KARE 11, a Twin Cities television station, refused to renew C. Thomas Ryther's contract as lead sportscaster for a fifth three-year term. In 1991, when Ryther was terminated, he was fifty-three years old. Ryther sued KARE 11 and its parent, Gannett Co., Inc. (collectively "KARE 11"), alleging a violation of the Age Discrimination in Employment Act ("ADEA"), 29 U.S.C. §§ 621-634. Following a jury verdict in Ryther's favor, the district court, the Honorable David S. Doty presiding, denied a motion for a new trial and, alternatively, a motion for judgment as a matter of law. The court entered judgment awarding Ryther $1,254,535 in back pay, front pay, liquidated damages, and attorneys' fees. See Ryther v. KARE 11, 864 F.Supp. 1510 (D.Minn.1994). KARE 11 appeals. We affirm the judgment of the district court.

* Ryther served as a sports anchor for Channel 11 from December 1979 until July 1991, pursuant to a series of four three-year contracts. Gannett/KARE 11 purchased the station in 1983, and in 1988 Janet Mason became KARE 11's vice president of news. At that time, the sports department's members included Jeffrey Passolt and Randy Shaver, both under age 40. Ryther then appeared on the six o'clock and ten o'clock news and hosted a weekly show, "Prep Sports Extra," during the football season. Passolt did a sports feature on the five o'clock news and, along with Shaver, served as weekend sports anchor.

In the summer of 1988, Ryther was approximately fifty years of age. Ryther's responsibilities began changing that year, shortly after Mason's appointment to vice president. KARE 11 removed Ryther from Prep Sports Extra, which he then co-anchored with Shaver, and during 1989, the year in which Linda Rios Brook became station manager, Mason removed Ryther from the six o'clock news and assigned him to a recreational segment on the five o'clock news. Passolt replaced Ryther as sports anchor during the six o'clock time slot. In May 1990, Shaver was named executive producer of sports, a position to which Ryther was entitled under his contract. Shaver assumed many of Ryther's organizational and planning duties.

On March 6, 1991, shortly after Ryther discovered he was being excluded from promotional photos, Ryther confronted Mason about the status of his contract. Mason told him his contract would not be renewed because he had failed in the market research. After several events detailed in the district court's opinion, Ryther, 864 F.Supp. at 1515-16, Ryther left KARE 11 and filed this lawsuit.

The decision not to renew Ryther's contract was made by Rios Brook, Richard Modig, Vice President of Broadcast Operations, and Mason. When Rios Brook was asked at trial what market research she "relied on" in making the decision about Ryther, she responded that it was the "Gallup" research, in reference to a survey conducted for KARE 11 in June 1990 by the Gallup Organization ("1990 Gallup Survey"). Tr. IV-136. Mason, similarly, said that she arrived at that decision after she got the 1990 Gallup Survey. Tr. V-194, V-197.

In earlier years, 1981 and 1989, there had been other market research, performed by the Atkinson-Farris Communications research firm ("Atkinson"), to determine KARE 11's ratings. In 1981, the Atkinson research found that Ryther was "not impressive" and that his quality score was "extremely low." In 1989, Atkinson again did a survey for KARE 11 and found sports "the softest part of your team." Ryther had "virtually the same ratings" as he had in 1981.

In 1990, partially because KARE 11 found the Atkinson research incomplete, KARE 11 sought new market research by commissioning the 1990 Gallup Survey. The 1990 Gallup Survey reported that Ryther had seventy-six percent viewer recognition, whereas Mark Rosen, a sportscaster at competitor WCCO, had eighty-one percent recognition. Rosen was rated number one and Ryther number two in the overall Twin Cities' market. The 1990 Gallup Survey reported that Ryther "underperform[ed]" and that he was not a strong player for KARE 11.1

KARE 11 urges that, upon receipt of the 1990 Gallup Survey, Mason, Rios Brook, and Richard Modig, KARE 11's vice president of operations, made the decision not to renew Ryther's contract in August 1990. The primary issue at trial, and also on this appeal, is whether the overall market research was the true reason for Ryther's dismissal, or merely a pretext for age discrimination. Ryther asserts that he offered evidence to show that this was not true, that in fact the decision was made prior to that time, and that the research was biased and merely a pretext for unlawful age discrimination.

The district court, in overruling KARE 11's motion for judgment as a matter of law, carefully summarized the evidence from which a jury could reasonably find that the alleged nondiscriminatory reason for refusing to rehire Ryther was false. In this regard, Judge Doty found that there was sufficient evidence for the jury reasonably to conclude that: the defendants made the decision not to renew Ryther's contract before the 1990 Gallup Survey was undertaken; some of Ryther's duties had been transferred to younger people and that his contract was not renewed despite positive performance evaluations from KARE 11; KARE 11 deceived Ryther by leading him to believe that his work was commendable, in order to prevent him from improving upon his alleged deficiencies; the 1990 Gallup Survey was purposely designed so that Ryther would not get a fair rating, thus masking the discriminatory reason for his termination; and KARE 11 provided a hostile work environment for Ryther because of his age. Ryther, 864 F.Supp. at 1515-18.2

Our standard of review of this evidence is governed by settled rule:

[W]e must consider the evidence in the light most favorable to [Ryther], assume all conflicts in the evidence were resolved by the jury in [Ryther's] favor, and give [Ryther] the benefit of all favorable inferences that may reasonably be drawn from the proven facts. A judgment [as a matter of law] should be granted only when all the evidence points one way and is susceptible of no reasonable inferences sustaining [Ryther's] position.

