Recorded child labor violations rose 103 percent from fiscal 2019 to fiscal 2023 while federal investigators fell from 780 to 733

What is happening?

Federal investigators found 5,766 minors employed in violation of child labor law in fiscal year 2023, up 88 percent from 3,073 in fiscal year 2019. Over the same five years the count of recorded child labor law violations rose from 3,748 to 7,624, a 103 percent increase, and it rose in every one of the five Wage and Hour Division regions.

The number of people who do the looking moved the other way. Wage and Hour Division investigators fell from 780 in fiscal year 2019 to 733 in fiscal year 2023, and to 611 as of 14 May 2025 — reported as a 52-year low, against a 1978 peak of 1,232. That leaves roughly one investigator per 270,000 workers — the ratio that follows from dividing the 165 million covered workers by 611, although the release itself prints 278,000.

In fiscal year 2025 the division concluded more than 950 child labor cases and assessed more than $37 million in civil money penalties, up from more than $15.1 million the year before. Fiscal 2025 produced more child labor cases than any year since the Great Recession. Meanwhile four states enacted laws in 2026 that weaken child labor protections, and only three states introduced bills to strengthen them, down from 15 in 2025.

Whose problem is this?

The party that sets how many investigators exist is an appropriations committee, and the harm from having too few of them lands on children in kitchens and on kill floors. Nothing on the appropriations ledger records that harm.

Where does this problem end?

The boundary matters because two different things are being counted here. One is how many minors are working illegally, and the other is how many of them anyone finds.

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

Now

Needs a new measurementThe target state. No source we opened states an official federal target for how many minors employed in violation would be acceptable, nor a staffing floor for the investigator corps. The Wage and Hour Division states a priority rather than a number: the worst forms of child labor, and minors in dangerous occupations. The Office of Inspector General issued this review as an advisory report and made no recommendations, deferring targets to a follow-on audit announced 31 July 2025.

How big is it?

The affected population is the number of minors found employed in violation of federal child labor law in a year: 4,030 in fiscal 2024 and 5,766 in fiscal 2023, the two most recent years for which we confirmed a count of children rather than a count of cases.

This is a detection count, not a prevalence estimate. It is the number of children that 611 to 733 investigators happened to find while covering roughly 165 million workers. The true number of minors working in violation of the law is not measured by any source we opened, and it cannot be smaller than these figures.

Under what conditions does it arise?

1. The rule runs out before the conduct does. The Fair Labor Standards Act sets no maximum daily or weekly hours for 16- and 17-year-olds, and 30 states set none either. In those states a 16-year-old can lawfully work at any hour of the day or night during the school year, so there is no violation to detect on that axis. 2. Detection capacity is falling while the count rises. Investigators went from 780 to 733 across the review period and to 611 by May 2025. The approved Wage and Hour Division budget came in $25.5 million below request in fiscal 2022 and $47.7 million below request in fiscal 2023. 3. The records that enforcement runs on are being removed. Indiana eliminated its youth employment documentation system in 2026. Conduct that no one is required to write down is conduct an investigator has no ordinary way to find, so removing the record shrinks the detectable surface without changing what is lawful. 4. Federal and state rules disagree, and employers follow the wrong one. Iowa allows 14- and 15-year-olds to work until 9pm during the school year where federal law says 7pm. Iowa businesses were fined for following state rather than federal law, and that state government asked the Department of Labor to reconsider the fines.

What has been tried?

Enforcement effort and enforcement capacity moved in opposite directions over this period, which is why the table above reads as activity rather than as progress.

What was found?

Why is it still unsolved?

Enforcement absent — not in the sense that nobody enforces, but in the sense that matters: the apparatus which would make the rule bite has been thinned faster than the violations have grown.

The headline numbers look like the opposite of neglect. Penalties rose from more than $15.1 million to more than $37 million in a single year, and fiscal 2025 produced more child labor cases than any year since the Great Recession. But those records were produced by a corps of about 611 investigators covering roughly 165 million workers, one for every 270,000 of them, after the agency was funded $47.7 million below its own request in a single year. The Office of Inspector General recorded investigators saying plainly that staff are insufficient to meet the current volume of complaints, let alone to explore how far the violations extend. Concentrating a shrinking corps on child labor raises the child labor numbers and drains other programs at the same time.

The second reason is that the rule itself is retreating where the conduct is. Federal law sets no hour ceiling for 16- and 17-year-olds and 30 states set none either, so the most common form of overwork by older teenagers is not a violation anywhere. In 2026 four more states moved the line: Nebraska created a permanent subminimum wage for 14- and 15-year-olds, West Virginia removed from state code the list of hazardous occupations prohibited for minors, Washington doubled the ceiling for minors in approved work-based learning programs from four hours a day and 20 hours a week to eight hours a day and 48 hours a week, and Indiana switched off the records system that documented who was employed at all. Michigan moved by administrative rule rather than by statute, raising child-to-staff ratios so that 16-year-olds may care for young children without supervision. Conduct that stops being illegal stops appearing in the violation count, and conduct that stops being documented stops being findable.

