OPINION OF THE COURT

HUTCHINSON, Circuit Judge.

Rutgers, The State University of New Jersey (Rutgers) appeals from a judgment of the United States District Court for the District of New Jersey granting appellee, Dr. Alfred Bennun (Bennun) promotion to Full Professor with full back pay retroactive to the end of the University's 1980-1981 promotion review period. The back pay amounted to the difference between Bennun's earnings as a tenured associate professor and what he should have been paid as a full professor, a position the district court held Rutgers had wrongly denied him. The judgment was fashioned to give Bennun full relief from Rutgers' violations of his right to be free of unlawful discrimination under 42 U.S.C.A. § 1981 (West 1981) and his right to equal opportunity in employment under Title VII of the Civil Rights Act of 1964 (Title VII), 42 U.S.C.A. §§ 2000e to 2000e-17 (West 1981). We will affirm in part. In doing so we hold that Rutgers' treatment of Bennun in the promotion review process did not violate § 1981 and will thus reverse that portion of the district court's judgment; but, as we agree with that part of the district court's order imposing liability because Rutgers' actions violated Title VII, we will affirm that part of its decision. Moreover, because all of the relief the district court granted Bennun was available to cure the Title VII violation, our partial reversal of the district court on Bennun's § 1981 claim for unlawful discrimination does not affect the remedy that the district court provided. Accordingly, we will affirm its remedial order.

I.

Bennun filed a charge of discrimination with the Equal Employment Opportunity Commission (the Commission) in 1981. He charged that Rutgers' denial of his bid for full professorship during the 1980-81 review period was not only discriminatory, but also in retaliation for his earlier filing of discrimination charges against Rutgers when it denied him promotion to associate professor with academic tenure. The Commission's investigation of Bennun's instant charge lasted over three years and resulted in a Commission determination that there was reasonable cause to believe that Rutgers violated Title VII when it failed to grant Bennun the rank of full professor. The EEOC's right to sue letter authorized Bennun to sue in his own right, but the Commission itself decided not to sue Rutgers.

In the meantime Bennun had also filed a grievance against Rutgers' refusal to afford him the rank of full professor, as permitted by the collective bargaining agreement between Rutgers and the American Association of University Professors. In 1982 a university review panel found the 1981 decision denying Bennun tenure was arbitrary and capricious. The review panel recommended that an outside arbitrator hear the matter. Rutgers did not implement this recommendation.

Rutgers' failure to implement the review panel's recommendation led Bennun to file suit in the Superior Court of New Jersey, Chancery Division, in 1984. In that suit Bennun sought to compel Rutgers to appoint an outside neutral party to hear and decide his grievance. The Chancery Division did not grant Bennun the relief he requested, but instead remanded the matter to the university with a direction that Rutgers undertake a further evaluation of Bennun's claim. Rutgers reevaluated Bennun in March of 1985 and persisted in denying him the status of a full professor.

In November of 1984, without awaiting the results of the reevaluation the state court had ordered, Bennun started this action in the United States District Court for the District of New Jersey pursuant to the Commission's right to sue letter. In it he alleged Rutgers violated Title VII, specifically 42 U.S.C.A. §§ 2000e-2(a)(1), 2000e-3(a) and 42 U.S.C.A. § 1981 when it denied him the rank of full professor in 1980-81. In August 1985 Bennun filed a second federal action with allegations similar to those in his 1984 federal complaint. The 1985 federal action arose out of Rutgers' 1982-83 refusal to grant him a full professorship. An identical refusal in the 1984-85 review period led Bennun to initiate still a third federal action in February of 1986. This 1984-85 refusal was occasioned by the reevaluation ordered by the Superior Court of New Jersey's Chancery Division. Bennun's second and third federal actions were, like his first, preceded by a Commission-issued right to sue letter.

All three federal actions were consolidated by the district court. Rutgers moved for partial summary judgment claiming, among other things, that Bennun had no standing to maintain his § 1981 claims because he was not Hispanic and that Bennun's 1984 state-court action precluded his claims based on the 1980-81 and 1982 promotion denials because of New Jersey's entire controversy doctrine. The district court denied Rutgers' motion on these points, although it did grant other relief on points not material to this appeal.

After a bench trial, the district court held that Bennun's denial of promotion to full professor was actionable under § 1981. See Bennun v. Rutgers, The State University, 737 F. Supp. 1393, 1397-98 (D.N.J. 1990). It then denied Bennun's retaliation claim under § 1981 and Title VII, but found that Bennun had proven his disparate treatment claims under both statutes. Id. at 1400-01, 1408-09. In its disparate treatment analysis, the district court compared Bennun's review packets with those of Dr. Ethyl Somberg, a professor in the same department as Bennun. Somberg had been promoted from Associate Professor to Full Professor in 1979. Id. at 1404-1409. Relying in part on its comparison of Somberg's academic credentials with Bennun's, the district court held that Bennun had made out a prima facie case of disparate treatment and that Rutgers' proffered nondiscriminatory reason, failure to meet the university's high standards for full professorship in the judgment of his peers, was a pretext for discriminatory denial of the promotion Bennun requested. In support of its decision the district court, comparing Bennun's promotion treatment with that of other tenured associate professors, in particular, with that of Somberg, found the following facts:

1. A different standard was applied to Bennun in terms of number of publications. (Bennun with more publications was [found] moderately active while Somberg[, with less,] was [found] excellent in quantity.)

2. A different standard was applied to Bennun concerning what level of achievement was necessary to be promoted. (Bennun was required to become an international leader while other promoted candidates, whose letters indicated they were not international leaders, were promoted.)

3. A different standard was applied concerning grant support. (For Bennun this was a negative factor, for Somberg it was not relevant.)

4. A different standard was applied concerning the level of specificity required by peer reviewers. (Bennun's [letters] were not specific while other candidates had none.)

5. A different standard was applied concerning the age of the peer review letters. (Bennun's were old and therefore negative. Somberg's, although just as old were not considered dispositive.)

6. A different standard was applied concerning the number of peer review letters. (The University inferred a lack of interest in Bennun but not in others who did not have the continued support evidenced in Bennun's packet.)

7. The [Appointments Promotions] committee sua sponte concluded, without any evidence, that a damning letter was being withheld from Bennun's packet. No such inference was drawn with reference to other candidates with similar or less substantial peer review packets.

8. Somberg's research was considered in light of her teaching. Bennun who taught a similar load was not so evaluated.

9. The University's explanation that Bennun's research was inadequate is not worthy of credence in light of a comparative analysis with other promoted candidates whose research qualifications were judged excellent.

Id. at 1409. The court ordered that Bennun be granted the rank of full professor retroactive to July, 1981, with full back pay and benefits. Id. This timely appeal by Rutgers followed.

II. A.

Promotion and tenure decisions at Rutgers are made on the basis of five criteria as set out in University Regulation 3.30. These five criteria are: teaching effectiveness; scholarly/creative activity; research accomplishments; professional activity/public service; and general usefulness. In Regulation 3.30 Rutgers also sets forth a standard for application of the criteria when tenured associate professors are under consideration for award of a full professorship:

A full professorship is the highest academic rank. An individual promoted to this rank should have met with distinction one or more of the criteria indicated previously, and have made substantial progress beyond that for which he or she was recognized at the associate professor level.

Joint Appendix (Jt.App.) at 1443.

The process of evaluating candidates for a full professorship begins when the candidate submits to his department chair a curriculum vitae, a list of potential outside evaluators and any other documents or materials the candidate wants to have considered. The persons who will serve as outside evaluators are selected by the candidate's department. The candidate's packet includes these materials, plus any materials added to the candidate's packet as his candidacy goes through successive levels of review and evaluation forms that are completed upon review at each successive level.

The candidate is first evaluated by those persons in his department who are at or above the level for which he is being considered. In this case, Bennun's departmental evaluators were all the full professors in his department. Based on these departmental evaluations, the candidate's department chairperson passes a departmental recommendation to the Dean of the academic unit that encompasses that department. An academic unit is made up of a group of related departments.

The candidate is then evaluated by his unit's Appointments and Promotions Committee, an advisory committee to the Dean. This committee has four members. Two are selected by the Dean and two by the faculty. The Appointments and Promotions Committee independently evaluates the candidate's packet.

After the committee's independent review, the candidate's Dean evaluates the packet. At the same time that the Dean is evaluating the candidate, the section, a University-wide group of all faculty members in the candidate's discipline at or above the level for which the candidate is being considered, evaluates the candidate's research and scholarly activity. The packet the section receives does not contain the recommendations of either the Dean or the Appointments and Promotions Committee.

After section evaluation, the university's Promotion and Review Committee, consisting of the University Executive Vice-President and Chief Academic Officer, the Camden, New Brunswick and Newark provosts and four senior faculty who are appointed by the President, reviews the candidate's packet which now includes the recommendations of the Dean and the section. This committee reports to the President if and only if its recommendation about the candidate is positive. The President then makes his own independent evaluation of the candidate's packet and at last submits a recommendation to the Board of Governors. The Board of Governors makes the final decision based on all of the candidate's packet including the President's recommendation.

B.

