Pullman-Standard v. Swint

Opinion by White:

*275
Justice White
delivered the opinion of the Court.
Respondents were black employees at the Bessemer, Ala., plant of petitioner Pullman-Standard (the Company), a manufacturer of railway freight cars and parts. They brought suit against the Company and the union petitioners — the United Steelworkers of America, AFL-CIO-CLC, and its Local 1466 (collectively USW) — alleging violations of Title VII of the Civil Rights Act of 1964,
78 Stat. 263
, as amended, 42 U. S. C. § 2000e
et seq.
(1976 ed. and Supp. IV), and
42 U. S. C. §1981
.
1
As they come here, these cases involve only the validity, under Title VII, of a seniority system maintained by the Company and USW. The District Court found “that the differences in terms, conditions or privileges of employment resulting [from the seniority system] are ‘not the result of an intention to discriminate’ because of race or color,” App. to Pet. for Cert. in No. 80-1190, p. A-147 (hereinafter App.), and held, therefore, that the system satisfied the requirements of § 703(h) of the Act. The Court of Appeals for the Fifth Circuit reversed:
“Because we find that the differences in the terms, conditions and standards of employment for black workers and white workers at Pullman-Standard resulted from an intent to discriminate because of race, we hold that the system is not legally valid under section 703(h) of Title VII, 42 U. S. C. 2000e-2(h).”
624 F. 2d 525
, 533-534 (1980).
*276
We granted the petitions for certiorari filed by USW and by the Company,
451 U. S. 906
(1981), limited to the first question presented in each petition: whether a court of appeals is bound by the “clearly erroneous” rule of Federal Rule of Civil Procedure 52(a) in reviewing a district court’s findings of fact, arrived at after a lengthy trial, as to the motivation of the parties who negotiated a seniority system; and whether the court below applied wrong legal criteria in determining the bona fides of the seniority system. We conclude that the Court of Appeals erred in the course of its review and accordingly reverse its judgment and remand for further proceedings.
I
Title VII is a broad remedial measure, designed “to assure equality of employment opportunities. ”
McDonnell Douglas Corp.
v.
Green,
411 U. S. 792, 800
(1973). The Act was designed to bar not only overt employment discrimination, “but also practices that are fair in form, but discriminatory in operation.”
Griggs
v.
Duke Power Co.,
401 U. S. 424, 431
(1971). “Thus, the Court has repeatedly held that a prima facie Title VII violation may be established by policies or practices that are neutral on their face and in intent but that nonetheless discriminate in effect against a particular group.”
Teamsters
v.
United States,
431 U. S. 324, 349
(1977) (hereinafter
Teamsters).
The Act’s treatment of seniority systems, however, establishes an exception to these general principles. Section 703(h),
78 Stat. 257
, as set forth in 42 U. S. C. §2000e-2(h), provides in pertinent part:
“Notwithstanding any other provision of this subchap-ter, it shall not be an unlawful employment practice for an employer to apply different standards of compensation, or different terms, conditions, or privileges of employment pursuant to a bona fide seniority . . . system . . . provided that such differences are not the result of an intention to discriminate because of race.”
*277
Under this section, a showing of disparate impact is insufficient to invalidate a seniority system, even though the result may be to perpetuate pre-Act discrimination. In
Trans World Airlines, Inc.
v.
Hardison,
432 U. S. 63, 82
(1977), we summarized the effect of § 703(h) as follows: “[AJbsent a discriminatory purpose, the operation of a seniority system cannot be an unlawful employment practice even if the system has some discriminatory consequences.” Thus, any challenge to a seniority system under Title VII will require a trial on the issue of discriminatory intent: Was the system adopted because of its racially discriminatory impact?
This is precisely what happened in these cases. Following our decision in
Teamsters
,
the District Court held a new trial on the limited question of whether the seniority system was “instituted or maintained contrary to Section 703(h) of the new Civil Rights Act of 1964.” App. A-125.
2
That court concluded, as we noted above and will discuss below, that the system was adopted and maintained for purposes wholly independent of any discriminatory intent. The Court of Appeals for the Fifth Circuit reversed.
I — I I — I
Petitioners submit that the Court of Appeals failed to comply with the command of Rule 52(a) that the findings of fact of a district court may not be set aside unless clearly erroneous. We first describe the findings of the District Court and the Court of Appeals.
Certain facts are common ground for both the District Court and the Court of Appeals. The Company’s Bessemer plant was unionized in the early 1940’s. Both before and after unionization, the plant was divided into a number of different operational departments.
3
USW sought to represent
*278
all production and maintenance employees at the plant and was elected in 1941 as the bargaining representative of a bargaining unit consisting of most of these employees. At that same time, IAM became the bargaining representative of a unit consisting of five departments.
4
Between 1941 and 1944, IAM ceded certain workers in its bargaining unit to USW. As a result of this transfer, the IAM bargaining unit became all white.
Throughout the period of representation by USW, the plant was approximately half black. Prior to 1965, the Company openly pursued a racially discriminatory policy of job assignments. Most departments contained more than one job category and as a result most departments were racially mixed. There were no lines of progression or promotion within departments.
The seniority system at issue here was adopted in 1954.
5
Under that agreement, seniority was measured by length of continuous service in a particular department.
6
Seniority was originally exercised only for purposes of layoffs and hirings within particular departments. In 1956, seniority was formally recognized for promotional purposes as well. Again, however, seniority, with limited exceptions, was only exercised within departments; employees transferring to
*279
new departments forfeited their seniority. This seniority system remained virtually unchanged until after this suit was brought in 1971.
7
The District Court approached the question of discriminatory intent in the manner suggested by the Fifth Circuit in
James
v.
Stockham Valves & Fittings Co.,
559 F. 2d 310
(1977). There, the Court of Appeals stated that under
Teamsters
“the totality of the circumstances in the development and maintenance of the system is relevant to examining that issue.”
559 F. 2d, at 352
. There were, in its view, however, four particular factors that a court should focus on.
8
First, a court must determine whether the system “operates to discourage all employees equally from transferring between seniority units.”
Ibid.
The District Court held that the system here “was facially neutral and . . . was applied equally to all races and ethnic groups.” App. A-132. Although there were charges of racial discrimination in its application, the court held that these were “not substantiated by the evidence.”
Id.,
at A-133. It concluded that the system “applied equally and uniformly to all employees, black and white, and that, given the approximately equal number'
*280
of employees of the two groups, it was quantitatively neutral as well.”
Id.,
at A-134.
9
Second, a court must examine the rationality of the departmental structure, upon which the seniority system relies, in light of the general industry practice.
James, supra,
at 352
. The District Court found that linking seniority to “departmental age” was “the modal form of agreements generally, as well as with manufacturers of railroad equipment in particular.” App. A-137. Furthermore, it found the basic arrangement of departments at the plant to be rationally related to the nature of the work and to be “consistent with practices which were . . . generally followed at other unionized plants throughout the country.”
Id.,
at A-136—A-137. While questions could be raised about the necessity of certain departmental divisions, it found that all of the challenged lines of division grew out of historical circumstances at the plant that were unrelated to racial discrimination.
10
Although unionization did produce an all-white IAM bargaining unit, it found that USW “cannot be charged with racial bias in its response to the IAM situation. [USW] sought to represent all workers, black and white, in the plant.”
Id.,
at A-145. Nor could the Company be charged with any racial discrimination that may have existed in IAM:
“The company properly took a ‘hands-off’ approach towards the establishment of the election units .... It bargained with those unions which were afforded repre
*281
sentational status by the NLRB and did s.o without any discriminatory animus.”
Id.,
at A-146.
Third, a court had to consider “whether the seniority system had its genesis in racial discrimination,”
James, supra,
at 352
, by which it meant the relationship between the system and other racially discriminatory practices. Although finding ample discrimination by the Company in its employment practices and some discriminatory practices by the union,
11
the District Court concluded that the seniority system was in no way related to the discriminatory practices:
“The seniority system . . . had its genesis ... at a period when racial segregation was certainly being practiced; but this system was not itself the product of this bias. The system rather came about as a result of colorblind objectives of a union which — unlike most structures and institutions of the era — was not an arm of a segregated society. Nor did it foster the discrimination . . . which was being practiced by custom in the plant.” App. A-144.
Finally, a court must consider “whether the system was negotiated and has been maintained free from any illegal purpose.”
James, supra,
at 352
. Stating that it had “carefully considered the detailed record of negotiation sessions and contracts which span a period of some thirty-five years,” App. A-146, the court found that the system was untainted by any discriminatory purpose. Thus, although the District
*282
Court focused on particular factors in carrying out the analysis required by § 703(h), it also looked to the entire record and to the “totality of the system under attack.”
Id.,
at A-147.
The Court of Appeals addressed each of the four factors of the
James
test and reached the opposite conclusion. First, it held that the District Court erred in putting aside qualitative differences between the departments in which blacks were concentrated and those dominated by whites, in considering whether the system applied “equally” to whites and blacks.
12
This is a purported correction of a legal standard under which the evidence is to be evaluated.
Second, it rejected the District Court’s conclusion that the structure of departments was rational, in line with industry practice, and did not reflect any discriminatory intent. Its discussion is brief but focuses on the role of IAM and certain characteristics unique to the Bessemer plant. The court concluded:
“The record evidence, generally, indicates arbitrary creation of the departments by the company since unionization and an attendant adverse affect
[sic]
on black workers. The individual differences between the departmental structure at Pullman-Standard and that of other plants, and as compared with industry practice, are indicative of attempts to maintain one-race departments.” 624 F. 2d, at 532.
In reaching this conclusion, the Court of Appeals did not purport to be correcting a legal error, nor did it refer to or expressly apply the clearly-erroneous standard.
*283
Third, in considering the “genesis” of the system, the Court of Appeals held that the District Court erred in holding that the motives of IAM were not relevant.
13
This was the correction of a legal error on the part of the District Court in excluding relevant evidence. The court did not stop there, however. It went on to hold that IAM was acting out of discriminatory intent — an issue specifically not reached by the District Court — and that “considerations of race permeated the negotiation and the adoption of the seniority system in 1941 and subsequent negotiations thereafter.”
Ibid.
Fourth, despite this conclusion under the third
James
factor the Court of Appeals then recited, but did not expressly set aside or find clearly erroneous, the District Court’s findings with respect to the negotiation and maintenance of the seniority system.
The court then announced that “[hjaving carefully reviewed the evidence offered to show whether the departmental seniority system in the present case is ‘bona fide’ within the meaning of § 703(h) of Title VII, we reject the district court’s finding.” 624 F. 2d, at 533. Elaborating on its disagreement, the Court of Appeals stated:
“An analysis of the totality of the facts and circumstances surrounding the creation and continuance of the departmental system at Pullman-Standard leaves us with the definite and firm conviction that a mistake has been made. There is no doubt, based upon the record in this case, about the existence of a discriminatory purpose. The obvious principal aim of the I. A. M. in 1941 was to exclude black workers from its bargaining unit.
*284
That goal was ultimately reached when maneuvers by the I. A. M. and U. S. W. resulted in an all-white I. A. M. unit. The U. S. W., in the interest of increased membership, acquiesced in the discrimination while succeeding in significantly segregating the departments within its own unit.
“The district court might have reached a different conclusion had it given the I. A. M.’s role in the creation and establishment of the seniority system its due consideration.”
Ibid,
(footnote omitted).
Having rejected the District Court’s finding, the court made its own findings as to whether the USW seniority system was protected by § 703(h):
“We consider significant in our decision the manner by which the two seniority units were set up, the creation of the various all-white and all-black departments within the U. S. W. unit at the time of certification and in the years thereafter, conditions of racial discrimination which affected the negotiation and renegotiation of the system, and the extent to which the system and the attendant no-transfer rule locked blacks into the least remunerative positions within the company. Because we find that the differences in the terms, conditions and standards of employment for black workers and white workers at Pullman-Standard resulted from an intent to discriminate because of race, we hold that the system is not legally valid under section 703(h) of Title VII, 42 U. S. C. §2000e-2(h).”
Id.,
at 533-534.
In connection with its assertion that it was convinced that a mistake had been made, the Court of Appeals, in a footnote, referred to the clearly-erroneous standard of Rule 52(a).
Id.,
at 533, n. 6.
14
It pointed out, however, that if findings
*285
“are made under an erroneous view of controlling legal principles, the clearly erroneous rule does not apply, and the findings may not stand.”
Ibid.
Finally, quoting from
East
v.
Romine, Inc.,
518 F. 2d 332, 339
(CA5 1975), the Court of Appeals repeated the following view of its appellate function in Title VII cases where purposeful discrimination is at issue:
“‘Although discrimination
vel non
is essentially a question of fact it is, at the same time, the ultimate issue for resolution in this case, being expressly proscribed by 42 U. S. C. A. § 2000e-2(a). As such, a finding of discrimination or non-discrimination is a finding of ultimate fact. [Cites omitted.] In reviewing the district court’s findings, therefore, we will proceed to make an independent determination of appellant’s allegations of discrimination, though bound by findings of subsidiary fact which are themselves not clearly erroneous.’” 624 F. 2d, at 533, n. 6.
III
Pointing to the above statement of the Court of Appeals and to similar statements in other Title VII cases coming from that court,
15
petitioners submit that the Court of Ap
*286
peals made an independent determination of discriminatory-purpose, the “ultimate fact” in this case, and that this was error under Rule 52(a). We agree with petitioners that if the Court of Appeals followed what seems to be the accepted rule in that Circuit, its judgment must be reversed.
