Workers' Compensation: Dying at Work

Workplace fatality data, political interference, and the workers left behind.

The Numbers Behind the Headlines

Each year, the U.S. Bureau of Labor Statistics (BLS) publishes its Census of Fatal Occupational Injuries (CFOI) — the most comprehensive federal accounting of workers who die on the job. The 2024 report, released February 19, 2026, found 5,070 fatal work injuries in the United States, down 4.0 percent from 5,283 in 2023. On its face, that sounds like progress. A worker dies every 104 minutes from a work-related injury. Transportation incidents remain the leading killer (38.2% of all fatalities). Drug and alcohol overdoses on the job declined sharply. The fatal injury rate fell to 3.3 per 100,000 full-time equivalent workers.

But here's the question no press release answers: Are these numbers real?

The Long Shadow of Underreporting

Labor economists and occupational safety researchers have raised concerns about underreporting of workplace injuries and fatalities for decades, long before politics entered the picture.

The CFOI captures fatal injuries through a multi-source verification system, drawing on OSHA records, death certificates, medical examiners, police reports, and state workers' compensation filings. For deaths, the data is relatively robust. But the broader injury and illness landscape — tracked separately by the Survey of Occupational Injuries and Illnesses (SOII) — relies on employer self-reporting. And therein lies the problem.

Why underreporting happens:

Employer incentives: OSHA recordkeeping requirements create financial and reputational pressure to minimize reported incidents. Injury rates can affect insurance premiums, OSHA inspection frequency, and contract eligibility.
Worker fear: Undocumented workers, gig economy workers, and employees in precarious jobs often do not report injuries for fear of retaliation, deportation, or job loss. Of the 1,229 Hispanic or Latino workers killed on the job in 2024, 68.5% were foreign-born — a population particularly vulnerable to underreporting pressure.
Classification games: Employers and insurers have long been accused of reclassifying injuries to avoid OSHA recordable thresholds, shifting workers to "modified duty" to avoid "days away from work" counts, or pressuring workers to use personal health insurance rather than file workers' compensation claims.
Coverage gaps: Independent contractors, the self-employed, and workers in the informal economy fall into statistical gray zones. The gig economy has dramatically expanded this population.

Research published in peer-reviewed journals — including studies by the National Institute for Occupational Safety and Health (NIOSH) — has consistently found that official BLS injury counts may capture only a fraction of actual workplace harm. Some estimates suggest the true burden of occupational illness and injury is 3 to 10 times higher than official statistics reflect.

For workers' compensation purposes, this is not academic. When injuries go unreported, workers lose access to medical care, wage replacement, and rehabilitation. When the underlying data is compromised — for whatever reason — the entire system of workplace safety regulation, actuarial pricing, and legal accountability becomes distorted.

The Unitary Executive Shadow Over the BLS

Underreporting driven by employer self-interest is troubling enough. But a new and more sweeping threat has emerged: direct political control over the federal statistical agencies themselves.

The Trump Administration has aggressively advanced the unitary executive theory — the constitutional argument that the President possesses plenary control over all executive branch officers and agencies, including those traditionally insulated from political interference by statute. Under this view, independent agency heads, career civil servants performing quasi-independent functions, and even statistical professionals at agencies like the BLS serve entirely at the President's pleasure and must execute his policy preferences.

The implications for data integrity are severe.

The BLS, housed within the Department of Labor, has historically operated with a firewall between political leadership and career statisticians. Data release schedules, methodology decisions, and publication protocols have been governed by professional standards — not electoral cycles. That firewall is now under pressure.

Following broad federal workforce reductions in early 2025 and the reassertion of presidential control over agency personnel, concerns have mounted within the statistical community about the independence of federal data collection. Researchers, former BLS officials, and academic economists have publicly warned that politicized interference with economic and labor statistics — including employment data, injury counts, and inflation measures — could corrupt the evidentiary foundation upon which workplace policy, litigation, and compensation systems depend.

If the President's team can direct what gets counted, how it gets classified, and when (or whether) it gets released, then the CFOI fatality figures — however well-intentioned the career staff who compile them — are vulnerable to political shaping.

Trump v. Slaughter and the Supreme Court's Looming Answer

The legal scaffolding for this shift is currently before the United States Supreme Court in Trump v. Slaughter — a case directly testing the limits of presidential removal power over officials at independent agencies.

