Plaintiff Douglas M. Mills (Mills) appeals the district court's grant of summary judgment to Health Care Service Corporation (HCSC) in his Title VII gender discrimination suit. The district court found that Mills was unable to prove that HCSC's proffered reasons for failing to promote him were pretextual. For the reasons set out below, we affirm the trial court's decision.

We give a brief outline of the facts here, and will discuss them in greater detail where relevant to our analysis. HCSC is a health care claims processing company with offices throughout Illinois. The plaintiff worked in a variety of divisions and positions at HCSC's Quincy office starting in 1988. In general, Mills received favorable work reviews.

In 1995, a co-manager of the Quincy office, Sandy Frillman resigned, leaving Linda Amburn as the office's sole manager. In response to Frillman's departure, the company created a new assistant manager position, which Amburn was responsible for filling. When the company posted the minimum requirements for the position, four individuals applied for the job: Teri Provine, Kathy Cluver, Darlene Butler, and the plaintiff. After Butler and the plaintiff were interviewed, Amburn offered the position to Butler. The plaintiff, citing his own qualifications, which he alleged were better than Butler's, sued, claiming gender discrimination. The district court granted summary judgment to the defendant, bringing us to this appeal.

We review the trial court's decision de novo, drawing all inferences in the light most favorable to the non-moving party--here, the plaintiff. Biolchini v. General Electric Co., 167 F.3d 1151, 1153 (7th Cir. 1999). Only if there are no genuine issues of material fact in dispute is summary judgment appropriate. Anderson v. Liberty Lobby, 477 U.S. 242, 255, 106 S. Ct. 2505, 91 L. Ed. 2d 202 (1986). In employment discrimination cases, we apply this standard with "added rigor" before granting summary judgment. Sarsha v. Sears, Roebuck & Co., 3 F.3d 1035, 1038 (7th Cir.1993).

Plaintiffs in employment discrimination cases can avert summary judgment in one of two ways. First, they can present direct evidence showing discriminatory intent by the defendant or its agents. Troupe v. May Dept. Stores, 20 F.3d 734, 736 (7th Cir.1994). A direct evidence assertion must be supported by allegations which, "if believed by the trier of fact, will prove the particular fact in question without reliance upon inference or presumption." Eiland v. Trinity Hospital, 150 F.3d 747, 751 (7th Cir.1998) (citations omitted).

Second, plaintiffs may offer indirect evidence that they have been discriminated against. This takes the form of the familiar three part burden shifting test established in McDonnell Douglas Corp. v. Green, 411 U.S. 792, 93 S. Ct. 1817, 36 L. Ed. 2d 668 (1973) and refined in Texas Dept. of Community Affairs v. Burdine, 450 U.S. 248, 101 S. Ct. 1089, 67 L. Ed. 2d 207 (1981). The McDonnell Douglas test requires that a plaintiff establish a prima facie case, satisfied by showing that the plaintiff: 1) is a member of a protected minority class or a female; 2) applied for, and was qualified for, an open position; 3) was rejected; and 4) the employer filled the position with a person not in the plaintiff's protected class, or the position remained open.1 EEOC v. Our Lady of the Resurrection Med. Ctr., 77 F.3d 145, 148 (7th Cir.1996). If the plaintiff establishes these elements, the burden shifts to the defendant to articulate legitimate, nondiscriminatory reasons for its employment action. Should the defendant meet its burden, the plaintiff then must cast doubt on whether these reasons are credible, or whether they are merely pretext. Courtney v. Biosound, Inc., 42 F.3d 414, 418 (7th Cir.1994).

Initially, we must determine whether this plaintiff meets the requirements of the prima facie test. Because Douglas Mills is a white male, he clearly does not satisfy prong one. Indeed, if strictly applied, the prima facie test would eliminate all reverse discrimination suits. However, the test is not so inflexible--it is well settled law that the protections of Title VII are not limited to members of historically discriminated-against groups. McDonald v. Santa Fe Trail Transp. Co., 427 U.S. 273, 280, 96 S. Ct. 2574, 49 L. Ed. 2d 493 (1976); Greenslade v. Chicago Sun-Times, Inc., 112 F.3d 853, 863 (7th Cir.1997); Hill v. Burrell Communications Grp., Inc., 67 F.3d 665, 667 (7th Cir.1995).

