treating plaintiff's claim as an individual case,
however, it is clear that he met his burden under the McDonnell

79/Douglas - Burdine standards. When the evidence is examined
in its totality, rather than in fragments as did the trial court,
a finding in his favor is compelled. To begin with, it can not
be seriously denied that plaintiff made out a classic prima facie

80/case of retaliation. (1) He complained he had been denied a 
promotion to GS-13 because of his race; (2) Mr. Baldovin, the per­
son in control of his future advancement, knew of the complaint;

78/ Teamsters v . United States, 431 U.S. 324, 360 (1977).
Trout v. Lehman, 702 F.2d 1094, 1107 (D.C. Cir. 1983); Stewart 
v. General Motors Corp., 542 F.2d 445, 453 (7th Cir. 1976); 
Baxter v. Savannah Sugar Refining Corp., 495 F.2d 437, 444-45 
(5th Cir. 1974).
79/ McDonnell Douglas Corp. v. Green, 411 U.S. 792 
(1973); Furnco Construction Corp. v. Waters, 438 U.S. 567 
(1978); Board of Trustees of Keene State College v. Sweeney,
439 U.S. 24 (1978); Texas Dept, of Community Affairs v. Bur­
dine, 450 U.S. 248 (1981); U.S. Postal Service Bd. of Governors 
v. Aikens, U.S. ___, 75 L.Ed.2d 403 (1983).
80/ The three elements are, (1) protected activity; (2) ad­
verse employment action; (3) a causal connection which may be 
shown by knowledge of the protected activity closely followed 
by retaliatory action. Womack v. Munson, 619 F.2d 1292, 1296 
and n.6 (8th Cir. 1980). See also Payne v. McLemore's Whole­
sale and Retail Stores, 654 F.2d 1130 (5th Cir. 1981).

59



(3) Baldovin acted adversely against him at the earliest oppor­
tunity by failing to give him a promised quality increase and by 
downgrading his appraisal so as to render plaintiff ineligible 
for the next available promotion.

Moreover, the proferred explanations were not legitimate, 
non-discriminatory, or worthy of credence. U.S. Postal Service 
v. Aikens, 75 L.Ed. 2d at 410-411, 412. First, prior to Mr.
Lewis' filing his EEO complaint in March 1975, he had been rated 
highly qualified for promotion, he had received no negative 
comments on his work, he had received a quality increase for 
outstanding work in 1974, and he had been promised another in 
January, 1975. After filing, he was shifted to a new team whose 
leader was instructed to build a record against him. The promised 
quality increase was "forgotten" until he complained to the 
agency EEO officer. Negative comments were placed in his file, 
and the recommendation of his immediate supervisor was, for the 
first time, overruled and he was rated ineligible for a promotion 
at the next rating period in January, 1976. As a result, the 
promotion was given to a White co-worker and plaintiff was not 
promoted until 1978. He has not been promoted since, despite his 
being rated well qualified.

All of the proferred explanations are belied by irrefutable 
documentary evidence. The delay in giving the quality increase 
was stated to be because such increases were usually given in 
June. The records of other non-Black employees conclusively 
demonstrated that such was not the case. The rating of plaintiff 
as not ready for a promotion in January 1976, was sought to be

60



explained away by claiming that he would not have gotten a promo­
tion anyway since the White who was promoted had received his 
first well qualified rating earlier than had plaintiff. This 
was shown not be true by the employee record cards, which demon­
strated that plaintiff had been rated well qualified in January 
1975 (before his EEO complaint was filed), while the other 
employee had been first so rated in August 1975, eight months 
later. Therefore, if plaintiff had continued to be rated well 
qualified in January 1976, he would have been entitled to the 
promotion to GS-13 in February. Since the proferred explanations 
can be shown to be "unworthy of credence" (Aikens, supra) by 
official and uncontestable agency records, plaintiff is entitled 
to a decision in his favor as a matter of law.