The third reason is that nobody knows the denominator. Every figure in this dossier is a count of what was caught. No source we opened estimates what share of illegally employed minors is ever detected, so a fall in the numbers next year would be unreadable.

What observation would mean it is solved?

Candidates — (a) minors found employed in violation falls for several consecutive years (b) the investigator corps recovers toward its historical level and the workers-per-investigator ratio drops (c) the share of minors working under a documented, hour-limited regime rises as states restore permits and ceilings.

(a) alone is the trap. The count can fall three ways: fewer children are working illegally, fewer investigators are looking, or the conduct was legalized. In 2026 all three forces were active at once, so a decline would be uninterpretable without (b) and (c) beside it.

(b) alone is not enough either. More investigators would raise the detected count before lowering it, which would read as the problem getting worse. And (c) measures the legal frame rather than the behavior, so a state could satisfy it on paper while nothing changes in the workplace. No single one of these three is readable on its own.

What is it connected to?

Fills with researchlikely connected to adult wage theft enforcement, unaccompanied minor sponsorship and labor trafficking, meat and poultry processing subcontracting, and school attendance. The Office of Inspector General flagged the handling of possible trafficking of minors as a concern in the same review, but we did not investigate relation type or evidence grade.

What these sources do not say

• How many minors were found employed in violation in fiscal 2025. Case counts, hazardous-occupation counts and penalty totals for fiscal 2025 were reported, but not the number of children. The federal data page carrying that figure returned HTTP 403 on every attempt.
• Whether the rise is in the conduct or in the looking. The Wage and Hour Division attributed the fiscal 2023 jump to its own strategic enforcement initiative, and the Inspector General explicitly wrote that the limited scope of the review did not confirm this as the primary contributing factor.
• What fraction of illegal child employment is ever detected. Without a detection rate, none of these counts can be turned into a population, and no source we opened offers one.
• That the three series are the same unit. Cases, violations, and minors are counted separately — 736 cases and 4,030 minors in fiscal 2024, 7,624 violations and 5,766 minors in fiscal 2023 — and no source reconciles them or explains the ratio between them.
• Whether any specific state rollback changed any measured outcome. The state law tallies and the federal violation counts appear in different documents, and nothing we opened links a named state law to a measured change in violations or injuries.
• What the penalties mean to the employers who pay them. Penalty totals are published in aggregate, but no source we opened sets them against firm revenue, so whether $37 million deters anything is unaddressed.

See the evidence

The Inspector General report was read directly — all figures attributed to it above come from pages we opened, including Table 1, Table 2, and Exhibits 1 and 2. The year-by-year counts of minors are not printed in the running text of that report; they appear only as data labels on Figure 1, which is a chart, and they were read off the rendered page rather than taken from a source that lists them. Everything else is secondary: the fiscal 2024 and fiscal 2025 federal figures reach us through legal and trade press citing Department of Labor data, because the underlying data page returned HTTP 403.

Three counting bases must not be confused, and the sources do not always separate them. The Inspector General counts violations and minors as distinct series, while the fiscal 2024 and 2025 press figures count cases. A single case can carry many violations and many minors, which is why 736 cases in fiscal 2024 and 7,624 violations in fiscal 2023 are not comparable quantities. The 2026 state legislative tally reached us through republications rather than through the original Economic Policy Institute analysis, which returned HTTP 403, as did the Stateline report of it. A syndicated copy of that Stateline report opened, and it names the four enacting states as Indiana, Nebraska, Washington and West Virginia. The People's World republication does not contradict that list, because it describes the Michigan change of the same year as administrative rules effective April 2026 rather than as a law passed by legislators. Michigan is therefore counted in this dossier as a rule change and not as one of the four.

People affected

Estimated range 4,030–5,766 As of FY2023~FY2024

Sensitivity The range is not a confidence interval. It is the span between the two most recent fiscal years for which a count of minors was confirmed, and the lower bound is the later year, so the two endpoints do not describe a trend. The more important limit is the axis: this counts minors that federal investigators found, not minors who were illegally employed. Detection is the binding constraint on this number, and it is shrinking. Wage and Hour Division investigators fell from 780 in FY2019 to 733 in FY2023 and to 611 as of 2025-05-14, roughly one for every 270,000 covered workers, so the figure is a floor whose distance from the true population is unknown. No source we opened offers a detection rate that would let that distance be estimated. The FY2025 value is probably above this range, since concluded child labor cases rose from 736 in FY2024 to more than 950 in FY2025 and penalties rose from more than $15.1 million to more than $37 million, but the FY2025 count of minors was not confirmed and is deliberately excluded rather than inferred from the case count. A second limit runs the other way: where states and the FLSA set no hour ceiling for 16- and 17-year-olds, which is the case in 30 states, overwork by older teenagers never becomes a violation and so never enters this count at all

Regional breakdown The OIG publishes counts by the five WHD regions rather than by state, and it does so only for child labor law violations, not for the number of minors that this population uses. Regions are multi-state groupings that do not map onto state codes, and splitting a national count of minors across states by population share would be proportional allocation, which is forbidden and is especially baseless here because the violations concentrate in specific industries such as limited-service restaurants rather than following population