Rebecca Slaughter is a Commissioner of the Federal Trade Commission whom President Trump sought to remove in defiance of the statutory "for cause" removal protection Congress established for FTC commissioners. The case builds on the Court's own fractured precedent in Seila Law v. CFPB (2020) and Collins v. Yellen (2021), which began chipping away at protections for single-director independent agencies. Trump v. Slaughter asks whether Congress can insulate members of multi-member commissions from at-will presidential removal.

The stakes extend far beyond the FTC. A ruling in the President's favor would effectively confirm that no federal official — including those overseeing the integrity of labor statistics — enjoys meaningful independence from presidential control. It would validate the unitary executive theory in its most expansive form, giving the executive branch direct leverage over every corner of the administrative state, including agencies like the BLS and OSHA whose data and enforcement actions directly govern workplace safety.

For workers and the workers' compensation industry, this is not a procedural footnote. It is an existential question about whether the data they rely on can be trusted.

What Corrupted Data Means for Workers and the Industry

The workers' compensation system is built on statistics. Actuaries price insurance premiums using injury frequency and severity data. State regulators set benefit schedules based on incidence rates. Attorneys litigate causation based on occupational exposure records. Employers design safety programs in response to loss experience data. Every link in this chain depends on the integrity of federal reporting.

When that data is unreliable — whether through employer underreporting or political manipulation — the consequences cascade:

For injured workers: Claims are denied at higher rates when injury patterns are obscured. Workers in high-risk occupations — truck drivers, construction laborers, grounds maintenance workers — face skeptical adjusters and defense counsel armed with official statistics that minimize their industry's true hazard profile. The 2024 CFOI shows construction and extraction workers suffered 1,032 fatalities — but how many non-fatal, disabling injuries went uncounted in the underlying SOII data that insurance companies use to price risk?
For the workers' compensation insurance industry: Actuarial models depend on stable, accurate loss data. If official injury counts systematically understate frequency or severity — or if political intervention distorts trend lines — pricing becomes unreliable. Reserves may be inadequate. Carriers can face unexpected losses. The entire rate-setting process, already complex across 50 state regulatory regimes, becomes even more treacherous.
For workplace safety: OSHA enforcement, inspection targeting, and regulatory rulemaking all depend on reliable injury and illness data to identify high-hazard industries and interventions that work. A politicized BLS that produces favorable statistics for politically connected industries is not merely an academic problem — it is a worker safety emergency. When the data says an industry is getting safer but workers keep dying, the gap between the official story and lived reality becomes a matter of life and death.
For the legal system: Workers' compensation attorneys, plaintiff's lawyers in tort cases, and defense counsel all rely on government statistics to establish causation, argue comparative fault, and value claims. If those statistics are compromised, every courtroom argument built upon them is built on sand.

The 104-Minute Clock Keeps Ticking

A worker died every 104 minutes from a work-related injury in 2024. That is the BLS's own figure — compiled by dedicated career professionals doing their best within an increasingly politicized institutional environment.

But if the unitary executive theory prevails in Trump v. Slaughter, if political appointees gain unfettered control over what those professionals count and how they count it, and if employer underreporting continues unchallenged by an under-resourced OSHA, the true toll will grow invisible — not because workers stop dying, but because we stop counting them honestly.

The families of the 5,070 workers who did not come home in 2024 deserve better. So do the workers still on the clock.

U.S. Bureau of Labor Statistics, National Census of Fatal Occupational Injuries in 2024, USDL-26-0230, released February 19, 2026. https://www.bls.gov/iif/fatal-injuries-tables.htm
U.S. Bureau of Labor Statistics, Survey of Occupational Injuries and Illnesses (SOII), https://www.bls.gov/iif/
Rosenman, K.D., et al., "How Much Work-Related Injury and Illness is Missed by the Current National Surveillance System?" Journal of Occupational and Environmental Medicine, 2006. (Foundational underreporting study, NIOSH-affiliated research.)
Leigh, J.P., "Economic Burden of Occupational Injury and Illness in the United States," Milbank Quarterly, 2011. (Estimates true burden of occupational harm.)
Seila Law LLC v. Consumer Financial Protection Bureau, 591 U.S. 197 (2020). (Supreme Court removal power precedent.)
Collins v. Yellen, 594 U.S. 220 (2021). (Extension of removal power doctrine.)
Trump v. Slaughter, pending before the U.S. Supreme Court (2025-2026 term). (Case challenging "for cause" removal protections for FTC commissioners under unitary executive theory.)