16
*287
Rule 52(a) broadly requires that findings of fact not be set aside unless clearly erroneous. It does not make exceptions or purport to exclude certain categories of factual findings from the obligation of a court of appeals to accept a district court’s findings unless clearly erroneous. It does not divide facts into categories; in particular, it does not divide findings of fact into those that deal with “ultimate” and those that deal with “subsidiary” facts.
The Rule does not apply to conclusions of law. The Court of Appeals, therefore, was quite right in saying that if a district court’s findings rest on an erroneous view of the law, they may be set aside on that basis. But here the District Court was not faulted for misunderstanding or applying an erroneous definition of intentional discrimination.
17
It was reversed for arriving at what the Court of Appeals thought was an erroneous finding as to whether the differential impact of the seniority system reflected an intent to discriminate on account of race. That question, as we see it, is a
*288
pure question of fact, subject to Rule 52(a)’s clearly-erroneous standard. It is not a question of law and not a mixed question of law and fact.
The Court has previously noted the vexing nature of the distinction between questions of fact and questions of law.
See Baumgartner v. United States,
322 U. S. 665, 671
(1944). Rule 52(a) does not furnish particular guidance with respect to distinguishing law from fact. Nor do we yet know of any other rule or principle that will unerringly distinguish a factual finding from a legal conclusion. For the reasons that follow, however, we have little doubt about the factual nature of § 703(h)’s requirement that a seniority system be free of an intent to discriminate.
Treating issues of intent as factual matters for the trier of fact is commonplace. In
Dayton Board of Education
v.
Brinkman,
443 U. S. 526, 534
(1979), the principal question was whether the defendants had intentionally maintained a racially segregated school system at a specified time in the past. We recognized that issue as essentially factual, subject to the clearly-erroneous rule. In
Commissioner
v.
Duberstein,
363 U. S. 278
(1960), the Court held that the principal criterion for identifying a gift under the applicable provision of the Internal Revenue Code was the intent or motive of the donor — “one that inquires what the basic reason for his conduct was in fact.”
Id.,
at 286
. Resolution of that issue determined the ultimate issue of whether a gift had been made. Both issues were held to be questions of fact subject to the clearly-erroneous rule. In
United States
v.
Yellow Cab Co.,
338 U. S. 338, 341
(1949), an antitrust case, the Court referred to “[findings as to the design, motive and intent with which men act” as peculiarly factual issues for the trier of fact and therefore subject to appellate review under Rule 52.
Justice Black’s dissent in
Yellow Cab
suggested a contrary approach. Relying on
United States
v.
Griffith,
334 U. S.
*289
100
(1948), he argued that it is not always necessary to prove “specific intent” to restrain trade; it is enough if a restraint is the result or consequence of a defendant’s conduct or business arrangements. Such an approach, however, is specifically precluded by § 703(h) in Title VII cases challenging seniority systems. Differentials among employees that result from a seniority system are not unlawful employment practices unless the product of an intent to discriminate. It would make no sense, therefore, to say that the intent to discriminate required by § 703(h) may be presumed from such an impact. As § 703(h) was construed in
Teamsters
,
there must be a finding of actual intent to discriminate on racial grounds on the part of those who negotiated or maintained the system. That finding appears to us to be a pure question of fact.
This is not to say that discriminatory impact is not part of the evidence to be considered by the trial court in reaching a finding on whether there was such a discriminatory intent as a factual matter.
18
We do assert, however, that under § 703(h) discriminatory intent is a finding of fact to be made by the trial court; it is not a question of law and not a mixed question of law and fact of the kind that in some cases may allow an appellate court to review the facts to see if they satisfy some legal concept of discriminatory intent.
19
Discrimi
*290
natory intent here means actual motive; it is not a legal presumption to be drawn from a factual showing of something less than actual motive. Thus, a court of appeals may only reverse a district court’s finding on discriminatory intent if it concludes that the finding is clearly erroneous under Rule 52(a). Insofar as the Fifth Circuit assumed otherwise, it erred.
IV
Respondents do not directly defend the Fifth Circuit rule that a trial court’s finding on discriminatory intent is not subject to the clearly-erroneous standard of Rule 52(a).
20
Rather, among other things, they submit that the Court of Appeals recognized and, where appropriate, properly applied Rule 52(a) in setting aside the findings of the District Court. This position has force, but for two reasons it is not persuasive.
First, although the Court of Appeals acknowledged and correctly stated the controlling standard of Rule 52(a), the acknowledgment came late in the court’s opinion. The court had not expressly referred to or applied Rule 52(a) in the course of disagreeing with the District Court’s resolution of the factual issues deemed relevant under
James
v.
Stockham
*291
Valves & Fittings Co.,
559 P. 2d 310
(1977).
21
Furthermore, the paragraph in which the court finally concludes that the USW seniority system is unprotected by § 703(h) strongly suggests that the outcome was the product of the court’s independent consideration of the totality of the circumstances it found in the record.
Second and more fundamentally, when the court stated that it was convinced that a mistake had been made, it then identified not only the mistake but also the source of that mistake. The mistake of the District Court was that on the record there could be no doubt about the existence of a discriminatory purpose. The source of the mistake was the District Court’s failure to recognize the relevance of the racial purposes of IAM. Had the District Court “given the I. A. M.’s role in the creation and establishment of the seniority system its due consideration,” it “might have reached a different conclusion.”
Supra,
at 284.
When an appellate court discerns that a district court has failed to make a finding because of an erroneous view of the law, the usual rule is that there should be a remand for further proceedings to permit the trial court to make the missing findings:
“[FJactfinding is the basic responsibility of district courts, rather than appellate courts, and . . . the Court of Appeals should not have resolved in the first instance
*292
this factual dispute which had not been considered by the District Court.”
DeMarco
v.
United States,
415 U. S. 449, 450
, n. (1974).
22
Likewise, where findings are infirm because of an erroneous view of the law, a remand is the proper course unless the record permits only one resolution of the factual issue.
Kelley
v.
Southern Pacific Co.,
419 U. S. 318, 331-332
(1974). All of this is elementary. Yet the Court of Appeals, after holding that the District Court had failed to consider relevant evidence and indicating that the District Court might have come to a different conclusion had it considered that evidence, failed to remand for further proceedings as to the intent of IAM and the significance, if any, of such a finding with respect to the intent of USW itself. Instead, the Court of Appeals made its own determination as to the motives of IAM, found that USW had acquiesced in the IAM conduct, and apparently concluded that the foregoing was sufficient to remove the system from the protection of § 703(h).
23
*293
Proceeding in this manner seems to us incredible unless the Court of Appeals construed its own well-established Circuit rule with respect to its authority to arrive at independent findings on ultimate facts free of the strictures of Rule 52(a) also to permit it to examine the record and make its own independent findings with respect to those issues on which the district court’s findings are set aside for an error of law. As we have previously said, however, the premise for this conclusion is infirm: whether an ultimate fact or not, discriminatory intent under § 703(h) is a factual matter subject to the clearly-erroneous standard of Rule 52(a). It follows that when a district court’s finding on such an ultimate fact is set aside for an error of law, the court of appeals is not relieved of the usual requirement of remanding for further proceedings to the tribunal charged with the task of factfinding in the first instance.
Accordingly, the judgment of the Court of Appeals is reversed, and the cases are remanded to that court for further proceedings consistent with this opinion.
So ordered.
In their original complaint, besides challenging the seniority system discussed in this opinion, plaintiffs also alleged discrimination in job assignments and promotions and the failure to post publicly a list of changes in assignments. These were all brought as “class” issues. Two charges of individual discrimination were also brought. The Court of Appeals held that the Company had violated Title VII in making job assignments and in selecting foremen. In granting certiorari, we declined to review those aspects of the decision.
The procedural history of these cases is rather complex. The original complaint was filed in 1971. Since that time the case has been tried three times and has twice been reviewed by the Court of Appeals.
In 1941, prior to unionization, the Bessemer plant was divided into 20
*278
departments. By 1954, there were 28 departments—26 USW units and 2 International Association of Machinists and Aerospace Workers (IAM) units. The departments remained essentially unchanged after 1954.
The International Brotherhood of Electrical Workers (IBEW) gained representation status for two small departments. The IBEW unit was all white. IBEW, however was decertified in 1946 and its members were reabsorbed into a department represented by USW.
A departmental seniority system was part of the initial collective-bargaining agreement between the Company and USW in 1942. Between 1947 and 1954, however, the seniority system changed from one based on departments to one based upon particular occupations within departments. In 1954, the system went back to a departmental base.
The only exceptions, until 1972 (see n. 7,
infra),
were for employees transferring at the request of the Company or for those electing transfer in lieu of layoff.
In 1972, the Company entered into an agreement with the Department of Labor to bring its employment practices into compliance with Executive Order No. 11246, 3 CFR 339 (1964-1965 Comp.). This provided an exception to the departmental limit on seniority, allowing certain black employees to make interdepartmental transfers without any loss of seniority.
The Fifth Circuit relied upon the following passage in
Teamsters,
431 U. S., at 355
-356:
“The seniority system in this litigation is entirely bona fide. It applies equally to all races and ethnic groups. To the extent that it ‘locks’ employees into non-line-driver jobs, it does so for all. . . . The placing of line drivers in a separate bargaining unit from other employees is rational, in accord with the industry practice, and consistent with National Labor Relation Board precedents. It is conceded that that seniority system did not have its genesis in racial discrimination, and that it was negotiated and has been maintained free from any illegal purpose.”
This passage was of course not meant to be an exhaustive list of all the factors that a district court might or should consider in making a finding of discriminatory intent.
The court specifically declined to make any finding on whether the no-transfer provision of the seniority system had a greater relative effect on blacks than on whites, because of qualitative differences in the departments in which they were concentrated. It believed that such an inquiry would have been inconsistent with the earlier Fifth Circuit opinion in this case.
In particular, the court focused on the history of the unionization process at the plant and found certain of the departmental divisions to be based on the evolving relationship between USW and IAM.
With respect to USW, the District Court found that “[u]nion meetings were conducted with different sides of the hall for white and black members, and social functions of the union were also segregated.” App. A-142. It also found, however, that “[w]hile possessing some of the trappings taken from an otherwise segregated society, the USW local was one of the few institutions in the area which did not function in fact to foster and maintain segregation; rather, it served a joint interest of white and black workers which had a higher priority than racial considerations.”
Id.,
at A-143.
It does not appear to us that the District Court actually found a qualitative difference but held it to be irrelevant. The relevant passage of the District Court opinion read as follows: “By ranking the twenty-eight USW and IAM departments according to some perceived order of desirability, one could . . . attempt to measure the relative effect of the no-transfer rule on white and black employees .... It may well be that a somewhat greater impact was felt by blacks than whites although . . . this conclusion is by no means certain.”
Id.,
at A-134.
The original complaint in this case did not mention IAM. Prior to the first trial, respondents sought and received leave to amend their complaint to add IAM as a Rule 19 defendant, “insofar as the relief requested may involve or infringe upon the provisions of such Union’s collective bargaining agreement with the Company.” Order of the District Court, June 4, 1974 (App. 29).
In
United States
v.
United States Gypsum Co.,
333 U. S. 364, 395
(1948), this Court characterized the clearly-erroneous standard as follows:
*285
“A finding is ‘clearly erroneous’ when although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed.”
We note that the Court of Appeals quoted this passage at the conclusion of its analysis of the District Court opinion.
Supra,
at 283.
See
Jackson
v.
City of Killeen,
654 F. 2d 1181, 1184
(1981);
Payne
v.
McLemore’s Wholesale & Retail Stores,
654 F. 2d 1130, 1147
(1981);
Wilkins
v.
University of Houston,
654 F. 2d 388, 390
(1981);
Lindsey
v.
Mississippi Research & Development Center,
652 F. 2d 488, 492
(1981);
Rohde
v.
K. O. Steel Castings, Inc.,
649 F. 2d 317, 320
(1981);
Joshi
v.
Florida State University,
646 F. 2d 981, 986
(1981);
Phillips
v.
Joint Legislative Committee,
637 F. 2d 1014
, 1024 (1981);
Danner
v.
United States Civil Service Comm’n,
635 F. 2d 427
, 430-431 (1981);
Thompson
v.
Leland Police Dept.,
633 F. 2d 1111, 1112
(1980);
Crawford
v.
Western Electric Co.,
614 F. 2d 1300, 1311
(1980);
Burdine
v.
Texas Dept. of Community Af
*286
fairs,
608 F. 2d 563, 566
(1979);
Williams
v.
Tallahassee Motors, Inc.,
607 F. 2d 689, 690
(1979);
Parson
v.
Kaiser Aluminum & Chemical Corp.,
575 F. 2d 1374, 1382
(1978);
Causey
v.
Ford Motor Co.,
516 F. 2d 416, 420-421
(1975);
East
v.
Romine, Inc.,
518 F. 2d 332, 338-339
(1975).
There is some indication in the opinions of the Court of Appeals for the Fifth Circuit (see n.
15,
supra)
that the Circuit rule with respect to “ultimate facts” is only another way of stating a standard of review with respect to mixed questions of law and fact — the ultimate “fact” is the statutory, legally determinative consideration (here, intentional discrimination) which is or is not satisfied by subsidiary facts admitted or found by the trier of fact. As indicated in the text, however, the question of intentional discrimination under § 703(h) is a pure question of fact. Furthermore, the Court of Appeals’ opinion in this case appears to address the issue as a question of fact unmixed with legal considerations.
