The Equal Employment Opportunity Commission (EEOC) brought suit against Ford Motor Company for violating Title VII, 42 U.S.C. § 2000e-2. EEOC alleged that Ford had discriminated against several women in its hiring practices at a parts warehouse in Charlotte, North Carolina. The district court found that Ford had discriminated against' ten individual women and awarded them back pay. We affirm, but remand the case to the district court for consideration of additional relief.
I.
Ford operates a parts warehouse in Charlotte. A number of the workers at the warehouse are “picker-packers.” These employees fill orders for automotive parts by “picking” out ordered parts from storage and “packing” them for delivery. Until 1972 when two women were hired as temporary workers, no woman had ever worked as a picker-packer at the warehouse. Ford did not hire a woman to work as a picker-packer on a permanent basis until 1975.
In 1971, Judy Gaddis applied for employment as a picker-packer at the warehouse. When she did not receive a job, she filed a sex discrimination charge with the EEOC. The EEOC, after unsuccessfully trying to secure an acceptable conciliation agreement, sued Ford under Title VII on behalf of two groups of women. The first group *186consisted of Gaddis and two others who applied for picker-packer jobs which Ford filled with men in the summer of 1971. The second group consisted of eleven women who applied for a single picker-packer opening in 1973. Again in 1973 Ford had chosen a man, not a woman, to work at the warehouse.
The 1971 woman applicants were Gaddis, Rebecca Starr and Zettie Smith. In response to proof that as of 1971 it had never hired women as picker-packers and proof of a discriminatory hiring policy, Ford attempted to demonstrate that it had not in fact discriminated against women in its hiring practices at the' Charlotte warehouse. Further, Ford tried to show that these particular women had not applied before the 1971 openings were filled. The district court rejected Ford’s arguments, finding that Ford had followed a discriminatory hiring policy and that the three women had applied prior to the time in 1971 when Ford hired three men.
With the aid of a special master, the district court calculated the back pay due Gaddis, Starr, and Smith. Based upon the employment history of the three men who were hired in 1971,1 the district court decided that, if the women had been hired in 1971, they would have worked at Ford until the time of trial. The special master thus began the back pay period with the date of the 1971 hirings and ended it at the date of trial in 1977.
With respect to Gaddis and Starr only, Ford attempted to terminate its back pay liability at three different junctures: in 1973 when they were recalled for one year by their former employer General Motors, later in 1973 when they rejected job offers from Ford, and in 1975 when they entered a CETA nurses training program. The district court refused to truncate the back pay periods at any of these points. Further, the district court rejected Ford’s request to subtract unemployment compensation from the back pay awards.
In 1973, Ford filled a vacancy by hiring a man, Robert Simpson, even though eleven women had applied for the position. The district court ruled that EEOC proved a prima facie case of sex discrimination, in this instance based upon Ford’s hiring procedures. After referral, the special master found that Simpson was hired in part because a warehouse employee took an interest in his application and that Ford did not seriously consider the applications of the eleven women. The special master determined that Ford did not intentionally discriminate against the women when it hired Simpson. But the special master concluded that Ford’s practice of hiring new employees on the recommendation of present workers, who were all male, without giving other applications full consideration, had a discriminatory effect upon the women applicants. The district court accepted the conclusion of the special master.
Pursuant to the district court’s instructions, the special master determined the back pay due each of the women who had applied for the 1973 opening. Smith was among the eleven who had applied, but she was not included in the 1973 group because her injury was remedied by the award she received as a member of the 1971 group. Of the remaining ten applicants, seven appeared before the special master and demonstrated their financial injury.2 Under the district court’s directions, the special master divided each woman’s award by seven, since the seven women had applied for a single vacancy.
Ford attempted to reduce its back pay liability by arguing that the back pay peri*187ods of the seven women should terminate with the end of Simpson’s employment at Ford. Simpson was laid off by Ford in 1974. He was not recalled; instead, Ford hired a woman, its first permanent female picker-packer, to fill his job. The special master found, however, that it was Ford’s usual practice to recall employees who had been laid off. Therefore, the special master refused to end the back pay period with Simpson’s termination, and the district court approved that decision.
The district court granted the two groups of women no additional relief beyond back pay. No mention of additional relief was made in the report of the special master, nor did the district court discuss other remedies in its findings of fact and conclusions of law, its referral order, or its final order.
Finally, the district court ordered Ford to report in the future about the company’s efforts to hire women. The court declined however, to grant general prospective relief against sex discrimination at the warehouse, due to the vigorous affirmative action program Ford had pursued since 1975.
II.
Because the 1971 openings and the 1973 opening involve different issues, we examine these two incidents of alleged discrimination separately.
A. 1971 Openings
(1) Discrimination
(a) Prima Facie Case
Under Title VII, EEOC bore the burden of establishing a prima facie ease that Ford discriminated against Smith, Gaddis, and Starr. McDonnell Douglas Corp. v. Green, 411 U.S. 792, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973). In this case, EEOC amply fulfilled its obligation.
EEOC showed that Ford had never hired a woman to work in its parts warehouse until 1972, when Gaddis and Starr were hired on a temporary basis, and that Ford hired its first woman permanent employee in 1975. In addition, Gaddis and Starr testified that when they applied in 1971 the receptionist told them that Ford did not hire women to work in the warehouse. Ford disputes this version of the facts and contends that the women were told that Ford “did not have females working [in the warehouse].” This version by itself supports an inference that Ford was discriminating on the basis of sex, but, in any event, the district court chose to believe the testimony of Gaddis and Starr. On these facts, the district court certainly made no clearly erroneous factual findings nor any improper legal determinations when it decided that EEOC established a prima facie case of sex discrimination by Ford against Gaddis, Starr, and Smith.
Once EEOC proved this prima facie case, Ford had the burden of proving a legitimate, nondiscriminatory reason for hiring the three men instead of Gaddis, Starr, and Smith. Board of Trustees of Keene State College v. Sweeney, 439 U.S. 24, 99 S.Ct. 295, 58 L.Ed.2d 216 (1978) (per curiam); McDonnell Douglas. Ford argues that Gad-dis, Starr, and Smith had not applied when it filled the three vacancies and that therefore they were not in reality victims of discrimination as a result of Ford’s 1971 hiring decisions.
(b) Rebuttal: Gaddis and Starr
Ford insists that Gaddis and Starr applied a few days after the vacancies had been filled. Instead of Gaddis and Starr, Ford hired James Blanchard and Phillip Coe. Ford’s employment records indicated that Blanchard was hired on August 5 and that Coe was hired on August 16. It is undisputed that Gaddis and Starr applied on July 21. Thus, it would appear that Gaddis and Starr applied before Blanchard and Coe were hired.
