In this article and in the ones to follow in the next few months 1, I will attempt to focus the discussion on those absurdities, anomalies, uncertainties, and paradoxes. My hope is to move the debate on discrimination in pay out of its current platitude, confusion, and direct the inquiry and the focus on new promising paths.
In this article, I will try to answer the question whether disparate impact theory of liability is available as a matter of law for a plaintiff or an enforcement agency to make a case of sex-based discrimination in compensation 2 under the Equal Pay Act of 1963 (“EPA” hereinafter) and/or under Title VII of the Civil Rights Act of 1964. I will approach this question from the perspective of the enforcement agencies (OFCCP and EEOC) and the Supreme Court, showing where the position of the enforcement agencies conflict with the Supreme Court relevant jurisprudence.
But before we try to answer this main question, I will briefly present the current state of the law of disparate impact. Since EO 11246 follows federal court interpretation of Title VII, it is taken for granted that all references to Title VII apply to EO 11246.
Disparate Impact in Title VII and EO 11246
Under Title VII of the Civil Rights Act of 1964, it is unlawful for an employer to “limit, segregate, or classify his employees or applicants for employment in any way which would deprive or tend to deprive any individual of employment opportunities or otherwise adversely affect his status as an employee, because of such individual’s race, color, religion, sex or national origin.” 42 U.S.C. § 2000e-2(a)(2).
This prohibition against disparate impact is distinct from disparate treatment by an employer, which requires a showing of discriminatory intent. Under Section 2000e-2(a)(2), an otherwise facially neutral business practice that disproportionately affects or impacts a protected group may be unlawful. Griggs v. Duke Power Co., 401 U.S. 424 (1971).
Reacting to the Supreme Court dismantling of the disparate impact theory of liability in Wards Cove Packing Co. v. Atonio, 490 U.S. 642 (1989) and Watson v. Fort Worth Bank & Trust, 487 U.S. 977 (1988), Congress codified disparate impact theory of liability in the Civil Rights Act (CRA) of 1991. Section 703 of Title VII of the CRA of 1991 provides:
- (1)(A) An unlawful employment practice based on disparate impact is established under this subchapter only if:
- (i) a complaining party demonstrates that a respondent uses a particular employment practice that causes a disparate impact on the basis of race, color, religion, sex, or national origin and the respondent fails to demonstrate that the challenged practice is job related for the position in question and consistent with business necessity; or
- (ii) the complaining party makes the demonstration described in subparagraph (C) with respect to an alternative employment practice and the respondent refuses to adopt such alternative employment practice.
Phase 1 (Prima Facie Case): To establish a prima facie case of disparate impact in a Title VII case, a plaintiff must (1) identify a specific employment policy or practice of the employer and (2) proffer evidence, typically statistical evidence, (3) of a kind and degree sufficient to show that the practice in question has caused exclusion of applicants for jobs or employees for promotions (4) because of their membership in a protected group.
Recently in Texas Department of Housing and Community Affairs v. Inclusive Communities Project, Inc., 2015 WL 2473449 (2015), the Supreme Court held that a disparate-impact claim that relies on a statistical disparity must fail if the plaintiff cannot point to a defendant’s policy or policies causing that disparity. Notably, the Supreme Court explained a policy is more than just a one-time decision – which may not necessarily be classified as a policy at all.
The Court went as far as stating that a “plaintiff who fails to allege facts at the pleading stage or produce statistical evidence demonstrating a causal connection cannot make a prima facie case of disparate impact.” Id. at *15. The Supreme Court explained that a robust causation requirement ensures that racial imbalance does not, in and of itself, establish a prima facie case of disparate impact and thus protects defendants from being held liable for racial disparities they did not create. The Supreme Court warned that without adequate safeguards at the prima facie stage, disparate-impact liability might cause race to be used and considered in a pervasive way and “would almost inexorably lead” governmental or private entities to use “numerical quotas,” and serious constitutional questions then could arise. The highest court instructed lower courts that they must examine with care whether a plaintiff has made out a prima facie case of disparate impact and favor a prompt resolution of these cases. In the Supreme Court’s view, the higher bar that must be set for the prima facie case of disparate impact-liability is necessary to protect potential defendants against abusive disparate-impact claims. Otherwise, businesses will be prevented from achieving their legitimate objectives and the free-market system will suffer from this interference.
