Withhold Release Orders vs Findings: US Forced-Labor Enforcement Since 1953, in One Register

Last updated October 23, 2026·Data: CBP Withhold Release Orders & Findings register (69 actions)

Two instruments, one statute family

A Withhold Release Order directs CBP personnel to detain specific merchandise from a specific company, region or production stage at the border: shipments arrive, get held in bonded space, and the importer either proves clean supply-chain origin or loses the goods. A Finding, by contrast, is the formal published determination under 19 USC 1307 that forced labor was used in producing a class of merchandise — it is broader, stickier and historically much rarer: 9 Findings against 60 WROs in the current register. Findings also survive longer; several pre-2000 Findings still appear in the register while most dormant WROs are formally revoked rather than left open.

Status matters operationally. In the register, 63 actions are fully Active, 5 carry a Partially Modified flag (typically narrowed scope after litigation or review) and 1 is Inactive pending litigation. An inactive entry is not a green light — it is an unresolved legal question that can flip back, which is why serious screening programs track status changes rather than snapshots.

Seventy years, three eras

The register's age distribution tells the enforcement story better than any single number. Forced-labor exclusion authority existed since the Tariff Act of 1930, but it was used sparingly — a handful of actions in the 1950s, then a burst of 1990s Findings mostly tied to Mexican and Chinese detention cases:

Source: https://web.archive.org/web/20261002200141id_/https://www.cbp.gov/sites/default/files/2026-09/withhold-release-orders-findings-fy26-2026-09-29.csv (fetched 2026-10-22).

The register this table derives from is our structured mirror of CBP's official CSV; counts are grouped by issue year.

Read it as three eras: a thin legacy layer (1950s–1990s, when Findings were the dominant instrument), a quiet 2000s, and the modern wave that begins around 2019 — driven by the shift toward region-wide presumptions rather than single-factory cases, culminating in the UFLPA regime from June 2022. For a shipper, the practical reading is that the modern entries behave differently: they are commodity- and region-based, they compound with entity-list designations, and they generate far more detentions per action than the legacy Findings ever did.

Geography: concentrated, with a long tail

Source: https://web.archive.org/web/20261002200141id_/https://www.cbp.gov/sites/default/files/2026-09/withhold-release-orders-findings-fy26-2026-09-29.csv (fetched 2026-10-22).

One publishing quirk matters when you read the economy column: "Fishing Vessels" is a flag-of-convenience grouping CBP publishes as its own row rather than a country, so economy-level aggregations should treat it as a sector bucket.

China dominates with 41 of 69 actions (59%) — 36 WROs and 5 Findings. But the tail is instructive: Mexico's four actions are all 1990s-era Findings, Turkmenistan carries a standalone cotton WRO, and the six fishing-vessel entries cover distant-water fleets flagged across several oceans. If your sourcing map stops at China, you are screened for roughly six-in-ten of these actions — and exposed to the rest.

What merchandise gets caught

Source: https://web.archive.org/web/20261002200141id_/https://www.cbp.gov/sites/default/files/2026-09/withhold-release-orders-findings-fy26-2026-09-29.csv (fetched 2026-10-22).

Cluster counts come from keyword classification of the merchandise column; every action matches at most its first rule, so the column sums to the register size.

Cotton and apparel inputs lead — unsurprising given XUAR's role in world cotton — followed by the food-crop lane (tomatoes and derivatives) and the polysilicon/silica chain that feeds solar panels. Seafood deserves special attention because the vessel-based entries make it the only cluster where the vessel itself, not just the product, is the enforcement target. Hair products and wigs look niche until you notice they were among the highest-volume e-commerce categories swept into detentions. Each cluster maps to a different paperwork trail; our compliance checklist walks the evidence file for each.

What a WRO actually does at the border

Abstract registers become concrete money very quickly. When a shipment covered by an active WRO arrives, CBP detains it rather than releasing it into commerce. The importer receives a detention notice and a window to submit a rebuttal evidence package; goods sit in bonded or general-order space while the case is examined, and storage, demurrage and capital costs accumulate on the importer's side of the ledger regardless of the eventual outcome. If the evidence clears the clear-and-convincing bar, the shipment moves; if it does not, the goods face exportation or destruction and the entry is denied. Planning for that scenario before purchase — instead of after a hold notice — is the single cheapest risk control available, because traceability documents assembled at sourcing time cost a fraction of documents reconstructed under detention pressure.

How the register relates to the UFLPA Entity List

Importers often conflate two artifacts that answer different questions. This register — WROs and Findings — attaches to merchandise: a product class from a named company, region or production stage. The UFLPA Entity List, maintained by DHS through Federal Register notices, attaches to entities: once an organization is listed, the statutory presumption applies to its goods and downstream articles even without a dedicated WRO. In practice they compound. A facility named on the entity list can pull otherwise-unlisted commodities into enforcement scope, and a commodity WRO can sweep in suppliers nobody has named individually. Complete screening therefore runs both checks: entities first, then the commodity register for every HS6 lane you ship. Our entity-list browser covers the roster side of that pair.

A screening workflow that matches the register's shape

1. Screen counterparties against the entity list — the register and the UFLPA Entity List overlap but are not identical; run both. Our free entity search covers the roster.
2. Classify early. Category risk is knowable before you order: look up the HTS code in the HS browse and check whether the chapter already appears in enforcement statistics.
3. Model the cost of a stop. A detained container accrues storage and capital costs while evidence is assembled; price that scenario in the landed-cost calculator before booking.
4. Watch for status flips. Partially Modified and Inactive-pending-litigation rows move in both directions; monitoring beats annual reviews.

Methodology and data freshness

All figures derive from our structured mirror of CBP's official "Withhold Release Orders and Findings" CSV (upstream file, retrieved 2026-10-22; 69 rows with per-row source URLs and SHA-256 digests in the sidecar). Decade, economy and product-cluster tables are computed at generation time from those rows, so this page updates itself whenever the pipeline refreshes. Cluster counts use the keyword rules shown above and can double-count multi-product actions only once by construction. Nothing here is legal advice — consult your broker or counsel for binding determinations. For the demand-side counterpart of enforcement, see our August 2026 US import snapshot.