VI. THE COURT'S DECISION TO AWARD COSTS TO THE DEFENDANT WAS 
IMPROPER
Subsequent to the entering of the trial courts' Final Judge­

ment on December 1, 1982, the defendants on December 10th filed 
a Motion to Amend or Alter Judgement and Extend Time to Fill Bill
of Costs pursuant to Rule 54(a) and 59(e) of the Federal Rules

81/
of Civil Procedure (R 2). The lower court granted this motion
on December 13th and entered an Order on May 9, 1983, holding that

A review of the Court's findings and conclusions clearly 
supports a finding that Plaintiffs' claims both individu­
ally and as to the class, were baseless ... [t]he Government 
was compelled to expend countless personnel hours and 
significant sums of money to prepare for trial and attempt 
to negotiate a settlement over a period of six years.

81/ The circumstances surrounding the trial courts' expedited 
and injudicious decision to grant defendants motion for costs 
warrants scrutiny. Defendants' motion was filed on December 
10th, and granted by the lower court on December 13th; notably, 
the motion was not received by plaintiff until December 16th 
and the order from the court was received on December 20th.

61



Rule 54(d) of the Federal Rules of Civil Procedure permits 
the taxation of costs to the prevailing party, but the courts must 
exercise sound discretion in their awarding. Delta Air Lines, Inc 
v. August, 450 U.S. 346 (1981). In accordance with these precepts 
there are several legal, statutory, and administrative impediments 
to the government receiving costs.

In 1978 the Department of Justice issued a directive pro­
mulgating guidelines governing the federal government in seeking 
costs in Title VII actions. The directive precludes an award of 
costs to the defendants (R 58-59). The guidelines require that 
the standards decreed in Christianburg Garment Co. v. Equal
Employment Opportunity Commission, 434 U.S. 412 (1978) be followed

82/
when the government seeks costs. Accordingly, costs should not
be sought unless: (1) the plaintiff instituted the action in bad
faith; (2) the plaintiff pursued the claim in a harassing or
vexatious manner; (3) the plaintiff litigated a claim which was

83/
patently groundless or frivolous.

In response to plaintiffs' contention that the April 14,
1978 Memorandum precluded the defendants from seeking costs unless 
the Christianburg standards have been met, defendants simply 
asserted that the Memorandum as issued by the United States 
Department of Justice had been rescinded. When plaintiffs replied

82/ The government memorandum cited with approval the decision 
of the District Court in Jaspers v. Alexander, 15 FEP Cases 1238 
(D.D.C. 1977), which applied similar standards in deciding not to 
award costs.
83/ The government is bound by its own regulations and guidelines 
See United States ex. rel. Accardi v. Shaughnessy, 347 U.S. 260 
(1954), Service v. Dulles, 354 U.S. 363 (1957); Vitarelli v. 
Seaton, 359 U.S. 535 (1959); United States v. Nixon, 418 U.S. 683 
(1974). To allow the defendant government to discard its guide-

62



that such a bald assertion was insufficient to contradict or
nulify a written and formal agency directive (R 27), defendants
produced an affidavit from J. Paul McGrath, Assistant Attorney
General for the Civil Division proclaiming that the government
had formally rescinded the April 14, 1978 directive as of a
directive dated April 1, 1983, after the costs here were incurred.
(R. Vol. IV, 15-17.) Plaintiffs appropriately responded by citing

84/
decisions from the Supreme Court which established that the
defendant's contemporaneous renouncement of its policy should not
be effectuated in the present action, since to do so would
perpetuate a manifest injustice on the plaintiffs who relied on

85standards ennunciated in the 1978 Memoranda.
The exercise of statutory power to grant fees to a prevail­

ing Title VII defendant must be applied under a much stricter
standard than that applied when considering awards to prevail-

