by Art Gutman Ph.D., Professor, Florida Institute of Technology
The purpose of this alert is to further explain the distinction between the McDonnell Douglas scenario for individual disparate treatment claims versus the Teamsters scenario for classwide pattern or practice claims.
The McDonnell Douglas scenario is also referenced in court cases as McDonnell-Burdine. It is so named after Supreme Court’s landmark ruling in McDonnell Douglas v. Green (1973) [422 U.S. 792], which established the scenario and a later ruling Texas Dept. of Community Affairs v. Burdine (1981) [450 US 248] in which the scenario was reaffirmed. In McDonnell v. Green, a black male (Percy Green) was legally engaged in civil rights activities during his prior employment at McDonnell Douglas. He participated in two illegal acts after a layoff: (1) a "stall-in", where cars parked to prevent employees from entering the plant, and (2) a "lock-in”, where an exit was padlocked to prevent employees from exiting the plant. When the company subsequently advertised for mechanics (Green's prior position), Green applied, was not rehired, and sued alleging race discrimination. The question reduced to whether Green was not rehired because of his legal civil rights activities prior to the layoff or his participation in illegal acts against the company after the layoff. To answer the question, the Supreme Court created the three-phase scenario depicted in the following table.
The McDonnell-Burdine Scenario
- Phase 1 - Plaintiff presents presumptive evidence of a violation: (protected group membership, qualification to perform the job, application for the job, rejection followed by continuation of the search)
- Phase 2 - Defendant must articulate a legitimate reason for the selection decision made in phase 1; proof is unnecessary (burden of production)
- Phase 3 - Plaintiff must provide direct or indirect evidence that the articulation by defendant in phase 2 is a pretext for discrimination (burden of persuasion)
The evidence in Phase 1 (also termed the “prima facie” phase) is light and presumptive. Or in Justice Powell’s words, Green had to establish the following:
(i) that he belongs to a racial minority; (ii) that he applied and was qualified for the job for which the employer was seeking applications; (iii) that, despite his qualifications, he was rejected; and (iv) that, after his rejection, the position remained open and the employer continued to seek applicants from persons of complainant's qualifications.
Green easily passed each step: (i) he was black; (ii) his qualifications were acknowledged based on his prior work; (iii) he was not rehired; and (iv) the search continued after Green was excluded. In other words, factual proof of a violation is not required in phase 1; only a presumption of a violation that is easy for the plaintiff to make and equally easy for the defendant to refute.
In phase 2 (also known as the “defense” phase), there is a production in response to the plaintiff’s presumptive evidence of a violation. In Powell’s words, "the burden then must shift to the employer to articulate some legitimate, nondiscriminatory reason for the employee's rejection." In other words, the defendant need only verbalize, without proving, a legal reason for the challenged selection decision. To pass its burden, McDonnell-Douglas articulated that Green was not rehired because of his involvement in illegal activities during the layoff.
That leaves the only burden of persuasion on the plaintiff in Phase 3 (also known as the “pretext” phase) to prove with direct or indirect evidence that the articulation in Phase 2 is not the true reason for the selection decision. In Justice Powell’s words:
Especially relevant to such a showing would be evidence that white employees involved in acts against petitioner of comparable seriousness to "stall-in" were nevertheless retained or rehired. ..... Other evidence ..... includes facts as to the petitioner's treatment of respondent during his prior term of employment; petitioner's reaction, if any, to respondent's legitimate civil rights actives; and petitioner's general policy and practice with respect to minority employment.
In this example, proof that similarly situated white workers also committed an illegal act but were nevertheless rehired would constitute indirect (or circumstantial) evidence, whereas proof of prior racially discriminatory acts against Green and other minorities (e.g., documents, eyewitnesses, etc.) constitutes direct evidence. Generally, most disparate treatment claims are resolved in Phase 3 based on indirect evidence. However, where there is direct evidence of a violation, the court may require a so-called “mixed-motive” instruction in which the direct evidence is taken as proof of a violation in Phase 1 and the defendant must prove, in kind, a legal reason for the selection decision. The mixed motive scenario, though a disparate treatment scenario itself, should not be confused with the McDonnell-Burdine scenario. By the way, ultimately, Percy Green could not prove pretext and lost.
