Coffeehouses and the Rise of Science

Historians have credited the seventeenth-century coffeehouse with fostering the exchange that powered the Scientific Revolution. Bounadi and Oehlen’s 2026 job-market paper is, to their knowledge, the first causal quantitative test of that claim. Working from hand-geocoded records across London’s 162 parishes, they find that a parish’s first coffeehouse was followed by a persistent rise in mathematical practitioners and scientific instrument shops, concentrated among people outside the Oxford–Cambridge orbit, and a calibrated urban model attributes 31 percent of London’s applied scientists between the 1650s and 1690s to coffeehouses.1

The claim and its pedigree

The idea is older than the economics. John Houghton, a Fellow of the Royal Society, wrote in 1701 that coffee-houses “make all sorts of people sociable, the rich and the poor meet together, as also do the learned and unlearned,” where an inquisitive man “may get more in an evening than he shall by Books in a month.” Habermas placed the coffeehouse at the origin of the bourgeois public sphere — a space where reason rather than rank shaped debate — and Mokyr counts it among the social institutions that lowered the cost of accessing “useful knowledge.” For a penny, about a tenth of a day’s wage, strangers could talk across ranks: the “penny universities.”2

The experimental science of the period ran on instruments, and the skill to make them was tacit — learned by apprenticeship, rarely written down — so face-to-face contact was the transmission channel that print could not replace. That is the paper’s premise: universities filter access to knowledge exchange to maintain quality, but filtering narrows the pool of contributors; an open venue tests whether broadened access raises production.3

Data and design

The authors geocode every recorded coffeehouse, mathematical practitioner workplace, and scientific instrument shop in seventeenth-century London into a parish-by-decade panel from the 1600s through the 1690s. The three record sets come from separate compilations by different historians (Lillywhite for coffeehouses, Taylor for practitioners, Clifton and Hay for shops), so the results are not an artifact of one author’s framing. Practitioner biographies supply a second layer: reconstructed social connections — friendship, apprenticeship, correspondence, consultation. Identification uses a staggered, heterogeneity-robust event study (Wooldridge 2023): 61 treated parishes against never-treated ones, and the parishes are small — 97 of 162 lie within the City walls, so treated and untreated parishes were often a few streets apart.4

Findings

Treated parishes averaged 0.62 mathematical practitioners in the decade before a coffeehouse opened; over the entry decade and the four after, the count roughly doubles to 1.30. Scientific instrument shops rise from 0.33 to 1.57 — nearly fivefold. Summed across treated parishes, the estimated effects equal 49 percent of London’s practitioner growth and 63 percent of its instrument-shop growth from the 1640s to the 1690s.5

Three mechanism results narrow what the coffeehouse did. Effects are strongest among practitioners whose surnames were least represented at Oxford and Cambridge — the access channel, drawing in talent the universities filtered out. Connections between craftsmen and theory-oriented practitioners rise, linking instrument skill to the development and teaching of theory. And Hooke’s diary shows him meeting two more people on coffeehouse days than on comparable days elsewhere in the same year — about two-thirds of them from outside the Republic of Letters’ elite correspondence network. Shops producing experimental instruments (microscopes, air pumps) rise proportionally faster than shops producing practical measurement and navigation instruments, from a smaller base — makers learning what the new experimental science demanded.6

Outside London

Across 200 European cities, applied mathematics titles and scientific instrument makers rise after a city’s first coffeehouse, relative to cities of similar size and growth: in the 37 cities with a coffeehouse before 1700, titles rose by 4.97 against a pre-entry mean of 5.95, and makers by 4.89 against a mean of 3.73. The authors read these as conditional correlations rather than causal effects — but they lack pre-trends, and they answer the relocation objection: activity can shuffle between adjacent London parishes, but not between London, Paris, and Munich.7

What the model adds

A quantitative urban model formalizes the coffeehouse as a meeting technology — meeting rates fall with distance and rise with a coffeehouse at either end — with entry into science chosen given the returns each parish offers. Calibrated to the parish and city estimates, coffeehouses account for 31 percent of London’s applied scientists from the 1650s to the 1690s; 23 percent of the gain in coffeehouse parishes came from elsewhere in London, the rest from new entrants the model counts as otherwise-lost talent.8

Robustness and limits

The threats the authors address: pre-trends (none), commercial booms coinciding with openings (controlled; entry followed commerce, so treated parishes started with more of it), record survival, spillovers onto neighboring parishes (positive spillovers, weighing against simple displacement), and the “just another meeting place” objection — earlier tavern openings rooted in traditional English sociability produce no comparable increase. The limits are the authors’ own: the data end in the 1690s and show nothing directly about later British technological change, the European city results are not identified, and the instrument trade’s secrecy is why patents cannot serve as the outcome measure. The paper is a job-market manuscript, not yet through peer review.9

Where the institution came from

Cowan’s history supplies the setting the econometrics abstracts from. The coffeehouse was not designed for science; it was a new kind of public house built on familiar templates — alehouse, inn, tavern — usually one room in the keeper’s own domicile, serving weak, unfiltered coffee alongside tea, chocolate, and brandy-fortified oddities. Its link to news and learned talk “had to be invented”: the template was set by the virtuosi, gentlemen whose culture of curiosity and wide-ranging discourse patronized the first coffeehouses and fixed the expectation of what a coffeehouse would be. Once public, the institution escaped that elite template and mixed ranks. The two accounts sit at an interesting angle: Cowan shows the coffeehouse began as an elite virtuoso institution and doubts Habermas’s rank-blind public-sphere reading, while Bounadi and Oehlen’s parish evidence shows its measurable effect concentrated precisely among practitioners outside the university elite — the institution’s later clientele, not its founders, are where the scientific action was.1011

  • industrial-revolution-antecedents — coffeehouses arriving in 1652 sit at the onset of sustained English productivity growth, and feed the same useful-knowledge supply side that Hoffman’s antecedents argument runs on
  • knowledge-problem — the coffeehouse as a pre-industrial solution to transmitting tacit knowledge that print could not carry
  • gradualism-and-revolutionary-change — a case where an institutional change accelerates a “revolution” without any single revolutionary moment
  • industrious-revolution — Cowan’s account reshapes the consumer-revolution story those pages build on

Sources

Footnotes

  1. Monir Bounadi, Jens Oehlen 2026 — Coffeehouses and the Rise of Science ↩

  2. Monir Bounadi, Jens Oehlen 2026 — Coffeehouses and the Rise of Science ↩

  3. Monir Bounadi, Jens Oehlen 2026 — Coffeehouses and the Rise of Science ↩

  4. Monir Bounadi, Jens Oehlen 2026 — Coffeehouses and the Rise of Science ↩

  5. Monir Bounadi, Jens Oehlen 2026 — Coffeehouses and the Rise of Science ↩

  6. Monir Bounadi, Jens Oehlen 2026 — Coffeehouses and the Rise of Science ↩

  7. Monir Bounadi, Jens Oehlen 2026 — Coffeehouses and the Rise of Science ↩

  8. Monir Bounadi, Jens Oehlen 2026 — Coffeehouses and the Rise of Science ↩

  9. Monir Bounadi, Jens Oehlen 2026 — Coffeehouses and the Rise of Science ↩

  10. Brian Cowan 2016 — On the Origin of the Coffeehouse ↩

  11. Monir Bounadi, Jens Oehlen 2026 — Coffeehouses and the Rise of Science ↩