magnitude · pre 2024-06..2025-03 vs post 2025-06..2026-03, 15.7M HS6 records · verdict: CONFIRMED — MECHANISM OPEN · VIETNAM REROUTING TESTED, RULED MINOR
Registered 2026-08-07 after an exploratory HS6 sweep, before confirmation: the customs value per kilogram of Chinese goods declines materially after the April 2025 tariff wave WITHIN identical HS6 codes — not as composition — by more than double the within-good decline of India and Vietnam.
Confirmed everywhere we looked. Across 136 matched HS6 goods, the declared value of
Chinese cargo fell 23% per kilogram between the year before Liberation Day and the year
after — the same physical goods, worth a quarter less on paper. India (−9.5%) and
Vietnam (−7.5%) show only the mild general deflation of the period; China’s decline
is three times deeper. It replicates on three slices not used to find it: West Coast only
−22.0%, East & Gulf only −22.1%, coarse HS2 grain −24.1%. The
steepest drops sit exactly where tariffs bite hardest: ornaments −61%, upholstered seats
−52%, plastic tableware −45%, wooden furniture −34%. And the same-good gap to
Vietnam is now extraordinary: wooden furniture clears US customs at $0.65/kg from China and
$2.36/kg from Vietnam — same HS6 code, 3.6× the declared value.
What we do not claim: the mechanism. Four candidates fit the data and this study
does not pick one: exporters discounting to hold American buyers; customs undervaluation
(at 50%+ ad valorem, every declared dollar shaved halves the duty — and the incentive
arrived exactly when the decline did); quiet downgrading to cheaper variants inside each code; or
origin-washing — if the expensive goods inside a code now arrive relabeled through
third countries while cheap ones still ship direct, China’s within-code value falls with no
price change at all. Amendment 2026-08-07: the fourth candidate was added after publication; the
original three-candidate text stands above per house rules. The observation is published; the
mechanism is an open case, like F-001’s root cause.
The best alternative, tested (2026-08-07). The strongest objection — that the
collapse is Chinese cargo hiding behind Vietnamese labels rather than falling prices — was put
to the data the day it was raised. If relabeling were the story, Vietnam’s shipments should
mirror China’s losses code-by-code at rerouting speed, with Vietnam’s own declared values
diluting toward China’s. Scanning every code where China lost volume: only two codes mirror
at all, and their matched gains cover under 5% of China’s decline. One shows the full
washing signature and is flagged — metal-frame seats (9401.79): China −216 kt, Vietnam
+83 kt, Vietnam’s unit value diluting $3.70→$3.33/kg. The other — monitors
(8528.52) — mirrors in volume but Vietnam’s value ROSE, the signature of genuine
production shift. Scope of the test, stated: it covers the Vietnam channel at code level; diffuse
washing or routing via third countries is untested and the fourth candidate stays open at reduced
weight. The invoice collapse survives its best alternative explanation.
Largest within-good declines, matched Chinese HS6 goods above 200 kt, year before vs year after Liberation Day.
Matched non-tariffed origins declining within 5pp of China would kill the China-specific claim; CBP quantity-based unit values (units, not kg) showing no decline would expose a density artifact.
This study is a frozen artifact: the ruling and its exhibits do not change when new data lands. If the verdict is ever revised, the revision is printed here with its reason, and the original stays on the record.
← all studies · sources: TT-008 port ledger · method & rules