Two small towns in the Swiss Jura mountains, La Chaux-de-Fonds and Le Locle, sit on terrain that was always too cold and rocky for agriculture. What they built instead was an industry that has lasted three centuries: watchmaking.
After a fire destroyed two-thirds of La Chaux-de-Fonds in 1794, the town was rebuilt on a grid plan. The streets were designed to mix housing with watchmaker workshops that needed north-facing light. The form of the city was engineered around a production process. By the time Karl Marx was writing Das Kapital, he analyzed La Chaux-de-Fonds as "a huge factory-town" — a case study in the division of labor. UNESCO inscribed both towns as World Heritage Sites in 2009, citing the physical urban plan itself as the legacy.
Population of La Chaux-de-Fonds today: approximately 40,000. Le Locle: approximately 10,000. Two towns, one mountain valley, marginal agricultural land, three centuries of globally competitive production.
No single company did this. Individual firms came and went — Omega, Jaquet Droz, Girard-Perregaux, Ulysse Nardin, Tissot all operated in or around the valley. But the industry endured through collapses that destroyed individual firms, because the knowledge wasn't housed in any firm. It was embedded in the population. The children of watchmakers became watchmakers. The skills circulated through the valley regardless of which company happened to hold them at any given moment.
Alfred Marshall described this phenomenon in his Principles of Economics in 1890. He called it the "industrial atmosphere" — the knowledge that "is in the air." A concentration of specialized producers in a particular place generates collective knowledge that exceeds what any individual firm possesses. The knowledge becomes external to the firm. It lives in the place.
This is not a historical curiosity. It is the most resilient economic form ever built.
The Hidden Champions
Germany has somewhere between 1,300 and 1,500 companies known as "hidden champions" — small and medium-sized firms that are global market leaders in narrow industrial niches. They represent roughly 40-47% of all hidden champions in the world, despite Germany having only 28 companies in the global Fortune 500.
The contrast is diagnostic. Germany punches far below its weight in mega-corporations. It punches far above its weight in niche world-beaters. A German company might make 60% of the global market for bottle-labeling machines, or 50% of the world's professional combi-ovens, from a town of 30,000 people that nobody outside the industry has ever heard of. That is the hidden champion pattern.
The key characteristics are worth examining together, because they reinforce each other:
Narrow market focus with global reach. The niche is too small for large corporations to bother disrupting, which means the hidden champion can dominate without a defense budget proportional to a multinational. Their competitive advantage is accumulated knowledge, not capital scale.
Long-term family ownership. Decisions are calibrated to the next generation, not the next quarter. The owner's children will inherit the reputation and the community relationships. This changes the time horizon of every decision.
The dual apprenticeship system. About 88% of German vocational learners at upper-secondary level are in dual programs — part classroom, part in-company training. Some 430,000 companies participate. Average employee tenure at Mittelstand firms runs 15 to 20 years. The apprenticeship system binds young workers to local employers from age 16. The knowledge pipeline runs from the firm into the next generation and back.
A 2024 study of 1,645 hidden champions across 401 German districts found that districts with more hidden champions show systematically higher median incomes, lower unemployment, higher apprenticeship rates, and more regional patents. Hidden champion density directly predicts regional prosperity. This is not complicated. The knowledge that makes a firm competitive lives in the town.
The Structural Variable
Consider Würth, founded in Künzelsau, Baden-Württemberg in 1945, grown from a two-person screws trading operation to a global network of over 400 companies and 81,000 employees. The founding family has spent approximately 70 million euros on a local innovation center, a university now named in the founder's honor, a major culture and convention center, and a foundation-funded school.
Künzelsau has a population of roughly 14,000 people.
A 14,000-person town has a privately-funded university, a major concert hall, a world-class innovation campus, and a foundation supporting children's education. Why? Because Reinhold Würth cannot exit. His wealth and his reputation are in the town. His grandchildren go to school there. The civic investment that looks like philanthropy is structurally rational: it is investment in the environment where the family will live.
This is the crucial variable that separates the Mittelstand model from the failed American company town: ownership alignment. In the Mittelstand model, the owners of capital live in the town they could destroy by leaving. In the company town model, the corporation's owners are elsewhere and have full exit optionality.
Pullman, Illinois (est. 1880): George Pullman built a model town entirely owned and controlled by the Pullman Company. Workers rented housing from Pullman, bought goods at Pullman stores, had no path to home ownership. When the 1893 depression hit, wages were cut but rents were not. The result was the 1894 Pullman Strike — 250,000 workers in 27 states over two months. The experiment was declared a failure before the company collapsed entirely.
Gary, Indiana (est. 1906): Founded by U.S. Steel on undeveloped southern Lake Michigan shore, named after the company's chairman. Peak population in 1960: 178,000. Current population: approximately 70,000. When U.S. Steel declined, Gary had nothing — no local ownership, no embedded knowledge economy, nothing to hold production in place. The owners were in Pittsburgh and New York. They left.
