Section 10(c) names no fine for an unlawful discharge — the Board tried to raise that ceiling by interpretation and three circuits vacated the result by 2026-02-04

What is happening?

The National Labor Relations Act gives the Board one lever against an employer that fires a worker for organizing, and that lever has a ceiling written into it. Section 10(c), codified at 29 U.S.C. 160(c), lets the Board order a violator to cease and desist and to take affirmative action including reinstatement of employees with or without back pay. The subsection carries no fine, no civil penalty and no damages language. Meanwhile the wait has more than tripled — the median number of days from the filing of an unfair labor practice charge to the issuance of a complaint stood at 456 as of the end of June 2026, against 441 days for calendar 2025 and a band of 100 to 150 days across 2012 to 2021.

The machinery around that ceiling is busy. In fiscal year 2025 the agency received 19,754 unfair labor practice charges, the Regional Offices issued 704 complaints, more than USD 64 million was recovered on behalf of employees, and 854 offers of reinstatement were secured. What is bounded is the remedy, not the presence of an enforcer.

The obvious explanation for the delay does not hold. The Board sat without a quorum from 2025-01-28, but issuing a complaint is a function of the General Counsel and the Regional Offices and does not require a Board quorum. A quorum returned when two members were sworn in on 2026-01-07, a new General Counsel named the backlog as the first priority, and through the end of June 2026 the offices issued 334 complaints against roughly 668 by the same point in earlier years while dismissing 4,418 cases. Incoming charge volume did not fall. The charge-to-complaint median rose rather than fell, and no source opened here names a cause.

Whose problem is this?

The body that would have to raise the price cannot. The Board may order only what Section 10(c) names, and between 2024 and 2026 it lost in three circuits for trying to read more into that sentence. Everyone with the authority to change the number sits outside the agency that hears the case.

Where does this problem end?

The clock that runs against the worker is six months and nothing comparable runs against the agency. Section 10(b) bars a complaint on conduct occurring more than six months before the charge was filed, which is a limit on the person who was fired rather than on the institution that decides.

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

Now

Needs a new measurementthe target state: no source opened here names a target figure for the charge-to-complaint interval, and none says what that interval should be. The only published target found is a Board goal of 180 days or less for the median age of pending cases, which measures a different quantity — the age of the inventory rather than the wait for a complaint. Nothing opened here proposes a target for the monetary remedy either, other than the penalty amounts written into a bill that has not been enacted.

How big is it?

Between about 854 and about 5,173 workers a year, as of 2025. The two ends sit on different clocks and answer different questions, and both labels matter. The low end is the number of workers for whom the Board secured an offer of reinstatement in fiscal year 2025. The high end is an estimate of the number of employer charges alleging a discharge or discipline filed in calendar year 2025, obtained by taking the 11,246 non-postal charges filed against employers that year and applying the 46 percent share that discharge and discipline allegations hold among settled employer charges, which gives about 5,173.

The distance between those two numbers is the distance between an allegation entering the system and a remedy leaving it.

Both ends are wrong in known directions. The high end rests on a share measured on settled charges, which are the merit-favorable subset, and no source opened here gives the allegation mix for filed charges, so the bias has an unmeasured size and an unknown sign. The high end is also itself a floor, because a single charge can name several workers. And neither end counts the people the act was aimed at: the co-workers who see a colleague fired and stop organizing leave no docket entry, and no source opened here has tried to size that group.

The low end overstates relief received, because it counts offers rather than returns. If the long-run declination rate of 76 percent for 2016 through 2023 held in fiscal year 2025, the number of workers actually restored to a job that year would be closer to about 205. No source opened here publishes the fiscal year 2025 split, so that figure is an inference drawn here and not a reported value.

Under what conditions does it arise?

1. The ceiling is in the statute rather than in a policy. Section 10(c) authorizes equitable relief and names no penalty. An agency cannot charge a price the statute does not name, and the attempt to read foreseeable pecuniary harms into the make-whole remedy was rejected by the Third Circuit in late 2024, the Fifth Circuit on 2025-10-31 and the Sixth Circuit on 2025-11-05, and again by the Fifth Circuit on 2026-02-04.

2. The bill is netted down before it is paid. Under the traditional method, interim earnings are deducted from the gross back pay owed to a wrongfully discharged employee. King Soopers, Inc., decided 2016-02-19, changed only the treatment of search-for-work expenses and left the offset itself intact. The maximum monetary exposure is therefore the wage differential the worker actually suffered, with no multiplier and no floor.

3. The bill arrives years after the act. The charge-to-complaint median was 456 days as of the end of June 2026, and complaint to decision by an administrative law judge added a further 245 days in calendar 2025, with Board review and court enforcement after that.

4. The remedy that repairs the campaign decays fastest. Of the three tools available for an illegal firing — notice postings, back pay and reinstatement — only reinstatement repairs the organizing drive. Since 2016, 76 percent of workers offered reinstatement declined it, from 15,328 offers and 11,627 declinations recorded between 2016 and 2023. Studies collected in 2000 found acceptance falling from 93 percent when the offer arrived within two weeks to 5 percent when it arrived after six months.

