Charted: Top-3 Capitals Own 63% of Industrial-Policy Stock — Package Dollars Hit 93%

Our H-NIPO research ledger answered the stock question: jurisdictions using trade-distorting subsidies nearly doubled from 36% to 59%, 34,248 interventions accumulated through 2023, and China, the EU, and the United States together accounted for roughly 53% of that historical tape. The Q3 chokepoint update then answered the targeting question: strategic / dual-use subsidy shares inside the Big Three now sit at 76–98%. This post answers a different distribution question desks actually trade when they hear “industrial policy is everywhere”: how concentrated is the system at the top?

The interactive dashboard above is built as a concentration lens. Toggle Concentration ladder, Package dollars, Strategic intensity, and Monthly flow. The punchline is deliberately two-sided. On counts, the top of the distribution is thick but not monopolistic — top-1 near 24%, top-3 near 63% on the Teneo cumulative stock framing. On major fiscal-package dollars, concentration is extreme: the United States alone is about 71% of the CHIPS/IRA/EU/CN/JP/KR war-chest universe we track, and the same Big Three clear roughly 93%. Monthly flow sits in between — and often looks less concentrated than the stock, because rest-of-world still prints most of the June Roundup tape.

The headline ladder: top-1 and top-3 across lenses

Read the table as a family of market shares, not one number. Count-based top-3 sits in a 48–63% band depending on vintage and coding. Dollar-based top-3 among large packages sits near 93%. Flow-based top-3 in a single busy month can fall to 38%. Analysts who quote only the Big Three stock share understate how dollarised US packages dominate fiscal capacity; analysts who quote only IRA headlines understate how many capitals still generate the daily intervention count.

Count concentration: thick top, living tail

Filter the ladder to Stock counts. The cumulative curve rises to about 24% at top-1, 45% at top-2, and 63% at top-3 — the Teneo Big Three share of cumulative industrial-policy interventions. Japan and Korea add a thin next rung; rest-of-world still holds roughly 28% of the stock in our six-bucket reconstruction.

That is concentrated relative to an equal six-way split. It is not a winner-take-all market. A system where the top three hold three-fifths of recorded interventions still leaves a long tail of capitals that can start subsidy races, copy HS6 product lists, or free-ride on templates written in Washington, Brussels, and Beijing. Our August 202608 toolkit update already flagged the same asymmetry in monthly data: Big Three headlines dominate narrative; rest-of-world counts still dominate parts of the flow.

IMF’s H-NIPO check sits a notch lower — China+EU+US at 53% of the 2009–2023 stock — and the 2023 census alone put the same trio at about 48% of 2,580 measures. Different windows, different coding intensity, same qualitative story: the top is thick; the middle and tail remain politically alive.

Dollar concentration: the war chest is a different distribution

Switch the view to Package dollars. Among the major semiconductor and clean-tech fiscal headlines we roll up — US IRA clean-energy tax expenditures (~$370B original score), US CHIPS appropriations ($52.7B) and ITC (~$24B), EU Chips mobilisation (~$47B), EU IPCEI state aid (~$40B), China’s Big Fund III (~$48B), plus Japan and Korea semiconductor envelopes — the United States alone is ~71% of the dollar universe and the top three jurisdictions clear ~93%.

This is the sharpest top-1 / top-3 print in the dashboard, and it answers a different policy question than NIPO counts. Counts ask who intervenes how often. Package dollars ask who can put nine- and ten-figure claims on factory geography. A jurisdiction that records many firm-level awards can look “active” on the count ladder 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.

Treat the package pie as a war-chest index, not an outlay tracker. Appropriations, mobilisation targets, state-aid approvals, equity-fund raises, and tax-credit scores are not interchangeable cash. They are comparable as political commitments: fiscal industrial policy, measured in headline dollars among the large packages, is more concentrated than the intervention stock.

Strategic intensity: concentration inside the Big Three

Concentration is not only across jurisdictions. Toggle Strategic intensity. GTA’s ZG #88 print shows the share of subsidy-based industrial-policy actions covering dual-use / advanced tech rising to 98% in China, 76% in the United States, and 70% in the EU in the 2025–26 window — from early-period baselines of 85%, 33%, and 50%.