At the same time, this Court has on occasion itself indicated that findings on “ultimate facts” are independently reviewable. In
Baumgartner
v.
United States,
322 U. S. 665
(1944), the issue was whether or not the findings of the two lower courts satisfied the clear-and-convincing standard of proof necessary to sustain a denaturalization decree. The Court held that the conclusion of the two lower courts that the exacting standard of proof had been satisfied was not an unreviewable finding of fact but one that a reviewing court could independently assess. The Court referred to the finding as one of “ultimate” fact, which in that case involved an appraisal of the strength of the entire body of evidence. The Court said that the significance of the clear-and-convincing proof standard “would be lost” if the ascertainment by the lower courts whether that exacting standard of proof had been satisfied on the whole record were to be deemed a “fact” of the same order as all other “facts not open to review here.”
Id.,
at 671
.
The Fifth Circuit’s rule on appellate consideration of “ultimate facts” has its roots in this discussion in
Baumgartner
.
In
Galena Oaks Corp.
v.
Scofield,
218 F. 2d 217
(CA5 1954), in which the question was whether the gain derived from the sale of a number of houses was to be treated as capital gain or ordinary income, the Court of Appeals relied directly on
Baum-gartner
in holding that this was an issue of “ultimate fact” that an appellate
*287
court may review free of the elearly-erroneous rule.
Causey
v.
Ford Motor Co., supra,
at 421, relying on
Galena Oaks Corp.
v.
Scofield, supra,
said that “although discrimination
vel non
is essentially a question of fact, it is, at the same time, the ultimate issue for resolution in this case” and as such, was deemed to be independently reviewable. The passage from
East
v.
Romine, Inc., supra,
at 339, which was repeated in the cases before us now,
supra,
at 285, rested on the opinion in
Causey
v.
Ford Motor Co.
Whatever
Baumgartner
may have meant by its discussion of “ultimate facts,” it surely did not mean that whenever the result in a case turns on a factual finding, an appellate court need not remain within the constraints of Rule 52(a). Baumgartner’s discussion of “ultimate facts” referred not to pure findings of fact — as we find discriminatory intent to be in this context — but to findings that “clearly impl[y] the application of standards of law.”
322 U. S., at 671
.
As we noted above, the Court of Appeals did at certain points purport to correct what it viewed as legal errors on the part of the District Court. The presence of such legal errors may justify a remand by the Court of Appeals to the District Court for additional factfinding under the correct legal standard.
Infra,
at 291-292.
See,
e. g., Furnco Construction Corp.
v.
Waters,
438 U. S. 567, 580
(1978): “Proof that [an employer’s] work force was racially balanced or that it contained a disproportionately high percentage of minority employees is not wholly irrelevant on the issue of intent when that issue is yet to be decided.”
We need not, therefore, address the much-mooted issue of the applicability of the Rule 52(a) standard to mixed questions of law and
fact
— i.
e.,
questions in which the historical facts are admitted or established, the rule of law is undisputed, and the issue is whether the facts satisfy the statutory standard, or to put it another way, whether the rule of law as applied to the established facts is or is not violated. There is substantial authority
*290
in the Circuits on both sides of this question. Compare
United States ex rel. Johnson
v.
Johnson,
531 F. 2d 169, 174, n. 12
(CA3 1976);
Stafos
v.
Jarvis,
477 F. 2d 369, 372
(CA10 1973); and
Johnson
v.
Salisbury,
448 F. 2d 374, 377
(CA6 1971), with
Rogers
v.
Bates,
431 F. 2d 16, 18
(CA8 1970); and
Pennsylvania Casualty Co.
v.
McCoy,
167 F. 2d 132, 133
(CA5 1948). There is also support in decisions of this Court for the proposition that conclusions on mixed questions of law and fact are independently reviewable by an appellate court,
e. g., Bogardus
v.
Commissioner,
302 U. S. 34, 39
(1937);
Helvering
v.
Tex-Penn Oil Co.,
300 U. S. 481, 491
(1937);
Helvering
v.
Rankin,
295 U. S. 123, 131
(1935). But cf.,
Commissioner
v.
Duberstein,
363 U. S. 278, 289
(1960);
Commissioner
v.
Heininger,
320 U. S. 467, 475
(1943).
Neither does the dissent contend that Rule 52(a) is inapplicable to findings of discriminatory intent. Rather, it contends, that the Rule was properly applied by the Court of Appeals.
In particular, in regard to the second
James
factor — whether the departmental structure was rational or in line with industry practice — the Court of Appeals did not focus on the evidentiary basis for any particular finding of the District Court. It appeared to make an independent examination of the record and arrive at its own conclusion contrary to that of the District Court. Likewise, in dealing with the genesis of the seniority system and whether or not the negotiation or maintenance of the system was tainted with racial discrimination, the Court of Appeals, while identifying what it thought was legal error in failing to consider the racial practices and intentions of IAM, did not otherwise overturn any of the District Court’s findings as clearly erroneous.
See 5A J. Moore
&
J. Lucas, Moore’s Federal Practice §52.06[2] (1982) (“Where the trial court fails to make findings, or to find on a material issue, and an appeal is taken, the appellate court will normally vacate the judgment and remand the action for appropriate findings to be made”);
Rule
v.
International Assn. of Bridge Workers,
568 F. 2d 558, 568
(CA8 1978);
Chicano Police Officer’s Assn.
v.
Stover,
552 F. 2d 918
, 921 (CA10 1977);
O’Neal
v.
Gresham,
519 F. 2d 803
, 805 (CA4 1975);
Burch
v.
International Assn. of Machinists & Aerospace Workers, AFL-CIO,
433 F. 2d 561
(CA5 1970);
General Electric Credit Corp.
v.
Robbins,
414 F. 2d 208
(CA8 1969).
IAM’s discriminatory motivation, if it existed, cannot be imputed to USW. It is relevant only to the extent that it may shed some light on the purpose of USW or the Company in creating and maintaining the separate seniority system at issue in these cases. A discriminatory intent on the part of IAM, therefore, does not control the outcome of these cases. Neither does the fact, if true, that USW acquiesced in racially discriminatory conduct on the part of IAM. Such acquiescence is not the equivalent of a discriminatory purpose on the part of USW.

---

Concurrence by Stevens:

Justice Stevens,
concurring in part.
Except to the extent that the Court’s preliminary comments on the burden of sustaining “any challenge to a seniority system under Title VII,”
ante,
at 277, are inconsistent with the views I expressed separately in
American Tobacco Co.
v.
Patterson, ante,
p. 86, I join the Court’s opinion.

---

Dissent by Marshall:

Justice Marshall,
with whom Justice Blackmun joins except as to Part I, dissenting.
In 1971, a group of Negro employees at Pullman-Standard’s Bessemer, Ala., plant brought this class action against Pullman-Standard, the United Steelworkers of America and its Local 1466 (USW), and the International Association of Machinists and its Local 372 (IAM). The plaintiffs alleged,
inter alia,
that the departmental seniority system negotiated
*294
by both unions discriminated against Negroes in violation of Title VII of the Civil Rights Act of 1964, 42 U. S. C. § 2000e
et seq.
(1976 ed. and Supp. IV), and the Civil Rights Act of 1866,
42 U. S. C. § 1981
. In 1974, the District Court for the Northern District of Alabama concluded that the seniority system did not operate to discriminate against Negroes. A unanimous panel of the Fifth Circuit reversed. The court ruled that the District Court had committed several errors of law, including failure to give proper weight to the role of the IAM, and had relied on patently inaccurate factual conclusions.
Swint
v.
Pullman-Standard,
539 F. 2d 77, 95-96
(1976). On remand, the District Court again ruled that the seniority system was immune from attack under Title VII, this time finding that respondents had failed to show discriminatory intent as required by this Court’s decision in
Teamsters
v.
United States,
431 U. S. 324
(1977).
Ante,
at 275. The Fifth Circuit again unanimously rejected the conclusion of the District Court.
624 F. 2d 525
(1980). The majority now reverses the Fifth Circuit’s second unanimous decision on the ground that the Court of Appeals did not pay sufficient homage to the “clearly erroneous” rule, Fed. Rule Civ. Proc. 52(a), in concluding that the seniority system at Pullman-Standard was the product of intentional discrimination against Negroes. Because I cannot agree with the premise of the majority’s decision to remand these cases for yet another trial, or with its application of that premise to the facts of this case, I respectfully dissent.
HH
The majority premises its holding on the assumption that “ ‘absent a discriminatory purpose, the operation of a seniority system cannot be an unlawful employment practice even if the system has some discriminatory consequences.’”
Ante,
at 277, quoting
Trans World Airlines, Inc.
v.
Hardison,
432 U. S. 63, 82
(1977). As I have previously indicated, I do not find anything in the relevant statutory language or legislative
*295
history to support the proposition that § 703(h) of Title VII immunizes a seniority system that perpetuates past discrimination, as the system at issue here clearly does, simply because the plaintiffs are unable to demonstrate to this Court’s satisfaction that the system was adopted or maintained for an invidious purpose. See
Teamsters
v.
United States, supra,
at 377-394
(opinion of Marshall, J.). In my opinion, placing such a burden on plaintiffs who challenge seniority systems with admitted discriminatory impact, a burden never before imposed in civil suits brought under Title VII, frustrates the clearly expressed will of Congress and effectively “freeze[s] an entire generation of Negro employees into discriminatory patterns that existed before the Act.”
Quarles
v.
Philip Morris, Inc.,
279 F. Supp. 505, 516
(ED Va. 1968) (Butzner, J.).
II
Even if I were to accept this Court’s decision to impose this novel burden on Title VII plaintiffs, I would still be unable to concur in its conclusion that the Fifth Circuit’s decision should be reversed for failing to abide by Rule 52(a). The majority asserts that the Court of Appeals in this action ignored the clearly-erroneous rule and made an independent determination of discriminatory purpose. I disagree. In my view, the court below followed well-established legal principles both in rejecting the District Court’s finding of no discriminatory purpose and in concluding that a finding of such a purpose was compelled by all of the relevant evidence.
The majority concedes, as it must, that the “Court of Appeals acknowledged and correctly stated the controlling standard of Rule 52(a).”
Ante,
at 290. In a footnote to its opinion, the Court of Appeals plainly states that findings of fact may be overturned only if they are either “clearly erroneous” or “made under an erroneous view of controlling legal principles.” 624 F. 2d, at 533, n. 6. Furthermore, as the majority notes,
ante,
at 283, the Court of Appeals justified its decision to reject the District Court’s finding that the senior
*296
ity system was not the result of purposeful discrimination by stating: “An analysis of the totality of the facts and circumstances surrounding the creation and continuance of the departmental system at Pullman-Standard leaves us with the
definite and firm, conviction that a mistake has been made”
624 F. 2d, at 533 (emphasis added; footnote omitted).
1
I frankly am at a loss to understand how the Court of Appeals could have expressed its conclusion that the District Court’s finding on the issue of intent was clearly erroneous with any more precision or clarity.
The majority rejects the Court of Appeals’ clear articulation and implementation of the clearly-erroneous rule on the apparent ground that in the course of correctly setting forth the requirements of Rule 52(a), the court also included the following quotation from its prior decision in
East
v.
Romine, Inc.,
518 F. 2d 332, 339
(1975):
“‘Although discrimination
vel non
is essentially a question of fact it is, at the same time, the ultimate issue for resolution in this case, being expressly proscribed by 42 U. S. C. A. § 2000e-2(a). As such, a finding of discrimination or nondiscrimination is a finding of ultimate fact. [Cites omitted]. In reviewing the district court’s findings, therefore, we will proceed to make an independent determination of appellant’s allegations of discrimination, though bound by findings of subsidiary fact which are themselves not clearly erroneous.’” 624 F. 2d, at 533, n. 6.
The only question presented by this case, therefore, is whether this reference to
East
v.
Romine, Inc.,
should be read as negating the Court of Appeals’ unambiguous ac
*297
knowledgment of the “controlling standard of Rule 52.”
Ante,
at 290. The majority bases its affirmative answer to that question on two factors. First, the majority contends that the Court of Appeals must not have properly respected the clearly-erroneous rule because its acknowledgment that Rule 52(a) supplied the controlling standard “came late in the court’s opinion.”
Ante,
at 290. Second, the Court of Appeals “identified not only the mistake” that it felt had been made, “but also the source of that mistake.”
Ante,
at 291. If the Court of Appeals had really been applying the clearly-erroneous rule, it should have abided by the “usual requirement of remanding for further proceedings to the tribunal charged with the task of factfinding in the first instance.”
Ante,
at 293.
Neither of these arguments justifies the majority’s conclusion that these cases must be remanded for a fourth trial on the merits. I am aware of no rule of decision embraced by this or any other court that places dispositive weight on whether an accurate statement of controlling principle appears “early” or late in a court’s opinion. Nor does the majority suggest a basis for this unique rule of interpretation. So long as a court acknowledges the proper legal standard, I should think it irrelevant whether it chooses to set forth that standard at the beginning or at the end of its opinion. The heart of the majority’s argument, therefore, is that the failure to remand the action to the District Court after rejecting its conclusion that the seniority system was “bona fide” within the meaning of § 703(h) indicates that the Court of Appeals did not properly follow the clearly-erroneous rule. Before addressing this issue, however, it is necessary to examine the nature of the finding of “intent” required by this Court in
Teamsters
,
the procedure that courts of appeals should follow in reviewing a district court’s finding on intent, and the extent to which the court below adhered to that procedure in this case.
