Import duty on 8705.40 in thirteen markets, and the four things a single rate hides

Fourteen economies were queried for HS 8705.40, the subheading for motorised concrete-mixer vehicles, and thirteen of them returned a reported most-favoured-nation duty. In ascending order those readings are Peru 0%, Morocco 2.5%, the Philippines 3%, Saudi Arabia 5%, Nigeria 5%, the United Arab Emirates 5%, Indonesia 5%, Kazakhstan 10%, Russia 15%, Viet Nam 15%, Malaysia 30%, Mexico 35% and Thailand 40%. Four of the thirteen sit at exactly 5%, which in this pull makes Saudi Arabia, Nigeria and the United Arab Emirates indistinguishable from Indonesia on the duty line alone. Seven of the thirteen sit at or below 5% and six at 10% or above, with no reading anywhere between those two points. Thailand at 40% is the highest reading and Peru at 0% the lowest; that comparison was made by ranking these thirteen economies against each other and covers no market outside the fourteen queried.

Every one of those figures is the simple average of the tariff lines the reporting country filed under that six-digit subheading. It is not the duty on a specific eight- or ten-digit national code, and it is not a quotation for any consignment. Nine of the thirteen readings rest on a single reported line, so for those nine the averaging step changes nothing: Peru, the Philippines, Saudi Arabia, Nigeria, the United Arab Emirates, Indonesia, Viet Nam, Malaysia and Thailand. The remaining four each rest on two reported lines — Morocco, Kazakhstan, Russia and Mexico — and for three of those four the two lines carry the same rate, so the average still equals the rate. Mexico is the exception, and it is the reason this page does not present any of these numbers as the duty a buyer will pay.

Mexico is the one 8705.40 reading in this pull where the average conceals a range. Its row records two tariff lines, a reported minimum of 20% and a reported maximum of 50%, which average to 35%. No consignment pays 35%. A consignment pays 20% or it pays 50%, and which of the two applies depends on the eight-digit national code the entry is filed under — a code this dataset does not carry at six-digit resolution and this page therefore cannot supply. Quoting a Mexican buyer a landed cost built on 35% would understate the duty by fifteen percentage points or overstate it by fifteen, with no way to know in advance which of the two errors had been made. On a machine of any real value that is not a rounding difference, and it is not resolvable from this source.

The same trap appears once more in the file, on the other subheading queried. Kazakhstan's reading for 8429.11, self-propelled track-laying bulldozers, averages 1.67% across three reported lines running from a minimum of 0% to a maximum of 5%. Every other reading in the file, on both subheadings, has a reported minimum equal to its reported maximum, which is why the averaging step is invisible for them and consequential only for these two. The operating rule that follows is narrow and worth stating plainly: before any landed-cost figure goes to a buyer, the eight- or ten-digit national code has to come from a customs broker in the destination country, and the duty has to be read against that code rather than against a six-digit average compiled for statistical comparison.

Twelve of the thirteen MFN readings carry the reporting year 2023. Russia's carries 2021, so the 15% attributed to Russia describes a schedule two years older than the twelve figures it is being listed beside, and five years older than the date the data was pulled, 3 September 2026. Listing all thirteen in a single ranked series is a presentational convenience, not a claim that they were measured at one moment. Nor is the newest of them current: a 2023 schedule is the most recent reading available here for any of the thirteen, and tariff schedules are amended between reporting years without that amendment appearing in a dataset built from what countries have filed. Anything time-sensitive has to be checked against the destination's own published schedule on the day it matters.

The preferential rows are further out of step. All four of them carry 2021, two years behind the MFN rows they would naturally be compared against. Nomenclature differs as well: ten of the thirteen MFN rows were filed on the revision this dataset labels H6, while Mexico, Peru and Russia were filed on the earlier revision it labels H5, and all four preferential rows are H5 too. Subheading definitions can be redrawn between nomenclature revisions, so an MFN row filed on one revision and a preferential row filed on another are not guaranteed to describe an identical basket of goods, even when both are labelled 8705.40. Nothing on this page corrects for that mismatch, and no correction should be inferred from the fact that the two numbers appear in the same sentence.

Algeria was queried on exactly the same terms as the other thirteen economies and returned no reported row for the years 2019 to 2023 — not for the MFN rate, not for the preferential rate, and on neither of the two subheadings. All four of its cells come back as a failed lookup rather than as a value. That is a different outcome from a lookup that succeeded and found a rate of zero, and the two are recorded differently in the file for exactly that reason. Algeria therefore has no place in the thirteen-market ranking, no place in the comparison between subheadings, and no duty rate anywhere on this page. Its absence from the list is a fact about the query, not a finding about Algerian customs.

