Containerised import value and weight by US port of entry, origin country and commodity, from US customs declarations. Built to test the cargo signal TT-001 infers from draught — the first independent count of mass at the same ports.
Containerisation rate by commodity chapter, imports — COVERAGE INCOMPLETE. The 2010-2023 backfill is still loading, so the archive holds twelve months for 2024 and 2025 but only a partial 2010. Any comparison across years today weighs one loaded month against twelve — an apparent rise from 38% to 63% on import containerisation is currently indistinguishable from that loading artifact. Lift when every month from 2010-01 to the last complete month is present, verified by a per-year month count, not by row totals.
· from 1970-01-01, open-ended
Unit value within commodity line (HS6) — COVERAGE INCOMPLETE. HS6 is fetched on demand, not as a standing archive — it is roughly 12.6x the rows of HS2 and the full history would exceed the free disk. The archive currently holds a single HS6 month for two ports. Lift when the intended window is present for the ports a claim is about, verified by a month count per port.
· from 1970-01-01, open-ended
Containerisation rate, exports, by gateway — COVERAGE INCOMPLETE. The 2010-2023 backfill is still loading, so the archive holds twelve months for 2024 and 2025 but only a partial 2010. Any comparison across years today weighs one loaded month against twelve — an apparent rise from 38% to 63% on import containerisation is currently indistinguishable from that loading artifact. Lift when every month from 2010-01 to the last complete month is present, verified by a per-year month count, not by row totals.
· from 1970-01-01, open-ended
Containerisation rate, imports, by gateway — COVERAGE INCOMPLETE. The 2010-2023 backfill is still loading, so the archive holds twelve months for 2024 and 2025 but only a partial 2010. Any comparison across years today weighs one loaded month against twelve — an apparent rise from 38% to 63% on import containerisation is currently indistinguishable from that loading artifact. Lift when every month from 2010-01 to the last complete month is present, verified by a per-year month count, not by row totals.
· from 1970-01-01, open-ended
Import value density by gateway — LEVEL ONLY, AND THE SERIES CANNOT SAY SO BY ITSELF. Comparable across ports in a given month; its movement over time is a mix index wearing a price label. Suspended until it is published together with the HS6 decomposition that separates the two.
· from 1970-01-01, open-ended
Import vs export value density, by gateway — MOVEMENT NOT YET INTERPRETABLE. The LEVEL is sound — every gateway imports goods worth more per kilo than it exports, and measurement bias is common-mode across the two sides. But the ratio has moved 2.03x (May 2024) to 2.02x (May 2025) to 1.83x (May 2026), and at this aggregation that is indistinguishable from a shift in the commodity mix. Publishing the trend would revive exactly the bias Census abandoned the method for in 1989. Lift once the HS6 within-commodity series can decompose the move into a within-line component and a mix component, and the two are published side by side.
· from 1970-01-01, open-ended
Census restatements observed, by month of data — NO OBSERVATIONS YET. The archive holds 20,618 entries and every one is a `backfill` — our own air_weight_kg column arriving for the first time — with zero genuine restatements. A revision can only be observed by re-fetching a month we already hold and finding it changed, and until 2026-07-30 nothing did that: the weekly timer ran a superseded collector that crashed against the migrated schema. It now re-syncs the last six months every Friday. Lift when the archive holds genuine `revision` or `withdrawn` entries across at least a few distinct months — not on the first one, which could be a one-off.
· from 1970-01-01, open-ended
Days from our first fetch to a restatement — NO OBSERVATIONS YET. The archive holds 20,618 entries and every one is a `backfill` — our own air_weight_kg column arriving for the first time — with zero genuine restatements. A revision can only be observed by re-fetching a month we already hold and finding it changed, and until 2026-07-30 nothing did that: the weekly timer ran a superseded collector that crashed against the migrated schema. It now re-syncs the last six months every Friday. Lift when the archive holds genuine `revision` or `withdrawn` entries across at least a few distinct months — not on the first one, which could be a one-off.
· from 1970-01-01, open-ended
Size of Census restatements, containerised weight — NO OBSERVATIONS YET. The archive holds 20,618 entries and every one is a `backfill` — our own air_weight_kg column arriving for the first time — with zero genuine restatements. A revision can only be observed by re-fetching a month we already hold and finding it changed, and until 2026-07-30 nothing did that: the weekly timer ran a superseded collector that crashed against the migrated schema. It now re-syncs the last six months every Friday. Lift when the archive holds genuine `revision` or `withdrawn` entries across at least a few distinct months — not on the first one, which could be a one-off.
· from 1970-01-01, open-ended
Their pages stay up and their data is not served. A correction is worth publishing.
Source: US Census Bureau, intltrade/imports/porths. The full method — what is measured, what is inferred, and what can mislead — is published as TT-008 method.
describe_series, query_series, check_agreementtrimtab://method/TT-008