The District Court examined the four factors approved by the Fifth Circuit in
James
v.
Stockham Valves & Fittings
*298
Co.,
559 F. 2d 310
(1977), cert. denied,
434 U. S. 1034
(1978), to determine whether the departmental seniority system at Pullman-Standard was adopted or maintained for a discriminatory purpose. Although indicating that these four factors are not the only way to demonstrate the existence of discriminatory intent,
2
the Court today implicitly acknowledges that proof of these factors satisfies the requirements of Teamsters.
3
In particular, the majority agrees that a finding of discriminatory intent sufficient to satisfy
Teamsters
can be based on circumstantial evidence, including evidence of discriminatory impact. See
ante,
at 289; see also
Arlington Heights
v.
Metropolitan Housing Dev. Corp.,
429 U. S. 252, 266, 267
(1977).
Given the nature of this factual inquiry, the court of appeals must first determine whether the district court applied correct legal principles and therefore
considered
all of the legally relevant evidence presented by the parties. This, as the majority acknowledges, is a “legal” function that the court of appeals must perform in the first instance.
Ante,
at 282, 283. Second, the court of appeals must determine whether the district court’s finding with respect to intent is
supported
by all of the legally relevant evidence. This, the Court holds today, is generally a factual determination limited by the dictates of Rule 52(a). Finally, if the court of appeals sets aside the district court’s finding with respect to intent, either because that finding is clearly erroneous or because it is based on an erroneous legal standard, it may determine, in the interest of judicial economy, whether the le-
*299
gaily relevant evidence presented to the district court “permits only one resolution of the factual issue.” Ante, at 292. If only one conclusion is possible, the reviewing court is free to find the existence of the fact in question as a matter of law. See
Bigelow
v.
Virginia,
421 U. S. 809, 826-827
(1975);
Levin
v.
Mississippi River Fuel Corp.,
386 U. S. 162, 170
(1967).
A common-sense reading of the opinion below demonstrates that the Court of Appeals followed precisely this course in examining the issue of discriminatory intent. Even the majority concedes that the Court of Appeals determined that the District Court committed “legal error” by failing to consider all of the relevant evidence in resolving the first and the third
James
factors.
Ante,
at 282, 283. With respect to the first
James
factor — whether the system inhibits all employees equally from transferring between seniority units — the District Court found that the departmental system “locked” both Negro and white workers into departments by discouraging transfers. The District Court acknowledged that Negroes might suffer a greater impact because the company’s previous discriminatory policy of openly maintaining “Negro” jobs and “white” jobs had caused Negroes to be concentrated in less desirable positions. The District Court concluded, however, that this differential impact was irrelevant in determining whether the seniority system operated neutrally. The Court of Appeals properly held that the District Court erred in failing to consider the fact that the departmental system locked Negroes into less desirable jobs.
Similarly, as for the third
James
factor — whether the seniority system had its genesis in racial discrimination — the District Court rejected respondents’ argument that the motives of the IAM were relevant. It concluded that the USW could not be charged with the racial bias of the IAM. The Court of Appeals held that this conclusion was erroneous because the “motives and intent of the I. A. M. in 1941 and 1942
*300
are significant in consideration of whether the seniority system has its genesis in racial discrimination.” 624 F. 2d, at 532.
4
As the majority acknowledges, where findings of fact “‘are made under an erroneous view of controlling legal principles, the clearly erroneous rule does not apply, and the findings may not stand.’”
Ante,
at 285, quoting 624 F. 2d, at 533, n. 6; see also
Kelley
v.
Southern Pacific Co.,
419 U. S. 318, 323
(1974);
United States
v.
General Motors Corp.,
384 U. S. 127, 141, n. 16
(1966);
United States
v.
Singer Manufacturing Co.,
374 U. S. 174, 194, n. 9
(1963);
United States
v.
Parke, Davis & Co.,
362 U. S. 29, 44
(1960);
Rowe
v.
General Motors Corp.,
457 F. 2d 348, 356, n. 15
(CA5 1972). Having found that the District Court’s findings as to the first and third
James
factors were made under an erroneous view of controlling legal principles, the Court of Appeals was
compelled
to set aside those findings free of the requirements of the clearly-erroneous rule.
5
But once these two findings were set aside, the District Court’s conclusion that the departmental system was bona fide within the meaning of § 703(h) also had to be rejected, since that conclusion was based at least in part on its erroneous determinations concerning the first and the third
James
factors.
At the very least, therefore, the Court of Appeals was entitled to remand this action to the District Court for the pur
*301
pose of reexamining the bona fides of the seniority system under proper legal standards. However, as we have often noted, in some cases a remand is inappropriate where the facts on the record are susceptible to only one reasonable interpretation. See
Dayton Board of Education
v.
Brinkman,
443 U. S. 526, 534-537
(1979);
Bigelow
v.
Virginia, supra,
at 826-827
. In such cases, “[effective judicial administration” requires that the court of appeals draw the inescapable factual conclusion itself, rather than remand the case to the district court for further needless proceedings.
Levin
v.
Mississippi River Fuel Corp.,
386 U. S., at 170
. Such action is particularly appropriate where the court of appeals is in as good a position to evaluate the record evidence as the district court. The major premise behind the deference to trial courts expressed in Rule 52(a) is that findings of fact “depend peculiarly upon the credit given to witnesses by those who see and hear them.”
United States
v.
Yellow Cab Co.,
338 U. S. 338, 341
(1949); see also
United States
v.
Oregon State Medical Society,
343 U. S. 326, 332
(1952). Indeed Rule 52(a) expressly acknowledges the importance of this factor by stating that “due regard shall be given to the opportunity of the trial court to judge of the credibility of the witnesses.” Consequently, this Court has been especially reluctant to resolve factual issues which depend on the credibility of witnesses. See generally
United States
v.
Oregon State Medical Society, supra,
at 332.
In the cases before the Court today this usual deference is not required because the District Court’s findings of fact were entirely based on documentary evidence.
6
As we
*302
noted in
United States
v.
General Motors Corp., supra,
at 141, n. 16, “the trial court’s customary opportunity to evaluate the demeanor and thus the credibility of the witnesses, which is the rationale behind Rule 52(a). . . , plays only a restricted role [in] a ‘paper case.’” See also
Jennings
v.
General Medical Corp.,
604 F. 2d 1300, 1305
(CA10 1979) (“When the findings of a trial court are based on documentary, rather than oral evidence, they do not carry the same weight on appellate review”);
Orvis
v.
Higgins,
180 F. 2d 537, 539
(CA2 1950).
7
I believe that the Court of Appeals correctly determined that a finding of discriminatory intent was compelled by the documentary record presented to the District Court. With respect to three of the four
James
factors, the Court of Appeals found overwhelming evidence of discriminatory intent. First, in ruling that the District Court erred by not acknowledging the legal significance of the fact that the seniority system locked Negroes into the least remunerative jobs in the company, the Court of Appeals determined that such disproportionate impact demonstrated that the system did not “ ‘operat[e] to discourage all employees equally from transferring between seniority units.’” 624 F. 2d, at 530, quoting
*303
James
v.
Stockham Valves & Fittings Co.,
559 F. 2d, at 352
. Second, noting that “[n]o credible explanation ha[d] been advanced to sufficiently justify” the existence of two separate Die and Tool Departments and two separate Maintenance Departments, a condition not found at any other Pullman-Standard plant, or the creation of all-white and all-Negro departments at the time of unionization and in subsequent years, the Court of Appeals concluded that the second
James
factor had not been satisfied.
8
624 F. 2d, at 533. Finally, with respect to the third
James
factor the Court of Appeals found that once the role of the IAM was properly recognized, it was “crystal clear that considerations of race permeated the negotiation and the adoption of the seniority system in 1941 and subsequent negotiations thereafter.” 624 F. 2d, at 532.
9
*304
After reviewing all of the relevant record evidence presented to the District Court, the Court of Appeals concluded: “There is no doubt, based upon the record in this case, about the existence of a discriminatory purpose.” Id., at 533. Because I fail to see how the Court of Appeals erred in carrying out its appellate function, I respectfully dissent from the majority’s decision to prolong respondents’ 11-year quest for the vindication of their rights by requiring yet another trial.
As the majority acknowledges,
ante,
at 284-285, n. 14, this Court stated in
United States
v.
United States Gypsum Co.,
333 U. S. 364, 395
(1948), that a finding of fact is clearly erroneous if “the reviewing court on the entire evidence is left with
the definite and firm conviction that a mistake has been committed”
(emphasis added).
Contrary to the majority’s suggestion,
ante,
at 279, n. 8, I find nothing in the Fifth Circuit’s decision in
James
v.
Stockham Valves & Fittings Co.
to imply that these factors constitute the only relevant criteria for determining discriminatory intent.
This conclusion would seem to be compelled since, as the majority notes, the
James
factors are nothing more than a summary of the criteria examined by this Court in
Teamsters,
431 U. S., at 355-356
.
As the majority indicates in a footnote,
ante,
at 292, n. 23, the discriminatory motive of the IAM is “relevant... to the extent that it may shed light on the purpose of USW or the Company in creating and maintaining the separate seniority system at issue in this case.” I do not read the Court of Appeals opinion in this action as holding anything more than that if the USW participated in establishing a system that was designed for the purpose of perpetuating past discrimination, the third
James
factor would be satisfied. Given that the IAM is a party to this litigation, its participation in the creation of the seniority system can hardly be deemed irrelevant.
It is therefore irrelevant that the Court of Appeals did not specifically hold that the District Court’s other factual findings were clearly erroneous.
Only two witnesses testified during the brief hearing that the District Court conducted on the question whether the seniority system at Pullman-Standard was immune under § 703(h). Both of these witnesses were longtime Negro employees of Pullman-Standard who testified on behalf of
respondents
concerning racial segregation at the plant and by the USW. There is no indication in the District Court’s opinion that it relied upon the
*302
testimony of these two witnesses in concluding that the system was bona fide within the meaning of § 703(h). The remainder of the record before the District Court consisted entirely of 139 exhibits submitted by respondents, the company, and the unions concerning the development and maintenance of the seniority system from 1940 through the 1970’s.
This is not to say that the clearly-erroneous rule does not apply to “document” cases. See
United States
v.
Singer Manufacturing Co.,
374 U. S. 174, 194, n. 9
(1963). However, “when the decision of the court below rests upon an incorrect reading of an undisputed document, [the appellate] court is free to substitute its own reading of the document.”
Eutectic Corp.
v.
Metco, Inc.,
579 F. 2d 1, 5
(CA2 1978). See also
McKensie
v.
Sea Land Service,
551 F. 2d 91
(CA5 1977);
Best Medium Pub. Co.
v.
National Insider, Inc., 385
F. 2d 384 (CA7 1967), cert. denied,
390 U. S. 955
(1968);
United States ex rel. Binion
v.
O’Brien,
273 F. 2d 495
(CA3 1959), cert. denied,
363 U. S. 812
(1960).
Although the majority is correct in stating that the Court of Appeals did not “refer to or
expressly
apply the clearly-erroneous standard” in reaching this conclusion,
ante,
at 282 (emphasis added), the appellate court’s adherence to the requirements of Rule 52(a) is nevertheless apparent from the following statement:
“The record evidence indicates that a significant number of one-race departments were established upon unionization at Pullman-Standard, and during the next twenty five years, one-race departments were carved out of previously mixed departments. The establishment and maintenance of the segregated departments appear to be based on
no other considerations than the objective to separate the races.”
624 F. 2d, at 531 (emphasis added).
In my opinion, this statement is sufficient to satisfy the requirements of Rule 52(a), particularly in light of the Court of Appeals’ general acknowledgment that it was bound by the clearly-erroneous rule. See
supra,
at 296-297.
Whether or not the Court of Appeals expressly ruled on the fourth
James
factor is irrelevant. As the Court of Appeals clearly stated, its conclusion was based on “the totality of the facts and circumstances surrounding the creation and continuance of the departmental system at Pullman-Standard.” 624 F. 2d, at 533; see also
id.,
at 532 (“It is crystal clear that considerations of race permeated the negotiation and the adoption of the seniority system in 1941 and subsequent negotiations thereafter”), and
id.,
*304
at 533 (“We consider significant in our decision . . . conditions of racial discrimination which affected the negotiation and renegotiation of the system . . .”). Even assuming that the District Court was correct in concluding that the system had been
maintained
free of any illegal purpose, the Court of Appeals was entitled to conclude that discriminatory intent had been demonstrated on the basis of other relevant evidence.

---

456 U.S. 273 (1982)
PULLMAN-STANDARD, A DIVISION OF PULLMAN, INC.
v.
SWINT ET AL.
No. 80-1190.
Supreme Court of United States.
Argued January 19, 1982.
Decided April 27, 1982.
[*]
CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE FIFTH CIRCUIT.
*274
Michael H. Gottesman
argued the cause for petitioners. With him on the briefs for petitioners in No. 80-1193 were
Robert M. Weinberg, Laurence Gold, Jerome A. Cooper, Bernard Kleiman,
and
Carl B. Frankel. Samuel H. Burr
and
C. V. Stelzenmuller
filed briefs for petitioner in No. 80-1190.