The distinction is worth holding onto because it is easy to lose. A missing row means the reporting country did not file that year's schedule into this dataset; it says nothing whatever about what Algerian customs charges on arrival, which may be high, low or zero. Treating the blank as a zero, or quietly dropping the country so the table looks complete, both produce a document that reads as authoritative and is not. Any Algerian enquiry has to be answered from the Algerian tariff schedule itself, or from a broker working in that market, and a figure sourced from a gap in an international dataset would be an invention that arrives with a citation attached to it. The same caution applies to any market outside the fourteen queried: silence in this pull is not evidence of a low rate, or of any rate.

A preferential rate of 0% on goods of Chinese origin appears under 8705.40 for four of the fourteen economies queried: Indonesia, the Philippines, Thailand and Viet Nam. Each is a single preferential line reported for 2021. Five of the fourteen are ASEAN member states — those four plus Malaysia — so four of the five returned a preferential row and one did not. On the bulldozer subheading 8429.11 the picture narrows further: the same preferential zero is recorded for Indonesia alone, and the Philippines, Thailand and Viet Nam carry no reported preferential row there at all. Each of the five preferential rows in the file — four on 8705.40 and one on 8429.11 — records exactly one preferential line and carries the earlier nomenclature label. Whether the missing rows mean no preference exists on that subheading, or only that none was filed, cannot be settled from this file.

Nine of the remaining economies — Kazakhstan, Malaysia, Mexico, Morocco, Nigeria, Peru, Russia, Saudi Arabia and the United Arab Emirates — returned no preferential row for 2019 to 2023, and Algeria, the tenth, returned nothing at all. That absence describes what was filed into the dataset, not what the destination charges. Malaysia is the case that shows why the difference is expensive: it is an ASEAN member state covered by the ASEAN–China free trade agreement under which Form E certificates are issued, its reported MFN duty under this heading is 30%, and this dataset carries no preferential row for it. Reading that as "Malaysia charges 30% on Chinese-origin machines" would be a conclusion drawn from a hole in a file rather than from a tariff schedule.

Where a preferential rate does apply, it applies only if a Form E certificate of origin is raised and accepted at the border; without an accepted certificate the entry falls back to the MFN rate, which is the whole reason the MFN column still matters in markets that have a recorded zero. Form E is normally raised before or at the time of shipment. If it was not, and not within three days of shipment either, the ACFTA operational certification procedures say it shall still be issued at the exporter's request, within twelve months of the shipment date, marked "ISSUED RETROACTIVELY" in Box 13. So a missed certificate is not automatically a lost preference. How each importing customs authority treats a certificate carrying that mark has not been verified and is not claimed.

The same pull covered a second subheading, 8429.11 for self-propelled track-laying bulldozers, across the same fourteen economies, and comparing the two market by market shows how much of a duty rate is a property of the classification decision rather than of the machine. In seven of the thirteen the bulldozer subheading carries the lower reading: Mexico 0% against an average of 35%, Thailand 0% against 40%, Viet Nam 0% against 15%, the Philippines 0% against 3%, Malaysia 20% against 30%, Russia 5% against 15%, and Kazakhstan an average of 1.67% against 10%. In five the two readings are identical — Peru at 0%, Morocco at 2.5%, Saudi Arabia, Nigeria and the United Arab Emirates at 5%. Indonesia is the only one of the thirteen where the bulldozer subheading is the higher of the two, at 10% against 5%.

Two subheadings are not the whole question for a self-loading concrete mixer, which can also answer to 8474.31, the subheading for concrete and mortar mixers. That subheading was not among those queried in this file, so no rate for it is stated anywhere on this page, and none should be read into the numbers that are. Where the classification of a self-loading mixer is contested between headings — and it is contested — the duty consequence of settling it cannot be worked out from the figures here, because one of the three candidate subheadings is simply missing from the source they come from. That gap is the reason the classification question and the duty question are handled as two separate pieces of work rather than one.

On a declared customs value of USD 30,000 — a round number used here to make the arithmetic legible, not a quotation and not a price — the difference between the reported MFN rate and the recorded 0% preferential rate under 8705.40 comes to USD 12,000 in Thailand, USD 4,500 in Viet Nam, USD 1,500 in Indonesia and USD 900 in the Philippines. That is the duty line and nothing else. Read carefully, it is what an accepted Form E is worth per unit in those four markets, on those four reported rates, at that assumed value. Read carelessly, it is a landed-cost estimate, which it is not. The ranking of the four follows directly from the MFN spread rather than from anything about the destinations themselves.

Three things are deliberately left out of that subtraction, and each of them moves the answer. Value-added tax and local turnover taxes are charged in many jurisdictions on a duty-inclusive base, so a duty saving pulls a tax saving behind it; those rates are not in this dataset and no figure for them appears here. Whether the customs value is struck on a CIF or an FOB basis differs by country and is not recorded either, so the USD 30,000 base is an assumption rather than a defined quantity. And the MFN rates used in the subtraction are 2023 readings while the preferential rates are 2021 readings, so the calculation crosses two reporting years. The four figures are an order of magnitude for a negotiation, not a landed cost.