The Nobel Pipeline

Every Nobel laureate leaves a paper trail through the world's universities, and Wikidata has quietly assembled it: who they are, when they won, and every institution that ever taught them. Read as a single dataset — 998 prize records across physics, chemistry, medicine, literature, peace, and economics — it becomes a longitudinal X-ray of how the scientific elite is made. Three findings organize this article.

Median age, science prizes 2001–25

vs. 50 in 1901–1950

Physics median age, 2001–25

71.5

vs. 45.5 in 1901–1950

Science laureates without university

every autodidact is in Literature or Peace

1 Genius waits longer

In the first half-century of the prizes, the median science laureate was 50 at the ceremony — and physics ran youngest of all, with a median of 45.5 and famous cases in their twenties and thirties (Lawrence Bragg won at 25, Heisenberg and Dirac at 31). Since 2001 the science median is 68, and physics — once the young genius's prize — is now the oldest category at 71.5. Part of this is queueing: the prize rewards work done decades earlier, and the backlog of confirmed discoveries lengthens. Part is structural: the "burden of knowledge" — each generation must climb further up the mountain before touching the frontier.

Age at the prize, 1901–2025

One dot per laureate · line = median by decade of award

Data table (decade medians)

Figure 1. Six panels, shared axes (age 20–100). The upward drift is universal but steepest in physics; peace is the only category whose laureates got younger in the 2010s (Malala Yousafzai won at 17, the youngest laureate in any category). Hover any dot for the laureate.

2 The great absorption

Where were science laureates educated? Before 1934, just 6% had touched an American university; the pipeline ran through Berlin, Munich, Göttingen, Cambridge and Paris. Between 1933 and 1945 the German university system expelled or drove out a generation of its scientists, and the American research university — flush with them, then with postwar money — became the pipeline's center. Of science prizes awarded 1967–2000, 55% of laureates had studied at a U.S. institution. The absorption has held, not grown, since 2000 — a plateau worth watching.

Share of science laureates educated in each country

Physics + Chemistry + Medicine · by decade of award · a laureate can count toward several countries

Data table

Figure 2. "Educated in" = any P69 institution in that country, undergraduate or doctoral. Shares can sum past 100% (many laureates trained in two countries — that mobility is itself the story). Germany's line collapses after the 1930s and never recovers its prewar share.

3 The few doors

The education graph is astonishingly concentrated. Harvard alone appears in the education history of 76 laureates; the top five institutions — Harvard, Columbia, Cambridge, MIT, Chicago — account for a substantial share of all science prizes ever given. Concentration compounds: laureates train students who become laureates, in the same buildings.

Institutions in the most laureates' education histories

All six prizes · a laureate with three alma maters counts once for each

Data table

Figure 3. "Trinity College" is Trinity College, Cambridge — listed separately from the University of Cambridge in Wikidata, so Cambridge's true total is higher than either bar alone. The same is true of other collegiate universities (Oxford colleges, University of California campuses are separate).

4 The slowest door of all

Sixty-nine of the 992 laureates in this data are women — 7%. The distribution across categories is itself the story: 21 in Peace, 18 in Literature, 14 in Medicine… and 5 in Physics in 125 years, two of them since 2018. The 21st century is measurably different (38 women among 297 laureates since 2001, 13%) but the physics pipeline remains, by a wide margin, the narrowest.

Women laureates by decade and category

Count per decade of award

Data table

Figure 4. Six panels, shared y-scale (0–14 per decade). Marie Curie is the 1900s physics bar and the 1910s chemistry bar — the only person to win two science prizes. The 2020s panels are partial (through 2025).

Put the three findings together and the pipeline's shape emerges: longer — the median laureate now spends two more decades between training and telegram; narrower — a handful of institutions mint an outsized share; and still, in its most mathematical corridor, almost entirely male. Whatever genius is, the route it takes has rarely been this legible.

Data & methods

Source. Wikidata (CC0), queried via the SPARQL endpoint 2026-07-09. Laureates = humans with an award statement (P166) for one of the six prizes; award year from the point-in-time qualifier (P585) on that statement. Organizations (Peace prize NGOs, IPCC, etc.) are excluded by the human filter — counts here are people, so category totals differ from official prize counts.
Education. "Educated at" (P69) as recorded by Wikidata editors — typically degree-granting institutions, occasionally secondary schools. Coverage is excellent for science laureates and thinner for Peace and Literature (all 32 laureates with no P69 are in those two categories — some genuinely autodidact, some under-documented).
Countries. An institution's country is its current P17 value — Strasbourg counts as France even for laureates who studied there under the German Empire; Soviet-era institutions count as their successor states. This smooths borders, not the migration story.
Age = award year − birth year (±1 year precision); implausible values (outside 15–100) dropped.
Collegiate ambiguity. Cambridge/Oxford colleges are separate Wikidata entities from their universities; Figure 3 does not merge them (noted in caption).
Multi-prize people (Curie, Pauling, Bardeen, Sanger…) appear once per prize — the unit is the laureate-prize record.