OECD Science, Technology and Innovation Outlook 2025

Foreword

The OECD Science, Technology and Innovation Outlook 2025 reviews key trends in science, technology and innovation (STI) policy in OECD countries and major partner economies. This edition comes at a time of accelerating technological change, intensifying geopolitical tensions and urgent demands for transformative responses to economic and societal challenges.

The Outlook shows the importance of improving the effectiveness and efficiency of STI policies as they aim to tackle broad goals and multiple priorities in a context of growing resource constraints. STI policies should leverage synergies among goals, deploying complementary policy measures, promoting cross-government cooperation, and fostering public-private funding models.

A central theme of this edition is how countries can reconfigure scientific cooperation in an increasingly fragmented geopolitical landscape, ensuring the openness that drives scientific advances while simultaneously protecting economic security concerns. Another is how science systems themselves must adapt – with new institutional arrangements, skills and incentives, if they are to contribute effectively to transformative change through more multi-disciplinary approaches.

The Outlook also explores the growing convergence of technologies such as artificial intelligence, biotechnology and quantum computing, which are reshaping innovation processes and demand novel types of policy support. It highlights the potential of more granular approaches that better appreciate industrial structures and assess the impact of policy interventions to mobilise diverse actors around shared missions. It also shows how governments can strengthen their capacity for foresight, policy experimentation and strategic intelligence to remain agile in the face of uncertainty.

Taken together, these insights underline that STI policy is at a turning point. The ability of governments to mobilise science, technology and innovation for transformative change, while navigating geopolitical pressures and rapid technological shifts, will be decisive in shaping the future.

Acknowledgements

The 2025 edition of the OECD Science, Technology and Innovation Outlook was prepared under the aegis of the OECD Committee for Scientific and Technological Policy (CSTP), with input from its working parties. The 2025 edition is a collective effort co-ordinated by the Science and Technology Policy Division of the OECD Directorate for Science, Technology and Innovation. Michael Keenan served as overall co-ordinator and Sylvain Fraccola as publication and communication co-ordinator. Blandine Serve co-ordinated statistical inputs. The publication was produced under the guidance of Alessandra Colecchia, Head of the Science and Technology Policy Division.

Chapter 1, “Mobilising science, technology and innovation policies for transformative change”, was prepared by Michael Keenan, Jessica Ambler, Mario Cervantes and Blandine Serve, with contributions by Philippe Larrue and Charles McIvor (OECD Directorate for Science, Technology and Innovation). The chapter benefited from review and comments by Simon Bennett (International Energy Agency).

Chapter 2, “Reconfiguring scientific co-operation in a changing geopolitical environment”, was prepared by Michael Keenan, Yoran Beldengrun, Carthage Smith and Blandine Serve, with contributions by Alan Paic and Hyunkyeong Yun (OECD Directorate for Science, Technology and Innovation). The chapter benefited from review and comments by Joachim Pohl (OECD Directorate for Financial and Enterprise Affairs). The authors are grateful to the presenters at two CSTP workshops related to the chapter’s topic, and in particular to Andrew James (University of Manchester), who provided written inputs and comments on the chapter.

Chapter 3, “Expanding the benefits of STI investments”, was prepared by Caroline Paunov, Sandra Planes-Satorra and Luke Mackle (OECD Directorate for Science, Technology and Innovation). It builds on work carried out by the Working Party on Innovation and Technology Policy. The chapter benefited from review and comments from Andrew Paterson (OECD Centre for Entrepreneurship, SMEs, Regions and Cities).

Chapter 4, “How science systems need to adapt to support transformative change”, was prepared by Carthage Smith, Andrea-Rosalinde Hofer, Frederic Sgard, Masatoshi Shimosuka and Gemma Volpicelli (OECD Directorate for Science, Technology and Innovation). It builds on work carried out by the OECD Global Science Forum.

Chapter 5, “Technology convergence: Trends, prospects and policies”, was prepared by David Winickoff, Claire Jolly, Alistair Nolan, Douglas Robinson and Marit Undseth (OECD Directorate for Science, Technology and Innovation), with contributions by Daniel Nadal. The chapter draws on recent work by the Working Party on Biotechnology, Nanotechnology and Converging Technologies and the OECD Space Forum.

Chapter 6, “An ecosystems approach to industrial policy”, was prepared by Damiano Morando, under the supervision and guidance of Antoine Dechezleprêtre and Guy Lalanne (Productivity, Innovation and Entrepreneurship Division of the OECD Directorate for Science, Technology and Innovation). The authors are grateful to Hélène Dernis for her support in preparing some of the key figures included in the chapter, to Charles McIvor for valuable discussions, and to the authors of the ecosystem papers on which this chapter draws: Antoine Dechezleprêtre, Hélène Dernis, Luis Diaz, Milenko Fadic, Guy Lalanne, Francesco Losma, Sara Romaniega Sancho and Lea Samek.

Chapter 7, “Tools for agility: Actionable strategic intelligence and policy experimentation”, was prepared by Caroline Paunov, Douglas Robinson, Sandra Planes-Satorra and Isabella López Trejos (OECD Directorate for Science, Technology and Innovation). It builds on work carried out on strategic intelligence by the Working Party on Biotechnology, Nanotechnology and Converging Technologies and on policy experimentation by the Working Party on Innovation and Technology Policy.

All chapters were reviewed by Alessandra Colecchia, Jens Lundsgaard and Jerry Sheehan (OECD Directorate for Science, Technology and Innovation).

The Outlook also benefited from comments by CSTP delegates and from discussions at the 125th and 126th Sessions of the CSTP in November 2024 and April 2025, respectively.

Thanks are also due to Silvia Appelt, Leonidas Aristodemou, Brigitte van Beuzekom, Hélène Dernis, Fernando Galindo-Rueda, Petra Kelly, Guillaume Kpodar, Kuniko Matsumoto, Laurent Moussiegt and Fabien Verger (OECD Directorate for Science, Technology and Innovation) for their helpful advice and statistical inputs.

The authors are grateful to Emily Acas and Kyriakos Vogiatzis (OECD Directorate for Science, Technology and Innovation) for their secretarial assistance and to Sebastian Ordelheide (OECD Directorate for Science, Technology and Innovation) for communications support. Special thanks are extended to Jennifer Allain for editorial contributions and bibliographic research.

Abbreviations and acronyms

Executive summary

Driving Change in a Shifting Landscape

Global challenges, rising economic security concerns, and disruptive emerging and converging technologies signify a new context for STI policy. Ensuring that STI policy remains fit-for-purpose in this new and rapidly changing environment requires fundamental structural reforms that can improve the effectiveness and efficiency of policy interventions, as well as continued attention to enhancing the evidence base. The OECD Science, Technology and Innovation Outlook 2025 analyses the shifting landscape and its implications for STI systems, providing recommendations across seven chapters that lay out STI policy reforms needed to drive ambitious change.

Leveraging policy complementarities to boost efficiency

Ambitious policy agendas, together with growing resource constraints, highlight the importance of enhancing STI policy efficiencies. With annual government allocations for R&D falling by 1.9% in 2024 in the OECD area, STI policies need to intentionally leverage synergies and mitigate trade-offs between different policy priorities so that STI support to national competitiveness, for example, can also contribute to sustainability transitions. Governments also need to balance and exploit synergies between their direct and indirect (e.g., tax incentives) support measures for R&D, since both can help accelerate transformative change in complementary ways. Co-ordination between STI and non-STI policy areas should also be strengthened.