Ford’s warehouse manager, however, testified that he thought that he hired the two men by phone before July 21, so that they could give their employers notice before joining Ford. The testimony of the two men themselves indicated otherwise. Coe testified that was hired “only just two or three weeks” before he went to work. Since Coe began work on August 16, according to his testimony he was hired no *188earlier than July 27, after Gaddis and Starr applied. Blanchard testified that he had been contacted by Miller “a week or two weeks” before he started his new job on August 5. He, then, was hired no earlier than July 23, also after Gaddis and Starr applied.
The district court, while noting that there was conflicting evidence, concluded that “[a]t least two of the men employed in August — Coe and Blanchard — were offered their jobs after Gaddis and Starr applied” (emphasis in original). Obviously, the district court compared the testimony of the warehouse manager with the testimony of Blanchard and Coe and found the recollections of Blanchard and Coe more persuasive. Credibility determinations are the province of the district judge, and, in light of this evidence, we do not find his conclusions clearly erroneous.3
(c) Rebuttal: Smith
Ford challenges Smith’s recovery by contending that she did not apply at all during 1971. Smith had originally been considered only as one of the victims of possible discrimination in 1973, because she had applied for a warehouse job in 1973. Before the district court, however, Smith testified that she had also applied for a warehouse job in April and June of 1971. Smith said she applied in 1971 because early in that year her employer had warned its workers that, due to a change in plant ownership, their jobs might not be secure. Smith stated that she applied in April when she began searching for a new job, and in June, during her vacation.
Ford relies on the fact that it found no 1971 application from Smith in its files. Ford’s records included only the 1973 application. The absence of a 1971 application carries little weight, however, because Ford routinely discarded applications after the passage of one year.
Further, Ford points to the testimony of its warehouse manager and its receptionist, who stated that Gaddis and Starr, who applied in July, were the first women applicants for warehouse jobs. But, in the face of Smith’s testimony concerning her June application and her reasons for applying, the general recollections of the warehouse manager and the receptionist are not particularly persuasive. The district court, after observing all three witnesses, credited Smith’s testimony. This credibility determination was for the district court and was not clearly erroneous.4
*189In addition to deciding that Smith applied in 1971, the district court concluded that “although she was qualified to be employed in the warehouse, her application was never seriously considered because she is a woman.” Ford suggests that there is no evidence in the record to show that Smith’s application was not considered. Contrary to Ford’s position, there was sufficient evidence to support the decision of the district court.
EEOC showed that Ford never hired women as permanent employees until 1975 and that Gaddis and Starr had been told in 1971 that Ford did not hire women to work in the warehouse. Based on these two compelling pieces of proof, the district judge could reasonably infer that Smith’s application, like the applications of Gaddis and Starr, did not receive serious consideration because of her sex. We cannot say that this factual determination was clearly erroneous.5
(2) Back Pay
In order to remedy the effects of discrimination against Gaddis, Starr, and Smith, the district court awarded them back pay. With the help of a special master, following the direction of this court in UTU v. Norfolk & Western Railway, 532 F.2d 336, 341 (4 Cir. 1975), cert. denied, 425 U.S. 934, 96 S.Ct. 1664, 48 L.Ed.2d 175 (1976), the court constructed the hypothetical work histories which the three women would have had at Ford if they had not been victims of discrimination. The court then used these hypothetical work histories to calculate back pay. The district court determined that if the women had been hired in 1971, they would have been working at Ford at the time of trial in 1977, as were two of the men who were hired in 1971. Accordingly, the court granted the women back pay from the time the men were hired in 1971 through 1977.
Ford urges that the back pay periods of Gaddis and Starr should have been terminated at three different points short of the date of trial. In addition, Ford insists that the awards of all three women should be reduced by the amount of unemployment compensation they received during the period.
(a) Subsequent Recall
Before Gaddis and Starr applied for the Ford warehouse jobs, they had worked in similar positions for General Motors. In 1971 Gaddis and Starr applied for work at the Ford warehouse because they had been laid off from their General Motors jobs. In January, 1973, Gaddis and Starr were recalled to General Motors. They worked at General Motors for approximately one year and were then laid off for a final time.
The district court subtracted their earnings at General Motors from their back pay awards. Ford insists that the district court *190should have gone further. Ford contends that the back pay period of Gaddis and Starr should have been terminated when they rejoined General Motors.
Ford’s argument misses the point of a Title VII back pay award. Under Title VII, a district court awards back pay in order to put the discrimination victim in the position she would have occupied if the defendant had not discriminated against her. Albemarle Paper Co. v. Moody, 422 U.S. 405, 95 S.Ct. 2362, 45 L.Ed.2d 280 (1975). Pursuant to this purpose, the district court constructed the claimants’ hypothetical employment histories, calculated what they would have earned if they had not been victims of discrimination, subtracted what they actually earned during the period, and awarded the difference in back pay. We prescribed this formula in UTU, 532 F.2d at 341, because it makes the discrimination victim whole while denying her a double recovery. If the back pay period ended when a discrimination victim obtained substantially equivalent temporary employment, as Ford suggests, the discrimination victim would not be made whole, because her back pay period would be shorter than her hypothetical work history. Ford’s proposed rule is thus contrary to the intent of Title VII, since it does not fully compensate the discrimination victim.
Moreover, the procedure followed by the district court is completely in accord with the statute. In 42 U.S.C. § 2000e-5(g), Congress provided: “Interim earnings or amounts earnable with reasonable diligence by the person or persons discriminated against shall operate to reduce the back pay otherwise allowable.” Thus, the statute requires only that interim earnings be deducted from the back pay award, the very procedure implemented by the district court; but it does not require, as Ford urges, that the back pay period end when a discrimination victim accepts other temporary employment. The back pay period may continue past the acceptance of another job by a discrimination victim, as long as the victim’s earnings are subtracted from the back pay award.
In this case, the district court properly found that the hypothetical employment histories of the women would extend beyond their employment at General Motors. The two men who were hired instead of Gaddis and Starr worked until the time of trial. Unlike the women, who were unable to continue at General Motors, the two male Ford workers were steadily employed from 1971 until 1977. Thus, if Ford had hired Gaddis and Starr in 1971, in all probability they would have worked until 1977.6
Ford also argues that the result reached by the district court is contrary to the method followed by the National Labor Relations Board under the National Labor Relations Act, 29 U.S.C. § 160. Ford contends that the NLRB has implemented a rule of ending back pay periods when a discrimination victim obtains substantially equivalent employment, even if the discrimination victim is subsequently laid off.