It is clear that the Supreme Court is restricting the role, the weight and the probative value of statistics in establishing causation in disparate impact cases. Statistics are more critical and determinative in establishing causation at the prima facie phase, and ultimately to the question of liability in disparate impact cases, but statistical imbalance does not automatically satisfy legal causation.
The Supreme Court is also insisting on the importance of linking causation to the actions of a specific employer or defendant. An employer can’t be held liable for inequalities and discrimination that exist in society that the employer did not create or cause. For example, if access to quality education is unequal or discriminatory in society, an employer’s use of education as a factor to determine pay may have a statistical adverse impact but as long as the employer has consistently applied the education factor, it should not be held liable. This position obviously conflicts with the teaching of Griggs, but as we will show below, it is the correct interpretation of the law in sex-based discrimination in compensation.
Phase 2: If a plaintiff succeeds in making a prima facie case of disparate impact, the burden shifts to the employer to show that the employment practice at issue is job related for the position in question and is consistent with business necessity. As the Supreme Court explained in Ricci v. DeStefano, an entity “could be liable for disparate-impact discrimination only if the [challenged practices] were not job related and consistent with business necessity.” 557 U.S. 557, at 587 (2009). See also Texas Department of Housing and Community Affairs, 2015 WL 2473449 (2015) (an employer may maintain a workplace requirement that causes a disparate impact if that requirement is a reasonable measurement of job performance or necessary to achieve a valid interest).
Phase 3: Although an employer may successfully show that a test or an employment practice is job related and consistent with business necessity, the plaintiff may still prevail if it can show that there is an available alternative employment practice with less adverse impact that serves the employer’s legitimate needs and that the employer refused to adopt it. See Ricci, 557 US 557 (2009). As a general matter, there is no case where a plaintiff won a disparate impact case on the basis of the alternative argument alone. The 6th Circuit overturned the only ruling in which the alternative argument was the sole basis for a plaintiff victory. (See Johnson v. City of Memphis, 2014 U.S. App. LEXIS 20644).
What Constitutes a Compensation Adverse Action or Injury?
Almost all adverse employment actions will have an impact on an employee’s pay. When an employer fails to hire or to promote an employee, the employee will be harmed, injured or adversely affected in the form of less or no compensation. When an employer steers an employee towards a lower paying job, the compensation of the employee will be adversely affected. Similarly, when an employer terminates the employment of an employee, the latter will not receive any more paychecks. In these types of employment actions, injury is measured (among other things) by less or no pay. However, there are not, per se, compensation adverse actions or injuries.
What is an injury or adverse action in pay? Does it have something to do with differences in pay? What difference in pay could be considered an adverse action? Is paying a woman and a man, performing the same job, different salaries an adverse compensation action? Can a woman earning substantially higher pay or equal pay than all the men in her job group still claim an adverse compensation action and have standing to bring a claim against her employer? Could the female employee who is the highest paid employee in her job group claim a concrete, particularized injury, or injury in fact?
Questions regarding a compensation adverse action or injury are not as straightforward as other types of employment actions (e.g. failure to hire, failure to promote, terminations). Differences in pay (negative as well as positive) or zero difference between an employee and his/her comparator may or may not constitute an actionable harm or injury. The controlling factor in determining a compensation injury is not the difference or no difference in pay (the dependent variable), but how an employee or group of employees are “rated” under the other independent factors or variables that influence or correlate with pay. Presence of injury, harm or adverse compensation action is not determined on the basis of pay considered in isolation; rather it is based on the interaction of pay with the relevant factors that determine and influence pay.
To illustrate this point I will use a simple example with three scenarios. Let’s consider the hypothetical case of an employer that has a policy and a practice of offering all new hires exactly the same starting pay and an annual pay increase based on a fixed percentage. Under this employer’s policy and practice the pay is determined by time with the company and time with the company only. Three scenarios could happen:
- Scenario #1: If a female employee earns 10% less than her male comparator, but has more or equal time with the company than her male comparator, she could claim an injury or adverse compensation action.