Making research security proportionate, precise and shaped with partners

Rising geopolitical tensions and strategic competition in emerging technologies are contributing to a growing securitisation of STI that is reconfiguring international STI collaborations. Public research systems are increasingly affected as governments seek to simultaneously: promote advanced capabilities and strategic autonomy in critical technology fields; protect sensitive knowledge through research security measures; and project national interests through selective partnerships and science diplomacy. Protecting sensitive research or academic collaborations can be done in ways that do not compromise research quality, undermine innovation and fragment co-operation on shared global challenges. To do so, research security policies must be proportionate, precise and developed in close partnership with scientists, businesses and other parts of government.

Broadening benefits through enhanced diffusion

Innovation activities typically cluster among leading firms, sectors, and regions due to economies of scale and knowledge spillovers. Such clustering can lead to concentration of economic and societal benefits in limited geographic areas. To broaden the impact, STI policies need to place greater emphasis on policies and investments to promote diffusion and to translate innovations into economy-wide productivity gains and societal benefits. Widening participation in innovation is a key lever for expanding its benefits, since it can enhance both the quality and societal relevance of technological development. Frontier-oriented STI policies should also consider how diffusion and adoption policies can be integrated into development efforts pushing at the technological boundary.

Adapting public science systems

Structural reforms are also needed to enable national science system to better respond to the changing policy context and enhance their contributions to major societal challenges. Key to such reforms is enabling and valuing multidisciplinary research that can generate solutions to complex socio-economic challenges that cut across disciplines and sectors. Reforms are also needed to develop a variety of transparent career-paths that recognise and enable mobility between academia and other sectors. Research infrastructures need more flexible support and governance mechanisms to enable them to operate together to address shared goals and promote transformative change. Academic research also needs to embrace greater direct engagement with society through improved communication measures and citizen science programmes. To ensure these structural reforms take root, performance assessment and incentive structures need to better recognise the variety of contributions to, and outputs from, science that are necessary to promote innovation. At the same time, governments should continue to ensure the freedom and autonomy of research, advance open science, and promote public trust in science.

Harnessing technology convergence

Promoting multi-disciplinary and cross-sectoral research becomes even more important as the convergence of technologies drives forward innovation. Four important technology areas – synthetic biology, neurotechnology, quantum technologies and earth observation from space – illustrate these processes. For example, artificial intelligence is enabling protein design to create molecules with novel properties with the potential to enable personalised therapies, while its convergence with immersive technologies offers opportunities to treat mental illness. Convergence is a process of integration involving different disciplines and communities. Governments can enhance convergence by supporting “convergence spaces”, which are physical, digital and technological infrastructures and platforms that can foster deep forms of interdisciplinary research, engineering and innovation.

Adopting an ecosystems approach

Adopting an industrial ecosystem perspective that goes beyond sectoral boundaries to consider both upstream and downstream industries can contribute to designing more effective industrial policies. It can also help governments to identify the full range of relevant stakeholders, including firms, start-ups, workers, investors, suppliers and trade partners, to design policies that better reflect the true complexity of the industrial landscape. Using the approach, however, entails developing a robust data infrastructure that brings together granular data from multiple sources to capture the ecosystem’s complexity.

Boosting policy agility through strategic intelligence and experimentation

To drive change in a shifting landscape, STI policymaking must be increasingly anticipatory and agile under conditions of high uncertainty. Practices such as strategic foresight, technology assessment, and policy evaluation can provide timely insights through anticipatory and real-time evidence production, while policy experimentation can enable testing of new ideas and critical evaluation of policy impacts. Together, these approaches support evidence-based policymaking and boost policy agility. Fostering their use among policymakers requires embedding them in national programmes and frameworks, increasing flexibility and adaptability within bureaucratic structures, and investing in training programmes for public sector officials. Ensuring there is a clear pathway for scaling up interventions that prove successful or phase down those that fail is also key.

Through these reforms, STI systems can help drive ambitious change

These policy reforms will strengthen national innovation systems, helping them drive change that responds to the shifting policy landscape and tackles future challenges.

1. Mobilising science, technology and innovation policies for transformative change

Abstract

Science, technology and innovation (STI) plays a prominent role in promoting greater economic competitiveness, resilience and security, and sustainability. To realise their potential, STI systems need to be reformed to generate and deploy relevant knowledge, technologies and innovation at an unprecedented pace and scale. This chapter proposes five key actions STI policymakers can take: promote a policy agenda that contributes to broad transformative change; balance direct and indirect support to research and development; strengthen co-ordination between STI policies and non-STI areas; mobilise public funding to crowd-in private finance; and promote transformative change that goes beyond “business-as-usual” outcomes. The chapter emphasises the need for governments to experiment with and adopt innovative policy mechanisms and tools, and to better appreciate and leverage innovation dynamics to accelerate transformative change.

Key messages

• Global challenges are placing increasing pressure on governments, firms and society more broadly to rethink how our economies and societies can better operate. There is a growing need for transformative change that promotes economic competitiveness, resilience and security, and sustainability transitions.

• Science, technology and innovation (STI) systems are expected to play a prominent role in advancing transformative change. They need to be reformed to generate and deploy relevant knowledge, technologies and innovations at an unprecedented pace and scale, under conditions of uncertainty and complexity. Many of these reforms are already well-known within the STI policy community yet pose significant implementation challenges.

• Governments should consider a range of policy actions when reforming their STI policy mix to better contribute to transformative change agendas. First, STI policy agendas should intentionally leverage synergies and mitigate trade-offs between a range of policy priorities that contribute to broad transformative change. For example, policy support for national competitiveness can also contribute to resilience and security as well as sustainability transitions, if designed appropriately.

• Second, policymakers should strike an appropriate balance and exploit synergies between direct and indirect support measures for research and development (R&D) to promote transformative change. While direct measures can support more ambitious R&D and technological breakthroughs, non-directed measures encourage R&D activities with near-market potential that can help accelerate transformative change.

• Third, governments should strengthen co-ordination between STI and non-STI policy areas in pursuing transformative change. The fragmentation of state structures can hinder governments’ ability to deliver the needed cross-cutting priorities and interventions to foster transformative change. Governments should continue to experiment with novel policy instruments, such as challenge-based funding and mission-oriented innovation policies, to bring together multiple actors to co-create and collaborate across innovation chains on transformative pathways.

• Fourth, governments should mobilise public funding to crowd-in private finance for transformative change. Several capital market failures discourage the allocation of private investment into technologies that promote transformative change. Governments should continue to experiment with instruments like blended finance to deploy public financial resources to leverage or attract private capital.

• Finally, governments should seek to promote transformative change rather than “business-as-usual” outcomes. To help steward fundamental, radical and possibly rapid changes, they must appreciate and embrace the nature of transformative change and how it differs from and relates to incremental change. STI policymakers should identify “leverage points” for interventions that can trigger and accelerate the sorts of system-wide changes needed for transformations.

• These five policy actions cover issues that have preoccupied STI policymakers in one form or another for several decades and in this sense are not unique to the pursuit of transformative change. However, the urgent need for transformative change means reforms like these should be implemented quickly if STI is to remain relevant and contribute to future economic and societal advancement.

Introduction

Growing geopolitical tensions, the accelerating climate crisis, biodiversity loss, rising inequality and rapid technological change: these and other challenges are placing increasing pressure on governments, firms and society more broadly to rethink how our economies and societies can better operate for the greater good. There is growing recognition of the need for transformative change, in which STI is expected to play a prominent role. To fulfil this promise, however, STI systems need to be reformed to generate and deploy relevant knowledge, technologies and innovation at an unprecedented pace and scale. This needs to be done in conjunction with reforms in other systems, including energy, health, agriculture and industrial production, where success will also depend on a range of framework conditions, including finance, skills and regulations.

