Critical Materials Atlas
Report 06 · August 2026 · Validation

Does a public-data risk index recover the IEA’s?

Critical Materials Atlas · independent public-data research · a falsification test of the atlas’s public-data method against the IEA Global Critical Minerals Outlook 2026
Abstract. A supply-risk score is only as trustworthy as the data under it, and the atlas is built entirely from public data — USGS/BGS production, World Bank governance, EU-CRM substitution and recycling. The IEA’s Global Critical Minerals Outlook 2026 reaches its own highest-risk list from proprietary market data the atlas cannot see. If the two disagree, the public-data approach is suspect; if they converge, it is validated. They converge: of the IEA’s six discrete high-risk minerals, four rank in the atlas’s top third and all six in the upper half — two independent methods landing on the same by-product and built-capability metals. The one honest difference is scope, not contradiction: the atlas’s wider 32-material aperture and governance weighting also elevate minerals the IEA’s energy-transition lens does not foreground.

Key findings

1 · Why this is a test worth running

The atlas’s entire claim is that you can see the critical-materials risk map without proprietary data — from production statistics, governance indicators and substitution factors that anyone can download. That claim is falsifiable. The IEA, with access to asset-level market intelligence (S&P Global, Wood Mackenzie) the atlas will never have, publishes its own assessment of which minerals are highest-risk. If a public-data index cannot recover that assessment, the public-data thesis fails. So we run the index blind and check whether the IEA’s named minerals rise to the top on their own.

2 · The result: the atlas ranking, IEA high-risk minerals highlighted

The atlas’s supply-risk index ranks 32 EU-CRM materials by a composite of production concentration (HHI), governance of the producing countries (World Bank WGI), substitutability and recyclability. The IEA’s GCMO 2026 risk-assessment section names eight minerals as highest-risk — gallium, magnet rare earths, yttrium, graphite, tungsten, germanium, tellurium, cobalt — of which six exist as discrete atlas materials (yttrium and tellurium sit inside the rare-earth basket and copper by-products). Below, the atlas’s top 14, with the IEA-named minerals marked.

Four of the six discrete IEA high-risk minerals rank in the atlas’s top third (ranks 1–11 of 32); all six in the upper half. gallium (#5), magnets (#8), graphite (#9) and germanium (#11) cluster near the top; tungsten (#13) and cobalt (#17) follow. Two methods, two data sources, one signal.

3 · What the convergence means — and what it doesn’t

The agreement is specific, not generic. Both methods independently elevate the same by-product metals — gallium and germanium, recovered as trace outputs of aluminium and zinc refining, structurally impossible to scale on their own — and the same built-capability chokepoints, where the concentration is in a hard-to-replicate processing step (rare-earth separation for magnets, battery-grade spheroidisation for graphite, APT/carbide for tungsten). That is the atlas’s own mechanism taxonomy (Report 04) recovering the IEA’s risk list from first principles.

The one real difference is honest and worth stating plainly: the atlas ranks a broader 32-material set and weights governance and recyclability, so it also puts beryllium, niobium, lithium and magnesium high — materials the IEA’s energy-transition-demand lens does not foreground. This is scope, not contradiction: a wider aperture and a different weighting, converging with the IEA exactly where the two overlap.

A public-data method that recovers a paid-data agency’s risk ranking is validated for what both cover. The convergence does not prove the atlas’s exact scores are right — the IEA’s weights are unpublished, so this is a test of rank agreement, not identical numbers. But it does show the free, reproducible approach is not missing the signal. Where the atlas ranks minerals the IEA does not foreground, that is a wider lens to interrogate, not an error to discount.

Method

Atlas index: the site’s live supply-risk model (risk.html) — production shares from USGS/BGS → Herfindahl concentration, governance via World Bank Worldwide Governance Indicators, substitutability and recycling factors from EU CRM 2023; see risk methods for the exact composite. IEA high-risk list: GCMO 2026 risk-assessment section (supply concentration + alternative-supply routes + strategic importance), © IEA, CC BY 4.0. The IEA’s weights are proprietary, so this tests rank agreement, not identical scores. “Top third” = ranks 1–11 of the 32-material set. Two of the IEA’s eight named minerals (yttrium, tellurium) have no discrete atlas material and are excluded from the six-of-eight count. The live, self-updating version is at iea-risk-check.html.

How to cite. Critical Materials Atlas (2026). Does a public-data risk index recover the IEA’s? Report 06. Zenodo. https://doi.org/10.5281/zenodo.21948855 @techreport{cma_report06_2026, title = {Does a public-data risk index recover the IEA's?}, author = {{Critical Materials Atlas}}, institution = {Critical Materials Atlas}, number = {Report 06}, year = {2026}, publisher = {Zenodo}, doi = {10.5281/zenodo.21948855}, url = {https://criticalmaterialsatlas.org/report-iea-crosscheck.html} }
Archived on Zenodo — concept DOI 10.5281/zenodo.21948855 (always resolves to the latest version). Cross-check against the IEA Global Critical Minerals Outlook 2026 (CC BY 4.0).