Bennun was born in Argentina and earned his bachelor's, master's and doctoral degrees from the University of Cordoba in that country. He joined the Department of Zoology and Physiology on Rutgers' Newark campus in 1969 as an untenured Associate Professor of Biochemistry. He became a tenured Associate Professor in 1972. Bennun's road to tenure was not smooth. Though he was unanimously backed for tenure in 1970 by his department he was ultimately denied tenure and given only a one-year terminal contract under the usual academic policy of up or out. Bennun secured tenure only after pursuing his legal remedies in court.

During the academic year 1978-79 Bennun applied for the position of Full Professor. He was turned down. His department had voted in favor of granting him full professorship by a plurality of three votes to two with one abstention. The Dean concurred in the department's recommendation. However, both the Appointments and Promotions Committee and the section voted against Bennun's promotion. The Appointments and Promotions Committee voted against Bennun zero to four. The section vote was two to six against Bennun with an abstention.

In 1980-81, Bennun again sought full professorship. His curriculum vitae then listed twenty-two published articles, including one being readied for publication, twenty-six abstracts, thirty-nine invited lectures, twenty-six paper presentations at professional meetings, sixteen different courses that he had taught at various times at Rutgers, and twenty-nine grants that he had received. Since his 1978-79 review, Bennun had readied one article for publication, given four invited lectures, made one paper presentation at a professional meeting and received two more grants.

The outside evaluations in Bennun's packet included four which had also been included in the 1978-79 packet. Three of these four evaluators were suggested by Bennun and one was solicited by Rutgers. Alfred Lehninger (Lehninger), the evaluator selected by Rutgers, said, while Bennun's work was not directly in his field, he was intrigued by it, thought it original and well supported. Because Lehninger was not directly involved in Bennun's specific academic field he referred Rutgers to Ephraim Racker (Racker). Racker had already been suggested by Bennun as an evaluator. He was described by Dr. James Hall, Chair of the Department of Zoology and Physiology at Rutgers' Newark Campus from 1962 to 1969, as "one of the, maybe 6 to 12 best, most widely known biochemists in the country." Jt.App. at 83.

Racker wholeheartedly supported Bennun for promotion. He wrote that "recent work on the effect of non-adrenaline on brain adenylate cyclase represents a significant contribution to this important field." Similarly, Sabatini, Chair of the Department of Cell Biology at NYU Medical Center, stated that Bennun was "well deserving promotion" based upon his "original scientific work. . . ." Finally, de la LLOSA noted that Bennun's research work on adenylate cyclase was "remarkable."

Bennun, 737 F. Supp. at 1402.

This time Bennun's department supported his application by a vote of five to zero with one abstention. The department rated Bennun outstanding in scholarly/creative activity and research accomplishments, above average in professional activity, above average to average in general usefulness and average in teaching effectiveness. The Dean concurred in the department's recommendation. He rated Bennun's performance as outstanding in scholarly/creative activity, outstanding to above average in research accomplishments, above average in professional activity and average in teaching effectiveness and general usefulness.

The remaining committees remained opposed to Bennun's candidacy. The Appointments and Promotions Committee voted unanimously against promotion, zero to four, ranking Bennun above average in professional activity but average in all other categories. It stated that Bennun's lack of outside grant funding since coming to Rutgers was puzzling and that his research accomplishments were tainted by his lack of experimental follow-up to prove his theoretical papers, the "apparent unwillingness of outside evaluators to commit themselves to detailed and informative judgment," Jt. App. at 1253, and a drop off in his publication rate since coming to Rutgers. The section voted against full professorship. Two members favored Bennun's promotion and four opposed his candidacy. The section concluded that although some of his research work was "substantial" its total was "insufficient to support promotion." Jt.App. at 1256. The Promotion Review Committee said "[t]he record indicates that he hasn't established a scholarly and [sic] research production to reach the level required for the rank of Professor." Jt.App. at 1257. The Promotion Review Committee rated Bennun as active professionally and useful to the graduate program, but only slightly above average in teaching effectiveness, The Board of Governors officially denied Bennun's request on May 15, 1981.

In 1982, undeterred, Bennun again sought the rank of full professor. In the two terms since he had last been considered Bennun had taught five different courses and overseen two research classes. Other additions to his file included publication of an article he had been readying for publication during his last review, another published abstract, two invited lectures, and another paper presented at a professional meeting.

The department recommended that Bennun be awarded a full professorship on a vote of four to zero with three abstentions. The department rated his scholarly/creative activity and research accomplishments as outstanding, his professional activity above average, his general usefulness slightly above average and his teaching effectiveness average. Again, the Dean concurred in the department's recommendation. He rated Bennun's scholarly/creative activity as outstanding to above average, his research accomplishments and professional activity as above average and his teaching effectiveness and general usefulness as average.

All four members of the Appointments and Promotions Committee voted against Bennun's candidacy. The Committee rated him above average in professional activity but average in all other categories just as it had during Bennun's previous review. The Committee criticized Bennun's lack of outside grant support and his diminishing research production. The Appointments and Promotions Committee described his outside evaluations as mixed and criticized Bennun because "his experiments, though carefully done, have not made an impact in his fields." Jt.App. at 1330.

Only three of the eighteen eligible section members voted. Two of the three voted against Bennun. One voted for him. The section said Bennun's publication record was unsatisfactory for a person seeking the rank of full professor. Thereafter, the Promotion and Review Committee failed to recommend Bennun be granted the status of full professor. Eventually the Board of Governors again denied Bennun's bid for promotion.

Bennun was again denied a full professorship after a 1983-84 review. Bennun has not sued Rutgers in connection with this review. Ironically, in 1983-84, for the first time, the section recommended Bennun as a full professor seven to two; but Bennun failed to get a recommendation to that effect from either his Dean or department chair. The department unanimously recommended Bennun but the Appointments and Promotions Committee narrowly voted not to recommend him with one member voting for him, two members against and one abstaining. The Promotion and Review Committee again did not recommend Bennun, and the Board of Governors reached the same decision.

In 1984-85, as stated, a reevaluation of Bennun's 1980-81 bid for promotion was undertaken as a result of Bennun's 1984 sojourn in state court. The first levels in the evaluation process relied on the 1982 department and section evaluations. Thus, the department again recommended Bennun for full professor, but the section did not. In a new evaluation, the Dean did not recommend Bennun: rating him above average in professional activity, above average to average for scholarly/creative activity, average in research accomplishments and general usefulness and average to below average in teaching effectiveness. The Dean expressed concern about Bennun's research, remarking that it "is not characterized by great originality or much impact on the profession," and what he said was Bennun's unremarkable publication output. Jt.App. at 1388. The Appointments and Promotions Committee did not recommend Bennun after rating him outstanding in professional activity, slightly above average in scholarly/creative activity, average in research accomplishments and general usefulness and average to below average in teaching effectiveness.

The Promotion and Review Committee also did not recommend promotion. It criticized Bennun primarily on his post-tenure research and said it did "not sufficiently reflect the recognition for original contributions to the discipline and leadership in its development which is necessary to justify this promotion." Jt.App. at 1375. The Board of Governors also denied Bennun's application.

III.

The district court had subject matter jurisdiction over Bennun's Title VII and § 1981 claims through its power to hear federal questions, see 28 U.S.C.A. § 1331 (West Supp. 1991), and its power to hear civil rights cases, see 28 U.S.C.A. § 1343(a)(3), (4) (West Supp. 1991). We have jurisdiction over the final order of the district court under 28 U.S.C.A. § 1291 (West Supp. 1991). Our scope of review varies with the individual issues the parties present. We will therefore separately set forth the scope of review at the outset of our analysis of each issue.

IV.

We begin with Rutgers' contention that Bennun's federal actions are precluded by New Jersey's entire controversy doctrine. Our review of the district court's conclusion that Bennun's present action was not barred by New Jersey's entire controversy doctrine is plenary. See Doerinkel v. Hillsborough, 835 F.2d 1052, 1054 n. 4 (3d Cir. 1987). The New Jersey entire controversy doctrine is a particularly strict application of the rule against splitting a cause of action. Like all versions of that rule its purpose is to increase judicial efficiency. Thus it precludes not only claims which were actually brought in previous litigation, but also claims that could have been litigated in the previous litigation. Bernardsville Quarry v. Borough of Bernardsville, 929 F.2d 927, 930 (3d Cir. 1991). Under settled federal case law we must give the 1984 New Jersey state-court judgment the same preclusive effect that New Jersey would give that judgment. See id. at 929-930 (quoting Migra v. Warren City School Dist. Bd. of Educ., 465 U.S. 75, 81, 104 S.Ct. 892, 896, 79 L.Ed.2d 56 (1984)). Accordingly, we have held that New Jersey's entire controversy doctrine applies to bar claims in a federal-court when there was a previous state-court action involving the same transaction. See, e.g., O'Shea v. Amoco Oil Co., 886 F.2d 584, 590-594 (3d Cir. 1989).

The district court rejected Rutgers' claim that Bennun's 1984 state-court action barred him from bringing the claims in this present case. It held the entire controversy doctrine did not foreclose any of Bennun's federal actions because "the 1984 state action sought relief relating solely to the employment agreement and not as to protection of [Bennun's] Constitutional and Civil rights." Jt.App. at 2267. Although we concur in the conclusion, we rely on a different rationale.