Many of the necessary reforms are well-known within the STI policy community yet pose significant implementation challenges. In response, the OECD Committee for Scientific and Technological Policy has developed the Agenda for Transformative Science, Technology and Innovation Policy to provide high-level guidance to policymakers on their STI policy reforms (OECD, 2024[1]).1 While the Transformative Agenda’s framework can be applied to any transformative goals, it highlights three that capture many contemporary STI policy concerns:

Promoting economic competitiveness that is fair and inclusive. Many OECD Member countries’ STI policies place renewed emphasis on productivity growth and international competitiveness. At the same time, income inequality has a sizeable and statistically significant impact on growth and is a key strategic consideration for economic development and societal outcomes.

Fostering resilience and security against risks and uncertainties posed by the growing emergence of systemic threats. Abrupt shocks, such as the COVID-19 pandemic, demonstrate the importance of resilience to anticipate, absorb, recover from and adapt to disruptive change. On the security side, rising strategic competition between countries in critical technologies and resources that underpin economic competitiveness and national security have led governments to increasingly pursue greater strategic autonomy.

Advancing sustainability transitions that mitigate and adapt to a legacy of unsustainable development from climate change, pollution and biodiversity loss. Advancing sustainability calls for accelerated transitions in specific industries, technologies, and established models of production and consumption.

Sustainability transitions have been a prominent feature in most national STI strategies for the last decade, though there are signs this may now be changing. The evolving geopolitical context has brought growing attention to national security through means of strategic autonomy and technological sovereignty (see Chapter 2), while economic competitiveness is again emerging as the pre-eminent concern of research and innovation policy. Since most OECD Member countries are interested in pursuing these goals simultaneously, this chapter considers whether they imply trade-offs or can be complementary and synergistic. It proposes that STI policy can support the transformative change needed to achieve a range of goals by intentionally leveraging synergies and mitigating trade-offs between them.

Much of the chapter focuses on the funding and financing of STI, a core concern for policymakers. Transformative change calls for greater directionality in STI systems, including in their allocation of resources. The chapter therefore explores data on R&D funding and governments’ research priorities. It also considers the policy instruments governments use to direct R&D expenditures towards the chosen priorities. These measures are part of a broader policy portfolio that also provides non-directed support, e.g. through R&D tax incentives to firms. Governments face challenges to balance this portfolio and promote synergies between different measures, particularly under conditions of uncertainty and complexity that demand agility and diversity. The chapter describes a simple schema for mapping policy portfolios along the innovation chain and according to their degree of directedness. It also highlights examples of selected policy innovations by several governments to foster more responsive R&D, more breakthrough research and innovation, and more integrated policy support across the innovation chain.

The chapter also considers measures to better co-ordinate STI policy with non-STI policy areas to promote transformative change. The fragmentation of state structures can hinder governments’ ability to deliver the sorts of cross-cutting priorities and interventions that are needed. Cross-government co-ordination is especially important, since market and structural conditions, such as regulations and standards, should be aligned to facilitate technology diffusion and phase-out, while the substantial scope of investments needed to facilitate transformations will necessitate buy-in from across government to co-invest in and co-manage coherent portfolios of activities. The chapter outlines how governments are experimenting with novel policy instruments, such as challenge-based funding and mission-oriented innovation policies (MOIPs), to bring together multiple actors, including from different policy domains, to co-create and collaborate across innovation chains on transformative pathways.

Since firms account for around two-thirds of R&D expenditures across the OECD and the private sector is the main source of R&D funding, they have an important role to play in promoting transformative change through STI. However, several capital market failures discourage the allocation of private investment to technologies that promote transformative change. Governments can use risk-mitigation tools to help firms cross “valleys of death” at various stages of the innovation chain. These include “blended finance”, with a view to deploying public financial resources to leverage or attract private capital. The chapter argues that governments should continue to experiment with such tools, which have the potential to direct STI finance and help scale-up private investments in research, development and innovation (RDI) and innovation to promote transformative change.

All these issues are broad and long-standing and have preoccupied STI policymakers in one form or another for several decades. In this sense, they are not unique to the pursuit of transformative change. However, since transformative change refers to a radical and permanent qualitative shift in current socio‑economic systems, new policy approaches to steward fundamental, radical and possibly rapid changes are needed. As a starting point, an appreciation of the nature of transformative change – and how it differs from and relates to incremental change – is essential. This chapter proposes that governments map and target multiple innovation system feedback cycles in their policy interventions to accelerate transformative change.

The chapter is structured around five proposed policy “actions” that cover these issues:

• Action 1: Promote a policy agenda that contributes to broad transformative change.

• Action 2: Direct R&D funding for transformations in combination with non-directed measures.

• Action 3: Strengthen co-ordination with non-STI policy areas on transformative change.

• Action 4: Mobilise public funding to crowd-in private finance for transformative change.

• Action 5: Promote transformative change rather than “business-as-usual” outcomes.

This chapter offers a brief overview of each action and provides selected examples of countries’ policies, particularly where these involve innovative approaches that offer lessons to other policymakers.

Action 1: Promote a policy agenda that contributes to broad transformative change

Transformative change calls for ambitious levels of STI investment over a long period

R&D investment is a key driver of growth and a core concern in STI policy. Transformative change calls for ambitious levels of investment over a long period, covering all parts of the innovation chain, from exploratory fundamental research to the deployment and diffusion of tested technologies. These investments are distributed among a variety of different actors within public research and innovation systems as well as private industry. As such, they include public funding for STI from research and innovation ministries and agencies, as well as from sectoral ministries and agencies in areas like energy, transport, agriculture and health. They also cover private financing for STI.

There has been a recent slowdown in R&D expenditure growth in the OECD

While R&D expenditures have increased markedly over the last two decades, there are concerns that debt burdens and inflationary pressures will lead to a slowdown in this growth or even an absolute decline. Recent policy uncertainty and economic activity indicators also signal the potential for rising levels of inflation and a softening of global growth (OECD, 2025[2]). The latest year for which internationally comparable OECD data on R&D expenditures are available is 2023, showing a 2.4% increase in inflation-adjusted terms on the previous year in the OECD, down from 3.6% in 2022. This growth was again driven by the business sector (Figure 1.1), which experienced a 2.7% increase from 2022 to 2023, compared to 2.5% for R&D performed in government sector institutes and 1.7% in the higher education sector. The business sector accordingly accounted for 73.6% of total gross domestic expenditure on R&D (GERD) in the OECD in 2023, up from 66% in 2010. Among the largest spending countries, the share of business-performed R&D increased in the People’s Republic of China (hereafter “China”) from 60% in 2000 to 77.7% in 2023, which is close to the proportion in the United States (78.4%) and higher than that of the EU27 (66.0%).

Figure 1.1. R&D trends by performing sectors in OECD countries, 2007-2023

2007=100

Source: OECD (2025), Main Science and Technology Indicators Database, http://oe.cd/msti (accessed in March 2025).

Statlink

Within the OECD, Israel (6.3%) and Korea (5%) continued to display the highest levels of R&D intensity as a percentage of gross domestic product (GDP) in 2023 (Figure 1.2). R&D intensity in the OECD climbed from 2.3% in 2013 to 2.7% of GDP in 2023. Growth in inflation-adjusted R&D expenditure in the OECD was distributed across several countries but with notable differences among them. In the United States, it stood at 1.7% and in the European Union (EU) at 1.6% in 2023. The European Union’s largest economies slowed the area’s overall growth: Germany’s R&D rose by 0.8%, while France’s fell by 0.5%. In contrast, R&D expenditure in Poland and Spain increased by over 8%. R&D growth in Japan (2.7%) and Korea (3.7%) exceeded the OECD average. At 8.7%, growth in R&D expenditure in China in 2023 surpassed that of the OECD (OECD, 2025[3]).2

Figure 1.2. R&D intensities, selected economies, 2013-2023

As a percentage of GDP

Note: 2023 data correspond to 2022 for the United Kingdom and 2024 for Canada.