The Teamsters scenario for pattern or practice claims is depicted in the next table below. Notice that it can involve both classwide and individual claims. This follows because classwide claims are always accompanied by individual claims, and those individual claims are resolved using McDonnell-Burdine rules. The pattern or practice scenario was established by the Supreme Court in the aforementioned Teamsters case, and Hazelwood School District v. US (1977) [433 US 299]. These two were companion cases decided on the same day. The Department of Justice (DOJ), which was at the time responsible for pattern or practice cases, prosecuted both cases, and won in Teamsters but lost in Hazelwood.
The Teamsters Scenario
- Phase 1 Plaintiff provides statistical evidence of underrepresentation of minorities/females in the workforce or overrepresentation minorities/females in less desirable jobs
- Phase 2 Defendant can use McDonnell-Burdine defense for pattern statistics; defense for individual claims generally follows McDonnell-Burdine rules
- Phase 3 Plaintiff's requirement for showing pretext is same as previously discussed for McDonnell-Burdine cases for both the overall claim and individuals claims
The pattern or practice scenario differs from McDonnell-Burdine primarily in Phase 1, which, for all intents and purposes requires either of two types of stock statistical comparisons: (1) underrepresentation of protected group members in the workforce as compared to a relevant labor pool (also known as “composition” statistics) or (2) overrepresentation within the company of protected class members in less desirable jobs (also known as “cross-job” disparities). Beyond that, the defendant must explain why such disparities exist in Phase 2 and plaintiffs must discredit the explanation in Phase 3.
The Teamsters case involved cross-job disparities. Prior to and beyond July 2, 1965, the effective date of Title VII, many blacks and Hispanics were "city drivers, " a lower paying (or less desirable) job, but none were "line drivers," a higher paying (or more desirable) job. Speaking for a unanimous Supreme Court, Justice Stewart ruled that the Government had the initial burden to present prima facie evidence of discrimination and the ultimate burden to persuade the court that the pattern or practice of discrimination was a "standard operating procedure." Or in Stewart’s words:
As the plaintiff, the Government bore the initial burden of making out a prima facie case of discrimination .... and because it alleged a system-wide pattern or practice of resistance to the full enjoyment of Title VII rights, the Government ultimately had to prove more than the mere occurrence of isolated or "accidental" or sporadic discriminatory acts. It had to establish by a preponderance of the evidence that racial discrimination was the company's standard operating procedure - the regular rather than the unusual practice.
This case was an easy victory based on the statistical evidence alone. Among 6,742 bus drivers, 1,828 were line drivers and 4,914 were city drivers. Collectively, 571 (or 8.5%) of all drivers were minorities. The key, however, was that for higher paying line driving job, there were only 13 minorities (or 2.2% out of 571), each hired after the lawsuit was filed. Thus, prior to the lawsuit, there was an inexorable zero for the better paying of two similar jobs. The defense argued that “statistics can never in and of themselves prove the existence of a pattern or practice of discrimination, or even establish a prima facie case shifting to the employer the burden of rebutting the inference raised by the figures”, to which Justice Stewart responded that Stewart responded that "gross" statistical disparities are "probative", and require "explanation", as in McDonnell-Burdine. And in this regard, Stewart ruled that the “company’s inability to rebut the inference of discrimination came not from the misuse of statistics, but from the inexorable zero.” In short, the Teamsters had a facially discriminatory policy that left no plausible legal explanation of why more than 500 minorities were segregated into the lower paying job, with none in the higher paying job.
Hazelwood is a landmark case because it illustrates what happens when there is a stock statistical disparity that falls far short of an “inexorable zero.” The Supreme Court therefore needed an operational definition of “gross” disparity, which it imported from Castaneda v. Partida (1977) [430 US 482], a case decided three months prior to Teamsters and Hazelwood. Castaneda was a pattern or practice jury pool case in which the charge was underrepresentation of Mexican Americans on jury pool in a county that was predominantly Mexican American.
