LEVAL, Circuit Judge:
The New School for Social Research (the “New School”) appeals the judgment of the United States District Court of the Southern District of New York (Haight, /.), entered pursuant to jury verdict, awarding damages to Dr. Adamantia Pollis, a retired professor of political science. The New School contests (i) the sufficiency of evidence in support of the jury’s finding of willfulness, with respect to its violation of the Equal Pay Act, 29 U.S.C. § 206(d), in paying Pollis less than comparable male faculty members; (ii) the court’s use of the doctrine of continuing violation to award damages for nineteen years of unequal pay in the face of a three-year limitations period, 29 U.S.C. § 255(a); and (iii) the sufficiency of the evidence to support the jury’s finding that the New School discriminated on the basis of gender in violation of Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e
et seq.
(“Title VII”), and § 296 of the New York Human Rights Law, N.Y. Exec. Law § 296, by failing to give Pollis full-time or more substantial part-time employment after she reached the mandatory retirement age of 70.
We affirm the jury’s finding that the New School’s violation of the Equal Pay Act was willful or reckless, but vacate the judgment and remand for recalculation of the award, which should have been limited to the amount of damages incurred within the limitations period. We reverse the award of damages for intentional gender discrimination.
Background
Pollis was hired as a professor of political science at the Graduate Faculty of the New School in 1964. She was granted tenure in 1966, and promoted to full professor in 1976. During her employment at the New School, she twice served as chair of the political science department. Her primary areas of specialty were human rights and Greek politics. According to evidence Pollis submitted at trial, her salary was lower than the salaries of five male teachers who were comparable to her.
The New School by-laws in effect between 1968 and 1994 provided that all full-time faculty members were required to retire at the end of the school year in which they turned 70. Under this rule, Pollis was required to retire at the end of the 1992/93 academic year.
In December 1992, Pollis requested that she be allowed to stay on in her tenured position' after June 1993. She also applied for a vacant full-time teaching position for comparative politics in the political science department. She testified at trial that she offered to forego tenure if she were given the position. Her requests were rejected. She was offered, instead, an adjunct position as senior lecturer, teaching two courses per year for $4,000 per course. She accepted this position under protest and filed suit alleging violations of Title VII, the New York Human Rights Law, and the Equal Pay Act.
After trial, the jury found in favor of Pollis on her' claim that the New School’s refusal to offer her a full-time or substantial part-time post-retirement position violated Title VII and the New York Human Rights Law. It also found in Pollis’s favor on her claim that the New School violated the Equal Pay Act by paying her less than comparable male faculty, and made the additional finding that the Equal Pay Act violation was committed willfully or with reckless disregard for Pol-lis’s rights. However, the jury found in favor of the New School on the claim that Pollis’s lesser rate of pay resulted from intentional gender discrimination. In a special verdict, the jury expressly found that Pollis’s lower
rate of pay was not motivated in any part by intent to discriminate on the basis of gender.
The district court awarded nineteen years of back pay under the Equal Pay Act, doubled because of the jury finding of willful or reckless violation. The court awarded compensatory and punitive damages for the refusal of more substantial post-retirement employment. Finally, the court awarded substantial attorneys’ fees, costs, and expenses.
Discussion
I.
Equal Pay Act Claim
Pollis showed at trial that, during a nineteen year period, her salary was less than that paid to five comparable male teachers of the New School. On that basis, the'jury found a violation of the Equal Pay Act.
The Equal Pay Act is violated if an employer whose employees are subject to the Fair Labor Standards Act pays wages to an employee
at a rate less than the rate at which he pays wages to employees of the opposite sex ... for equal work on jobs the performance of which requires equal skill, effort, and responsibility, and which are performed under similar- working conditions ....
29 U.S.C. § 206(d).
A violation occurs when an employer pays lower wages to an employee of one gender than to substantially equivalent employees of the opposite gender in similar circumstances. A plaintiff need not prove that the pay disparity was motivated by an intention to discriminate on the basis of gender.