Source: OECD (2025), Main Science and Technology Indicators Database, http://oe.cd/msti (accessed in March 2025).

Statlink

What policy goals are governments prioritising in their R&D expenditures?

Since there are always more ideas and prospective projects to fund than there are available resources, setting research priorities and selecting R&D performers have long been recognised as key policy concerns (see, for example, Weinberg (1963[4])). Furthermore, significant proportions of government-funded R&D target specific economic and societal goals, which are subject to priority-setting processes.

Data on government budget allocations for R&D can be usefully broken down to provide insights on the areas being funded by the public sector (Figure 1.3). Data for the OECD show that support has grown most strongly for health objectives (reflecting changing societal expectations on healthy living and ageing) and general advancement of knowledge (reflecting a relative retreat by governments to set research objectives) over the last few decades. However, R&D investments targeting health-related objectives have declined steadily between 2020 and 2024. After reaching USD 97.4 billion in constant purchasing power parities (PPP) in 2020 – at the height of the COVID-19 pandemic – investments fell to USD 86.3 billion in 2024, a decline of 11.5%. By contrast, support for energy R&D (USD 31.9 billion in constant PPP in 2024) and defence R&D (USD 111.17 billion constant PPP in 2024) increased sharply over the same 2020-24 period, by 51% and 17%, respectively, reflecting policy goals to reduce greenhouse gas emissions and enhance national security. There is some variety across OECD Member countries on the relative weight of these areas in their R&D budget portfolios, as shown in Figure 1.4. These reflect, in part, different institutional set-ups and R&D funding arrangements across countries, as well as their sectoral specialisation.

Figure 1.3. Trends and broad spending categories of government R&D budgets, OECD, 1991-2024

1991 = 100

Note: GUF: general university funds; GBARD: government budget allocations for R&D.

Source: OECD calculations based on OECD (2025), Main Science and Technology Indicators Database, https://oe.cd/msti (accessed on 17 October 2025).

Statlink

Figure 1.4. R&D budget by broad spending categories, selected economies, 2024

As a percentage of total government budget allocations for R&D

Notes: GUF: general university funds. Public GUF refer to the R&D funding share from the general grant that universities receive from the central (federal) Ministry of Education or corresponding provincial (state) or local (municipal) authorities in support of their overall research/teaching activities. General advancement of knowledge (financed from sources other than GUF) is R&D funding from general grants that cannot be attributed to an objective and are financed by sources other than GUF. 2024 data corresponds to 2023 for Chile, Israel, Korea United Kingdom and 2022 for Canada.

Source: OECD calculations based on OECD (2025), Research and Development Statistics, https://www.oecd.org/en/data/datasets/research-and-development-statistics.html (accessed on 17 October 2025).

Statlink

Insights on the directionality of public R&D funding can also be gleaned from analysis of the administrative data of research and innovation funding bodies. Focusing on societal goals, an analysis of R&D project funding data in the OECD Fundstat database (version: 2024, May 2025) (Aristodemou et al., forthcoming[5])3 shows that public R&D funding grew across all major goals from 2015 to 2023 (Figure 1.5). Among the societal goals categories used in the analysis, “Prosperity” accounted for the largest amount of R&D funding in 2023, followed by “Health” and “Planet”. In terms of growth over the 2015-2023 period, “Energy” grew 2.3 times, “Prosperity” 2.1 times, “Planet” 1.8 times, “Peace” 1.7 times and “Health” 1.6 times. “Education” saw the lowest growth, at 1.2 times. These patterns highlight that much of government R&D remains focused on promoting economic competitiveness, and that although support to sustainability transitions has risen in recent years, it remains modest in comparison.

Figure 1.5. Estimates of R&D funding to societal goals, 2015-2023

R&D funding awards for 19 OECD countries and EC-EU programmes

Notes: The OECD Fundstat database includes R&D project-level data from 19 OECD countries (Australia, Austria, Belgium, Canada, Czechia, Estonia, Finland, France, Germany, Japan, Ireland, Latvia, Lithuania, Norway, Portugal, Sweden, Switzerland, the United Kingdom and the United States) and the European Commission programmes. For 2021, the data for these 19 countries represent approximately 51% of the total government budget allocations for R&D, excluding general university funds, for these countries as reported in the OECD Main Science and Technology Indicators Database. Sustainable Development Goal (SDG) categories are mutually exclusive with fractional allocations using the SDG classifier on R&D project descriptions (Aristodemou et al., forthcoming[5]). The SDG (https://sdgs.un.org/goals) categories are defined as follows: Prosperity includes SDG 8, SDG 9, SDG 10 and SDG 11; Health includes SDG 1, SDG 2, SDG  and SDG 5; Planet includes SDG 6, SDG 13, SDG 14 and SDG 15; Energy comprises SDG 7; Peace covers SDG 16 and SDG 17; Education corresponds to SDG 4 more closely resembling scholarship-driven research as opposed to research on education; and No relevance projects are without identifiable alignment to any specific SDG. R&D funding award data reflect authorisation rather than actual commitments or expenditure. The prominent peak in 2020 largely results from increased R&D funding related to the COVID-19 pandemic response, along with the inclusion of Japan’s Green Innovation Fund in the database.

Source: OECD analysis of the OECD Fundstat database (v. 2024) (accessed in May 2025).

Statlink

Tensions and synergies in pursuing a range of priorities

While there are strong synergies and interdependencies between the priorities of economic competitiveness, security and sustainability, insular efforts to advance specific goals may compromise others. Some of the relationships between these policy goals are universal while others may be specific to a particular geographic context or sector Table 1.1 provides a comprehensive, albeit likely incomplete, overview of some of the synergies and tensions between sustainability and other policy priorities related to economic competitiveness, inclusive development, and national security and resilience.

Table 1.1. Synergies and tensions between diverse science, technology and innovation policy priorities

This chapter unpacks some of the connections between sustainability transitions and economic competitiveness and inclusion (Chapter 2 discusses connections with security). The chapter first outlines synergies and trade-offs then discusses them in reference to specific policy interventions from a selection of countries to provide examples of how such synergies are leveraged, trade-offs are mitigated or where goals may be in dissonance. In this regard, the analysis considers a selection of the latest STI policy announcements and budgets from four countries – Australia, Canada, Korea and the United Kingdom – as well as from the European Union, as outlined in Table 1.2. These countries were chosen primarily for their global geographical spread.

Table 1.2. Examples of science, technology and innovation policy priorities

Evidence suggests that climate action and economic development strategies are mutually reinforcing. Over the next decade, ambitious targets and policies to reduce greenhouse gas emissions could result in a net gain to global GDP (OECD/UNDP, 2025[8]). Without government intervention, industry-led STI activities tend to focus on optimising the profitability and efficiency of established solutions rather than the emergence of new alternatives (Garsous, Bourny and Smith, 2023[9]; OECD, 2025[10]). However, challenging fiscal positions may require governments to direct support towards priority areas to balance short-term growth and long-term sustainable development (OECD, 2025[2]).

The scale-up of industries and technologies that are less destructive or carbon-intensive, like the use of natural gas and liquefied natural gas as bridge fuels, can provide interim solutions. However, they can also slow or draw resources away from more sustainable alternatives (Meadowcroft, 2011[11]). Australia recognises this reality in its Future Gas Strategy, which is pragmatic about the necessity of gas-powered generation for electricity grid security and reliability while identifying the commercialisation of net zero alternatives as a means to reduce demand (Australian Government, 2024[12]).