5. The fast path is not fast. Section 10(j) allows a petition to a district court for temporary relief once a complaint has issued, which places it downstream of the 456 days. Reaching court has averaged 296 days with a further 108 days in district court proceedings, and on 2024-06-13 the Supreme Court required district courts to apply the traditional four factor test, including a showing of likely irreparable harm, to those petitions.

A price that is capped by law and paid years late is not a price that changes conduct.

What has been tried?

Two directions have been tried and both aimed at the same ceiling. The Board tried to raise it by interpretation and lost in three circuits, and Congress has a drafted bill that would raise it by statute and has not passed it. Nothing tried so far has moved the number a violator can expect to pay.

What was found?

Why is it still unsolved?

Cost structure — the statute prices an unlawful discharge at wage replacement, and the wage replacement arrives years later and shrinks on the way.

The first movement is the ceiling. Section 10(c) lets the Board order a violator to stop and to take affirmative action including reinstatement with or without back pay, and it names no penalty. When the Board tried in 2022 to read foreseeable pecuniary harms into the make-whole remedy, three circuits vacated the result between late 2024 and 2026-02-04, with the Fifth Circuit reasoning that Section 10(c) allows equitable remedies rather than legal ones and that ninety years had passed before the agency first claimed the wider power. The maximum monetary exposure for firing an organizer is the wage differential the worker actually suffered, reduced by whatever the worker earned elsewhere.

The second movement is time, and it works on both halves of the remedy. The median charge-to-complaint interval was 100 to 150 days across 2012 to 2021, 441 days for calendar 2025 and 456 days as of the end of June 2026, with a further 245 days from complaint to decision by an administrative law judge in calendar 2025. By the time an order lands, the thing the discharge was aimed at has already been settled. This is where the non-monetary half dies: 76 percent of workers offered reinstatement since 2016 declined it, and acceptance in the studies collected in 2000 fell from 93 percent within two weeks to 5 percent after six months. The remedy that repairs an organizing drive is the one that delay destroys first.

The third movement is that the escape hatch was narrowed inside the same window. Section 10(j) exists because Board proceedings are slow, and the Supreme Court recorded that reasoning on 2024-06-13 while nonetheless requiring the full four factor preliminary injunction test. The tool built to outrun the delay now runs at roughly the speed of the delay.

The cost of the act is capped at wage replacement while the benefit of the act is the defeat of an organizing campaign. Nobody defects from this arrangement because nobody in it can. The worker cannot, because the six month limitation in Section 10(b) runs against the worker while nothing comparable runs against the agency. The agency cannot, because it has spent 2022 through 2026 losing in three of the four circuits to rule on it. Congress could, and the drafted bill setting USD 50,000 and USD 100,000 penalties sat in committee as of the source opened. On the other side of the ledger an advocacy side research institute and a nonprofit labor research group estimate that roughly USD 1.7 billion a year flows to union avoidance consultants and law firms. That estimate is modelled spending on lawyers and consultants, and the USD 64 million recovered in fiscal year 2025 is agency recovery across every kind of charge. They are not the same quantity and nothing here divides one by the other, but the order of magnitude of the first is what a market looks like when the legal downside is bounded and known.

What observation would mean it is solved?

Candidates — (a) the charge-to-complaint median returns to the 100 to 150 day band of 2012 to 2021 (b) a monetary penalty for an unfair labor practice exists in statute and is actually assessed (c) the share of reinstatement offers that workers accept rises materially above the 24 percent implied by the 2016 through 2023 figures.

(a) alone is weak. A faster path to a bounded remedy is still a bounded remedy. The first half of 2026 also shows how the median can be driven by disposal policy rather than by speed, since 4,418 cases were dismissed while only 334 complaints issued and the median still rose.

(b) alone is weak. A penalty that nobody assesses within 456 days is discounted to something close to nothing by the time it arrives. The penalty amounts in the drafted bill have never been scored, so no source opened here says how many violations would draw one or how much would ever be collected.

(c) alone is weak. Acceptance of a reinstatement offer moves with the labor market and with what the employer did after the firing. A rise could mean the remedy improved, or it could mean workers had fewer outside options that year.

Each of the three can move for reasons that have nothing to do with what a violation actually costs. They have to be read together, and the second has to be read against how long it took to assess.

What is it connected to?

Fills with researchfirst contract bargaining delay, the misclassification of workers as independent contractors, and the exclusion of agricultural and domestic workers from the Act. Relation type and evidence grade were not confirmed in this round.

