Last updated 7 November 2026 · every figure recomputed from the sources listed at the foot of this page
Paper Antidumping Duty: 30 Cases, 16 HTS Headings, and a 14.78× Caliber Gap
The anti dumping duty on paper is the cleanest illustration of a trap that runs through the whole antidumping dataset. Ask what the duty is on imported paper and the answer depends on which of two numbers you read: across the 30 paper cases in this corpus the dumping margin's median is 115.29% while the cash-deposit rate's median is 7.8% — a gap of 14.78×, wider than the steel family's. This page prints both, maps the family's unusually narrow HTS footprint, and states plainly where its own sample is too small to support a conclusion.
What the 30 paper cases actually are
The slice is 30 case numbers whose product scope names paper, producing 88 Federal Register order-lifecycle records. 30 of the 30 cases are active; the rest are revoked. The records split 68 antidumping, 10 countervailing-duty and 10 combined AD/CVD, so the stacking question exists here too, just at smaller scale.
What makes paper structurally different is the tariff footprint. The family resolves to 122 eight-digit lines inside only 16 four-digit headings, and the largest single heading accounts for 15 of the 30 cases. Compare that with steel's 30 scattered six-digit candidates: paper is a narrow, deep family where a handful of headings carry almost the whole story, and that is good news for anyone trying to determine whether their product is inside an order.
The caliber gap, and why it is wider here
Of the 88 records, 30 state a single unambiguous percentage (34.1 percent) — a slightly higher share than the steel slice, and for the same reason: most lifecycle documents do not print a rate.
This table is wider than the screen: scroll it sideways to reach every column.
| Rate caliber | Records | Median | Mean | Minimum | Maximum |
|---|---|---|---|---|---|
| Cash-deposit rate | 18 | 7.80% | 15.59% | 2.90% | 92.34% |
| Dumping margin | 12 | 115.29% | 105.64% | 3.91% | 266.83% |
Source: Federal Register API (federalregister.gov), AD/CVD order-lifecycle corpus built by scripts/build_adcvd.py · checked .
The two rows are two different quantities. The cash-deposit rate is what CBP collects at entry. The dumping margin is the finding the determination rests on. Here the medians are 7.8% and 115.29%, a ratio of 14.78× — and note the ranges: cash-deposit rates span 2.9% to 92.34%, margins span 3.91% to 266.83%. The margin column reaches far higher than the deposit column because a margin can be stated for an order as a whole, as a country-wide or all-others figure, while a cash-deposit rate is the instruction attached to a particular exporter.
A practical warning follows from the means rather than the medians. The mean cash-deposit rate is 15.59% against a median of 7.8%, and the mean margin is 105.64% against a median of 115.29%. Both distributions are right-skewed, so the mean is being pulled by a handful of triple-digit determinations. If you summarize this family by its average you will describe a case that does not exist.
The 7 document types, and the 58 records that state no rate
Counting document types is the quickest explanation for why 58 of 88 records carry no percentage: continuations, revocations and combined orders are lifecycle events, not rate statements.
This table is wider than the screen: scroll it sideways to reach every column.
| Document type | Records | Cases | Rated records | Median of rated |
|---|---|---|---|---|
| Administrative review results | 49 | 13 | 22 | 7.80% |
| Sunset review results | 9 | 7 | 4 | 196.88% |
| Antidumping order | 9 | 7 | 3 | 112.64% |
| Amended final determination | 4 | 4 | 0 | — |
| Countervailing duty order | 3 | 3 | 1 | 10.00% |
| Combined AD/CVD order | 6 | 6 | 0 | — |
| Continuation | 8 | 5 | 0 | — |
Source: Federal Register API (federalregister.gov), AD/CVD order-lifecycle corpus built by scripts/build_adcvd.py · checked .
Two rows deserve attention. Sunset review results show a median of 196.88% across 3 cases against 7.8% for administrative review results across 8 — an even wider apparent spread than steel's. And the antidumping-order row, which records the original determinations, shows 196.88% too, because in this slice the original order is the document that carried the high margin in the first place.