In fact, the approach under the NLRA is identical to the approach adopted by the district court. In NLRB v. Mastro Plastics Corp., 354 F.2d 170 (2 Cir. 1965), cert. denied, 384 U.S. 972, 86 S.Ct. 1862, 16 L.Ed.2d 682 (1966),7 the guilty employer made the *191same argument advanced by Ford. Both the Board and the court rejected the employer’s position. The court explained:
Several discriminatees secured substantially equivalent employment or became self-employed during the back pay period and yet suffered additional losses thereafter. Mastro argues that the taking of such employment terminates completely its back pay obligation. We do not agree. The only issue here is whether the dis-criminatee willfully incurred a loss of earnings. It would be unjust to require him to mitigate his damages to the greatest extent possible but then to penalize him for substantial but short-lived success. Unless in taking substantially equivalent or self-employment the dis-criminatee willfully forwent greater earnings, his back pay should not be reduced beyond the interim earnings he in fact received.
Id. at 179.
This passage highlights the central weakness of Ford’s argument. If the back pay period ended completely when a discrimination victim accepted substantially equivalent employment, then a discrimination victim would be discouraged from mitigating damages. She could accept equivalent employment only by forfeiting continued back pay rights, with the risk that her new job might end and leave her without income from any source. Many discrimination victims might choose unemployment and continuing back pay over employment which might be short-lived. Title VII, like the NLRA, should be interpreted so as to encourage mitigation of damages by discrimination victims in order to reduce the back pay liability of Title VII defendants would be ill-served by a rule which encouraged discrimination victims to remain idle.8
Therefore, by refusing to terminate the back pay period when Gaddis and Starr returned to General Motors, the district court furthered two policies underlying Title VII. It made Gaddis and Starr whole, by awarding them back pay during their entire hypothetical work histories at Ford. At the same time, the district court followed a rule which induces Title VII discrimination victims to mitigate damages. Without reservation, we affirm the district court’s decision.
(b) Job Offers
Later in 1973, while Gaddis and Starr were working at General Motors, Ford offered a permanent warehouse job to Gaddis. She declined the offer. Then Ford offered the job. to Starr, who likewise declined. Ford asserts that its back pay liability cannot extend past the time when Gad-dis and Starr turned down the offers.
The district court found that one principal reason why Gaddis and Starr rejected Ford’s offer was because they would have been forced to abandon their seniority at General Motors. By accepting a job with Ford and leaving General Motors, Gaddis and Starr would have forfeited all their rights at General Motors. Ford’s offer did not include retroactive seniority to the time *192of the 1971 hirings. Ford presented Gaddis and Starr with an offer of mere beginning employment.
We have twice held under Title VII that victims of discrimination need not accept job offers which include a loss of seniority in order to preserve their back pay rights. In UTU, 532 F.2d at 340, we explained:
A refusal to commit seniority suicide is not an acceptable reason to deny back pay. Victims of discrimination should not be required to forfeit wage and security benefits accruing because of their seniority in order to remain eligible for purely speculative future back pay relief. A general rule requiring such action would frustrate the central purposes of Title VII.
Accord, Hairston v. McLean Trucking Co., 520 F.2d 226, 232 (4 Cir. 1975) (plaintiff’s refusal of promotion or transfer because of a discriminatory loss of seniority is not a defense to plaintiff’s back pay claim).
Ford would present Gaddis and Starr with a similarly intolerable choice. Under Ford’s legal theory, if Gaddis and Starr rejected Ford’s offer and stayed at General Motors, they would forego their rights to further back pay benefits. On the other hand, if they accepted the job offered by Ford, which they had not held for the previous two years because of Ford’s discriminatory hiring policy, they would lose their seniority rights at General Motors. Thus, Ford’s job offer was tainted by the effects of the discrimination it had practiced in 1971. Gaddis and Starr could accept the offer only by forfeiting the seniority they had accumulated at General Motors and without a compensating offer of seniority at Ford to alleviate the effects of the discrimination against them in 1971. As in Hairston and UTU, the result advocated by Ford would “frustrate the central purposes of Title VII.”
Our decision is in accord with the decisions of three other circuits. In Comacho v. Colorado Electronic Technical College, Inc., 590 F.2d 887 (10 Cir. 1979) (per curiam), a woman sued a college under Title VII for discriminatorily firing her. She prevailed and was awarded back pay. The college argued that her right to back pay was terminated when the college made her an offer of reinstatement. The Tenth Circuit replied:
42 U.S.C. § 2000e-5(g) ... gives the trial court discretion as to the remedies. The obvious purpose of the Act is to compensate persons for injuries suffered on account of unlawful discrimination. Congress clearly intended that the remedies employed would “make whole” as nearly as possible any person injured under the Act.... An offer of reinstatement without back pay does not “make whole” the person injured by an illegal firing. Such a holding would circumvent the objectives of Title VII. The college thus did not offer to fully compensate plaintiff, but merely offered to reemploy her.... Thus we must hold that a reinstatement offer without back pay does not relieve a guilty employer from further liability.
Id. at 889.
In Jurinko v. Edwin L. Wiegand Co., 477 F.2d 1038 (3 Cir.), vacated on other grounds, 414 U.S. 970, 94 S.Ct. 293, 38 L.Ed.2d 214 (1973), the district court held that discrimination against the plaintiffs ceased when they were offered jobs by the defendant. The Third Circuit rejected this portion of the district court’s opinion, stating:
The terms of the 1969 job offers were within Wiegand’s control, and it did not offer the plaintiffs back seniority or back pay. The offer that was made did not rectify the effects of its past discrimination, and the plaintiffs were under no duty to accept such an offer.
Id. at 1047.
Similarly, in Claiborne v. Illinois Central Railroad, 583 F.2d 143, 153 (5 Cir. 1978), cert. denied, 442 U.S. 934, 99 S.Ct. 2869, 61 L.Ed.2d 303 (1979), the Fifth Circuit held that: “Failure to accept an offer of reinstatement in a Title VII case does not necessarily terminate the right to relief, so long as those amounts earned elsewhere or eamable with reasonable diligence are deducted from each plaintiff’s award.” Fur*193thermore, the Fifth Circuit explained that “the disadvantages to an employee returning to the railroad without retroactive seniority” would “more than justify” an employee’s decision to refuse an offer of reinstatement. Thus, the Fifth Circuit, like the Third and the Tenth, would find, as we do, that Ford’s job offers to Gaddis and Starr did not end their back pay periods.9
Further, it should be noted that Ford was benefited by the employment of Gaddis and Starr at General Motors. Their earnings at General Motors were deducted in the calculation of their back pay awards. Gaddis and Starr reduced Ford’s back pay liability by working at General Motors.
