The French state grants companies nearly eight billion euros annually in tax breaks for research—and yet only 3.9% of work carried out in public laboratories is funded by industrial contracts. In Germany, that figure exceeds 12%. It is not the volume of public money that is lacking. It is the channel through which it flows.

Three Decades of Public Funding Without Industrial Demand

France devotes 0.72% of its GDP to public research: on par with Germany (0.92%) and the United Kingdom (0.78%), and above the OECD average [1]. That is hardly negligible. But behind this overall figure lies an architecture profoundly different from that of its neighbors.

In France, 58% of public research spending is oriented toward fundamental research. In the United Kingdom, Finland, and the Netherlands, more than 40% goes to applied research [2]. This allocation is not happenstance: it reflects a postwar model in which large national bodies (CNRS, CEA, INSERM) absorbed the majority of resources, driven by a logic of autonomous scientific production, weakly connected to industrial needs. The share of public research executed by national agencies represents 51% of the state sector’s spending, compared to 43% for universities—whereas in Germany, universities execute 55% of that same spending [3].

The result: French companies place few research orders with public laboratories. In 2019, the most recent year for which detailed comparative data are available, German industrialists committed over 4 billion euros in contracts with public research—against a structurally lower level in France, where the share of public research funded by companies stood at 3.9% of spending [4]. The gap has not closed.

The Transfer Circuit: From Hope to Trimmed Budget

To bridge this gap, France has built since 2010 a comprehensive apparatus, financed by future investment programs. The thirteen SATT (societies whose mission is to move laboratory inventions to companies) received 856 million euros in public funding. As of January 1, 2026, their cumulative record since creation stood at 4,650 patents filed, 2,357 licenses signed with companies, and 987 startups created [5]. The startups supported have raised 4.4 billion euros, with their economic valuation estimated at over 6 billion [5].

These figures are real. But they merit contextualization. In 2024, the coverage rate of SATT total expenses by their own revenue reached only 50%, when the initial target was 39%: signaling improvement, but also that financial balance without subsidy remains out of reach [6]. The Senate had flagged this as early as 2017: results remained globally below expectations, with financial returns that business plans could not deliver [7].

More telling still: in 2025, SATT funding shifts from future investment programs to the Research Programming Act, with a budget slashed from 80 to 45 million euros [6]. Project freezes, reduction in the number of support cases. The minister signaled his intention to hand control to territorial actors [6]. This is not a signal of the apparatus’s maturity: it is a budget cut dressed up as strategy.

Research technology institutes (IRT) and institutes for energy transition (ITE)—fifteen structures in total, anchored in industry-research partnerships—received 450 million euros for 2020-2024. Their record over fourteen years is solid: 1,900 employees hired, over 1,220 doctoral students trained, over 700 technologies transferred to industry [8]. But their project-based funding has never been secured beyond 2024: the government’s response to the Senate in September 2026 confirms the end of the multi-year modality, with no numerical commitment for the future [8].

Carnot institutes form the third tier of the architecture. With 116 million euros in public funding in 2024, they represent barely 0.7% of public research financing, yet concentrate 55% of R&D outsourced by companies to laboratories [9]. Their contract revenues grew from 425 million to 595 million euros between 2016 and 2023—a 42% increase over seven years [10]. But the evaluation conducted by the Inspection Générale des Finances task force in December 2024 is clear: “the leverage effect of the apparatus on company performance could not be demonstrated statistically,” and its impact on competitiveness remains “difficult to evaluate, in the absence of reliable and standardized data” [9].

Money Flows; Connection Does Not Establish

The problem is not the absence of intermediary institutions. France has built more than most of its neighbors. It lies in the funding architecture that runs through them.

Everything transiting through SATT, IRT, and Carnot is financed from state envelopes with short horizons, renewed by tranche, with no anchor in industrial actors’ revenues. Industrialists collaborating benefit from the service without bearing maturation risk. Direct industrial contracting—the kind that structures transfer in Germany, where companies themselves commit 4 billion euros to public research—remains marginal in France. This is not a gap of will: it is the logical consequence of a research tax credit of nearly eight billion euros distributed “horizontally” in the state’s own terms [11]—that is, with no obligation of reciprocal cooperation with public research.

Philippe Aghion, whose work on creative destruction underpins growth-through-innovation theory and who received the Nobel Prize in Economics in 2025, has documented for thirty years that firms close to the technological frontier innovate more intensely when exposed to competition, and that rapid technology diffusion between research and industry is one of the pillars of effective innovation policy [12]. Daron Acemoglu and Simon Johnson, awarded the Nobel Prize in 2024, show that institutions determine who captures gains from technological progress, and that when institutions distribute resources without constraining reciprocity, gains concentrate without spreading [13]. Both diagnoses converge on the same French blind spot: public financing of transfer is not conditioned on industrial co-decision over what is produced.

Nicolas Dufourcq, chief executive of Bpifrance, named the problem in 2020: “The translation of French science coming out of universities into lasting enterprises is poor” [14]. Olivier Lluansi, in his work on reindustrialization, emphasizes that funds are directed toward breakthrough technologies at the expense of modernizing the existing industrial base, which concentrates two-thirds of real potential [15].

