Mariana Mazzucato: The State Took the Risk, the Market Took the Reward

The first time I read Mariana Mazzucato, I had the slightly annoying experience of finding out that an argument I thought was mine had already been made, better, with footnotes. Mazzucato is an Italian-American economist at University College London, where she is Professor in the Economics of Innovation and Public Value and the founding director of the Institute for Innovation and Public Purpose. She is best known for two books — The Entrepreneurial State (2013) and Mission Economy (2021) — and for a claim that sounds mild when you state it flatly and turns out to be explosive once you follow it anywhere: that governments have not merely fixed markets at the edges, they have been the risk-taking investors at the front of the riskiest technologies of the last seventy years. And that they have been paid almost nothing for it.

I want to set out what she actually argues, where the evidence is strong, where her critics land real blows, and what she says should change. She is worth doing carefully, because she is one of the few economists whose ideas have travelled out of the seminar room and into actual policy documents, and because the sloppy version of her argument — the government invented everything, tech founders are frauds — is not what she says, and is easy to knock down.

Who she is, and how she got here

Mazzucato was born in Italy in 1968 and raised largely in the United States, where her father was a physicist at Princeton. She trained as an economist, spent years in the innovation-economics tradition — the unfashionable corner of the field that studies where new things actually come from rather than how existing things clear at a price — and arrived at UCL in 2017 to found the institute she still runs.

She has since written two more books that matter to the argument: The Value of Everything (2018), which asks how economics decided what counts as value creation and what gets quietly reclassified as value extraction, and The Big Con (2023), written with Rosie Collington, on how the consulting industry hollowed out the state's own capacity to do things. Read in sequence, the four books are one continuous case. Public institutions do more than we credit them for; the accounting we use makes their contribution invisible; and once it is invisible, it becomes easy to outsource, underfund, and finally to believe it was never there.

She has also done the thing most academics never do, which is to take the argument into rooms where it has consequences — advising governments and international bodies on innovation and industrial strategy. That has made her unusually influential and unusually exposed. People who would ignore a journal article will attack a policy framework.

The core claim: the state as investor, not referee

Standard economics gives the state a narrow job. Markets fail in specific ways — public goods get underprovided, research has spillovers nobody can capture, pollution goes unpriced — and the state exists to patch those failures and then get out of the way. Fix the market, then leave. It is a modest, defensible theory, and Mazzucato's argument is that it describes almost nothing about how breakthrough technology has actually been financed.

Her account is that public agencies have not been patching failures at the margin. They have been doing the thing we praise venture capital for: putting money into ventures with long horizons, no visible market, high probability of failure, and no way to know in advance which bets will pay. Except the public agencies did it earlier, for longer, and at a stage when the uncertainty was highest and the private appetite lowest. By the time private capital arrives, she argues, the deepest uncertainty has usually been eaten by someone else.

Her marquee illustration is the smartphone. The line associated with her is that every component that makes the iPhone smart was publicly funded: the internet protocols, GPS, the touchscreen, voice assistance, key work on the microchip and the battery chemistry — each traceable to defence research agencies, national science funders, public universities and state laboratories, often decades before there was a product. She makes a parallel case about pharmaceuticals, where publicly funded basic research underpins a striking share of the science behind new molecular entities that are then priced as private discoveries.

Her point was never that Steve Jobs did nothing. It was that assembling publicly funded components into something people love is a real achievement — and a different achievement from bearing the risk of inventing them.

That distinction is the whole argument, and it is the part that gets lost. Mazzucato is not claiming bureaucrats built the iPhone. She is claiming that a story in which the private sector is dynamic and risk-taking while the public sector is slow and rent-seeking has the risk profile exactly backwards, and that this misdescription has real consequences for who gets paid.

The clearest single case for this is artificial intelligence, where the funding record runs for half a century before the first venture round. I have set out that evidence separately in who funded AI, and I will not re-run it here — Mazzucato's argument is older and broader than AI, and treating her as an AI commentator shrinks her.