- Scenario #2: If a female employee earns an equal amount of pay as her male comparator, but has more time with the company than her male comparator, she could claim an injury or adverse compensation action.
- Scenario #3: If a female employee earns 10% more than her male comparator, but has proportionately higher percentage (e.g. 50%) of time with the company than her male comparator, she could claim an injury or adverse compensation action.
To our knowledge, plaintiff’s bar or federal enforcement agencies have not taken advantage of the broader possibilities that Title VII offers in this area over the Equal Pay Act (EPA). Accepting that all three scenarios could constitute a compensation injury will require many to revise their methodological assumptions and models when analyzing compensation. However, under the EPA, the language of the statute limits actionable claims to scenario #1, where plaintiffs must succeed in making a prima facie case by demonstrating that they performed work substantially equal to that of someone of the opposite sex (the comparator) and that they were paid less than the comparator.
The Building Blocks of Disparate Impact Analysis Applied to Pay
When we explore the application of disparate impact to pay, what exactly are we talking about? Are we talking about applying disparate impact to base pay (a mega dependent variable in its own right) or are we talking about applying disparate impact to the independent variables that determine base pay jointly or severally?
This issue will be addressed in a future publication; it is worth mentioning though that OFCCP is now talking about auditing W-2 earnings (the super-mega dependent variable) that aggregate many discrete mega dependent variables (e.g. base pay; overtime pay; allowances and reimbursements; signing bonuses; retention bonuses; performance bonuses; equity adjustments; deferred compensation plans; profit-sharing; etc.). In this case, the question becomes: should we be talking about applying disparate impact jointly or severally to the different forms and types of earnings listed in the IRS code that are reported on a W-2? Each type of compensation will be determined by a number of explanatory factors or independent variables – consequently how would a disparate impact analysis apply to all these sets of independent variables?
We will address these questions in a forthcoming article, but suffice to say that legal causation will be forever lost in this maze. And frankly, the issues that exist when applying disparate impact analysis to base pay will be compounded when pay is defined as W-2 earnings that include so many discrete types of incomes. Additional issues will be raised by the blender of aggregating discrete events that will make it impossible to infer any legal causation. This aggregation of earnings, added to the aggregation of dissimilar jobs (discussed below), will make the OFCCP drift even further from the Supreme Court insistence on the principle of causation. For the Supreme Court, a statistical imbalance alone is not enough to meet the necessary causal nexus needed for the prima facie case. On the other hand, OFCCP is entering into conciliation agreements on the basis of statistical imbalances alone.
Causation and aggregation are relatively small issues when compared to the effect of the Supreme Court holdings in County of Washington v. Gunther and Smith v. City of Jackson. In the sections below we will show that, as a matter of law, disparate impact is not available in sex-based discrimination in compensation as well as where the OFCCP and the EEOC positions conflict with federal law.
The Poison Pill of County of Washington v. Gunther
Despite the strong dissent by Justice Rehnquist, describing the narrowness of the majority decision as a “restricted railroad ticket, good for this day and train only,” County of Washington v. Gunther 452 U.S. 161 (1981) is an important case in the history of anti-discrimination laws in compensation. The Court resolved a seventeen-year-old controversy involving interpretations of the Equal Pay Act of 1963, Title VII of the Civil Rights Act of 1964, and the Bennett Amendment to Title VII, without having to define the precise contours of lawsuits challenging sex-based discrimination in compensation under Title VII.
County of Washington v. Gunther has not been always read fairly or accurately. In some instances, the case has been erroneously described as the “comparable worth” case. From their perspective, the OFCCP and the EEOC have been reading the case superficially as standing for the proposition that, since 1970, the Supreme Court has been expanding the scope of the compensation practices that constitute unlawful sex-based discrimination under Title VII.
A fair reading of the case should lead to the opposite conclusion: notably that the court put a stop to the expansion of sex-based discrimination in compensation in many critical ways including the planting of a seed for the dismantling of disparate impact in sex-based discrimination in pay.
Defendant County of Washington won the case at the district level. The district court found that the job of the female matrons in the county jail were different (not substantially equal) from the males’ job of jail guards, terminated the inquiry under Title VII and entered a judgment as a matter of law for the defendant. The district court held that like the EPA, only claims of intra-job (similar jobs) sex-based discrimination in compensation are actionable under Title VII.
