The Entrepreneurial Ecosystem

In brief

Entrepreneurship does not spread evenly across the map. It piles up in a handful of places: Silicon Valley for software and semiconductors, before it the textile towns and cutlery districts that Alfred Marshall studied, later the biotech corridors around Boston and the film cluster of Hollywood. The entrepreneurial ecosystem is the explanation for that lumpiness. A region develops a dense, self-reinforcing network of skilled workers, investors, suppliers, universities, law firms, and, above all, shared informal knowledge. Once the network is thick enough, it produces new companies at a rate no single firm, however well run, could match on its own. The advantage lives in the place, not in any one business inside it. AnnaLee Saxenian's study of why an open Silicon Valley pulled ahead of a closed Route 128 is the sharpest evidence that how the network is wired matters as much as what is in it.

The full treatment

The problem it answers

Standard economics long treated the firm as the unit of analysis and location as an afterthought: a company chose where to sit by weighing rent, wages, and transport costs. That view cannot explain why startups keep clustering in the same few expensive cities when land and labor are cheaper almost everywhere else. If a software engineer costs twice as much in Palo Alto, why does the next founder still move there? The entrepreneurial ecosystem answers that the high-cost location supplies something the cheap one cannot: proximity to everything a young firm needs before it can afford to build any of it. The question the concept answers is why the productivity of a founder depends so heavily on the founder's neighbors.

How it works: agglomeration and spillovers

Alfred Marshall (1842 to 1924), in Principles of Economics (1890), gave the first systematic account. He observed that firms in the same trade tended to gather in "industrial districts," and identified three reasons that concentration pays. First, a shared labor pool: a thick local market for a specialized skill protects both workers, who can find another job without moving, and firms, which can hire fast. Second, specialized suppliers: a dense cluster supports vendors of intermediate goods and services too niche to survive on a single customer. Third, and most subtle, knowledge spillovers. Marshall wrote that in such a district the secrets of the trade are "in the air," picked up by proximity, gossip, and the movement of people between firms. Modern economists call these three forces agglomeration economies: benefits that accrue to firms simply from being near other firms.

The third force is the one that makes the ecosystem specifically entrepreneurial. Much of what a founder needs to know is tacit: how to price a product, when a market is ready, which investor actually returns calls, how a chip is really manufactured. Tacit knowledge does not travel well in documents; it moves through conversation, mentorship, and the churn of people changing jobs. A region where engineers switch employers freely, where a failed startup scatters its veterans into new ones, becomes a machine for redistributing know-how. Paul Krugman (born 1953), in Geography and Trade (1991), showed that once such concentration begins, increasing returns can lock it in: firms cluster because other firms are already there, and the advantage compounds. This is where the ecosystem connects to network effects: each new participant, another specialist lawyer, another angel investor, another experienced operator, makes the location more valuable to everyone else.

The key example: Saxenian on Silicon Valley versus Route 128

The most influential single study is AnnaLee Saxenian's Regional Advantage (1994). In 1970 the two leading American technology regions, Silicon Valley in California and the Route 128 corridor around Boston, looked comparable, both anchored by great universities and defense contracts. By the 1990s Silicon Valley had pulled decisively ahead. Saxenian argued the difference was not in the ingredients but in the industrial culture, the way firms and people were connected.

Route 128 was dominated by large, vertically integrated, secretive companies such as Digital Equipment Corporation. They kept their technology in-house, discouraged employees from leaving, and treated other firms as pure rivals. Silicon Valley developed the opposite ethic: porous firm boundaries, constant job-hopping, informal information sharing across companies, and dense collaboration with suppliers and universities. California law reinforced this by refusing to enforce the non-compete clauses that bound engineers in Massachusetts, so knowledge and people flowed freely. When a technology or a firm failed in the Valley, its talent recombined into new ventures; in the closed Boston system, that knowledge stayed locked up or died with the company. Saxenian's conclusion was that a decentralized, networked region adapts and generates startups faster than one built of self-sufficient fortresses, even when the fortresses are individually strong. Openness, not size, was the advantage.