86/ing plaintiffs. District courts must not "engage in post hoc

83/ Contd.
TTnes in this case would be particularly inappropriate in that 
the memorandum was formulated as a pairt of a settlement agreement 
to resolve the question generally of the appropriate standard to 
govern the award of costs in actions brought pursuant to Title
VII. Thomas v. Department of the Navy, No. 77-1965 (D.D.C.,
1978)(R 31-34.)
84/ Green v. United States, 376 U.S. 149 (1964); See, e.g. Hutto 
v. Finney, 437 U.S. 678 (1978); Bradley v. Richmond School Board, 
416 U.S. 696 (1974); Thorpe v. Housing Authority of Durham, 393 
U.S. 268 (1969); United States v. Schooner Peggy, 1 Circuit 103 
(1801) .
85/ See Hughes v. Heylt Patterson, Inc., 647 F.2d 452, 454 (4th 
Cir. 1981); Iowa Public Service Co. v. I.C.C., 643 F.2d 542 (8th 
Cir. 1981); Smart v. Porter Paint Co., 630 F.2d 490, 497 (7th 
Cir. 1980)
86/ See, e,g. , Stenseth v. Greater Fort Worth, etc., 673 F.2d 842 
(5th Cir. 1982); United States v. Terminal Transport Co., Inc.,

63



reasoning by concluding that, because a plaintiff did not ulti­
mately prevail, his action must have been unreasonable or without 
foundation. . . 434 U.S. at 421-22." Jones v Texas Tech Univer­
sity, 656 F. 2d 1 137, 1 145 (5th Cir. 1981), citing Christianburg 
434 U.S. at 421-22.

The lower court in its final judgement neglected to discuss
the effect of the 1978 Memorandum and simply concluded that this
lawsuit falls under the Christianburg standards. (R. Excerpts
176-78) This determination is out of kilter with the facts in
the record and is incongruous with the court's previous actions

87/in the litigation. It cannot be seriously argued that
plaintiffs' positions were legally barren or factually unsub­
stantiated, and particularly in light of the court's defective 
final judgement, there is absolutely no indication that the 
Court made the inquiry required under Christianburg. See, e,g., 
Anthony v. Marion City General Hospital, 617 F.2d 1164 (5th 
Cir. 1980). Notably, the defendants never sought costs on

86/ Contd.
653 F.2d 1016, 1021-1022 (5th Cir. 1981); Jones v. Texas Tech. 
University, 656 F.2d 1137 (5th Cir. 1981); Luna v. Inter­
national Assn, of Machinists and Aerospace Workers Local No.
36, 614 F.2d 259 (5th Cir. 1980). See also Sanders v. Monsanto 
Co., 529 F.Supp. 704 (S.D. Tex. 1981); EEOC v. Autumn Leaves 
Nursing Home, 521 F.Supp. 1052 (S.D. Miss. 1981); Mosley v. St. 
Louis Southwestern Ry. Co., 22 FEP Cases 835 (E.D. Tex. 1980).
87/ 1. The trial court certified the class from the bench at the
conclusion of an extensive evidentiary based on documentary and 
testimonial evidence obtained during pre-trial discovery (R. Ex­
cerpts p.5). 2. The court did not grant defendants' motion
to dismiss presented at the conclusion of the plaintiff's case.
(R 105). 3. The court's own findings of fact clearly state that 
certain aspects of the case support a showing of racial discrimina­
tion with respect to plaintiff's statistical proof and testimonial 
evidence. See, e,g., Finding of Fact Nos. 66 and 189; Conclusions 
of Law Nos. 11 and 12. 4. The lower court failed to consider sub­
stantial aspects of plaintiff's evidence. The court's assessment 
of the case and final judgement which formed the basis of the 
court's subsequent determination to award costs to defendants was

64



Christianburg grounds but' only on the basis of the revocation of
the 1978 directive.