The crux of the entire controversy issue, according to Rutgers, is whether the events giving rise to Bennun's Title VII and § 1981 actions based on the university's 1980-81 and 1982 denials of full professorship arise out of the same single occurrence or transaction, or a related series of transactions or occurrences that gave rise to Bennun's 1984 state-court suit in the Chancery Division. Rutgers concedes that even if we reverse the district court on this point, Bennun's claims relating to the 1984-85 evaluation remain. See Reply Brief of Appellant at 14 n. 8. It makes this concession because the events that led to Bennun's 1984-85 denial of status as a full professor did not happen until after his state-court suit had ended. Thus, a reversal of the district court's holding on the entire controversy doctrine will not fully dispose of this case.

Bennun argues that application of the entire controversy doctrine is moot because even if it does apply, it would not bar his action concerning the 1984-85 denial. He reasons that the district court finding that he was qualified for promotion in 1980-81, determined, a fortiori, that he was qualified in 1984-85 when the remanded evaluation of the 1980-81 promotion denial occurred. Bennun further notes that Rutgers had committed itself to promote him retroactively to 1981 if the 1984-85 reevaluation resulted in a decision in Bennun's favor. Jt.App. at 1377. Accordingly, he says the district court could not have acted contrary to the entire controversy doctrine when it held he was entitled to an order directing Rutgers to grant him, retroactive to 1981, the status of full professor on its finding that the 1984-85 denial was unlawfully discriminatory.

"The Supreme Court [of the United States] frequently has explained that the mootness doctrine is derived from Article III's prohibition against federal courts issuing advisory opinions." See E. Chemerinsky, Federal Jurisdiction § 2.5.1, at 110 (1989). As a result, the doctrine has a jurisdictional tenor of constitutional dimensions. See Liner v. Jafco, Inc., 375 U.S. 301, 306 n. 3, 84 S.Ct. 391, 394 n. 3, 11 L.Ed.2d 347 (1964). If the specific facts or circumstances of this case renders Rutgers' entire controversy argument moot we cannot decide the entire controversy issue. We therefore address first Bennun's argument that the entire controversy issue is moot.

The district court found that Bennun was qualified for promotion to Full Professor at the time of the 1980-81 review and that he was discriminatorily denied the promotion. See Bennun, 737 F. Supp. at 1408-09. Bennun's argument that this finding means that he was a fortiori qualified for promotion in 1984-85 has some appeal. Because Rutgers had agreed to promote Bennun retroactively to 1981 if the 1984-85 evaluation was resolved in Bennun's favor, the remedy that Bennun would receive by winning his action based on the 1984-85 promotion denial would be the same as he would receive if he won his action based on the 1980-81 promotion denial. Thus, Rutgers' concession that the entire controversy doctrine does not preclude the claim based on the 1984-85 promotion denial at first appears to moot the issue. A closer examination though, casts doubt on this conclusion.

Assuming, without deciding, that the district court's determination that Bennun was qualified for promotion in 1980-81 necessarily implies that he was also qualified in 1984-85, the entire controversy doctrine would still be relevant to this case if we were to decide that the district court erred when it "implicitly" found Bennun was qualified for promotion in 1984-85 or that it erred when it found Bennun was discriminated against in the 1984-85 review. Based on the record relating to the 1984-85 denial, we might possibly decide that the district court's ruling on it could not withstand scrutiny. If so, in order to fully evaluate Bennun's claims, we would then have to see whether the record on the 1980-81 or 1982 denials supports the district court's finding; but the 1980-81 and 1982 denials are the very ones that Rutgers claims we cannot examine under the entire controversy doctrine. Thus, the entire controversy doctrine may still determine the viability of Bennun's claims and therefore its application to our decision on the earlier appeals is not moot.

On the merits of the entire controversy doctrine, Rutgers contends that the district court mistakenly focused on the different nature of the state and federal claims in failing to dismiss those arising out of the 1980-81 and 1982 denials. The claim regarding the 1980-81 denial was the subject of the grievance that led to the 1984 state-court action. In his state-court complaint Bennun alleged that he was "considered for promotion to the rank of full Professor and has continuously been denied said promotion" over the course of many years. Jt.App. at 2151. Because the 1982 decision was part of Bennun's string of promotion denials and was known to him before he filed the 1984 state-court action, Rutgers claims that the 1982 claim is also barred. At bottom, Rutgers argues that the entire controversy between Bennun and Rutgers is its repeated refusal to promote him to Full Professor and that he could have raised all of the issues relating to the 1980-81 and 1982 actions before the state court in 1984, and that his later claims are so related to the earlier denials as to bring the doctrine into play even on them.

In contrast, Bennun argues that he was seeking to vindicate different rights in the state, as opposed to the federal actions. He says the 1984 state-court action sought to compel Rutgers to stick to the terms of the collective bargaining agreement and the present suit seeks to address claims of discrimination. The earlier suit was not a part of the extended transaction regarding the merits of Bennun's promotion claims. Rather, he argues, the grievance it concerned made that transaction, in which he sought only to vindicate his procedural right to a full airing of his grievance, a transaction distinctly different and separate from those in which he sought relief from unlawful discrimination, especially since relief on the merits of his discrimination claim was not available on the state action to compel arbitration.

The Commission, arguing in support of Bennun, as amicus curiae, takes a slightly different tack. It argues that the entire controversy doctrine applies only if further litigation would later be needed to resolve the controversy at issue. If so, whatever is not originally litigated is barred. The Commission contends that no further litigation was necessary in 1984 to decide Bennun's right to arbitration when he brought the state court action to vindicate his asserted right to an impartial outside arbitrator under the collective bargaining agreement. Additionally, the Commission notes that the state and federal claims involved different facts and requested different relief.

According to the Commission, application of the entire controversy doctrine to Bennun's § 1981 and Title VII claims would ignore the fact that the merits of an underlying dispute cannot be addressed in an action to compel arbitration. The Commission says that if the entire controversy doctrine applies to this case, Bennun would have been compelled to request prematurely a right-to-sue letter from the Commission in order to litigate all of his possible claims at one time and so comply with the New Jersey's preclusive rule. Because a right-to-sue letter terminates administrative attempts to settle the dispute through conciliation, a means of resolving Title VII disputes that Congress sought to encourage, the Commission argues application of the doctrine to Bennun's case would frustrate congressional intent.

Rutgers responds that the cases the Commission refers us to on the judicial policy against resolving the merits of a claim during an action to compel arbitration are inapposite because Bennun's "1984 state-court action simply represent[ed] a step in his efforts to redress the alleged discriminatory denials of promotion." Supplemental Reply Brief of Appellants at 7. On the supremacy claims it also argues that in this case the time period for administrative action had passed by the time Bennun had filed the 1984 action so that the Commission's argument that the doctrine interferes with the Title VII conciliation process is not material. For the reasons that follow we find it unnecessary to decide the issue the Commission raises concerning potential conflict between New Jersey's judge-made entire controversy doctrine and Congress's intent to foster conciliation in Title VII cases.

We begin with an evaluation of the leading New Jersey cases on the entire controversy doctrine and this Court's interpretation of them. In Aetna Ins. Co. v. Gilchrist Bros., Inc., 85 N.J. 550, 428 A.2d 1254 (1981), the Supreme Court of New Jersey stated that the entire controversy doctrine "requires that a party include in the action all related claims against an adversary and its failure to do so precludes the maintenance of a second action." Id. 428 A.2d at 1257. The Aetna court noted that the doctrine traces its roots to Ajamian v. Schlanger, 14 N.J. 483, 103 A.2d 9 cert. denied, 348 U.S. 835, 75 S.Ct. 58, 99 L.Ed. 659 (1954), an opinion authored by then-Supreme Court of New Jersey Justice Brennan. The Ajamian court barred a plaintiff's damage claim against a defendant who had allegedly induced the plaintiff to enter into an agreement by deceit. The court held this claim was barred by an earlier action of the plaintiff for rescission of the contract. Justice Brennan wrote that the Ajamian holding was necessary to achieve judicial efficiency, a fundamental goal of New Jersey's then recent constitutional reform of its judicial system, by providing for the "expeditious determination in a single action of the ultimate merits of an entire controversy between litigants." Ajamian, 103 A.2d at 10, cited in Aetna, 428 F.2d at 1257.

Many cases since Ajamian have discussed what relation among claims is necessary to trigger the entire controversy doctrine. In Falcone v. Middlesex County Med. Soc., 47 N.J. 92, 219 A.2d 505 (1966) (per curiam), the plaintiff brought an action for damages because the defendant would not allow the plaintiff to join its society. The plaintiff had previously sued the defendant in order to compel it to accept him as a member. The court found that the second action was barred by the first and stated that:

[The] elemental considerations of fairness to the other party and the urgent need for eliminating the delay and wastage incident to the fragmentation of litigation dictated that all of the aspects of the plaintiff's controversy be included within [one] legal proceeding.

Id. 219 A.2d at 506.

Quoting from a Superior Court of New Jersey, Appellate Division case, we have said:

[If] the litigants in the action as framed will, after final judgment therein is entered, be likely to have to engage in additional litigation in order to conclusively dispose of their respective bundles of rights and liabilities which derive from a single transaction or related series of transactions, then the omitted component must be regarded as constituting an element of the minimum mandatory unit of litigation. That result must obtain whether or not the component constitutes either an independent cause of action by technical common-law definition or an independent claim which, in the abstract, is separately adjudicable.