Frieze v. Boatmen's Bank, 950 F.2d 538, 540 (8th Cir.1991) (internal citations and quotations omitted).3

Stated another way, it is well settled that we will not reverse a jury's verdict for insufficient evidence unless, after viewing the evidence in the light most favorable to the verdict, no reasonable juror could have returned a verdict for the non-moving party. Gardner v. Buerger, 82 F.3d 248, 251 (8th Cir. 1996).

The law governing the allocation of evidentiary burdens in age discrimination cases like this one is well established. See generally St. Mary's Honor Ctr. v. Hicks, 509 U.S. 502, 510 & n. 4, 113 S.Ct. 2742, 2749 & n. 4, 125 L.Ed.2d 407 (1993); Texas Dep't of Community Affairs v. Burdine, 450 U.S. 248, 252-56, 101 S.Ct. 1089, 1093-95, 67 L.Ed.2d 207 (1981); McDonnell Douglas Corp. v. Green, 411 U.S. 792, 800-06, 93 S.Ct. 1817, 1823-26, 36 L.Ed.2d 668 (1973).4

KARE 11 does not contend that Ryther failed to establish a prima facie case of age discrimination. There exists ample evidence that the jury could reasonably believe that (1) Ryther was within the protected age group (he was fifty-three years old); (2) as manifested by his contract renewals and KARE 11's own evaluations, he had been performing his job at a satisfactory level for over twelve years; (3) his contract in 1991 was not renewed;5 and (4) KARE 11 replaced him with a younger person. (Jeff Passolt was only thirty-three years of age and did not have as high of a performance rating as Ryther.)

The dissent urges that, once KARE 11 articulated a non-discriminatory reason for its actions, it was entitled to a judgment as a matter of law because it destroyed plaintiff's prima facie case. The articulation of a non-discriminatory reason by the employer destroys the legal presumption of plaintiff's prima facie case, and plaintiff is therefore no longer entitled to a judgment as a matter of law. Hicks, 509 U.S. at 510, 113 S.Ct. at 2749. However, as Hicks and Burdine make clear, if the employee can demonstrate that the reason given for the employer's action is pretextual, then the case moves to a new level of factual inquiry. Id. at 515, 113 S.Ct. at 2752 (citing Burdine, 450 U.S. at 255, 101 S.Ct. at 1094). Under these circumstances, assuming that there exists credible and substantial evidence of pretext, the plaintiff may still rely upon the elements of the prima facie case and the substantial evidence of pretextuality to urge to the trier of fact that the employer was guilty of intentional discrimination. Id. There is no synergistic formula to determine whether the elements of a prima facie case and pretext are sufficient to allow a jury to make a finding of intentional discrimination; we simply inquire whether there is substantial evidence which would allow a finder of fact to determine that the employer's reasons were a subterfuge or mask to conceal age-based animus.

With this in mind, we again turn to the fundamental issue in this case: whether Ryther produced sufficient evidence to allow a jury reasonably to find that KARE 11 intentionally discriminated against him on the basis of his age. Although much of the evidence is circumstantial, we agree with the district court's careful analysis that a reasonable jury could infer that KARE 11's asserted reason for discharge was false, and that the evidence was sufficient to allow a jury to find that KARE 11 engaged in age discrimination.6

Ryther urges that the record is replete with evidence that his research ratings reflected not his abilities, but KARE 11's failure to emphasize sports. The plenary evidence to this effect included the testimony of Ryther that, just days before his dismissal, Paul Baldwin, KARE 11's assistant news director, told him, "[t]he research isn't your fault," and explained that Ryther's showing relative to WCCO's Mark Rosen was the result of WCCO's promotion of Rosen, its ownership of broadcast rights in several major sporting events, and its emphasis on sports generally. Other evidence showed that Ryther continued to ask for better sports promotions, but was denied. In fact, Rios Brook admitted that "[sports] was not an area that I was concerned about," and Mason testified that "sports was relatively unimportant" in comparison to other parts of the newscast. Relatedly, Gallup Vice President Dr. Frank Newport admitted that Ryther's showing might be due in part to KARE 11's poor promotion of sports and noted that Rosen's recognition was "unusual" for a sportscaster. Yet despite KARE 11's own lack of sports promotion, Ryther remained the number two sportscaster in the market, second only to Rosen, and above KARE 11's own Jeff Passolt and Randy Shaver.7

There can be little doubt that, although it had before it the 1981 and 1989 Atkinson research, the jury could reasonably reject KARE 11's alleged reliance upon Ryther's low market ratings on the ground that KARE 11 kept rewarding Ryther for his performance. In fact, KARE 11 negotiated and awarded Ryther with substantial salary increases in three different interim threeyear contracts. These contract renewals could easily justify a finding that, in this interim time period, Ryther's performance was more than adequate to fulfill KARE 11's programming interests.8

In May 1989, Lilyan Wilder, a training consultant to KARE 11, copied a letter to Janet Mason, written to Ryther after reviewing his performance in a training session. The letter read in part:

It was a pleasure to see you again and to work with you. Your authority, your sense of "sports" and the essence of it, are excellent. Your timing, your play-by-play and your good, strong voice are all positive.

Appellee's App. at E2 (emphasis added). Likewise, as late as August 1, 1990, Barry Nash, a talent coach hired by KARE 11 in 1990, wrote about Ryther to Mason and Baldwin:

Hats off to Tom for the effort to create reports with more universal appeal. Innovations like the Three Musketeers footage he used to begin his piece on fencing are certainly a step in the right direction.