What is less settled is how to apply the traditional prima facie test to this situation. One option is to simply drop the first prong. On the other hand, at least two circuits use a different standard in "reverse discrimination" cases. See Taken v. Oklahoma Corp. Comm'n, 125 F.3d 1366, 1369 (10th Cir.1997); Harding v. Gray, 9 F.3d 150, 152-53 (D.C.Cir.1993).2 Those cases operate from the premise that "invidious discrimination against white [men] is relatively uncommon in our society, and so there is nothing inherently suspicious in an employer's decision to promote a qualified minority [or female] applicant instead of a qualified white [male] applicant." Harding, 9 F.3d at 153. This approach has also been justified because the presumption that arises once the McDonnell Douglas prima facie test is met--that unless otherwise explained discrimination is more likely than not the reason for the challenged decision--is "not necessarily justified when the plaintiff is a member of an historically favored group." Notari v. Denver Water Dept., 971 F.2d 585, 589 (10th Cir.1992).

Thus, these circuits have modified the prima facie test and added various substitutes (referred to as "background circumstances") for the burden imposed on minority or women plaintiffs to show that they are members of a protected class. Harding, 9 F.3d at 153. The contours of what constitutes a background circumstance are not precise. In Harding, the D.C. Circuit enumerated two categories of circumstances: "evidence indicating that the particular employer at issue has some reason or inclination to discriminate invidiously against whites [or men]," and "evidence indicating that there is something 'fishy' about the facts of the case at hand." Id. (citations omitted). Plaintiffs in that circuit have shown something was 'fishy' when they presented evidence of schemes to fix performance ratings to their detriment, that the hiring system seemed rigged against them because it departed from the usual procedures in an "unprecedented fashion," or that they were passed over despite superior qualifications. Id. at 154 (citations omitted). Another court, applying a variant of the McDonnell Douglas test in a reverse discrimination Bivens action, held that background circumstances could include situations in which: the person ultimately hired was clearly less qualified than the plaintiff, the hiring authority expressed intense interest in hiring a woman, and there was a pattern of hiring women in the past. Duffy v. Wolle, 123 F.3d 1026, 1036-37 (8th Cir.1997), cert. denied, --- U.S. ----, 118 S. Ct. 1839, 140 L. Ed. 2d 1090 (1998).3 Additionally, in Reynolds v. School Dist. No. 1, Denver, Colorado, the plaintiff successfully showed background circumstances where she was the only white employee in the department and nearly all of the decision makers were Hispanic. 69 F.3d 1523, 1534 (10th Cir.1995). All of these background circumstances "support an inference that the defendant is one of those unusual employers who discriminates against the majority," thus shifting the burden to the defendant to articulate a legitimate reason for its hiring decision. Taken v. Oklahoma Corp. Comm'n, 125 F.3d 1366, 1369.

These approaches are not meant to foreclose pursuit of legitimate Title VII claims by white men. Harding, 9 F.3d at 154 (background circumstances not an additional hurdle, but rather a substitute for minority plaintiff's burden under first prong of McDonnell Douglas test). Nothing in the modified prima facie formulations alters the fact that a majority plaintiff can always use direct evidence of prohibited discrimination to defeat a summary judgment motion. Id. Moreover, if a plaintiff fails to make a prima facie case under the modified test, he can demonstrate illegal discrimination by any other "indirect evidence sufficient to support a reasonable probability ... that but for [his] status [as a white male] the challenged employment decision" would not have occurred. Taken, 125 F.3d at 1369, quoting Notari, 971 F.2d at 590 (regardless of race, an employee who is the victim of intentional discrimination "who adduces sufficient evidence of that discrimination should be permitted to proceed beyond the prima facie stage of litigation"); see also Duffy, 123 F.3d at 1036 ("[J]ust because a reverse discrimination plaintiff cannot show background circumstances ... does not inexorably mean that his employer has not intentionally discriminated against him," and if that employee has evidence of discrimination he should be permitted to shift the burden to the defendant to prove a legitimate reason for the challenged employment decision.).

