Self-contained, dated, citable write-ups of the atlas’s core findings — each with its method, its caveats, and a downloadable version. Built on public data; meant to be checked.
China’s October-2025 export-control wave — expanded rare earths + the battery supply chain — was suspended for one year, to November 2026. A scheduled snap-back, not a resolution. Three years of controls land exactly where the mechanism map says the leverage is: by-product and built-capability chokepoints, never diffuse chains.
A falsification test survived: the atlas’s free, public-data supply-risk index independently recovers the IEA’s proprietary-data high-risk list — four of six discrete minerals in the top third. Convergence on the specifics; the only difference is a wider aperture.
Concentration is not static — it moves, and the direction tracks the mechanism. Built-capability chokepoints erode (rare-earth refining 84→73% by 2035), policy and thermodynamic ones entrench. Tested against the IEA GCMO 2026 projections for six minerals.
Knowing China refines what it doesn’t mine is only half an answer. This report turns the critical-material chokepoints into decisions: which kind of moat holds each one, who could realistically break it, whether the world could absorb a cut — and the counterintuitive result that for 8 of 31 materials, removing the leading exporter makes concentration worse.
In 17 of 32 critical materials the top exporter is not the top miner; in several, a country that mines under 5% of world supply exports more than a quarter of it. Import-origin statistics systematically misidentify where supply is concentrated — a measurement gap between reconciled trade and USGS mine production, stated precisely.
We tested whether economic-complexity density forecasts which countries acquire downstream capability. It looked strong (AUC ~0.86) — until the controls: density is 95–98% a country’s overall diversity, placebos like whisky score the same, and the residual collapses toward chance. A negative result, published because the discipline is the point.
Across 58 end-use chains, the bottleneck is rarely the mine — only 4 chokepoints are geological; the other 54 were made. This report sorts them by the mechanism that concentrates each binding stage — thermodynamic, by-product, built capability, geological, governance, policy, or none — so the concentration is predicted from the process, and each mechanism names its own fix.
Permanently archived on Zenodo — concept DOI 10.5281/zenodo.21948855 (always resolves to the latest version). Each release also carries its own version DOI.