Charted: East Asia Holds 33% of Industrial-Policy Stock — North America Owns 71% of Package Dollars

Our H-NIPO research ledger mapped the stock and toolkit of modern industrial policy. The concentration companion then asked how thick the top of that distribution is — top-1 near 24%, Big Three near 63% on counts, and roughly 93% of major package dollars in three capitals. This post answers a different desk question: where does activity, risk, and fiscal capacity sit on the map?

The interactive dashboard above is built as a geography lens. Toggle Regions, Count × $, US states, and Sectors. Filter regions by stock counts, package dollars, or June flow; filter sectors by semis, clean energy, EV/battery, dual-use, or minerals. The punchline is deliberately multi-map. On cumulative counts, East Asia leads at about 33% and the top-three regions (East Asia + North America + Europe) clear roughly 72%. On major fiscal packages, North America alone is about 71% of the war-chest universe we track. On US subnational awards, Arizona, Texas, and New York together hold about 58% of a ~$39B tracked CHIPS megaproject tip. And on June 2026 monthly flow, rest-of-world still prints 62% of the tape.

The headline map: four meters, four geographies

Read the table as a family of maps, not one slogan. Count geography tells you who generates intervention frequency. Package geography tells you who can put nine- and ten-figure claims on factory sites. Flow geography tells you that a busy month can still be RoW-heavy even when the multi-year stock is Big-Three thick. Subnational geography tells you that even inside the US dollar tip, awards cluster into a handful of fab corridors. Averaging these rows into a single “industrial policy is concentrated in X” sentence is a category error.

Regional stock: East Asia leads the count map

Filter Regions to Stock counts. East Asia — China plus estimated Japan and Korea shares — sits near 33% of the cumulative industrial-policy stock. North America (US-led) follows near 21%, Europe (EU + member states as one bloc) near 18%, and rest-of-world still holds about 28%. That ordering is the geographic twin of the concentration ladder: China remains the largest single jurisdiction; rolling Asia’s next rung on top of China makes East Asia the thickest region.

The stock pie is therefore Asia-weighted without being Asia-monopoly. A system where East Asia holds one-third of recorded interventions still leaves North America and Europe with nearly two-fifths combined, and a long RoW tail that can start subsidy races, copy HS6 lists, or free-ride on templates written in Beijing, Washington, and Brussels. IMF’s H-NIPO check (China+EU+US ≈ 53% of the 2009–2023 stock) and the 2023 NIPO census (same trio ≈ 48% of 2,580 measures) sit one notch below the Teneo cumulative Big Three framing — different windows, same qualitative map: the tip is regional, not planetary.

Package dollars: North America owns the war chest

Switch the region metric to Package $. The map flips. North America jumps to about 71% of the major CHIPS/IRA/EU/CN/JP/KR war-chest universe (~$627B statutory / mobilisation / tax-credit tip). Europe and East Asia together hold most of the remainder; rest-of-world barely registers on this particular dollar pie because the tracked packages are the large, disclosed industrial-policy war chests — not every SME voucher on earth.

That asymmetry is the core geography insight for capital allocation. Counts ask who intervenes how often. Package dollars ask who can relocate a fab, battery plant, or advanced-packaging line with a fiscal claim large enough to move board decisions. A jurisdiction that records many firm-level awards can look “active” on the count map while still lacking US-scale tax-credit and appropriation capacity. Conversely, a single IRA-sized tax-expenditure envelope can dominate a dollar pie without matching China’s historical intervention count share. Desks that only watch NIPO counts will underweight North American fiscal capacity; desks that only watch IRA headlines will underweight East Asian intervention frequency.

Count × dollar scatter: two powers, two axes

Open Count × $. The scatter places jurisdictions on stock-count share (x) versus package-dollar share (y). The United States sits mid-count (~21%) and extreme-dollar (~71%). China sits high-count (~24%) and mid-dollar (~8% on the Big Fund III tip we score). The EU sits mid on both axes. Japan and Korea are thin on counts and small-but-nonzero on packages. Rest-of-world is the high-count / zero-package tip on this war-chest construction — politically alive on the tape, fiscally invisible in the mega-package pie.

Bubble size tracks package dollars, so the US circle dominates the visual field even when its count share is not first. That is intentional. Industrial-policy geography is not a single choropleth; it is at least two overlays — frequency and fiscal firepower — that disagree about which region is “on top.”