The defendant, HCSC, asks us to adopt one of these modified forms of the prima facie test. This court has never actually decided whether doing so is appropriate. Some of our opinions have noted the need for modifications to the McDonnell Douglas test. Kirk v. Board of Educ. of Bremen Community H.S. Dist., No. 228, 811 F.2d 347, 354 n. 10 (7th Cir.1987) (noting in dicta "At a minimum, it seems that some modification of the 'minority group' circumstance is necessary in a reverse discrimination case." Murray v. Thistledown Racing Club, Inc., 770 F.2d 63 (6th Cir.1985)); see also Ustrak v. Fairman, 781 F.2d 573, 577 (7th Cir.1986) (McDonnell Douglas framework designed to protect women and minorities, not whites); and Christensen v. Equitable Life Assur. Soc. of U.S., 767 F.2d 340, 343 (7th Cir.1985) (discussing distinction between using McDonnell Douglas paradigm to show discrimination against the minority as opposed to the majority).4

However, in cases where we have had the opportunity to formally adopt the test, we have not taken that step. In Hill v. Burrell Communications Grp., Inc., 67 F.3d 665 (7th Cir.1995), we assumed that the ordinary standard applied because "it was never meant to be applied rigidly." 67 F.3d at 668. We observed that we could not discern any reason for "applying one formula for the allocation of burdens of production in discrimination cases involving minorities and a different formula in reverse discrimination cases...." Id. at 668 n. 2. Nevertheless, we declined to resolve the question of the appropriate standard. Id.; see also Pilditch v. Board of Educ. of City of Chicago, 3 F.3d 1113, 1116 n. 1 (7th Cir.1993).

We think that it is now appropriate to answer this question. We agree with much of the analysis in Murray, Hill and Notari which operates from the presumption that it is the unusual employer who discriminates against majority employees.5 We also believe that if majority plaintiffs have to show less to prove their prima facie burden than minorities (who have historically suffered the type of discrimination Title VII sought to prevent), employers lose the "screening out benefits" that the prima facie test was intended to provide. See Jayasinghe v. Bethlehem Steel Corp., 760 F.2d 132, 134 (7th Cir.1985) (prima facie test serves to screen out cases "where the plaintiff fails to distinguish his or her case from the ordinary, legitimate kind of adverse personnel decision"). Thus, a plaintiff in a reverse discrimination case must show at least one of the background circumstances these other courts have alluded to. We note, however, that this modified test is not to be interpreted in a constricting fashion. It in no way precludes any plaintiff with direct evidence of discrimination from bringing his claim. Moreover, if a plaintiff cannot show background circumstances, but "has established a logical reason to believe that the [employer's] decision rests on a legally forbidden ground," such as his race or gender, he may shift the burden to the defendant to prove that the challenged employment action was actually based on legitimate, non-discriminatory reasons. See Carson v. Bethlehem Steel Corp., 82 F.3d 157 (7th Cir.1996) (per curiam).

Now that the nature of the prima facie test is properly defined we must determine whether Mills can meet the four prongs. The defendant does not dispute prongs two, three and four; it is clear that Mills met the minimum qualifications for the position of associate manager, he was not hired, and a woman, Darlene Butler, was selected for the job. The only question is whether the plaintiff can show any background circumstances which give rise to an inference of discrimination. The defendant asserts that because Mills cannot demonstrate that his qualifications were clearly superior to Butler's, the plaintiff cannot establish a prima facie case.