Distinctions that matter

The concept has been given several names, and they are not quite interchangeable. Marshall's industrial districts describe geographic concentration of a single trade. Michael Porter (born 1947), in The Competitive Advantage of Nations (1990), rebranded and broadened the idea as the cluster: a geographic concentration of interconnected companies, suppliers, and institutions in a field, whose rivalry and cooperation drive productivity and innovation. Cluster theory is often about established industries and national competitiveness. The entrepreneurial ecosystem, a term that spread in the 2010s through writers such as Daniel Isenberg and researchers such as Erik Stam, narrows the lens to new-firm creation specifically, and adds elements Marshall did not stress: venture capital, serial entrepreneurs who reinvest and mentor, and a local culture that tolerates risk and failure. All three describe the same underlying gravity. They differ in whether the payoff is measured as productivity, national advantage, or the birth rate of startups.

Lineage

The intellectual line runs from Alfred Marshall's industrial districts (1890), through the mid-century urban economics of Jane Jacobs (1916 to 2006), who in The Economy of Cities (1969) argued that dense, diverse cities generate growth by letting new work sprout from old, to Michael Porter's clusters (1990) and Paul Krugman's formal "new economic geography" (1991), which won him the Nobel Memorial Prize in Economic Sciences in 2008. Saxenian's Regional Advantage (1994) supplied the decisive case study. The concept also descends from Joseph Schumpeter's account of creative destruction: if the entrepreneur is the engine of capitalist renewal, the ecosystem is the account of the habitat that engine needs to run. Where Schumpeter located dynamism in the heroic individual founder, ecosystem theory relocates it to the surrounding web.

The strongest case for it

The empirical record is hard to argue with. Innovative industries are radically concentrated: a large share of American venture capital and patenting clusters in a few metropolitan areas, and the pattern repeats worldwide, from Bangalore to Shenzhen to Tel Aviv. The concept explains this without appeal to luck: dense networks lower the cost of finding talent, capital, and customers, and speed the diffusion of the tacit knowledge that written manuals cannot carry. It makes falsifiable predictions, since regions that enforce non-competes and lock knowledge inside firms should generate fewer startups, and Saxenian's comparison bears this out. It also gives policymakers something to work with. Rather than subsidizing individual companies, they can try to thicken the network: fund research universities, ease the movement of skilled workers, and let capital and people recombine. The idea reframes economic development as cultivating a habitat rather than picking winners.

The strongest case against it

The concept has real critics, and the objections bite.

The first is that it is often unfalsifiable folk theory. Because a successful region always has, in hindsight, dense networks and spillovers, the story can be told after any success and explains nothing before it. The economist Josh Lerner, in Boulevard of Broken Dreams (2009), documented how governments squandered fortunes trying to build "the next Silicon Valley" by copying its visible features, science parks, state venture funds, incubators, almost always in vain. If the recipe cannot be executed, the theory may be describing a correlation, not a mechanism.

The second objection targets the spillover claim directly. Enrico Moretti (born 1968), in The New Geography of Jobs (2012), and much of the productivity literature confirm that agglomeration is real, but skeptics note that the tacit-knowledge channel is hard to measure and easy to overstate. Some concentration may simply reflect firms chasing the same scarce input, such as talented graduates, rather than any magic in the air.

The third is that ecosystems may be path-dependent accidents rather than reproducible systems. Silicon Valley's rise owed much to specific, unrepeatable history: Cold War defense spending, Stanford's early licensing, the transistor arriving at Shockley's lab, California's peculiar non-compete law. If the cause is a chain of contingencies, the concept offers a description dressed as a blueprint.

Finally, critics point to the costs the framing hides. Successful ecosystems generate ruinous housing prices, displacement, and sharp regional inequality, and the very openness Saxenian praised can coexist with precarious work and monoculture. Celebrating the ecosystem, on this view, can flatter a region's self-image while obscuring who pays for the concentration.

Where it stands now

Agglomeration is one of the most robust findings in economic geography, and the entrepreneurial ecosystem is now the default frame for how cities and governments think about growth. It has been institutionalized, in the work of the OECD, the World Bank, and countless regional development agencies that try to map and grow local ecosystems. The scholarly frontier has moved from asserting that ecosystems matter to the harder questions of how they form, why they decay, and whether they can be deliberately created or only cultivated. The honest consensus is split: the advantages of dense, open networks are real and measurable, but the record of engineering them from scratch is poor. The idea explains why entrepreneurship clusters far better than it explains how to make it cluster somewhere new.

Test yourself

Think of a place known for one kind of enterprise, and ask what a newcomer inherits there that they could not buy or build alone. Then ask the harder question: is that advantage something a government could deliberately reproduce elsewhere, or is it the residue of history that happened only once? Your answer marks exactly where the theory is strong and where its critics have a point.