Opinion
No. 88-1647.
Heard January 9, 1989.
Decided May 19, 1989.
Nancy Gertner with whom Silverglate, Gertner, Fine Good, Boston, Mass., was on brief, for appellant.
Peter J. McGinn with whom Christopher H. Little, Tillinghast, Collins Graham, and Beverly Ledbetter, Gen. Counsel, Providence, R.I., were on brief, for defendant, appellee Brown University.
Appeal from the United States District Court for the District of Rhode Island.
Before CAMPBELL, Chief Judge, COFFIN and BREYER, Circuit Judges.
The appellant, Ann Seidman, claims that Brown University discriminated against her on the basis of sex. Her claim is based on the following events, which occurred about ten years ago: (1) Brown opened a search for candidates for the Henry R. Luce Professorship of The Comparative Study of Development; (2) it narrowed the field to three candidates, of whom Seidman was the third choice; (3) it offered the job to the first two choices, both men; and (4) after the two men turned down Brown's offer, it did not offer Seidman the job, but reopened the search instead.
Seidman's factual case rests upon detailed evidentiary arguments designed to show that Brown has no good answer to the question: "Why, if not for sex-based reasons, did you reopen the search?" Seidman's legal case rests upon the standards set forth in a consent decree that Brown had signed, a decree that says, in relevant part, that if Brown does not fill a tenured position, such as the Luce Professorship, "after the prescribed search procedure," Brown must "demonstrate by clear and convincing evidence that the decision not to fill the position was nondiscriminatory as to sex." See Lamphere v. Brown University, 491 F. Supp. 232, 240 (D.R.I. 1980) (reprinting consent decree), aff'd, 685 F.2d 743 (1st Cir. 1982). We have held that the words "demonstrate by clear and convincing evidence," in this context, require Brown:
1) to articulate a legitimate, non-discriminatory reason for its actions, and
2) to show by clear and convincing evidence that this reason is not a pretext.
Lamphere v. Brown University, 798 F.2d 532, 536 (1st Cir. 1986) ( "Seidman I").
Seidman argued her case in district court and lost. The district court thought that Brown had not discriminated on grounds of sex; rather, it had reopened the search for legitimate reasons. Seidman appealed. We read the entire 2200 page record. We decided the various legal points she raised, and we concluded that the district court should reevaluate the evidence, but only in respect to one particular factual matter. Seidman I, 798 F.2d at 542-43 (remanding case for district court to decide whether Brown's reasons for reopening the search were pretextual). The district court did so, and it reached the same conclusion, 690 F. Supp. 125 (1988). Seidman appeals again. We have reviewed the district court's reevaluation, limiting ourselves to consideration of the specific question we previously left open. We conclude that the district court's redetermination was not legally erroneous, and so we must affirm its judgment.
I.
To understand this appeal, one must read the facts set forth in Seidman I. We shall not repeat them here. Rather, we shall simply set forth an outline of several key facts and events, as we have gathered them from the record:
1. Brown meant the Henry R. Luce Professorship to be a distinguished position, carrying with it a tenured place at the University. Brown also hoped that its holder would help to unite, to some degree, the disciplines of sociology and economics at Brown. It created a Search Committee that included three members of the Sociology Department, one member of the Economics Department, and two alternates, one from each department.
2. On December 18, 1978, the Search Committee made an initial recommendation. It recommended that Brown first offer the position to Professor Bardhan, then, should he turn it down, to Professor Streeten. If Professor Streeten also declined the offer, the Committee recommended that Brown offer the job to Professor Seidman, but only as a three to five year, nontenured appointment.
3. University officials told the Committee that Brown would not offer the position on a nontenured basis; its holder must be given tenure.
4. Professor Bardhan turned down the position.
5. As of January 22, 1979, Professor Rueschemeyer, the Search Committee Chair, thought he had the Committee's agreement to recommend a tenured offer to Seidman (if Streeten turned down the job). He conveyed this impression to the Sociology Department. But at least one other member of the Committee thought there was no such agreement.
6. On January 22, 1979, the Sociology Department members met and decided to recommend Streeten for the job. They also decided to offer the job to Seidman if Streeten turned it down.
7. Streeten declined Brown's offer.
8. On March 15, 1979, members of the Search Committee met with the University Provost. They explained that the Department had not fully understood the position of the Search Committee; that the formal report of the Department's position, made to the Provost, incorrectly said that the Search Committee's recommendation of Seidman had been unanimous; and that the Department had never formally voted on Seidman's candidacy. The Provost then told the Department to reconsider and vote formally.