The issue presented by the plaintiffs to the 9th circuit was whether plaintiffs were denied equal pay for work substantially equal to that performed by male guards, and even if the work was not substantially equal, could some of the inequality in pay be explained only by intentional sex discrimination. Plaintiffs’ alternative contention was that even if their jobs were not substantially equal (as the district court found them to be different from those of the male jailers), they should be allowed to proceed and prove that some inequality in wages was due to intentional sex discrimination.
In order to evaluate plaintiffs’ contention that, even if the work was not substantially equal, the defendant nevertheless violated Title VII if some of the difference in salary between the female plaintiffs and the male guards can be attributed to intentional sex discrimination, both the 9th circuit and the Supreme Court had to consider the interrelationship of Title VII and the Equal Pay Act, focusing on the Bennett Amendment.
The EPA applies to intra-jobs comparisons or situations where a plaintiff contends there has been a denial of equal pay for equal work (similar jobs). It does not apply to inter-jobs comparison (dissimilar jobs), for instance, where the plaintiff is performing comparable (but not substantially equal) work or comparable worth (jobs with equal “values” to the organization), or where a position held by the plaintiff is unique or does not have a comparator. On the other hand, the language of § 703(a) of Title VII defines discrimination much more broadly than the language contained in the EPA. Title VII, however, is not silent on its relationship with the EPA. An amendment to § 703(h) of Title VII, 42 U.S.C. § 2000e-2(h), commonly known as the Bennett Amendment, provides:
- “It shall not be an unlawful employment practice under this subchapter for any employer to differentiate upon the basis of sex in determining the amount of the wages or compensation paid or to be paid to employees of such employer if such differentiation is authorized by the provisions of [the Equal Pay Act].”
With this stage set as a background, the Supreme Court handed down its ruling in County of Washington that became seminal in determining and firmly drawing the law of sex-based discrimination in compensation at least in the following critical areas:
- County of Washington limited Title VII sex-based claims of discrimination in compensation to disparate treatment or intentional discrimination and, in dictum, the court expressed the view that the incorporation of the fourth affirmative defense of the EPA may conflict with the application of Griggs’ disparate impact to sex-based pay discrimination.
- When plaintiff alleges and proves intra-job discrimination in pay (unequal pay for equal work or jobs), the EPA and Title VII are coextensive. In this case, the EPA standards will apply in Title VII suits and the decisions interpreting the EPA will be authoritative on plaintiffs suing under Title VII.
- Like the 9th circuit before it, the Supreme Court accepted and left undisturbed the district court’s findings that the work of the female matrons was not substantially equal to the work of the male jail guards (dissimilar work or jobs). The majority and the dissent both agreed and determined that this was not a case of intra-job comparison (comparison of similar work or jobs) and went out of their way to reject any inter-jobs comparison (comparison of dissimilar work or jobs) such as “comparable worth” for purpose of establishing a case of sex-based pay discrimination under Title VII. The court did not need the comparison of the male guard job to the female guard job for its reasoning. The court could have reached the same conclusion even if no male guard worked in the county jail. In absence of a relevant comparator, the female plaintiffs were left with a challenge of their employer compensation action based on direct evidence of intentional bias. The majority opinion agreed with the plaintiffs that the fact that their employer, County of Washington, depressed their wages by setting a wage scale for female guards at a lower level than its own market or wage survey because of their sex, constitutes direct evidence of intentional sex discrimination. It follows from the Court’s opinion that when a plaintiff raises a claim of intentional sex-based discrimination in compensation under Title VII without an allegation of equal work or jobs, but is able to present direct evidence of intentional sex discrimination, plaintiff can proceed under the direct evidence burden of proof that may or may not require the plaintiff to identify any comparator.