The Mittelstand firms stayed because their owners could not leave without losing everything that mattered to them beyond the money.
The District as Organism
The Italian industrial districts extend Marshall's insight in a different direction. Where the Mittelstand concentrates a broad firm capability in a single family over generations, the Italian district distributes knowledge across a dense network of small, specialized firms operating in the same narrow supply chain.
Prato, Tuscany's second-largest city (population approximately 198,000), hosts roughly 7,000 companies in the fashion sector. No single firm dominates. Knowledge and trust circulate through the district. Spinning firms know weaving firms; weaving firms know dyeing firms; dyeing firms know finishing firms; finishing firms know designers. The knowledge of how Prato textiles are made is not in any firm. It is in the relationships between firms, and in the fact that the people running those relationships live near each other, eat at the same restaurants, send their children to the same schools.
The Sassuolo ceramic tile district in Emilia-Romagna produces approximately 80% of all Italian ceramic tiles from a cluster spanning the provinces of Modena and Reggio Emilia. The district recently invested 400 million euros in Industry 4.0 automation — evidence that the district model is not a relic. It can absorb systematic technological change because the decision to invest is distributed across dozens of firms that collectively understand their own production process.
Emilia-Romagna adds an institutional layer that reinforces the district structure: cooperatives. In this region, cooperatives account for roughly a third to 40% of regional GDP; roughly two out of three inhabitants are cooperative members. Bologna's per capita disposable income is the highest of any of Italy's 103 provinces.
The cooperative structure matters because it changes the firm's objective function. Cooperative firms optimize a mix of employment and profit, weighting employment more heavily during downturns. During recessions, they reduce hours and wages before reducing headcount. This produces significantly better employment stability than conventional firms. The knowledge does not walk out the door when the market turns.
What Cities Destroy
Major cities attract educated workers and generate real agglomeration economies. The returns are measurable. College-educated workers in large US cities earn approximately 40% more than equivalent workers in small towns. Cumulative job growth since 2009 has been nearly 14% in cities over a million residents versus 2.7% in rural areas and small cities.
But the talent concentration also destroys a category of value that does not show up in those numbers.
The knowledge that makes Sassuolo tiles, Montebelluna ski boots, or La Chaux-de-Fonds movements globally competitive is not general-purpose knowledge you can acquire at a university. It is accumulated in a specific place among people who have worked alongside each other across generations. This knowledge cannot be relocated to Milan or Zurich. Move the watchmakers out of the Jura valley and you dissolve the industry atmosphere that makes them capable of producing what they produce.
Jane Jacobs wrote about this in a different register. Her skepticism of branch-office economies was exactly this: places that attract corporate facilities without developing genuine local production capacity are economically fragile because the owners are elsewhere, the knowledge is elsewhere, and "corporations might suddenly leave the area, driven by global economic fluctuations, abandoning workers and families." The Mittelstand towns are the opposite of this. The owners cannot be somewhere else. The knowledge cannot be somewhere else. The production is anchored.
When Toyota built manufacturing plants outside Toyota City, it found it difficult to replicate the embedded knowledge environment. The city government responded by building a manufacturing innovation center specifically to support new business development and the preservation of monozukuri — the ethic of making things with deep care. The knowledge is so spatially anchored that it required a civic institution to preserve and transmit it. The city became a co-producer of the firm's competitive advantage.
What the Physical Form Carries
There is a pattern in what the successful anchor towns look like.
La Chaux-de-Fonds was literally rebuilt for walkers and workshop workers. Housing mixed directly into production strips. Sassuolo and Prato are dense pre-automobile cities where workers and workshops are proximate. German Mittelstand towns are predominantly European urban in form: town center, walkable core, housing near employment. Würth's art museum, concert hall, and innovation campus in a 14,000-person town are the contemporary equivalent of the civic patronage that underwrote Italian Renaissance urban form.
The American company town looked different from the inside: worker housing, factory, company store. All owned by the same entity. The spatial monoculture is the architectural expression of the economic fragility. When the company leaves, the town has no structure of its own.
The anchor towns that endured built civic life alongside economic life. Not instead of it, not after it, but alongside it, as an expression of the same embedded commitment that kept the firm in place. Würth's foundation schools and Billund's LEGO-funded airport and children's foundation are not corporate social responsibility in the contemporary branding sense. They are the mark of owners who cannot leave.
The knowledge is in the air of these towns. It is also in the streets, in the civic institutions, in the apprenticeship relationships, in the fact that the owner's grandchildren will be educated in the schools the owner funded. What looks like generosity is actually the structural consequence of alignment between those who hold capital and those who bear the consequences of how it is deployed.
That alignment is not a personal virtue. It is a product of spatial embeddedness. The owner who cannot exit builds differently than the owner who can.
This is what the Swiss watch valley, the Emilian cooperative district, and the Westphalian family firm have in common. The knowledge is in the air. The owners are in the town. The town endures.