An operational definition of gross disparity was not required in Teamsters because of the inexorable zero. However, a definition was required in Hazelwood, which lacked the inexorable zero. The charge was underrepresentation of black teachers in the workforce in comparison to availability of licensed black teachers in the relevant labor pool. The Hazelwood Court established rules for determining (a) how much of a disparity is necessary for a prima facie case, and (b) the relevant labor pool for assessing the alleged composition disparity (the percentage of adult Mexican-Americans in the county was 79.1% compared to 39% in the jury pool). The Castaneda Court accepted social scientific testimony to establish the now well-known two-standard deviation rule. Accordingly:
As a general rule for such large samples, if the difference between the expected value and the observed number is greater than two or three standard deviations, then the hypothesis that the jury drawing was random would be suspect to a social scientist.
The disparity between expected and observed Mexican-Americans in the jury pool was roughly 29 standard deviations, thus establishing the prima facie case.
Returning to Hazelwood, the charges in this case sounded as compelling as those in Teamsters. In the words of Justice Stewart:
The Government mounted its "pattern or practice" attack in the District Court upon four different fronts. It adduced evidence of (1) a history of alleged racial discriminatory practices, (2) statistical disparities in hiring, (3) standardless and largely subjective hiring practices, and (4) specific instances of alleged discrimination against 55 unsuccessful applicants for teaching jobs.
Ultimately, however, two factors favored the defendants. The first factor related to the question of the proper composition disparity. Prior to the EEO Act of 1972, which added Title VII coverage of public entities, only 0.6% of the teachers in the Hazelwood School District were black. This percentage grew to 1.8% in 1972, and to 3.7% at the time the suit was filed. The percentage of licensed black teachers in the Hazelwood School District was 5.7% as compared to 15.4% in the county as a whole, which included St. Louis. Statistically, the percentage of qualified black teachers hired when the suit was filed (3.7%) was significantly less than the percentage living in the county as a whole (15.4%), but was not significantly less than the percentage living within and around the Hazelwood School District (5.7%). Justice Stewart ruled that the relevant labor pool included only the immediate area surrounding the Hazelwood School District, and the DOJ lost because the disparity in question (3.7% vs. 5.7) was not statistically significant.
Even though the DOJ lost on the prima facie claim, the Supreme Court discussed a second factor, which was whether the defendant could explain the disparity relative to the entire county. Justice Stewart stated that assuming even that a prima facie case could be made, the defense had two legitimate articulations: (1) that competition from the city of St. Louis made it difficult to recruit black teachers; and (2) that the school district’s affirmative action plan significantly raised minority hiring rates after the EEO Act of 1972 was passed. Had the case proceeded past this point (to Phase 3), it would have been the DOJ’s burden to prove that these articulations were pretext.
There are three additional points to note. First, it easy to confuse pattern or practice with adverse impact because both involve classwide assertions based on statistical evidence. However, critically, in Title VII case law adverse impact cases rely on flow statistics (i.e., statistically significant disparities in selection ratios between minority or female applicants based on an identified selection test or procedure, whereas Title VII pattern or practice cases follow from the aforementioned stock statistical comparisons. Second, adding to the potential confusion, the two-standard deviation rule established in Castaneda has been incorporated into adverse impact as well as pattern or practice cases.
Lastly, when comparing title VII case law to OFCCP enforcement of Executive Order 11246, the difference between pattern or practice and adverse impact becomes ambiguous. Based on OFCCP conciliation agreements and consent decrees, it appears that OFCCP consistently applies a pattern or practice theory of discrimination to applicant flow disparities, particularly when a selection system is ambiguous and not clearly organized into specific steps that could be challenged under an adverse impact theory. For example, perhaps a hiring process is not organized into clear steps, but at some point work experience is evaluated subjectively and used as a pass/fail criterion. However, the employer only applies the experience requirement for female applicants and not for male applicants, and adverse impact exists for the overall selection system. In this scenario, applying a pattern and practice theory of discrimination may lead to a reasonable inference of discrimination (it isn’t an adverse impact scenario), yet this is not consistent with Title VII case law.
I could say more, but if you’ve gotten this far, you’ve shown incredible patience. If you want more on the adverse impact vs. pattern or practice issue, let me know.