Most significant, however, in the consideration of the conflicting evidence, notwithstanding the earlier Atkinson reports, is Mason's personal review of Ryther's performance in March 1990. She gave him the rating of "commendable," the second highest mark possible, and indicated that "his work is done quickly and accurately; total job responsibilities are met." Mason's 1990 review of Ryther also stated in part: "As anchor: knows the market & key players/contacts[;] he wants to put on a good product--open to trying new ideas.... As sports director--has developed good working relationship with the movers & shakers of the professional & college sports world." With this kind of commendation written as late as March 1990, it is readily understandable how the jury could reject the prior market research of "underperformance" as the reason for Ryther's termination.

Even assuming that the "research" allegedly relied upon included both the Atkinson reports and the 1990 Gallup Survey, we conclude that there is sufficient evidence for a reasonable jury to find that it played little or no role in KARE 11's decision not to retain Ryther in 1991.

This then brings us to the 1990 Gallup Survey--the market research upon which Mason and Rios Brook specifically say they relied in making the decision not to rehire Ryther.

B. Ryther's Claims that the 1990 Gallup Survey Was Biased

Ryther testified that the 1990 Gallup Survey questions were both designed and interpreted to provide an incomplete picture of viewers' perceptions of his performance. He initially challenged the 1990 Gallup Survey's methods as an incomplete means of obtaining research concerning his performance.

Gallup surveyed a random sample of viewers using two methods: a "Q score technique" and open-ended questions. The Q score technique employs multiple questions to measure audience recognition and approval (particularly strong like and dislike) of the selected personalities. Ryther was among twenty-five on-air personalities included in this portion of the survey.

Open-ended questions, by contrast, allow viewers to describe identified persons in their own words, and by Gallup's description are "designed to help [stations] gain a more complete understanding of what viewers think about key personalities." For example, the 1990 Gallup Survey asked viewers, "How would you describe Jeff Passolt? What comes to mind that you particularly like or dislike about him as a newscaster?" The ten key personalities included Rosen, Passolt, and KARE 11's other lead anchors, among others. Ryther, however, was excluded from this portion of the research.

Ryther also notes Janet Mason's admission that in advance of the research she told Gallup that one of the "important issues" about which KARE 11 sought information was "the sportscaster position." Although Mason identified Passolt and Rosen as "key personalities" for purposes of the research project, she did not so characterize Ryther. Rather, she justified the omission of open-ended questions about Ryther on the grounds that their inclusion would have made the survey "too long" and that similar questions had been asked about him in 1989 research conducted by Atkinson. Mason also admitted, however, that the 1989 Atkinson project asked such "free response" questions concerning each of the ten other "key personalities."

KARE 11 dismisses Ryther's argument that KARE 11 designed the 1990 Gallup Survey in a manner unfavorable to him as an irrelevant argument that is "without foundation and intrusive of KARE's business judgment." Reply Br. at 8. KARE 11's statement not only mischaracterizes Ryther's attack on the survey, which is plainly a claim that the survey was biased, but is incorrect as a matter of law. As the Supreme Court unanimously observed in Burdine, the fact "that the employer misjudged the qualifications of the [plaintiff] does not in itself expose [the employer] to ... liability, although this may be probative of whether the employer's reasons are pretexts for discrimination." 450 U.S. at 259, 101 S.Ct. at 1097 (emphasis added). The jury may thus consider as wholly relevant both whether the 1990 Gallup Survey was designed in a manner that from the outset disfavored Ryther, and whether the survey was actually a sound--as opposed to pretextual--basis upon which to make employment decisions.

It remains an open question whether, standing alone, this evidence would support the jury's verdict. But we are concerned with whether the overall evidence supports a reasonable inference that age motivated KARE 11's actions. Hicks, 509 U.S. at 510, 113 S.Ct. at 2749. To that end, Ryther's attack on the survey is probative. The ultimate concern, of course, is whether the employer gave an honest explanation of its behavior. See Harvey v. Anheuser-Busch, Inc., 38 F.3d 968, 973 (8th Cir.1994). Yet, in the nature of things, evidence that the defendant employer says it relied on inaccurate market research may assist the finder of fact in determining whether the employer is giving an honest explanation of its actions. See Burdine, 450 U.S. at 259, 101 S.Ct. at 1097.

As the district court held, the jury reasonably could have found KARE 11's explanations to be "trivial" and inferred that the real reason defendants omitted Ryther from the open-ended questions was a fear that the results of the survey would undermine their age-based decision not to renew his contract. Relatedly, a reasonable jury might also infer that, if it was unwieldy or redundant to repeat such questions about Ryther, KARE 11 ought to have excluded such repetitious questions about Passolt and Rosen as well. In other words, a reasonable jury could have reasoned that, if it was redundant and costly to ask open-ended questions about Ryther, it was redundant and costly to ask open-ended questions about Rosen, Passolt, and the other eight "key personalities," all of whom were included in the 1989 Atkinson research. The fact that KARE 11 did not include Ryther in this portion of the 1990 Gallup Survey reasonably suggests that KARE 11 had already decided to terminate Ryther. Moreover, as the district court stated, the long delay between the research results and the time of Ryther's notice of dismissal reasonably suggests the defendants did not want to provide Ryther an opportunity to address his weaknesses, and thus supports the inference that KARE 11 had an age-based agenda to terminate Ryther. The jury had every right to believe that the survey was inadequate, biased, and in fact a subterfuge to mask KARE 11's age-based animus against Ryther.