Elaine Jones
argued the cause for respondents. With her on the brief were
Jack Greenberg, James M. Nabrit III, Patrick O. Patterson, Judith Reed, Barry L. Goldstein,
and
C. Lani Gunier.
[†]
Robert E. Williams
and
Douglas S. McDowell
filed a brief for the Equal Employment Advisory Council as
amicus curiae.
*275
JUSTICE WHITE delivered the opinion of the Court.
Respondents were black employees at the Bessemer, Ala., plant of petitioner Pullman-Standard (the Company), a manufacturer of railway freight cars and parts. They brought suit against the Company and the union petitioners — the United Steelworkers of America, AFL-CIO-CLC, and its Local 1466 (collectively USW) — alleging violations of Title VII of the Civil Rights Act of 1964, 78 Stat. 253, as amended, 42 U. S. C. § 2000e
et seq.
(1976 ed. and Supp. IV), and 42 U. S. C. § 1981.
[1]
As they come here, these cases involve only the validity, under Title VII, of a seniority system maintained by the Company and USW. The District Court found "that the differences in terms, conditions or privileges of employment resulting [from the seniority system] are `not the result of an intention to discriminate' because of race or color," App. to Pet. for Cert. in No. 80-1190, p. A-147 (hereinafter App.), and held, therefore, that the system satisfied the requirements of § 703(h) of the Act. The Court of Appeals for the Fifth Circuit reversed:
"Because we find that the differences in the terms, conditions and standards of employment for black workers and white workers at Pullman-Standard resulted from an intent to discriminate because of race, we hold that the system is not legally valid under section 703(h) of Title VII, 42 U. S. C. 2000e-2(h)." 624 F. 2d 525, 533-534 (1980).
*276
We granted the petitions for certiorari filed by USW and by the Company, 451 U. S. 906 (1981), limited to the first question presented in each petition: whether a court of appeals is bound by the "clearly erroneous" rule of Federal Rule of Civil Procedure 52(a) in reviewing a district court's findings of fact, arrived at after a lengthy trial, as to the motivation of the parties who negotiated a seniority system; and whether the court below applied wrong legal criteria in determining the bona fides of the seniority system. We conclude that the Court of Appeals erred in the course of its review and accordingly reverse its judgment and remand for further proceedings.
I
Title VII is a broad remedial measure, designed "to assure equality of employment opportunities."
McDonnell Douglas Corp.
v.
Green,
411 U. S. 792, 800 (1973). The Act was designed to bar not only overt employment discrimination, "but also practices that are fair in form, but discriminatory in operation."
Griggs
v.
Duke Power Co.,
401 U. S. 424, 431 (1971). "Thus, the Court has repeatedly held that a prima facie Title VII violation may be established by policies or practices that are neutral on their face and in intent but that nonetheless discriminate in effect against a particular group."
Teamsters
v.
United States,
431 U. S. 324, 349 (1977) (hereinafter
Teamsters
). The Act's treatment of seniority systems, however, establishes an exception to these general principles. Section 703(h), 78 Stat. 257, as set forth in 42 U. S. C. § 2000e-2(h), provides in pertinent part:
"Notwithstanding any other provision of this subchapter, it shall not be an unlawful employment practice for an employer to apply different standards of compensation, or different terms, conditions, or privileges of employment pursuant to a bona fide seniority . . . system. . . provided that such differences are not the result of an intention to discriminate because of race."
*277
Under this section, a showing of disparate impact is insufficient to invalidate a seniority system, even though the result may be to perpetuate pre-Act discrimination. In
Trans World Airlines, Inc.
v.
Hardison,
432 U. S. 63, 82 (1977), we summarized the effect of § 703(h) as follows: "[A]bsent a discriminatory purpose, the operation of a seniority system cannot be an unlawful employment practice even if the system has some discriminatory consequences." Thus, any challenge to a seniority system under Title VII will require a trial on the issue of discriminatory intent: Was the system adopted because of its racially discriminatory impact?
This is precisely what happened in these cases. Following our decision in
Teamsters,
the District Court held a new trial on the limited question of whether the seniority system was "instituted or maintained contrary to Section 703(h) of the new Civil Rights Act of 1964." App. A-125.
[2]
That court concluded, as we noted above and will discuss below, that the system was adopted and maintained for purposes wholly independent of any discriminatory intent. The Court of Appeals for the Fifth Circuit reversed.
II
Petitioners submit that the Court of Appeals failed to comply with the command of Rule 52(a) that the findings of fact of a district court may not be set aside unless clearly erroneous. We first describe the findings of the District Court and the Court of Appeals.
Certain facts are common ground for both the District Court and the Court of Appeals. The Company's Bessemer plant was unionized in the early 1940's. Both before and after unionization, the plant was divided into a number of different operational departments.
[3]
USW sought to represent
*278
all production and maintenance employees at the plant and was elected in 1941 as the bargaining representative of a bargaining unit consisting of most of these employees. At that same time, IAM became the bargaining representative of a unit consisting of five departments.
[4]
Between 1941 and 1944, IAM ceded certain workers in its bargaining unit to USW. As a result of this transfer, the IAM bargaining unit became all white.
Throughout the period of representation by USW, the plant was approximately half black. Prior to 1965, the Company openly pursued a racially discriminatory policy of job assignments. Most departments contained more than one job category and as a result most departments were racially mixed. There were no lines of progression or promotion within departments.
The seniority system at issue here was adopted in 1954.
[5]
Under that agreement, seniority was measured by length of continuous service in a particular department.
[6]
Seniority was originally exercised only for purposes of layoffs and hirings within particular departments. In 1956, seniority was formally recognized for promotional purposes as well. Again, however, seniority, with limited exceptions, was only exercised within departments; employees transferring to
*279
new departments forfeited their seniority. This seniority system remained virtually unchanged until after this suit was brought in 1971.
[7]
The District Court approached the question of discriminatory intent in the manner suggested by the Fifth Circuit in
James
v.
Stockham Valves & Fittings Co.,
559 F. 2d 310 (1977). There, the Court of Appeals stated that under
Teamsters
"the totality of the circumstances in the development and maintenance of the system is relevant to examining that issue." 559 F. 2d, at 352. There were, in its view, however, four particular factors that a court should focus on.
[8]
First, a court must determine whether the system "operates to discourage all employees equally from transferring between seniority units."
Ibid.
The District Court held that the system here "was facially neutral and . . . was applied equally to all races and ethnic groups." App. A-132. Although there were charges of racial discrimination in its application, the court held that these were "not substantiated by the evidence."
Id.,
at A-133. It concluded that the system "applied equally and uniformly to all employees, black and white, and that, given the approximately equal number
*280
of employees of the two groups, it was quantitatively neutral as well."
Id.,
at A-134.
[9]
Second, a court must examine the rationality of the departmental structure, upon which the seniority system relies, in light of the general industry practice.
James, supra,
at 352. The District Court found that linking seniority to "departmental age" was "the modal form of agreements generally, as well as with manufacturers of railroad equipment in particular." App. A-137. Furthermore, it found the basic arrangement of departments at the plant to be rationally related to the nature of the work and to be "consistent with practices which were . . . generally followed at other unionized plants throughout the country."
Id.,
at A-136 — A-137. While questions could be raised about the necessity of certain departmental divisions, it found that all of the challenged lines of division grew out of historical circumstances at the plant that were unrelated to racial discrimination.
[10]
Although unionization did produce an all-white IAM bargaining unit, it found that USW "cannot be charged with racial bias in its response to the IAM situation. [USW] sought to represent all workers, black and white, in the plant."
Id.,
at A-145. Nor could the Company be charged with any racial discrimination that may have existed in IAM:
"The company properly took a `hands-off' approach towards the establishment of the election units . . . . It bargained with those unions which were afforded representational
*281
status by the NLRB and did so without any discriminatory animus."
Id.,
at A-146.
Third, a court had to consider "whether the seniority system had its genesis in racial discrimination,"
James, supra,
at 352, by which it meant the relationship between the system and other racially discriminatory practices. Although finding ample discrimination by the Company in its employment practices and some discriminatory practices by the union,
[11]
the District Court concluded that the seniority system was in no way related to the discriminatory practices:
"The seniority system . . . had its genesis . . . at a period when racial segregation was certainly being practiced; but this system was not itself the product of this bias. The system rather came about as a result of colorblind objectives of a union which — unlike most structures and institutions of the era — was not an arm of a segregated society. Nor did it foster the discrimination. . . which was being practiced by custom in the plant." App. A-144.
Finally, a court must consider "whether the system was negotiated and has been maintained free from any illegal purpose."
James, supra,
at 352. Stating that it had "carefully considered the detailed record of negotiation sessions and contracts which span a period of some thirty-five years," App. A-146, the court found that the system was untainted by any discriminatory purpose. Thus, although the District
*282
Court focused on particular factors in carrying out the analysis required by § 703(h), it also looked to the entire record and to the "totality of the system under attack."
Id.,
at A-147.
The Court of Appeals addressed each of the four factors of the
James
test and reached the opposite conclusion. First, it held that the District Court erred in putting aside qualitative differences between the departments in which blacks were concentrated and those dominated by whites, in considering whether the system applied "equally" to whites and blacks.
[12]
This is a purported correction of a legal standard under which the evidence is to be evaluated.
Second, it rejected the District Court's conclusion that the structure of departments was rational, in line with industry practice, and did not reflect any discriminatory intent. Its discussion is brief but focuses on the role of IAM and certain characteristics unique to the Bessemer plant. The court concluded:
"The record evidence, generally, indicates arbitrary creation of the departments by the company since unionization and an attendant adverse affect
[sic]
on black workers. The individual differences between the departmental structure at Pullman-Standard and that of other plants, and as compared with industry practice, are indicative of attempts to maintain one-race departments." 624 F. 2d, at 532.
In reaching this conclusion, the Court of Appeals did not purport to be correcting a legal error, nor did it refer to or expressly apply the clearly-erroneous standard.
*283
Third, in considering the "genesis" of the system, the Court of Appeals held that the District Court erred in holding that the motives of IAM were not relevant.
[13]
This was the correction of a legal error on the part of the District Court in excluding relevant evidence. The court did not stop there, however. It went on to hold that IAM was acting out of discriminatory intent — an issue specifically not reached by the District Court — and that "considerations of race permeated the negotiation and the adoption of the seniority system in 1941 and subsequent negotiations thereafter."
Ibid.
Fourth, despite this conclusion under the third
James
factor the Court of Appeals then recited, but did not expressly set aside or find clearly erroneous, the District Court's findings with respect to the negotiation and maintenance of the seniority system.
The court then announced that "[h]aving carefully reviewed the evidence offered to show whether the departmental seniority system in the present case is `bona fide' within the meaning of § 703(h) of Title VII, we reject the district court's finding." 624 F. 2d, at 533. Elaborating on its disagreement, the Court of Appeals stated:
"An analysis of the totality of the facts and circumstances surrounding the creation and continuance of the departmental system at Pullman-Standard leaves us with the definite and firm conviction that a mistake has been made. There is no doubt, based upon the record in this case, about the existence of a discriminatory purpose. The obvious principal aim of the I. A. M. in 1941 was to exclude black workers from its bargaining unit.
*284
That goal was ultimately reached when maneuvers by the I. A. M. and U. S. W. resulted in an all-white I. A. M. unit. The U. S. W., in the interest of increased membership, acquiesced in the discrimination while succeeding in significantly segregating the departments within its own unit.
"The district court might have reached a different conclusion had it given the I. A. M.'s role in the creation and establishment of the seniority system its due consideration."
Ibid.
(footnote omitted).
Having rejected the District Court's finding, the court made its own findings as to whether the USW seniority system was protected by § 703(h):
"We consider significant in our decision the manner by which the two seniority units were set up, the creation of the various all-white and all-black departments within the U. S. W. unit at the time of certification and in the years thereafter, conditions of racial discrimination which affected the negotiation and renegotiation of the system, and the extent to which the system and the attendant no-transfer rule locked blacks into the least remunerative positions within the company. Because we find that the differences in the terms, conditions and standards of employment for black workers and white workers at Pullman-Standard resulted from an intent to discriminate because of race, we hold that the system is not legally valid under section 703(h) of Title VII, 42 U. S. C. § 2000e-2(h)."
Id.,
at 533-534.
In connection with its assertion that it was convinced that a mistake had been made, the Court of Appeals, in a footnote, referred to the clearly-erroneous standard of Rule 52(a).
Id.,
at 533, n. 6.
[14]
It pointed out, however, that if findings
*285
"are made under an erroneous view of controlling legal principles, the clearly erroneous rule does not apply, and the findings may not stand."
Ibid.
Finally, quoting from
East
v.
Romine, Inc.,
518 F. 2d 332, 339 (CA5 1975), the Court of Appeals repeated the following view of its appellate function in Title VII cases where purposeful discrimination is at issue:
" `Although discrimination
vel non
is essentially a question of fact it is, at the same time, the ultimate issue for resolution in this case, being expressly proscribed by 42 U. S. C. A. § 2000e-2(a). As such, a finding of discrimination or non-discrimination is a finding of ultimate fact. [Cites omitted.] In reviewing the district court's findings, therefore, we will proceed to make an independent determination of appellant's allegations of discrimination, though bound by findings of subsidiary fact which are themselves not clearly erroneous.' " 624 F. 2d, at 533, n. 6.