This is too narrow a view of background circumstances. While HCSC's contention is true, Mills can pursue other avenues in support of his claim. Mills points to "disproportionate hiring patterns" favoring women at HCSC's Quincy office. Between 1988-1995, nearly all promotions at the office went to women, and at the time the challenged hiring decision was made, females dominated the supervisory positions in the relevant office. As other cases have held, these facts, while far from actual evidence of discrimination, are enough to overcome the background presumption that a white man was not subject to employment discrimination. Reynolds v. School Dist. No. 1, Denver, Colorado, 69 F.3d 1523, 1534 (10th Cir.1995); Daye v. Harris, 655 F.2d 258, 261 (D.C.Cir.1981). Thus the burden shifts to the defendant to articulate legitimate, non-discriminatory reasons for its decision to hire Butler instead of Mills.

HCSC offers just such reasons in defense of its decision to hire Butler instead of Mills. The company avers that Butler had clearly superior qualifications to Mills, and was a better fit for the position of associate manager.

Regarding the "superior qualifications" HCSC notes that while the two candidates have comparable business degrees, Butler had an additional Associate's degree in computer science. The defendant also points to Linda Amburn's belief that Butler had superior qualifications in the areas of creativity, flexibility, innovativeness, interpersonal skills, leadership qualities, communications skills, teamwork and product knowledge. These beliefs were culled from both oral and written interviews. Amburn testified in depositions that she was more impressed with Butler's style and content in both types of interviews. In particular, she pointed to answers to a written interview question asking the candidates to propose structural changes in the office, and the fact that in the oral interview, Butler smiled more than Mills, who did not appear very happy, and had a stern countenance.

Amburn also believed that the breadth of Butler's experience was greater than the plaintiff's, because, although they had both been at HCSC for eight years, Butler had worked her way up to supervisor while Mills remained in the same position throughout. Additionally, Amburn pointed to Butler's success as an "inquiry supervisor," a position Amburn believed was one of the most important and difficult at the company. Although Mills had some experience as an inquiry supervisor, his tenure was for a shorter duration and, apparently, less successful. Finally, HCSC presents other reasons to buttress its contentions. However, because these reasons are peripheral to its central contention that it made its decision based on Butler's superior qualifications, and we need not address them.

Once an employer articulates legitimate, nondiscriminatory reasons for the employee's termination, the employee must present enough evidence to support a finding that the stated reasons are pretextual. Wolf v. Buss (America) Inc., 77 F.3d 914, 919 (7th Cir.1996). "Because a Title VII claim requires intentional discrimination, the pretext inquiry focuses on whether the employer's stated reason was honest, not whether it was accurate." Helland v. South Bend Community Sch. Corp., 93 F.3d 327, 330 (7th Cir.1996), cert. denied, 519 U.S. 1092, 117 S. Ct. 769, 136 L. Ed. 2d 715 (1997). In short, Mills has the burden of showing that HCSC's reasons for hiring Butler were either lies (see Russell v. Acme-Evans Co., 51 F.3d 64, 68 (7th Cir.1995)) or had no basis in fact (see Johnson v. City of Fort Wayne, 91 F.3d 922, 931 (7th Cir.1996)).

As Mills correctly contends, at summary judgment the plaintiff is not required to establish pretext and provide evidence of a discriminatory motive by the defendant, as the district court mistakenly apprehended. This level of proof is only required when a plaintiff's case is submitted to a finder of fact. St. Mary's Honor Center v. Hicks, 509 U.S. 502, 508, 113 S. Ct. 2742, 125 L. Ed. 2d 407 (1993). The plaintiff who survives summary judgment and goes to trial may be "well advised to present additional evidence of discrimination [beyond pretext] because the factfinder is not required to find in [the plaintiff's favor] simply because she establishes a prima facie case and shows that the employer's proffered reasons are false." Anderson v. Baxter Healthcare Corp., 13 F.3d 1120, 1124 (7th Cir.1994) (emphasis omitted) (citation omitted). However, as we have held, for a plaintiff to prevail on a defendant's motion for summary judgment, he need not show "pretext plus," and "he must only 'produce evidence from which a rational trier of fact could infer that the company lied' about its proffered reasons for his dismissal .... [i]f an inference of improper motive can be drawn, there must be a trial." Id. (emphasis omitted) (citations omitted); see also Loyd v. Phillips Bros., Inc., 25 F.3d 518, 522 (7th Cir.1994).