In sum, the district court reached an eminently reasonable result. It did not permit Ford to cut off the back pay period by making Gaddis and Starr an incomplete and unacceptable offer, and it denied Gaddis and Starr a double recovery by deducting their General Motors wages from their back pay awards. We affirm this aspect of the district court’s decision.10
(c) Nurses Training
In September, 1975, Gaddis and Starr entered a nurses training program, a CETA program for the unemployed. Ford asked the district court to terminate the back pay period of Gaddis and Starr upon their entry into the program. The district court refused. Instead, the court accepted the special master’s report which treated the training program as employment and reduced the back pay awards by forty hours of minimum wage earnings for each week the women spent in the program.
On appeal, Ford again presses the argument that its back pay liability should not extend beyond the date when Gaddis and Starr entered the nurses training program. We think that there are four reasons why Ford should not prevail.
*194First of all, Gaddis and Starr did not remove themselves from the labor market when they entered the CETA program. Both testified that they would have accepted work had it been offered to them. The special master’s report credited their testimony. He found that the two women had not removed themselves from the labor market, and his finding was adopted by the district court. Because this determination was a reasonable credibility determination, we cannot treat it as clearly erroneous.
Two courts have decided similar cases in which Title VII claimants entered school after suffering discrimination at the hands of an employer. Those courts have decided that the critical determination is whether the claimant remained “ready, willing, and available” for employment. Taylor v. Safeway Stores, Inc., 524 F.2d 263, 267 (10 Cir. 1975); United States v. Wood, Wire & Metal Lathers Local 46, 328 F.Supp. 429, 439, 443 — 14 (S.D.N.Y.1971). If the claimant is ready, willing, and available for work, these courts have suggested, the back pay period does not come to a close when the claimant enters school.
The district court in this case followed the rationale of these cases by instructing the special master to ascertain whether Gaddis and Starr had removed themselves from the labor market. We conclude that the district court acted correctly in ordering the special master to examine the availability of these two particular women for work, rather than mechanically ending the back pay period upon their entry into the CETA training program. The “ready, willing, and able” test followed by the district court permits the relief in each case to be fitted to fit the circumstances of particular claimants.
Second, the CETA training program closely resembled employment. Both women received the minimum wage for the time they spent in the training program. Both testified that they regarded the program as employment. Indeed, the apparent intent of the CETA program is to provide the unemployed with work while at the same time preparing them for future employment.
The special master treated the training program as employment like the claimants’ temporary employment at General Motors. As with the wages the women earned at General Motors, the special master subtracted the wages the women earned in the training program from Ford’s back pay liability. 42 U.S.C. § 2000e-5(g). In addition, the district court took every precaution to insure that Ford was not unfairly burdened by back pay liability. Even though Gaddis and Starr sometimes attended the program for less than forty hours per week, the district court subtracted forty-hours of earnings for each week they spent in the program. In this fashion, the district court properly treated the nurses training program as the equivalent of employment and also guaranteed that Gaddis and Starr did not receive more than one full recovery.
Third, Gaddis and Starr were enrolled in the program as a result of their presence on the unemployment compensation rolls. Gaddis and Starr were receiving unemployment compensation when they began the training program. Both testified that they did not sign up for the CETA program on their own and that they enrolled at the urging of the state unemployment compensation office. In fact, a representative of the state agency testified that their unemployment benefits would have been can-celled if they had not enrolled in the CETA program. As the special master found, “[T]his course of action was recommended very strongly by the North Carolina Employment Security Commission.”
Gaddis and Starr were unemployed as a direct result of the discrimination against them by Ford. As the employment histories of the men hired in 1971 demonstrate, if Ford had not discriminated against Gad-dis and Starr in 1971, they probably would not have been receiving unemployment compensation. But for Ford’s discriminatory hiring practices, they would have had no need of a CETA program. However, because they were unable to find work and because they were a burden on the state unemployment compensation fund, Gaddis *195and Starr were convinced by the state to enroll in the CETA training program. Ford cannot equitably be relieved of Title VII back pay liability because of an event which flowed directly from its discriminatory actions. The district court reached the proper result when it refused to terminate Ford’s back pay liability upon the entry of Gaddis and Starr into the CETA training program.
Fourth, as has been detailed above, the district court did not abuse its discretion, see Franks v. Bowman Transportation Co., 424 U.S. 747, 96 S.Ct. 1251, 47 L.Ed.2d 444 (1976), when it chose not to end these claimants’ back pay period with their enrollment in the nurses training program. As the Tenth Circuit indicated in Taylor, 524 F.2d at 268, district courts have discretion to examine the particulars of a claimant’s entry into school and to decide whether the employer’s back pay liability should end or continue. In this case, the enrollment of Gaddis and Starr into nurses training is colored by several facts, including their availability for work, the similarity between the CETA program and employment, and the close connection between the state unemployment compensation agency and the CETA training program. When all these factors are considered, it is apparent that the district court did not abuse its discretion when it permitted Ford’s back pay liability to continue during the training program, while giving Ford credit for the claimants’ possible minimum wage earnings during the period.
(d) Unemployment Compensation
Ford argues that the awards of all three women should be reduced by the amount of unemployment compensation they received during the back pay period. The district court did not deduct unemployment compensation from the back pay awards. We conclude that the district court ruled correctly.
There is a division of authority over the deduction of unemployment compensation from Title VII back pay awards. Some district courts have subtracted unemployment compensation in computing back pay. See, e. g., Heelan v. Johns-Manville Corp., 451 F.Supp. 1382 (D.Colo.1978).
On the other hand, one district court in this circuit has decided that monies received as “state unemployment compensation ‘are made to carry out an independent social policy’ and are not deductible from a back pay award under Title VIL” Abron v. Black & Decker Manufacturing Co., 439 F.Supp. 1095, 1115 (D.Md.1977), quoting Inda v. United Air Line, Inc., 405 F.Supp. 426 (N.D.Calif.1975) aff’d in part and vacated in part, 565 F.2d 554 (9 Cir. 1977), cert. denied, 435 U.S. 1007, 98 S.Ct. 1877, 56 L.Ed.2d 388 (1978). See also Tidwell v. American Oil Co., 332 F.Supp. 424 (D.Utah 1971).
In an analogous context, the Supreme Court has decided that unemployment compensation need not be deducted from back pay awarded under the NLRA. NLRB v. Gullett Gin Co., 340 U.S. 361, 71 S.Ct. 337, 95 L.Ed. 337 (1951). The Supreme Court’s analysis is particularly convincing:
... To decline to deduct state unemployment compensation benefits in computing back pay is not to make the employees more than whole, as contended by respondent. Since no consideration has been given to collateral losses in framing an order to reimburse employees for their lost earnings, manifestly no consideration need be given to collateral benefits which employees may have received.