- Lower courts subsequently interpreted County of Washington to mean that when plaintiff raises a claim of intentional sex-based discrimination in compensation under Title VII in the absence of an allegation of equal work or jobs and when the plaintiff does not present direct evidence of discrimination, then the case will proceed under the McDonnell Douglas/Burdine approach to disparate treatment to evaluate the strength of plaintiffs’ circumstantial evidence. In this case, the lower court created a standard of “similarly situated employee in all material respects” that is more stringent than the statutory comparator in the EPA. At least when the court is evaluating plaintiffs EPA prima facie case, it compares the jobs held by the female and male employees but does not compare the skills, qualifications and other characteristics of the individual employees holding those jobs. In fact this approach under EPA is less stringent than the court created standard of “similarly situated employee in all material respects” that became part of Title VII law.
In contrast with Title VII, the statutory language in EPA on its face clearly defines discrimination in pay in a comparative way. Thus the comparison referred to in EPA is not a mere court-created evidentiary or heuristic device but rather is a part of a substantive and statutory claim for relief. Therefore, many commentators hastily concluded, as wishful thinking, that the standards for comparing comparators under the EPA are narrower than the standards under Title VII, relying on the fact that the statute's narrow standards must be strictly enforced, whereas the broad focus of Title VII must mandate an attack against any form of sex-based discrimination in compensation. Our contention here is that the federal court did not follow this path. In absence of direct evidence of discrimination in pay and when plaintiff does not allege intra-job sex-based pay discrimination, the federal court has applied Title VII “similarly situated employee in all material respects” or “nearly identical” comparators standards that are at least as narrow or stringent as the statutory comparison mandated by the statutory language in the EPA.
“Footnote 11” in Smith v. City of Jackson - The “coup de grace” to Disparate Impact
The majority opinion in County of Washington ruled that Title VII incorporated the EPA's four affirmative defenses. Briefly, the EPA makes it illegal for employers to pay unequal wages to men and women who perform substantially equal work. A plaintiff claiming wage discrimination must establish that: “(1) the employer pays different wages to employees of the opposite sex; (2) the employees perform equal work on jobs requiring equal skill, effort and responsibility; and (3) the jobs are performed under similar working conditions.” After the plaintiff makes each of these showings, the defendant employer may avoid liability by proving that the wage disparity is justified by one of four affirmative defenses — that is, that the employer has set the challenged wages pursuant to “(1) a seniority system; (2) a merit system; (3) a system which measures earnings by quantity or quality of production; or (4) a differential based on any other factor other than sex.”
As we discussed in the above section, the Bennett Amendment codified in section 703(h) of Title VII, incorporated the EPA four affirmative defenses including the fourth one “any other factor other than sex” (“AOFOTS” hereinafter) into Title VII. In doing so, the Court in County of Washington had expressed doubt in dictum as to whether sex-based, disparate impact claims of discrimination in pay are viable under Title VII in light of the Bennett Amendment incorporation of the AOFOTS affirmative defense:
- “More importantly, incorporation of the fourth affirmative defense could have significant consequences for Title VII litigation. Title VII’s prohibition of discriminatory employment practices was intended to be broadly inclusive, proscribing “not only overt discrimination but also practices that are fair in form, but discriminatory in operation.” Griggs v. Duke Power Co., 401 U.S. 424, 431, 91 S. Ct. 849, 853, 28 L. Ed. 2d 158 (1971). The structure of Title VII litigation, including presumptions, burdens of proof, and defenses, has been designed to reflect this approach. The fourth affirmative defense of the Equal Pay Act, however, was designed differently, to confine the application of the Act to wage differentials attributable to sex discrimination. H.R. Rep. No. 309, 88th Cong., 1st Sess., 3 (1963), U.S. Code Cong. & Admin. News 1963, p. 687. Equal Pay Act litigation, therefore, has been structured to permit employers to defend against charges of discrimination where their pay differentials are based on a bona fide use of “other factors other than sex.””
The question that was presented to the US Supreme Court in Smith v. City of Jackson is whether the disparate impact theory of liability was available under the Age Discrimination in Employment Act (ADEA). A plurality of the Supreme Court held that a modified disparate impact theory of liability that does not use Griggs’ job relatedness and consistency with business necessity is available under the ADEA.
