I concur in part and dissent in part.
I concur in Parts I and II of the majority opinion. However, I respectfully dissent from the majority’s conclusion in Part III that the district court should be reversed.
I disagree with the majority’s conclusion that the evidence offered by plaintiffs was insufficient to rebut the City’s claim that it assigned police officers to desk duty solely on the basis of business necessity. The district court did not base its conclusion that plaintiffs’ evidence was sufficient to rebut this claim on the statistics that the majority criticizes and finds irrelevant, but on “a thorough review of the record as a whole.” Joint Appendix at 153.
I agree with the district court that the record provides evidence that plaintiffs, the only two women on their “turn” or shift at the Youngstown Police Dept., were disproportionately assigned to desk duty, and the City offered no legitimate nondiscriminatory reason for this disproportionate assignment. The testimony of Eugene Sabatino, Captain of the Youngstown Police Department, indicated that plaintiffs were treated differently than similarly situated qualified male officers on the basis of sex. Sabatino stated that he was responsible for making assignments to desk duty for his “turn” or shift. He stated that the qualifications for desk duty were ability to type, astuteness, and congeniality. He stated that the officers who had these qualifications were six men (Officers Hol-zsehuh, Casanta, Marciano, LoSasso, Malone, and Centric) and two women — the plaintiffs herein.
Sabatino stated that desk assignment for police officers was not a favorable assignment, but some officers volunteered to work the desk in exchange for receiving better days off (Saturday and Sunday). Two male officers on his turn, Centric and Malone, were two men who chose to volunteer for desk duty in exchange for better days off.1 He testified that plaintiffs, in contrast, were not offered better days off when they were given desk assignments because they did not volunteer for desk duty. He also stated that a lot of male officers claimed they could not type, but he believed it might have been a *73ploy to avoid desk duty, and he believed male officers had been prepped not to be able to type so they would not have to do desk duty. However, Sabatino stated that if he found out a male officer could type, he would assign him to the desk.
Sabatino then reviewed the work schedules for the year 1985 and stated that no male officer, who was qualified for desk duty, was assigned as frequently to desk duty as plaintiff Simon2 except for male officers who volunteered or were assigned to desk duty for disciplinary reasons. After reviewing the schedules of plaintiff Simon’s assignments to desk duty in 1986, Sabatino stated that no male officer, who qualified for desk duty, was assigned to this position as frequently as she was except for Officer Centric, who had volunteered for desk duty and was receiving better days off in exchange for his assignment to desk duty. Officer Simon did not receive better days off. Sabatino also conceded that often a person with less seniority was given most of the desk assignments, but in 1986, in spite of the fact that Officer Simon had the most seniority, she was assigned to the desk more often than any qualified male officer except for Officer Centric, who volunteered.
I believe Sabatino’s testimony indicates that women who could type were treated differently than similarly situated qualified male officers, and that plaintiffs were assigned to desk duty disproportionately, which established a prima facie case of sex discrimination. Kent County Sheriff’s Ass’n v. County of Kent, 826 F.2d 1485, 1492 (6th Cir.1987) (citing McDonnell Douglas Corp. v. Green, 411 U.S. 792, 801, 93 S.Ct. 1817, 1823-24, 36 L.Ed.2d 668 (1973)); Pierce v. Commonwealth Life Ins. Co., 40 F.3d 796, 800-803 (6th Cir.1994). See also E.E.O.C. v. Romeo Community Schools, 976 F.2d 985 (6th Cir.1992). The City offered no nondiserimi-natory reason for assigning women to desk duty more frequently than similarly situated qualified male officers because there was no business necessity to assign qualified women to desk duty instead of qualified males, who could type. Therefore, I believe the district court should be affirmed on this issue.
