Catch documentation reaches 32 percent of US seafood imports by volume — the rule that would widen it was withdrawn in 2023

What is happening?

The United States requires catch documentation for imported seafood in 13 species groups and in none of the rest. In fiscal year 2023 that scope reached 773,238,048 kilograms out of roughly 2.4 billion kilograms of seafood the country imported — 32 percent by volume and 30 percent by value, on the figures the National Marine Fisheries Service reported to Congress in May 2024. The other two-thirds by volume entered US commerce under no catch-documentation requirement at all.

The instrument is the Seafood Import Monitoring Program, created by a final rule of 2016-12-09 under the Magnuson-Stevens Act and phased in during 2018. It requires the importer of record to hold an international fisheries trade permit, to report harvest and landing data at entry through the Customs and Border Protection filing channel, and to keep a chain-of-custody record for two years. It is not a labeling program, the data it collects are confidential, and the agency does not use it to attest that any particular shipment is legal.

Widening the species list was tried once. The Fisheries Service proposed an expansion on 2022-12-28 that would have moved coverage from about 1,100 species toward about 1,670, drew more than 2,000 public comments, and then withdrew the proposal on 2023-11-16 in order to run a comprehensive program review. The review finished in the summer of 2024. On 2024-11-14 the agency published an action plan whose first commitment was to extend traceability requirements to all US seafood imports through a two-tier system. That plan contains no date, no deadline and no milestone.

Then the direction reversed. Executive Order 14276 of 2025-04-17, published 2025-04-22, directs the Secretary of Commerce to immediately consider revising or rescinding recent expansions of the program to unnecessary species. Neither the action plan nor the executive order has produced a rule. A Federal Register query on the rulemaking identifier for this program returns three documents in total, the last of them the 2023 withdrawal. The federal regulatory agenda moved the replacement proposal from 12/00/2025 to 08/00/2026, which is the month now underway, and as of 2026-08-08 nothing has published. Coverage is unchanged at 13 species groups.

Whose problem is this?

The body that would have to widen the list is the same body that published a plan to widen it to everything. Nothing outside that body can make it finish, and the shipments that arrive in the meantime are not recorded anywhere as a decision by anyone.

Where does this problem end?

The boundary matters here because the machinery already exists and only the list of species is missing. A shipment outside that list is not lightly checked or partly checked. It carries no catch-documentation requirement at all, so there is nothing on it to check.

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

Now

Needs a new measurementthe target state, as of 2026-08-08: no source opened here names a coverage figure the program is trying to reach. The November 2024 action plan commits to all US seafood imports and sets no date. The Unified Agenda entry carries a month for a proposal and no target for coverage. The executive order points the other way and names no figure either. Because there is no number the program is meant to reach, there is also no published basis on which to say it is behind.

How big is it?

About 340 million people, which is the entire US resident population, put at 340,003,797 for 2024. That headcount is not a claim about how many people ate illegally caught fish. It is a claim about who cannot check, and the reason it takes in everyone is that the program is not consumer facing and produces no signal a buyer can read.

The exposure behind that headcount is a per-person quantity. NOAA reports per-capita seafood consumption of 19.1 pounds in calendar year 2023 and estimates that 80 percent of the seafood eaten in the United States that year was imported, which gives about 15.28 pounds of imported seafood per resident per year. Applying the FY2023 uncovered share of 68 percent by volume gives about 10.4 pounds per resident per year of imported seafood that entered under no catch-documentation requirement. Applying instead the about-half figure carried on the public program page gives about 7.6 pounds.

That spread is not measurement noise. It exists because the agency does not agree with itself about how much of the import stream its own program covers, and the widest, most favourable figure is the one on the public-facing page while the narrowest is the one in the report to Congress.

A narrower headcount, counting only residents who actually eat seafood, is not derivable from anything opened here. The one source that would supply a seafood-consumer prevalence returned HTTP 403 on every path tried, so a lower headcount would have to be invented, and it is left out rather than guessed.

Under what conditions does it arise?

1. Scope lives in rulemaking, and rulemaking is where it died. The 13 species groups are set by regulation, so widening them requires notice and comment. The single attempt drew more than 2,000 submissions from constituencies that wanted opposite things — importers who called the expansion burdensome, and conservation groups and members of Congress who called it far too narrow, as recorded by the Congressional Research Service — and the response to irreconcilable comment was withdrawal rather than choice. A rule that satisfies neither side can be withdrawn. The status quo needs no rule and therefore needs no agreement.

2. A species list is the wrong shape for the evasion it invites. Coverage attaches to the declared species, so a different declaration moves a shipment out of the program entirely. The agency review summary records stakeholders making exactly that point, and the FY2023 audit results show that the declared country of harvest was frequently unverifiable from the documents supplied. The declared species rests on the same documents.

