Dr. Zoe Spencer, a sociology professor at Virginia State University, sued the University under the Equal Pay Act and Title VII for paying her less than two male professors, allegedly because she is a woman.

Spencer earned about $ 70,000 per year-a median salary when compared to the men who were also full professors in the Department of Sociology, Social Work, and Criminal Justice. But Spencer's lawsuit proposes comparing her pay to that of two former University administrators, Drs. Michael Shackleford and Cortez Dial, who each earned over $ 100,000 per year as professors in other departments. While Spencer asserts that the difference in pay was due to her sex, the University provides a different explanation: Shackleford's and Dial's jobs differed from Spencer's and, as former administrators, their pay was set as a prorated portion of their previous salaries.

After discovery, the district court granted summary judgment for the University (and its former president, Dr. Keith Miller). We affirm. Though Spencer establishes a pay disparity, she fails to present evidence that creates a genuine issue of material fact that Shackleford and Dial are appropriate comparators. In any event, unrebutted evidence shows that the University based Shackleford's and Dial's higher pay on their prior service as University administrators, not their sex. 1

I. Equal Pay Act

Spencer first claims that the disparity between her salary and her chosen comparators' violates the Equal Pay Act. The statute forbids the University (like other employers) from:

Discriminat[ing] ... between employees on the basis of sex by paying wages to employees ... at a rate less than the rate at which [the employer] 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, except where such payment is made pursuant to (i) a seniority system; (ii) a merit system; (iii) a system which measures earnings by quantity or quality of production; or (iv) a differential based on any other factor other than sex ....

29 U.S.C. § 206 (d)(1). To prove a violation of the Act, Spencer must make an initial ( i.e. , prima facie) showing of three elements: (1) the University paid higher wages to an employee of the opposite sex who (2) performed equal work on jobs requiring equal skill, effort, and responsibility (3) under similar working conditions. EEOC v. Maryland Ins. Admin. , 879 F.3d 114 , 120 (4th Cir. 2018) (citing Corning Glass Works v. Brennan , 417 U.S. 188 , 195, 94 S.Ct. 2223 , 41 L.Ed.2d 1 (1974) ).

This initial showing permits an inference that a pay disparity was based on sex discrimination. Maryland Ins. Admin. , 879 F.3d at 120 . The inference of discrimination stands even without the support of any evidence of discriminatory intent. Id. Only once this inference exists does the burden shift to the employer to show that the pay differential was based on a factor other than sex. Id .

Spencer's choice of Shackleford and Dial as comparators establishes the first element of her initial showing-the existence of a wage differential. By choosing two of the highest-paid professors at the University, Spencer ensured that her wages were much lower. Yet that same decision to pick Shackleford and Dial precludes her from establishing, as the second element requires, that she and they performed "equal" work requiring "equal skill, effort, and responsibility."

Equality under the Act is a demanding threshold requirement. It requires a comparator to have performed work "virtually identical" (or the apparent synonym, "substantially equal") to the plaintiff's in skill, effort, and responsibility.

Wheatley v. Wicomico Cty. , 390 F.3d 328 , 332-33 (4th Cir. 2004). Similarity of work is not enough; the Act explicitly distinguishes between the work itself (which must be "equal") and the conditions of work (which need only be "similar"). 29 U.S.C. § 206 (d)(1). The Act does not provide courts with a way of evaluating whether distinct work might have "comparable" value to the work the plaintiff performed. See Wheatley , 390 F.3d at 333 ; see also Sims-Fingers v. City of Indianapolis , 493 F.3d 768 , 771 (7th Cir. 2007) (Posner, J.) (explaining that, when trying to identify "comparable" pay for unequal work, there are "no good answers that are within the competence of judges to give"). Instead, the Act's inference of discrimination may arise only when the comparator's work is equal to the plaintiff's.

In alleging this necessary equality, a plaintiff may not rely on broad generalizations at a high level of abstraction. Wheatley , 390 F.3d at 332 . But Spencer attempts just such an impermissibly general comparison. In Spencer's view, all University professors perform equal work because they all perform the same essential tasks: preparing syllabi and lessons, instructing students, tracking student progress, managing the classroom, providing feedback, and inputting grades. See Appellant's Brief at 9. The performance of these tasks, Spencer posits, requires the same skills: studying, preparing, presenting, discussing, and so forth. See id. at 9-10. But these tasks and skills are shared by middle-school teachers and law-school professors, pre-algebra teachers and biomedical-engineering professors.