However, I believe the district court erred in awarding damages for economic loss that were too speculative. The district court awarded damages for economic loss due to plaintiffs’ loss of AT time because of disproportionate assignment to desk duty. In this regard, he stated the following:
Damages or other remedies are difficult to quantify for this injury; the primary tangible injury noted by the plaintiffs was their relative inability to gain accumulated time (“AT”) while working at the desk. Testimony was presented that the primary method for gaining AT was making an arrest while working on the street on one of the evening or night turns, and then going to court on that arrest in the daytime. Because plaintiffs lost the opportunity to gain AT rather than losing a specific amount of AT, it is difficult to pinpoint the amount of the plaintiffs’ loss.
In their post-hearing brief, plaintiffs offer this court a formula for awarding damages in this context; they request $1,000 damages per plaintiff per year that the officer worked the desk more than she would have done if desk duty had been assigned on a strict rotational basis to all officers on the turn. This seems to be a reasonable formula; it offers the plaintiff a nominal amount for each year they were actually injured. This formula, therefore, should be followed. Therefore, because plaintiff Simon worked the desk more often than average from 1979 through 1982, 1984 through 1987, and 1991, a total of nine years, she should be awarded $9,000 compensatory damages for lost AT opportunity. Plaintiff Dellick worked the desk more than average in 1982, and from 1984 through 1987, a total of five years, and she should therefore receive $5,000 compensatory damages for lost AT opportunity.
The district court found that plaintiffs could have accumulated additional pay during this lost AT time, and the calculation of $1,000 per year was not clearly erroneous, but was a “fair reflection of damages incurred by plaintiffs because of defendant’s discriminatory activities.” Therefore, the district court ap*74proved the recommendation of the magistrate in this regard.
Defendant argues that the district court committed clear error by awarding these damages to plaintiffs because the amount of damages was speculative. For example, there was no evidence and no testimony about the number of arrests made or the AT time earned by plaintiffs while they were not on desk duty. There was no evidence presented about the average AT time earned by other officers who earned AT time by making arrests at night and going to court during the day, but were not assigned to the desk disproportionately. Defendant argues that the award was merely based on a suggested amount introduced for the first time by plaintiffs in their post-trial brief filed on November 2,1992, and was merely plucked from the air.
I agree with defendant’s argument. In order to receive damages under Title VII or 42 U.S.C. § 1983, the principles of common law torts are ordinarily used. Memphis Community School District v. Stachura, 477 U.S. 299, 305, 106 S.Ct. 2537, 2541-42, 91 L.Ed.2d 249 (1986). Back-pay awards are usually the make-whole remedy used to redress any economic harm suffered. Gutzwiller v. Fenik, 860 F.2d 1317, 1333 (6th Cir.1988). Although courts are granted discretion in their damage awards, damages still must be proved and speculative awards may not be given. Bigelow v. RKO Radio Pictures, 327 U.S. 251, 263-64, 66 S.Ct. 574, 579-80, 90 L.Ed. 652 (1946). I believe that in the present case, there is a way to determine approximately how much AT time was lost by being assigned to desk duty, but the district court failed to require that this amount be determined. Instead, the court merely accepted the figure offered by plaintiffs without any knowledge if the amount of AT time earned by officers working the streets and making arrests at night was equal to $1,000 a year. The record indicates that an officer could be compensated at $4 per hour for accumulating AT time. To warrant $1000 per year, plaintiffs would have had to accumulate 250 hours of AT time. There was no evidence to indicate that this number of AT hours could be accumulated during the time when plaintiffs were disproportionately assigned to the desk instead of being out on the street where AT time could be accumulated. I believe the amount of AT time earned by male officers who were also assigned to desk duty can be determined to provide an equitable damage award by comparison. Because plaintiffs failed to provide this calculation, I believe the district court must be reversed on this issue, and the case must be remanded for a calculation of damages that is not based on speculation.
To conclude, I believe the district court’s determination that plaintiffs were disproportionately assigned to desk duty on the basis of sex should be affirmed, but this case must be remanded to the district court for a more specific determination of damages.
. The majority's statement that males were permanently assigned to desk duty is not supported by the record. Officers were only assigned to desk duty if civilian personnel were not available to work on a substitution basis.
. Officer Dellick was not under Sabatino's command until 1988.