Mission-oriented policy: what she wants instead

Mission Economy takes the Apollo programme as its organising image. Her point is not that we should go to the moon again, or that governments should run companies. It is that Apollo was a case of the state defining a hard, concrete, time-bound objective, and then organising procurement, research contracts and private firms around achieving it — the public sector setting direction, not just correcting prices.

She wants the same posture applied to problems we actually have: decarbonisation, antimicrobial resistance, dementia, affordable housing. Not fund the green sector, which is a subsidy programme, but a defined mission with a deadline, inviting bottom-up solutions from anyone who can deliver. The difference she insists on is between picking sectors and picking a direction — between choosing which firm wins and choosing which problem is worth solving.

On the returns side, her proposals are specific:

• Equity stakes. Where the state funds early, it should be able to hold an upside position rather than only absorbing the downside.
• Retained royalties. Public research bodies that license discoveries should keep a meaningful share of what those discoveries earn.
• Conditions on public money. Grants, guarantees and bailouts should carry obligations — on pricing, on reinvestment, on emissions, on limits to share buybacks — rather than arriving unconditional.
• Reinvesting the returns. Whatever comes back funds the next round of risk, so the innovation fund replenishes itself instead of depending on the annual budget fight.

Underneath all four is a single conviction: that the problem is not the quantity of public money but its structure. Money that flows out with no claim attached produces exactly the outcome we observe — public risk, private reward — and then that outcome is cited as proof the public sector is bad at this.

Where the critics have a case

I do not think you can take Mazzucato seriously without taking her critics seriously, and the criticism is not merely ideological reflex.

The strongest objection, made carefully by Alberto Mingardi and echoed by Deirdre McCloskey, is about causation. Much of the public funding she cites was defence spending, aimed at military objectives, which happened to spill over into civilian use. Calling that mission-oriented innovation policy reads intent backwards into what was often luck. If the payoff came as an unintended spillover, that is an argument for funding basic research broadly — not for the state steering technology toward chosen destinations.

The second objection is selection. If you assemble the technologies that worked and trace them back, you will find public funding in the ancestry of nearly everything, because public funding is in the ancestry of nearly everything. The programmes that produced nothing do not get a chapter. Critics point to a broad empirical literature on industrial policy whose results are considerably more mixed than her cases suggest.

Third: picking winners is genuinely hard, and failure is not costless. States face pressure from incumbents, electoral cycles and their own procurement habits. An equity-stake model concentrates the risk of political capture precisely where the money is. It is fair to ask whether the institution she is proposing to empower is the institution that actually exists.

The honest position, I think, is that she has proved the reward side of her thesis beyond argument, and made the risk side plausible without fully closing it.

What survives all three objections is the distributional half. Even if the causal story is looser than she claims — even if spillovers explain more than direction — the public still bore enormous early cost and retained almost no claim on the result. You can dispute why the state was there. You cannot really dispute that it was there and left empty-handed.

Why this argument keeps mattering

Mazzucato is describing one mechanism inside a much larger pattern: the steady movement of things built collectively into private ownership, each step defensible on its own terms, the cumulative result something nobody voted for. That pattern is what I trace in how technology gets captured, and it does not depend on anyone behaving badly — it works best when everyone behaves normally, which is also what makes it so hard to see while it is happening, a dynamic I take apart in the composite man.

She belongs, too, in a small group of economists arguing that the direction of technology is a political choice rather than a natural process — a case Acemoglu on power and progress makes from a different angle and with different evidence. Where Acemoglu asks who captures the gains once a technology exists, Mazzucato asks who paid for it to exist at all. They are the same question, taken from opposite ends.

The practical value of reading her is that she makes an invisible thing visible. Once you have seen the public investment sitting underneath a product you were told was a private miracle, you cannot stop seeing it — and the question of what the public got back stops being an ideological question and becomes an accounting one. That is a more useful place to argue from than where most of us were standing before she wrote.