This attempted comparison ultimately relies on the common title of "professor" plus some generalized responsibilities ( e.g. , teaching students). Yet we have rejected an analogous claim that jobs with the same title and only vaguely corresponding responsibilities can be considered equal. In Wheatley we concluded that the plaintiffs, supervisors in a county's emergency-services department, failed to meet their burden to show that supervisors in different departments performed equal work because they could not demonstrate that the different jobs were equal in skill and responsibility. 390 F.3d at 334 ; see also Sims-Fingers , 493 F.3d at 771 . Spencer's case suffers from a near-identical flaw.

Spencer's bird's-eye view is particularly unpersuasive given the inherent features of academia. Professors are not interchangeable like widgets. Various considerations influence the hiring, promotion, and compensation of different professorial jobs. Cf. Zahorik v. Cornell Univ. , 729 F.2d 85 , 93 (2d Cir. 1984) (discussing the tenure process). As a result, faculty salary decisions require a complex balancing of factors. Among other things, those decisions account for the differences in skill and responsibility attendant to different jobs. For instance, an engineering professorship requires different skills, effort, and responsibility than professorships in other fields, such as sociology. Evidence offered by Spencer proves this very point: The University systematically pays engineering professors more than humanities professors. J.A. at 136. This reflects differences in skill along with market forces that compensate engineers more highly. This market reality confirms that Spencer's broad generalizations about tasks and skills, which apply to virtually all teachers, fail to satisfy her burden to show equal work.

In contrast to Spencer's generalized tasks and skills, a litany of concrete differences underscore that Spencer does not perform work equal to that of Shackleford and Dial. First, Shackleford and Dial taught in different departments than Spencer did. While comparisons might be drawn between some departments, any such comparison requires the plaintiff to articulate with specificity why the work performed and skills needed by a professor in one department are virtually identical-and not just generally related or of comparable worth-to those in another. As our precedents recognize, the differences between academic departments generally involve differences in skill and responsibility. See Strag v. Bd. of Trustees, Craven Cmty. Coll. , 55 F.3d 943 , 950 (4th Cir. 1995) (finding that the Biology and Mathematics departments required instructors to have different skills and responsibilities); Soble v. Univ. of Maryland , 778 F.2d 164 , 167 (4th Cir. 1985) (finding that the specialized nature of certain university departments called for distinct skills); cf. Wheatley , 390 F.3d at 332-33 (recognizing job-related differences for directors of different county government departments).

There are still more differences. Along with serving in different departments, the three professors taught at different class levels at the University. Spencer taught mainly undergraduate courses, while Shackleford and Dial taught more graduate courses. And Shackleford also supervised doctoral dissertations. Contrary to Spencer's assertion, the fact that the University credited hours spent supervising dissertations in a similar manner to hours spent teaching regular courses does nothing to establish the equivalency of supervising dissertations and teaching undergraduates. Nor did the professors work equal hours, as the record shows that Shackleford and Dial worked more than Spencer did week to week. 2

None of this is to say that the Equal Pay Act cannot apply in the higher-education context. But in that context-one where the work is an exercise in intellectual creativity that can be judged only according to intricate, field-specific, and often subjective criteria-Spencer must provide the court with more than broad generalities to meet her burden. She must present evidence on which a jury could rely to decide that she, Shackleford, and Dial had equal jobs, not just that they all performed vaguely related tasks using nominally comparable skills. That is, there must be evidence showing the jobs were equal in the strict sense of involving "virtually identical" work, skill, effort, and responsibility, not in the loose sense of having some comparative value. Wheatley , 390 F.3d at 333 .

Despite all of these issues, Spencer claims her expert, Joseph Rosenberg, found "that Shackleford and Dial were significantly overpaid in comparison to Spencer ." Appellant's Brief at 44 (emphasis added). Not only is this irrelevant to establishing equal work, this claim is a bit misleading: Rosenberg asserted that Shackleford and Dial were overpaid relative to all other professors, both men and women. 3

In addition to looking at her chosen comparators, Spencer's expert tried to identify a general disparity between the pay of men and women at the University. But his efforts revealed no statistically significant disparity within each "school." If anything, this evidence undermines Spencer's claimed inference of discrimination. See Strag , 55 F.3d at 950 (suggesting that "isolated incidents or random comparisons demonstrating disparities in treatment may be insufficient to draw a prima facie inference of discrimination without additional evidence that the alleged phenomenon of inequality also exists with respect to the entire relevant group of employees " (quoting Houck v. Virginia Polytechnic Ins. , 10 F.3d 204 , 206-07 (4th Cir. 1993) )).