.. . Payments of unemployment compensation were not made to the employees by respondent but by the state out of state funds derived from taxation. True, these taxes were paid by employers, and thus to some extent respondent helped to create the fund. However, the payments to the employees were not made to discharge any liability or obligation of respondent, but to carry out a policy of social betterment for the benefit of the entire state, [citations omitted]. We think these facts plainly show the benefits to be collateral. It is thus apparent from what we have already said that failure to take them into account in ordering back pay does not make the employees more than “whole” as that phrase has been understood and applied.
*196Id. at 364, 71 S.Ct. at 339 (emphasis in original).11
Based on the Abron decision, the analysis of the Supreme Court, and the remedial policies of Title VII, we agree with the district court that “awards of back pay under Title VII should not be affected by a system of compensation which is designed to serve a wholly independent social policy.” To decide otherwise would undercut to some degree the corrective force of a Title VII back pay award.
(3) Conclusion
Many of Ford’s objections contest the factual findings and credibility determinations of the district court and the special master. Because we find none of them clearly erroneous, we do not overturn any of these resolutions of conflicting facts and testimony. Most of the rest of Ford’s contentions challenge the district court’s exercise of discretion in framing a remedy. As indicated above, we find that the district court did not abuse its discretion but, on the contrary, closely examined each issue to determine an equitable adjustment of Ford’s back pay liability. Consequently, we affirm the district court’s decision with respect to Gaddis, Starr, and Smith.
B. 1973 Opening
(1) Discrimination
In 1973, Ford hired Roger Simpson to fill a single vacancy at the warehouse. Ford passed over eleven women who had applied for the job. In its order of reference to the special master, the district court found that EEOC had established a prima facie case of discrimination against the eleven. Upon referral, while the special master concluded that Ford did not act with discriminatory intent when it hired Simpson, he did conclude that Ford had acted pursuant to a pattern or practice of discrimination. The district court approved the special master’s report. Ford raises several challenges to the court’s conclusion that it discriminated against the eleven female applicants.
(a) Disparate Impact or Disparate Treatment
First, Ford argues that this case from the beginning has been a disparate treatment case founded on an allegation that Ford intentionally discriminated against the eleven women. Because of this assumption, Ford insists that the special master in effect ruled that EEOC had failed to set forth a prima facie case, by finding that Ford acted with no discriminatory intent when it passed over the eleven women.
It is true that the Title VII case against Ford with respect to the 1971 openings was founded on a disparate treatment theory. The district court found that as of 1971 Ford followed vague and subjective hiring criteria, and that Ford’s personnel stated that Ford did not hire women to work in the warehouse. And, under McDonnell Douglas, Ford failed to articulate a legitimate nondiscriminatory reason for its decision to hire three men to fill the 1971 vacancies. The district court thus uncovered discriminatory intent behind the 1971 hir-ings.
The district court’s Title VII ruling with respect to the 1973 opening, however, was grounded on a finding of disparate impact, not disparate treatment. When it referred the case to the special master, the district court found that “the plaintiff [had] established a prima facie case of discrimination” as to the eleven women. The special master interpreted this language as a finding of disparate impact rather than disparate treatment, and accordingly ruled against Ford because of its finding that Ford’s employment policies resulted in a pattern or practice of discrimination against the eleven women.
In its order approving the special master’s report, the district court responded to the argument Ford makes here and confirmed that the special master correctly interpreted the referral order. The district court explained that its order contained “a finding by the court of a pattern and prac*197tice of discrimination against women. This was the correct interpretation of the court’s conclusion of law, although the word ‘pattern’ was not used by the court.”
Of course, as the district court perceived, Title VII condemns both intentionally discriminatory employment policies and employment practices which only have a disparate impact on women. Albemarle Paper Co.; Griggs v. Duke Power Co., 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971). As well, in any given case, disparate treatment and disparate impact may serve as alternative theories upon which a right to relief may be established. Wright v. National Archives & Records Service, 609 F.2d 702, 710-11 (4 Cir. 1979) (in banc).
Consequently, Ford cannot challenge the plaintiff’s prima facie case by pointing to the special master’s finding of no discriminatory intent. The district court could determine that Ford violated Title VII based upon a disparate impact theory, which does not require a finding of discriminatory intent. Nor can Ford object to the district court’s reliance on a disparate impact theory, even if a disparate treatment theory was also considered as a basis for relief in connection with the 1971 openings, the 1973 opening, or both. As we explained in Wright, a district court is free to rely on either theory.
(b) Pattern or Practice
Second, Ford contends that even if this case rests on a finding of disparate impact, EEOC did not establish a prima facie ease because it failed to prove a pattern of discrimination. We find, however, that EEOC showed both that Ford’s hiring history revealed a pattern of discrimination against women and, whatever the probative force of the statistics, that the employment practice followed by Ford in filling the 1973 opening had a disparate impact upon women.
As of 1973, Ford had never hired a woman to work in its warehouse as a permanent employee. Between 1971 and 1973, Ford had two openings for permanent employees. The first opening was filled by David Martin, whom the district court found was hired because of his qualifications. The second position was offered to Gaddis and Starr, and Ford urges that these offers demonstrate that Ford was not using a discriminatory employment policy during the period. The district court discovered, however, that the offers were made to Gaddis and Starr only after Ford learned that Gaddis had filed a sex discrimination charge against it with the EEOC. Consequently, these offers are entitled to little weight. Subsequently, Simpson, a man, was hired to fill the position.
EEOC’s case is less than .totally convincing, because Ford made only one contested employment decision between 1971 and 1973. Nevertheless, this history amounts to a total exclusion of women from permanent employment at the warehouse, except when Ford felt the threat of an EEOC complaint. And, Ford’s record of pre-1972 employment discrimination gives these facts more force. Consequently, while there were few hiring decisions during the period, we think that the district court reasonably inferred from the record a pattern of discrimination against women. •
But even if the bare record of Ford’s hiring decisions is insufficient, the method Ford used to hire Simpson was clearly an employment practice which had a disparate impact on women. Ford considered applications inactive after six months unless the applicant contacted the warehouse to update his application. Ford did not inform applicants of this policy. Simpson had filed his application in 1971, approximately two years prior to the date Ford hired him in 1973. Simpson had done nothing to update his application. Instead, as the special master found, an employee at the warehouse, who was Simpson’s friend, “birddogged” his application. Simpson was hired, the special master concluded, at least in part to accommodate his friend.