3. Even inside the covered third, verification runs at a quarter of one percent. In FY2023 the agency selected 329 of 130,710 covered filings for audit, and about 56 percent of what it audited came back noncompliant. The design leans on deterrence by documentation, because detection at that rate cannot be doing the work.

4. The direction of travel reversed between administrations while the file was open. The November 2024 plan says extend to everything. The April 2025 executive order says consider rescinding expansions to unnecessary species. Neither has produced a rule, the agenda date slipped by eight months, and the list is now held in place by the absence of agreement rather than by a decision to keep it as it is.

What has been tried?

Two directions were tried at once — widen the list, and improve what happens inside it. The first was withdrawn and never reissued, and the second is measured only by audits that reach a quarter of one percent of covered filings.

What was found?

Why is it still unsolved?

Institutional gap — the requirement exists, the machinery to run it exists, and the scope of the requirement does not reach most of what arrives.

Start with what is actually missing. This is not a case of a rule being broken. For the roughly two-thirds of import volume outside the list there is no rule to break. An importer of an uncovered species files an entry with no harvest data, no landing data and no chain of custody, and has complied fully. Enforcement cannot reach into that space, and no amount of audit capacity would change it, because there is nothing in the space to audit. That is why the failure sits at scope rather than at enforcement, even though a second and real enforcement hole sits inside the covered third.

The second movement is that scope can be changed only through the one process that has already failed at it. Species coverage lives in regulation, so widening it means notice and comment, and notice and comment on this program produced more than 2,000 submissions pulling in opposite directions. The agency withdrew rather than decide. Withdrawal carries an asymmetry that is easy to miss. Extending the list requires a completed rulemaking, and leaving the list alone requires nothing whatsoever. The path of least effort points at the status quo every single time, and it does so without anyone having to choose it.

The third movement is that nothing about the delay is visible as an event. There is no missed statutory deadline, no court order and no expiring authority. The program keeps running, the reports to Congress keep arriving, the agenda entry keeps a month in it, and that month keeps moving. A plan published in November 2024 promised complete coverage and named no date, so it cannot be late. An executive order five months later pointed the opposite way and named no date either, so it cannot be late. Two commitments in opposite directions, neither of them enforceable against a calendar, leave the species list exactly where it stood in 2018.

What observation would mean it is solved?

Candidates — (a) a final rule takes effect extending catch-documentation requirements to all US seafood imports (b) the audited noncompliance rate falls and stays low across successive annual reports (c) a published estimate of the illegally sourced share of US seafood imports falls from the 2019 level.

(a) alone counts scope, not verification. A second tier that requires only the country of harvest for most of the volume would move coverage to 100 percent on paper while leaving the audit rate where it is. The November 2024 plan proposes exactly that shape, with a smaller data set for the second tier. Full coverage checked at a quarter of one percent is a different object from full coverage checked.

(b) alone can fall for the wrong reason. The audited sample is not known to be representative, the agency says so, and a 2022 statute has required representative sampling without any opened source stating that it was achieved. Noncompliance measured across 329 filings can move because behaviour changed, because selection changed, or because fewer audits were run. Audits fell from 1,131 in FY2020 to 390 in FY2023 while the compliant share also fell, and no opened source separates those causes.

(c) alone has no measurement behind it. The only estimate the US government has of the illegal share is seven years old and refers to 2019, and GAO recorded in May 2023 that no successor had been published. An indicator that is not being measured cannot register either an improvement or a decline. The three have to be read together, and (b) has to be read against the number of audits it came from.

What is it connected to?

Fills with researchforced labor and human trafficking in seafood supply chains, which this program does not collect data on and which the November 2024 action plan proposes to touch indirectly through vessel trip and transshipment data · customs enforcement and country-of-origin rules more broadly · fishery stock assessment and management in exporting states · the import traceability regimes of the European Union and Japan. Relation type and evidence grade were not confirmed in this round.