9. On March 16, 1979, the Sociology Department voted, 7-2, not to offer the job to Seidman, but to reopen the search instead.
In our previous opinion, we decided that the evidence in the record showed no improper motive on the part of the Provost, when, on March 15, he sent the recommendation back to the Sociology Department for reconsideration and for a formal vote. Seidman I, 798 F.2d at 542. We remanded the case to the district court, however, so that it could reconsider the key question of why the Sociology Department changed its mind about Seidman. Why, after recommending her for the position on January 22, did it not do so on March 16? Did Seidman's gender play a role in this change of mind? Or, more precisely, did Brown prove that this impermissible factor did not play a role in its decision? We wrote:
[W]e are not now willing to decide — without an initial district court decision — the key factual question about the Sociology Department's decision on March 16 to reopen the search. Brown has presented witnesses, testimony, and documents offering a legitimate, nondiscriminatory explanation of this event. Professor Seidman has presented evidence of her own, including one witness, Professor Evans, who testified that Professor Marsh opposed Seidman because of "attitudes toward women, with the sex of the candidate." The district court, which has heard the witnesses and can better evaluate their credibility, should decide in the first instance whether or not Brown has shown that its explanation is not a pretext. In doing so, the court should decide whether Brown has articulated legitimate reasons for reopening the search, and has proved clearly and convincingly that those reasons are the true ones. At the same time, it must take care not to interpret the decree in a way that would impose an unreasonably difficult burden of proof upon Brown. See supra, pp. 536-37. The district court may decide this case on the existing record should it believe it appropriate to do so.
Id., 798 F.2d at 542-43. On remand, the district court decided, on the basis of the preexisting evidentiary record, that Brown had articulated legitimate, nondiscriminatory reasons for reopening the search, and that Brown had proved by "clear and convincing evidence" that those reasons were the true ones. Seidman now appeals, arguing that these findings are legally incorrect.
II.
Before turning to Seidman's evidentiary arguments, we shall consider three more purely legal arguments that she raises. First, she notes that one factor in the Sociology Department's decision was the fact that she, unlike the two male candidates, had not previously held a tenured position at a major university. She argues that: (1) Brown cannot lawfully rely upon such a reason, because prior tenure, though apparently a neutral hiring criterion, is not really neutral at all, since universities, in the past, were far more likely to have given tenure to men than to women. Thus, to use prior tenure as a factor in hiring will have a "disparate impact" upon women. See Watson v. Fort Worth Bank Trust, ___ U.S. ___, 108 S.Ct. 2777, 2786, 101 L.Ed.2d 827 (1988) (disparate impact analysis may be applied in discrimination cases involving subjective decisionmaking). (2) The law requires an employer to show that a hiring criterion which has a disparate impact bears a "manifest relationship to the employment in question." Dothard v. Rawlinson, 433 U.S. 321, 329, 97 S.Ct. 2720, 2727, 53 L.Ed.2d 786 (1977); see also Price Waterhouse v. Hopkins, ___ U.S. ___, 109 S.Ct. 1775, 104 L.Ed.2d 268 (1989) (employer may not condition employment on facially neutral qualifications that have a disparate impact "when those . . . qualifications are not required for performance of the job"). If the employer makes this showing, it may lawfully use such a factor in hiring, unless the employee can show that "other selection devices without a similar discriminatory effect would also 'serve the employer's legitimate interest.'" Dothard, 433 U.S. at 329, 97 S.Ct. at 2727. (3) In this case, however, the consent decree, by imposing on Brown the burden of proving that its decision was nondiscriminatory, requires Brown, the employer, to prove "by clear and convincing evidence" both that prior tenure bears a "manifest relationship" to the job in question, and that there is no other, equally effective "selection device" that it could use instead. (4) Brown has not shown, by "clear and convincing evidence," that prior tenure bears a "manifest relationship" to qualification for the Luce professorship; and it certainly has not shown that it could not instead rely exclusively on other criteria (such as publications, teaching evaluations, etc.). (5) Brown, therefore, did not "clearly and convincingly" show that this factor in its decision was non-discriminatory.