See Tomka v. Seiler Corp.,
66 F.3d 1296, 1310 (2d Cir.1995). The New School does not contest the sufficiency of the evidence to support the finding of a violation.
A.
“Continuing Violation” Doctrine
Under 29 U.S.C. § 265(a), however, a claim under the Equal Pay Act must be commenced within two year’s of its accrual, or three years if the violation is willful. As the district court noted, back pay awards for Equal Pay Act violations are typically limited to damages sustained within this period.
See Ashley v. Boyle’s Famous Corned Beef Co.,
66 F.3d 164, 168 (8th Cir.1995) (en banc);
Brinkley-Obu v. Hughes Training, Inc.,
36 F.3d 336, 351 (4th Cir.1994);
Gandy v. Sullivan County,
24
F.3d 861, 865 (6th
Cir.1994).
The district court held the statutory period was not applicable because the New School’s payment of unequal wages constituted a “continuing violation.” The continuing violation doctrine allows a plaintiff in certain circumstances to recover on the basis of an ongoing policy or practice of illegal activity initiated prior to the limitations period.
See Hanover Shoe, Inc. v. United Shoe Mach. Corp.,
392 U.S. 481, 502 n. 15, 88 S.Ct. 2224, 2236 n. 15, 20 L.Ed.2d 1231 (1968).
In
Acha v. Beame, 570
F.2d 57, 65 (2d Cir.1978), we recognized that the “continuing violation” doctrine might in some instances justify a remedy for discrimination occurring prior to the limitations period. A group of female police officers, suing under Title VII, claimed that the use of discriminatory hiring-policies had denied them seniority and made them vulnerable to layoff. In dicta, we explained that the district court had the authority to grant “retroactive seniority” pre-dating the limitations period if the plaintiffs could prove that a discriminatory policy which had deprived them of seniority was still in effect within the limitations period and that the plaintiffs .continued to suffer as a result. “[Wjhere an illegal policy is so maintained,” we explained, “relief for injuries sustained even before the beginning of the limitations period is appropriate.” 570 F.2d at 65.
We believe that Acha’s reasoning is inapplicable to the present case. Seniority rights are not of immediate value, but have determinative effect on the future terms of one’s employment, such as work assignments or insulation from layoff. As a consequence, a discriminatory policy that results in a wrongful denial of seniority causes harm not so much at the moment of denial as in the future, and thus only by an award of retroactive seniority can future continuing injury be avoided. A discriminatory pay scale, in contrast, has immediate effect; prospective relief requires nothing more than discontinuance of the unlawful conduct.
Furthermore, a claim of discriminatory pay is fundamentally unlike other claims of ongoing discriminatory treatment because it involves a series of discrete, individual wrongs rather than a single and indivisible course of wrongful action. As the Supreme Court explained in
Bazemore v. Friday,
characterizing the harm imposed by a racially discriminatory pay scale, “Each week’s paycheck that delivers less to a [disadvantaged class member] than to a similarly situated [favored class member] is a wrong actionable under Title VII, regardless of the fact that this pattern was begun prior to the effective date” of limitation. 478 U.S. 385, 395-96, 106 S.Ct. 3000, 3006, 92 L.Ed.2d 315 (1986) (Brennan, J., concurring in part, joined by all other members of the Court).
Under this view, “each continuation or repetition of the wrongful conduct may be regarded as a separate cause of action for which suit must be brought within the period beginning with its occurrence.”
Developments in the Lam
— Statute
of Limitations,
63 Harv. L.Rev. 1177, 1205 (1950);
see also Klehr v.
AO.
Smith Corp.,
— U.S. -, -, 117 S.Ct. 1984, 1990, 138 L.Ed.2d 373 (1997). Thus, a cause of action based on receipt of a paycheck prior to the limitations period is untimely and recovery for pay differentials prior to the limitations period is barred irrespective of subsequent, similar timely violations.