In addition, the greening of some industries will be more difficult. This could be the case where low-carbon solutions are far from commercial or competitive, where production or deployment infrastructure requires substantial upfront investment, or where significant disruption of industrial operations would damage economic competitiveness. Many countries provide public funding or incentivise private financing for carbon capture, utilisation and storage (CCUS) technologies and infrastructures, including Canada’s tax credit for CCUS (Government of Canada, 2024[13]) and Korea’s Carbon Capture and Utilisation (CCU) Demonstration Support Center and supports for businesses to apply CCU technologies (Government of Korea, 2024[14]). The United Kingdom also has plans to leverage GBP 8 billion of private investment in CCUS infrastructure (HM House of Commons, 2024[15]).

Several governments are also taking steps to streamline bureaucracy or harmonise regulation to advance sustainability transitions and contribute to economic growth. For instance, Australia, Canada and the European Union have introduced initiatives to cut red tape and accelerate approval processes for clean growth projects (Commonwealth of Australia, 2024[16]; Government of Canada, 2024[13]; European Commission, 2025[6]).

Analysis of the synergies and trade-offs between policy priorities (Table 1.1), paired with policy examples from the countries analysed (Table 1.2), yields a range of policy options available for the design of integrated STI approaches needed to optimise or navigate interdependencies between different goals. These are outlined in Table 1.3 and Table 1.4 and organised loosely according to different phases of transformation and corresponding intervention points (Ghosh et al., 2020[17]; Kanger, Sovacool and Noorkõiv, 2020[18]).

Table 1.3. Policy options to support competitive sustainability transitions

Technology-driven development often co-exists with, or may reinforce, absolute and relative poverty (Chataway, Hanlin and Kaplinsky, 2014[20]). Without appropriate distributive or inclusion measures, contemporary innovation pathways can exclude large segments of the global population as both producers and beneficiaries of change (Planes-Satorra and Paunov, 2017[21]). Inequality is correlated with slower growth and constrained innovation and can impair sustainable development by limiting demand for less competitive net zero solutions (Mazarr, 2022[22]; Ostry, Berg and Tsangarides, 2014[23]). This is a central focus of some of the initiatives outlined in the STI strategy documents.

For instance, Canada is making efforts to improve data on the national clean technology industry to better understand and address the needs of under-represented groups (ISED, 2025[24]). The United Kingdom has embedded sustainability and inclusive development into the Innovation Accelerator programme, which supports the development of innovation clusters to advance the development of greener technologies and helps to address regional income and productivity disparities (DSIT, 2024[25]; UKRI, 2025[26]). Recent evaluations of the EU Bioeconomy Strategy have revealed uneven distribution of activities across EU regions (BioRural, 2024[27]; European Commission, 2022[28]). This has spurred efforts to update the strategy to take advantage of the growth potential of the expanding bioeconomy while reducing reliance on fossil fuels and improving economic outcomes for rural areas.

Embedding inclusion and fairness as key considerations within sustainability initiatives is necessary to facilitate fair transitions.4 However, engagement is largely absent as a topic addressed at the strategic level in the country examples examined, apart from the Canadian and Australian strategy documents, which make mention of indigenous consultations. For example, Canada has allocated CAD 800 million to support indigenous-led environmental conservation efforts (Government of Canada, 2024[13]). The Australian government commits to drawing on the expertise of Aboriginal and Torres Strait Islander communities to mitigate climate change and the transition to net zero (Australian Government, 2024[29]).5

In addition, action may be necessary to balance potential conflicts between sustainable and inclusive development.6 Many of the strategy documents reference national supports for communities and workers affected by sustainability transitions, which emphasises the necessity of connectivity and coherence between STI and other policy domains. This is further discussed below. More generally, policies focused on skills accumulation and lifelong learning may also contribute to productivity increases.7 Chapter 3 discusses these and other related issues extensively.

Table 1.4. Policy options to support fair sustainability transitions

Governments use a range of funding instruments to support RDI, many of which can be used to promote activities across a wide range of the innovation chain (Figure 1.6). Nevertheless, some funding instruments are preferred over others to promote either (breakthrough) R&D or demonstration, deployment and diffusion of technologies. Both are essential, since a mix of knowledge, innovation, and novel and existing technologies is needed to promote transformative change.8 The challenge for governments is to strike an appropriate balance, which will vary depending on, for example, a technology’s maturity and the domestic capabilities of firms and universities to develop and adopt related novel science and technologies.

Figure 1.6. Mix of selected policy instruments for public funding of research, development and innovation

Notes: R&D: research and development. This figure shows a selection of R&D and innovation funding instruments used by governments and their typical range. A more comprehensive taxonomy developed by the EC-OECD STIP Compass database can be found at: https://stip.oecd.org/stip.

Another important STI policy debate concerns striking an appropriate balance between directed and non‑directed support to RDI activities performed in both the public and private sectors. As shown in Figure 1.3, much government support for R&D is non-directed and serves general economic development and the advancement of knowledge. In the public sector, non-directed support typically takes the form of institutional “core” funding for universities (including GUF) and public research institutes; and “responsive” R&D grants where researchers propose research projects “bottom-up” for funding. This contrasts with “managed” R&D programme project grant calls, where funding bodies define, with varying degrees of precision, the areas of research they will fund. These “top-down” calls are co‑designed with the research community and support a mix of R&D: typically, basic research in universities and some public research institutes and applied research and experimental development in public research institutes dedicated to supporting firms’ technological upgrading.

There is a natural tension between promoting scientific research that is explicitly oriented towards solving practical challenges and encouraging a broad-based development of scientific capabilities that might ultimately contribute to such goals. This is because research for nominally different purposes can help to achieve transformative goals in unexpected ways. For instance, analysis of low-carbon and other environmental management patents indicates that core scientific disciplines like chemistry and physics, together with material sciences and biology, are among the most heavily cited sources of scientific knowledge relevant for new inventions by inventors and examiners. The wide-ranging nature of these scientific influences underscores the challenge of pinpointing a single dominant field driving low-carbon innovation. This suggests that policymakers should avoid a crude classification of scientific domains as relevant for tackling specific transformative goals (OECD, 2025[30]). It is also often the case that significant breakthroughs emerge from the accumulation and combination of decades of curiosity-driven research across various fields. This was shown most recently in the rapid development of COVID-19 vaccines,9 demonstrating how long-term investments in R&D contribute to societal resilience (see Chapter 4).

Similar tensions play out in debates around public support to private sector R&D. Not only do governments vary in the level of support they offer businesses to encourage them to perform R&D and innovate, they also vary in the policy instrument portfolio they use (Figure 1.7).10 Among directed funding instruments, governments offer grants, loans, credits and debt guarantees to support businesses in their RDI activities and use public procurement to promote firms’ innovation and technological upgrading. Among non-directed instruments are business R&D tax incentives and innovation vouchers (Figure 1.6).

There has been considerable change in the business R&D support policy mix over the last two decades, with a near-universal shift from directed support instruments to a greater reliance on indirect R&D tax incentives. In 2022, 32 of the 38 OECD Member countries gave preferential tax treatment to business R&D expenditures. R&D tax incentives represented around 56% of total government support for business R&D in 2022, compared to 35% in 2006 (Figure 1.8).

Figure 1.7. Direct government funding and government tax support for business R&D, 2023

As a percentage of GDP

Notes: For Austria, Bulgaria, Chile, China, Germany, Ireland, Luxembourg, OECD average, Portugal, the Slovak Republic, South Africa and the United Kingdom, the latest available figures of direct and tax support for business R&D refer to 2022 instead of 2023. For Australia, EU-27 area, France, New Zealand, Switzerland and the United States, figures refer to 2021. For Brazil, Colombia, Denmark and Romania, data refer to 2020. Preliminary OECD estimate of government tax relief for R&D expenditures for the OECD in 2022. For general and country-specific notes on the estimates of government tax relief for R&D expenditures, see https://stats.oecd.org/wbos/fileview2.aspx?IDFile=7bac5f9d-e557-4938-8928-dda26fb93a19. Data on government tax relief for business R&D also includes subnational tax support for Canada, Hungary and Japan.