III
Pointing to the above statement of the Court of Appeals and to similar statements in other Title VII cases coming from that court,
[15]
petitioners submit that the Court of Appeals
*286
made an independent determination of discriminatory purpose, the "ultimate fact" in this case, and that this was error under Rule 52(a). We agree with petitioners that if the Court of Appeals followed what seems to be the accepted rule in that Circuit, its judgment must be reversed.
[16]
*287
Rule 52(a) broadly requires that findings of fact not be set aside unless clearly erroneous. It does not make exceptions or purport to exclude certain categories of factual findings from the obligation of a court of appeals to accept a district court's findings unless clearly erroneous. It does not divide facts into categories; in particular, it does not divide findings of fact into those that deal with "ultimate" and those that deal with "subsidiary" facts.
The Rule does not apply to conclusions of law. The Court of Appeals, therefore, was quite right in saying that if a district court's findings rest on an erroneous view of the law, they may be set aside on that basis. But here the District Court was not faulted for misunderstanding or applying an erroneous definition of intentional discrimination.
[17]
It was reversed for arriving at what the Court of Appeals thought was an erroneous finding as to whether the differential impact of the seniority system reflected an intent to discriminate on account of race. That question, as we see it, is a
*288
pure question of fact, subject to Rule 52(a)'s clearly-erroneous standard. It is not a question of law and not a mixed question of law and fact.
The Court has previously noted the vexing nature of the distinction between questions of fact and questions of law. See
Baumgartner
v.
United States,
322 U. S. 665, 671 (1944). Rule 52(a) does not furnish particular guidance with respect to distinguishing law from fact. Nor do we yet know of any other rule or principle that will unerringly distinguish a factual finding from a legal conclusion. For the reasons that follow, however, we have little doubt about the factual nature of § 703(h)'s requirement that a seniority system be free of an intent to discriminate.
Treating issues of intent as factual matters for the trier of fact is commonplace. In
Dayton Board of Education
v.
Brinkman,
443 U. S. 526, 534 (1979), the principal question was whether the defendants had intentionally maintained a racially segregated school system at a specified time in the past. We recognized that issue as essentially factual, subject to the clearly-erroneous rule. In
Commissioner
v.
Duberstein,
363 U. S. 278 (1960), the Court held that the principal criterion for identifying a gift under the applicable provision of the Internal Revenue Code was the intent or motive of the donor — "one that inquires what the basic reason for his conduct was in fact."
Id.,
at 286. Resolution of that issue determined the ultimate issue of whether a gift had been made. Both issues were held to be questions of fact subject to the clearly-erroneous rule. In
United States
v.
Yellow Cab Co.,
338 U. S. 338, 341 (1949), an antitrust case, the Court referred to "[f]indings as to the design, motive and intent with which men act" as peculiarly factual issues for the trier of fact and therefore subject to appellate review under Rule 52.
Justice Black's dissent in
Yellow Cab
suggested a contrary approach. Relying on
United States
v.
Griffith,
334 U. S.
*289
100 (1948), he argued that it is not always necessary to prove "specific intent" to restrain trade; it is enough if a restraint is the result or consequence of a defendant's conduct or business arrangements. Such an approach, however, is specifically precluded by § 703(h) in Title VII cases challenging seniority systems. Differentials among employees that result from a seniority system are not unlawful employment practices unless the product of an intent to discriminate. It would make no sense, therefore, to say that the intent to discriminate required by § 703(h) may be presumed from such an impact. As § 703(h) was construed in
Teamsters,
there must be a finding of actual intent to discriminate on racial grounds on the part of those who negotiated or maintained the system. That finding appears to us to be a pure question of fact.
This is not to say that discriminatory impact is not part of the evidence to be considered by the trial court in reaching a finding on whether there was such a discriminatory intent as a factual matter.
[18]
We do assert, however, that under § 703(h) discriminatory intent is a finding of fact to be made by the trial court; it is not a question of law and not a mixed question of law and fact of the kind that in some cases may allow an appellate court to review the facts to see if they satisfy some legal concept of discriminatory intent.
[19]
Discriminatory
*290
intent here means actual motive; it is not a legal presumption to be drawn from a factual showing of something less than actual motive. Thus, a court of appeals may only reverse a district court's finding on discriminatory intent if it concludes that the finding is clearly erroneous under Rule 52(a). Insofar as the Fifth Circuit assumed otherwise, it erred.
IV
Respondents do not directly defend the Fifth Circuit rule that a trial court's finding on discriminatory intent is not subject to the clearly-erroneous standard of Rule 52(a).
[20]
Rather, among other things, they submit that the Court of Appeals recognized and, where appropriate, properly applied Rule 52(a) in setting aside the findings of the District Court. This position has force, but for two reasons it is not persuasive.
First, although the Court of Appeals acknowledged and correctly stated the controlling standard of Rule 52(a), the acknowledgment came late in the court's opinion. The court had not expressly referred to or applied Rule 52(a) in the course of disagreeing with the District Court's resolution of the factual issues deemed relevant under
James
v.
Stockham
*291
Valves & Fittings Co.,
559 F. 2d 310 (1977).
[21]
Furthermore, the paragraph in which the court finally concludes that the USW seniority system is unprotected by § 703(h) strongly suggests that the outcome was the product of the court's independent consideration of the totality of the circumstances it found in the record.
Second and more fundamentally, when the court stated that it was convinced that a mistake had been made, it then identified not only the mistake but also the source of that mistake. The mistake of the District Court was that on the record there could be no doubt about the existence of a discriminatory purpose. The source of the mistake was the District Court's failure to recognize the relevance of the racial purposes of IAM. Had the District Court "given the I. A. M.'s role in the creation and establishment of the seniority system its due consideration," it "might have reached a different conclusion."
Supra,
at 284.
When an appellate court discerns that a district court has failed to make a finding because of an erroneous view of the law, the usual rule is that there should be a remand for further proceedings to permit the trial court to make the missing findings:
"[F]actfinding is the basic responsibility of district courts, rather than appellate courts, and . . . the Court of Appeals should not have resolved in the first instance
*292
this factual dispute which had not been considered by the District Court."
DeMarco
v.
United States,
415 U. S. 449, 450, n. (1974).
[22]
Likewise, where findings are infirm because of an erroneous view of the law, a remand is the proper course unless the record permits only one resolution of the factual issue.
Kelley
v.
Southern Pacific Co.,
419 U. S. 318, 331-332 (1974). All of this is elementary. Yet the Court of Appeals, after holding that the District Court had failed to consider relevant evidence and indicating that the District Court might have come to a different conclusion had it considered that evidence, failed to remand for further proceedings as to the intent of IAM and the significance, if any, of such a finding with respect to the intent of USW itself. Instead, the Court of Appeals made its own determination as to the motives of IAM, found that USW had acquiesced in the IAM conduct, and apparently concluded that the foregoing was sufficient to remove the system from the protection of § 703(h).
[23]
*293
Proceeding in this manner seems to us incredible unless the Court of Appeals construed its own well-established Circuit rule with respect to its authority to arrive at independent findings on ultimate facts free of the strictures of Rule 52(a) also to permit it to examine the record and make its own independent findings with respect to those issues on which the district court's findings are set aside for an error of law. As we have previously said, however, the premise for this conclusion is infirm: whether an ultimate fact or not, discriminatory intent under § 703(h) is a factual matter subject to the clearly-erroneous standard of Rule 52(a). It follows that when a district court's finding on such an ultimate fact is set aside for an error of law, the court of appeals is not relieved of the usual requirement of remanding for further proceedings to the tribunal charged with the task of factfinding in the first instance.
Accordingly, the judgment of the Court of Appeals is reversed, and the cases are remanded to that court for further proceedings consistent with this opinion.
So ordered.
JUSTICE STEVENS, concurring in part.
Except to the extent that the Court's preliminary comments on the burden of sustaining "any challenge to a seniority system under Title VII,"
ante,
at 277, are inconsistent with the views I expressed separately in
American Tobacco Co.
v.
Patterson, ante,
p. 86, I join the Court's opinion.
JUSTICE MARSHALL, with whom JUSTICE BLACKMUN joins except as to Part I, dissenting.
In 1971, a group of Negro employees at Pullman-Standard's Bessemer, Ala., plant brought this class action against Pullman-Standard, the United Steelworkers of America and its Local 1466 (USW), and the International Association of Machinists and its Local 372 (IAM). The plaintiffs alleged,
inter alia,
that the departmental seniority system negotiated
*294
by both unions discriminated against Negroes in violation of Title VII of the Civil Rights Act of 1964, 42 U. S. C. § 2000e
et seq.
(1976 ed. and Supp. IV), and the Civil Rights Act of 1866, 42 U. S. C. § 1981. In 1974, the District Court for the Northern District of Alabama concluded that the seniority system did not operate to discriminate against Negroes. A unanimous panel of the Fifth Circuit reversed. The court ruled that the District Court had committed several errors of law, including failure to give proper weight to the role of the IAM, and had relied on patently inaccurate factual conclusions.
Swint
v.
Pullman-Standard,
539 F. 2d 77, 95-96 (1976). On remand, the District Court again ruled that the seniority system was immune from attack under Title VII, this time finding that respondents had failed to show discriminatory intent as required by this Court's decision in
Teamsters
v.
United States,
431 U. S. 324 (1977).
Ante,
at 275. The Fifth Circuit again unanimously rejected the conclusion of the District Court. 624 F. 2d 525 (1980). The majority now reverses the Fifth Circuit's second unanimous decision on the ground that the Court of Appeals did not pay sufficient homage to the "clearly erroneous" rule, Fed. Rule Civ. Proc. 52(a), in concluding that the seniority system at Pullman-Standard was the product of intentional discrimination against Negroes. Because I cannot agree with the premise of the majority's decision to remand these cases for yet another trial, or with its application of that premise to the facts of this case, I respectfully dissent.
I
The majority premises its holding on the assumption that " `absent a discriminatory purpose, the operation of a seniority system cannot be an unlawful employment practice even if the system has some discriminatory consequences.' "
Ante,
at 277, quoting
Trans World Airlines, Inc.
v.
Hardison,
432 U. S. 63, 82 (1977). As I have previously indicated, I do not find anything in the relevant statutory language or legislative
*295
history to support the proposition that § 703(h) of Title VII immunizes a seniority system that perpetuates past discrimination, as the system at issue here clearly does, simply because the plaintiffs are unable to demonstrate to this Court's satisfaction that the system was adopted or maintained for an invidious purpose. See
Teamsters
v.
United States, supra,
at 377-394 (opinion of MARSHALL, J.). In my opinion, placing such a burden on plaintiffs who challenge seniority systems with admitted discriminatory impact, a burden never before imposed in civil suits brought under Title VII, frustrates the clearly expressed will of Congress and effectively "freeze[s] an entire generation of Negro employees into discriminatory patterns that existed before the Act."
Quarles
v.
Philip Morris, Inc.,
279 F. Supp. 505, 516 (ED Va. 1968) (Butzner, J.).
II
Even if I were to accept this Court's decision to impose this novel burden on Title VII plaintiffs, I would still be unable to concur in its conclusion that the Fifth Circuit's decision should be reversed for failing to abide by Rule 52(a). The majority asserts that the Court of Appeals in this action ignored the clearly-erroneous rule and made an independent determination of discriminatory purpose. I disagree. In my view, the court below followed well-established legal principles both in rejecting the District Court's finding of no discriminatory purpose and in concluding that a finding of such a purpose was compelled by all of the relevant evidence.
The majority concedes, as it must, that the "Court of Appeals acknowledged and correctly stated the controlling standard of Rule 52(a)."
Ante,
at 290. In a footnote to its opinion, the Court of Appeals plainly states that findings of fact may be overturned only if they are either "clearly erroneous" or "made under an erroneous view of controlling legal principles." 624 F. 2d, at 533, n. 6. Furthermore, as the majority notes,
ante,
at 283, the Court of Appeals justified its decision to reject the District Court's finding that the seniority
*296
system was not the result of purposeful discrimination by stating: "An analysis of the totality of the facts and circumstances surrounding the creation and continuance of the departmental system at Pullman-Standard leaves us with the
definite and firm conviction that a mistake has been made.
" 624 F. 2d, at 533 (emphasis added; footnote omitted).
[1]
I frankly am at a loss to understand how the Court of Appeals could have expressed its conclusion that the District Court's finding on the issue of intent was clearly erroneous with any more precision or clarity.
The majority rejects the Court of Appeals' clear articulation and implementation of the clearly-erroneous rule on the apparent ground that in the course of correctly setting forth the requirements of Rule 52(a), the court also included the following quotation from its prior decision in
East
v.
Romine, Inc.,
518 F. 2d 332, 339 (1975):
" `Although discrimination
vel non
is essentially a question of fact it is, at the same time, the ultimate issue for resolution in this case, being expressly proscribed by 42 U. S. C. A. § 2000e-2(a). As such, a finding of discrimination or nondiscrimination is a finding of ultimate fact. [Cites omitted]. In reviewing the district court's findings, therefore, we will proceed to make an independent determination of appellant's allegations of discrimination, though bound by findings of subsidiary fact which are themselves not clearly erroneous.'" 624 F. 2d, at 533, n. 6.
The only question presented by this case, therefore, is whether this reference to
East
v.
Romine, Inc
.,
should be read as negating the Court of Appeals' unambiguous acknowledgment
*297
of the "controlling standard of Rule 52."