What these sources do not say

• What an unlawful discharge actually costs an employer. The agency reports an aggregate recovery of more than USD 64 million for fiscal year 2025 and separately reports 854 offers of reinstatement, and nothing opened here allows dividing one by the other, because the aggregate covers every kind of charge rather than discharges alone. The number a fired worker would want before filing does not exist in public form.
• How much of a gross award the interim earnings offset removes. The deduction is the operative rule and it is the mechanism that bounds the cost, and its size is published nowhere reached in this round.
• What the charge-to-complaint interval ought to be. The only published target found is a Board goal of 180 days or less for the median age of pending cases, which is a different measure. No source opened here names a target for the 456 day figure.
• What happens in the cases that do not reach a complaint. The source that produced the 441 and 456 day medians states that the vast majority of cases settle before a complaint issues, and no source opened here gives a processing time distribution for that settling majority.
• Why the interval kept growing after the quorum was restored. Complaint issuance does not require a Board quorum, and the median rose from 441 to 456 during the first half of 2026 while a new General Counsel dismissed 4,418 cases and issued about half the usual number of complaints. No source opened here names a cause for that movement.
• Whether delay explains the declination rate today. The only figures relating waiting time to acceptance come from studies collected in 2000, and the 76 percent declination rate for 2016 through 2023 is published with no breakdown by how long the worker waited.
• How many of the fiscal year 2025 offers were accepted. The 2016 through 2023 totals give 15,328 offers and 11,627 declinations, while the 854 offers reported for fiscal year 2025 come with no acceptance figure at all.
• What the two figures of 245 days have in common. One source published on 2026-01-20 gives 245 days as the calendar 2025 median from complaint issuance to decision by an administrative law judge, and another published on 2026-01-27 gives 245 days as the median age of pending cases at fiscal year end. Neither cites the other and neither acknowledges the coincidence.

See the evidence

One primary document was read end to end and it is the load-bearing one. The statutory text of 29 U.S.C. 160 was opened directly, and it is the source of the central claim that the subsection names no fine, no civil penalty and no damages. The Supreme Court opinion in Starbucks Corp. v. McKinney and the text of S. 852 were also read directly. Everything else reaches the reader second hand. The most important consequence is that no figure published by the agency itself was verified at the agency: every NLRB number here — 19,754 charges, 704 complaints, USD 64 million, 854 offers of reinstatement, 76 percent declinations — arrives through a reporter, because the agency domains refused connections throughout this round. The 456 day and 441 day medians rest on a single analyst working from a single private docket dataset, and the same holds for the 245 day complaint-to-decision figure. Two publishers here argue theses and are named in the source column for it — the backlog series comes from a union avoidance consultancy that attributes the slowdown to enforcement priorities set in 2021, and the union avoidance spending estimate comes from an advocacy side research institute that attributes delay to employers exploiting Board procedure. Neither of them explains why the median rose again in the first half of 2026 after the guidance was rescinded and the docket was being cleared, and that gap is left open rather than filled. Four disagreements are left visible rather than resolved. Two sources published a week apart attach 245 days to two different measures and neither acknowledges the other. Charge counts sit on different clocks and different scopes, with 19,754 for fiscal year 2025 covering all unfair labor practice charges and 11,246 for calendar 2025 covering only non-postal charges against employers, which is a subset on a different calendar rather than a contradiction. Quorum restoration is dated to the confirmation on 2025-12-18 by one publisher and to the swearing in on 2026-01-07 by another, and both are correct about different events. And the cause of the delay is contested between the two advocacy sources named above. One figure in this document is an inference drawn here rather than a reported value, and it is labelled as such in block 5 — the estimate of about 205 workers actually restored to a job in fiscal year 2025. This is a Path A output (research-based definition), so observation_refs is empty and provenance_mode: press-derived.

People affected

Estimated range 854–5,173 As of 2025 · fiscal year 2025 for the low bound and calendar year 2025 for the high bound

Sensitivity The width of this interval is not statistical uncertainty. The low bound of 854 is the number of workers for whom the Board secured an offer of reinstatement in fiscal year 2025, reported by a publisher summarising agency figures that could not be opened at the agency in this round. The high bound of 5173 is an estimate of discharge and discipline allegations filed against employers in calendar year 2025. The two ends therefore sit on different clocks, and the distance between them is the distance between an allegation entering the system and a remedy leaving it. Three corrections push in different directions and none of them can be computed from the sources opened here. Upward, because the high bound counts charges rather than workers and a single charge can cover many people at once, and because neither bound counts workers who never filed at all, whether because the six month limitation in Section 10(b) had already run or because the wait was not judged worth it. Upward again, and by far the largest omission, because neither bound counts the group the discharge was actually aimed at: the co-workers who watch a colleague be fired and stop organizing leave no docket entry by construction, and no source opened here has tried to size that group. Downward, because the low bound counts offers rather than returns. If the long-run declination rate of 76 percent recorded for 2016 through 2023 held in fiscal year 2025, the number of workers actually restored to a job that year would be closer to about 205. No source opened here publishes the fiscal year 2025 acceptance split, so that figure is an inference drawn in this document rather than a reported value, and it is labelled as one. The bias attached to the 46 percent share is separately unresolved, because that share is measured on settled charges and applied to filed charges, and its sign is unknown.

Regional breakdown No source opened in this round publishes a regional or state breakdown of discharge allegations or of reinstatement offers. The underlying agency case records do carry a Region field and a state field, so the absence is a silence in the published analyses rather than a limit of the data, and the agency domains that would host such a breakdown returned connection refused at the network level on every attempt in this round. Splitting either bound by state population would be proportional allocation, and organizing activity and the discharges that follow it are concentrated by industry and by employer rather than distributed with population.