Source: OECD (2025), OECD Tax Incentives Database, https://oe.cd/rdtax (accessed in April 2025).

Statlink

Figure 1.8. Shift in the government policy support mix for business R&D, 2000-2022

OECD, constant PPP USD, 2007=100

Notes: For general and country-specific notes on the estimates of government tax relief for R&D expenditures, see: www.oecd.org/sti/rd-tax-stats-gtard-ts-notes.pdf. Data on government tax relief for business R&D also includes subnational tax support for Canada, Hungary and Japan.

Sources: OECD (2025), R&D Tax Incentives Database, http://oe.cd/rdtax (accessed in April 2025); OECD (2025), Main Science Technology Indicators Database, https://oe.cd/msti (accessed in March 2025).

Statlink

Neutrality and reduced policy discretion of tax incentives have several desirable features when funding R&D. They are less costly to administer and, when neutrally designed and available on demand, are more easily compliant with state aid rules (OECD, 2024[31]). However, after two decades of widespread deployment, there is broad consensus that tax incentives are more suited, in principle, to encouraging R&D activities with near-market potential and the shortest payback time. By contrast, direct measures, such as grants, are more suitable for supporting longer term, high-risk R&D, and targeting specific areas that either generate public goods or have particularly high potential for spillovers. Both types of measures provide useful support, but the growing urgency to promote transformative change may point to the need for a rebalanced approach in some countries that gives greater prominence to more ambitious direct measures (González Cabral, Appelt and Hanappi, 2021[32]).11

Given that scientific research and technological innovation are inherently uncertain, policy support should “spread bets” on a diversity of solutions using a portfolio approach. This will help avoid technological lock‑in and develop the absorptive capacities to access knowledge and technologies developed elsewhere. A portfolio approach should also balance funding support across stages of the innovation chain and promote interactions and complementarities between stages to help steward ideas from conception to application and bridge particular “valleys of death”. There is no one-size-fits-all solution and the composition of these portfolios and the research areas, technologies, industries and other forms of innovation that are prioritised will depend significantly on the current context of individual countries and their desired future visions.

In the meantime, governments are experimenting with novel funding mechanisms and arrangements to promote more responsive R&D, more breakthrough research and innovation, and more integrated support across the innovation chain. These are briefly discussed below.

Innovative funding mechanisms to promote responsive R&D

Various policy innovations are emerging that aim to make funding more agile and responsive to changing conditions. Some countries have introduced funding initiatives that consider a broader set of societal considerations in their award decisions. For instance, with the Strategic Innovation Partnership programme in Sweden, the innovation agency Vinnova ranks proposals based on traditional criteria (i.e. the business case and degree of scientific excellence), and, for those that pass this initial evaluation, non-government partnerships select projects to fund that best align with their “theories of change”. Under the National Research and Innovation Strategy for Smart Specialisation of the Czech Republic 2021-2027, standard RDI calls for proposals provide a “bonus” during the assessment process for projects that are relevant to the missions, which increases their probability of procuring funding.12 The Austrian Research Promotion Agency considers the sustainability of each project in addition to the substantive and economic aspects. This includes emissions, pollution, resource and energy consumption, and socio-economic impact (e.g. effects on poverty reduction, health, education, gender, working conditions and fighting corruption).13 Canada’s Strategic Innovation Fund also includes social considerations in its funding decisions, including if a recipient commits to its 50-30 Challenge for board diversity, inclusive hiring practices, environmental practices, indigenous consultations and investment in local communities14 (McIvor, forthcoming[33]).

Governments are also experimenting with flexible organisational structures to ensure funding agencies are better equipped to respond to emerging opportunities and challenges. For instance, some governments are using network delegation to crowd in private sector investment and take funding decisions. Under this model, governments competitively select associations, networks or consortia, who prove their connections to the STI systems to play a role in the funding process. Although the model is not entirely novel, to focus them more on transformative goals, governments are running competitions to select these types of organisations based on their co-developed visions or roadmaps for transformative change. Examples include the Netherlands’ Top Consortia for Knowledge and Innovation, which provide an ongoing matchmaking role within their sectors to help develop consortia of partners to apply for funding opportunities that tackle aspects of the Dutch missions and help disseminate the results of these projects within their sectors. The government still takes the funding decisions under these arrangements. By contrast, Denmark’s Innomissions programme uses a more decentralised model.15 After issuing a call for proposals to develop roadmaps to address four mission areas, the government has delegated control to the winning consortia to issue calls for proposals, review the proposals and allocate funding, and itself performs just a state aid check of the approved projects16 (McIvor, forthcoming[33]).

In other developments, some funding bodies are experimenting with randomisation and lotteries for taking funding decisions to test whether they can achieve more inclusive and ambitious outcomes compared to traditional allocation methods. There are variations to randomisation, including: partial randomisation, which initially vets proposals before selecting those that pass an initial set of criteria at random; weighted randomisation, which ranks proposals, with the better ranked ones awarded more “tickets” in the randomised selection; and tiering, which is similar to weighted randomisation but with less granularity in the ranking.17 Some funding bodies are using these approaches to advance on the transformative goals. For instance, the British Academy used partial randomisation for its small research grants. It found that using partial randomisation could lead to a more ethnically and institutionally diverse cohort of award-holders.18 The Austrian Science Fund also used partial randomisation through pilot grants, which provided seed funding for radical new and bold research ideas that have the potential to transform established scientific knowledge in all disciplines (McIvor, forthcoming[33]).

Organisational innovations to promote breakthrough R&D

Several governments are paying particular attention to the ideal organisational structures to accelerate breakthrough, or transformative, research and innovation. An increasing concern of the scientific community in recent years is that research funding processes have become too conservative and only encourage incremental advances in STI. Failure to encourage and support research on risky, “out of the box” ideas may jeopardise a country’s longer term ability to compete economically and to harness science for solving national and global challenges (OECD, 2021[34]) (see Chapter 4). At the same time, there are worrisome claims that research productivity has been falling in recent decades. Multiple explanations have been offered for this phenomenon, including changes in scientific incentives that reward incremental science, the growing need for but outstanding challenges of supporting interdisciplinarity, and the declining share of public research, which tends to be more supportive of breakthrough R&D (Ciaffi, Deleidi and Di Bucchianico, 2024[35]; OECD, 2023[36]).

These concerns have led several countries to establish new public bodies to pursue focused breakthrough research and innovation, broadly inspired by, and in many cases modelled on, the United States Defense Advanced Research Projects Agency, with various degrees of adaptation. Examples include the Federal Agency for Disruptive Innovation (Germany); the Moonshot Research and Development Program (Japan); the Advanced Research and Invention Agency (United Kingdom); the High-risk, High-gain Research Programme (France); and the Advanced Research Projects Agency for Health (United States). Some of these initiatives are outlined in table 1.5 together with others that use established or open funding calls to identify high‑potential projects that might be outside of current funding priorities or mandates.

