A willful workplace safety violation is priced at a maximum of USD 165,514 per citation — a statutory base last set in 1990

What is happening?

United States federal law prices the worst category of workplace safety violation at a maximum of USD 165,514 per citation as of 2026-08-08. That ceiling covers a willful violation and a repeated violation. A serious violation, the next class down, is capped at USD 16,550. Neither figure has moved since 2025-01-15.

The ceiling is per violation and not per case, and that distinction has to be stated before anything else. Both the statute and the regulation attach the amount to each violation, and OSHA has issued many citations out of a single inspection. The largest entry on the OSHA list of top enforcement cases is a single employer cited on 2009-10-29 at USD 81,340,000, which is 491 times the ceiling that now applies to one willful citation. So USD 165,514 is the price of one citation, not a limit on what one employer can be charged for one event.

What produces that price is a formula with no term for harm, no term for firm size, no target level and exactly one input. The statutory base in 29 U.S.C. 666 still reads USD 70,000 for a willful or repeated violation and USD 7,000 for a serious one, set by Congress on 1990-11-05. Everything above those figures is inflation indexing — a one-time catch-up multiplier of 1.78156 effective 2016-08-01, then an annual step keyed to the consumer price index for urban consumers. The ratio between today and the base is 2.365 for the willful ceiling and 2.364 for the serious one, which is one multiplier applied to a thirty-six-year-old number.

In 2026 that machinery produced zero. The Office of Management and Budget cancelled the annual adjustment on 2026-04-17 because the lapse in appropriations of October and November 2025 stopped the Bureau of Labor Statistics from producing the October 2025 price data the formula requires, and the statute names no alternative. OSHA carried the cancellation into an agency memo on 2026-05-21 and the Department of Labor made it final in a rule effective 2026-05-27. The 2026 ceiling is therefore not adjusted for inflation. It is frozen at the 2025 nominal level in a year with positive inflation.

Below the ceiling, the amounts actually assessed sit an order of magnitude lower. In FY2025 the average federal penalty for a serious violation was USD 4,678, and the median federal penalty in a case where a worker was killed was USD 16,550 — exactly one maximum serious citation, and about 10 percent of the willful ceiling.

Whose problem is this?

The body that sets the ceiling and the body that sets the amount actually paid are not the same body, and neither of them is asked to state a level of deterrence. Congress owns a number it has not touched since 1990, and the agency owns a schedule of reductions it moved downward in 2025.

Where does this problem end?

The boundary matters here because the schedule already exists and is applied every working day. What is in question is not whether there is a penalty but how the number is arrived at, and whether anything in that arithmetic responds to what the violation cost.

What is the state now, and what should it be?

Now

Needs a new measurementthe target state: no source opened here names a penalty level that would be adequate to deter, or any criterion by which adequacy would be judged. The 2015 indexing statute, the OMB memorandum, the OSHA memo, the Department of Labor rule and the OSHA penalties page all describe a mechanical price-index computation and nothing else. No document opened here says that a target was ever set, or considered, or rejected.

How big is it?

Between 130 million and 155 million workers stand under this penalty schedule. The two ends are two published counts of the same thing that do not agree. OSHA states approximately 130 million workers at more than 8 million worksites, on a page whose other figures are FY2024 inspections and 2023 fatalities. A national labor federation report of April 2026 states 155 million workers under OSH Act jurisdiction at 12 million workplaces. Neither source mentions the other, and no document opened here reconciles them.

That figure counts people standing under the ceiling, not people to whom it was ever applied. The ceiling attaches only to willful and repeated citations, and on the only breakdown found there were 650 willful and 5,035 repeated citations nationwide in calendar 2025, out of 132,762 citations across 69,996 inspections. The narrowest count is people for whom the worst outcome actually occurred: 5,070 fatal work injuries in 2024, a rate of 3.3 per 100,000 full-time equivalent workers.

The gap between those quantities is the finding rather than a defect of the estimate. A ceiling that nominally governs a national workforce is actually tested a few thousand times a year. The number of workers employed at the establishments that received a willful or repeated citation is not derivable from anything opened here — OSHA inspection records carry employee counts, but no summary page publishes them and the enforcement summary reports inspections only, so the middle of the range cannot be pinned.

Under what conditions does it arise?

1. The base is statutory. The numbers in 29 U.S.C. 666 are USD 70,000 for a willful or repeated violation and USD 7,000 for a serious one, and they are still the operative text of the Code. Only Congress can change them, and it last did so on 1990-11-05. Between November 1990 and October 2015 the ceiling stood still in nominal terms while prices rose 78.2 percent.

2. The only automatic motion is a consumer price index. The 2015 statute indexes the base to a measure of household consumption prices and applies it to a sanction on employers. Nothing in the formula refers to firm size, revenue, the number of workers exposed, the severity of the injury, or the cost of the abatement that was not performed.