Melikian v. Corradetti, 791 F.2d 274, 279-80 (3d Cir. 1986) (quoting Wm. Blanchard Co. v. Beach Concrete Co., 150 N.J. Super. 277, 375 A.2d 675, 683-84 (App. Div.), certif. denied, 75 N.J. 528, 384 A.2d 507 (1977)). The first suit in Melikian involved a breach of contract claim. We held that New Jersey's entire controversy doctrine barred a subsequent suit against the same defendant for fraudulently inducing the plaintiff to sign the contract. Id. at 280.

In Wm. Blanchard Co., the Appellate Division held that the claims of an owner of a building project against a general contractor were barred by the entire controversy doctrine. Wm. Blanchard Co., 375 A.2d at 684. The owner and the contractor asserted claims against one another and also sought to assert their rights to contribution and indemnification against a number of subcontractors. Id. at 678. A plethora of earlier suits among subcontractors on the project resulted in the owner and general contractor's being named as third-party defendants. Id. at 679-81. While the litigation had begun as early as 1970, the owner did not raise the claims in question until mid-1975. Id. at 679-82. The separate lawsuit initiated by the owner was barred by the entire controversy doctrine. Because all of the claims arose out of a construction project that was completed in 1971, the court held that the owner's claims were "a component of the controversy [that] may not be fairly withheld," id. at 684 (citing State v. Gregory, 66 N.J. 510, 333 A.2d 257, 264 (1975)), and "withholdable from the litigation only on penalty of forfeiture," id. at 685.

In Woodward-Clyde Consultants v. Chemical Pollution Sciences, Inc., 105 N.J. 464, 523 A.2d 131, 135 (1987), the court articulated various standards by which a court should determine if certain claims should have been brought together, including whether one claim was germane or related to the other claims. The court set these standards out against "the twin goals of judicial administration and fairness to litigants" that underlie the doctrine. Id.

These cases lend some plausibility to Rutgers' argument that the state-court action and Bennun's instant federal actions are part of a running dispute between him and Rutgers. While the state-court action only sought to compel arbitration, it did arguably arise from the same "related series of transactions," see Melikian, 791 F.2d at 279 (quoting Wm. Blanchard Co., 375 A.2d at 683-84), that are the basis for this case: the repeated denial of a promotion to full professorship for Bennun. In fact, both the 1984 state-court action and the first of Bennun's three suits that were eventually consolidated to form this case are based on the 1980-81 promotion denial.

Still, the touchstone of the entire controversy doctrine is judicial efficiency, as Justice Brennan noted in Ajamian and the Supreme Court of New Jersey recently reaffirmed in Woodward Clyde. The application of the entire controversy doctrine normally promotes judicial efficiency, as the cases outlined above suggest. In this case though, the application of the doctrine to Bennun's federal claims would frustrate the doctrine's basic purpose by short-circuiting otherwise available methods of non-judicial dispute resolution. Were we to accept Rutgers' position, a person who has a contractually based right to non-judicial process would have to forego vindication of that right to preserve his judicial remedies. If he did seek to compel non-judicial resolution of the dispute, the court in which he sought that relief could not, under this interpretation of the entire controversy doctrine, efficiently resolve the limited issue presented and effectuate the parties' purpose of avoiding judicial resolution of the merits. Instead, the court would be drawn (inefficiently) into the merits of the dispute. Parties entitled to judicial enforcement of their contractual rights to arbitration would find that entitlement provided only at the expense of their contractual rights. This result stands in stark contrast to the purpose underlying the entire controversy doctrine.

Thus, even assuming that the claims arising from the 1980-81 and 1982 promotion denials were related to the 1984 state-court action, we hold that New Jersey's courts would not apply the entire controversy doctrine to this case because Bennun's course of action was the most judicially efficient option open to him short of the Hobson's choice of either giving up his right to arbitration under the collective bargaining agreement or his right to sue in federal or state court on the merits of his unlawful discrimination claims. Bennun's state-court action was not "the trigger which . . . would start the chain reaction of other litigation to resolve the balance of the issue raised by the entire controversy," Wm. Blanchard Co., 375 A.2d at 684 (quoting Vacca v. Stika, 21 N.J. 471, 122 A.2d 619, 622 (1956)), but rather an attempt to avoid litigation altogether.

V.

Because the entire controversy doctrine does not bar Bennun's § 1981 claims or his Title VII claims we proceed to their consideration on the merits.

A.

We begin with Bennun's § 1981 claims. At the threshold Rutgers attacks the district court's conclusion that Bennun had standing to sue under 42 U.S.C.A. § 1981 (West 1981). It argues that Bennun is not Hispanic under the test set forth in Saint Francis College v. Al-Khazraji, 481 U.S. 604, 107 S.Ct. 2022, 95 L.Ed.2d 582 (1987). We have concluded, however, that it is not necessary to decide the question. We can instead dispose of Bennun's § 1981 claim on other grounds. The question of Bennun's standing to bring this § 1981 action is not one of those constitutional and prudential limitations that restrict a court's power to act and so must be resolved before we can proceed further on the merits. See Valley Forge Christian College v. Americans United for Separation of Church and State, Inc., 454 U.S. 464, 471-76, 102 S.Ct. 752, 757-61, 70 L.Ed.2d 700 (1982). Standing deficiencies of the kind Rutgers asserts are instead analogous to a party's failure to satisfy one element of a particular cause of action. See Bell v. Hood, 327 U.S. 678, 679-684, 66 S.Ct. 773, 774-777, 90 L.Ed. 939 (1946) (discussing the difference between matters that go to a court's jurisdiction and those that go to a party's cause of action); see also Air Courier Conf. of America v. American Postal Workers Union, ___ U.S. ___, ___ n. 3, 111 S.Ct. 913, 917 n. 3, 112 L.Ed.2d 1125 (1991) ("Whether a cause of action exists is not a question of jurisdiction, and may be assumed without being decided.") (citing Burks v. Lasker, 441 U.S. 471, 476 n. 5, 99 S.Ct. 1831, 60 L.Ed.2d 404 (1979)). Indeed, our jurisdiction to hear Bennun's § 1981 claim is based on other statutes. See, e.g., 28 U.S.C.A. § 1331; 28 U.S.C.A. § 1343(a)(3), (4). Thus we do not have to reach the question of § 1981 standing to decide Bennun's § 1981 claim.

B.

Relying on Patterson v. McLean Credit Union, 491 U.S. 164, 109 S.Ct. 2363, 105 L.Ed.2d 132 (1989) Rutgers contends that the district court erred when it ruled that Bennun stated a cognizable § 1981 claim because the change in status from tenured associate professor to full professor is not, at Rutgers, a change in position that creates or offers an opportunity for a new contract. The district court based its decision on four factors: (1) the "significant and substantial differences between the associate and full professor in terms of prestige and perception;" (2) the "differences in terms of function between the associate and full professor;" (3) the value of the position as evidenced by the elaborate promotion procedure; and (4) the stigma that accompanies the perennial rejection for promotion. Bennun, 737 F. Supp. at 1398. Since the parties do not dispute the facts as found by the district court but rather argue about the application of Patterson, the question, as presented, is one of law and our review is plenary. See Doerinkel, 835 F.2d at 1054 n. 4.

Patterson limited § 1981's reach to racial discrimination in the making and enforcing of contracts. Patterson, 109 S.Ct. at 2375. Brenda Patterson was a teller and file coordinator at a credit union. Id. 109 S.Ct. at 2368-69. She alleged that she was fired, harassed and, at one point, not promoted to an intermediate accounting clerk position because she was black. She further alleged that she was entitled to relief under § 1981 and North Carolina state law. Id. at 2369. The issue of importance to us is that she was not promoted because she was black. The Supreme Court held that denial of such a promotion would only be actionable when it "rises to the level of an opportunity for a new and distinct relation between the employee and the employer." Id. at 2377. Effectively, the Court held that in the promotion context a new "contract" is not made for purposes of § 1981 unless such a new and distinct relationship would flow from the promotion.

Rutgers takes issue with the district court's use of all but the second factor. It argues that "stigma," "perception" and "value" are all irrelevant to the analysis under the Supreme Court's decision in Patterson. It also disagrees with the district court's conclusion that the position of tenured associate professor was so functionally different from the position of professor that it met the Patterson requirements. Rutgers says both positions are supervisory and while there may be a new contract within the Patterson rationale when there is a change from a non-supervisory to a supervisory position, that is not so when the change is just from one supervisory position to another. We agree with Rutgers' criticism of the district court's use of the value, stigma and perception factors. None of them indicate the creation of a new contract because they do not touch any of the terms of the contract as to duties, tenure, compensation or essential function. Virtually all of the research and student-oriented responsibilities of Full Professors and tenured Associate Professors are the same. Holders of these two positions do nothing different day-in and day-out except for those few weeks out of the year when they evaluate candidates, if any, for a full professorship.