That is a second concentration story: even when the count ladder still leaves room for rest-of-world activity, the money and rules inside the Big Three are increasingly aimed at the same chokepoint product map. The US catch-up (+43 pp from the early window) is the steepest path; China never left the high-nineties. Pair this panel with the Q3 update’s follow-rate story: same-product subsidy races after 2020 often sit in a 60–80% band. Concentration of targets amplifies concentration of actors.

Monthly flow: when the tape looks less top-heavy

Open Monthly flow. June 2026’s GTA Roundup split is disclosed and blunt: United States 163 developments (20%), EU + member states 100 (12%), China 47 (6%), rest of world 513 (62%). Top-1 is the US; top-3 is only 38%.

That does not contradict the 63% stock figure. Stocks accumulate years of Big Three activism; a single month’s all-jurisdiction tape still includes routine trade and industrial developments far beyond CHIPS press releases. The analytical mistake is to treat every Section 301 notice as the entire industrial-policy cycle. The analytical mistake in the other direction is to treat rest-of-world count dominance as evidence that fiscal capacity is evenly distributed — the package panel says it is not.

Import-barrier concentration adds a coercive wrinkle. In the 2025 distortive toolkit, import barriers were about 27% of actions, and the United States alone was roughly 20% of those barrier actions. Count concentration and coercive-tool concentration can diverge: a capital can punch above its stock share on tariffs and screens even when it does not lead every subsidy race.

Who is exposed — and what would change the story

Exposed: mid-sized economies that must follow Big Three HS6 product lists without matching US/EU/CN fiscal envelopes; exporters into markets where package-dollar concentration funds local capacity and import-barrier concentration raises walls; desks that underwrite “industrial policy is everywhere” as an equalising story when the dollar war chest remains US-heavy; firms that win announcement subsidies in the count ledger but still sit outside the strategic product maps that now absorb 70–98% of Big Three subsidy IP.

Relative winners under current rules: incumbents already inside dual-use and advanced-technology lists; capitals that can stack multi-year tax credits, appropriations, and export controls; upstream equipment and critical-mineral suppliers sitting at the intersection of concentrated package dollars and concentrated strategic targeting; rest-of-world policymakers who still generate most of some monthly tapes and can copy templates without funding them at IRA scale.

What would change the story: a sustained rise in rest-of-world package dollars that breaks the ~71% US share of the war-chest universe; a collapse of Big Three strategic shares back toward early-2010s levels; or a multi-year drop in cumulative stock concentration below the 50% top-3 band. None of those appear in the 2023–2026 ledgers summarised here.

Caveats and methodology

• Counts ≠ dollars. One firm award and one multi-year statute both count as interventions; fiscal impact differs by orders of magnitude.
• Individual count shares inside the Big Three are estimated to sum to disclosed bloc totals (Teneo 63%, H-NIPO 53%, 2023 census 48%). Treat top-1 count figures as order-of-magnitude, not census cells.
• Package USD figures mix metrics (appropriations, mobilisation targets, state-aid totals, tax-credit scores, equity raises) and FX conversions; the 71% / 93% prints are concentration inside this constructed universe, not global fiscal accounts.
• EU is coded as a bloc (institutions + member states) to match Teneo / Roundup geography; member-state aid drives much of the fiscal reality.
• June Roundup coverage is broader than NIPO-only selective IP, so flow shares are not a monthly restatement of the cumulative stock.
• Strategic shares are within-bloc percentages, not shares of world subsidiesChina at 98% means nearly all Chinese subsidy IP is strategic-coded, not that China owns 98% of world strategic subsidies.

The shareable takeaway

Industrial policy is concentrated at the top — but which top depends on the meter. On cumulative counts, top-1 is near 24% and top-3 near 63%. On major package dollars, top-1 (US) is about 71% and top-3 about 93%. On a busy monthly tape, top-3 can fall to 38% while rest-of-world still prints 62%. Inside the Big Three, strategic targeting itself concentrates toward dual-use product maps at 70–98%. Subsidies, tariffs, and industrial policy reshape economies not only by spreading activism across more capitals, but by piling fiscal capacity and chokepoint logic at the thick end of the distribution.

Related reading: Fiscal & industrial policy research (H-NIPO/2023) and Q3 update — strategic subsidy targeting.