Where this page's sample runs out
The steel page in this series can price the composition effect, because 10 steel cases carry both review types and the paired ratio can be computed on them. Paper cannot, and saying so is more useful than smoothing over it.
This table is wider than the screen: scroll it sideways to reach every column.
| Case group | Cases | Median latest rate |
|---|---|---|
| Carries a sunset-review rate | 3 | 196.88% |
| Carries an administrative-review rate | 8 | 7.80% |
| Carries only a sunset-review rate | 2 | 196.88% |
| Carries only an administrative-review rate | 7 | 7.80% |
| Carries both (the paired sample) | 1 | paired ratio median 1.000x |
Source: Federal Register API (federalregister.gov), AD/CVD order-lifecycle corpus built by scripts/build_adcvd.py · checked .
The paired sample here is 1 case. One case cannot distinguish a composition effect from a rule, and a single paired ratio of 1.0× is not evidence of anything. So this page makes no claim that sunset reviews are or are not harsher than administrative reviews in the paper family. What the table does show is the same structural warning as steel's: the cases carrying a sunset rate (3) and the cases carrying an administrative rate (8) are largely different cases, with only 1 in common, so any comparison between those two medians is a comparison between two case mixes rather than between two procedures.
That is a reusable rule for reading any AD/CVD table: before comparing two group medians, check how many units appear in both groups. If the overlap is small, the comparison describes the groups, not the groups' labels.
Where the rates concentrate
This table is wider than the screen: scroll it sideways to reach every column.
| Rate band | Cash-deposit records | Margin records |
|---|---|---|
| 0% | 0 | 0 |
| 0.01-9.99% | 13 | 2 |
| 10-24.99% | 0 | 2 |
| 25-49.99% | 4 | 0 |
| 50-99.99% | 1 | 0 |
| 100% and above | 0 | 8 |
Source: Federal Register API (federalregister.gov), AD/CVD order-lifecycle corpus built by scripts/build_adcvd.py · checked .
The same bottom-heavy deposit column and top-heavy margin column as steel, with thinner cells. Several bands hold only a few records, which is the sample size talking rather than a finding — a band with two records in it is not a mode.
The HTS footprint: 16 headings carry the family
This is the part of the page an importer can act on. Because the family sits in 16 four-digit headings, checking whether a paper product is covered is a bounded search rather than an open one.
This table is wider than the screen: scroll it sideways to reach every column.
| HTS heading | Cases naming a line in it |
|---|---|
| 4811 | 15 |
| 4802 | 9 |
| 4820 | 8 |
| 4819 | 6 |
| 4810 | 5 |
| 4823 | 5 |
| 3703 | 3 |
| 4804 | 2 |
| 4806 | 2 |
| 4808 | 2 |
| 3919 | 1 |
| 4805 | 1 |
| 4818 | 1 |
| 4911 | 1 |
| 8305 | 1 |
| 9505 | 1 |
Source: AD/CVD case register built by scripts/build_trade_remedies.py · checked .
The concentration is the point. Heading 4811 alone appears in 15 of the 30 cases, and the top few headings account for the large majority of the family. If your product classifies in one of the headings near the top of that table, the probability that some order touches your line is high and the scope language is worth reading carefully. If it classifies in a heading near the bottom, or in a chapter absent from the table altogether, the family is unlikely to reach you — which is a different and equally useful answer.
Two cautions. The column counts cases that name at least one line inside the heading, so a heading with a high count is one that many orders touch rather than one that carries many lines. And 1 of the 30 cases carry no HTS linkage at all, so the footprint is built from the 29 that do.