Based on its argument that no substantial difference exists between the two positions at Rutgers, the University would have us conclude that there is no analogy between a change to the position Bennun covets from the one he holds. To resolve this question we turn to the record. It shows that both Associate and Full Professors evaluate all faculty candidates for promotion up to the level of Associate Professor, but Associate Professors have no role in deciding who get the status of Full Professor. The teaching, research and advisory functions of tenured Associate Professors and Full Professors are the same and both groups are members of the same collective bargaining unit. The single difference between the two positions is participation in the review of candidates for promotion to Full Professor. This, Rutgers says adds only one advisory capacity and cannot of itself create a new contract for tenured associates who attain a full professorship within the meaning of Patterson.

Bennun, however, points out one other difference between Full Professors and tenured Associate Professors. Only Full Professors are eligible to become members of, or to review candidates for, the four President-appointed spots on the Promotion Review Committee. This is the committee that ultimately makes recommendations to the President about changes in faculty status. In addition, the process used to evaluate candidates for Full Professor is also used to evaluate candidates for tenure. Since the grant of tenure creates a new contract, Bennun asks us to infer that use of the same process in evaluating candidates for Full Professor shows attainment of status as a full professor also creates a new contract. We reject at once this last stated argument and move quickly on in our analysis. The end result of a tenure award is vastly increased responsibilities and benefits. The tenure award represents a change in the fundamental terms of the contract. No commensurate change occurs with an award of full professorship.

42 U.S.C.A. § 1981 states in part that "[a]ll persons . . . shall have the same right . . . to make and enforce contracts. . . ." In Patterson, the Supreme Court supplied a working definition of when a promotion entails a new contract. See Patterson, 109 S.Ct. at 2372. Since § 1981 does not protect post-formation discrimination this definition is crucial. See id. at 2377. A promotion becomes actionable under § 1981 when "the nature of the change in position was such that it involved the opportunity to enter into a new contract with the employer." Id. In further explanation the Supreme Court referred to Hishon v. King Spalding, 467 U.S. 69, 104 S.Ct. 2229, 81 L.Ed.2d 59 (1984). See Patterson, 109 S.Ct. at 2377. Hishon was a Title VII case. It involved a law firm's discriminatory refusal to promote an associate to partner. Hishon, 467 U.S. at 71-72, 104 S.Ct. at 2231-32. The Court cited Hishon as an example of a "new and distinct relation between the employee and the employer." Patterson, 109 S.Ct. at 2377. The Fifth Circuit has thoughtfully considered the Supreme Court's citation to Hishon:

The difference between rights, duties, and compensation of associates and partners in a law firm are significant. While both the associate and partner are attorneys, the role of partner is supervisory and carries increased responsibility. It may also involve personal liability for the affairs of the firm. The method of compensation and amount of benefits attendant to an associate position differ from those of a partner.

Harrison v. Associates Corp. of North America, 917 F.2d 195, 198 (5th Cir. 1990) (finding no new contract in a promotion from computer operator to lead computer operator where there were "no significant change in duties and responsibilities"). We do not think that the change in status from tenured Associate Professor to Full Professor at Rutgers is like the change from associate to partner in a law firm. The change in status in Bennun's case is much closer, essentially, to the change in positions in Harrison. The differences in the duties between the two academic ranks and the jobs of the workers are, in both cases, too small for the creation of a new contract under Patterson. For this reason, we hold that the promotion of a tenured Associate Professor to Full Professor at Rutgers is not a new contract for purposes of § 1981. Therefore, Bennun's § 1981 claim must fail.

VI.

Rutgers' final contention is that the district court erred in holding that Rutgers violated Title VII when it denied Bennun a promotion to full professor. This attack, largely fact bound, proceeds on several fronts.

A.

Before considering Rutgers' arguments on the merits of Bennun's Title VII claims we will outline in general terms the law relating to burdens of proof and sufficiency of the evidence in Title VII cases.

Insofar as Rutgers' challenges go to the sufficiency of the evidence to support the district court's finding of discrimination, we review for clear error. See Bellissimo v. Westinghouse Elec. Corp., 764 F.2d 175, 179 (3d Cir. 1985) (citing Fed.R.Civ.P. 52(a)), cert. denied, 475 U.S. 1035, 106 S.Ct. 1244, 89 L.Ed.2d 353 (1986), overruled on other grounds, Price Waterhouse v. Hopkins, 490 U.S. 228, 109 S.Ct. 1775, 104 L.Ed.2d 268 (1989). To the extent it challenges the district court's application of law to the facts as found, our review is plenary. See Logue v. International Rehabilitation Assocs., Inc., 837 F.2d 150, 152 (3d Cir. 1988).

The alternating burdens of proof in a Title VII discrimination case are allocated as follows. First, the plaintiff must establish a prima facie case of discrimination. See Texas Dept. of Community Affairs v. Burdine, 450 U.S. 248, 252-253, 101 S.Ct. 1089, 1093-94, 67 L.Ed.2d 207 (1981). The proof has three prongs. The plaintiff must prove by a preponderance of the evidence that he is a member of a minority; that he applied for, is qualified for and was rejected for the position sought, and that non-members of the protected class were treated more favorably. Roebuck v. Drexel Univ., 852 F.2d 715 (3d Cir. 1988) (quoting McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802, 93 S.Ct. 1817, 1824, 36 L.Ed.2d 668 (1973)). After this showing is made, the defendant may rebut the discriminatory presumption by showing that there was a "legitimate, nondiscriminatory reason" why someone else was preferred. McDonnell Douglas Corp., 411 U.S. at 802, 93 S.Ct. at 1824. If the defendant does so, then the burden, now an ultimate one, is reassigned to the plaintiff and the plaintiff must show that the reasons proffered by the defendant are pretextual. See id. at 804, 93 S.Ct. at 1825; see also Kunda v. Muhlenberg College, 621 F.2d 532, 541-42 (3d Cir. 1980) (quoting McDonnell Douglas Corp., 411 U.S. at 802, 93 S.Ct. at 1824) (discussing Title VII burdens of production).

This Court has held that "more than a denial of promotion as a result of a dispute over qualifications" must be shown to prove pretext. Molthan v. Temple Univ., 778 F.2d 955, 962 (3d Cir. 1985). In Molthan, a case also involving a tenured associate professor who sought a promotion to full professor, we decided that the plaintiff's prima facie case had been rebutted by Temple's proffered nondiscriminatory reason for failing to promote her. Temple argued that the plaintiff was denied promotion because "the majority of those who voted on each of [the plaintiff's] promotion decisions believed that she was not qualified for promotion." Id. (citing Zahorik v. Cornell Univ., 729 F.2d 85, 94 (2d Cir. 1984)). The plaintiff attempted to show pretext by establishing that she had been recruited under the implication that tenure would follow, that she was granted tenure, that Temple had tried to persuade her to change the department she was assigned to and that the majority of full professors at Temple were male. Id. We held that the plaintiff's evidence "raise[d] no inference of sex discrimination," id. at 963, and that:

[F]or a plaintiff to succeed in carrying the burden of persuasion, the evidence as a whole must show more than a denial of tenure [or promotion] in the context of disagreement about the scholarly merits of the candidate's academic work, the candidate's teaching abilities or the academic needs of the department or university.

Id. at 962 (quoting Zahorik, 729 F.2d at 94). We have, however, held that such a dispute will satisfy the earlier hurdle of establishing the qualifications of the professor, as long as the plaintiff demonstrates that "he was sufficiently qualified to be among those persons from whom a selection, to some extent discretionary, would be made." Roebuck, 852 F.2d at 726 (quoting Banerjee v. Board of Trustees of Smith College, 648 F.2d 61, 63 (1st Cir.), cert. denied, 454 U.S. 1098, 102 S.Ct. 671, 70 L.Ed.2d 639 (1981)). We also noted, citing to Burdine, 450 U.S. at 256, 101 S.Ct. at 1095, that actual evidence of discrimination is not needed to disprove pretext. Rather, proof that the employer did not act for the reason asserted is enough. Roebuck, 852 F.2d at 726.

B.

With this general outline of the shifting burdens of proof in Title VII cases, we pass to consider Rutgers' individual arguments against Title VII liability.

1.

We consider first Rutgers' argument, renewed in the Title VII context, that Bennun cannot recover because he failed to show that he is Hispanic and therefore has not established the first element necessary for a Title VII prima facie case, that he is a member of a protected class. See Brief for Appellants at 29, n. 18. Citing to Bennun's testimony, Rutgers argues:

His mother, Ann Odesser, was born in what is now Romania and first spoke Romanian [Jt.App. at 209]. His father, Leon Bennun, was born in what is now Israel and first spoke Ladino, a dialect used by Sephardic Jews [Jt.Ap. at 210]. Plaintiff conceded that neither his mother nor his father is Hispanic [Jt.App. at 209-10].

Brief of Appellant at 26. Therefore, Rutgers says that Bennun cannot be Hispanic because his forebears did not come from that ethnic group. Rutgers made this argument in its attack on Bennun's § 1981 claim but by reference adopted the argument that Bennun was not Hispanic as a basis for defeating Bennun's Title VII prima facie case. Id. at 29, n. 18; Reply Brief of Appellant at 24, n. 14. Though Patterson v. McLean made it unnecessary to decide Bennun's ethnic origin in the context of his § 1981 claim, we must face it in the Title VII context.

Initially, we reject Bennun's contention that Rutgers waived the argument that he is not Hispanic by its concession to the district court that he had standing under Title VII. See Jt.App. at 2260. This concession did not relieve Bennun of the need to establish a prima facie case on the merits of his Title VII claim. One element of that prima facie case is proof of membership in a class the statute protects.