C. Mason's Treatment of Ryther Before the Gallup Survey of 1990

The district court found sufficient evidence for the jury to conclude that Janet Mason's decision not to renew Ryther's contract was made before the 1990 Gallup Survey was commissioned. The evidence to this effect included Ryther's testimony that: (1) between 1988 and 1990, KARE 11 transferred his duties to younger members of the sports department; (2) when Mason assumed her role as Ryther's supervisor in 1988, KARE 11's managing editor Marie Kurken told him to "watch [his] back" because Mason "was out to get" him, and he "was number one on her list, on her hit list, to get out of that news room"; (3) Mason treated Ryther as though he "couldn't seem to do anything right"; and (4) when Mason took over, he "went from being a valued member of the news staff sports department to almost a--in Janet Mason's eyes, as an incompetent. And incidents kept happening that underlined and verified those words of Marie Kurken. It kept happening and happening and happening, so I noted them." In addition, there was documentary and testimonial evidence that Mason, in March 1990, gave Ryther the rating of "commendable," stating that his "work is done quickly and accurately; total job responsibilities are met," but shortly thereafter, when notifying him of his dismissal, explained the decision as based on the showing of earlier research that Ryther was a "failure" in the market.

In McDonnell Douglas Corp., the Court observed that "evidence that may be relevant to any showing of pretext includes facts as to the [employer's] treatment of [plaintiff] during his prior term of employment." 411 U.S. at 804, 93 S.Ct. at 1825. As the unanimous McDonnell Douglas Court understood, evidence that the defendant treated the plaintiff, whose performance remained stable throughout the relevant period, differently upon a change in supervisors may, together with the elements of the prima facie case and evidence that the new supervisor "was out to get" him, support a reasonable inference that age motivated that difference in treatment. Id.; see also Hicks, 509 U.S. at 510, 113 S.Ct. at 2749 (As the Court explained, a jury's disbelief of employer's explanation, together with prima facie case, suffices to show intentional discrimination "particularly if disbelief is accompanied by a suspicion of mendacity." (emphasis added)).10

A jury might reasonably infer from Ryther's "unimproved showing" that KARE 11 felt his long-term performance justified the non-renewal of his contract. But a reasonable jury might also infer that KARE 11's continuous approval and commendable ratings of that performance belie that claim. There exists substantial evidence that, after Janet Mason became Ryther's supervisor (and before the 1990 Gallup Survey), KARE 11 determined that Ryther's contract should not be renewed. Moreover, it cannot be said that no reasonable jury could have rejected as contrived Mason's explanation that she rated Ryther favorably in March 1990 out of fear that rating him unfavorably would cause him to fall apart emotionally. Such a statement may appear untruthful to reasonable sensibilities. A reasonable jury could also infer that Mason failed to notify Ryther of his alleged deficiencies for fear that he might correct them,11 or that Mason treated Ryther as "an incompetent" because she harbored an age-based animus against him. See Hicks, 509 U.S. at 510, 113 S.Ct. at 2749. In sum, a reasonable jury could infer that Mason had made a decision to terminate Ryther before the 1990 Gallup survey was conducted.

D. Ryther's Claims that KARE 11 Had a Corporate Atmosphere Unfavorable to Older Employees

The district court relied on several portions of the record in holding that Ryther's evidence of a corporate atmosphere unfavorable toward older employees could reasonably support the jury's inference that Ryther was the subject of age discrimination. KARE 11 contends this evidence is insufficient, noting that statements made by employees not involved in Ryther's non-renewal and stray remarks in the workplace do not give rise to a reasonable inference of discrimination. Not only is KARE 11's reduction of this evidence to a few "stray remarks" factually incorrect, but, more importantly, such evidence can, if sufficient together with other evidence of pretext, support a reasonable inference of age discrimination. As the Supreme Court stated in McDonnell Douglas:

Other evidence that may be relevant to any showing of pretext includes facts as to the [employer's] ... general policy and practice with respect to [older persons'] employment. On the latter point, statistics as to [defendant's] employment policy and practice may be helpful to a determination of whether [its] refusal to rehire [plaintiff] conformed to a general pattern of discrimination against [older employees].

411 U.S. at 804-05, 93 S.Ct. at 1825 (footnote and citations omitted).

Although Ryther did not present his case in the form of statistical evidence, he did offer testimony suggesting KARE 11's actions "conformed to a general pattern of discrimination" against older employees. Id. at 805, 93 S.Ct. at 1825. This evidence included: Ryther's testimony that he was criticized for the bags under his eyes; Mason's testimony that she once considered allowing Ryther to wear glasses because she felt they might help cover them; testimony that several older employees were suddenly given poor performance ratings and forced to choose between early retirement and demotions; testimony that others in the sports department made cutting remarks about Ryther's age, calling him an "old fart," an "old man," and saying he was "too old to be on the air," and "had no business being in the industry any more for his age"; testimony that Shaver and Mason had frequent discussions about Ryther; and testimony that Shaver complained about Ryther to Mason on ostensibly age-related grounds. In the latter connection, the following excerpt from the testimony of Edward Villaume, a former sports department intern, is illuminating:

Q: Did you ever hear Randy Shaver make comments about Tom's age?

A: Yes, I did.

Q: And what comments did you hear Randy make about Tom's age?

A: Randy Shaver called Tom Ryther an old man, an old fart, and said he was too old to be on the air.

Q: Did you hear Jeff Passolt make any comments about Tom's age?

A: Yes, I did.

Q: And what comments did you hear Jeff Passolt make about Tom's age?

A: That Tom was an old man. He called him too old to be on the air, couldn't figure out why Randy and himself, Jeff, were not number one, and that Tom had no business being in the industry any more for his age, called him an old fart as well.

Q: Did you hear Randy Shaver make his comments on more than one occasion?

A: Yes, I did.

Q: Approximately how many times did you hear Randy Shaver make those comments?

A: I would say approximately ten or more.

