How it’s computed
For each material and year I compute the export Herfindahl and China’s export share in value (USD) and in volume (metric tons, from BACI quantity, ~94% coverage), 2002–2024. When value-HHI sits above volume-HHI — or value-China-share above volume-China-share — a high-priced producer or material is weighing more in dollars than in tonnes.
Caveat: ~6% of flows lack reported quantity (the volume series is computed over reported-tonnage flows); BACI quantities are partly estimated. Computed by build_vq.py.
Implied price = export value ÷ tonnage (USD/tonne), derived from the trade data itself. China vs the rest of the world: where China's $/tonne sits above the rest, it is exporting the higher-value processed form rather than ore — which is precisely why its value-share can exceed its volume-share. (Unit values from trade data are noisy — quality/product mix varies — so read trends, not the exact level.)
Where value overstates concentration (the price effect)
Materials ranked by how much more concentrated they look in value than in volume in the latest year (HHI gap, ×100, and China-share gap). A large positive gap means the value figure is partly price.
| Material | value HHI | volume HHI | price gap (×100) | China value% | China volume% |
|---|
Computed from BACI HS02 + HS17 (value & quantity) → volume.json. Volume = metric tons; value = USD. The dotted line in the chart marks the 2017 HS-vintage join.