See Knight v. Columbus,
19 F.3d 579, 582 (11th Cir.),
cert. denied,
513 U.S. 929, 115 S.Ct. 318, 130 L.Ed.2d 280 (1994);
see also
Douglas Laycock,
Continuing Violations, Disparate Impact in Compensation, and Other Title VII Issues,
Law & Contemp. Probs., Autumn 1986 at 53, 61.
With this holding, we join the other circuits to have considered the issue, which have unanimously adopted the reasoning of
Bazemore
and concluded that back pay cannot be recovered under the Equal Pay Act for salary differentials outside the limitations period. The Sixth Circuit, for example, has explained that a plaintiff who complains of willful violation of the Equal Pay Act over a ten-year period of employment is not barred from bringing suit but “should not be allowed to collect for damages outside the three-year limitations period.”
Gandy,
24 F.3d at 8651 The Eighth Circuit, reaching the same conclusion, has reasoned that this approach “strikes a reasonable balance between permitting redress of an ongoing wrong and imposing liability for conduct long past.”
Ashley,
66 F.3d at 168.
See also Brinkley-Obu,
36 F.3d at 351;
E.E.O.C. v. McCarthy,
768 F.2d 1, 3 n. 4 (1st Cir.1985).
Pollis had the right to bring an Equal Pay Act claim at any time during the course of her employment. The statute of limitations requires that such claims be brought promptly to protect the defendant against stale or fabricated claims. The fact that it might have been awkward for Pollis to bring her suit against the New School while continuing to teach there is not a sufficient reason to exempt her from the statute of limitations. Her recovery should have been limited to damages incurred within the three-year limitations period for willful or reckless violations.
B.
‘Willful” Violation
The New School contends that there is insufficient evidence to support the jury’s finding that the New School willfully violated the Equal Pay Act.
A defendant’s violation of the Equal Pay Act is willful or reckless within the meaning of § 255(a) if “the employer either knew or showed reckless disregard for the matter of whether its conduct was prohibited by the statute.”
Reich v. Waldbaum, Inc.,
52 F.3d 35, 39 (2d Cir.1995) (quoting
McLaughlin v. Richland Shoe Co.,
486 U.S. 128, 133, 108 S.Ct. 1677, 1681, 100 L.Ed.2d 115 (1988));
see also Hazen Paper Co. v. Biggins,
507 U.S. 604, 617, 113 S.Ct. 1701, 1710, 123 L.Ed.2d 338 (1993). A plaintiff need not show that an employer acted with intent to discriminate or in bad faith.
See Trans World Airlines, Inc. v. Thurston,
469 U.S. 111, 126 n. 19, 105 S.Ct. 613, 624 n. 19, 83 L.Ed.2d 523 (1985):
Pollis testified that on multiple occasions over several years, she complained to New School decisionmakers about discrepancies between her salary and the salaries of male professors. Responses she received, construed in the light most favorable to Pol-
lis, indicated an awareness on the part of the administration that her salary level was below that of comparable male teachers. Nonetheless, the School continued to pay Pollis less than comparable male teachers.
This evidence — that the New School knew that Pollis was paid less than comparable males, but did not rectify the situation — is sufficient to support the jury’s finding of reckless or willful violation of the Equal Pay Act.
See E.E.O.C. v. Delaware Dep’t of Health and Soc. Servs.,
865 F.2d 1408, 1419 (3rd Cir.1989).
Under 29 U.S.C. § 206(d), therefore, compensatory damages for the Equal Pay Act violation should have been calculated by reference to the three-year limitations period for willful violations, and the resulting compensatory award should be doubled pursuant to the Fair Labor Standards Act’s liquidated damages provision, 29 U.S.C. § 260.
See E.E.O.C. v. Detroit Health Dep’t,
920 F.2d 355, 358 (6th Cir.1990);
Brinkman v. Department of Corrections,
21 F.3d 370, 373 (10th Cir.),
cert. denied,
513 U.S. 927, 115 S.Ct. 315, 130 L.Ed.2d 277 (1994).
II.
