The full essay, online
about an hour · no download
The complete essay in your browser, with 16 live figures you can explore — toggle the West bloc, read every chart, follow the long view. Free, nothing to install.
Read the essay online →
GDP shares, population, and power in world history — five centuries of economic transformation.
Size is headcount; weight is productivity.
It takes a memorable claim about the economic weight of nations and subjects it to scrutiny rather than repeating or debunking it. Behind each share-point is a life the headline forgets — a clinic with a queue, a haircut in Delhi worth less at the exchange rate than the same hour in Zurich. The whole method is one identity:
A nation's share of world GDP is population × price-level × productivity. People switch silently between two rulers — market-rate nominal and price-adjusted PPP — and between size and prosperity. Separate the rulers, divide out population, and five centuries of rise and fall collapse onto one durable variable: productivity.
Read the ruler first, the direction second, and the decimal point last.
Every edition is rendered from a single source of truth — one data file, one manuscript — so the formats can never drift. Same essay, three formats.
about an hour · no download
The complete essay in your browser, with 16 live figures you can explore — toggle the West bloc, read every chart, follow the long view. Free, nothing to install.
Read the essay online →62 pages · to keep & cite
The same essay, typeset with the cover — to download, print, quote, and cite. Embedded figures and reference tables; the citable artifact.
Download the PDF →spoken aloud · bring a TTS app
Prose stripped of tables and figure scaffolding so a text-to-speech app (Speechify, Peech, or your screen reader) reads it cleanly aloud. A document for TTS, not a recorded audio track.
Download the TTS edition →Short on time? A 5-minute version is on Medium and Substack — the argument and the key figures, condensed.
Three layers keep the claim honest: dated editions to quote and cite, a public repository as the openly-licensed source of truth, and a skill that codifies the reasoning. When a new IMF, World Bank, or UN release lands, the numbers change in one place — build/data.py — a 52-check suite re-runs, and every edition rebuilds. The argument, a method and a set of directions, does not expire when a denominator is revised.
Earlier editions are never overwritten: each is a dated tagged release, and a new one stands beside the old. Anyone may read, copy, fork, or translate it under the licence below — here is how to update when new figures land. But the canonical edition, main, and this site are altered only by the author: a fork can go its own way, it cannot take over the original.
Data is drawn from public sources, kept in separate lanes and never spliced: IMF World Economic Outlook, World Bank WDI, UN World Population Prospects, the Maddison Project, and Bairoch (manufacturing). Every public claim is logged and ruler-labelled in the claim registry. An independent analysis built from a public Uncommon Knowledge conversation (Stephen Kotkin with Peter Robinson) — not a transcript, an endorsement, or a Hoover publication.
PPP figures carry roughly ±5–10% structural uncertainty; deep-history figures are reconstructions with wide error bars. Read the ruler first, the direction second, and the decimal point last.