This is an appeal from a judgment awarding relief in a Title VII discrimination suit against the Mississippi State Employment Service (MSES), a state employment referral agency. Allegedly, MSES discriminated on the basis of race and gender in referring applicants to job openings listed with it. The district court awarded the plaintiffs, a class of black and female applicants to the MSES Bolivar County office, over $2.8 million in back pay and prejudgment interest. The parties challenge various aspects of the components of the award and the calculation of the amount due. We AFFIRM in part and REVERSE in part.

MSES refers applicants to private sector job openings listed with it. In 1972, the plaintiffs sued MSES,1 primarily basing the suit on Title VII,2 alleging that the agency discriminated on the basis of race and gender in classifying, referring, and testing applicants. The district court certified a class of blacks and women who have been or may hereafter be victims of discrimination in seeking referrals from the MSES Bolivar County office. After a three week bench trial in 1979, in a 48 page opinion, the trial court dismissed the individual and class claims on the merits.3 On appeal, this Court reversed in part, finding that the action of MSES toward the plaintiff class had established a pattern and practice of race and sex-based disparate treatment with respect to job referrals, in violation of Title VII, and that named plaintiff Rebecca Gillespie had succeeded in proving her individual claim of discrimination.4 We remanded the case for determination of an appropriate remedy.5

Because many of MSES's records were no longer in existence and because of the large number of claimants,6 the parties entered into a 48 page stipulation, with close to a hundred pages of attached exhibits, one of the stipulations was that a reasonable method of calculating back pay would be to base the award on the economic value of the relevant job orders listed with MSES, rather than to proceed on the basis of identifiable acts of discrimination against individuals.7 As will be discussed in greater detail, the parties agreed on a formula to calculate gross back pay that considered the number of job orders and vacancies, the percentage of eligible class members, and the actual referrals of class members. From this, they made adjustments for interim earnings and work-force attrition.8

In 1984 the district court issued an injunction prohibiting discriminatory out-of-code referrals and gender-specific job orders. Then, in 1988, largely relying on the parties' stipulations, the district court awarded the plaintiffs $2,873,274.94 in back pay and prejudgment interest.9 Both parties appealed.

As a state agency, MSES argues that the Eleventh Amendment prohibits the imposition of prejudgment interest on a back pay award against it. It also contends that the district court erred in the following respects: 1) by not tolling the accrual of back pay when MSES ceased all discriminatory practices, 2) by not reducing the total award by sixty-eight percent, since there was just a thirty-two percent chance that the average referral by MSES would lead to a job offer, and 3) by not reducing the award by the value of twenty-five job orders that were never filled.10

The plaintiffs appeal the judgment only insofar as the district court lowered the award for a particular job order to account for a presumed benefit to the plaintiffs for a disproportionately high number of referrals of women to another, lower-paying job.

The Eleventh Amendment bars an award of retroactive monetary relief against an unconsenting state in federal court.11 The Amendment removes jurisdiction over such suits from the federal courts.12 There are two important exceptions to this principle. First, a state may explicitly waive its immunity and consent to be sued in federal court.13 Second, Congress has the power under section 5 of the Fourteenth Amendment to abrogate the state's immunity to enforce the Amendment's protections.14 Congress exercised this power in enacting the Civil Rights Act of 1964.

The Civil Rights Act of 1964 specifically prohibits discrimination on the basis of race or sex by an employment agency in making referrals.15 As a member of the national network of state employment services established by the Wagner-Peyser Act,16 MSES comes under the definition of "employment agency" as used in that Civil Rights Act.17 The Act also provides for remedies. When a court finds that an employment agency has intentionally engaged in an unlawful employment practice, it may

enjoin the respondent from engaging in such unlawful employment practice, and order such affirmative action as may be appropriate, which may include, but is not limited to, ... hiring of employees, with or without back pay (payable by the employer, employment agency ... responsible for the unlawful employment practice), or any other equitable relief as the Court deems appropriate.18

The Act clearly abrogates the states' Eleventh Amendment immunity by providing that courts may award retroactive monetary relief against state defendants which violate Title VII.

The ability of courts to award monetary relief to victims of discrimination serves useful purposes: it deters unlawful discrimination and it compensates those who have suffered because of it.19 With respect to its compensatory purpose, the Act specifies that back pay is an appropriate remedy once a violation is established, but it clearly states that the affirmative relief a court may order is not limited to back pay. Congress vested broad equitable powers in Title VII courts "to make possible the 'fashion[ing] [of] the most complete relief possible' ".20 The measure of relief is that amount necessary to make persons whole or put them in their "rightful place".21 In tailoring appropriate relief to successful Title VII plaintiffs, courts have awarded straight back pay, the value of lost fringe benefits, and prejudgment interest.22

Despite the broad language of the Act, MSES argues that as an arm of the state, it is immune from an award of interest because the Act does not expressly mention interest as a component of a monetary award. We addressed this issue in Whiting v. Jackson State University,23 and we found, under Title VII, no Eleventh Amendment bar to the assessment of prejudgment interest on a back pay award against an "arm of the state". We reaffirm that decision today.