In reaching this conclusion, the court’s plurality, when discussing the legislative history and purpose for including the Reasonable Factors Other Than Age (RFOA) in ADEA, provided an example of a contrasting text in EPA that prohibits all disparate impact claims. The court’s detour through the EPA was important to its reasoning and to the decision reached in the ADEA case. It was not dictum. Indeed, a noteworthy aspect of the plurality’s reasoning is found in the famous Footnote 11, where the plurality discussed the contrast between the text of the ADEA and the text of the EPA:
- “We note that if Congress intended to prohibit all disparate-impact claims [in the text of the ADEA 1975], it certainly could have done so. For instance, in the Equal Pay Act of 1963, 29 U.S.C. § 206(d)(1), Congress barred recovery if a pay differential was based “on any other factor”—reasonable or unreasonable—“other than sex.” The fact that Congress provided that employer could use only reasonable factors in defending a suit under the ADEA is therefore instructive.”
As we know from Griggs, the disparate impact theory of liability focuses on factors that are neutral (unrelated to sex) but harshly impact one sex more than the other in their operation or application. Subsequently, after the plaintiff establishes its prima facie case in a typical disparate impact case, the burden shifts to the employer to show that those neutral factors are job related and consistent with business necessity. Footnote 11 creates an incompatibility, antagonism and an undeniable conflict with Griggs. From the rejection by the Court’s plurality of the lower standard of reasonableness in EPA, we can logically infer that the court will certainly reject the Griggs higher standard of job relatedness and consistency with business necessity in Title VII. The rejection of the application of the reasonableness standard by the court’s plurality, when interpreting AOFOTS in the EPA, foreclosed the application of any form or variation of the adverse impact theory of liability in Title VII compensation cases.
Our contention is that the discussion by the Smith court of Footnote 11’s impact on EPA claims has significant implications for sex-based Title VII discrimination in compensation claims. For a plaintiff, the clear implications are that sex-based disparate impact claims in all their forms and variations are not available under Title VII whether they are predicated on the ADEA standard of reasonableness or the Griggs standard of job relatedness and consistency with business necessity. If Smith’s Footnote 11 implies that the disparate impact theory in all its forms and variations are unavailable under Title VII with respect to sex-based pay discrimination claims, presumably this forecloses claims based on the “alternative employment practice” phase of the Griggs disparate impact theory.
Note that when asked to consider a writ of certiorari to the Second Circuit in Aldrich, the Supreme Court denied the application. It did so over the objection of three justices, who thought the split in the circuits warranted the granting of the writ. See 506 US 965 (1992) (J. White)(“I would grant certiorari to resolve the acknowledged conflict among Circuits regarding the interpretation of the federal Equal Pay Act.”)
As I will discuss in a forthcoming publication, several courts of appeals have disagreed on whether the incorporation of the fourth affirmative defense AOFOTS into Title VII, totally foreclosed the application of a theory of disparate impact or just substantially changed the burden of proof in the theory of disparate impact when it is applied to discrimination in sex-based compensation claims (e.g. a reasonable standard may be used to replace the job relatedness and consistency with business necessity standard in phase 2 of a disparate impact burden of proof) . Notwithstanding the areas of disagreement, no court of appeal has held that the Griggs’ standard of job relatedness and consistency with business necessity was applicable to sex-based discrimination in compensation claims under Title VII. As I will discuss in the next section, why enforcement agencies are still taking the position that a Griggs’ type of disparate impact theory is available in sex-based pay discrimination claims under Title VII is anyone’s guess.
Disparate Impact in Compensation According to the Enforcement Agencies
From the perspective of the enforcement agencies, notably the EEOC and the OFCCP, we will consider two important documents that relate to compensation discrimination: OFCCP Directive 307 (Feb 28, 2013) and EEOC Directive on Compensation Discrimination (EEOC Directive No. 915.003 (Dec. 5, 2000)). Both of these documents are Directives meaning that they are internal guidance documents that provide the agencies’ staff with direction on how to approach auditing or investigating contractors’ or employers’ compensation practices.
Both EEOC and OFCCP took the position in both of their Directives that Griggs’ adverse impact theory applies to both compensation as a dependent variable as well as to the independent variables that determine compensation (e.g. performance, time with the company, education, experience etc.). The EEOC at least acknowledged in its Directive in Footnote 34 that some courts have rejected the agency’s position on availability of disparate impact in sex-based discrimination in pay. With all fairness to the EEOC, its Directive on Compensation Discrimination was written prior to the US Supreme Court deciding Smith v. City of Jackson.
