This matter is before us on the plaintiffs appeal from the dismissal, as time-barred, of his suit under the Age Discrimination in Employment Act, 29 U.S.C. §§ 621 et seq. In 1978, when he was 58 years old, Alexander Davidson was hired by Western Illinois University, a state institution, as a professor in its college of business administration. On October 5, 1985, the university signed a collective bargaining agreement with a union representing faculty members, including Davidson, which provided that initial salaries of faculty members would be determined by individual negotiation when the faculty member was hired, but that thereafter a faculty member could obtain an individual raise (as distinct from participating in a general raise for all faculty members) only by producing a bona fide written offer of employment, whether from another university, a consulting firm, or some other employer. Upon verifying the genuineness of the offer, the president of Western Illinois University is authorized, although not required, to grant the faculty member a raise not to exceed the salary in the offer.
On March 9, 1987 — seventeen months after the signing of the collective bargaining agreement — Davidson filed a complaint with the Equal Employment Opportunity Commission challenging the compensation system established by the agreement. He claimed that under it younger professors than he- with no greater experience, academic credentials, or even faculty rank were being paid higher salaries. Later he supplemented his complaint by instancing the cases of two assistant professors hired in August 1987 at salaries higher than his. He filed his complaint in district court on February 22, 1989. Relying primarily on
Delaware State College v. Ricks,
The defendants (the university and others, but we can ignore the others) argue that the judge was correct to dismiss the suit as time-barred. But they also argue that it is barred on other grounds as well, including the Eleventh Amendment, a ground that we must discuss first because it goes to the existence of federal jurisdiction over the case.
Edelman v. Jordan,
The Act defines “employer” (the class of potential defendants in age discrimination cases) to include “a State or political subdivision of a State and any agency or instrumentality of a State or a political subdivision of a state,” 29 U.S.C. § 630(b), and is explicit that an employer who violates the Act is liable for legal and equitable relief. 29 U.S.C. §§ 626(b), (c). Unless Congress had said in so many words that it was abrogating the states’ sovereign immunity in age discrimination cases — and that degree of explicitness is not required,
Pennsylvania v. Union Gas Co., supra; Dellmuth v. Muth, supra,
The next question is whether this suit was filed too late, which in turn depends on whether the administrative complaint was filed too late. The Supreme Court has held that the time for attacking a discriminatory practice begins to run when the practice is first applied to you, not when the application blossoms into a full-blown injury. In
Delaware State College v. Ricks, supra,
the plaintiff was denied tenure, allegedly for improper reasons, a year before his employment contract expired. The Court held that the statute of limitations started to run when tenure was denied, not when the plaintiff lost his job because, as a consequence of being denied tenure, his employment contract was allowed to run out.
Lorance v. AT & T Technologies, Inc.,
*444 Davidson contends that a compensation scheme under which an existing employee to get a raise must come up with an offer from another employer, while the starting salary of a new employee is not subject to any such limitation, discriminates against older workers because existing employees are likely to be, on average, older than new ones. As Davidson sees it, his university’s compensation scheme set in motion a chain of events certain to result eventually in his receiving a lower salary than some younger faculty members of no greater experience or ability. He can get raises only by producing offers from other employers, while new and presumably younger employees can sign on at any salary that the university is generous enough to offer them. The scheme doomed him to eventual salary inferiority just as the denial of tenure doomed Ricks to eventual loss of his job.
The analysis and conclusion can be questioned. Davidson himself was 58 when hired. There was no limit on his initial salary and we may conjecture that it exceeded that of some — perhaps of most or all — of his younger colleagues. And it is not the case that once hired you can get a raise only if you gin up an offer from the outside. You participate in all university-wide raises for faculty members. No doubt there are diminishing returns to years of service (a point to which we shall return), but whether Western Illinois University’s compensation scheme magnifies this phenomenon is conjectural.
However, we are not considering — not yet, anyway — the merits of Davidson’s claim, but only whether it was filed too late. And here the problem in upholding the district court’s ruling is precisely that the claim remained conjectural until new and younger faculty members were hired at higher salaries than Davidson. It was certain that Ricks had been denied tenure. It was an untested hypothesis that Davidson would eventually be passed by a younger colleague. If Davidson was a hot commodity in the academic or consulting markets — as some older professors are — he might be able to obtain outside offers that would persuade the Western Illinois University administration to keep raising his salary; he might end up both the highest-paid and the oldest member of the faculty.