Rutgers' argument that Bennun cannot recover on his Title VII claim unless he proves he is Hispanic implies that the minorities protected by Title VII are the same as those § 1981 covers. It fails to confront differences between groups protected by § 1981 and Title VII. Rutgers does not cite any authority for its assertion that the minorities Title VII protects are only those who have standing under § 1981. It relies on Saint Francis College v. Al-Khazraji, 481 U.S. at 607, 107 S.Ct. at 2025, a case decided in the context of § 1981. Al-Khazraji did not hold that a lack of standing under § 1981 implies a lack of protection under Title VII. The Equal Employment Opportunity Commission (EEOC), the agency charged with interpreting Title VII, has concluded otherwise and its regulations would allow Bennun to rest his claim solely on his national origin, see 29 C.F.R. § 1606.1 (1990), a basis that would not be sufficient for a § 1981 claim under Al-Khazraji. See Al-Khazraji, 481 U.S. at 613, 107 S.Ct. at 2028.

Rutgers' contention that Bennun cannot prevail upon his Title VII claim if he is not Hispanic also ignores the language of § 2000e-2. It provides, in material part:

(a) It shall be an unlawful employment practice for an employer —

(1) to fail or refuse to hire or to discharge any individual, or otherwise to discriminate against any individual with respect to his compensation, terms, conditions, or privileges of employment, because of such individual's race, color, religion, sex, or national origin; or

(2) to limit, segregate, or classify his employees or applicants for employment in any way which would deprive or tend to deprive any individual of employment opportunities or otherwise adversely affect his status as an employee, because of such individual's race, color, religion, sex, or national origin.

42 U.S.C.A. § 2000e-2(a). Section 1981 does not mention national origin. It provides:

All persons within the jurisdiction of the United States shall have the same right in every State and Territory to make and enforce contracts, to sue, be parties, give evidence, and to the full and equal benefit of all laws and proceedings for the security of persons and property as is enjoyed by white citizens, and shall be subject to like punishment, pains, penalties, taxes, licenses, and exactions of every kind, and to no other.

42 U.S.C.A. § 1981. If Al-Khazraji does not control in cases arising under § 2000e-2, Bennun's status as a member of a protected group could be based on his conceded national origin as a native of Argentina.

On this record, however, we are unwilling to rest our holding that Bennun has proven that he is a member of a minority protected by Title VII on his national origin. There are several reasons. First, it would be unfair to Rutgers to base our decision on that theory. The parties did not try this case on the theory that Rutgers unlawfully discriminated against Bennun because of his birth in Argentina.

Secondly, the district court's findings and opinion indicate to us that Bennun had persuaded it that he was discriminated against as a Hispanic. They give no hint that the district court was also persuaded Bennun would have been denied promotion simply because he came from Argentina. We are therefore unable to conclude that there would have been the same ultimate finding of unlawful discrimination of Bennun were Argentinian by birth, but not Hispanic in culture, language and appearance. Accordingly, we will not rely on his national origin as an Argentinian to show he is a member of a protected minority under Title VII.

Rutgers' argument that Bennun has not shown he has an ethnic basis for his Title VII claim, as a Hispanic, nevertheless fails in this case. The district court's finding that Bennun was Hispanic can be disturbed by this Court only if it is clearly erroneous. See Bellissimo, 764 F.2d at 179. Although this finding was made in addressing Rutgers' § 1981 attack, if it is not clearly erroneous, it necessarily supports the district court's ruling on Bennun's Title VII claim. The district court noted Bennun's uncontradicted assertion in an affidavit that his father was a Sephardic Jew who, like all Sephardic Jews, traced his lineage to those Jews who were expelled from Spain during the Spanish Inquisition of 1492. The court, however, based its finding that Bennun is Hispanic on his birth in Argentina, his belief that he is Hispanic, identifies with and continues to adopt Spanish culture in his life and speaks Spanish in his home.

The district court's determination that Bennun is Hispanic is not clearly erroneous. The word "Hispanic" has been defined as "of, or derived from Spain or the Spanish." Webster's New Dictionary and Thesaurus 262 (Concise ed. 1990). Bennun's uncontroverted assertion that his father is a Sephardic Jew meets the dictionary definition of Hispanic. Rutgers' argument that Bennun conceded that his father was not Hispanic subtly misconstrues the record. At his deposition, Bennun, in response to Rutgers' counsel's question, "Do you contend that your father is Hispanic?," Jt.App. at 210, answered "No," id. However, Bennun went on to explain that he did not contend that his father was Hispanic because it is a fact that he was by virtue of his Sephardic roots.

Bennun's birth in a Latin American country where Hispanic culture predominates, his immersion in Spanish ways of life and the fact that he speaks Spanish in the home also support this conclusion. Census information often relies on the primary language spoken in the home as a means of classifying persons by race. See, e.g., Castaneda v. Partida, 430 U.S. 482, 486 n. 5, 97 S.Ct. 1272, 1276 n. 5, 51 L.Ed.2d 498 (1977). In addition, the district court was able to see Bennun firsthand. The district court's opportunity for observation about Bennun's appearance, speech and mannerism is informative as we apply the clearly erroneous standard in reviewing the finding by the district court that Bennun in Hispanic. We think unlawful discrimination must be based on Bennun's objective appearance to others, not his subjective feeling about his own ethnicity. Discrimination stems from a reliance on immaterial outward appearances that stereotype an individual with imagined, usually undesirable, characteristics thought to be common to members of the group that shares these superficial traits. It results in a stubborn refusal to judge a person on his merits as a human being. Our various statutes against discrimination express the policy that this refusal to judge people who belong to various, particularly disadvantaged, groups is too costly to be tolerated in a society committed to equal individual liberty and opportunity. Thus, coupling the evidence the cold record presents with the district court's opportunity to observe Bennun, we hold that its determination that he was Hispanic is not clearly erroneous.

2.

Having found that Bennun was Hispanic for purposes of Title VII, the district court went on consider whether Bennun had established the other elements necessary for a prima facie case. It began this analysis by noting that Bennun conceded that there was no direct evidence to support a prima facie case of discrimination, but relied instead on circumstantial evidence of disparate treatment. See Bennun, 737 F. Supp. at 1400. We examine the record with respect to the remaining elements of Bennun's prima facie case in that context.

One of the elements Bennun must show to make out his prima facie case is "that he applied and was qualified for a job for which the employer was seeking applicants." McDonnell Douglas Corp., 411 U.S. at 802, 93 S.Ct. at 1824. The district court recognized that, in an academic setting, decisions about promotion are best left to the academics. See Bennun, 737 F. Supp. at 1409. Rutgers complains that the court nevertheless functioned as a "Super-Tenure Board" and determined that Bennun was qualified for the position of Full Professor without giving any deference to the judgments of Bennun's academic evaluators. Rutgers notes that of the seven levels of review available in 1980-81, Bennun was viewed favorably by only two. Of the ten other professors whose promotion packets were submitted into evidence, all received approval at at least five levels while seven, including Somberg, received approval at all seven.

Bennun dismisses Rutgers's argument that the court should have paid more heed to the academic judgments of the evaluators because it is these very judgments that are poisoned by discriminatory animus. Bennun contends that the district court was correct in looking at the evaluations made by the other professionals in his own department and his international recommendations to determine that he was qualified for the position of full professor. He points out that once the district court had reason to question the objectivity and neutrality of Bennun's evaluators, it appropriately "evaluated the underlying evidence contained in the promotion packets and the peer review letters to determine whether reasonable, non-discriminators could plausibly reach and recite the conclusions university spokespeople expressed." Brief for Appellee at 29-30.

Kunda v. Muhlenberg College is the seminal case on this topic in our Court. The plaintiff in Kunda was an instructor who, after five years at Muhlenberg, was considered for tenure and promotion to Associate Professor. Kunda, 621 F.2d at 536-37. She was turned down and, despite several appeals in her favor within the college's reassessment procedures, received a terminal contract for the 1974-75 academic year. The President of the school stated that she was denied promotion because she lacked a terminal degree in her field. However, the plaintiff was able to show that three people without terminal degrees were promoted during the period of her employment at Muhlenberg. The district court determined that Kunda had set forth a case of discrimination. Id. at 536-40. This Court affirmed. In Kunda we stated that no special deference is to be paid to the tenure and promotion decisions of universities when they are scrutinized under Title VII because:

The fact that the discrimination in this case took place in an academic rather than commercial setting does not permit the court to abdicate its responsibility to ensure the award of a meaningful remedy. Congress did not intend that those institutions which employ persons who work primarily with their mental faculties should enjoy a different status under Title VII than those which employ persons who work primarily with their hands.

Id. at 550. We did, however, caution courts not to "substitute their judgment for that of the college with respect to the qualifications of faculty members for promotion and tenure." Id. at 548.