Our problem with this argument lies in steps 3 and 4. We believe, in respect to the "manifest relationship" of prior tenure to the position, that Seidman's argument fails at step 4. That is to say, the fact that other major universities have given a person tenure would seem obviously related to that person's fitness for a tenured position at Brown. It does not prove that the person is qualified, but it does indicate similar experience, and it suggests that the person's colleagues at another university have thought her qualified for a similar level of appointment. The record here makes obvious, as is commonly known, that university faculty members consider prior tenure as one among several key indications that a person is fit for similar or more advanced work. The relationship of prior tenure to qualification for the Luce professorship, in our view, is "manifest."
We believe, in respect to the availability of an alternative "selection device," that Seidman's argument fails at step 3. We cannot read the consent decree as imposing upon Brown the burden of proving the nonexistence of any alternative to the use of prior tenure in the academic hiring process. The decree focuses on modifying the "basic" steps of an anti-discrimination lawsuit, requiring Brown to show the absence of "pretext," rather than requiring the employee to prove the presence of pretext, as in the usual Title VII case. See Texas Department of Community Affairs v. Burdine, 450 U.S. 248, 252-53, 101 S.Ct. 1089, 10934-94, 67 L.Ed.2d 207 (1981) (under Title VII, if the employer articulates a neutral reason for its actions, the employee must then show, by a preponderance of the evidence, that the asserted reason is a pretext for discrimination). Neither the language of the decree, nor any discussion of it in the relevant judicial opinions, suggests that it was meant to alter the academic hiring process itself, or so significant a part of it as the consideration of prior tenure. The burden of proving that a university could not develop a workable "tenure-blind" selection system would be formidable. In addition to the general difficulty of "proving a negative," see Seidman I, 798 F.2d at 536, the highly subjective nature of the academic appointment process would make it especially hard to show, clearly and convincingly, that no tenure-blind selection process would work equally well, i.e., would result in the selection of faculty members of equal quality. See id. at 537 (consent decree's shifting of burden of proof is permissible remedy "so long as the rules are applied sensibly so as not to impose an unreasonably difficult burden [of proof] on Brown"). If the authors of the decree had intended the decree to impose so formidable a burden of proof, or, indirectly, to have so revolutionary an effect on the academic hiring process, someone would have mentioned it.
It is true that the broad language of the decree might be read literally, as imposing on Brown the burden of proof, by clear and convincing evidence, as to any and all issues that might arise in any future discrimination case — including, here, the burden of justifying the use of prior tenure as a factor in hiring. But, a consent decree must be interpreted in much the same way as a contract. See Local Number 93, International Association of Firefighters v. City of Cleveland, 478 U.S. 501, 519, 106 S.Ct. 3063, 3074, 92 L.Ed.2d 405 (1986) (although consent decrees are treated like judgments for some purposes, "because their terms are arrived at through mutual agreement of the parties, consent decrees also closely resemble contracts"); United States v. ITT Continental Baking Co., 420 U.S. 223, 236-37, 95 S.Ct. 926, 934-35, 43 L.Ed.2d 148 (1975) (since consent decrees "have many of the attributes of ordinary contracts, they should be construed basically as contracts"); Arrieta-Gimenez v. Arrieta-Negron, 859 F.2d 1033, 1041 (1st Cir. 1988); AMF Inc. v. Jewett, 711 F.2d 1096, 1101 (1st Cir. 1983). Its general language must be read as limited by its purpose and context. See Massachusetts Association for Retarded Citizens v. King, 668 F.2d 602, 607 (1st Cir. 1981) (the interpretation of consent decrees, like that of contracts, "depends in part upon language, in part upon the circumstances surrounding their formation, and in part upon the basic purposes of the decree"); see also Waters v. Massey-Ferguson, Inc., 775 F.2d 587, 592 (4th Cir. 1985); 4 Williston on Contracts § 610B (3d ed. 1961). And, the context surrounding this decree suggests that the burden of proof rules were designed to compel Brown to prove, clearly and convincingly, absence of pretext, but not to prove the absence of any conceivable alternative to the use of prior tenure in hiring. Since we conclude that the consent decree does not shift the burden of proof as to "other selection devices," it was Seidman's burden to explain how selection devices apart from prior tenure could adequately serve Brown's interests, and she did not do so. The district court did not err in considering Seidman's lack of prior tenure as one of Brown's legitimate, nondiscriminatory reasons.
