Supply Chain Action Points
Read this first — the conclusion, and the moves to make:
- Within five working days of the executing text, record the definition of item, the operative event for consignments spanning July 1, and the base of the 2 euro fee; each entry must cite a clause, and any of the three still unresolved goes on the open legal list at the next review.
- Restate landed cost by unit-value band rather than as a sector average, with band edges set where the flat charge crosses 5% and 10% of value, and keep that table as the working document until the text is issued.
- Hold the 2 euro fee in a separate contingency line with no operative date attached, and keep it out of any published price until the instrument carrying it exists.
- For the first ten working days after July 1, log every consignment spanning the date with the rule applied and the basis for it; the open count should be zero before the first month closes.
The European Union abolished its €150 customs-duty exemption for low-value imports on July 1, replacing it with a flat €3 per-item duty and a further €2 handling fee expected from November. The change is cooling small-parcel e-commerce air cargo, with Hong Kong–Europe tonnages down about 20% over three weeks per WorldACD.
The Analysis
On July 1 the European Union abolished its 150 euro customs-duty exemption for low-value imports, replacing it with a flat 3 euro duty per item and a further 2 euro handling fee expected from November. Hong Kong to Europe air tonnage fell about 20% over three weeks, according to WorldACD. The headline reads as a tax on cheap parcels. The instrument is narrower and sharper than that.
Two things changed at once and only one of them is about money. The duty base moved from the value of a consignment to the number of items in it. Everything that follows, including the shape of the air cargo response, comes from that substitution, not the size of the charge.
Start with what was abolished. The exemption removed was a customs-duty exemption at a value threshold. It was not the treatment of import charges generally, and the briefing speaks only to duty. Anyone who reads it as the end of relief as such is over-scoped, and will rebuild a model around an instrument that did not touch what they rebuilt. The two treatments have different bases and different recovery routes, and a landed-cost line that merges them cannot be pulled apart later.
The sharper change is the base. A flat duty per item means the assessment no longer scales with value. On a 5 euro item, 3 euros is 60% of the goods value. On a 140 euro item it is a little over 2%. The measure is therefore regressive by unit value: heaviest where the parcel mix is cheapest and most numerous, the segment that dominates item counts on the channel. Any average impact quoted for this measure is close to meaningless: the same charge is a fifth of the price on one item and a fiftieth on another. The unit of analysis is the item, not the shipment and not the sector.
The ambiguity that decides the most money is what counts as an item. The briefing says per item and does not resolve whether a consignment of four units is assessed four times or once. On one reading the charge multiplies with quantity; on the other it is a fixed cost per parcel that rewards consolidation. The two differ by a factor equal to the number of units, which for small-parcel e-commerce is routinely in the single digits. Obtain the executing text and register the exposure item by item. Until then a landed cost built on this measure is a range, and the width of the range is the item count.
The 2 euro fee needs its own line because of its legal status. The briefing describes it as expected from November. Expected is not enacted. There is no operative date in that sentence, and a fee without one is a contingency, not a cost. Register it separately with the status visible, and keep it out of a price list as though it were law. The same question applies to its base, per item or per parcel. If per item, it compounds the regressivity of the duty; if per parcel, it works the other way and rewards the consolidated shipment. One clause decides which.
Where the exposure sits, by operator. Sellers whose mix is high-count and low-value take the full weight immediately, at the first entry after July 1. Sellers in the 60 to 150 euro band absorb a few percent and will mostly not notice until margin review. Importers moving consolidated air freight on commercial entries are outside the duty change on its face, though they share the capacity consequences. Anyone already holding stock inside the union is untouched at import, because the charge attaches at the border and their goods have already crossed it. That last group's position changes without any action on their part, and coverage misses them.
The clocks. The duty is live from July 1. The fee has no date, only an expectation. The air cargo response is already visible three weeks in, which is faster than purchase behaviour can move. That timing is itself the evidence: a 20% tonnage drop inside three weeks is not demand destruction from a 3 euro charge. It is a channel repricing, with integrators adjusting what they carry and at what price, and shippers holding volume while they wait for the text. The number is real. The attribution in the headline is not in the number.
What the coverage missed is the interaction between the base change and the tonnage figure. Because the charge is per item, the parcels that stop first are those with the worst ratio of items to value: light, cheap, multi-unit orders. Those are also the parcels with the lowest value per kilo, which is my assumption rather than a reported figure, though it holds for most of the low-value mix. Removing them takes weight out of the bellies faster than it takes value out of the trade. A 20% drop in tonnage is therefore consistent with a much smaller drop in value, and almost no drop in the orders that matter commercially. Reading the 20% as e-commerce collapsing overstates it by an unknown factor. Reading it as noise understates it. The channel's cheapest traffic has been repriced out, and the tonnage line shows it first.
Put figures on it, and take them as mine rather than the article's. A four-unit order at 6 euros each is a 24 euro consignment. Under the exemption, duty was nil. Now the duty is 12 euros, half the goods value, and with a 2 euro fee assessed per parcel it becomes 14 euros, or 58%. If the fee is assessed per item, the total is 20 euros, or 83%. Same goods, same order, a spread of 25 percentage points, decided by one unresolved clause. Compare a single 140 euro item: 3 euros is 2.1%, a rounding error. That is the regressivity in one line, and it is why the sector average tells you nothing.
Hold the other branch. Had the Union kept the value threshold and added a percentage charge, the burden would have scaled with price, the cheap end would have been spared, and the tonnage response would have been smaller. It chose a flat charge on an item base, which concentrates the burden at the bottom of the price range. That is a distributive choice, and the market reacts to the shape, not the level.
The pit is the consignment declared before July 1 and released after it. Which rule governs depends on the operative event the instrument names, and member state practice is not uniform. That file will be the first dispute and the one most likely to be settled quietly, which is why it should be documented.
All of the above assumes the charge is assessed on the item as described. If the text defines the unit differently, the percentages move, and move most for the smallest orders.
↑ Back to the key points