Of the seven women who appeared before the special master, three had applied over six months before the opening was filled and therefore would not have been considered. Under Ford’s policy, they were unaware that they had been eliminated by *198the passage of time and did not benefit from the special treatment Simpson received. The remaining four had applied within six months and therefore should have been considered for the job. Yet, they had no advantage over Simpson even though their applications were current and his had long been inactive.
It is apparent that Ford’s hiring policy had a discriminatory impact upon women applicants. First of all, their applications were subject to the six-months rule, without their knowledge. Friends of warehouse workers, such as Simpson, had the benefit of assistance from inside the warehouse to keep their applications current. The influence of warehouse employees was critical in the selection of new workers, at least in this instance. Because male workers are likely to recommend their male friends as prospective employees, male applicants were more likely to be hired at the warehouse than female applicants. As the special master found, this informal secret selection process led to “an environment with loose employment procedures” which “operated to perpetuate the status quo — i. e., as all-male warehouse.” Ford’s hiring habits, then, served to exclude women from the warehouse and foreclosed eleven women from serious consideration for the 1973 vacancy.12
The decision of the district court virtually is compelled by our holding in Barnett v. W. T. Grant Co., 518 F.2d 543, 549 (4 Cir. 1975). In that case, the Title VII plaintiff attempted to prove discrimination through statistics, which we found “quite suggestive,” if not “overwhelming.” As in this case, the plaintiff’s proof went beyond statistics:
In addition to statistics, a court must also examine “patterns, practices, and general policies to ascertain whether racial discrimination exists.” [citations omitted]. Word-of-mouth hiring, which the district court found to be Grant’s primary method of recruiting new [employees], is discriminatory because of its tendency to perpetuate the all-white composition of a work force.
In harmony with Barnett, we find that Ford’s informal hiring procedure, which relied heavily on the recommendation of a current employee, was discriminatory because Ford’s workforce consisted only of men.
(c) Relative Qualifications
Ford makes one final argument to overturn the decision of the district court. Ford contends in response to this prima facie case that it hired Simpson because he was better qualified than the women applicants. To begin with, the special master found that the women applicants, like Simpson, were qualified to work as picker-packers.13 All had previous clerical or industrial work experience, ample training for “picking” and “packing;” and thus the special master’s finding was not clearly erroneous.
More to the point, as the special master found, the applications of the women did not receive a “fair comparison” with Simpson’s because of the recommendation of the male warehouse employee. Simpson was *199the only applicant who received an interview; even the women who had current applications were not interviewed. The influence of the warehouse employee was thus an important motivating reason for Simpson’s selection. Ford cannot rely on relative qualifications when the recommendation of a warehouse employee played a key role in Simpson’s selection and when the female applicant did not receive serious consideration.
(2) Back Pay
The district court awarded the 1973 women applicants back pay. Seven appeared before the special master to prove their damages. Under the instructions of the district court, the special master calculated each woman’s back pay and awarded one-seventh of each sum, because Ford only filled one vacancy in 1973. Ford does not contest this aspect of the district court’s decision.
Ford does object to the length of the back pay period. After working several months, Simpson was laid off on February 6, 1974. Simpson’s recall rights expired by the time Ford was ready to fill his vacant position. At that time, in September, 1975, Ford did not recall Simpson, but in his place hired Doris Baumgardner, who was Ford’s first female permanent picker-packer.
The special master found that in the usual situation Ford recalled laid-off employees, even when their recall rights had expired. Because the hiring of Baumgardner was an unusual case, the special master did not terminate the back pay period with Simpson’s layoff. Instead, the special master included Baumgardner’s employment within the back pay period for the eleven women.
Ford argues that this extension of the back pay period was improper. According to Ford, “the court was required to track the actual work experience of the employee allegedly favored by Ford’s alleged discrimination.” Ford’s proposed rule is unacceptable because it bears no relation to the purpose of a Title VII back pay award.
The purpose of a back pay award is to make the discrimination victim whole, to erase the effects of illegal discrimination. To accomplish that purpose, courts construct the hypothetical employment history of the discrimination victim to determine the appropriate back pay period. UTU, 532 F.2d at 341.
Frequently, the employment history of the applicant who was actually hired will provide the best guide to the hypothetical employment history of the discrimination victim. In fact, the district court in this case used the employment history of the men hired in 1971 to determine the back pay periods for Gaddis, Starr, and Smith. But the employment history of the employee is not controlling in and of itself. Actual employment histories are used only when they are the best available guide to the hypothetical employment history of the discrimination victim. The focus is on the probable job career of the victim, who is to be made whole, not on the career of the actual employee, who is of course not a party to the Title VII case.
In this instance, the special master tracked Simpson’s work history, but only to a point. Simpson’s employment history was used as the best indicator of the probable work history of the female applicants until Doris Baumgardner was hired to fill his job. At that point, the special master found that Simpson’s employment history was abnormal. The testimony of Ford’s warehouse manager revealed, as the special master found, that normally employees are recalled even after their recall rights had expired. In this case, Simpson was not recalled, and a woman was hired in his place. Ford’s departure from its usual practice took place after EEOC filed this sex discrimination suit.
It was apparent, then, that Ford’s decision not to recall Simpson was out of the ordinary, a product of unusual circumstances. As a result, Simpson’s work history failed as an indicator of the probable work history that one of the seven women would have had. Since laid-off employees were usually recalled, the special master made the only reasonable decision when he aban*200doned Simpson’s actual work history and concluded that one of the seven women would have probably been recalled if she had been hired. If the special master had instead followed Simpson’s work history, the special master would have based the back pay period not on the hypothetical work history of the discrimination victim, but on the peculiarities of Simpson’s situation. That approach would have been contrary to UTU and to the “make whole” doctrine of Albemarle Paper Co.
The special master adhered to the dictates of UTU and Albemarle Paper Co., by constructing the hypothetical work histories of the women. In our view, the special master reasonably decided that if one of the women had been hired in 1973 she would have been recalled in 1975 even though Simpson was not.
III.
In its complaint, EEOC asked the district court to “make whole those persons adversely affected by [sex discrimination], by providing appropriate back pay, with interest, ... and other affirmative relief necessary to eradicate the effects of unlawful employment practices.” As described above, the district court meticulously formulated a back pay award to correct the effects of sex discrimination against the women applicants.
But the district court offered the women no further remedies. Under Title VII, district courts have discretion in the formulation of remedial orders. E. g., Franks v. Bowman Transportation Co. At the same time, district courts have a duty to grant the “fullest relief possible to the victims of discrimination,” to insure that the effects of discrimination are totally removed from the lives of discrimination victims. Albemarle Paper Co., 422 U.S. at 418, 95 S.Ct. at 2372, see also UTU, 532 F.2d at 341.
