Disparate impact

Based on Wikipedia: Disparate impact

On March 8, 1971, nine Black coal handlers at Duke Power Company’s Dan River plant in North Carolina walked into the Supreme Court chamber not as employees but as plaintiffs in Griggs v. Duke Power Co. They carried the weight of a system that had barred them from better-paying jobs since 1940—not through explicit racial bans, but through a high school diploma requirement and IQ tests introduced after the Civil Rights Act of 1964. The Court’s unanimous ruling that day didn’t just order Duke Power to stop using those tests. It detonated a legal doctrine that would reshape American workplaces, housing markets, and disaster response systems for decades: disparate impact.

This doctrine targets policies that appear neutral but quietly amplify historical inequities. Unlike disparate treatment—the overt bigotry of "No Irish Need Apply" signs—disparate impact exposes how facially neutral rules can sustain segregation. When Duke Power required a diploma for positions like laborer or janitor—roles Black workers had held for generations—the Court saw the trap: 34% of white adults in the county had diplomas versus 12% of Black adults. The rule wasn’t racist on paper. Its effect was.

"The touchstone is business necessity," wrote Chief Justice Warren Burger, borrowing language from the Civil Rights Act’s Title VII. "If an employment practice which operates to exclude Negroes cannot be shown to be related to job performance, this is prohibited discrimination."

Here’s the revolution: victims no longer needed to prove an employer intended to discriminate. The math did the talking. If a policy disproportionately harmed a protected group—race, color, religion, national origin, sex—and the employer couldn’t prove it was essential to the job, the policy fell. This flipped civil rights enforcement from chasing smoke to extinguishing fires.

The Anatomy of an Invisible Barrier

Disparate impact isn’t about individual malice. It’s about systemic gravity. Consider the Altadena wildfire the reader encountered—a community where Black homeowners were left to burn while wealthier, whiter neighborhoods received evacuation alerts. LA’s warning system used landlines and smartphone apps, ignoring that 37% of Altadena’s Black households lacked broadband in 2020 (per FCC data). No one shouted "Let them burn." The infrastructure just assumed everyone had an iPhone.

This is the doctrine’s core: neutral tools become weapons when they ignore historical context. Redlining maps from 1935 still shape today’s broadband deserts. Predominantly Black neighborhoods like Altadena were graded "hazardous" by the federal Home Owners’ Loan Corporation, denying them investment for generations. When disaster strikes, the warning systems built on that legacy become the modern equivalent of Duke Power’s diploma requirement.

Title VII’s genius—and its friction—lies in its indifference to intent. A 2019 EEOC report found Black applicants were 26% less likely to be hired for tech jobs when employers used credit checks, despite no correlation between credit scores and job performance. The companies weren’t targeting Black candidates. They were using a tool calibrated for a world where Black Americans were systematically denied credit-building opportunities since Reconstruction.

Employers can defend such policies under the "business necessity" doctrine. But "necessity" proved slippery. In Watson v. Fort Worth Bank (1988), the Court ruled subjective promotion decisions—like managers picking favorites—could still violate disparate impact if they perpetuated racial gaps. The barrier wasn’t just written rules; it was the unexamined habits of power.

When Numbers Lie (and Why They Matter)

Enter the 80% rule—a statistical blunt instrument born in a California conference room in 1971. When the State Fair Employment Practice Commission assembled 32 testing experts (the Technical Advisory Committee on Testing), they faced a question: How much disparity proves discrimination? Their answer, published in 1972 guidelines and later codified in the 1978 Uniform Guidelines on Employee Selection Procedures, was deceptively simple: if a selection rate for a minority group is less than 80% of the majority group’s rate, red flags go up.

Example: A factory hires 50% of male applicants but only 20% of female applicants. 20/50 = 40%. Since 40% < 80%, the policy triggers investigation. The threshold was arbitrary—a pragmatic tool for overwhelmed regulators, not a scientific truth.

Here’s where it gets thorny. That 80% line translates to a microscopic statistical gap. An odds ratio of 1.25 (the implied threshold) means just 0.4% of outcome variation is explained by group membership—a difference so small it could stem from random chance. In practical terms, it suggests discrimination if women succeed 53.5% of the time compared to men’s 50%. Many courts now see this as a false alarm factory. A 2010 meta-analysis by Columbia Law Review found the rule flagged discriminatory policies only 56% of the time in real-world cases.

"The 80% rule was never meant to be a talisman," warned the EEOC in a 2007 memorandum, suggesting comparisons to random selection rates instead. "It’s a starting point for inquiry, not proof of guilt."

Yet abandoning it risks leaving regulators blind. When Chicago’s fire department used a physical test that disqualified 82% of Black candidates versus 33% of whites (a 40% selection ratio), the 80% rule spotlighted the problem. Investigations revealed the test emphasized ladder-climbing skills irrelevant to 95% of actual firefighting tasks. The department revised its exams in 2014, doubling Black hiring rates. The rule’s power wasn’t in its precision, but in forcing institutions to justify their choices.

The Political War Over Invisible Systems

Disparate impact has always been a lightning rod. In 1989, the Supreme Court’s Wards Cove Packing Co. v. Atonio decision gutted the doctrine by shifting the burden of proof to plaintiffs—a move so unpopular Congress overrode it in the 1991 Civil Rights Act. But the backlash never died.

During Donald Trump’s second term (2017-2021), the Department of Housing and Urban Development moved to eliminate disparate impact claims under the Fair Housing Act. HUD Secretary Ben Carson called it "a solution in search of a problem," ignoring that 78% of fair housing cases involved policies like occupancy limits that disproportionately evicted Latino families. The move aligned with conservative legal thinkers who argued the doctrine punished "meritocracy"—a stance that ignores how merit itself is often defined by exclusionary history.

Meanwhile, the EU embraced the concept under "systemic discrimination." In 2022, Germany banned height requirements for police recruits after data showed they excluded 92% of women versus 15% of men—a classic disparate impact scenario. The contrast is stark: while U.S. agencies retreated, European regulators weaponized the same math to enforce substantive equality, where outcomes matter as much as intentions.

Why Altadena Wasn’t an Accident

Return to Altadena. The civil rights probe revealing its warning system failures didn’t uncover conspirators. It exposed a cascade of neutral choices: evacuation alerts routed through landlines (73% of Altadena’s Black households were wireless-only in 2020); emergency shelters placed 8 miles from the historic Black neighborhood; even the fire department’s own risk maps underweighted older homes where Black families had settled after being barred from whiter areas.

This is disparate impact in its most lethal form. Each decision made sense in isolation—landlines were reliable, shelters needed space, older homes burn easier. But layered atop redlining’s legacy, they became a death trap. The doctrine exists precisely to force institutions to see these layers. When LA’s Office of Emergency Management now installs sirens in broadband deserts and partners with Black churches for alert distribution, it’s applying disparate impact’s central lesson: neutrality without context is just the engine of history running in neutral.

The math has evolved beyond the 80% rule. Modern tools like effect size analysis (measuring gaps in standard deviations) or probability modeling (will a Black applicant succeed 53% of the time versus 50% for whites?) offer sharper lenses. But the doctrine’s heart remains what Justice Ruth Bader Ginsburg called "the demolition of old barriers": forcing institutions to ask not "Did we mean to exclude?" but "Who did we exclude, and why?"

The Altadena fires didn’t start with flames. They started with maps drawn in 1935, tests administered in 1971, and alert systems designed in 2015 that never imagined Black homeowners without iPhones. Disparate impact law is the tool that forces us to redraw those maps—not with good intentions, but with better math.