concurring.
I concur in the judgment, but I arrive at that conclusion by a slightly different route. Unlike the majority I believe the district court correctly found that plaintiffs made out a prima facie case. I conclude, however, that it was successfully rebutted.
Since Bryant and Lillibridge were hired as teachers this is a compensation case, not a hiring practices case. We proceed on the assumption that they and other male and female teachers on the faculty of the American School were equally qualified, or at least that differences in compensation other than the special allowances complained of reflected differences in qualifications or responsibilities. The narrow issue is whether either Bryant or Lillibridge made out a case that their sex was the reason they were denied equal compensation.
The evidence establishes that during the four years in issue there were 574 contracts awarded to teachers at the American School. Of these 476 were ISS sponsored, and thus paid higher benefits, and 98 were local-hire, and thus paid lower benefits. Of the 476 ISS sponsored contracts 240 were awarded to women and 236 to men. It thus appears that sex probably was not a factor in determining whether to award the ISS sponsored contracts. Of the 98 local-hire contracts, however, 97 were with females and only one with a male. All local hire contracts were to married persons, but in the instance of the one male awarded such a contract the evidence suggests that he received the lower compensation because for one year his qualifications were deficient. Thus, for purposes of our analysis, his contract may be disregarded. What then emerges is that in a pool of employees of equal qualifications, where two levels of compensation are paid, only females have been paid at the lower rate. This fact, without more, is circumstantial evidence sufficient to support the inference either that sex discrimination is the intended reason for that result (disparate treatment) or that some factor other than sex is causing a disparate impact on women.
I agree with the district court that these statistics about compensation sufficed to place on the employer the burden of producing evidence that a business reason other than sex produced the observed disparity. ISS did produce evidence that its dual contract policy was necessary because: (1) extra benefits were necessary to recruit persons coming to Iran primarily to teach; (2) the Iranian government objected to paying duplicate benefits to spouses of employees receiving benefits from another contractor; and (3) the extra benefits gave ISS a better hold on employees in Iran primarily to teach. The trial court credited this evidence, finding:
*578I conclude that these reasons, individually and in the aggregate, constituted legitimate, non-discriminatory reasons for awarding ISS-sponsored contracts only to persons whose primary purpose for being in Iran was to teach at the American School, and I also conclude that this basis for allocating ISS-sponsored and local-hire contracts had a manifest relationship to the employment in question.
Bryant v. International School Services, Inc., 502 F.Supp. 472, 487 (D.N.J.1980). This finding is not clearly erroneous, and suffices to overcome the inference from circumstantial evidence that sex discrimination in the award of local-hire contracts was intended.
Bryant and Lillibridge urge, however, that one additional factor must be taken into account. With one exception not relevant, all persons who were awarded local-hire contracts were married women. Thus, they claim, even if ISS’s business reasons are credited those practices had a disparate impact on women, and no sufficient justification was advanced for imposing that impact. The trial court concluded that awarding ISS contracts on the basis that the teacher’s primary reason for being in Iran was to teach there was a sufficient business reason for permitting such disparate impact as occurred. An inference can be drawn from the fact that only married women were given local-hire contracts that the pool of persons qualified as teachers who were in Iran primarily for purposes other than teaching there consisted largely of married women. But even accepting that inference, it in no way detracts from the legitimacy of the business purpose behind the local-hire differential. The makeup of the local-hire pool of qualified persons was determined not by the ISS business purpose, but by circumstances over which ISS had no control. The trial court held, and I agree, that the local-hire vs. ISS sponsored contract differential benefits program reflected a valid non-sexual business policy.
The trial court concluded, however, that this valid policy was implemented in a manner which had a discriminating effect upon Bryant and Lillibridge, because in each case ISS failed to award increased benefits to them when their primary purpose for being in Iran changed.1 I understand the court’s holding to be based on disparate impact rather than disparate treatment, for there is no evidence that the post-employment personnel practices it discussed were in any way different for male or female teachers. I do not think the court’s conclusion can withstand analysis.
The evidence is that ISS awarded contracts annually. Bryant entered into a local-hire contract in October 1976. In March 1977 she and her husband decided to separate and seek a divorce. She completed the 1976-77 school year and returned to the United States. The court concluded that from March of 1977 forward her primary purpose in being in Iran was to teach, and awarded higher benefits for that period. If there was any disparate impact in her case it arose not from the valid differential between local-hire and ISS sponsored contracts, but from the practice of the American School of awarding annual contracts. At the time she contracted the local-hire policy was correctly applied, and the business purpose for contracting with teachers on an annual basis is patently valid.
Lillibridge’s case is similar. She contends that her husband extended his commitment to Bell Helicopter in April, 1977 so that she could fulfill her contract for the 1976-77 *579year and continue teaching in the 1977-78 year. The Court awarded extra compensation from April 1977 through June 1978. As to the 1976-77 contract, the analysis we make in Bryant’s case applies. When the local-hire contract was awarded her primary purpose for being in Iran was other than teaching. That is not the case with respect to the 1977-78 contract. However the evidence on which the trial court relied in finding that the pay differential should be paid was that Lillibridge was not adequately informed of the availability of an ISS sponsored contract for persons situated as she was. At best, however, this lack of information establishes poor personnel administration, impacting equally on both men and women in the teaching staff of the American School. The sex-based impact of poor personnel practices is in her case happenstance. It does not support a charge of disparate sexual impact.
Thus I agree that the judgment appealed from must be reversed and judgment entered for ISS.
. The trial court’s conclusion was based on the fact that:
ISS failed to describe or disclose the basis of its policy to persons hired locally. Further, ISS ascertained the primary purpose for a person being in Iran by the simple expedient of determining if that person’s spouse worked in Iran for Bell Helicopter, Grumman, or perhaps another American company. If the teacher’s spouse did work for an American company in Iran, ISS assumed that the teacher’s primary purpose for being in Iran was not to teach at the American School but to accompany the spouse. ISS’s failure to inform its teachers or local applicants for teaching positions of the primary purpose test, prevented such a teacher from establishing her qualifications for an ISS-sponsored contract in those situations when her primary purpose for being in Iran was, in fact, to teach at the American School.
502 F.Supp. at 487.