Q3 Concentration: GERD Top-1 Is 29.4% China — Gap Shrinks to 0.3 pp; Non-OECD Hits 92%

Our 2026 concentration lens answered the distribution question on four perimeters: GERD Top-1 at 29.4% China, publications Top-3 at 50%, CET AI patents near 75%, and KTI VA still US-led at 28%. The Q3 OECD/AAAS restatement then showed that finer shares shrink the China–US GERD gap from 1.0 pp to 0.3 pp without breaking the duopoly. This post is the concentration vintage of that restatement: did converter noise change who sits at the top, or only how fragile the ranking looks?

The interactive dashboard above is a Q3 concentration lens — not another KPI strip. Toggle HHI / scoreboard, Ladder + Lorenz, Converter frames, and Vintage + impact. The punchline is deliberately frame-dependent. On headline PPP, Top-1 is still China at 29.4% and Top-3 clears 65.2%. On the non-OECD ledger alone, Top-1 hits 92.3%. On alternate PPP or market exchange rates, China falls back below the United States and the world Top-1 label flips. Same 2024 expenditure year; three different concentration stories.

The Q3 concentration scoreboard

Read the table as a family of market shares under different meters, not one number. The prior concentration print already showed that funding, papers, patents, and output disagree on who is Top-1. Q3 adds a fifth disagreement: whether China is Top-1 at all depends on which converter you trust for the same GERD year.

Analysts who quote only the NSF rounded 30% / 29% overstate the lead. Analysts who quote only EXR understate how thick the PPP duopoly is. Analysts who ignore the non-OECD ledger miss where concentration is most extreme once OECD peers are stripped out.

Vintage slope: NSF rounded → AAAS finer shares

Open Vintage + impact. The slope from the NSF State of S&E rounded print to the OECD/AAAS Q3 restatement is the first hinge of this post. China share moves 30% → 29.4%; US share 29% → 29.1%; the gap collapses 1.0 → 0.3 pp; US+China eases 59% → 58.5%; analytical GERD HHI eases ~1,920 → ~1,890. Top-3 stays near 65%.

That is not diversification. Three names still clear about two-thirds of world PPP R&D. It is a restatement that makes the overtake fragile without making the duopoly thin. The concentration shape — steep Top-2, Japan as a distant third, RoW closing the ladder — barely moves. What moves is the political readability of a 0.3 pp lead.

Pair this panel with the Q3 update: the same vintage that narrowed the headline also documented China’s business BERD lead of about +$100B (constant 2020 PPP) and a US government-sector lead of about +$19B. Concentration of funding effort and concentration of who wins the converter debate are different questions.

Converter frames: when Top-1 disappears

Toggle Converter frames. Three meters of the same China-vs-US GERD ratio:

• Headline PPP (OECD/AAAS): China at about 102% of USworld Top-1 stays China; US+China still 58.5%.
• Alternate PPP sensitivity: China at about 90–95% of USovertake vanishes; US remains ahead.
• Market exchange rates: China near ~50% of USUS clear Top-1; the “China overtook” headline never prints.

This is the second hinge. A desk that only watches PPP world shares sees a knife-edge concentration story with China on top. A desk that only watches EXR sees a US-led funding system where China is large but not first. Both desks can be reading honest OECD/AAAS tables. The concentration answer is frame-conditioned.

Treat converter frames as a ranking stress test, not a claim that PPP is wrong. The prior concentration post already warned that EXR paints a different China-vs-US picture; this Q3 lens makes the boolean explicit: does world Top-1 stay China? Only under headline PPP.

Sector split: business BERD funds the knife-edge

Stay on Converter frames or return to HHI / scoreboard for the sector bars. China’s business BERD lead of about +$100B against the United States (constant 2020 PPP) is the micro-foundation of the 0.3 pp world-share edge. The US still leads government-sector R&D by about +$19B. OECD-area funding mix remains business-heavy (~64% business-funded vs ~23% government-funded in the Q3 OECD framing).

That is the third hinge. Concentration is not only which country sits at Top-1; it is which sector inside those countries is pulling the ladder. A government-budget desk that watches GBARD defence reorientation can miss that the world-share knife-edge is a business BERD story. A corporate-R&D desk that watches only BERD can miss that US government dollars still print a lead on the public ledger.

The non-OECD panel sharpens the same point from another angle: once you leave the OECD perimeter, China alone is about 92.3% of non-OECD R&D — an HHI near 8,500 on the disclosed-plus-residual basket. Mid-tier emerging innovators exist; they do not yet thicken the non-OECD ladder the way India thickens the publication Top-3.