Second, Seidman points to a memorandum prepared by Professor Speare, the chair of the Sociology Department (who became the new chair of the Search Committee), in which Speare listed the reasons why the Sociology Department, at its meeting of March 16, decided to reopen the search. That memorandum (which we have printed as an appendix) sets forth five "major reasons given by faculty members for not offering the Luce Professorship to Ann Seidman on a permanent basis at this time." Those five reasons are:
1) "Seidman would not provide the type of bridge between sociology and economics which was envisioned" for the Luce chair;
2) Seidman's "methodology" would not make a "distinctive contribution" to the department;
3) Seidman's appointment "would not provide the desired level of recognition for Brown" in the field;
4) Seidman was "not distinguished enough for this appointment because she had not previously held a tenured appointment at a major university;"
5) To reopen the search "would attract several new candidates."
The district court concluded that the "essential uncontroverted reason" for the department's actions "was the failure of Ann Seidman to obtain the approval of the Selection Committee" and that a "subsidiary basis" for the Department's actions was that Seidman "does not meet the unique requirements of a prestigious chair . . . in terms of accomplishment or peer recognition." Seidman argues that the district court's interpretation of Brown's reasons is unsupported by the record, because it conflicts with the Speare memorandum. She also says it was unfair for the district court to find that Brown had nondiscriminatory reasons other than those articulated in the Speare memorandum, since Seidman had focused her efforts to prove pretext on those particular reasons.
We cannot accept Seidman's arguments. For one thing, we see no serious inconsistency between the district court's statements and the Speare memorandum. The memorandum, before listing the five reasons, noted that "there never was unanimous support for Ann Seidman's nomination. . . . the search committee was divided." And, the district court's description of the "subsidiary basis" for reopening the search sounds somewhat like a summary of the five specific reasons that the Speare memorandum sets forth. The differences are not significant enough to have unfairly surprised Seidman, or deprived her of a "'full and fair opportunity' to demonstrate pretext." See Burdine, 450 U.S. at 258, 101 S.Ct. at 1096.
For another thing, the Speare memorandum did not, in any relevant sense, "bind" either Brown or the district court. It purports only to be an effort by one faculty member to summarize the views of his colleagues. As such, it is one piece of evidence (though an important one) among many pieces of evidence that Brown introduced to show the various reasons motivating the various members of the Sociology Department. The memorandum rests on the same footing as other, equally competent evidence on this point, such as the testimony of Department members about their concerns and their reasons for voting as they did at the March 16 meeting.
The best one can say for Seidman is that the Speare memorandum, when examined in light of the Department members' testimony and other evidence, suggests that different individuals may have had different reasons, sometimes conflicting reasons, sometimes changing reasons, for voting to reopen the search. But life, unlike law, does not always present its reasons in neat packages. The issue is not whether the hiring committee (composed of nonlawyers) presented, at the time, a single, clear set of reasons for its decisions. The issue is, instead, whether Brown can show that the committee's actions had nothing to do with sex discrimination. See Price Waterhouse, ___ U.S. ___, 109 S.Ct. 1775, 104 L.Ed.2d 268 (Title VII requires that "gender must be irrelevant to employment decisions"). Brown must clearly and convincingly explain what happened, but its explanation may involve complex, shifting, and sometimes contradictory individual motivations. The essential question is whether the series of events as a whole, the story that they tell — muddled and conflicting as it may be — was inconsistent with sex discrimination. Of course, changing reasons, failure to articulate reasons contemporaneously, conflicts among individuals' reasons, etc. might suggest a subterfuge concealing sex discrimination. But, such conflict and confusion does not necessarily mean that Brown failed to prove that there was no discrimination. It might, instead, simply reflect a group decision-making process at work in the highly subjective area of faculty hiring. In Banerjee v. Board of Trustees of Smith College, 648 F.2d 61 (1st Cir.), cert. denied, 454 U.S. 1098, 102 S.Ct. 671, 70 L.Ed.2d 639 (1981), for example, we upheld a finding that a denial of tenure was not discriminatory, even though the district court found that "no single reason or set of reasons can be provided as to why the Committee voted as it did." Id. at 63. We noted that the "articulation of reasons for refusing tenure by . . . collegial decision-making apparatus" may lack "clarity," but such clarity is not legally required; all members of the hiring committee need not "agree on a single reason, except in an ultimate conclusion; the requirement is that they should not reach this [conclusion] by improper [reasons]." Id. at 64. Brown need not prove that the Department members were perfectly consistent, rational, or in some objective sense "correct," but only that their actions were free from sex discrimination. See Burdine, 450 U.S. at 258-59, 101 S.Ct. at 1096-97 (defendant need not show, by "objective evidence," that it hired the most qualified candidate); Kumar v. Board of Trustees, University of Mass., 774 F.2d 1, 11 (1st Cir. 1985), (academic hiring may involve subjective, discretionary concerns), cert. denied, 475 U.S. 1097, 106 S.Ct. 1496, 89 L.Ed.2d 896 (1986); Riley v. University of Lowell, 651 F.2d 822, 824 (1st Cir.) ("subjective discretionary elements" are "traditional" in academic hiring), cert. denied, 454 U.S. 1125, 102 S.Ct. 974, 71 L.Ed.2d 112 (1981). The district court did not err; its identification of Brown's reasons was supported by the record, as was its conclusion that some combination of those neutral reasons was the true motivation for the Department members' votes at the March 16 meeting. See Part III, infra.
