How demand-by-bloc is read (and why imports are only a proxy)
For each material we aggregate import value by destination (the “to” side of every trade flow, ) into blocs, and read the share each bloc pulls. Overlaid: the industrial-policy driver and each bloc’s import reliance.
Big caveat — two flaws, not one. Imports are not final consumption. (1) Re-export of the same metal: China imports ores it refines and re-exports; Hong Kong, the Netherlands and Singapore trans-ship. The net-demand page fixes this by subtracting exports. (2) Demand embodied in finished goods — gallium consumed inside an imported chip never appears as a gallium import at all — and for critical metals this is the larger channel. Netting does not touch it. Two routes at it, both tested and both limited. A material-footprint model (Raw Material Equivalents via a multi-region input-output table) captures all tiers but is sector-coarse — it sees “electronics”, not gallium. A bottom-up estimate (product trade × metal intensity) sounds per-metal, and we prototyped it for cobalt on battery trade — but stress-testing showed it answers a trade-stage question, not a demand one: it reveals that intermediate cobalt is China-bound while finished cells are net-imported by the US/EU, yet calling that “cobalt demand” is a category error (it erases China’s own large use, mixes product-weight with contained metal, and misses cobalt riding in imported cars). So even cobalt resists a per-country final-demand figure. A bloc’s share here is revealed trade pull — a stage fact, not consumption. Inputs: flows × demand.json → bloc_demand.json.
Who pulls each squeezed metal
Each bloc’s exposure — and its policy answer
| Bloc | import pull, key metals | most-pulled squeezed metals |
|---|
What this reconnects, and opens
The supply-structure arm said which metals can’t scale; the demand arm said how fast the pull is growing; this says who is pulling — and the three together turn a materials list into a map of industrial-policy contests. The honest limit is the proxy: import pull conflates processing with consumption, which is precisely why China tops the demand table for metals it also refines for the world. The clean next child is an apparent-consumption estimate (production + imports − exports) per bloc, and a coupling of these demand shares to each bloc’s stated 2030 policy targets — does the pull match the promise?