3. The formula has one input and no fallback. The multiplier is computed from the price index for October of the prior year. In 2026 that month of data did not exist, and the statute names no alternative, so the legally correct output was zero rather than an estimate. OMB recorded that 2026 is the first year adjustments are not required, and that deviating from the statute would create significant litigation risk.

4. The number actually assessed is set by a separate chain. Classification, the reduction factors in the Field Operations Manual for size, good faith and history, and settlement all sit between the ceiling and what an employer pays. Those factors moved downward on 2025-07-14, when the history-based reduction was raised from 10 percent to 20 percent and size-based reductions for smaller employers were expanded.

5. The ceiling is rarely reached because an inspector rarely arrives. OSHA reports 30,273 federal inspections in FY2025 and approximately 1,850 inspectors; the labor federation report counts 1,651 inspectors across federal and state programs for 12 million workplaces, which it renders as one inspection per workplace every 191 years. A price only bites where somebody charges it.

What has been tried?

Two directions have been running at the same time. The indexing machinery moves the ceiling up by a small step in most years, and the enforcement manual moved the assessed amount down in 2025. The one attempt that would change the base itself is a bill that has not left committee.

What was found?

Why is it still unsolved?

Cost structure — the price of the worst category of employer conduct is produced by a formula that contains no term for harm, no term for firm size and no target level, and the realized price sits an order of magnitude below even that formula.

The first movement is the base. Congress wrote USD 70,000 and USD 7,000 into the statute in November 1990 and has not returned to them. From that date to October 2015 the ceiling did not move at all in nominal terms while prices rose 78.2 percent, which left roughly USD 39,300 of 1990 purchasing power in a USD 70,000 ceiling by the time the catch-up rule restored it once in 2016. Since then the only motion has been an annual step of about 2.6 percent, roughly USD 4,200 on the willful ceiling. A base that only Congress can move, and that Congress moves once every few decades, is a base that spends most of its life falling.

The second movement is the mechanism that replaced Congress. Indexing was the answer to a frozen base, and its virtue is that it is mechanical. The 2026 cancellation is the price of that virtue. When one month of price data did not exist, the statute offered no substitute and the agencies concluded they had no authority to invent one, so the adjustment for the entire federal government came out as zero. The number that prices a workplace death now depends on whether a statistical agency was funded in a particular October. Nothing in that chain asks how much deterrence is needed, and OMB treats the whole national penalty adjustment as economically insignificant because its annual impact falls under USD 100 million.

The third movement is what happens under the ceiling, and it points the same way. Classification, the reduction factors and settlement set the amount actually assessed, and that chain moved downward in July 2025 while the ceiling stood still. The observable result is an FY2025 average of USD 4,678 for a serious violation and a median of USD 16,550 where a worker was killed. Then the probability term multiplies it down again: on the only breakdown found, willful citations were 650 for the entire country in a year, and at the reported staffing each workplace would be inspected once every 191 years. A sanction that is small, capped, mechanically predictable and rarely charged is a sanction that can be budgeted for in advance.

One thing this document does not claim. No source opened here compares an OSHA penalty to the capital cost of the abatement that was not performed, so the familiar assertion that the fine is cheaper than the fix is left unstated rather than assumed.

What observation would mean it is solved?

Candidates — (a) Congress raises the statutory base in 29 U.S.C. 666 (b) the average and median penalties actually assessed rise toward the ceiling (c) the fatal work injury rate falls for several consecutive years.

(a) alone is weaker than it looks. Raising a ceiling changes the maximum, not the amount assessed. The chain below the ceiling is untouched by the number in the statute, and it moved downward in July 2025 while the ceiling stood still. H.R. 6597 would take the willful maximum to USD 800,000 and the willful minimum to USD 60,000, and the minimum is the part that would actually bind on ordinary cases, but no source opened here models what either figure would do to what employers pay.

(b) alone is worse. Assessed amounts can rise because inspectors write more willful citations, because a few very large cases land in one year, or because settlement practice changed, and an aggregate cannot separate those from a change in deterrence. The counts are also thin enough that a handful of cases moves the average, and the only source found for the classification breakdown is an unofficial aggregator that says of itself that it is not an official OSHA resource.

(c) alone counts the wrong thing at the wrong resolution. Fatality counts move with employment levels, industry mix and the business cycle, and they arrive a year late. They also cover traumatic injury only. The advocacy report estimates 135,000 deaths a year from occupational disease against the 5,070 traumatic injury deaths in the federal count, and a disease death separated from its exposure by decades rarely produces a citation at all, so no penalty schedule of any level would be tested against it. The three have to be read together, and (b) has to be read against the number of cases it rests on.

What is it connected to?

Fills with researchpenalty design in other federal safety and environmental regimes, State Plan penalty levels and how they are actually assessed, the criminal referral path under 29 U.S.C. 666(e) and why it is used so rarely, the effect of the 2026 cancellation on every other federal agency that indexes penalties the same way, and comparable ceilings in other countries. Relation type and evidence grade were not confirmed in this round.