Table 1.5. Selected examples of science, technology and innovation policy measures to promote funding agility and breakthroughs

The purported need for these new research funding organisations has been justified on a number of grounds, including that larger projects are required than academic laboratories can undertake; more co‑ordination is needed than occurs in academic departments or across generic research consortia; the desired innovations might be insufficiently profitable to arise through start-ups funded by venture capital or industrial R&D projects; and a mismatch exists between time frames typical of academia and traditional research funders and the immediacy of some challenges (OECD, 2024[31]). Programme managers in these initiatives typically have broad freedom to design technical initiatives and redirect resources between their portfolio of projects through a large integrated budget (OECD, 2021[34]).19 Funding decisions can also be rapid, with organisations like Germany’s Federal Agency for Disruptive Innovation being able to take some initial decisions within two weeks.20

Integrating funding across the innovation chain

Some countries are implementing policy initiatives that support research, development and/or demonstration activities across the entire innovation chain. Table 1.6 provides some examples. In several instances, countries have developed two-part funding programmes to support R&D and subsequent demonstration of targeted technologies, such as CCUS, e.g. Norway’s CLIMIT Programme and Canada’s Agricultural Clean Technology Programme.

Table 1.6. Examples of policy initiatives targeting the entire innovation chain

Individual funding authorities with more expansive mandates are also supporting solutions across the innovation system and along the innovation chain to promote transformative change. Because these organisations have such a breadth of tools at their disposal, they are uniquely situated to address more systemic challenges. In the United Kingdom, the UKRI Challenge Fund21, for example, is addressing societal challenges through funding a range of activities, including collaborative cluster projects, R&D centres, research projects, demonstration projects, behavioural research, and other areas. While many of these agencies are funding transformative goals as part of their broader STI mandates, some funding authorities have more fundamentally incorporated these goals into how they are structured. For instance, the Netherlands Enterprise Agency has a range of instruments that support everything from proof of concept and investments in seed-stage companies to business growth and partnerships. It restructured itself around 3 thematic domains and 20 societal challenges. It then mapped out its programmes to identify how each one relates to its transformative goals to support the scale-up and phase-out of different technologies, as well as the gaps in its programme offerings. The Netherlands Enterprise Agency uses a “theory of change” to guide its investment decisions, and an annual Societal Challenge Cycle is used to update its overarching organisational strategy22 (McIvor, forthcoming[33]).

Action 3: Strengthen co-ordination with non-science, technology and innovation policy areas on transformative change

Public funding to support scientific and technological breakthroughs as well as their diffusion must come from several parts of government, including sectoral ministries and agencies in areas like energy, transport, agriculture and health. Ministries and authorities with formal STI policy responsibilities need to help orchestrate this effort and steer public and private investments to where they are needed the most. However, multidimensional issues like inclusive economic renewal, security and resilience, and sustainability transitions cannot be achieved or even be chiefly driven by STI policies. Other policy areas with regulatory and fiscal powers have often taken the lead. Such transformations require a more systematic and agile approach to contend with issues that cut across policy boundaries and require co‑ordination across subnational, national and international levels of governance.

The policy landscape in many countries is characterised by structural silos and disconnects between different policy domains, national and subnational counterparts, and different actors working at the interface between STI policy and the STI system (e.g. funding agencies). While this segmentation has enabled the management and even optimisation of different aspects of complex systems in isolation, it can be a barrier to the effective transformation of these systems to better address complex societal challenges.

Governments can deploy a range of cross-government and territorial co-ordination measures to alleviate fragmentation and better orchestrate their interventions, including shared national visions, roadmaps and missions; joint programming between research and innovation funding agencies; and strategic oversight by high-level cross-departmental committees. Some countries have also implemented structural and organisational changes, for example by merging funding agencies or ministries and territorial authorities for STI that cover different parts of the innovation chain (Halme et al., 2019[37]). Box 1.1 provides an overview of related policy measures found in the EC-OECD STIP Compass database.

Box 1.1. What cross-government coherence and co-ordination measures are governments taking?

An analysis of data from the STIP Compass database identified close to 400 unique cross-government coherence and co-ordination initiatives related to the transformative goals (EC-OECD, 2023[38]).* Many of the initiatives analysed target the optimisation of government operations by reducing bureaucracy, consolidating funding and activities, facilitating the co-development or co-funding of shared priorities and policy portfolios, harmonising a policymaking culture or processes in certain areas (e.g. procurement, experimentation, etc.), and making activities more responsive and flexible (e.g. programming concierge platforms, single-window funding applications, etc.).

Around 35% of the cross-government initiatives analysed include horizontal co-ordination bodies between national policy domains. Of these, the most engaged policy domains include economic affairs (37%), education (31%), environment (28%), culture (27%), energy, (27%), finance (25%) and agriculture (23%). While around a half of them issue specific recommendations to ministries to implement, a smaller proportion include the development of joint studies (18%) or the alignment of budget allocations (7%).

Around 40% of the cross-government initiatives analysed include national strategies. Many of these target objectives/challenges or themes that cut across several sectors or are universally relevant. Roughly 50% are related to climate change or environmental sustainability while over 25% target issues of socio‑economic security (energy/food) and other societal challenges (health, aging population). A smaller proportion (15%) relate to inclusiveness (e.g. inequality, job insecurity). Energy is the most represented sector, included in roughly one-third of the strategies analysed. Several other sectors are also reasonably well-represented. Health and healthcare, automotive and road transport, agriculture, food, and marine and ocean are each captured in 15-20% of strategies while education, telecommunications and IT, public administration, pharmaceuticals, and electronics are each represented in 10-15%.

The presence or absence of particular follow-up mechanisms can signal the level of formalised co‑ordination or concerted attention to the translation of strategies into policy action. Roughly half of the strategies analysed are introduced in parallel to periodic monitoring or evaluation mechanisms. At the same time, a smaller proportion are linked to targeted mechanisms or tools intended to support their implementation: 40% of strategies have an associated action plan, 25% have a dedicated co-ordinating or monitoring public body, 20% have dedicated budget allocations, and roughly 10% are linked to a new regulation or law. Additionally, some strategies are supplemented with complementary initiatives, such as the creation of new governance structures or bodies, policy intelligence or consultation bodies, and networks.

* Elements of the STIP Compass data taxonomy (e.g. policy themes related to green transitions, research security, and equity and inclusion) were used to identify policies aligned with the transformative goals.

Source: EC-OECD STIP Compass database, https://stip.oecd.org/stip (accessed on 10 March 2025).

Mission-oriented innovation policies

Among different types of STI policies with transformative ambitions, MOIPs form an internationally recognised policy approach, with distinct principles and features, and a growing body of practical and conceptual knowledge supporting their adoption (OECD, 2024[39]). They involve co-ordinated packages of policy and regulatory measures tailored to mobilising STI to address well-defined objectives related to a societal challenge, in a defined period.23 MOIPs can span various stages of the innovation chain from research to demonstration and market deployment. They can also mix supply-push and demand-pull instruments and cut across various policy fields, sectors and disciplines. While they confront many of the traditional challenges of national innovation systems, MOIPs tend to provide longer term and more consistent funding compared to traditional research and innovation schemes, reflecting their alignment with the long-term character of broader, transformative goals (OECD, 2024[40]).

Given that missions are often nested across different levels of government, the locations from which they are co-ordinated and operated play a critical role in shaping their governance dynamics. Different centres of gravity provide different opportunities and challenges to mission governance. A recently published OECD study of about 100 missions aiming to reduce greenhouse gas emissions has found that, despite significant achievements and progress, they fall short of leveraging the complementarities of various policy and regulatory interventions to scale-up broad and ambitious solutions (OECD, 2024[40]). Most remain narrowly focused on technological innovation, led by STI authorities and reliant on innovation policy funding.

Budgets can set powerful conditions on funding that may force groups to co-operate across silos on the delivery of certain budget items. However, while many missions are supported by a rather integrated co‑ordination structure, most of them are funded by different funding streams that correspond to the different instruments/activities they integrate into their portfolio, originating from different mission partners and beyond. A mission may have funding for core STI activities (development of the agenda, “orchestration” of the mission) then lack funding to support other activities from research and innovation to skills and infrastructure needed to make an impact. This fragmented funding structure has significant implications for the level of mission integration, since it hinders co-ordination and co-operation. Mission managers often find government budgets inflexible and cyclical, which makes it harder for missions to pivot. Box 1.2 outlines possible governance configurations to help make missions more transformative.