Q: Did you hear Jeff Passolt make those comments on more than one occasion?

A: Yes, I did.

Q: And approximately how many times did Jeff Passolt make those comments?

A: Somewhere around ten. Not as often as Randy.

Q: Did you ever hear Dave Levine, or Levine, make comments about Tom Ryther's age?

A: Yes, I did. Dave would often chime right in with Randy and Jeff, or would make a comment on his own about Tom's age.

* * *

Q: Had you ever heard Randy Shaver complain to Janet Mason in your presence?

A: Yes.

Q: Can you tell us about what was said on that occasion when you were present when Randy complained to Janet Mason?

A: Randy had said to Janet that Tom was never around any more, that he was on the phone, and that he just wasn't able to grasp the new computer system and couldn't handle the, kind of the newer technology.

Q: Did you ever hear any other staff members make comments about Tom's age?

A: Yes.

Q: And who was that?

A: Brian Singer, who was a camera man, had mentioned that more than once, and also had mentioned the fact that he could not understand how Tom was still in the business and why Randy and Jeff were not the number one anchor position there in the sports department.

KARE 11 argues that the statements referenced in this testimony were not those of persons responsible for the decision not to renew Ryther's contract. To the extent that these statements were made outside the presence of the decisionmakers, KARE 11 is correct that they do not, standing alone, raise an inference of discrimination. Compare Frieze, 950 F.2d at 541-42 (reversing denial of defendant's motion for JNOV) with Morgan v. Arkansas Gazette, 897 F.2d 945, 950-51 (8th Cir.1990) (affirming denial of defendant's motion for JNOV). The evidence also reveals, however, that Shaver and Mason had frequent discussions about Ryther, and that they discussed Ryther's ability to "grasp" some of the "newer" developments at the station. Furthermore, other evidence shows that Mason was generally responsive to Shaver's ideas and demands, including his request that Ryther be taken off Prep Sports Extra. The jury could thus reasonably infer that Mason formed her judgment about Ryther on the basis of the discriminatory comments frequently made by Shaver, Passolt, Levine, and Singer, and acted on them by terminating him.

KARE 11 dismisses the testimony of three former KARE 11 employees that the station was systematically ridding itself of older employees because those employees were dissimilarly situated and because " 'individual employees' opinions of actions taken by their employer, ... in themselves, are insufficient to support [Ryther's] argument that his age was a determining factor in his discharge.' " Appellants' Br. at 35 (quoting Morgan, 897 F.2d at 950 (alteration ours)).12 As to KARE 11's reliance on Morgan, we think Judge John R. Gibson's opinion for this Court in Morgan supports our conclusion:

Much of the testimony recited above can be described as no more than individual employees' opinions of actions taken by their employer, which, in themselves, are insufficient to support Morgan's argument that his age was a determining factor in his discharge. There was, however, evidence that, during Tinker's administration, a pattern of employees over the age of forty leaving the circulation department and being replaced by younger employees developed. As we observed in MacDissi v. Valmont Industries, Inc., 856 F.2d 1054 (8th Cir.1988), in a similar context, "[t]his fact is certainly not conclusive evidence of age discrimination in itself, but it is surely the kind of fact which could cause a reasonable trier of fact to raise an eyebrow, and proceed to assess the employer's explanation for this outcome." Id. at 1058.

897 F.2d at 950-51 (emphasis added) (footnote omitted). The Morgan court went on to conclude that "additional threads of evidence which can be gleaned from the record," including a reference to a former employer as an "old 'fuddyduddy' [who was] not smart enough to help" his department, and one employee's "observation of a trend away from older, more experienced employees toward younger ones," "support[ed] a finding that age was a determining factor in the decision to fire" the plaintiff. Id. at 951. Thus, while the statements of sports department employees are not, "in themselves," sufficient to uphold the district court, those statements were relevant to the jury and, together with other evidence of pretext, such as a "trend" toward younger employees, and the elements of the prima facie case, support a reasonable inference of age discrimination.

The dissent argues that the articulation of a nondiscriminatory reason (i.e., Ryther's low ratings from the market research, which allegedly show that he performed his job unsatisfactorily) destroys one of the elements of the prima facie case. This is wrong for several reasons. First, KARE 11 has never contended on appeal that Ryther failed to make a prima facie case; second, in determining whether KARE 11 is entitled to a judgment as a matter of law, it is incumbent upon the trial court and the judges of this court to give the benefit of all favorable inferences to Ryther, who is the verdict holder; third, a trier of fact, when considering all of the evidence could reasonably find that the market research and the low rating of KARE 11's sports department were based upon KARE 11's own admission that it did not promote sports (and that "sports was relatively unimportant"); in addition, the record contains evidence, which the jury could reasonably believe, that the 1990 Gallup Survey, upon which Mason and Brook said they specifically relied in making the decision not to renew Ryther's contract, was biased against Ryther; fourth, notwithstanding the earlier market research, KARE 11 as late as 1990 declared that Ryther performed his job in a "commendable" way and that his "total job responsibilities are met;" fifth, notwithstanding the earlier market research, Ryther's contract was renewed three times with substantial raises.

Clearly, under the state of this record, whether Ryther was in fact performing his job satisfactorily and whether his performance met the reasonable expectations of KARE 11 was a question of fact for the jury.

In summary, the record as a whole supports a reasonable inference that age, and not some other factor, motivated KARE 11's decision not to renew Ryther's contract. The plaintiff produced overwhelming evidence as to the elements of a prima facie case,13 and strong evidence of pretextuality, which, when considered with Ryther's work environment's indications of age-based animus, clearly provide sufficient evidence as a matter of law to allow the trier of fact to find intentional discrimination. If this evidence is not sufficient for the plaintiff's case to be submitted to a jury, it is difficult to hypothecate what evidence may ever achieve the threshold standard set forth by Burdine and Hicks. And, as the experienced district court judge stated, "[i]t is clear that the jury believed Ryther's evidence and did not believe defendants' proffered explanation." Ryther, 864 F.Supp. at 1517.