What these sources do not say

• What penalty level would be adequate. Not one source opened here states a target, or a criterion for judging adequacy, or a record that anybody ever set one. The whole chain from statute to agency page describes a computation and stops.
• Why a household price index is the right multiplier for an employer sanction. The measure that moves the ceiling tracks consumer consumption prices. No document opened here defends that choice or names an alternative that was considered and rejected.
• What the penalty costs against what abatement costs. No source opened here places an OSHA penalty next to the price of the safety measure that was not taken. That comparison is the one most often asserted about this subject and it is absent from everything read here.
• That the 2026 number did not move. The OSHA penalties page presents the current table as carrying the annual adjustment for inflation, while the OSHA memo of 2026-05-21 states there were no inflation-based increases for 2026. A reader who consults only the penalties page cannot learn that this year the adjustment was cancelled.
• How assessed penalties are distributed. OSHA publishes no distribution at all. The enforcement summary carries an inspection table for FY2020 through FY2025 and no counts of violations by classification and no penalty totals — no median, no share assessed at the maximum, no share reduced in settlement. Every average and median in this document that bears on deterrence comes from a union report or an unofficial aggregator, not from the agency that holds the records.
• Whether the largest-case list is current. The OSHA page of top enforcement cases carries no as-of date and its most recent entry is dated 2010. A reader cannot tell whether the sixteen-year gap is a fact about enforcement practice or a fact about page maintenance. The same page family is stale elsewhere: the public enforcement search tools state their data are current only through 2025-12-04.
• Whether a cancelled year is ever recovered. If the 2027 adjustment is computed from October 2026 data against the 2025 base, the 2026 increment is forfeited permanently. The catch-up authority in the 2015 statute was a one-time provision used in 2016. OMB says only that agencies continue using the 2025 levels.
• What the State Plans actually charge. State Plans must adopt maximum penalty levels at least as effective as the federal ones, but no source opened here publishes a state-by-state table of actual maximums and no source defines that standard numerically. The same OSHA page states that State Plans are not required to impose monetary penalties on state and local government employers, and no source quantifies how many workers are therefore covered by a safety regime with no monetary sanction behind it at all.

See the evidence

The primary documents in this file were read directly and they carry the load. The statute, the 2016 catch-up rule, the 2026 Department of Labor rule, the OMB memorandum, the regulation text, the OSHA memo and the bill text as introduced were all opened, and they agree with each other on every figure that appears in more than one of them — the base amounts, the 1990 date, the 2016 multiplier, the current maximums and the cancellation of the 2026 adjustment. Everything that bears on what is actually assessed is second hand. The average of USD 4,678 and the median of USD 16,550 come from an advocacy report, and that report could only be read through a text extraction proxy because the publisher returns HTTP 403 — what was verified is a proxy render rather than the response of the publisher, and the same applies to the federal fatality release. The calendar 2025 classification counts and the total penalties assessed rest on a single unofficial aggregator that describes itself as not an official OSHA resource, and no agency page publishes either figure, so they carry a low grade and no attempt has been made to reconcile them against the federal FY2025 inspection count. Two disagreements are left visible rather than resolved. OSHA reports approximately 1,850 inspectors, 130 million workers and more than 8 million worksites, while the labor federation report gives 1,651 inspectors, 155 million workers and 12 million workplaces, and neither source acknowledges the other or states whether its inspector count includes State Plan staff. And the OSHA penalties page presents the 2026 table as carrying the annual inflation adjustment while the OSHA memo of the same year states that no such adjustment occurred, which is a contradiction inside one agency rather than between two sources. This is a Path A output (research-based definition), so observation_refs is empty and provenance_mode: press-derived.

People affected

Estimated range 130,000,000–155,000,000 As of 2026 as published; the agency count carries no stated reference year for the worker figure

Sensitivity The width of the band is a disagreement between two published counts of the same covered workforce rather than a measurement interval, so it does not narrow with better arithmetic and no source opened here reconciles it. What the number fails to count is the distance between standing under the ceiling and ever meeting it: on the only breakdown found, an unofficial aggregator of Department of Labor enforcement data, only 5,685 of 132,762 citations in calendar 2025 were willful or repeated, which are the two classes the USD 165,514 ceiling attaches to, and at the reported staffing of 1,651 inspectors for 12 million workplaces each workplace would be inspected once every 191 years. The narrowest reading of the same population is the 5,070 people who died of a traumatic work injury in 2024, which is a count of outcomes rather than a subset of the citations the ceiling priced. The limit in the opposite direction is that the band counts covered workers only. Workers outside OSH Act coverage are not enumerated by any source opened here, state and local government employees in State Plan states may sit under a regime with no monetary penalty at all because the plans are not required to impose one on public employers, and the number of workers employed at the establishments that actually received a willful or repeated citation is published nowhere found, so the middle of the range cannot be pinned.

Regional breakdown No source opened here publishes covered-worker counts, citation classifications or assessed penalties by state. The federal OSHA and State Plan universes are reported separately only for inspector headcount, and no state-by-state table of actual maximum penalty levels was found.