Finally, Plaintiffs submit that to retroactively place the 
burden of costs on the unsucessful plaintiff, who is the only 
person capable of instituting a Title VII action against the 
federal government, would have a "chilling effect" and thus frus­
trate the sound Congressional policy of uprooting discrimination.

Conclusion
This Court may set aside the district court's findings of 

fact, which are clearly erroneous. A finding is clearly erroneous 
if it is not supported by substantial evidence, if it evolves 
from an erroneous conception of the applicable law, or if the 
reviewing court upon analyzing the entire record is left with a 
definite and firm conviction that an error has been made. See 
United States v. Gypsum Co., 333 U.S. 364 (1948). The trial 
court's decision is premised on fatal errors in reasoning and 
clearly erroneous factual findings which warrant a reversal 
of the decision below and remand of the proceedings to the dis­
trict court with the direction that it grant plaintiffs-appel- 
lants' request for declaratory injunctive and affirmative relief. 
In accordance with Rule 53(a) the appellate courts have mandated 
such a result. See, e.g. , Carroll v. Sears Roebuck & Co., 32 FEP 
Cases 286, 296-97 (5th Cir. 1983); United States v. Georgia Power 
Co., 695 F.2d 890, 892 (5th Cir. 1983).

86/ Contd.
egregriously defective. Consequently, the cost determination as 
based on the deficient findings of fact and conclusions of law 
is similarly defective. 5. Other than hearsay testimony barred 
by F.R.Ev. 801(d), there was not one shred of evidence to support 
the court's speculation that Black employees left the agency to obtain more lucrative promotions.

65



Respectfully submitted.

JACK GREENBERG 
CHARLES STEPHEN RALSTON 
GAIL J. WRIGHT

10 Columbus Circle 
Suite 2030
New York, New York 10019 
(212) 586-8397

ATTORNEYS FOR PLAINTIFFS-APPELLANTS

CERTIFICATE OF SERVICE

I hereby certify that I have served the Brief and Record 
Excerpts on counsel for defendants-appellees by Federal Express 
and by United States mail, first class postage prepaid, 
addressed to Javier Aguilar, Esq. and James R. Gough, Esq., 
Assistant U.S. Attorneys, Southern District of Texas, 12000 
Federal Building and U.S. Courthouse, 515 Rusk Avenue, Houston, 
Texas 77002.

//Dated: October t 1983.

Q . dkJyuCounsel cor Appellants



APPENDIX A

FINDINGS OF FACT WHICH ARE "IDENTICAL'' OR
"SUBSTANTIALLY THE SAME" AS 

BY DEFENDANT
THOSE PROPOSED

Paragraph of 
Opinion

Paragraph of Defendant's 
Proposed Findings of Fact

Introduction Introduction
1
2
34
5
6
7
8 
9
10
1112
1314
15
16
17
18
19
20 
21 
22
2324
25

1
2
34
5
6
7
8 
9

10
11
12
1314
15
16 
18
19
20 
22
23
24
2526

27 & 28



Opinion Defendant's Proposed Findingsof Fact
26 29
27 3028 31
29 3230 33
31 34
32 35
33 3634 37
35 3836 3937 40
38 41
39 42
40 4341 44
42 45
43 4644 47
45 48
46 49
47 50
48 51
49 52
50 53
51 54
52 55
53 56
54 57
55 58
56 59
57 60 & 61
58 62
59 63
60 64
61 65
62 66
63 70 & 71
64 78
65 79
66 80
67 81
68 82
69 83 & 84



Opinion Defendant's Proposed Findings
of Fact

70 8871 89i & 9172 9273 9374 9475 9576 9677 9778 9879 10280 10381 104 & 10582 10683 10784 10885 109 & 11086 11187 112 & 11388 11489 11590 116 & 11791 11892 11993 12094 12195 122
96 12397 12498 125
99 126

100 127
101 128
102 129 & 130
103 132104 133
105 134
106 135107 136 & 137
108 138 & 139
109 141110 142
111 143112 146
113 147