Mariana Mazzucato on the Entrepreneurial State

The Mazzucato Thesis

Mariana Mazzucato's The Entrepreneurial State (2013) argued that nearly every fundamental component of the iPhone — touchscreen, GPS, internet, voice recognition, lithium-ion battery, the underlying silicon manufacturing capability — originated in publicly funded research. The implication is that the standard narrative of Silicon Valley as a private-innovation success story substantially understates the public foundation on which it was built.

The thesis was politically inconvenient for both progressives (who had treated Silicon Valley as a libertarian counterexample to the "government creates value" frame) and conservatives (who had treated the tech sector as evidence that government should stay out of innovation policy). The empirical record, Mazzucato argued, supported neither narrative.

The Historical Record

The major public-research foundations of modern technology:

The internet: DARPA's ARPANET (1969 onward), the funding of TCP/IP development, the National Science Foundation's NSFNET backbone, and the public funding of the original web protocols at CERN.

GPS: a US Department of Defense system, originally built for military navigation, made available for civilian use in 1983 after the Korean Airlines 007 incident.

Touchscreen technology: developed at CERN in the 1970s, extended through DARPA and DOD-funded research at multiple universities through the 1990s and 2000s.

Voice recognition: DARPA funded the foundational research from the 1970s through the 1990s, including the work that eventually became Siri.

The lithium-ion battery: foundational research by John Goodenough at Oxford and University of Texas, funded substantially by US DOE grants. Goodenough's 2019 Nobel Prize recognized the public-funded origin of what is now a $50+ billion market.

HTML and the web: developed at CERN by Tim Berners-Lee, who explicitly placed the protocols in the public domain.

The economic logic is that fundamental research has positive externalities — the discoveries benefit many actors beyond the researcher — and private firms therefore underinvest. Public funding captures the social return that private funding cannot. The empirical return on US public R&D investment has been estimated at multiples of the private return on comparable spending.

What Private Firms Did Add

The mistake in some readings of Mazzucato is to conclude that private firms add nothing. The actual claim is narrower: the fundamental science is publicly funded, but the productization, distribution, commercialization, and ongoing improvement are largely private. Apple did not invent the touchscreen, but Apple's iPhone team integrated the components, designed the user experience, and built the supply chain that turned the underlying technology into a mass-market product. Private firms add substantial value; they also build on a foundation they did not pay for.

The political-economy question is whether the public capture of the resulting value is adequate. If public funding produces the foundational technology and private firms capture nearly all the returns, the political-economy support for continued public funding erodes. The Bayh-Dole Act (1980), which gave universities ownership of patents arising from federally funded research, is the most contested instance of this question — it accelerated technology transfer but reduced the public return on public investment.

The DARPA Model

The Defense Advanced Research Projects Agency, founded after Sputnik in 1958, is the most-studied model for productive public research funding. Key features:

• Program managers with substantial autonomy to fund projects of their own choosing.
• Short program durations (typically 3-5 years) with explicit deliverables.
• Mix of academic and industry recipients selected for capability rather than political distribution.
• Acceptance of high failure rates as the cost of high-risk funding.
• Mission-driven framing tied to specific military or strategic needs.

The DARPA model has been adapted in other domains — ARPA-E for energy, IARPA for intelligence, and the proposed ARPA-H for health. The empirical record on whether the model travels well outside its original domain is mixed but encouraging.

The NIH Comparison

The National Institutes of Health is the other large public-research funder with a substantial track record. The structure is different — peer review, longer grant durations, less mission-driven — but the outcomes have been substantial. Most modern pharmaceutical innovation relies on basic biological research funded through NIH. The infrastructure of clinical trials, biomedical equipment, and trained researchers depends substantially on NIH support.

Bhaven Sampat's work has documented that roughly 80% of recently approved drugs trace foundational research to NIH funding, even when the drugs were ultimately developed and commercialized by private firms. The pattern is similar to the iPhone analysis but extended across an entire industry.

The Counterargument

Critics of the Mazzucato framework argue that public R&D has substantial waste, captures rent through bureaucratic processes, and crowds out private investment that would otherwise have occurred. The strongest version of the argument is that the successes are post-hoc rationalizations — DARPA produced some successes among many failures, and counting only the successes overstates the program's return.

The counter-counterargument is that the failures are integral to the model — high-risk research requires accepting that most projects will fail, and the expected return on the portfolio justifies the funding even when most individual grants do not produce visible outcomes. The empirical comparison of US public-R&D returns against the returns from comparable private R&D supports the public-R&D side of the argument by substantial margins.

The Honest Reading

Mazzucato's thesis is empirically robust and politically inconvenient for both major US political parties. Public funding produces foundational technology that private firms then commercialize; the standard narratives of either pure-public or pure-private innovation are both wrong. The policy implications — sustained public R&D funding, better capture of resulting value through equity stakes or royalty arrangements, deliberate technology-transfer architecture — have been adopted only partially. The current generation of US industrial policy (CHIPS, IRA, ARPA-H) substantially follows the Mazzucato framework, which makes the next decade an empirical test of whether the approach can produce comparable returns at a much larger scale.

Reading List

• Mariana Mazzucato, The Entrepreneurial State (2013)
• Mariana Mazzucato, The Value of Everything (2018)
• Mariana Mazzucato, Mission Economy (2021)
• Margaret O'Mara, The Code (2019)
• Erik Hurst on US R&D returns (various NBER papers)