Interest is compensation for the use of funds; it is not awarded as a penalty against a defendant.24 We have stated in this Circuit that under Title VII interest is an item that "should be included in back pay" to make a victim whole.25 Recently in Sellers v. Delgado Community College26, we found that a magistrate abused his discretion in denying prejudgment interest on a back pay award. The case before us shows that the make-whole purpose of Title VII would be frustrated by the failure to award interest. The defendant's wrongful conduct occurred in 1969 and 1970, approximately twenty years ago. It is obvious that an award of back pay geared to a 1969 or 1970 salary will be insufficient to put the plaintiffs in their rightful place, and it is also true that the defendant has had the use of this money during the intervening years.

MSES argues that Library of Congress v. Shaw27 controls this case. In Shaw, decided after our decision in Whiting, the Supreme Court held that pursuant to the "no-interest rule", the United States was immune from an award of interest on attorney's fees absent express congressional assent to the award of interest. Title VII, the Court concluded, did not expressly allow interest against the government. Just last term, however, the Court stated that the "added gloss of strictness"28 imposed by the no-interest rule applies only when the liability of the federal government is at issue.29 MSES is an arm of the state; it thus does not benefit from the rule applied in Shaw.30

We reaffirm our holding in Whiting that the Eleventh Amendment does not bar the award of prejudgment interest on a Title VII back pay award against a state agency because Congress exercised its section 5 power to abrogate broadly the state's immunity. Congress did not limit the tools with which courts could fashion relief to victims of discrimination perpetrated by state defendants. The award of prejudgment interest is one of those tools. We hold that the district court properly assessed interest against MSES as part of the plaintiffs' make-whole relief.31

The district court's judgment awarding back pay to the class necessarily must rest on the parties' stipulations regarding the calculation of back pay. Accordingly, we quote liberally from the stipulations to establish the framework under which the disputed calculations must be made:

27. Plaintiffs and the MSES defendants have agreed that it would be useful to calculate the economic benefit which could have resulted from each job order involved in the back pay proceedings.... This information was calculated by multiplying the number of hours worked a week on a job by its hourly pay rate, multiplying the product by the number of vacancies specified in the job order, multiplying that product by the estimated number of weeks worked in each year from the date of the job order to the end of 1984 (taking into consideration the expected duration of the job as specified on the job order). This figure for each year is the estimated earnings from the job order before the effects of employee attrition are taken into account.

30. [I]t is necessary to make a reasonable determination or estimation of the length of time the claimants would have remained employed if they had been hired. The parties have stipulated that it would be extremely difficult to reconstruct individual employers' turnover rates....

33. On average ... 4.38 percent of covered employees per calendar quarter are affected by terminations of one kind or another (excluding layoffs). The parties know of no better or more accurate information to use for purposes of making classwide determinations of the length of time class member claimants would have continued to be employed by an employer if they had been referred and hired on a job order....

34. [P]laintiffs and the MSES defendants stipulate that applying the attrition rate [4.38 percent per quarter] to the earnings which would have resulted from the referral and hire of class members, rather than to the number of positions they would still occupy at the end of each successive quarter, will work out fairly to all concerned ... and will be a workable manner of taking turnover fully into account in making calculations of back pay.

37. The parties agree that the amount of the economic benefit available from a job order is not the same as estimated gross back pay, because blacks and women would not have received all of the referrals and hires in the absence of discrimination.... In order to calculate estimated gross back pay, it is necessary to multiply the economic benefit available from a job order in each year by the percentage representing a nondiscriminatory rate of referral for the job order in question....

38. [T]he MSES defendants should be given credit for the percentage of referrals who were black, if any, as to each job order involved in the racial discrimination count. (No women were referred on any job order subject to the sexual discrimination court)....

40. [T]he gross back pay figures ... do not ... take into account any interim earnings, or any defenses which the MSES defendants may have as to any of the job orders....32

MSES argues that back pay should not accrue beyond the date on which its discrimination ceased. It contends that since it did not discriminate in referrals after July 1973, "that is the date the monetary award of back pay should cease as to the class". The district court used the agreed-upon formula to calculate the award through 1984. MSES relies on Ford Motor Co. v. EEOC33 for support. Ford held that an employer's unqualified offer of employment to one previously discriminated against tolls the accrual of back pay.

We review challenges to the district court's calculation of the back pay award,34 including its determination of the length of the back pay period,35 under the "clearly erroneous" standard. We uphold the court's decision. A major purpose of a monetary award under Title VII is to make victims of discrimination whole.36 This goal is not necessarily achieved when the defendant stops inflicting new injuries on other applicants. In Ford, the employer offered the plaintiff a position that had previously been denied to him because of discrimination. At that point, the plaintiff ceased to suffer further injuries from the employer's wrongful denial of the job. In this case, that MSES began a nondiscriminatory referral policy shielded it from new liability, but did not repair the economic damage done to the earlier victims of discrimination by the wrongful denial of employment opportunities. As the district court observed, "MSES has not offered the individual plaintiffs or class members specific referrals to positions comparable to those for which defendants had earlier denied referrals."37 Ford does not apply.