Kunda was followed by EEOC v. Franklin Marshall College, 775 F.2d 110 (3d Cir. 1985), cert. denied, 476 U.S. 1163, 106 S.Ct. 2288, 90 L.Ed.2d 729 (1986). In Franklin Marshall, the main issue was whether the school had any privilege to withhold discovery of peer review materials. There the Commission was investigating an allegation of discrimination by the College in violation of Title VII. Id. at 111. It sought to compel the disclosure of confidential peer review materials "[f]or each individual granted or denied tenure" over approximately a five year period. Id. at 112 (quoting EEOC subpoena) (emphasis in quoted material). In Franklin Marshall, this Court rejected any special treatment of academic institutions, unlike the Second Circuit in Gray v. Board of Higher Educ., City of New York, 692 F.2d 901, 904 (2d Cir. 1982) (adopting a balancing approach towards compelling discovery in the academic context), and the Seventh Circuit in EEOC v. University of Notre Dame Du Lac, 715 F.2d 331, 337-38 (7th Cir. 1983) (recognizing a qualified privilege protecting materials related to the peer review process). Instead, we held that there is no privilege in such materials and moreover, no balancing process is needed before disclosure of peer review materials relevant to the investigation is required. We held that "[n]o inference can be drawn from the legislative history of Title VII, as amended, that Congress intended or would permit academic institutions to bar the EEOC's access to material relevant to an investigation." Id. at 115. The relevancy of these peer review materials to an investigation of unlawful discrimination within the academy logically implies that they are to be used in determining whether discrimination has taken place. Indeed, the Supreme Court vindicated this Court's position just last year when it held that there was no special privilege that exempted university peer review materials from the scope of the Commission's subpoena powers. University of Pa. v. EEOC, 493 U.S. 182, 110 S.Ct. 577, 582, 107 L.Ed.2d 571 (1990) ("[W]e cannot accept the University's invitation to create a new privilege against disclosure of peer review materials.").

Accordingly, the district court sought to determine if Bennun had demonstrated that he was qualified for a full professorship. In summarizing Bennun's bid for a full professor's title the district court began with the 1978-79 review. In that review Bennun's research and scholarly activity was rated outstanding by his department, his professional activity was rated above average and his teaching and general usefulness were regarded as average to below average. The department voted in favor of promotion. The Appointments and Promotions Committee, voting against promotion, rated him above average in scholarly activity and research, primarily because his publication in refereed journals was modest, average to below average in teaching and general usefulness, but this committee noted he was "`extraordinarily' active" in his professional activity. Id. at 1401-02. The Dean, who also voted in favor of promotion, rated him above average to outstanding in scholarly activity, professional activity and research, although he noted that there "`continues to be a question' about his research," id. at 1402, and rated him average to below average in teaching and general usefulness. The Promotion Review Committee found "`insufficient evidence of distinction in teaching and scholarly work to warrant promotion.'" Id.

Bennun's credentials at the time of the 1978-79 review included twenty refereed articles, eight of which had been written in the eight years since he was granted tenure. Bennun had also published a large number of abstracts and papers which are shorter than articles, delivered a substantial number of lectures and continued to receive significant grant support and teach classes covering a wide variety of subjects. Bennun had outside letters submitted by four internationally known persons in his field who wholeheartedly recommended him. See supra at 161. Three of these letters were solicited by Bennun and one was solicited by the University. Id.

The district court then reviewed Bennun's credentials as of his 1980-81 review for promotion from tenured associate to full professor. This was one of the reviews at issue. Bennun's lectures, presentations and abstracts continued to increase. The department again recommended him for promotion, this time 5-0-1. His ratings were the same as in 1978-79, except that his teaching and general usefulness had increased to average and average to above average respectively. The Appointments and Promotions Committee unanimously rejected Bennun, rating him average in every category except professional activity where he was rated above average. This committee was critical of the lack of outside acknowledgment of Bennun's research, his failure to receive outside funding, and was convinced that the department was withholding a letter that spoke unfavorably of Bennun. The Dean recommended Bennun, while the section, critical of Bennun's output, and the Promotion and Review Committee recommended against promotion. Id. at 1402-03.

Finally, after reciting "a similar pattern" in the 1983-84 application, the district court found that Bennun had proved that he was qualified for promotion. Id. at 1403-04.

Rutgers takes issue with the district court's conclusion that Bennun is qualified for promotion to full professor. As we have explained:

[I]n order to satisfy the qualifications element of [a prima facie case], plaintiff "need only show that he was sufficiently qualified to be among those persons from whom a selection, to some extent discretionary, would be made. That is, he need show only that his qualifications were at least sufficient to place him in the middle group of tenure candidates as to whom both a decision granting tenure and a decision denying tenure could be justified as a reasonable exercise of discretion by the tenure-decision making body."

Roebuck, 852 F.2d at 726 (quoting Banerjee, 648 F.2d at 63) (citations omitted); see Fowle v. C C Cola, 868 F.2d 59, 64 (3d Cir. 1989). The positive votes from his own department year after year and the favorable letters of recommendation that Bennun received from various world renowned scientists in his field show objectively that Bennun was at least in this middle group of candidates for full professor. In addition, the Dean also favored Bennun's promotion in 1980-81 and 1982 as did the section in 1984-85.

Rutgers maintains that Bennun cannot be in the middle group of tenure candidates because he received support at only a few levels of review while the ten professors he was compared with received support from almost every level of review. See Brief of Appellant at 33. Rutgers misunderstands the analysis set forth in Roebuck concerning the "middle group" test. Whether a candidate is in the middle group for prima facie case purposes is not to be determined by whether he was in the middle group of those applicants whose credentials are in evidence. Rather, the middle group is the group whose members' bids for advancement are debatable.

There is no need to sift through Bennun's credentials in any detail in connection with his prima facie case. The divergent votes of the various review groups show the sufficiency of Bennun's scholarly and academic credentials for attainment of the rank of full professor was debatable among the peers who evaluated him. See Molthan, 778 F.2d at 962 (quoting Zahorik, 729 F.2d at 94). Accordingly, the district court correctly concluded that Bennun's showing was sufficient to meet the qualification prong of his required prima facie case.

3.

The presence of the final requirements of Bennun's prima facie case are apparent from the record. Despite his qualifications for the position of Full Professor, Bennun was rejected, and Full Professorships were available at the time he was refused promotion. See McDonnell Douglas Corp., 411 U.S. at 802, 93 S.Ct. at 1824; Roebuck, 852 F.2d at 715 (quoting Banerjee, 648 F.2d at 62). The district court then sought to determine if he had proved that non-minorities were treated more favorably than he was, the third prong of his prima facie case. In doing so, the district court compared Bennun's file to that of Dr. Ethel Somberg, a member of the same department as Bennun. Somberg was promoted to full professor in 1978-79. Id. at 1404. Her candidacy received support at all levels. The department rated her outstanding in teaching, scholarly activity and general usefulness and above average in research and professional activity. The Appointments and Promotions Committee gave Somberg the same marks that the department did except that it rated her above average to average in research. The Dean rated her outstanding in general usefulness, above average to outstanding in teaching, above average in scholarly activity, average to above average in research and average in professional activity. The Dean thought that her research was considerable after taking account of her teaching load. He also discounted the fact that some of her supporting letters were old and that few spoke to her overall contribution to her discipline. Id. Somberg had a total of seventeen publications. Only six of them were in refereed journals. She had no invited presentations and no outside letters from internationally known persons working in her field. Id. at 1404-05. All of this amply shows that Somberg was treated more favorably then Bennun. The district court did not err when it concluded Bennun had met his burden of making out a prima facie case of discrimination under Title VII.

C.

Rutgers insists that the district court did not properly consider the non-discriminatory reason it says prevented Bennun's promotion. By not paying sufficient attention to Bennun's teaching effectiveness, general usefulness and professional activity, Rutgers argues that the court ignored an objective non-discriminatory basis for promoting Bennun. The district court concedes that the University met its production burden at the second step of the McDonnell-Douglas-Burdine analytical framework by articulating Bennun's asserted lack of qualification as a non-discriminatory reason for not promoting Bennun to full professor. After reviewing all of the packets, it held that Rutgers' assertion that Bennun's research was "uncreative, moderate in quantity and insufficiently recognized in his field", constituted the proffer of a legitimate non-discriminatory reason for denying him tenure. Id. at 1407. Under Rutgers' criteria, Bennun had to be rated outstanding in one area, and short of that, he could not be promoted.

The district court, having decided that Rutgers had shown a nondiscriminatory reason for denying Bennun tenure, namely that Bennun was not promoted because the University's evaluators thought his research was deficient, see id., went on to consider pretext. On appeal, Rutgers further contends that the district court erred in not considering Bennun's lower marks in teaching effectiveness, general usefulness and professional activity in evaluating its reasons for not promoting Bennun. Because it would not alter our decision we will accept these reasons for purposes of analysis and gauge the district court's finding of pretext against them.

D.

The pretext issue is the dispositive one. The district court concluded that Bennun had carried his ultimate burden of showing pretext. The court was persuaded that the reason Rutgers assigned for denying Bennun's promotion, the poor quality and insufficient quantity of his research, was a pretext concealing the fact that Rutgers had applied a different standard to Bennun than it had to Somberg who was a member of a favored class.

1.

