Nearly 100 years ago, Ernest Hemingway wrote that a character went bankrupt “Gradually and then suddenly.”  Likewise, the disparate impact doctrine in civil rights law is ending first gradually, and then suddenly.

In civil rights law, disparate impact occurs when the demographics of selected individuals (e.g., hired applicants or admitted students) does not reflect the population of applicants.

Disparate impact makes organizations vulnerable to legal challenges of discrimination.  For example, employers sued for disparate impact must prove that their practices serve a legitimate business purpose and are the least discriminatory option available.  At no point in this legal process does intent come into play.  That means that employers can be found liable even when they had no intention of discriminating.

The problem with disparate impact doctrine is that disparities happen throughout society.  Social clubs, sports teams, legislatures, and many other organizations do not have demographics that reflect applicants or the general population.  But disparate impact doctrine assumes that when disparities occur in employment and education, discrimination is at work and must be corrected.  The problem is that correcting these disparities means noticing demographic characteristics (e.g., race and sex) and then treating some applicants different because of them.  Disparate impact doctrine is a major hurdle to creating a colorblind society.

What started in the 1960s as well intentioned efforts to remedy recent discrimination blossomed into an entire DIE industry dedicated to stamping out disparate impact through race-conscious practices.  For decades, opponents fought disparate impact doctrine, with some minor victories but no fundamental changes.

The beginning of the end for disparate impact was Students for Fair Admissions v. Harvard.  In that case, the Supreme Court ruled that race-based affirmative action admissions practices at universities are unconstitutional.

The Students for Fair Admissions ruling was confined to higher education, but the implications were obvious.  “Eliminating racial discrimination means eliminating all of it,” Justice Roberts wrote in his opinion.  Students for Fair Admissions also dispensed with the idea that discrimination with good intentions is acceptable.  Such practices still qualify as discrimination and are illegal.  Very quickly, litigation against race-conscious policies proliferated, and DIE in corporations is in retreat.

When Trump returned to office, the disparate impact doctrine suffered more body blows.  The president issued executive orders banning affirmative action, DIE programs, and other race-conscious practices in government agencies, government contractors, and recipients of federal grants.  Virtually overnight, large segments of the economy were told to dispense with the policies designed to respond to disparate impact.  A later executive order attacked disparate impact doctrine as being “wholly inconsistent with the Constitution” and ordered agencies to deprioritize enforcement of disparate impact violations.  Agencies immediately began revising definitions and guidance to defang the doctrine.  Another major change occurred in July, when a federal judge dismissed the Luevano consent decree, which had made it impossible for the government to use standardized tests for many civil service jobs because of the disparate impact the tests caused.

What is surprising is how easy the recent string of victories has been under the Trump administration.  Most of the federal apparatus designed to mitigate disparate impact was based on prior presidents’ executive orders, which Trump revoked with the stroke of a pen.  The Luevano consent decree was ended when no opposing party raised an objection to ending it.

Consequently, disparate impact doctrine is the weakest it has ever been.  The doctrine’s strongest remaining foundation is Griggs v. Duke Power Company, in which the Supreme Court unanimously ruled against a company that used standardized testing for personnel decisions because of the disparate impact it created.  The Manhattan Institute’s Dan Morenoff argues that even this foundation is rickety.  Griggs is ripe for overturning because Griggs’s approach to civil rights law is contradicted by a recent Supreme Court case (Muldrow v. City of St. Louis), and the Court’s textualists will likely be skeptical of the Griggs ruling, which was not based in a plain reading of the relevant statute.

With disparate impact doctrine on the ropes, it is time to finish the fight.  Morenoff provides a roadmap for doing so by weakening and then overturning Griggs.  First, the Equal Employment Opportunity Commission (EEOC) can issue guidance that states that intent is now required to prove discrimination when adverse impact arises.  To make the repudiation of Griggs clearer, the EEOC can explicitly state that cognitive ability tests are permitted, even if they result in disparate impact.  Because the EEOC issues “guidance” and not rules, the Trump administration can do this unilaterally without going through the rule-making process.  (Full disclosure: I work for a company, RIOT IQ, that could benefit if legal barriers to cognitive testing in employment settings were removed.)

The knockout blow for disparate impact theory would be to overturn Griggs.  One option is the traditional method of waiting for an employer to be sued and then appealing the case all of the way to the Supreme Court.  According to Morenoff, another route to overturning Griggs would be for an employer that desires to implement a policy that would result in disparate impact to sue the EEOC to ask for a declarative judgment that the policy is acceptable.

The latter strategy allows legal strategists to craft a clean test case that has the greatest potential to make it to the Supreme Court and requires only an employer willing to front it.  In addition to psychological tests, A.I. systems for evaluating employees or selecting job applicants are well suited to challenge Griggs because forcing them to produce demographically diverse outputs reduces their usefulness.  This was shown in an absurd manner in 2024, when Google’s A.I. Image generator Gemini valued diversity so much that it created images of black Nazis and Native American Founding Fathers.  It would probably not be hard to show that legal requirements to avoid disparate impact in A.I. systems for screening job applicants will produce results just as useless and disconnected from reality.

After decades fighting disparate impact theory and its fruits (like affirmative action), the process of change has shifted from occurring “gradually” to “suddenly.”  Just three years ago, disparate impact theory was the law of the land, and fear of running afoul of it was fueling much of the diversity industry and forcing American society to obsess over race and sex.  But the victories are piling up, and the end of disparate impact doctrine seems within reach.  It is time to finish the job and end this constitutional distortion at last.

Dr. Russell T. Warne is a psychologist who works as the chief scientist for RIOT IQ and is the creator of the Reasoning and Intelligence Online Test.  Readers interested in Dr. Warne’s work can find his professional website at https://russellwarne.com and follow him on social media on Facebook and X.

Image via Pexels.