Who the cases run against
This table is wider than the screen: scroll it sideways to reach every column.
| Origin | Rated records | Median of rated |
|---|---|---|
| China | 8 | 115.29% |
| Portugal | 6 | 7.80% |
| India | 5 | 3.91% |
| Brazil | 4 | 27.11% |
| Korea | 3 | 6.19% |
| Cambodia | 1 | 10.00% |
| Federal Republic of Germany | 1 | 2.90% |
| Sri Lanka | 1 | 10.00% |
| Vietnam | 1 | 92.34% |
Source: Federal Register API (federalregister.gov), AD/CVD order-lifecycle corpus built by scripts/build_adcvd.py · checked .
9 origin countries appear in the rated paper records. As with steel, a case can name several countries and a country's median is a median over records rather than the rate in force for that country today.
How to check your own paper product
One. Classify to an eight-digit line. The family's narrowness means this step is unusually informative here: 122 lines cover the whole slice.
Two. Take the four-digit heading and look it up in the footprint table above. If the heading is absent, the antidumping family is unlikely to apply and the question becomes the base tariff rate.
Three. Read the product scope of each case naming a line in that heading. Scope language describes products; it can include or exclude your article regardless of the code, and it is the operative text.
Four. Identify which quantity the current document states, then check whether a countervailing duty applies alongside — 10 of the 88 records here are combined AD/CVD documents.
What this page cannot tell you
The corpus is a snapshot dated 2026-10-29. The 58 records that state no percentage are documents that print no rate, not zeroes. The paired sample is 1 case, so the review-type comparison is explicitly not made. And no figure here is a determination for a specific shipment: that needs the classification, the case, the exporter and the entry date.
Frequently asked questions
- What is the anti dumping duty on paper from China?
- There is no single figure. 30 case numbers in this corpus have a product scope naming paper, spanning 9 origin countries and 16 HTS headings. A duty attaches to a case and an exporter. Across all rated paper records the cash-deposit median is 7.8% and the margin median is 115.29%.
- Why is the dumping margin so much higher than the cash-deposit rate?
- They measure different things. The margin is the finding of how far the US price fell below normal value; the cash-deposit rate is what CBP collects at entry against the final liability. In the paper slice the medians are 115.29% and 7.8%, a ratio of 14.78×.
- How many HTS headings does the paper antidumping family cover?
- 16 four-digit headings, containing 122 eight-digit lines across the 29 cases that carry an HTS linkage. Heading 4811 is the most frequently named, appearing in 15 cases.
- Is a sunset review rate higher than an administrative review rate for paper?
- This page does not claim so. The medians differ sharply (196.88% against 7.8%), but only 1 paper case carries both review types, so the paired sample is 1 case and cannot separate a composition effect from a rule. The steel page in this series has enough overlapping cases to make that comparison; paper does not.
- What is the cash-deposit rate for paper?
- The median across the 18 paper records stating a cash-deposit rate is 7.8%, mean 15.59%, range 2.9% to 92.34%. That is a distribution across cases and exporters, not a rate applicable to a shipment.
- Does a countervailing duty apply on top of the antidumping duty on paper?
- Sometimes: 10 of the 88 paper records are combined AD/CVD documents. Where both apply, the countervailing duty is a separate percentage on the same entered value, so the combined burden is the sum plus the base rate.
Sources
- Federal Register API (federalregister.gov), AD/CVD order-lifecycle corpus built by scripts/build_adcvd.py — data/adcvd-orders.json -- 2,756 records, one per Federal Register order-lifecycle document x attributable case number (2,972 documents kept of 4,615 listed) (read 2026-10-29T08:46:20Z, sha256 7fc488813c2a2e49…)
- AD/CVD case register built by scripts/build_trade_remedies.py — data/trade-remedies.json -- 555 case records with product scope text and HTS linkage expanded against the HTS vintage of 2026-10-19 (read 2026-10-29T08:46:19Z, sha256 1786841611e226e1…)
Every figure above is recomputed from the datasets listed here at render time; the dataset vintage and each source's check timestamp are printed with it. This page extends the prototype that already earns search traffic on the same mechanism. Rates are informational, not a binding customs determination: confirm the classification and the entry with CBP before shipping.