The country that exports it usually isn't the one that mines it.

In 18 of 32 materials, the top exporter is not the top miner. In 4 of 32, a country that mines under 5% of world supply nonetheless exports more than a quarter of it — a near-pure refiner or trans-shipment effect.

Read this precisely. The origin gap is a measurement gap between two public datasets — reconciled bilateral trade vs USGS mine production — not a fully traced physical supply chain. Part of any gap is legitimate: a refiner genuinely exports a different product than the mined ore (cobalt chemicals are not cobalt ore). And customs data cannot separate a true refiner from a re-export hub. The claim is that import-origin statistics misidentify where supply is concentrated — not that any single country secretly mined the material.

This is the single number behind the atlas's thesis: the refiner is not the source. An apparent dependence on the country that ships a material is, one layer upstream, a dependence on whoever mined the ore it was made from. Customs rules assign "origin" at the point of last substantial transformation, so the trade record stops at the refinery — and the real geography of supply disappears.

The measure

For each material, in the same year, we compare two public measures: where it is exported from (reconciled bilateral trade, CEPII BACI) and where it is mined (USGS production shares).

A large positive gap means a country sells far more than it digs out of the ground — it is processing or trans-shipping someone else's ore.

Where the gap is largest (2024)

The 14 largest gaps among the 18 materials where exporter ≠ miner. ⛓ gallium, germanium and hafnium share one HS6 code (811292) and cannot be separated in trade.

The twist: the illusion is not only China's

The intuitive story is "China hides behind refineries." The data only half-supports it. For many materials China is both the lead miner and the lead exporter — its chokehold is largely genuine, not an accounting artefact. The materials where exporter and miner diverge are instead fronted by industrial refiners and entrepôts: Finland for Congolese cobalt, Japan for largely Chinese-mined titanium sponge, Germany for Iranian strontium, Norway for Indonesian nickel, the United States for Congolese tantalum.

So the corrective cuts two ways. It deflates apparent dependence on refiner countries — a German strontium "supply" is Iranian rock. And it reveals that a genuinely concentrated upstream — DR Congo cobalt, Indonesian nickel, Chinese rare earths — is more concentrated than the diversified-looking trade ledger suggests, because the ore is laundered through several different refiners before it ships.

What this is — and isn't

• It is a like-for-like comparison of two public measures (CEPII-BACI reconciled trade shares; USGS mine-production shares), per material, every year 2018–2024.
• It isn't a claim that the refiner adds no value, nor that the traded form equals the mined form: some gap is legitimate — Finland really does export refined cobalt chemicals, which sit under a different product than mined cobalt. The gap measures where customs would mislead you about origin, not fraud.

Why it matters

Supply-risk and friend-shoring policy is often written off import-origin tables. This finding says those tables, taken at face value, misidentify the chokepoint in a majority of critical materials — sometimes flattering a refiner, sometimes hiding how concentrated the real mine base is. The fix is cheap: reconcile the bilateral trade, then subtract the mine layer. That is what this atlas does.

This finding across the atlas: follow the ore past the refiner in the origin trace; watch the gap widen over two decades on the trends page; check which rising-concentration trends survive an autocorrelation-robust test under robustness; and see where the refiner sits as a network chokepoint.

Reproduce: findings.py → results/findings.json. Full method and validation in the method note.