In this case, the district court did not discuss possible remedies other than back pay. In its findings of fact and conclusions of law, in its referral order to the special master, and in its order adopting the special master’s report, the district court did not address other remedial steps. Under Albemarle Paper Co.; then, the district court was obliged to consider other remedies and if it rejected them to articulate its reasons.
The district court may not have discussed other possible remedies because it concluded that Ford had made diligent efforts since 1975 to hire women. Although Ford’s affirmative action program might justify the district court’s decision not to order general affirmative relief against sex discrimination, Ford’s anti-discrimination efforts do not alter the district court’s responsibility to redress the discrimination suffered by these individual women. Ford’s good faith since 1975 does not remove the burden placed upon the ten women by discrimination practices in 1971 and 1973. See UTU, 532 F.2d at 340.
In fact, these women are not aided in any fashion by Ford’s program'. They still do not possess the seniority they would have accumulated had they been hired in earlier years, and they do not yet have the jobs they might well have had if Ford had not discriminated against them. Ford’s affirmative action program, then, does not excuse the district court’s failure to consider additional relief for the individual women.
One seemingly desirable remedy in this instance would be to grant the women hiring preferences at the parts warehouse, with retroactive seniority. See 42 U.S.C. § 2000e-5(g); Teamsters v. United States, 431 U.S. 324, 325, 97 S.Ct. 1843, 1849, 52 L.Ed.2d 396 (1977); Franks v. Bowman Transportation Co.; Cathey v. Johnson Motor Lines, Inc., 398 F.Supp. 1107, 1122 (W.D. N.C.1974).
This remedy may be particularly appropriate for the three women who were not hired in 1971. Gaddis and Starr have had difficulty finding jobs since 1971. After their final layoffs from General Motors, they were forced to enter a CETA training program, apparently unable to find work in the auto parts field. If Ford had not discriminated against them in 1971, they might *201well have avoided this checkered employment career. Smith had been employed during much of the time since 1971, but at relatively low wages. She also might be entitled to a hiring preference with retroactive seniority.
The seven women who were not hired in 1973 present a more difficult case. Ford filled only one opening in 1973, and thus all seven cannot reasonably claim that they would have been employed by Ford since 1973 if Ford had not discriminated against them. However, the district court may be able to fashion a remedy, such as hiring preferences without seniority, which would grant them more complete relief than a back pay award.
If the district court decides to order Ford to hire any or all of these discrimination victims as positions come open, it should also consider whether the women should receive an additional monetary award to compensate them for the estimated earnings lost between the date of the district court’s decision and the date of employment at the warehouse. See White v. Carolina Paperboard Corp., 564 F.2d 1073, 1091 (4 Cir. 1977). Such an award, along with back pay, preferential hiring, and retroactive seniority, would constitute the “fullest relief possible.”
We emphasize that the terms and conditions of relief for discrimination victims is determined by district courts, not by this court. We only instruct the district court to consider these possible remedies; not necessarily to provide them. On remand, the district court should articulate its reasons for rejecting or accepting the additional relief proposed by EEOC.
IV.
We affirm the district court’s decision as a proper exercise of discretion founded on not clearly erroneous factual determinations. We remand for consideration of further remedies, so that the district court can fully exercise its discretion to formulate the most complete relief possible for these discrimination victims.
AFFIRMED AND REMANDED.
. Ford filled four positions with men in August, 1971. One of those positions was created especially for a man, Parks, who transferred from a Ford warehouse in another city. The district court instructed the special master to base the back pay periods of the three women on the employment histories of Blanchard, Coe, and Rogers, the three men hired to fill the other three positions in which vacancies had occurred in the regular course of operations.
. The seven women were Gloria Gaither, Betty Erwin, Mary Luckey, Frances McGinnis Gantz, Mary Cox, Arlene Brown, and Frankie Benson Grier. Because Mary Luckey had died in the interim, a representative of her estate appeared before the special master.
. Ford also suggests that Blanchard and Coe were more qualified than Gaddis and Starr. Gaddis and Starr had worked as full-time parts pickers at a General Motors parts warehouse. Blanchard had worked for a chain saw manufacturer in a position similar to a picker-packer, and Coe had worked for a Ford dealership. However, both Blanchard and Coe had recently been high school students and had worked only part-time and during the summers. The district court found that Gaddis and Starr were qualified to work for Ford as picker-packers, and on this record it is certainly not clear that Blanchard and Coe were more qualified.
Moreover, the gravamen of the EEOC complaint is that Gaddis and Starr were not considered for the openings, even though they were qualified, merely because they were women. Ford can hardly defend itself on the basis of relative qualifications when it never evaluated the qualifications of the women.
. Ford claims that it was surprised by Smith’s testimony that she had applied in 1971. Smith testified as one of the women who applied in 1973, but mentioned during her testimony that she had applied in 1971. Ford requests a remand so that the district court can conduct a further hearing into Smith’s claim.
As explained above, the determination of when Smith first applied was purely a credibility determination. Smith said she applied in 1971. Ford’s personnel stated that Gaddis and Starr were the first women to apply for warehouse jobs. Thus, the basic question was whether Smith’s story was believable in the face of contrary testimony.
The district court heard Smith’s testimony and the testimony of Ford’s employees. Ford had ample opportunity to cross-examine Smith. The district court concluded that Smith was telling the truth and that under Ford’s employment practices her application was not seriously considered because of her sex. We see no need for further proceedings, since the district court has heard and evaluated the pertinent evidence.
Ford makes one last argument for a remand by contending that Smith stated that she did not specifically request a picker-packer job when she applied at the warehouse. As Ford points out, during her testimony Smith stated, “I didn’t indicate what job I applied for.” But, as *189Ford fails to mention, Smith immediately continued, “I just, unless it was something in my line of work or whatever.” Ford’s attorney then asked, “Your line of work was what?” Smith replied, “Stockroom or shipping or something like that.”
Smith’s testimony is not a model of clarity. But she indicated that she applied for a job in her line of work, which was stockroom or shipping work. The job at the warehouse which obviously fitted to her description was the picker-packer job. The district judge, who heard Smith’s testimony and observed her demeanor during cross-examination, found that she applied for a warehouse job. Based on Smith’s testimony, which Ford itself developed on cross-examination, that finding is not clearly erroneous. Here, again, there is no need for a remand.