Ante,
at 290. The majority bases its affirmative answer to that question on two factors. First, the majority contends that the Court of Appeals must not have properly respected the clearly-erroneous rule because its acknowledgment that Rule 52(a) supplied the controlling standard "came late in the court's opinion."
Ante,
at 290. Second, the Court of Appeals "identified not only the mistake" that it felt had been made, "but also the source of that mistake."
Ante,
at 291. If the Court of Appeals had really been applying the clearly-erroneous rule, it should have abided by the "usual requirement of remanding for further proceedings to the tribunal charged with the task of factfinding in the first instance."
Ante,
at 293.
Neither of these arguments justifies the majority's conclusion that these cases must be remanded for a fourth trial on the merits. I am aware of no rule of decision embraced by this or any other court that places dispositive weight on whether an accurate statement of controlling principle appears "early" or late in a court's opinion. Nor does the majority suggest a basis for this unique rule of interpretation. So long as a court acknowledges the proper legal standard, I should think it irrelevant whether it chooses to set forth that standard at the beginning or at the end of its opinion. The heart of the majority's argument, therefore, is that the failure to remand the action to the District Court after rejecting its conclusion that the seniority system was "bona fide" within the meaning of § 703(h) indicates that the Court of Appeals did not properly follow the clearly-erroneous rule. Before addressing this issue, however, it is necessary to examine the nature of the finding of "intent" required by this Court in
Teamsters,
the procedure that courts of appeals should follow in reviewing a district court's finding on intent, and the extent to which the court below adhered to that procedure in this case.
The District Court examined the four factors approved by the Fifth Circuit in
James
v.
Stockham Valves & Fittings
*298
Co.,
559 F. 2d 310 (1977), cert. denied, 434 U. S. 1034 (1978), to determine whether the departmental seniority system at Pullman-Standard was adopted or maintained for a discriminatory purpose. Although indicating that these four factors are not the only way to demonstrate the existence of discriminatory intent,
[2]
the Court today implicitly acknowledges that proof of these factors satisfies the requirements of
Teamsters.
[3]
In particular, the majority agrees that a finding of discriminatory intent sufficient to satisfy
Teamsters
can be based on circumstantial evidence, including evidence of discriminatory impact. See
ante,
at 289; see also
Arlington Heights
v.
Metropolitan Housing Dev. Corp.,
429 U. S. 252, 266, 267 (1977).
Given the nature of this factual inquiry, the court of appeals must first determine whether the district court applied correct legal principles and therefore
considered
all of the legally relevant evidence presented by the parties. This, as the majority acknowledges, is a "legal" function that the court of appeals must perform in the first instance.
Ante,
at 282, 283. Second, the court of appeals must determine whether the district court's finding with respect to intent is
supported
by all of the legally relevant evidence. This, the Court holds today, is generally a factual determination limited by the dictates of Rule 52(a). Finally, if the court of appeals sets aside the district court's finding with respect to intent, either because that finding is clearly erroneous or because it is based on an erroneous legal standard, it may determine, in the interest of judicial economy, whether the legally
*299
relevant evidence presented to the district court "permits only one resolution of the factual issue."
Ante,
at 292. If only one conclusion is possible, the reviewing court is free to find the existence of the fact in question as a matter of law. See
Bigelow
v.
Virginia,
421 U. S. 809, 826-827 (1975);
Levin
v.
Mississippi River Fuel Corp.,
386 U. S. 162, 170 (1967).
A common-sense reading of the opinion below demonstrates that the Court of Appeals followed precisely this course in examining the issue of discriminatory intent. Even the majority concedes that the Court of Appeals determined that the District Court committed "legal error" by failing to consider all of the relevant evidence in resolving the first and the third
James
factors.
Ante,
at 282, 283. With respect to the first
James
factor — whether the system inhibits all employees equally from transferring between seniority units — the District Court found that the departmental system "locked" both Negro and white workers into departments by discouraging transfers. The District Court acknowledged that Negroes might suffer a greater impact because the company's previous discriminatory policy of openly maintaining "Negro" jobs and "white" jobs had caused Negroes to be concentrated in less desirable positions. The District Court concluded, however, that this differential impact was irrelevant in determining whether the seniority system operated neutrally. The Court of Appeals properly held that the District Court erred in failing to consider the fact that the departmental system locked Negroes into less desirable jobs.
Similarly, as for the third
James
factor — whether the seniority system had its genesis in racial discrimination — the District Court rejected respondents' argument that the motives of the IAM were relevant. It concluded that the USW could not be charged with the racial bias of the IAM. The Court of Appeals held that this conclusion was erroneous because the "motives and intent of the I. A. M. in 1941 and 1942
*300
are significant in consideration of whether the seniority system has its genesis in racial discrimination." 624 F. 2d, at 532.
[4]
As the majority acknowledges, where findings of fact " `are made under an erroneous view of controlling legal principles, the clearly erroneous rule does not apply, and the findings may not stand.' "
Ante,
at 285, quoting 624 F. 2d, at 533, n. 6; see also
Kelley
v.
Southern Pacific Co.,
419 U. S. 318, 323 (1974);
United States
v.
General Motors Corp.,
384 U. S. 127, 141, n. 16 (1966);
United States
v.
Singer Manufacturing Co.,
374 U. S. 174, 194, n. 9 (1963);
United States
v.
Parke, Davis & Co.,
362 U. S. 29, 44 (1960);
Rowe
v.
General Motors Corp.,
457 F. 2d 348, 356, n. 15 (CA5 1972). Having found that the District Court's findings as to the first and third
James
factors were made under an erroneous view of controlling legal principles, the Court of Appeals was
compelled
to set aside those findings free of the requirements of the clearly-erroneous rule.
[5]
But once these two findings were set aside, the District Court's conclusion that the departmental system was bona fide within the meaning of § 703(h) also had to be rejected, since that conclusion was based at least in part on its erroneous determinations concerning the first and the third
James
factors.
At the very least, therefore, the Court of Appeals was entitled to remand this action to the District Court for the purpose
*301
of reexamining the bona fides of the seniority system under proper legal standards. However, as we have often noted, in some cases a remand is inappropriate where the facts on the record are susceptible to only one reasonable interpretation. See
Dayton Board of Education
v.
Brinkman,
443 U. S. 526, 534-537 (1979);
Bigelow
v.
Virginia, supra,
at 826-827. In such cases, "[e]ffective judicial administration" requires that the court of appeals draw the inescapable factual conclusion itself, rather than remand the case to the district court for further needless proceedings.
Levin
v.
Mississippi River Fuel Corp.,
386 U. S., at 170. Such action is particularly appropriate where the court of appeals is in as good a position to evaluate the record evidence as the district court. The major premise behind the deference to trial courts expressed in Rule 52(a) is that findings of fact "depend peculiarly upon the credit given to witnesses by those who see and hear them."
United States
v.
Yellow Cab Co.,
338 U. S. 338, 341 (1949); see also
United States
v.
Oregon State Medical Society,
343 U. S. 326, 332 (1952). Indeed Rule 52(a) expressly acknowledges the importance of this factor by stating that "due regard shall be given to the opportunity of the trial court to judge of the credibility of the witnesses." Consequently, this Court has been especially reluctant to resolve factual issues which depend on the credibility of witnesses. See generally
United States
v.
Oregon State Medical Society, supra,
at 332.
In the cases before the Court today this usual deference is not required because the District Court's findings of fact were entirely based on documentary evidence.
[6]
As we
*302
noted in
United States
v.
General Motors Corp., supra,
at 141, n. 16, "the trial court's customary opportunity to evaluate the demeanor and thus the credibility of the witnesses, which is the rationale behind Rule 52(a) . . . , plays only a restricted role [in] a `paper case.'" See also
Jennings
v.
General Medical Corp.,
604 F. 2d 1300, 1305 (CA10 1979) ("When the findings of a trial court are based on documentary, rather than oral evidence, they do not carry the same weight on appellate review");
Orvis
v.
Higgins,
180 F. 2d 537, 539 (CA2 1950).
[7]
I believe that the Court of Appeals correctly determined that a finding of discriminatory intent was compelled by the documentary record presented to the District Court. With respect to three of the four
James
factors, the Court of Appeals found overwhelming evidence of discriminatory intent. First, in ruling that the District Court erred by not acknowledging the legal significance of the fact that the seniority system locked Negroes into the least remunerative jobs in the company, the Court of Appeals determined that such disproportionate impact demonstrated that the system did not " `operat[e] to discourage all employees equally from transferring between seniority units.' " 624 F. 2d, at 530, quoting
*303
James
v.
Stockham Valves & Fittings Co
.,
559 F. 2d, at 352. Second, noting that "[n]o credible explanation ha[d] been advanced to sufficiently justify" the existence of two separate Die and Tool Departments and two separate Maintenance Departments, a condition not found at any other Pullman-Standard plant, or the creation of all-white and all-Negro departments at the time of unionization and in subsequent years, the Court of Appeals concluded that the second
James
factor had not been satisfied.
[8]
624 F. 2d, at 533. Finally, with respect to the third
James
factor the Court of Appeals found that once the role of the IAM was properly recognized, it was "crystal clear that considerations of race permeated the negotiation and the adoption of the seniority system in 1941 and subsequent negotiations thereafter." 624 F. 2d, at 532.
[9]
*304
After reviewing all of the relevant record evidence presented to the District Court, the Court of Appeals concluded: "There is no doubt, based upon the record in this case, about the existence of a discriminatory purpose."
Id.,
at 533. Because I fail to see how the Court of Appeals erred in carrying out its appellate function, I respectfully dissent from the majority's decision to prolong respondents' 11-year quest for the vindication of their rights by requiring yet another trial.
NOTES
[*]  Together with No. 80-1193,
United Steelworkers of America, AFLCIO, et al.
v.
Swint et al.,
also on certiorari to the same court.
[†]
Solicitor General Lee, Assistant Attorney General Reynolds, Deputy Solicitor General Wallace, Jessica Dunsay Silver, Marie E. Klimesz, Constance L. Dupre, Philip B. Sklover,
and
Vella M. Fink
filed a brief for the United States et al. as
amici curiae
urging affirmance.
[1]  In their original complaint, besides challenging the seniority system discussed in this opinion, plaintiffs also alleged discrimination in job assignments and promotions and the failure to post publicly a list of changes in assignments. These were all brought as "class" issues. Two charges of individual discrimination were also brought. The Court of Appeals held that the Company had violated Title VII in making job assignments and in selecting foremen. In granting certiorari, we declined to review those aspects of the decision.
[2]  The procedural history of these cases is rather complex. The original complaint was filed in 1971. Since that time the case has been tried three times and has twice been reviewed by the Court of Appeals.
[3]  In 1941, prior to unionization, the Bessemer plant was divided into 20 departments. By 1954, there were 28 departments — 26 USW units and 2 International Association of Machinists and Aerospace Workers (IAM) units. The departments remained essentially unchanged after 1954.
[4]  The International Brotherhood of Electrical Workers (IBEW) gained representation status for two small departments. The IBEW unit was all white. IBEW, however was decertified in 1946 and its members were reabsorbed into a department represented by USW.
[5]  A departmental seniority system was part of the initial collective-bargaining agreement between the Company and USW in 1942. Between 1947 and 1954, however, the seniority system changed from one based on departments to one based upon particular occupations within departments. In 1954, the system went back to a departmental base.
[6]  The only exceptions, until 1972 (see n. 7,
infra
), were for employees transferring at the request of the Company or for those electing transfer in lieu of layoff.
[7]  In 1972, the Company entered into an agreement with the Department of Labor to bring its employment practices into compliance with Executive Order No. 11246, 3 CFR 339 (1964-1965 Comp.). This provided an exception to the departmental limit on seniority, allowing certain black employees to make interdepartmental transfers without any loss of seniority.
[8]  The Fifth Circuit relied upon the following passage in
Teamsters,
431 U. S., at 355-356:
"The seniority system in this litigation is entirely bona fide. It applies equally to all races and ethnic groups. To the extent that it `locks' employees into non-line-driver jobs, it does so for all. . . . The placing of line drivers in a separate bargaining unit from other employees is rational, in accord with the industry practice, and consistent with National Labor Relation Board precedents. It is conceded that that seniority system did not have its genesis in racial discrimination, and that it was negotiated and has been maintained free from any illegal purpose."
This passage was of course not meant to be an exhaustive list of all the factors that a district court might or should consider in making a finding of discriminatory intent.
[9]  The court specifically declined to make any finding on whether the notransfer provision of the seniority system had a greater relative effect on blacks than on whites, because of qualitative differences in the departments in which they were concentrated. It believed that such an inquiry would have been inconsistent with the earlier Fifth Circuit opinion in this case.
[10]  In particular, the court focused on the history of the unionization process at the plant and found certain of the departmental divisions to be based on the evolving relationship between USW and IAM.
[11]  With respect to USW, the District Court found that "[u]nion meetings were conducted with different sides of the hall for white and black members, and social functions of the union were also segregated." App. A-142. It also found, however, that "[w]hile possessing some of the trappings taken from an otherwise segregated society, the USW local was one of the few institutions in the area which did not function in fact to foster and maintain segregation; rather, it served a joint interest of white and black workers which had a higher priority than racial considerations."
Id.,
at A-143.
[12]  It does not appear to us that the District Court actually found a qualitative difference but held it to be irrelevant. The relevant passage of the District Court opinion read as follows: "By ranking the twenty-eight USW and IAM departments according to some perceived order of desirability, one could . . . attempt to measure the relative effect of the no-transfer rule on white and black employees . . . . It may well be that a somewhat greater impact was felt by blacks than whites although . . . this conclusion is by no means certain."
