For six decades, the Equal Employment Opportunity Commission has compiled an annual demographic snapshot of American workers, creating a national record that reveals patterns no single complaint ever could. Last week, the agency's Republican majority voted 2-1 to dismantle that system.
The proposal, which entered the Federal Register on Thursday, would stop requiring employers to classify workers by race and sex in their annual reports. The agency argues that collecting such demographic information may itself violate constitutional equal protection guarantees. Public comment closes August 24, with a hearing scheduled for August 11.
Since 1966, private employers with at least 100 workers and certain federal contractors with at least 50 have submitted these reports covering job categories and demographic breakdowns. Today, those filings encompass more than 50 million workers across roughly 73,000 employers. A single worker passed over for promotion suspects bias but cannot see the company-wide pattern. The EEOC's spreadsheet could show it.
The data itself identifies no one and remains confidential. What it does is reveal what an individual complaint cannot: whether a wound reflects a system. When a Black manager loses a promotion or a pregnant worker gets sidelined with a polite excuse, they are left with suspicion. They have no way to know if others face the same invisible wall.
There are legitimate critiques of the current system. Employers spend an estimated $273 million annually complying. The categories are broad and sometimes outdated. A numerical imbalance can become mistaken proof of discrimination or shield illegal preferences. These concerns deserve serious attention, but the proposal to abandon the system entirely represents something different: a decision to stop looking rather than to look better.
Under the new approach, employers would still maintain their own hiring, promotion, pay and termination records. What would vanish is the standardized demographic framework that lets investigators compare those decisions across groups. The EEOC could still demand demographic data after a complaint or lawsuit begins, but without routine baseline reports, it would lack the consistent foundation for comparison. Employers might not have collected the same information in the same way.
The real shift is more subtle than blindness. The agency would retain the power to examine particular employers when its leaders choose. Routine reports arrived whether or not anyone suspected wrongdoing. Without them, the EEOC's leadership gains control over which workplaces get scrutinized and when.
EEOC Chair Andrea Lucas has already signaled where her attention lies. In December, she invited white men who believed they suffered race or sex discrimination to file complaints. She has framed diversity programs as producing widespread, systemic discrimination, primarily against white employees. That characterization preceded investigation, not the other way around.
The Nike case illustrates the selective approach. No employee complaint triggered that investigation. Lucas issued a commissioner's charge in May 2024 directing the agency to examine Nike. Notably, that charge cited Nike's publicly disclosed EEO-1 data, the product of the very reporting system Lucas now proposes to eliminate. The agency later subpoenaed extensive demographic records while investigating alleged bias against white employees. Investigation remains open.
The 2023 national workforce count offers a useful test of Lucas's theory. White men comprised roughly one-third of the surveyed workforce but held 52.7 percent of executive and senior management positions, making them the only group overrepresented at that level, according to an Associated Press analysis.
That does not mean white men cannot face discrimination at work. Title VII protects all workers, and the EEOC should enforce the law for everyone. But a nationwide record serves precisely this function: allowing evidence to test whether an official's theory matches reality. Lucas wants to end the measurement best equipped to complicate a conclusion she has already reached.
Without those annual counts, future years will have no baseline snapshot. A later administration could attempt to restore the form. It cannot retroactively obtain the data for years when no count existed. Title VII will remain law. What will disappear is the standardized evidence of whether employers obey it. Workers will be left knowing their own wounds, but the government will no longer maintain the record that reveals the system that created them.
Author James Rodriguez: "An agency stops counting discrimination not because counting is impossible, but because the count might show what leadership has decided not to see."
Comments