Despite her expert's efforts, Spencer's generalized claims cannot establish that she engaged in equal work, which categorically dooms her attempt to establish wage discrimination under the Equal Pay Act. Cf. Wheatley , 390 F.3d at 334 (a plaintiff may not "indiscriminately aim at all department supervisors collectively, and then expect to meet the EPA standard" for equal work).

But even if Spencer could meet her initial burden, her claim would still fail because the University established that the salary difference was based on a "factor other than sex." 29 U.S.C. § 206 (d)(1)(iv). As the defendant, the University bore the burden of establishing this affirmative defense. Maryland Ins. Admin. , 879 F.3d at 120 . Granting summary judgment on this ground required the district court to find that the proffered reason did in fact explain the wage disparity, not merely that it could . Id. at 121 .

Here, there is no dispute that the wage difference at issue resulted from the University setting Shackleford's and Dial's pay at 75% of their previous salaries as administrators. In practice, the University generally paid former administrators who became professors "9/12ths" of their administrator salary. This practice appears to rest on the theory that professors work nine months out of the year, while administrators work year-round. Indeed, Spencer admits that her comparators' pay during their short stints as University professors was set according to the 9/12ths practice. Appellant's Reply Brief at 16 ("Shackleford and Dial's salaries were set entirely on their prior salaries as administrators.").

In response to the University's explanation, Spencer claims that the 9/12ths practice should not have been used to calculate Shackleford's and Dial's salaries. According to her, the University's historical practice only applied to administrators who were previously tenured faculty. But even if the University erroneously applied its 9/12ths practice to overpay Shackleford and Dial, such an imprudent decision would still serve as a non-sex-based explanation for the pay disparity. See Anderson v. Westinghouse Savannah River Co. , 406 F.3d 248 , 272 (4th Cir. 2005) ("We do not sit as a 'super-personnel department weighing the prudence of employment decisions' made by the defendants." (quoting DeJarnette v. Corning, Inc. , 133 F.3d 293 , 299 (4th Cir. 1998) )); Smith v. Univ. of North Carolina , 632 F.2d 316 , 346 (4th Cir. 1980) ("[The] law does not require, in the first instance, that employment be rational, wise, or well-considered-only that it be nondiscriminatory." (quoting Powell v. Syracuse Univ. , 580 F.2d 1150 , 1156-57 (2d Cir. 1978) )).

The Equal Pay Act is a powerful tool, permitting an employee to prevail on a wage discrimination claim with no evidence of intentional discrimination. But this tool must be tempered by adherence to its provisions. Doing so requires that the work performed by the plaintiff and her comparators be equal and that the wage disparity not be based on a factor other than sex. Spencer's claim fails on both requirements.

II. Title VII

Having rejected Spencer's Equal Pay Act claim, we must separately consider her claim of Title VII sex-based wage discrimination. Under Title VII of the Civil Rights Act of 1964, an employer cannot "discriminate against any individual with respect to [her] compensation ... because of such individual's ... sex." 42 U.S.C. § 2000e-2(a)(1). Title VII, in contrast to the Equal Pay Act, requires establishing intentional discrimination. A Title VII plaintiff may make this showing of intentional discrimination using direct or circumstantial evidence. Alternatively, the plaintiff may use the burden-shifting framework of McDonnell Douglas Corp. v. Green , 411 U.S. 792 , 93 S.Ct. 1817 , 36 L.Ed.2d 668 (1973), "to develop an inferential case" of discriminatory intent. Chalmers v. Tulon Co. of Richmond , 101 F.3d 1012 , 1017 (4th Cir. 1996).

A prima facie pay-disparity case under McDonnell Douglas requires a plaintiff to establish (1) she is a member of a protected class, (2) she was performing her job satisfactorily, (3) an adverse employment action occurred, and (4) the circumstances suggest an unlawfully discriminatory motive. See McDonnell Douglas , 411 U.S. at 802 , 93 S.Ct. 1817 ; Texas Dep't of Cmty. Affairs v. Burdine , 450 U.S. 248 , 253, 101 S.Ct. 1089 , 67 L.Ed.2d 207 (1981). Where, as here, the prima facie case of wage discrimination is based on comparators, the plaintiff must show that she is paid less than men in similar jobs. See Brinkley-Obu v. Hughes Training, Inc. , 36 F.3d 336 , 343 (4th Cir. 1994) .