Ladder + Lorenz: five perimeters, one shape family

Switch to Ladder + Lorenz. Toggle GERD, publications, CET AI, KTI, and non-OECD. The Lorenz-style curves all sit above the equal-split line; how far above depends on the perimeter:

• GERD: Top-1 29.4%, Top-2 58.5%, Top-3 65.2%duopoly floor.
• Publications: Top-1 31%, Top-3 50% (China–US–India)steeper Top-1, half-world oligopoly at three.
• CET AI: Top-1 ~75%, Top-3 ~92%cliff, not knife-edge.
• KTI VA: Top-1 28% (US), Top-3 71%Top-1 label flips; concentration stays thick.
• Non-OECD: Top-1 92.3%the extreme Q3-specific perimeter.

HHI by lens in the scoreboard view ranks the same way: non-OECD and CET AI dominate; GERD and KTI sit in a mid-band near 1,800–1,900; publications are milder on HHI because the RoW residual is still large even after Top-3 clears half.

Open Vintage + impact for the volume×HCA scatter. China sits far right on publication share and only mid-pack on highly-cited article rate (~1.3% of own articles in the world top 1% cited vs US ~1.7%). Volume concentration and citation-intensity leadership still disagree — the same asymmetry the 2022 research ledger flagged and the August 202608 CET update sharpened for patents.

Who is exposed — and what would change the story

Exposed: ministries that treat “China overtook on GERD” as a settled fact without stating the converter; multilaterals that model non-OECD R&D as a diversified emerging-market club when one country is 92% of that ledger; industrial-policy desks that debate intensity gaps of a few tenths of a point while CET AI priority-family volume sits near 75%; research strategists who equate publication Top-1 with HCA leadership; investors who price KTI services leadership without noticing that manufacturing VA on the same $11.7T ledger is China-led.

Relative winners under current meters: China-scale business BERD platforms that funded the PPP knife-edge; US government and services franchises that still lead on public R&D dollars, HCA rates, and KTI services VA; labs and standards bodies that convert GERD into cited output rather than article counts alone; smaller high-intensity systems that punch above world-dollar share even when they are not Top-3.

What would change the story: a clean US reclaim on headline PPP (or a sustained China gap beyond converter noise); alt-PPP and EXR converging toward the 102% PPP ratio so frame debates stop flipping Top-1; non-OECD Top-1 falling well below ~70% as India and others scale R&D dollars, not only papers; CET AI China share compressing toward the semiconductor ~40% band; KTI total VA Top-1 flipping to China. Q3 shows the gap shrinking and the ranking remaining frame-fragile — not a break in concentration shape.

Caveats and methodology

• GERD shares use OECD/AAAS March–June 2026 restatement of 2024 PPP (China 29.4% / US 29.1%); NSF rounded prints (30% / 29%) are the prior vintage on the slope panel, not a second expenditure year.
• Converter ratios (China vs US ≈102% PPP, 90–95% alt PPP, ~50% EXR) follow the Q3 theme update; alt-PPP midpoints are analytical.
• Business BERD +$100B and government +$19B are constant-2020 PPP AAAS/OECD sector gapsnot nominal EXR gaps.
• Non-OECD China share 92.3% is disclosed AAAS; India+ and residual rungs are estimated closures for the ladder.
• Do not splice China NBS yuan GERD (¥3.93T domestic in 2025) into PPP world shares; the August 202608 update tracks that flow separately.
• Publication and HCA figures follow NSF Indicators framing; HCA is a rate, not a world HCA market share.
• CET AI ~75% is an NSF Translation volume anchor; US/EU residual ladder shares are estimated.
• HHI values are analytical on disclosed-plus-residual buckets (0–10,000), not official NSF or OECD statistics.
• KTI VA shares follow NSF State of S&E 2026; Japan/RoW are estimated closures.

The shareable takeaway

Measurement and science remain concentrated at the top — and Q3 mostly changed how fragile the Top-1 label is, not the shape of the ladder. On headline PPP GERD, Top-1 (China) is 29.4%, Top-3 about 65%, and the China–US gap is only 0.3 pp (from 1.0 pp on the NSF rounded print), with US+China still near 59%. On the non-OECD ledger, Top-1 hits 92.3%. On CET AI priority patents, Top-1 is about 75%. On KTI value-added, Top-1 flips to the United States at 28%. Business BERD (+$100B) funds the knife-edge; alternate PPP and EXR converters reverse the ranking. Progress is still funded and published through a system that looks contested in headlines and top-heavy once you rank the distribution — and converter-conditioned once you stress-test the meter.

Related reading: Concentration lens (2026) · Q3 OECD/AAAS restatement · August 202608 China flow + CET · 2022 research ledger.