These uncertainties become particularly important if we suppose that Davidson is" proceeding under a “disparate impact” rather than a “disparate treatment” theory of liability. He may not be able to prove that the compensation scheme was adopted
in order
to disadvantage the old, but — provided disparate impact is a permissible theory in age discrimination cases — he could still win his case if he could show both that the scheme had the effect of disadvantaging the old and that there was no good reason for it.
Wards Cove Packing Co. v. Atonio,
Sometimes the effect part of the disparate impact case — the disparate impact itself — is known from the start. We know the distribution of average heights of women relative to men, so we can assess the disparate impact of a minimum-height requirement for being a police officer. But it
*445
would be difficult to figure out the effect of Western Illinois University’s compensation scheme on older faculty members in any but the very crudest of terms (“the phenomenon of diminishing returns to years of experience is well documented in studies of academic salaries,”
Tagatz v. Marquette University,
There are two ways to handle this problem. The first is to hold that the claim of a plaintiff who wants to proceed on a disparate impact theory of liability does not arise (“accrue” is the common legalism) until the disparate impact occurs. The second is to hold that although the claim arises when the practice that he is challenging is first made applicable to him (when he is hired, if the practice was adopted earlier, or when the practice is adopted, if as here he had been hired before), provided only that he knows about the practice and its application to his situation, the statute of limitations will be tolled until he has had a reasonable opportunity to determine whether he has a claim. As we note in a companion case decided today,
Cada v. Baxter Healthcare Corp.,
We have found no cases that discuss the issue but we are satisfied that the second approach, which tolls the statute of limitations in appropriate cases but does not postpone the accrual of the plaintiffs cause of action, is superior. After all, it is the practice of which the plaintiff is complaining (here, the scheme of compensating new faculty differently from existing faculty), rather than the disparity of its impact upon the class to which he belongs; that disparity is just evidence, albeit essential evidence, that the practice is discriminatory, and hence actionable. What prevents his suing right off is not that the practice of which he complains has not been adopted as yet or has not been applied to him as yet but that he does not have enough evidence to determine whether the practice is unlawful and he could not obtain that evidence with due diligence — it doesn’t exist yet. For such cases, equitable tolling provides the tailored solution.
The district judge should not have dismissed the suit as time-barred, therefore, without determining that Davidson, when or shortly after the collective bargaining agreement was signed, had in hand, or could with due diligence have obtained, the minimum evidence he needed in order to determine whether he had a claim under the age discrimination law. But we agree with the defendants that the decision must be affirmed anyway, because the complaint fails to state a claim even if that law allows a plaintiff to proceed on the basis of disparate impact. The law forbids employment discrimination on the basis of age; it does not require that the old be paid higher salaries than the young. If the policy of Western Illinois University was to pay each of its faculty members the faculty member’s reservation price or opportunity cost — the salary he would command in a no less attractive job with some other employer — the university could not be accused of discrimination even if it turned out that, on average, younger faculty members were getting paid more than older ones.
It is true that an employer cannot defend a discriminatory wage pattern by pointing to the fact that, as a result of discrimination by other employers, blacks or women or members of some other statutorily protected group command lower wages, and it is therefore rational for him to pay them less than he pays white males.
Rucker v. Higher Educational Aids Bd.,
*446
Western Illinois University bases salary on the market. It will not give a faculty member a raise that raises him above the salary of his peers unless he provides proof of his market value, in the form of a bona fide written offer from another employer. The university does not apply this policy to new employees, since it could not hope to hire someone (like Davidson himself) away from that person’s existing employer without paying him more than his present salary. Because of the psychic and pecuniary costs of relocation, an existing employee can usually be retained if his offer from another employer is matched. This assumption cannot be made with a new employee. He is unlikely to leave his present job if his new employer offers him merely the same salary that he is getting in that job.
If disparate impact is a permissible theory in age discrimination cases, then a practice innocent on its face, as Western Illinois University’s compensation scheme is, could be shown to be discriminatory by considering its effect and the justifications offered for it. But not just by showing that one or two younger faculty members have been hired at higher -salaries than the plaintiff, an existing employee. For that is inevitable. Employees who stay longest are, on average though of course not in every case, those who have the fewest offers to go elsewhere. Their market alternatives are therefore relatively poor and they fare relatively badly under any market-based compensation scheme, irrespective of age. Employees newly hired must, as we have said, have been lured away from other jobs or job opportunities. Some of these employees will have a high market value and therefore command a high initial salary under Western Illinois University’s method of fixing starting salaries. The limited disparate impact that is all that Davidson has offered to show does not come close to demonstrating a taint of age discrimination.
So while disagreeing with the district judge’s ground of decision, we agree with his result.
AFFIRMED.