MSES's argument also contradicts the stipulations, which bind the parties absent manifest injustice.38 The parties agreed that gross back pay would be calculated on the basis of the number of weeks that a claimant would have worked on a job from the date of the job order until the end of 1984.39 Then the economic value calculated from that time period would be reduced each quarter to account for attrition. Although MSES reserved any defenses it may have "as to any of the job orders", its position on appeal ignores its stipulation that 1984, not the date on which it stopped discriminating, marks the end of the back pay period. We uphold the court's ruling on this point.

2. Sixty-eight Percent Reduction of the Award

Unlike the typical Title VII defendant, MSES is a referral agency, not an employer; it has denied the potential for jobs, not jobs, to the plaintiffs.40 It argues that because there was only a thirty-two percent chance on average that a job would result from a single referral, the back pay award should be reduced by sixty-eight percent. We reject this argument, largely on the basis of the parties' stipulations.

As previously mentioned, the parties did not base the back pay award on a known number of referrals denied because of discrimination. Rather, they agreed that class members were denied a percentage of the referrals on the job orders, which corresponds to a percentage of the economic value of the job orders. The parties also agreed on the expectation that if MSES referred enough qualified applicants for a job opening, all of the vacancies would be filled. Thus, that there is roughly a one-third chance of a job resulting from any single referral does not operate to reduce the back pay award, because the award is not based on a known number of lost referrals. Moreover, there is no reference to MSES's referral/hire argument in the stipulations that formed the basis for calculating back pay. Nor does this issue fall under its reservation of defenses as to the inclusion of particular job orders. We reject this challenge to the award.

3. The Economic Value of Twenty-five Job Orders Never Filled

MSES argues that a reduction should be made for twenty-five job orders on which it discriminated in referrals because the jobs were never filled. The parties agreed that "[t]here were openings which were not filled by the local office[,] although the local office might have been able to fill them by referring enough applicants".41 In the light of MSES's failure to refer such a high percentage of admittedly qualified applicants--the class members--for many job orders, and considering the noted stipulation, the district court's inclusion of the value of these job orders is not clearly erroneous.

4. Limitation of the Award for the Baxter-Travenol Job Order

Baxter-Travenol Laboratories is a private employer that sought referrals for two job categories from MSES: the higher-paying position of material handler and the lower-paying position of assembler. During the period December 1969 to May 1970, MSES discriminated on the basis of gender on this job order by referring only men to the material handler's job. It also referred a disproportionately high number of women to the assembler's job.42 The trial judge stated that the very individuals harmed by discrimination on the material handler's job order benefitted from discrimination in their favor on the assembler's job order. "[I]n the exercise of its equitable discretion", the district court concluded that a reduction in this component of the back pay award was necessary to prevent a windfall to the plaintiffs. The court limited the award to $190,000 plus interest from May 1970.43 We reverse the limitation on this job order because it is based on a clearly erroneous factual finding.

Although Title VII provides that a court "may" award back pay to a victim of discrimination,44 the Supreme Court has stated that when a Title VII violation is found, back pay is a "reasonably certain prospect".45 We recently observed that once a plaintiff establishes a violation of the statute, "the instances wherein such an award is not granted are exceedingly rare".46

The measure of a back pay award is ordinarily the amount needed to make the victim whole. As the Supreme Court explained in Albemarle Paper:

The general rule is, that when a wrong has been done, and the law gives a remedy, the compensation shall be equal to the injury. The latter is the standard by which the former is to be measured.47

This suit is against an employment referral agency, not an employer. The parties' stipulated on a reasonable method to calculate back pay that does not give each claimant full back pay, but rather divides the economic value of jobs denied to the class among the would-be referred applicants.

A Title VII plaintiff is not entitled to recover an amount greater than his pecuniary losses. Thus, the statute prescribes that interim earnings "shall operate to reduce the back pay otherwise allowable".48 In this case, the district court deducted interim earnings with respect to the Baxter-Travenol job order in addition to its further reduction.49

The district court based the equitable reduction on its finding that the same individuals were harmed and helped by different types of discrimination on the two job openings:As the two types of positions had identical educational requirements and were filled contemporaneously, the women referred to material handlers positions would necessarily have been the same women who were referred to assemblers positions.50

We reverse the limitation of the award because the finding that the same women were both harmed and helped by different acts of discrimination is clearly erroneous. The women who would have been referred to be material handlers are the claimants on this job order. Of the twenty claimants on the job order, only two were referred and hired as assemblers during the time period for which discrimination was found.51 The district court deducted the interim earnings of these two positions, as well as the interim earnings from all other claimants, from the award as required by the statute.

We remand the case to the district court for determination of the appropriate award on this job order under the framework of the parties' stipulations.

We affirm the district court's award of back pay and prejudgment interest against MSES except insofar as the court reduced the back pay award for the Baxter-Travenol job order. We reverse the reduction of the award because we conclude that it is based on an erroneous factual finding. We remand the case to the district court for recalculation of the monetary award for this job order.