Id.,
at A-134.
[13]  The original complaint in this case did not mention IAM. Prior to the first trial, respondents sought and received leave to amend their complaint to add IAM as a Rule 19 defendant, "insofar as the relief requested may involve or infringe upon the provisions of such Union's collective bargaining agreement with the Company." Order of the District Court, June 4, 1974 (App. 29).
[14]  In
United States
v.
United States Gypsum Co.,
333 U. S. 364, 395 (1948), this Court characterized the clearly-erroneous standard as follows:
"A finding is `clearly erroneous' when although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed."
We note that the Court of Appeals quoted this passage at the conclusion of its analysis of the District Court opinion.
Supra,
at 283.
[15]  See
Jackson
v.
City of Killeen,
654 F. 2d 1181, 1184 (1981);
Payne
v.
McLemore's Wholesale & Retail Stores,
654 F. 2d 1130, 1147 (1981);
Wilkins
v.
University of Houston,
654 F. 2d 388, 390 (1981);
Lindsey
v.
Mississippi Research & Development Center,
652 F. 2d 488, 492 (1981);
Rohde
v.
K. O. Steel Castings, Inc.,
649 F. 2d 317, 320 (1981);
Joshi
v.
Florida State University,
646 F. 2d 981, 986 (1981);
Phillips
v.
Joint Legislative Committee,
637 F. 2d 1014, 1024 (1981);
Danner
v.
United States Civil Service Comm'n,
635 F. 2d 427, 430-431 (1981);
Thompson
v.
Leland Police Dept.,
633 F. 2d 1111, 1112 (1980);
Crawford
v.
Western Electric Co.,
614 F. 2d 1300, 1311 (1980);
Burdine
v.
Texas Dept. of Community Affairs,
608 F. 2d 563, 566 (1979);
Williams
v.
Tallahassee Motors, Inc.,
607 F. 2d 689, 690 (1979);
Parson
v.
Kaiser Aluminum & Chemical Corp.,
575 F. 2d 1374, 1382 (1978);
Causey
v.
Ford Motor Co.,
516 F. 2d 416, 420-421 (1975);
East
v.
Romine, Inc.,
518 F. 2d 332, 338-339 (1975).
[16]  There is some indication in the opinions of the Court of Appeals for the Fifth Circuit (see n. 15,
supra
) that the Circuit rule with respect to "ultimate facts" is only another way of stating a standard of review with respect to mixed questions of law and fact — the ultimate "fact" is the statutory, legally determinative consideration (here, intentional discrimination) which is or is not satisfied by subsidiary facts admitted or found by the trier of fact. As indicated in the text, however, the question of intentional discrimination under § 703(h) is a pure question of fact. Furthermore, the Court of Appeals' opinion in this case appears to address the issue as a question of fact unmixed with legal considerations.
At the same time, this Court has on occasion itself indicated that findings on "ultimate facts" are independently reviewable. In
Baumgartner
v.
United States,
322 U. S. 665 (1944), the issue was whether or not the findings of the two lower courts satisfied the clear-and-convincing standard of proof necessary to sustain a denaturalization decree. The Court held that the conclusion of the two lower courts that the exacting standard of proof had been satisfied was not an unreviewable finding of fact but one that a reviewing court could independently assess. The Court referred to the finding as one of "ultimate" fact, which in that case involved an appraisal of the strength of the entire body of evidence. The Court said that the significance of the clear-and-convincing proof standard "would be lost" if the ascertainment by the lower courts whether that exacting standard of proof had been satisfied on the whole record were to be deemed a "fact" of the same order as all other "facts not open to review here."
Id.,
at 671.
The Fifth Circuit's rule on appellate consideration of "ultimate facts" has its roots in this discussion in
Baumgartner.
In
Galena Oaks Corp.
v.
Scofield,
218 F. 2d 217 (CA5 1954), in which the question was whether the gain derived from the sale of a number of houses was to be treated as capital gain or ordinary income, the Court of Appeals relied directly on
Baumgartner
in holding that this was an issue of "ultimate fact" that an appellate court may review free of the clearly-erroneous rule.
Causey
v.
Ford Motor Co., supra,
at 421, relying on
Galena Oaks Corp.
v.
Scofield, supra
,
said that "although discrimination
vel non
is essentially a question of fact, it is, at the same time, the ultimate issue for resolution in this case" and as such, was deemed to be independently reviewable. The passage from
East
v.
Romine, Inc., supra,
at 339, which was repeated in the cases before us now,
supra,
at 285, rested on the opinion in
Causey
v.
Ford Motor Co.
Whatever
Baumgartner
may have meant by its discussion of "ultimate facts," it surely did not mean that whenever the result in a case turns on a factual findings, an appellate court need not remain within the constraints of Rule 52(a).
Baumgartner's
discussion of "ultimate facts" referred not to pure findings of fact — as we find discriminatory intent to be in this context — but to findings that "clearly impl[y] the application of standards of law." 322 U. S., at 671.
[17]  As we noted above, the Court of Appeals did at certain points purport to correct what it viewed as legal errors on the part of the District Court. The presence of such legal errors may justify a remand by the Court of Appeals to the District Court for additional factfinding under the correct legal standard.
Infra,
at 291-292.
[18]  See,
e. g.,
Furnco Construction Corp.
v.
Waters,
438 U. S. 567, 580 (1978): "Proof that [an employer's] work force was racially balanced or that it contained a disproportionately high percentage of minority employees is not wholly irrelevant on the issue of intent when that issue is yet to be decided."
[19]  We need not, therefore, address the much-mooted issue of the applicability of the Rule 52(a) standard to mixed questions of law and fact —
i. e.,
questions in which the historical facts are admitted or established, the rule of law is undisputed, and the issue is whether the facts satisfy the statutory standard, or to put it another way, whether the rule of law as applied to the established facts is or is not violated. There is substantial authority in the Circuits on both sides of this question. Compare
United States ex rel. Johnson
v.
Johnson,
531 F. 2d 169, 174, n. 12 (CA3 1976);
Stafos
v.
Jarvis,
477 F. 2d 369, 372 (CA10 1973); and
Johnson
v.
Salisbury,
448 F. 2d 374, 377 (CA6 1971), with
Rogers
v.
Bates,
431 F. 2d 16, 18 (CA8 1970); and
Pennsylvania Casualty Co.
v.
McCoy,
167 F. 2d 132, 133 (CA5 1948). There is also support in decisions of this Court for the proposition that conclusions on mixed questions of law and fact are independently reviewable by an appellate court,
e. g.,
Bogardus
v.
Commissioner,
302 U. S. 34, 39 (1937);
Helvering
v.
Tex-Penn Oil Co.,
300 U. S. 481, 491 (1937);
Helvering
v.
Rankin,
295 U. S. 123, 131 (1935). But cf.,
Commissioner
v.
Duberstein,
363 U. S. 278, 289 (1960);
Commissioner
v.
Heininger,
320 U. S. 467, 475 (1943).
[20]  Neither does the dissent contend that Rule 52(a) is inapplicable to findings of discriminatory intent. Rather, it contends, that the Rule was properly applied by the Court of Appeals.
[21]  In particular, in regard to the second
James
factor — whether the departmental structure was rational or in line with industry practice — the Court of Appeals did not focus on the evidentiary basis for any particular finding of the District Court. It appeared to make an independent examination of the record and arrive at its own conclusion contrary to that of the District Court. Likewise, in dealing with the genesis of the seniority system and whether or not the negotiation or maintenance of the system was tainted with racial discrimination, the Court of Appeals, while identifying what it thought was legal error in failing to consider the racial practices and intentions of IAM, did not otherwise overturn any of the District Court's findings as clearly erroneous.
[22]  See 5A J. Moore & J. Lucas, Moore's Federal Practice § 52.06[2] (1982) ("Where the trial court fails to make findings, or to find on a material issue, and an appeal is taken, the appellate court will normally vacate the judgment and remand the action for appropriate findings to be made");
Rule
v.
International Assn. of Bridge Workers,
568 F. 2d 558, 568 (CA8 1978);
Chicano Police Officer's Assn.
v.
Stover,
552 F. 2d 918, 921 (CA10 1977);
O'Neal
v.
Gresham,
519 F. 2d 803, 805 (CA4 1975);
Burch
v.
International Assn. of Machinists & Aerospace Workers, AFL-CIO,
433 F. 2d 561 (CA5 1970);
General Electric Credit Corp.
v.
Robbins,
414 F. 2d 208 (CA8 1969).
[23]  IAM's discriminatory motivation, if it existed, cannot be imputed to USW. It is relevant only to the extent that it may shed some light on the purpose of USW or the Company in creating and maintaining the separate seniority system at issue in these cases. A discriminatory intent on the part of IAM, therefore, does not control the outcome of these cases. Neither does the fact, if true, that USW acquiesced in racially discriminatory conduct on the part of IAM. Such acquiescence is not the equivalent of a discriminatory purpose on the part of USW.
[1]  As the majority acknowledges,
ante,
at 284-285, n. 14, this Court stated in
United States
v.
United States Gypsum Co.,
333 U. S. 364, 395 (1948), that a finding of fact is clearly erroneous if "the reviewing court on the entire evidence is left with
the definite and firm conviction that a mistake has been committed
" (emphasis added).
[2]  Contrary to the majority's suggestion,
ante,
at 279, n. 8, I find nothing in the Fifth Circuit's decision in
James
v.
Stockham Valves & Fittings Co
.
to imply that these factors constitute the only relevant criteria for determining discriminatory intent.
[3]  This conclusion would seem to be compelled since, as the majority notes, the
James
factors are nothing more than a summary of the criteria examined by this Court in
Teamsters,
431 U. S., at 355-356.
[4]  As the majority indicates in a footnote,
ante,
at 292, n. 23, the discriminatory motive of the IAM is "relevant . . . to the extent that it may shed light on the purpose of USW or the Company in creating and maintaining the separate seniority system at issue in this case." I do not read the Court of Appeals opinion in this action as holding anything more than that if the USW participated in establishing a system that was designed for the purpose of perpetuating past discrimination, the third
James
factor would be satisfied. Given that the IAM is a party to this litigation, its participation in the creation of the seniority system can hardly be deemed irrelevant.
[5]  It is therefore irrelevant that the Court of Appeals did not specifically hold that the District Court's other factual findings were clearly erroneous.
[6]  Only two witnesses testified during the brief hearing that the District Court conducted on the question whether the seniority system at Pullman-Standard was immune under § 703(h). Both of these witnesses were longtime Negro employees of Pullman-Standard who testified on behalf of
respondents
concerning racial segregation at the plant and by the USW. There is no indication in the District Court's opinion that it relied upon the testimony of these two witnesses in concluding that the system was bona fide within the meaning of § 703(h). The remainder of the record before the District Court consisted entirely of 139 exhibits submitted by respondents, the company, and the unions concerning the development and maintenance of the seniority system from 1940 through the 1970's.
[7]  This is not to say that the clearly-erroneous rule does not apply to "document" cases. See
United States
v.
Singer Manufacturing Co.,
374 U. S. 174, 194, n. 9 (1963). However, "when the decision of the court below rests upon an incorrect reading of an undisputed document, [the appellate] court is free to substitute its own reading of the document."
Eutectic Corp.
v.
Metco, Inc.,
579 F. 2d 1, 5 (CA2 1978). See also
McKensie
v.
Sea Land Service,
551 F. 2d 91 (CA5 1977);
Best Medium Pub. Co.
v.
National Insider, Inc.,
385 F. 2d 384 (CA7 1967), cert. denied, 390 U. S. 955 (1968);
United States ex rel. Binion
v.
O'Brien,
273 F. 2d 495 (CA3 1959), cert. denied, 363 U. S. 812 (1960).
[8]  Although the majority is correct in stating that the Court of Appeals did not "refer to or
expressly
apply the clearly-erroneous standard" in reaching this conclusion,
ante,
at 282 (emphasis added), the appellate court's adherence to the requirements of Rule 52(a) is nevertheless apparent from the following statement:
"The record evidence indicates that a significant number of one-race departments were established upon unionization at Pullman-Standard, and during the next twenty five years, one-race departments were carved out of previously mixed departments. The establishment and maintenance of the segregated departments appear to be based on
no other considerations than the objective to separate the races.
" 624 F. 2d, at 531 (emphasis added).
In my opinion, this statement is sufficient to satisfy the requirements of Rule 52(a), particularly in light of the Court of Appeals' general acknowledgment that it was bound by the clearly-erroneous rule. See
supra,
at 296-297.
[9]  Whether or not the Court of Appeals expressly ruled on the fourth
James
factor is irrelevant. As the Court of Appeals clearly stated, its conclusion was based on "the totality of the facts and circumstances surrounding the creation and continuance of the departmental system at Pullman-Standard." 624 F. 2d, at 533; see also
id.,
at 532 ("It is crystal clear that considerations of race permeated the negotiation and the adoption of the seniority system in 1941 and subsequent negotiations thereafter"), and
id.,
at 533 ("We consider significant in our decision . . . conditions of racial discrimination which affected the negotiation and renegotiation of the system. . ."). Even assuming that the District Court was correct in concluding that the system had been
maintained
free of any illegal purpose, the Court of Appeals was entitled to conclude that discriminatory intent had been demonstrated on the basis of other relevant evidence.