In the alternative, KARE 11 requests a new trial. KARE 11 asserts it was prejudicial error to admit unconnected evidence and that the jury was given inaccurate and confusing instructions. We review both the district court's admission of the evidence, O'Dell v. Hercules, Inc., 904 F.2d 1194, 1200 (8th Cir.1990), and its choice of instructions, Resolution Trust Corp. v. Eason, 17 F.3d 1126, 1132 (8th Cir.1994), for clear abuses of discretion.

KARE 11 seeks a new trial on the basis of the district court's evidentiary rulings. In particular, KARE 11 challenges the admission of evidence concerning settlement agreements, "stray remarks," and anecdotal stories of other individuals, all of which, it asserts, were unconnected to KARE 11's decision to terminate Ryther.

A new trial based on evidentiary rulings is appropriate only when the challenged rulings were "so prejudicial as to require a new trial which would be likely to produce a different result." O'Dell, 904 F.2d at 1200. For the reasons stated by the district court, Ryther, 864 F.Supp. at 1523-25, we do not find that to be the case here.

A party is entitled to an instruction which accurately states the law and is supported by the evidence. See, e.g., EEOC v. Atlantic Community Sch. Dist., 879 F.2d 434, 436 (8th Cir.1989). The jury instructions below adhered to the Hicks standard. See Hicks, 509 U.S. at 510, 113 S.Ct. at 2749.14

KARE 11 asserts that these instructions failed to make clear that the jurors must find intentional age discrimination in order to return a verdict in Ryther's favor. We disagree. The instructions acknowledged KARE 11's proffered explanation for its decision and stated that Ryther was required to prove that its explanation was "merely a pretext or cover-up for intentional age discrimination." We cannot say the court abused its discretion in failing to instruct the jury otherwise.15

KARE 11 also asks the Court to reverse the jury's finding of willfulness and to vacate its corresponding award. The station suggests a finding that it was negligent or aware of the ADEA is insufficient, in and of itself, to support a finding of willful discrimination, and urges that the only evidence of willfulness here is testimony that KARE 11 had anti-discrimination employment policies and had instructed managers concerning how to execute those policies.

The Supreme Court has stated that a violation of section 7(b) of the ADEA, 29 U.S.C. § 626(b), is willful " 'if the employer knew or showed reckless disregard for the matter of whether its conduct was prohibited by the ADEA.' " Hazen Paper Co. v. Biggins, 507 U.S. 604, 614, 113 S.Ct. 1701, 1708, 123 L.Ed.2d 338 (1993) (quoting Trans World Airlines v. Thurston, 469 U.S. 111, 126, 105 S.Ct. 613, 624, 83 L.Ed.2d 523 (1985) (internal quotation marks and ellipsis omitted in original)). Upon review, the issue is whether a rational jury could have concluded KARE 11's conduct met this standard. Glover v. McDonnell Douglas Corp., 12 F.3d 845, 848 (8th Cir.1994), cert. denied, 511 U.S. 1070, 114 S.Ct. 1647, 128 L.Ed.2d 366 (1994).

KARE 11 is correct to suggest that conduct that is " 'merely negligent' " will not support a finding of willfulness. Biggins, 507 U.S. at 615, 113 S.Ct. at 1708 (quoting McLaughlin v. Richland Shoe Co., 486 U.S. 128, 133, 108 S.Ct. 1677, 1681, 100 L.Ed.2d 115 (1988)). It is also well established that our concern is not simply whether the employer acted voluntarily, but whether it consciously violated the ADEA. See Brown v. Stites Concrete, Inc., 994 F.2d 553, 560 (8th Cir.1993) (en banc); MacDissi v. Valmont Indus., Inc., 856 F.2d 1054, 1061 (8th Cir.1988). Thus, the Supreme Court has declined to hold "that a violation of the Act is 'willful' if the employer simply knew of the potential applicability of the ADEA." Thurston, 469 U.S. at 127, 105 S.Ct. at 624.

That said, the employee need not, to prove willfulness, show "that the employer's conduct was outrageous, or provide direct evidence of the employer's motivation, or prove that age was the predominant, rather than a determinative, factor in the employment decision." Biggins, 507 U.S. at 617, 113 S.Ct. at 1710. And, as the Biggins Court emphasized, "[i]t would be a wholly circular and self-defeating interpretation of the ADEA to hold that, in cases where an employer more likely knows its conduct to be illegal, knowledge alone does not suffice for liquidated damages." Id.

Ample evidence supports the jury's finding that KARE 11 "more likely [knew] its conduct to be illegal" here. The record reveals that KARE 11 had equal employment opportunity ("EEO") policies that clearly forbade age-based discrimination, and that its managers received EEO and affirmative action training. Other testimony showed that all of KARE 11's managers were instructed to prevent age discrimination. KARE 11's directives spelled out that, before terminating an older employee, the manager should document the difficulties and offer the employee help and retraining. Yet Ryther received no such assistance or offers.