Title VII Claim
The New School contests the sufficiency of evidence in support of the jury verdict in Pollis’s favor on her claims that the New School violated Title VII and the New York Human Rights Law by granting her only part-time, adjunct, non-tenured employment upon her attainment of the institution’s mandatory retirement age of 70. The New School asserts that the evidence presented at trial, and all reasonable inferences drawn from it, did not reasonably support the plaintiffs burden to show that the decision involved intentional discrimination based on plaintiffs gender.
See Fisher v. Vassar College,
114 F.3d 1332, 1347 (2d Cir.1997) (in banc),
petition for cert. filed,
66 U.S.L.W. 3178 (U.S. Sept. 2, 1997) (No. 97-404).
Pollis based her case on a statistical analysis of the eight tenured Graduate Faculty professors who were subject to the mandatory retirement provisions of the New School’s by-laws between 1974 and 1993.
Under the by-laws, full-time faculty members were required to relinquish tenure, full-time employment, and the right to participate in the governance of academic affairs at the end of the academic year in which they turned 70.
Six male professors (Dr. Hans Jonas, Dr. Saul Padover, Dr. Aaron Gur-witsch, Dr. Jerome Bruner, Dr. Robert Heil-broner, and Dr. Herbert Schlesinger) and two female professors (Dr. Mary Henle and the plaintiff) were subject to the policy during the relevant time period.
Three of the male professors in the comparison group turned 70 more than twenty years prior to Pollis, during the administration of President James Everett. Jonas turned 70 in 1966. He was then appointed as the Alvin Johnson Professor of Philosophy, a full-time, tenured position with no specified retirement date. Gurwitsch and Padover turned 70 in 1971. They were each awarded Distinguished Service Professorships, which allowed them to continue on as full-time faculty members.
The three other male New School professors in the comparison group reached the age of 70 during the administration of President Fanton, who was hired in 1982. Dr. Jerome Bruner was a world-famous psychologist, the author of internationally acclaimed studies in psychology, and one of the two or three most famous professors on the New School faculty. He turned 70 in 1985. Bruner requested that the New School continue his affiliation in order for Bruner to remain eligible for a research grant from an outside foundation which would extend beyond his mandatory retirement date. The terms of Bruner’s original appointment to the New School in 1981 (at the age of 65) had stated that “the institution has the power, without specific Board of Trustee review, to extend [the mandatory retirement age] to age 75 on a year by year basis.” Fanton testified that
this language “seemed to suggest that Professor Bruner had at least an opening if not an understanding that he could continue beyond age 70.” Although Bruner’s tenure was discontinued at the end of the academic year in which he turned 70, Fanton recommended, and the Board of Trustees of the New School approved, a two-year appointment to a nontenured, full-time position as Professor of Social Sciences and Humanities. At the conclusion of the two-year appointment, Bruner sought, but was denied, continuation in this position. The New School offered, as an alternative, a combined appointment with the Social Science Research Council characterized as a “phasing.down’* of Bruner’s affiliation. Bruner rejected the offer, and left the university.
Dr. Robert Heilbroner, an internationally famous economist widely published in both the academic and popular press, reached 70 in 1988. Like Bruner, he was one of the stars of the New School Graduate Faculty. Fanton asked Heilbroner to stay on at the New School in a non-tenured, part-time position, in which his duties included editing an alumni newsletter, serving as Senior Fellow in the Wolfson Center, fund-raising, and teaching. After the grant funds that supported publication of the newsletter were expended and the fundraising drive concluded, Heilbroner became an adjunct professor at the New School, where he continued to teach courses in the economics department on an ongoing, non-tenured basis.