US subnational: AZ · TX · NY as the fab corridor tip

Toggle US states. Inside the North American dollar tip, awards still cluster. On a tracked ~$39.2B CHIPS megaproject tip built from public facility notices, Arizona (~$8.5B, ~22%), Texas (~$8.0B, ~20%), and New York (~$6.2B, ~16%) together clear about 58%. Ohio and Oregon add another ~18%; the residual “other tracked states” bucket holds the remaining quarter.

This is facility geography, not a full Commerce outlay census, and it is not IRA clean-energy tax-credit claims by ZIP code. It still answers a practical location question: when Washington writes a semiconductor industrial-policy cheque, the first-order map is not “the United States” in the abstract — it is a short list of metro corridors with fab campuses, packaging plans, and supplier ecosystems. Supply-chain desks that only model “US onshoring” as a national switch will miss the state-level concentration that determines labour markets, power interconnect queues, and water/permitting risk.

Sector maps: semis lean Asia; clean energy leans North America

Open Sectors. Estimated regional intensity by strategic theme shows why a single world map fails. Semiconductors and advanced packaging remain East Asia–heavy (~41% of that sector’s IP geography in our allocation), with North America close behind (~32%) on the back of CHIPS awards. Clean energy and grid flip the other way: North America near 48%, Europe 24%, East Asia 22% — the IRA tax-credit machine showing up as geography. EV/battery and dual-use sit more balanced across the three industrial-policy poles. Critical minerals processing still puts East Asia near 44%, with a thicker RoW tail (~22%) reflecting mining and midstream jurisdictions outside the Big Three.

Filter the sector control to Semis or Clean energy to isolate those two maps. The policy implication is blunt: tariffs, subsidies, and local-content rules do not redistribute a single homogeneous “industrial capacity” — they push different sector maps in different directions. A semiconductor export-control stack that bites East Asia is not the same geography as an IRA domestic-content rule that pulls clean-energy capex into North American corridors.

Flow check: June still belongs to the rest of the world

Switch Regions to June flow. The monthly Roundup geography is the humility check. United States 20%, EU+MS 12%, China 6%, rest-of-world 62% of 823 interventions. Top-3 blocs together are only 38%. That does not falsify the East Asia stock lead or the North America package lead. It warns that stock concentration is not flow concentration: in a single busy month, capitals outside the Big Three can still dominate the raw count tape even when multi-year ledgers and statutory war chests look oligopolistic.

Our August 202608 toolkit update already flagged the same asymmetry in monthly data. Geography makes it visual: the RoW slice on the June pie is larger than any single industrial-policy pole. Narrative that “only three capitals matter” fails the monthly map even when it roughly fits the cumulative stock and package-dollar maps.

Caveats and what this map is not

• Regional stock shares roll jurisdiction buckets into continents using the Teneo Big Three (~63%) plus estimated Japan/Korea/RoW splits from the theme concentration ledgerthey are geography reconstructions, not a new GTA microdata extract.
• Package USD figures are statutory appropriations, mobilisation targets, state-aid approvals, or tax-credit scoresnot outlays-to-date — and are not comparable dollar-for-dollar across jurisdictions.
• US state award shares are estimated from public CHIPS megaproject / facility notices for a ~$39B tracked tip, not a complete Commerce disbursement census, and not IRA energy tax-credit claims.
• Sector × region intensities are estimated allocations of strategic / dual-use themes (GTA ZG #88 + NIPO sector coding), useful for relative geography, not audit-grade HS6 maps.
• June 2026 flow is one month’s Roundup geographya flow check, not a substitute for multi-year stock.
• EU is treated as one bloc (EU + member states) to match Teneo Big Three framing; intra-EU member-state geography is out of scope here.

What desks should take from the map

Industrial-policy geography is a stack of disagreeing maps. East Asia leads the cumulative count map (~33%). North America owns the major package-dollar map (~71%). AZ·TX·NY own the US fab-corridor map (~58% of tracked awards). Rest-of-world still owns much of the monthly flow map (62% in June). Semiconductors lean Asia; clean energy leans North America. Subsidies, tariffs, and industrial-policy toolkits reshape economies by moving these layers at different speeds — not by painting one colour onto one world map.

Related reading: research ledger · concentration companion · Q3 chokepoint update · August toolkit update.