Directive 307 contemplates the application of disparate impact analysis to both the independent variables that determine pay and to pay as the dependent variable. Section 7 (Review and Test Factors before Accepting the Factors for Analysis) of Directive 307 direct the compliance officer to determine “whether using the factors present adverse impact issues. Where identified employment practices such as performance review systems show disparities, COs should proceed to review for potential evidence regarding whether the practices result in disparate treatment or disparate impact – including inquiring about evidence of validation and/or the existence of best practices to reduce adverse impact.”
This is an interesting position taken by the OFCCP. As a matter of logic, any independent variable or explanatory factor that correlates or explains differences in pay will have, by definition, a disparate impact on the disfavored group. The greater is the explanatory power of the independent variable, the more likely it will have disparate impact. It is tautological for the OFCCP to direct its COs to reject or remove any factor that is neutral on its face and in its application because it significantly explains the difference in pay. In Directive 307, OFCCP gave itself the latitude to reject the factors the contractor includes in its multiple regression analysis for the simple reason that these factors explain away the significance of race or gender. But this is only the prima facie case of a disparate impact – to be true to a Griggs’ disparate impact type of approach, the next step for OFCCP is to determine whether the significant factors are job related and consistent with business necessity.
Obviously, the OFCCP can’t just reject the factors that determine pay based on a prima facie case of disparate impact without at least asking the contractor for evidence of validation. The overwhelming majority of the employers in the United States do not formally validate the factors or independent variables that determine their pay – for the simple reason that there is no known method for doing so. How would an employer validate time with the company or performance? The fact that OFCCP is asking for evidence of validation of any significant explanatory factor or independent variable is tantamount to reserving the right to arbitrarily reject any explanatory factor advanced by a contractor. This reasoning can invalidate any determinants of pay, including the ones that are especially important to explain pay variation between the groups, and, consequently, can produce a finding of discrimination in pay where none exists. To find contractors guilty of pay discrimination because they cannot prove the validity of a major determinant of pay, when there is no known method for doing so, places a harsh burden of proof on contractors that the federal court did not put on defendants, essentially requiring contractors to prove their innocence. To paraphrase William Bielby and Pamela Coukos, OFCCP is “dueling with unconventional weapons” in violation of “the norms of fair play”. The OFCCP approach to disparate impact in compensation is guaranteed to support the OFCCP’s position, regardless of the actual facts of the case.
Conclusion: Oversight is a Good Thing After All
Publicly, OFCCP has justified its Directive 307 as an attempt to align OFCCP’s interpretation of EO 11246 with Title VII law in the area of pay discrimination. To support this position, OFCCP has been emphasizing the flexibility that Directive 307 provides the agency in making a case of discrimination in pay. This flexibility has been presented as a proof of alignment with Title VII.
A plaintiff has always had the flexibility to develop the evidence he/she needs to make his/her case. This flexibility has been always available to plaintiff and to enforcement agencies and taken for granted by the court. There is nothing specific in Title VII that mandates this flexibility. What is concerning are the areas of conflict between Directive 307 and federal court interpretation of Title VII. There are many areas where Directive 307 is irreconcilable with Title VII.
In addition to the unavailability of disparate impact under Title VII, I will briefly mention Directive 307 aggregation tool using Pay Analysis Groups (PAGs):
PAGs in Directive 307 are unauthorized by Title VII and they conflict with federal court crafted “similarly situated employee in all material respects” or “nearly identical” comparator standards. Regardless of the theory of liability advanced by OFCCP (e.g. disparate impact or disparate treatment), aggregation of dissimilar jobs clashes with legal causation that federal court insists a plaintiff should be able to prove. The aggregation approach through PAGs and multiple regression analysis is an important tool in the OFCCP war against the pay gap.
The problem for OFCCP is that the pay gap has at least three sources:
- Women have low pay because they are concentrated in low paying jobs in the organization and/or society for reasons that may or may not be discriminatory;
- Women have low pay because the pay for the jobs where they are concentrated is set at a low level by the job evaluation systems of the employers and/or the market for reasons that may or may not be discriminatory;
- Women have low pay because, even when they hold the same jobs than men and possess the same qualifications than men, employers may pay them less than men for reasons that may or may not be discriminatory.