Dr. Herbert Schlesinger, a psychology professor and director of the New School’s clinical psychology program, reached age 70 in October 1991. Schlesinger was the largest grant recipient in the Graduate Faculty, with grants that continued past his scheduled retirement age. Schlesinger’s intended replacement as director of the clinical program turned down the position shortly before Schlesinger’s anticipated retirement. The New School- asked Schlesinger to stay on as acting director of the clinical program on a part-time basis until a replacement could be found. Schlesinger turned down the offer; however, he continued his association with the New School as an adjunct professor, teaching three courses in the first year of his retirement and one course in each of the following years.
Dr. Mary Henle, a well-recognized scholar in the field of Gestalt psychology, reached age 70 in 1983. Henle requested that she be allowed to continue teaching full-time, at a reduced salary. Her request was denied; she was offered, and declined, an adjunct position teaching one course per semester.
The plaintiff turned 70 in June 1993. Pri- or to her birthday, she informed the provost that she did not want to retire. She applied for a vacant tenured professorship in the New School’s political science department, and informed the department that she was willing to accept the position on a non-tenured basis. Her application was denied, and she was offered a position as adjunct professor, teaching two courses per year.
Pollis’s statistical evidence suffers from several serious flaws that render it insufficient to sustain a reasonable inference that her treatmerit by the New School was motivated by discriminatory intent.
First, the size of the group subjected to statistical analysis was tiny — especially considering that the comparisons encompassed a twenty-year period. A statistical showing of discrimination rests on the inherent improbability that the institution’s decisions would conform to the observed pattern unless intentional discrimination was present. The smaller the sample, the greater the likelihood that an observed pattern is attributable to other factors and accordingly the less persuasive the inference of discrimination to be drawn from it. Thus, several courts have ruled, as a matter of law, that discrimination may not be proved by statistics involving so small a pool.
See Mayor of Philadelphia v. Educational Equality League,
415 U.S. 605, 621, 94 S.Ct. 1323, 1333, 39 L.Ed.2d 630 (1974) (statistical evidence regarding 13-member panel insufficient, based on small sample size, to support inference of racial discrimination);
Haskell v. Kaman Corp.,
743 F.2d 113, 121 (2d Cir.1984) (10 terminations over 11-year period insufficient sample size to support inference of age discrimination);
Coble v. Hot Springs Sch. Dist. No. 6,
682 F.2d 721, 733-34 (8th Cir.1982) (15 decisions over 8 years insufficient sample size
to support inference of gender discrimination).
In addition to the small size of the group to which Pollis seeks comparison, each of the male members of the group differs so substantially from Pollis that no meaningful inference may be drawn from the statistics.
Three of the male faculty members included in the comparison group — Gurwitsch, Padover, and Jonas — reached the mandatory retirement age between 22 and 27 years before Pollis. Their appointments to full professorships at that time were recommended by a different university president, and ratified by a Board of Trustees from which, no doubt, few or no members remained to participate in the later retirement decisions.
Even more significant than the difference in decisionmakers, furthermore, is the huge lapse of time separating the decisions in question and hence the different circumstances in which the decisions were made. The evidence showed that during the early 1970s the New School had a far more permissive attitude generally toward extending teaching careers past age 70 than in 1993. In the 1970s
all
professors reaching 70 were exempted from mandatory retirement. Fan-ton testified that when he assumed the presidency in 1982, he adopted a stricter policy toward continued teaching after the mandatory retirement age. Furthermore in 1993, when Pollis’s case arose, there was a strong non-diseriminatory reason for the New School to look warily on such applications. Under the provisions of the Age Discrimination in Employment Act about to come into effect only months later, on January 1, 1994, educational institutions would lose the right to retire tenured professors mandatorily at the age of 70. Finally, we note that no requests by women in the 1970s were rejected; thus, the fact that three men were exempted supports no suggestion that gender-based discrimination was operating at that time.
Thus, Pollis’s statistical comparison of her case in 1993 with three grants of full-time teaching positions to males in the late 1960s and early 1970s furnishes no useful information as to whether the decision in her case was motivated by bias.