Mills argues that the decision to promote Butler because she was better qualified for the position is suspect for several reasons. First, he contends that HCSC's reliance on Butler's background as an "inquiry supervisor" is pretextual. This is pretext, according to Mills, because his performance in the same role was "well above average," and he was "highly effective in that position." This is hardly persuasive evidence of pretext, because even if HCSC was mistaken in evaluating the plaintiff's qualifications, Mills does not show that this evaluation was not honestly undertaken. Lenoir v. Roll Coater, Inc., 13 F.3d 1130, 1133 (7th Cir.1994). The only exception to this would be where the employer's evaluation was so egregiously mistaken that it had no basis in fact (Kaniff v. Allstate Ins. Co., 121 F.3d 258, 265 (7th Cir.1997)), but this is far from that case. More persuasive is the plaintiff's allegation that the company's contention that experience in the inquiry unit was an important factor in its decision is pretext. As Mills argues, two women who were previously promoted over him into assistant manager positions had no background at all in this area, while he had significant experience in this field. On summary judgment, where we must draw all reasonable inferences in the plaintiff's favor, we agree with him that this appears odd.

Even if this created a genuine issue of material fact about whether using inquiry unit experience was pretextual (although we are not convinced that it does), it would provide Mills with little succor. Because the defendant offered multiple reasons for its decision, Mills must show that all were pretextual in order to reverse the district court's grant of summary judgment. Wolf v. Buss (America) Inc., 77 F.3d 914, 919 (7th Cir.1996). He cannot show that the reams of evidence including appraisals of past experience, written evaluations, and the results of oral interviews were anything other than HCSC's honest opinion of which employee was a better fit for the assistant manager position. In the absence of such a showing, even if these evaluations were wrong, we do not sit as a "super-personnel department" with plenary authority to review HCSC's business decisions. Debs v. Northeastern Illinois University, 153 F.3d 390, 396 (7th Cir.1998).

Mills tries a different tack in arguing that HCSC's failure to articulate consistent reasons for the decision to hire Butler establish pretext. Immediately following the plaintiff's interview with Linda Amburn, she prepared an evaluation of him ranking him in ten categories, and included written comments. On a 1 to 5 scale, with 1 being the worst, Mills received two 2s (below average), six 3s (average) and one 4 (above average). In the written section, Amburn observed that she did not believe that assistant manager was truly the position that Mills sought. Mills avers that this is inconsistent with what Amburn told him at a meeting on August 28, 1995--two weeks after Butler was hired, when Amburn did not pinpoint any specific deficiencies in his performance, and failed to identify areas where he needed improvement. He also alleges these reasons do not comport with the "superior qualifications" rational asserted in this case. Mills asks us to draw an inference that a genuine issue exists as to whether these "inconsistencies" comprise pretext.

The district court did not draw such an inference. It observed that Amburn might not have told Mills about his performance deficiencies at the meeting because she did not want to embarrass the plaintiff, or she might not have wanted to talk about the matter. However, since Mills was the non-moving party on summary judgment, it was improper for the district court to draw such inferences in the defendant's favor. On the other hand, while any inferences drawn from the facts must be viewed in the light most favorable to the non-moving party, only reasonable inferences need be made. Felce v. Fiedler, 974 F.2d 1484, 1488 (7th Cir.1992). Particularly because Mills conceded that the purpose of the meeting with Amburn was not to find out the reasons why Butler was promoted, but instead to ascertain what he could to improve his chances in the future, we believe that this evidence is a wash, and is not inconsistent with what was said on any other occasions. Additionally, we see no inconsistencies between the plaintiff's low scores on his interview evaluation and HCSC's proffered reason that it hired Butler because she had superior qualifications. Drawing an inference of pretext from this evidence would be unreasonable.

Reviewing all of the evidence,6 we find that the plaintiff cannot show that the reasons HCSC proffered for hiring Darlene Butler over him were pretextual.

Although we disagree with part of the district court's analysis, its ultimate conclusion was correct. Accordingly, the grant of summary judgment to the defendant is

AFFIRMED.