This is not, however, the only evidence suggesting KARE 11 acted willfully. As noted above, Ryther testified that Mason repeatedly treated him "as though he couldn't seem to do anything right," and that he "went from being a valued member of the news staff sports department to almost a--in Janet Mason's eyes, as an incompetent." Such evidence of repeated harassment may, together with other evidence, support a finding of willfulness. See Kelewae v. Jim Meagher Chevrolet, Inc., 952 F.2d 1052, 1054 (8th Cir.1992) (per curiam) (repeated incidents of harassment about the quality of employee's work in conjunction with suggestions that he retire supports finding of wilfulness). Here, further evidence showed that Mason and the others responsible made the decision not to renew Ryther's contract prior to commissioning the 1990 Gallup Survey, but concealed that decision from him, perhaps because they did not want to give him a chance to improve his performance. This evidence, too, supports a finding of willfulness. See Tolan v. Levi Strauss & Co., 867 F.2d 467, 471 (8th Cir.1989) ("Evidence of concealment may show the employer knew its conduct violated the ADEA.").

As the district court stated, "[i]t is clear that defendants were more than merely aware of the ADEA statute," and "[t]his is not a case where the employer incorrectly, but in good faith, believed that the statute permitted a particular age-based decision." Ryther, 864 F.Supp. at 1520. The entirety of the evidence suggests they knew they were violating the law in terminating Ryther on the basis of his age. At a minimum, it suggests those responsible were recklessly indifferent to the ADEA's requirements.

The judgment of the district court is therefore affirmed.

LOKEN, Circuit Judge, dissenting.

I respectfully dissent. I conclude that KARE 11 is entitled to judgment as a matter of law or, at a minimum, a new trial.

As the court explains, when a discrimination case is submitted to the jury, the presumption created by plaintiff's prima facie case is no longer relevant. But if the plaintiff has no direct evidence of age discrimination, as in this case, the elements of the prima facie case remain relevant, for they may, along with proof of pretext, satisfy plaintiff's ultimate burden to prove age discrimination. Hicks, 509 U.S. at 510, 113 S.Ct. at 2749. Thus, in submitting this case, the district court instructed the jury on four elements of Ryther's prima facie case. See footnote 14, supra at p. 1087.

There is no rigid formula defining the elements of a prima facie case of discrimination. See United States Postal Serv. Bd. v. Aikens, 460 U.S. 711, 715, 103 S.Ct. 1478, 1481, 75 L.Ed.2d 403 (1983). The elements necessarily vary depending upon, for example, the type of adverse employment action that is challenged and the nature of the alleged discrimination. But in every case, those elements must be sufficient to raise a valid inference of unlawful discrimination. See Furnco Constr. Corp. v. Waters, 438 U.S. 567, 575-77, 98 S.Ct. 2943, 2948-50, 57 L.Ed.2d 957 (1978). Here, the district court in my view failed to properly define the elements of Ryther's prima facie case. To explain why, I must review some undisputed facts concerning Ryther's employment history with KARE 11.

Ryther was first hired as Channel 11's lead sportscaster in 1979, when he was 41 years old and the station was under different ownership. Ryther was given a three-year fixed term contract at an initial annual salary of $55,000. The contract was renewed in 1982. Gannett/KARE 11 acquired Channel 11 in 1983. KARE 11 renewed Ryther's contract for additional three-year terms in 1985 and 1988. During the final year of the 1988 contract, Ryther was paid an annual salary of $160,000.

Television broadcasting stations commonly conduct periodic market research to survey the popularity of both the station and its on-air "personalities" in the local market. In December 1981, the independent research firm of Atkinson-Farris Communications surveyed the Twin Cities television audience for Channel 11. The researchers reported that Channel 11 was a distant third in "rating" among the three Twin Cities network affiliates.1 Regarding the popularity of Ryther, Atkinson-Farris reported:

The situation for Tom Ryther is particularly unfortunate. First, his overall [Quality] Score is not impressive (17), but more important, viewers of [Channel 11] are not even enthusiastic. A Quality Score of 18 from supporters of a personality's own station is extremely low.

Between May 1986 and October 1988, Gannett's in-house research organization conducted a number of "Tracking Studies" of the Twin Cities market. These studies showed Ryther's market impact and ability to attract viewers holding steady, well below KARE 11's newly-hired news and weather anchors. The studies also revealed that Mark Rosen, the new lead sportscaster for arch-rival WCCO, had already passed Ryther in these popularity measures.

In 1989, the Atkinson research firm again surveyed the Twin Cities television audience for KARE 11. It found that KARE 11's news programs had made "massive progress" since 1981, pulling even with WCCO in overall customer preference, well ahead of the third network affiliate. It found that KARE 11's lead news anchor had "the broadest base of support among newscasters in the market," and that its weather anchor was "the on-air person with the broadest appeal and greatest pulling power." However, it characterized KARE 11 sports as "the softest part of your team." It found that Ryther had "virtually the same ratings he had in our 1981 project," while Mark Rosen "has been able to come into the market during that time and pass Ryther." The report also noted that sports was relatively unimportant to viewers in choosing a local news program; only five percent of the persons surveyed listed sports as a reason for their newscast preferences.

In 1990, KARE 11 retained the Gallup organization to again survey the Twin Cities market. The Gallup survey again found that sports was a "low interest" facet of KARE 11's broadcasts, with a value of six percent (compared, for example, to weather's 76 percent). Gallup also reported that Ryther had relatively high viewer recognition (76 percent, compared to 81 percent for Mark Rosen), but low net impact (28 percent, compared to 45 percent for Rosen). KARE 11's other sportscasters, Jeff Passolt and Randy Shaver, had low recognition and low net impact. Because a widely recognized personality should attract more viewers, Gallup's expert, Dr. Frank Newport, testified that Ryther's scores--high recognition but low net impact--placed him in Gallup's "penalty box." The Gallup report concluded:The data suggest that Tom Ryther is not a strongly positive factor for KARE. The sportscaster position is the only front four role which is not filled by a very strong player for KARE. Ryther himself does below average on many of the key indicators created in this research: he underperforms based on where we think he should be based on his recognition and years in market.