Transferring Funding Decision to Those Who Bear the Risk

The lever that reshapes the architecture holds in a single line: condition access to public financing of technology maturation on binding industrial co-investment, and transfer to public-private partnerships the right to decide transfer priorities—not to state operators alone.

Concretely, this means replacing administered grants by tranche (the PIA model) with a mechanism where IRT, SATT, and Carnot receive public funding proportional to industrial contracts they have actually generated—not to their business plan intentions. The Carnot model already moves in this direction, but with too narrow a base (116 million euros, 0.7% of public research financing) and no capacity to influence laboratories’ thematic directions.

The next step is to go further: give industrial consortiums a say in applied research priorities of public institutions (not fundamental research, which must remain free) in exchange for their multi-year financial commitment. This is not privatization of research: it is building a reciprocal contract between public money and industrial agenda. It is what gave strength to Germany’s Fraunhofer-Gesellschaft, which generates over 3 billion euros in research contracts annually, roughly 70% financed directly by industry, with mixed governance.

After this shift, public budget no longer finances an intermediary between research and industry: it finances the portion of work that industry cannot yet profitably do alone, on condition that industry co-decides which technologies deserve to be matured. Public money no longer finances a stock of patents; it finances real demand.

Sources

[1] OECD, comparative data, cited by FIPECO, François Ecalle, “Public Spending on Research and Innovation,” updated 2025, https://www.fipeco.fr/fiche/Les-dépenses-publiques-en-faveur-de-la-recherche (consulted 10/10/2026).

[2] DG Trésor, “French Public Research in International Comparison,” working note, https://www.tresor.economie.gouv.fr (consulted 10/10/2026).

[3] MESRE-DGESIP/DGRI-SIES, “State of Higher Education, Research and Innovation in France No. 19” (EESR19), sheet 44, 2025, https://publication.enseignementsup-recherche.gouv.fr/eesr/FR/T027 (consulted 10/10/2026).

[4] La Fabrique de l’industrie, “Research and Innovation: How to Bring Public and Private Spheres Together?”, 2021, https://www.la-fabrique.fr/fr/publication/recherche-et-innovation-comment-rapprocher-spheres-publique-et-privee-2/ (consulted 10/10/2026).

[5] SATT Network, “SATT Results as of January 1, 2026,” January 2026, https://www.satt.fr/resultats-satt-1er-janvier-2026/ (consulted 10/10/2026).

[6] Alcimed, “SATT in 2025: Results, Funding and Changing Governance,” 2025, https://www.alcimed.com/fr/insights/bilan-satt/ (consulted 10/10/2026).

[7] Senate, information report by M. Philippe Adnot on behalf of the Finance Committee, “SATT: Structures for Capitalizing Public Research That Still Must Prove Their Concept,” July 2017, https://www.senat.fr/rap/r16-683/r16-683_mono.html (consulted 10/10/2026).

[8] Senate, question and ministerial response from Mme Berthet, “Sustainability of IRT/ITE Financing Within France 2030,” published 02/04/2026, response 24/09/2026, https://www.senat.fr/questions/base/2026/qSEQ260408199.html (consulted 10/10/2026).

[9] Inspection Générale des Finances / Conseil Général de l’Économie, “Evaluation of the Carnot Institutes Apparatus,” December 2024, https://www.igf.finances.gouv.fr/files/live/sites/igf/files/contributed/Rapports%20de%20mission/2025/2024-M-047-03%20Rapport%20Instituts%20Carnot%20-%20web.pdf (consulted 10/10/2026); https://www.economie.gouv.fr/cge/instituts-carnot (consulted 10/10/2026).

[10] AEF info / AiCarnot, “First Barometer of Partnership Research,” October 2024, https://www.aefinfo.fr/depeche/719262 (consulted 10/10/2026).

[11] MESRE-DGESIP/DGRI-SIES, EESR19, sheet 50 (Research Tax Credit), 2025, https://publication.enseignementsup-recherche.gouv.fr/eesr/FR/T268 (consulted 10/10/2026); DG Budget, program 172, PLF 2023, https://www.budget.gouv.fr (consulted 10/10/2026).

[12] Philippe Aghion, Peter Howitt, Céline Antonin, Simon Bunel, The Power of Creative Destruction, Odile Jacob, 2021; lecture at the Collège de France, “Rethinking Capitalism: The Power of Creative Destruction,” October 15, 2024, https://www.college-de-france.fr (consulted 10/10/2026).

[13] Daron Acemoglu, Simon Johnson, “Nobel Lecture: Institutions, Technology, and Prosperity,” American Economic Review 115(6):1709-1748, June 2025, https://www.aeaweb.org/articles?id=10.1257/aer.115.6.1709 (consulted 10/10/2026).

[14] Nicolas Dufourcq, cited by AEF info, “Transfer: Nicolas Dufourcq (Bpifrance) Hopes LPPR Will Put France at Same Level as Israel,” February 6, 2020, https://www.aefinfo.fr/depeche/621726 (consulted 10/10/2026).

[15] Olivier Lluansi, Reindustrialize to Live Better in France: Act for All and on All Territories, 2025; remarks cited in Big Média / Bpifrance Le Lab, 2024, https://bigmedia.bpifrance.fr (consulted 10/10/2026).