Why Innovation Clusters: Silicon Valley and Beyond

Why inventors and tech companies crowd into a few places, and whether governments can create the next innovation hub.

With video calls and cloud software, it might seem that innovators could work anywhere. Yet innovation remains strikingly concentrated. A small number of places, such as the San Francisco Bay Area, Boston, Shenzhen, Tokyo and Bengaluru, account for a very large share of the world’s patents, tech start-ups and venture capital.

Why innovators cluster

Economists point to three main forces, first described by Alfred Marshall over a century ago:

Knowledge spillovers: ideas travel through conversations, job changes and chance meetings. Research using patent citations finds that inventors are more likely to cite patents from nearby inventors, suggesting knowledge still spreads locally.
A deep talent pool: firms can find specialised workers, and workers can change jobs without moving city.
Specialised suppliers and investors: lawyers, venture capitalists and component makers who understand the industry cluster nearby.

These advantages reinforce each other. Talent attracts firms, and firms attract more talent.

The history of Silicon Valley

Silicon Valley grew from a mix of factors: Stanford University’s engineering programmes, heavy U.S. military spending on electronics after the Second World War, early semiconductor firms whose employees left to found new companies, and a culture that tolerated failure. California’s general refusal to enforce non-compete agreements is often cited as helping workers move freely between firms, spreading knowledge.

Bengaluru's rise

Bengaluru became India's main technology hub after companies such as Infosys and Wipro grew there and multinational firms opened offices from the 1980s and 1990s. Engineering colleges supplied talent, and each successful firm trained workers who later founded or joined others. The cluster became self-reinforcing, even though other Indian cities had similar advantages on paper.

Can governments build clusters?

Many governments have tried to create the next Silicon Valley with science parks and subsidies. Results have been mixed. Economists note that successful clusters usually grow from existing strengths, such as a strong university or an established industry, rather than being built from nothing. Clusters also bring costs, especially very high housing prices that push out lower-paid workers.

Thinking the internet has made location irrelevant

Remote work has grown, but the most innovative activities remain highly concentrated. Face-to-face contact, local job markets and informal networks still matter for sharing complex, early-stage ideas.

Key takeaways

Innovation is concentrated in a small number of cities worldwide.
Knowledge spillovers, talent pools and specialised suppliers draw innovators together.
Silicon Valley grew from universities, military spending, spin-off firms and worker mobility.
Governments find it hard to create clusters from scratch; most grow from existing strengths.