What these sources do not say

• What the about-half figure on the public program page is a share of. It carries no unit — volume, value, entries or shipments — while the report the same agency sent to Congress gives volume and value separately and both are lower. That page was last updated 2025-05-30, a year after the report published, and nothing opened here reconciles the two.
• Why the import denominator halved between FY2022 and FY2023. The agency table shows 4.8 billion kilograms falling to 2.4 billion, and the report offers only that it continues to evaluate the data and to explore how the pandemic altered supply chains. Coverage is published as a share, so the denominator sets the headline. The FY2022 figure also does not sit with the separate NOAA series reporting 6.3 billion pounds of edible seafood imported in calendar year 2023, and no page opened here states which basis either number is on.
• How the 13 species groups were chosen. The 2016 final rule refers to seven principles. No source opened here publishes the scoring that produced 13 groups rather than 5 or 20, and members of Congress have criticized the failure to disclose the method used for selecting additions, as recorded by the Congressional Research Service.
• When any of it happens. A four-page federal roadmap published 2024-11-14, proposing to extend a program to the whole of a roughly USD 22 billion import stream, contains no deadline, no milestone and no target. Its only temporal language is that actions not requiring regulatory change will be implemented as soon as possible within available resources.
• Whether the statutory audit mandate has been met. A 2022 statute required procedures supporting statistically robust conclusions that audited samples are representative of covered imports. Four years later the agency reports a 0.25 percent selection rate and notes that at least 10 completed audits are needed before a country-level conclusion is significant, a threshold most countries do not reach. No source opened here states either compliance or non-compliance.
• What share of seafood, rather than of filings, was checked. CRS records that the reports to Congress do not give the volume audited per species group or the number of imports per species group, so the relationship between imports and audits cannot be computed from what is published.
• What became of the November 2024 action plan after the April 2025 executive order. No rescission, no reaffirmation and no revised plan appears in anything opened here. CRS raised the question in March 2025 and nothing since answers it.
• How large the illegal share is now. Advocacy and appropriations documents from 2024 through 2026 still cite the same 2019 estimate. GAO recorded in May 2023 that no successor had been published, and none was found on 2026-08-08.

See the evidence

The evidence tier stays secondary because this is a research-based definition with no linked observation record, not because the reading was shallow. Sixteen of the eighteen rows were opened directly and two were not. The spine of this document rests on primary federal material read in full — the report to Congress for FY2023, the raw text of the 2022 proposed rule, of the 2023 withdrawal notice, of Executive Order 14276 and of the 2025 marine mammal notice, the November 2024 action plan, the August 2024 review update, and the GAO report of May 2023. Two Federal Register queries and two Unified Agenda editions were used to establish a negative — that no replacement rule exists — as a documented count rather than an inference. Where sources disagree, the disagreement is left visible rather than resolved, because it is the finding. The share of US seafood imports the program covers is given as 32 percent by volume and 30 percent by value in the report to Congress, as about one-third by CRS reading that same report, as less than 40 percent in the November 2025 appropriations submission, as about 45 percent by agency officials speaking to GAO in May 2023, as nearly half in the November 2023 media release, and as about half on the program page as it stood on 2026-08-08 — and only the first two of those six state what the share is a share of. The import denominator in the report to Congress falls from 4.8 billion kilograms in FY2022 to 2.4 billion in FY2023 without explanation, and the earlier figure does not sit with the separate NOAA series that reports 6.3 billion pounds of edible seafood imported in calendar year 2023, so the two years are not compared here. Second-hand material is marked as such throughout: the USITC estimate reaches this document through GAO and an appropriations submission, and the four bills in block 7, the letters from members of Congress and the criticism of the species selection method are known only as citations inside CRS report R48469, with no bill page or letter opened for this document. This is a Path A output, a research-based definition, so observation_refs is empty and provenance_mode: press-derived.

People affected

Estimated range 340,003,797–340,003,797 As of 2024

Sensitivity The headcount has no width: low equals high at the 2024 US resident population because that is the only sourced denominator, and the value that would narrow it to seafood consumers was unreachable at HTTP 403. The width that the evidence does support sits on exposure, 7.6 to 10.4 pounds of uncovered imported seafood per resident per year, and it comes from the coverage figure disagreeing with itself across agency publications — 32 percent by volume in the report to Congress against about half on the public program page, with about 45 percent, nearly half and less than 40 percent also in circulation, and only the report to Congress stating a unit of measure. What this number fails to count: it counts exposure to an unverified supply, not harm, because uncovered is not the same as illegally sourced, and the only estimate of the illegal fraction is about 11 percent of import value for 2019, a figure about the whole import stream rather than the uncovered part. It cannot separate covered from uncovered at the plate, because customs codes do not map cleanly to species and generic prepared-product codes cannot be attributed at all. It counts no one outside the United States, so fishing crews, coastal communities in exporting states and the stocks themselves are absent, and it counts no displaced US fishermen, for whom the only figure is an aggregate of about USD 60.8 million a year in foregone income. The limit in the opposite direction: the step from imports to consumption assumes the uncovered share of what is eaten equals the uncovered share of what is imported, which re-export churn breaks, and NOAA reports 6.3 billion pounds of edible seafood imported in 2023 against roughly 5.2 billion pounds implied by per-capita consumption, a gap no opened source reconciles. The FY2022 import denominator in the same agency table is 4.8 billion kilograms against 2.4 billion in FY2023, unexplained, so no year-over-year comparison is attempted here.

Regional breakdown No source opened here breaks either seafood consumption or seafood import entry down by state. The FY2023 report to Congress reports entry filings, audits and permit holders at national level and publishes no geographic distribution, and per-capita consumption is published as a single national figure. Splitting the national headcount by state population share is apportionment, which invents a distribution the sources do not contain, so it is not done.