Title VII requires the compared jobs to be only "similar" rather than "equal," as required under the Equal Pay Act. See id. While there is no bright-line rule for what makes two jobs "similar" under Title VII, courts consider "whether the employees (i) held the same job description, (ii) were subject to the same standards, (iii) were subordinate to the same supervisor, and (iv) had comparable experience, education, and other qualifications-provided the employer considered these latter factors in making the personnel decision." Bio v. Fed. Express Corp. , 424 F.3d 593 , 597 (7th Cir. 2005) (internal quotation marks omitted); see also Herster v. Bd. of Supervisors of Louisiana State Univ. , 887 F.3d 177 , 185 (5th Cir. 2018) (stating that a "variety of factors are considered when determining whether a comparator is similarly situated, including job responsibility, experience, and qualifications."). While Title VII's "similarity" requirement demands less of plaintiffs than the Equal Pay Act's "equality" requirement, it is not toothless: the plaintiff must provide evidence that the proposed comparators are not just similar in some respects, but "similarly-situated in all re spects ." Mitchell v. Toledo Hosp. , 964 F.2d 577 , 583 (6th Cir. 1992). For the same reasons we discussed above, Spencer's broad generalizations cannot even show sufficient similarity to meet her burden under Title VII. 4

Even if we concluded that Spencer had established a prima facie case of Title VII wage discrimination, her case still could not withstand summary judgment. Once a plaintiff establishes a prima facie case, the burden of production shifts to the employer to proffer a legitimate, nondiscriminatory explanation for the wage disparity. Guessous v. Fairview Prop. Investments, LLC , 828 F.3d 208 , 216-17 (4th Cir. 2016) ; Maryland Ins. Admin. , 879 F.3d at 120 n.7. Here, the University satisfies this requirement through its practice of paying administrators 9/12ths of their previous salary. Just as this practice satisfies the Equal Pay Act's "factor other than sex" affirmative defense, it qualifies as a legitimate, nondiscriminatory explanation under Title VII. Cf. Cty. of Washington v. Gunther , 452 U.S. 161 , 169, 101 S.Ct. 2242 , 68 L.Ed.2d 751 (1981) (recognizing that the Bennett Amendment to Title VII incorporates the four affirmative defenses from the Equal Pay Act).

Having proffered a nondiscriminatory explanation, the University shifts the burden back to Spencer to prove that the explanation is merely pretextual for invidious discrimination. Guessous , 828 F.3d at 216 . Spencer cannot supply any evidence of this. Since the touchstone of discrimination is treating two groups differently based on characteristics only one possesses, it is vital for Spencer to provide evidence that the University has used the 9/12ths policy to pay men more than women. Instead, Spencer merely argues that the University misapplied the 9/12ths policy to Shackleford and Dial. But again, even if the University "erroneously or even purposely misapplied the ... policy, it is not proof of unlawful discrimination." Dugan v. Albemarle Cty. Sch. Bd. , 293 F.3d 716 , 722 (4th Cir. 2002).

As a final note, Spencer also alleges that the University, and its former provost, engaged in unlawful retaliation because of her complaints about pay disparities. As the district court noted, the facts supporting most of these allegations are exceedingly weak, and the allegations themselves are mostly conclusory. See Spencer v. Virginia State Univ. , No. 3:16-cv-989, 2018 WL 627558 , at *14-17 (E.D. Va. Jan. 30, 2018). Even if Spencer's characterization of the behavior is accurate, Spencer offers insufficient evidence that each action was both material and undertaken because of her complaints about salary equity. See Burlington N. & Santa Fe Ry. Co. v. White , 548 U.S. 53 , 68, 126 S.Ct. 2405 , 165 L.Ed.2d 345 (2006) (holding that "a plaintiff must show that a reasonable employee would have found the challenged action materially adverse ... [meaning] it might have 'dissuaded a reasonable worker from making or supporting a charge of discrimination' "). That is not to say no one harbored animus toward Spencer, but a personal conflict alone does not constitute retaliation. Cf. Hawkins v. PepsiCo, Inc. , 203 F.3d 274 , 281 (4th Cir. 2000) ("But Hawkins presents no facts that tend to show this allegedly disparate treatment was due to race rather than Price's admittedly low regard for Hawkins' individual performance."). Because the district court correctly found that Spencer cannot establish a prima facie case of retaliation, we do not address the merits of the University's defenses.

Accordingly, the judgment of the district court is

AFFIRMED .