concurring and dissenting:
I concur in the majority opinion except in its treatment of the discriminatory salaries claim.
There is no legal compulsion to treat the class members hired after 1965 differently from those hired before 1965 and every practical reason not to do so. I would hold that both sets are entitled to the remand for award of appropriate individual relief that the majority now orders only for those hired before 1965.
To require that those hired after 1965 now be put to the further expense and delay of making new proof of a pattern originating in the post-Act era and affecting only them is in my view both legally unjustified and most unfair.1 Taking into account the legal misapprehensions found by the Supreme Court to have tainted the district court’s finding that no pattern or practice had been established, that finding is as clearly erroneous with respect to the post-1965 hires as to the pre-1965 hires. Rightly viewed in light of the Supreme Court’s legal analysis, see 106 S.Ct. 3000, 3006-11, the evidence presently of record establishes beyond question that from at least 1968 2 until 1976 at a minimum, and possibly until as late as 1981, the Extension Service had engaged in a pattern or practice (or conceivably two technically different patterns) of race-based discrimination in the payment of salaries, surely affecting all pre-1965 hires and almost as surely affecting some post-1965 hires.
While it is obvious that more class members hired in the pre-1965 (pre-merger) era will be able easily to demonstrate specific disparities affecting them as individuals in the post-Act era, it is practically certain that so will some post-1965 hires. The evidence unmistakably showed that although post-1965 black hires started out in salary parity with their white counterparts, new disparities soon developed which showed up in the regression analyses of the work force well into the post-Act period. See 751 F.2d at 690 (dissenting opinion).
The proper way now to bring this protracted litigation over salary discrimination to a relatively speedy and eminently fair basis in light of the Supreme Court’s opinion is as follows. We should hold that the class claimants have established the existence of a Teamsters3 “pattern or practice” of race-based salary discrimination from as far back as 1968 to as late as 1981.4 On that basis, we should remand for Stage II proceedings with respect to all members of the class employed at any time during that period who desire to present individual claims.
In those Stage II proceedings, any employee presenting a claim would be entitled *484to benefit of the Teamsters presumption, see Teamsters, 431 U.S. at 359 n. 45, 97 S.Ct. at 1867 n. 45, that any disparity shown was traceable to the discriminatory pattern. But of course, only claimants able to show such a disparity could prevail. By this means the defendants would have ample protection against the individual claims of post-1965 hires. If, as defendants generally assert, no post-Act pattern of discrimination in respect of post-1965 hires ever developed, that will emerge in the efforts of such claimants to make the threshold showing of individual disparity.
In all fairness to these long-delayed claimants, defendants are entitled to no more, and they to no less. I therefore dissent from the majority’s treatment of the salary claims of the claimants hired after 1965.
.The majority thinks that different treatment of the two sub-sets is mandated by the Supreme Court’s discussion in footnote 8 of its opinion. The majority apparently reads this to require new proof by the post-1965 sub-set of a pattern or practice affecting only its members. With respect, I think this completely misreads the Court’s peripheral allusion to this subset. As I read it, the Court is simply saying that its textual discussion of the nature of a "continuing violation” obviously could not have reference to members of a class who only became employees after the Act made challenged conduct a "violation.” As to them there was no pre-Act conduct which by its “continuation" into the post-Act era became only then a "violation.” But the Court certainly is not saying that a “pattern or practice” might not be shown equally to have affected both pre-Act hires on a “continuing" basis and post-Act hires on an "original basis." Here, that is exactly what the evidence did show.
. The earliest date to which constitutional claims could run. See 751 F.2d at 690, 694 (dissenting opinion).
. See Teamsters v. United States, 431 U.S. 324, 336, 97 S.Ct. 1843, 1855, 52 L.Ed.2d 396 (1977).
. Concededly, the terminal date of the proven pattern or practice is not that clear on the evidence of record. Because the district court thought no pattern of any duration had been established, it did not address the question. The evidence, rightly assessed now in light of the Supreme Court's opinion, clearly establishes the existence of statistically significant salary disparities, hence a "pattern or practice,” down to at least 1976. See 106 S.Ct. at 3008; 751 F.2d at 691 (dissenting opinion). It simply is not clear how far past 1976 such statistically significant disparities, hence a proven pattern or practice, continued, though the opposing experts agreed *484that by 1981 the disparities, though still present, had become statistically insignificant. See 106 S.Ct. at 3008; 751 F.2d at 691-92 (dissenting opinion).
On this state of the record it would of course be possible on remand to require the district court to find the terminal date of the pattern or practice on the present or a reopened record. The more just disposition, however, is to assume its existence down to the latest date, 1981, when any general disparities were shown to exist. The effect of this would only be to extend the period during which individual claims could be shown to have arisen. They would still have to be established individually, and if the pattern did not actually exist at the time of their claims, they will presumably not be able to make the required threshold showing of individual disparity.