Third, Seidman says that the district court should have relied more heavily upon Brown's internal investigation of the relevant events, conducted by Brown's Affirmative Action Monitoring Committee. In our prior opinion, we said that the district court was "free to read, and to benefit from, the Monitoring Committee's opinions," but we also said that the district court was to try the matter "de novo." Seidman I, 798 F.2d at 538. We read the Monitoring Committee's report when this case was previously before us. We do not believe that the Committee's presentation of the facts, or its legal reasoning, is sufficiently compelling as to require the district court to reach the same conclusions as the Committee did.
III.
We turn now to Seidman's main, evidentiary argument, that the record does not support the district court's conclusion that Brown clearly and convincingly proved the absence of discrimination. At this stage in the proceedings Seidman is bound to have difficulty proving her point. For one thing, the relevant legal standard favors Brown. We can overturn the district court's factual findings only if we have "the definite and firm conviction that a mistake has been committed." Fortin v. Commissioner of Massachusetts Dep't of Public Welfare, 692 F.2d 790, 794 (1st Cir. 1982); see also Kumar, 774 F.2d at 9 ("we are not to set aside a finding of fact, even a Title VII case finding involving mixed questions of law and fact, unless it was 'clearly erroneous'"). For another thing, if we had not previously believed that the district court, on the basis of the record, might have reached a conclusion in Brown's favor, we would not have remanded the case as we did in Seidman I. We believed that the district court could reasonably believe either such witnesses as Professor Evans, who thought sex discrimination was involved, or such witnesses as Professor Anderson, who thought there was no such discrimination, and we thought that the outcome should turn on the court's weighing of the evidence and its assessment of the credibility of the witnesses. Although we could stop here, and simply say that we have no "firm conviction" that the district court made a "mistake" in finding that Brown had proved no discrimination, we shall go on briefly to indicate, in rough outline form, why we think the evidence was adequate.
There are two suspicious circumstances on which Seidman rests her basic case. First, just why did the Sociology Department change its mind about Seidman's candidacy between the January 22 meeting, when it decided to offer her the job, and the March 16 meeting, when it decided to reopen the search? Did it change its mind because it suddenly became apparent that Seidman, a woman, might actually obtain the job? Or, did it change its mind for legitimate, nondiscriminatory reasons? Second, leaving aside the change of mind, why did it suddenly reopen the search once it became apparent that, otherwise, Seidman might obtain the job? Again, were its reasons legitimate or discriminatory?
In each instance, Brown came forward with a legitimate, nondiscriminatory explanation. And, in each instance, Brown offered enough testimony and other evidence for the district court to have concluded that Brown had clearly and convincingly shown that its profered answer was the true one. The explanation of the "change of mind" lies in the intricate politics of the Search Committee and the Department. Evidently the Committee's chair, Professor Rueschemeyer, and one of its members, Professor Evans, wanted Professor Seidman to have the job. At the January meeting Professor Rueschemeyer indicated to the Department faculty that the "Selection Committee was prepared to accept the decision of the Sociology Department," a form of words that, in context, suggested that the Committee as a whole (not just Rueschemeyer and Evans) favored Seidman's appointment. The district court might have found, however, that Rueschemeyer did not fully and clearly inform the faculty about the differences of opinion within the Committee, as to Seidman's qualifications. At the March 16 meeting, the faculty learned more about the details of the Search Committee's views. They, or at least some of them, decided that the January consensus in Seidman's favor rested on an incorrect understanding of those views. Also, Professor Anderson testified that some faculty members (or at least she herself) had not realized in January that "the faculty was supposed to 'exercise their full professional judgment,'" and not just decide whether to accept the Search Committee's recommendation, and that, by March 16, they became aware of their full responsibility. A memo written by Professor Evans on March 22, 1979 also supports this point. Finally, different faculty members, with different views, were present at the March meeting than at the January meeting. In our view, the record permits the district court to find that these neutral, nondiscriminatory factors clearly and convincingly explain the Department's change of mind.