The statistical comparison involving five professors who became subject to the mandatory retirement policy in the 1980s and 1990s is no more probative of discriminatory intent. First, needless to say, a sample size of five is even smaller and therefore less probative than a sample of eight. Furthermore, the three males who were offered full-time positions in the 1980s and ‘90s were so different from Pollis, and from Henle, that the comparisons are virtually meaningless. As noted, Bruner and Heilbroner were internationally celebrated stars, who brought prestige to the institution and attracted students and other professors.
While Pollis offered evidence that she enjoyed a good scholarly reputation, she offered no suggestion that she was considered one of the most renowned members of the New School faculty or enjoyed a reputation akin to that of Bruner or Heilbroner. Bruner and Schlesinger, furthermore, were major grant recipients and thus represented important sources of funding for the New School. Pollis had no such economic importance to the School.
Finally, the New School desperately needed Schlesinger to stay on because his intended replacement as director of the clinical program had unexpectedly withdrawn. When Schlesinger declined the School’s plea that he continue' to serve full-time until a replacement could be found, moreover, he continued, like Pollis, as a part-time adjunct teacher.
Pollis made no showing that either her situation or Henle’s
was comparable in these
vital respects to Bruner’s, Heilbroner’s or Sehlesinger’s. The only male member of the statistical group reasonably comparable to Pollis is Schlesinger after he declined the School’s plea that he continue to run his clinical program; in that situation, the School gave him the same position as Pollis,- on terms not shown to have been significantly different.
In short, because her statistical group was so tiny, was spread over such a long period, and was composed So largely of individuals who were not fairly comparable to her, Pol-lis’s statistics do not support an inference about the School’s motivations in offering her less than full-time employment when she reached the mandatory retirement age. We will not impute discrimination in the terms of employment on the basis of circumstantial evidence that has no logical tendency to show that discrimination was present.
Once a Title VII defendant has answered the minimal requirements of the “pri-
ma facie” case by proffering a non-discriminatory explanation, plaintiff must point to evidence that can reasonably support the inference by a preponderance of the evidence that the defendant’s conduct was attributable to the prohibited discriminatory, motivation.
See Fisher,
114 F.3d at 1337. Where a plaintiff lacks evidence of employer conduct that directly expresses discrimination, the plaintiff may rely on statistical or other circumstantial evidence. If the plaintiff seeks to prove the discrimination by statistical evidence, however, the statistics must support reasonably the inference that the employer’s adverse decision would not have occurred but for discrimination. Pollis’s statistical evidence is too scant and too. flawed to support such an inference.
Finally, Pollis argues that an inference of discriminatory intent can be drawn from the fact that she was treated with “callous thoughtlessness bordering on hostility” by the dean and provost of the New
School and never granted the status of emeritus professor although that position was given to other professors, including Dr. Mary Henle, upon retirement. Absent some evidence that it was motivated by discriminatory intent, however, bad treatment does not establish a violation of Title VII.
We therefore reverse the jury verdict in favor of Pollis on her claim that the New School’s implementation of its mandatory retirement policy violated Title VII.
Since the standards under the New York Human Rights Law are the same as those under Title VII,
see Tomka,
66 F.3d at 1304 n. 4, we also reverse the jury verdict in Pollis’s favor on her state law claim.
We note, finally, that the district court’s award of attorneys’ fees was based, in part, on awards that we have overturned. Accordingly, we vacate the award and remand for reconsideration in the light of the extent to which Pollis’s suit was ultimately successful.
Conclusion
We vacate the jury verdict finding the New School liable for violation of Title VII and violation of the New York Human Rights Law. We vacate the district court award of damages covering nineteen years of violation of the Equal Pay Act, and remand for the district court to calculate the proper amount of damages incurred within the three-year limitations period. Finally, the award of attorneys’ fees is vacated and remanded for reconsideration in the light of the ultimate result of Pollis’s suit.