The reader will recall that the record shows the Appointment and Promotions Committee concluded that Somberg's research was acceptable in terms of both quantity and quality. The district court pointed out that Bennun had published an appreciably higher number of articles and that he had received favorable reviews from internationally known scholars. Yet the quantity of Bennun's research was rated as "moderate," somewhat lower than Somberg's "excellent." See id. at 1405. This discrepancy helped persuade the district court that Rutgers' proffered reason for denying Bennun promotion to full professor was pretextual. Accordingly it held that Rutgers had violated Title VII by discriminating against Bennun. Id. at 1409. Rutgers says that the district court erred in comparing Bennun to Somberg because they were not similarly situated and therefore, like the proverbial apples and oranges, cannot be compared. Rutgers notes that Somberg was promoted because she was rated outstanding in both teaching effectiveness and general usefulness while Bennun was not. For example, while Somberg had been rated uniformly higher than Bennun in teaching, the district court disregarded this because Bennun had taught a similar number of courses. The court pointed out that a single-lecture peer evaluation demonstrated no material difference between the two of them. Id. at 1406 n. 14. Rutgers contends that its comparison was misleading. Rutgers also notes that the packets of Somberg and Bennun the district court used for comparison were from different years. Thus the composition of some of their reviewing committees may have been different. In essence, Rutgers contends that Somberg was an outstanding teacher and did not need to be an outstanding researcher to be promoted. It says that Bennun conversely was not outstanding in either. However, because Bennun's main emphasis was on research, Rutgers concedes that his packet was evaluated in that light and this explains the differences between the evaluation of their research undermines any comparison between them.

Bennun takes issue with Rutgers' definition of "similarly situated." To define it as narrowly as Rutgers does, he says, would effectively isolate a university's promotion and tenure decisions from meaningful judicial review. Additionally, he notes that Rutgers' own policy encourages the comparison of colleagues having the same or similar duties. Bennun notes that the key factors for comparison are a candidate's length of service and his accomplishments in his particular field. Because both of these were similar, a comparison was appropriate.

2.

Before considering the pretext issue directly we must first confront the sub-issue of whether the district court's comparison of Bennun and Somberg's packets was proper. The district court based its decision that Bennun and Somberg were similarly situated on the fact that both Bennun and Somberg are biochemistry professors in the Zoology and Physiology Department at Rutgers' Newark campus. See Bennun, 737 F. Supp. at 1404. Rutgers contends that they were not similarly situated because Somberg was rated outstanding in two categories, teaching effectiveness and general usefulness, and Bennun was not rated as highly in these categories.

We cannot accept Rutgers' position. It would change "similarly situated" to "identically situated." In Franklin Marshall we allowed the Commission to subpoena a college for information on all persons considered for tenure over approximately a five-year period. See Franklin Marshall, 775 F.2d at 112-17. Although Franklin Marshall involved a determination of the relevancy of such information, see id. at 116, the broad sweep of relevancy we used to decide that case implies that two professors in the same specialty who are evaluated within two years of each other by the same college can sometimes have their credentials compared. The propriety of such comparison is case-specific. We must therefore determine whether comparison between Somberg and Bennun is appropriate here. We hold that even if Somberg was a "teaching-oriented" professor and Bennun was a "research-oriented" professor, as Rutgers contends, a comparison between the two can be made to determine if Rutgers' five objective criteria for advancement to full professor were evenly applied.

Though the district court focused only on the disparity between the standards applied to Bennun and Somberg with regard to their research qualifications, Rutgers' disparate treatment of Somberg and Bennun by itself is enough to preclude a holding that the district court's rejection of Rutgers' reason for not promoting Bennun was clear error. The district court simply found Rutgers' reason "unworthy of credence." Burdine, 450 U.S. at 256, 101 S.Ct. at 1095 (citing McDonnell Douglas Corp., 411 U.S. at 804-05, 93 S.Ct. at 1825-26). The district court, not this Court, finds facts and assesses credibility. Like any factfinder, it can accept some parts of a party's evidence and reject others. It may also, like any factfinder, assess credibility in light of the maxim, falsus in uno, falsus in omnibus. See Black's Law Dictionary 543 (5th ed. 1979) (defined as "false in one thing, false in everything"). We are unable to say, as a matter of law that it erred in its determinations.

To demonstrate the lack of clear error in the fact findings that underlie the ruling on pretext, we turn again to the record. One of the more glaring examples of disparate treatment was the Appointments and Promotions Committee's conclusion that Somberg's publications were "above average in quantity," Jt.App. at 1829, while Bennun's publication rate was questionable. Somberg had only seventeen publications of any type since becoming an Associate Professor in 1964, while Bennun had thirty-five since joining the faculty at Rutgers in 1969. Over a fifteen year period (1964-79) Somberg had eighteen fewer publications than Bennun did over a twelve year period (1969-81). These facts indicate that it is not an unwarranted invasion of the college's tenure process to determine that Bennun was held to higher standards in objective terms.

The difference in the treatment of the two candidates is even sharper in the section's ratings. The section rates candidates only on their research and scholarly activity. Yet the section wrote that Somberg's "recent publications [were] evidence of continuing research activity," Jt.App. at 1832, while Bennun was downgraded because his "total output [was] insufficient," Jt.App. at 1256. This conclusion is surprising because Somberg published only three pieces over the three years prior to her promotion (1976-79) while Bennun published eight pieces in the three years (1978-81) prior to his promotion denial in 1980-81. The section recommended Somberg for promotion despite the fact that her recent career production was inferior to Bennun's. This is another clear example of the disparate treatment that Bennun was subjected to.

Rutgers' argument that the Somberg and Bennun decisions are unassailable when all five criteria are considered together also fails to withstand analysis. Rutgers contends that Bennun was not rated outstanding in any category in the later stages of review and therefore could not have been promoted under any circumstances. But if throughout the promotion process the same standards applied to Somberg had been applied to Bennun, an unbiased observer might reasonably conclude that Bennun should have gotten at least one outstanding rating. If so, the district court, as factfinder, did not clearly err by drawing the conclusion that differing standards were applied to Bennun and Somberg, to Bennun's detriment, when the Appointments and Promotions Committee rated Bennun average in research accomplishments and rated Somberg above average.

Of course, we recognize the difference between Somberg, primarily a teacher, and Bennun, primarily a researcher. Still, whatever particular areas of strength or weakness one or the other promotion candidate has, an evaluation based on objective criteria requires all those criteria to be objectively applied. If that is done then one candidate's strength will offset his weakness in comparison with another's strength and the converse will permit objective comparison. Here, Bennun's strengths were not given the same overall weight as were Somberg's. In one of Bennun's stronger areas, research accomplishments, he was held to a higher standard than Somberg. Rutgers' articulated nondiscriminatory reliance on the five criteria, taken as a whole, in deciding not to promote Bennun is so undermined by its inconsistency in applying them that we cannot hold the district court's finding was clearly erroneous. The district court did not commit clear error in finding Rutgers's assigned reason for denying Bennun the rank of full professor was a pretext. Therefore, the district court did not err in holding Rutgers is liable to Bennun for unlawful discrimination in employment under Title VII.

VII.

Because of the many issues in this case we close with a summary of our holdings. Bennun's claims are properly before us. We believe that New Jersey's courts would not apply the entire controversy doctrine to bar Bennun's civil rights claims. To so apply it would be anomalous: a doctrine driven by a goal of judicial efficiency would be used to undercut the use of non-judicial methods of dispute resolution. On the merits, we hold that Bennun cannot prevail on his § 1981 claim because the promotion of a tenured associate professor to full professor at Rutgers does not create the new and distinct employment relationship needed to meet Patterson's holding that a § 1981 plaintiff must show he was discriminated against in the formation of a contract.

We also hold that the district court did not err in determining that Bennun successfully made out a claim of discrimination under Title VII. The differences of opinion about his qualifications among the various reviewing groups were sufficient to establish that his fitness for promotion was debateable. Because Rutgers did not challenge the other portions of Bennun's prima facie showing, resolution of this requirement in Bennun's favor completes his prima facie case. Rutgers articulated Bennun's lack of qualifications as its non-discriminatory reason for denying his promotion to Full Professorship. Bennun has carried his ultimate burden of persuading the district court, as factfinder, that Rutgers' proffered reasons were mere pretexts. In particular, Rutgers' disparate treatment of Bennun's research accomplishments bear this out. Since Bennun produced enough evidence of pretext to make out a prima facie case of discrimination, Rutgers is liable to him under Title VII. We will therefore affirm that part of the district court's order granting judgment for Bennun on his Title VII claim. We will reverse that part of the district court's order granting judgment for Bennun on his § 1981 claim. Finally, because all of the relief that Bennun received on both claims flowed from Rutgers' liability on Bennun's Title VII claim alone, we will not disturb the district court's remedial order. Costs will be taxed against appellant.

HUTCHINSON, Circuit Judge.

SUR PETITION FOR REHEARING

HUTCHINSON, Circuit Judge.

The petition for rehearing filed by appellants in the above captioned matter having been submitted to the judges who participated in the decision of this court and to all the other available circuit judges of the circuit in regular active service, and no judge who concurred in the decision having asked for rehearing, and a majority of the circuit judges of the circuit in regular active service not having voted for rehearing by the court in banc, the petition for rehearing is denied.

Chief Judge Sloviter would grant in banc rehearing for the reasons set forth in her attached Statement Sur Denial of Rehearing In Banc.

Judge Roth would grant in banc rehearing and joins in Chief Judge Sloviter's attached Statement.