Leaving aside the Department's change of mind, why did Brown reopen the search? What explains the fact that, when it reached the third person on the list, a woman, Brown decided to reopen the search? Brown's response concerns the Department's evaluation of Seidman's credentials. Brown says the Department concluded that Seidman was significantly less qualified than the two men; and, when it realized that the appointment carried both a title and tenure, it decided that it should reopen the search rather than appoint Seidman. The Speare memorandum lists five non-discriminatory reasons — elaborations on why Seidman's qualifications were thought inadequate — which, Brown says, motivated the Department's decision. Seidman attacks each of them.
First, Seidman says that the Department's reliance on the fact that she had not previously held tenure at a major university is, in itself, discriminatory. We have discussed, and rejected, this contention at pp. 919-921, supra.
Second, pointing to evidence in the record, Seidman claims that the reason that she "would not provide the type of bridge between sociology and economics which was envisioned" is but a "pretext" to hide discrimination. The record, however, contains evidence that the Economics Department preferred both Bardhan (to whom it was willing to offer a tenured joint appointment) and Streeten (with whom, its representative said, its members would be willing to work) to Seidman. The Speare memorandum, consistently with this evidence, says that "members of the Economics Department had indicated that they could work with Streeten, but not with Seidman." The district court could reasonably conclude that Brown had clearly and convincingly shown that some members of the Sociology Department believed just what the Speare memorandum said.
Similarly, the Speare memorandum says that members of the Department believed that Seidman's "overlapping methodology" was a negative factor ( i.e., that Seidman's approach to sociology did not differ enough from that of other Department members to offer the Department methodological diversity), as was her "inadequate professional recognition." The district court concluded that Brown had shown that the Department members did in fact hold these beliefs, and decided against Seidman partly because of them. Seidman seeks to reargue whether or not diversity of methodology ought to have played an important role in selection, and whether or not her professional recognition was as inadequate as the Speare memorandum suggests, but she fails to convince us that the district court could not reasonably have found that those considerations were neutral factors that truly did motivate the Department members' votes.
Seidman points to the Speare memorandum's statement that "it is also significant that we received several unsolicited letters suggesting Bardhan and Streeten for the position while few, if any, were received which mentioned Seidman." She points out that this statement is false: The Department received unsolicited letters about Streeten, but not about Bardhan. Bardhan's qualifications, however, were obviously superior to Seidman's; it was Streeten versus Seidman that presented the closer question. And, the memo correctly states that the Department received unsolicited letters in respect to Streeten, but not Seidman. Thus, the inaccuracy does not seem important enough to discredit the memo significantly.
Seidman also points to the memo's statement that "it was the consensus of those present that if we reopened the search, with a longer period of advertising, we would attract several new candidates since the pool of people interested in such positions is constantly changing." She notes that earlier, a different Brown committee, and Professors Speare and Reuschemeyer, had both suggested the opposite. The question, however, is what those at the Sociology Department meeting on March 16 believed. The faculty members present at that meeting overlapped only partially with the two groups of people who had said, on prior occasions, that a new search would not attract a larger pool. The district court could reasonably have thought that this fact, and not any bad-faith change of position, accounted for the apparent contradiction. The question whether a new search would attract a larger pool of applicants was speculative enough to warrant a variety of opinions on the subject.
Seidman goes on to compare her credentials to those of Streeten, and to argue that the difference is not significant enough to explain the reopening of the search. The district court, however, heard testimony from many of those present at the March 16 meeting. It had the opportunity to evaluate the testimony of key witnesses on both sides. It could reasonably have concluded that Seidman was less well known than Streeten, that her recommendations were less enthusiastic, that she was less likely to win acceptance among the economists. It could have taken account of the fact that the Search Committee was willing to recommend Streeten for tenure, but was hesitant and divided in respect to Seidman. One might reargue the facts of this case endlessly, but a district judge has heard the evidence and he has determined that Brown proved its point. Our review of the record leaves us without any firm conviction that he was wrong, even taking full account of the special standard of proof. Because of this conclusion, and because we reject as unfounded the various other arguments that Seidman makes on this appeal, we must affirm the district court's decision.
The judgment of the district court is
Affirmed.