Passolt, on the other hand, is no superstar either. His overall net impact, in fact, is roughly the same as Ryther's. The plus for Passolt would appear to be that he has a lower recognition, and thus is now performing at a higher level rel[ative] to where we think that he should be. Thus, it is our opinion that Passolt has the higher potential for the station.

At any rate, a change in sportscaster would appear to have a relatively low down-side risk for the station as Ryther is no star as is.

Shortly after receiving the Gallup report in the fall of 1990, KARE 11 decided not to renew Ryther's contract in July 1991. However, the station also decided to leave Ryther in place until his three-year contract expired in July 1991.2 Because KARE 11 permitted Ryther to complete his three-year contract term, the refusal to offer him a new contract in 1991 was, in essence, a refusal to hire. Thus, no matter how favorable Ryther's internal performance reviews at KARE 11 had been in the past, the relevant question in July 1991 was whether he was qualified to be hired (or rehired) as the lead sportscaster at a Twin Cities television station at a salary of $160,000 per year.

The district court's jury instruction took no account of this critical aspect of the case. Far worse, the instruction quoted in footnote 14 significantly misstated the elements of Ryther's prima facie case when it instructed the jury to find whether, "Second, plaintiff's job performance was satisfactory," and "Third, plaintiff was terminated from his job when his contract was not renewed." This instruction told the jury to ignore the fundamental difference between the decision whether to rehire an employee whose fixed-term contract has expired, and the decision whether to terminate an employee who has worked without the guaranteed but limited security of a fixed-term contract.

The instructions also ignored another essential aspect of the evidence in this case that impacts upon the elements of a prima facie case. The 1989 and 1990 research showed that (i) KARE 11 had gained substantial overall ratings despite a weak sports anchor, and (ii) sports attracts relatively few Twin Cities viewers. Following Ryther's non-renewal, his duties were spread among the remaining KARE 11 sportscasters; no one was added to the KARE 11 sports team. The independent market research justified KARE 11's decision to reduce this part of its newsroom force by not renewing an underperforming, highly compensated lead sportscaster and redistributing his job among the remaining staff. Compare Thomure v. Phillips Furniture Co., 30 F.3d 1020, 1024 (8th Cir.1994), cert. denied, 513 U.S. 1191, 115 S.Ct. 1255, 131 L.Ed.2d 135 (1995). Thus, the case had evidentiary elements of a reduction-in-force, and the jury should have been instructed accordingly.

In my view, this was prejudicial instruction error. Given ten years of market research showing that Ryther lacked the ability to attract local viewers, no Twin Cities station would have considered him qualified for his former position and salary.3 Thus, the independent market research gave KARE 11 a powerful, objective business reason for not renewing Ryther's contract that, in addition, refuted an essential element of his prima facie case. Compare Hayman v. National Acad. of Sciences, 23 F.3d 535, 538 (D.C.Cir.1994); Craft v. Metromedia, Inc., 766 F.2d 1205, 1216 (8th Cir.1985), cert. denied, 475 U.S. 1058, 106 S.Ct. 1285, 89 L.Ed.2d 592 (1986).4 And this research was disinterested, objective evidence gathered and offered at trial by third party professionals. See Dace v. ACF Indus., Inc., 722 F.2d 374, 377 n. 6 (8th Cir.1983).

In these circumstances, Ryther's purported pretext evidence failed to create a submissible case of age discrimination. When the employer's objective evidence not only tends to establish a legitimate, nondiscriminatory reason for the adverse employment action, but also effectively refutes the plaintiff's prima facie case, I think it highly unlikely that pretext evidence can support a reasonable inference of age discrimination. Evidence of pretext "is relevant only to the extent it contributes to an inference that [KARE 11] intentionally discriminated against [Ryther] because of his age. See Hicks, 509 U.S. at 510, 113 S.Ct. at 2749." Nelson v. Boatmen's Bancshares, Inc., 26 F.3d 796, 801 (8th Cir.1994).5

To summarize, ten years of independent market research established that Ryther was the overpaid, underperforming anchor of the least significant segment of KARE 11's news team. Gallup recommended a change, and KARE 11 acted on that recommendation. "[T]he issue is not whether the reason articulated by the employer warranted the discharge, but whether the employer acted for a nondiscriminatory reason." Halsell v. Kimberly-Clark Corp., 683 F.2d 285, 292 (8th Cir.1982), cert. denied, 459 U.S. 1205, 103 S.Ct. 1194, 75 L.Ed.2d 438 (1983). Although Ryther obviously persuaded the jury that KARE 11 treated him unfairly, I conclude he did not prove intentional age discrimination. Therefore, this case presents the same situation we faced in Barber v. American Airlines, Inc., 791 F.2d 658, 661 (8th Cir.), cert. denied, 479 U.S. 885, 107 S.Ct. 278, 93 L.Ed.2d 254 (1986):

We have carefully read every page of the testimony at this trial, and we are persuaded that this stringent standard [for setting aside a jury verdict] has been met. The jury could rationally have believed that plaintiffs ought in good conscience to have been permitted to stay in Little Rock ... but there is absolutely no substantial evidence in this record that would justify attributing American's actions to plaintiffs' age.

Like the panel in Barber, I would hold that the district court erred in denying KARE 11's post-trial motion for judgment as a matter of law. At a minimum, I believe that the district court's prejudicial instruction errors warrant a new trial.6 For these reasons, I respectfully dissent.