. Ford argues that the district court placed an impermissibly heavy burden upon it. According to Ford, the district court required it to “establish” or prove that it had not discriminated against the three women. Keene State, Ford contends, requires only that it “articulate” a legitimate nondiscriminatory reason for choosing the men instead of Gaddis, Starr, and White.
In this case, Ford neither “articulated” nor “established” a legitimate nondiscriminatory reason for hiring the men, since the district court found that the facts did not coincide with Ford’s representations. Ford insisted that the other women had not applied when the men were hired. The district court determined the truth to be that the women had applied beforehand. In this situation, Ford was left with no defense to the prima facie case of discrimination.
. Ford submits that this focus on the work histories of the men who were actually hired is punitive. But the best evidence of the female applicants’ probable work histories at Ford was the work histories of the men who held the 1971 positions. By constructing the hypothetical work histories of the women through reliance on the work histories of the men, the district court acted only to compensate the discrimination victims, not to punish Ford.
. See note 11, infra.
Furthermore, the back pay formula employed by the district court in this case is the back pay formula used under the NLRA. The Board estimates the back pay the discrimination victim would have earned but for the discrimination and subtracts actual earnings to determine the back pay award. NLRB v. Guilett Gin Co., 340 U.S. 361, 71 S.Ct. 337, 95 L.Ed. 337 (1951); Phelps Dodge Corp. v. NLRB, 313 U.S. 177, 61 S.Ct. 845, 85 L.Ed. 1271 (1941); F. W. Woolworth Co., 90 NLRB 289 (1950).
We also note that the NLRB decisions cited by Ford are inapposite to this case. In Circle Bindery, Inc., 232 NLRB 1185, 1187-88 (1977), the Board terminated the back pay period when *191the discrimination victim accepted permanent employment, but because the discrimination victim would have left the temporary job from which he had been discriminatorily discharged when he accepted the permanent job. Thus, the acceptance of the permanent job marked the end of his hypothetical work history for the temporary employer. In Star Baby Co., 140 NLRB 678, 683-84 (1963), the Board ended the back pay periods when the discrimination victims found permanent employment, but did not address the problem faced in the instant case. Star Baby was complicated by the fact that the employer had gone out of business.
. Of course, if a discrimination victim accepts permanent employment with substantially equivalent wages and remains at that job, the back pay period would end when she accepted the permanent job, because thereafter she would suffer no damages. Her actual earnings from that point on would be equal to or more than what she would have earned if she had not been a victim of discrimination. See Butta v. Anne Arundel County, 473 F.Supp. 83, 89 (D.Md. 1979); Milton v. Bell Laboratories, Inc., 428 F.Supp. 502, 515 & n. 19 (D.N.J.1977). In this case, however, Gaddis and Starr were unable to remain at General Motors, because they were once again laid off after working for about a year. They thus suffered additional damages during the period following their layoff from General Motors, a period when they would probably have been employed if Ford had not discriminated against them in 1971.
. Ford points to several NLRA decisions in which the NLRB has held that an offer of reinstatement terminates an employer’s back pay liability to a discriminatorily discharged employee. E. g., D’Armigene, Inc., 148 NLRB 2 (1964), enf’d as modified, 353 F.2d 406 (2 Cir. 1965). For several reasons we do not find these decisions persuasive.
The decisions cited by Ford discuss offers of reinstatement. In the instant case, Ford did not make the women offers of reinstatement, but only offers of beginning employment. An offer of reinstatement carries with it seniority rights, whereas an offer of beginning employment includes no seniority rights. This failure of Ford to offer retroactive seniority prevents Ford from relying on its offers to Gaddis and Starr.
In fact, our decision in this case is consistent with at least one decision of the NLRB. See note 11 infra. In NLRB v. Hilton Mobile Homes, 387 F.2d 7 (8 Cir. 1967), the employer offered a discharged employee reinstatement, but without seniority rights. The Board held that this offer did not terminate the company’s back pay liability because the offer did not include seniority. The court agreed and enforced the Board’s order. The offer in the instant case is almost identical to the offer in Hilton Mobile Homes, but significantly different from the offer of reinstatement addressed in D’Armigene.
Also, Gaddis and Starr mitigated damages. In some situations, refusal of an offer of employment or reinstatement might amount to a failure to mitigate damages, and, under 42 U.S.C. § 2000e-5(g), the discrimination victim’s recovery might be reduced by the amount of wages she could have earned in the offered job, as “amounts earnable with reasonable diligence.” In the case of Gaddis and Starr, they reduced Ford’s liability by continuing to work at General Motors.
. In addition to its finding that the women did not want to lose their General Motors seniority, the district court found that neither Gaddis nor Starr wanted to be the only woman employed at the warehouse. In fact, Ford’s warehouse manager testified that he warned Gaddis and Starr that they might be subject to sexual harassment if they worked at the warehouse. He told them that “they may very well [be] subjected to passes, offensive remarks, off-col- or language” and that “there are not now nor, .. . have there ever been women working in the warehouse.”
Incredibly, Ford tries to rely upon the discriminatory atmosphere of the warehouse to argue that Gaddis and Starr unreasonably rejected Ford’s job offers. It would be ironic indeed if Ford could defeat the remedial purposes of Title VII by resting upon sexual harassment of women in its warehouse. Ford cannot escape the fact that its offer included no seniority rights to make Gaddis and Starr whole.
We note further that the warehouse manager told Gaddis and Starr that if Ford hired one of them, it would hire the other to avoid subjecting a lone woman to sexual harassment. In 1973, Ford offered a single position to each of the women. It did not offer positions to both.
. This decision is'particularly persuasive, because Congress turned to the “make whole” provisions of the NLRA when it framed the relief provisions of Title VII. Albemarle Paper Co., 422 U.S. at 419-22, 95 S.Ct. at 2372-73.
. Our description in Wright, 609 F.2d at 711, of the disparate impact theory is directly applicable to the path Ford followed to fill its 1973 vacancy:
Central to proof of a prima facie case under this theory is proof of an employment policy or practice which, though facially neutral or even benign in actual purpose, nevertheless imposes a substantially disproportionate burden upon a claimant’s protected group as compared to a favored group within the total set of persons to whom it is applied.
. Simpson had worked for about three years as a material handler at a printing company. Of the eleven women, four had worked in warehouses for at least two years, including two who had worked in auto parts warehouses; one had worked for over five years as a distribution clerk in a post office; two had worked in textile mills; and one had spent nine years as an assistant manager of a drive-in.
Further, it should be noted that the district court found: “Although Ford contends that similar past experience is an important factor, the speed with which a new employee is expected to learn and usually does learn how to perform his job demonstrates that such past experience is not an essential job qualification.”