Getting Griggs Wrong

It’s amazing how everyone knows about Griggs v. Duke Power. It comes up frequently in online discussion despite being a topic that is not taught in college unless you are studying employment law, yet everyone seems to know about it, even on Reddit subs that have nothing to do with law. It’s also misunderstood, and I admit to being among the uninformed. For example, people will say:

“Griggs makes employer IQ testing illegal..”

“Disparate impact is illegal”

“Griggs makes it hard/impossible for all but the biggest of companies to use cognitive screening”

These are all wrong. Cognitive screening is illegal if it meets both conditions:

1. It leads to a disparate impact, as stipulated by the 4/5 rule:

The four-fifths rule prescribes that a selection rate for any group (classified by race, orientation or ethnicity) that is less than four-fifths of that for the group with the highest rate constitutes evidence of adverse impact (also called ‘disparate impact’), that is, discriminatory effects on a protected group. Adverse impact may occur in any of the company’s decisions related to employees – be it hiring, promotion, training, transfer or layoff. A particular test or selection procedure must be evaluated for effectiveness in terms of selection based only on characteristics relevant to the job or task. It may, however, render disadvantage to members of a race, gender or ethnic group.

2. However, disparate/adverse impact is not illegal if it’s unintentional:

The “inference of discrimination” is not sufficient for a plaintiff to prove his or her case. Once a plaintiff shows an inference of discrimination, an employer then may “rebut” that case by explaining that its policy or practice has a legitimate business justification. A valid business justification must “serve[], in a significant way, the legitimate employment goals of the employer.” Ward’s Cove Packing Co. v. Atonio, 490 U.S. 642, 659 (1989). The employer in Griggs, for example, would have possessed a valid business justification if high school diplomas or aptitude tests were truly necessary for employees to perform the work at issue.

So screening would only be illegal if it, first, leads to a disparate impact and then, second, it’s shown that there is no valid justification for such testing. The plaintiff must clear both hurdles. This makes cognitive screening more common than one may be inclined to believe, although it’s not necessarily called IQ testing, but rather proxies for cognitive ability, such as the Wonderlic test, are used.

Regarding, Griggs, only black candidates were singled out for testing, which is more discriminatory than giving everyone such tests and blacks scoring lower. The Civil Rights act “authorizes the use of any professionally developed ability test, provided that it is not designed, intended, or used to discriminate.” [1] It is recommended by the EEOC that at least two screening methods be used, such a brief interview and a test. Also in Watson v. Fort Worth Bank & Trust, 1988, the court ruled “Employers are not required, even when defending standardized or objective tests, to introduce formal “validation studies” showing that particular criteria predict actual on-the-job performance.” So tests are allowed if applied in a non-discriminatory manner, which was not the case with Griggs.

Employers have many ways of employing cognitive screening without explicitly using IQ tests. For example, there are phone interviews, Wonderlic tests, and so-called ‘brain teasers’. When one applies for a job, especially for one that is technical and exclusive/prestigious, the first step is often a phone interview in which the applicant goes through a gauntlet of increasingly difficult brain teasers, such as “how many cubes can fit in a box…” Those who make it are offered a much more comprehensive in-person interview, but the questions are effectively an IQ test designed to screen all but a handful applicants.

Wonderlic tests, which unlike phone interviews are done in person, are also surprisingly common and are another form of screening that tests for cognitive ability but without explicitly being an IQ test. The applicant has only 12 minutes to answer 50 questions, and the average score is only 20. Scores above 40 are rare, so one gets a normal distribution that appropriates that of an IQ score distribution.

A Google search yields a ton of information for Wonderlic-related results, so obviously, in spite of Griggs, these tests are common. According to the Wonderlic website, “The validity of a test is determined by how well it measures what it is intended to measure. Wonderlic has an extensive research base documenting the validity of our assessments.” A Google search also revealed few, if any recent or pending, cases involving the improper use of the test, suggesting that the commonly held belief that employers cannot use cognitive screening tests for fear of litigation, is wrong or overblown.

But what about college degrees? A common argument is that employers are stuck using degrees because alternatives don’t exist, yet employers use Wonderlics and phone interviews, so there is sorta a contradiction. I suspect for lesser-skilled jobs, the Wonderlic is used in lieu of a high school or college degree to screen candidates. Phone interviews are more effective for screening candidates for high-paying positions than degrees, because degrees do an inadequate job of differentiating quality, except for the most elite of degrees. If you have 100 applicants with BAs, and there is negligible performance difference between, say, a top-20 school grad with a 3.3 GPA in history versus a top 40 grad with a 3.6 GPA in economics, you need a more precise filtering mechanism, and giving in-person interviews for many candidates is impractical. Also, another problem is grade inflation, which has devalued college GPAs, just as it has for high school GPAs.

[1] Of Course It’s Legal, When Properly Administered.