The result of a multiple regression using an aggregation of dissimilar jobs (under the epithet of PAGs) will necessarily be ambiguous because the sex coefficient produced by the regression incorporates all three of the sources of the pay gap described above. Given this fact, it is not surprising that courts when considering regressions based on aggregation of different jobs were uncertain as to the particular discrimination claim that the regression was attempting to prove.
For example if an employer has a concentration of females in lower paying jobs because of a placement process that may or may not be discriminatory, the multiple regression analysis based on an aggregation of different jobs into PAGs may likely produce a sex coefficient that is statistically significant in determining pay (evidence of pay discrimination). The issues created by the aggregation of different jobs into PAGs in pay discrimination analysis will be addressed in a forthcoming publication.
While EEOC Directive No. 915.003 is more aligned with federal court jurisprudence than OFCCP Directive 307, in recent years, the Supreme Court has been very critical of the EEOC interpretations of the law. Indeed, the Supreme Court has consistently refused to define what level of deference the EEOC’s interpretations are owed (e.g. Chevron or Skidmore), preferring to retain a broad and undefined discretion to accept or reject agency analysis. Further, when the Court does apply a settled deference standard, it more often than not finds the EEOC’s interpretation unpersuasive.
While I sympathize with the EEOC that has to face in this situation an “arbitrary and capricious” Supreme Court, the high level of scrutiny that EEOC regularly faces from federal courts have kept the agency’s interpretations “as close as possible” to federal court jurisprudence. Even when EEOC was trying to advance social or policy changes, it has carefully considered federal court precedent. On the other side, OFCCP, when compared to EEOC, almost never has to defend its interpretations of the law in federal court. The majority of the contentious audits are settled before any type of litigation is initiated, and when litigation is initiated by OFCCP there is a long road of administrative litigation before a contractor can have its day in federal court. The absence of oversight has led the OFCCP to develop a rogue interpretation of discrimination in compensation in its Directive 307.
As an agency that has audit and enforcement authority, OFCCP does not need such a rogue interpretation of pay discrimination. It already has unfettered access to contractors’ data that was upheld by a federal judge in United Space Alliance. Contrary to the EEOC or private plaintiffs that have to play by the Federal Rules of Civil Procedures (e.g. the general gate-keeping function of Fed. R. Civ. P. 23 and other rules applicable to discovery) and frequently face federal court scrutiny, the OFCCP does not have to worry about any of these limitations. Regrettably, we can recognize in Directive 307 many of the statistical tools and techniques that were tried unsuccessfully by the plaintiffs’ bar in federal court. Directive 307 recycled many tricks, theories, doctrines, tools, analyses, statistical approaches, and other weapons that plaintiffs unsuccessfully tried in Title VII cases. The words that come to mind regarding Directive 307 in the context of an OFCCP audit are that the agency is ”dueling with unconventional weapons” and violating “the norms of fair play.”
If indeed I am correct on the unavailability of disparate impact theory of liability to prove sex-based discrimination in compensation as a matter of law, then OFCCP and EEOC need new Directives.
This article was prepared by the author in his personal capacity. The views and opinions expressed in this article are the author's own and do not reflect the views of his current or former employers. The author is not an attorney, his personal views and opinions are not legal advice, and no one should rely upon or construe them as such.
1. Next publications will address the following topics: The Problem with Starting Pay; What Can We Learn from the British Experience with the Equality Act and the Canadian Experience with the Canadian Human Rights Act (CHRA), the Equal Wages Guidelines (1986), and the Canada Labour Code (CLC)? What is Pay? Why Compensation Discrimination Experts Can’t Get it Right? The Legal Challenges of Correcting “OFCCP Style” Findings of Pay Discrimination; The Question of Comparators in Pay Discrimination; The Role, the Weight, the Probative Value of Statistics in Compensation Discrimination and Legal Causation; The Pay Gap is not Legally Actionable; Disparate Impact in Compensation According to the Circuit Courts. ↩
2. In this article when we use the words “pay” or “compensation”, we mean “base pay.”↩