Who really drives innovation? The surprising power of public science

When we talk about innovation, the imagery is predictable: venture capital, Silicon Valley, founders chasing scale. But new research from London Business School's Professor Paolo Surico suggests the story we tell ourselves is, at best, incomplete (Who Really Drives Innovation? CEPR & VideoVox Economics, 24 April 2026).

In a sweeping study of post-war US innovation, Surico and his co-authors trace the origins of transformative breakthroughs not to the private sector, but to publicly funded science. Their paper, The Public Origins of American Innovation (CEPR Discussion Paper 20788), offers a striking conclusion. Just two per cent of patents account for around 20 per cent of productivity growth and those patents are disproportionately rooted in government-funded research.

The implication is not subtle. The real engine of innovation is not who commercialises ideas, but where they begin.

The quiet architecture of discovery

To understand why, Surico’s work returns to a foundational moment. In 1945, US engineer and inventor Vannevar Bush set out a blueprint for the US innovation system in Science, The Endless Frontier. His model was elegantly simple: government funds basic research, universities push the frontiers of knowledge, and the private sector scales and commercialises the results.

Eighty years on, the data suggest that architecture still holds.

Using granular patent records, the researchers track whether innovations were publicly funded, privately developed, or jointly produced. Crucially, US patents disclose government involvement, allowing the team to link early-stage research to long-run economic outcomes. Periods with unusually high levels of government-backed patenting are followed, several years later, by surges in productivity.

This is not coincidence. It is structure.

Why small shares deliver outsized impact

Government-funded patents are few in number but different in nature. They are more likely to emerge from basic research, carry broader spillovers, and shape entire technological trajectories. In short, they are riskier, slower, and vastly more consequential.

That helps explain the 2-to-20 ratio.

Public funding does not aim for incremental gains or immediate returns. It absorbs uncertainty that private capital avoids. Many projects fail. But the successes are outsized “home runs” whose benefits ripple across the economy, from pharmaceuticals to computing to navigation systems.

Importantly, this is not a story of government picking winners. It is a story of governments funding uncertainty and allowing the distribution of outcomes to do the rest.

Not the Pentagon, but science and health

One of the study’s more counterintuitive findings is institutional. While defence spending looms large in public debates, it is agencies focused on basic science and health that drive the biggest productivity gains.

In the US, that means the National Institutes of Health and the National Science Foundation. Their edge is not sectoral, but methodological: a sustained commitment to fundamental research.

The lesson is subtle but important. Innovation is not inherently “military” or “medical” or “digital”. What matters is whether it originates in deep, curiosity-driven science.

Crowding in, not crowding out

For decades, economists have debated whether public R&D displaces private investment. Surico’s findings point firmly in the opposite direction.

Public research, particularly in universities and non-profits, appears to crowd in private capital. It expands the frontier, attracts talent, and creates opportunities that firms can later exploit. The result is a complementary system rather than a competitive one: the state leads on direction, the market on diffusion.

This dynamic is especially visible in startups. Young firms outperform incumbents when backed by public funding, combining mission-driven research with competitive pressure. Privately funded startups, by contrast, tend to optimise for profitability rather than broad productivity gains.

A fragile equilibrium

The model, however, is not guaranteed to persist. Surico points to two emerging risks.

First, cuts to US science funding in 2025 threaten the pipeline of basic research. The effects may not be immediate, but over time they could erode both productivity growth and technological leadership.

Second, the rise of AI is reshaping where talent sits. If leading researchers migrate from universities to private firms, the balance of innovation may tilt toward commercially driven outcomes, potentially weakening the broader societal gains that historically followed public investment.

Europe’s moment?

Across the Atlantic, the findings land with particular force. Europe’s challenge, Surico argues, is less about how much it spends and more about how it spends it. Too much goes on procurement, too little on foundational research and scaling innovative firms.

The recent Draghi Report on European competitiveness may offer a turning point. If it catalyses a shift toward sustained investment in basic science and later-stage innovation, it could play a role analogous to Bush’s post-war blueprint in the US.

Europe, in other words, does not lack ideas. It lacks the system to keep them.

Rethinking the narrative

The enduring contribution of Surico’s work is to move the debate beyond anecdotes and ideology. For every celebrated success or high-profile failure, the aggregate evidence tells a clearer story: public investment in basic research has been a foundational driver of modern economic growth.

The private sector remains indispensable. But it is not the starting point.

Innovation, it turns out, is less about the visible triumphs of markets and more about the quiet, patient work of public science.

Listen to the entire podcast here