Box 1.2. Making missions more transformative

While the challenge to design and, even more, implement missions are numerous and are well-documented, the options to make them more transformative are less clear. Based on previous and ongoing OECD work, five main pathways can be envisaged:

1.Gradual broadening and strengthening of missions by incrementally enlisting new actors; building trust; learning and attracting higher commitments from public authorities outside the science, technology and innovation (STI) realm; and higher investments from private actors. Given the legitimacy of and resources available to STI authorities, which define in large part their convening power, these missions might not extend and deepen much further.

2.Transfer of mission leadership from STI to sectoral authorities who “own” the challenges. Their mandate aligns more closely to the mission objectives (e.g. net zero, circular economy) and they hold essential intervention tools, resources and legal powers to realise them. STI authorities would need to ensure innovation remains a priority in the strategic agendas of these sector-led missions.

3.Ownership of missions by centre-of-government bodies such as a prime minister’s office or a powerful “transition” committee that can enforce a whole-of-government approach to realise the mission. In practice, this pathway has been challenged by a tendency for approaches to remain innovation-driven and a lack of buy-in from participating ministries. A carefully designed combination of carrots and sticks will be necessary to prevent departments from drawing on mission budgets without fully embracing objectives.

4.Smaller scale regional or local missions may be better equipped to define collective agendas and integrate different interventions while leveraging the benefits of place-based innovation and various forms of proximity (e.g. geographic, cultural). These “micro-missions” would still need to be articulated with bigger (national, global) transformative agendas to contribute meaningfully to grand challenges.

5.Dedicated mission agencies could be developed to “co-ordinate mission operations from the ground floor” and report to participating ministries. Once entrusted with one or several missions, these agencies should enjoy significant autonomy to protect activities from political short-term interference. They would also require a large portfolio of instruments or the possibility to co-operate with other agencies.

The one thing these pathways all have in common is adapting leadership structures and fostering co‑ordination and collaboration beyond STI authorities to unlock missions’ transformative potential. The choice between these pathways will depend on the trajectory of each mission, but also on underpinning national or regional institutional specificities.

Source: Adapted from OECD (2024[40]).

To help alleviate some of these challenges, separate funding authorities are issuing joint calls, which allows them to co-ordinate with a broader selection of funding instruments to support more activities across the innovation chain or the innovation system. In this model, the funding pots remain separated by authority. For instance, in Norway, the PILOT-E scheme24 funds business innovation from concept to market through a collaborative approach between five funding agencies that take a co-ordinated funding decision based on the programmes they have available and the technology readiness level of the proposal.  Similar flexibility is provided by Ireland’s Impact 2030 Steering Group25, which launches joint calls across the five largest STI funding departments. The funders then determine which instruments are best suited for supporting the different proposals26 (McIvor, forthcoming[33]).

Some governments are using central pots of funding to support activities from across government in a manner that transcends traditional ministerial structures and authorities. France’s Acceleration Strategies for Innovation27 has a central budget managed directly by a dedicated agency under the Prime Minister’s Office. They fund a broad portfolio of activities under various government agencies, without influence by their supervisory ministries, which cover a range of activities, including R&D, technology transfer, technology demonstration, infrastructure investment, and skills formation. Chile is taking a similar approach through its Sustainable Productive Development Program, which also combines investment and STI measures under the one programme. Ministers decide on a theme to focus on each year28 (McIvor, forthcoming[33]).

The ability to end a programme’s or project’s funding is an important aspect of agility but is often hard to achieve in practice within standard governance structures. Many MOIPs have built in either formal review processes within the life cycle of the project or taken stage-gated approaches to funding, where over a specified interval they reduce the number of projects and increase the amount of funding, e.g. Korea’s Alchemist programme, which funds six projects in the first year, three in the second and just one over five years (McIvor, forthcoming[33]).

Missions often grant key roles in the development of the strategic agenda and in implementation to incumbents within the sector(s) where the mission is located. These actors have resources and capabilities as well as infrastructure and networks that make their participation “unavoidable” in any change initiative. However, they also have vested interests in the currently established system that they may be tempted to preserve by advocating for incremental improvements rather than transformational change through alternative, more exploratory, solutions.29 Balancing the participation of incumbents in governance is therefore a key challenge for many missions, especially for ecosystem-based missions that rely on a high level of delegation of several governance functions (not least the development of the strategic agenda and the mobilisation and co-ordination of stakeholders) to ecosystem actors. Policymakers should be wary of becoming limited in their reach to established players within existing policy ecosystems, who already identify and know how to navigate this ecosystem. This calls for an important role for the state as a “moderator” to ensure a balanced and inclusive approach in the development and implementation of the mission’s strategic agenda to avoid mission capture (OECD, 2024[41]).

Action 4: Mobilise public funding to crowd-in private finance for transformative change

In recent years, there has been a growing focus among policymakers and funders to promote innovative financing mechanisms that can crowd in new sources of private financing for climate, clean energy, biodiversity and other sustainability challenges. According to the 2024 Financing for Sustainable Development report (United Nations Department of Economic and Social Affairs, 2024[42]), financing the SDGs requires trillions of dollars per year. Access to finance remains a critical obstacle, however, as shortfalls in funding constrain many countries’ ability to put forward and deliver ambitious climate commitments. The private financing gap is most evident in the energy sector, where, according to the International Energy Agency, 85% of the required investments in non-fossil fuel‑based energy will need to come from private sources (IEA, 2019[43]).

Several capital market failures discourage the allocation of private investment into technologies that promote transformative change.30 For example, there are often long-standing alternatives to low-carbon technologies, while deep technology solutions are well-known for being more intensive with timelines for development that do not align with private sector investment requirements. For emerging markets and developing economies, financing the implementation of their current climate plans remains particularly challenging in a context of high public debts and insufficient international support for climate finance (OECD/UNDP, 2025[8]). Achieving the SDGs will require co-operation between developed and developing economies where most of the impacts of the global challenges like climate change and global health are occurring. At COP29, for example, developed countries agreed to a plan in which developed countries committed to providing USD 300 billion annually by 2035 to assist poorer countries in combating climate change. This amount falls short of the USD 1.3 trillion annually that many developing countries believe is necessary to address climate challenges adequately (Bhattacharya et al., 2024[44]; CORDAID, 2024[45]).

Channelling STI financing for the SDGs requires more and new partnerships with multilateral development banks, charities and private foundations, and official development assistance, but also with private investors, pension funds and financial actors operating at the local level. Yet to direct STI financing for the SDGs at scale, a transformation in private investment and financing is needed. Governments can play critically important roles in promoting private investment in sustainability transitions through a range of economic and regulatory instruments. These are underpinned by a range of public policy goals, e.g. climate policies, industrial policies, energy security policies, and improving economic resilience and reducing dependence on global value chains.31 Many of these instruments target innovation and technology, even if they do not directly subsidise the costs of firm investments in R&D, by affecting the broader financial eco-system for innovation.

Towards blended finance?

Mobilising private capital rests primarily on managing risk. When public capital is used to mobilise private or commercial capital, it normally means to provide an investment situation in which risk and returns have found a balance that is acceptable to those investors. This will also depend on the project itself and to what extent the financial solution offered provides an acceptable risk-return profile in each case. Among the innovative approaches to crowd in private finance is “blended finance”, which has mainly been used in development finance (OECD, 2018[46]). With its focus on deploying public financial resources with the view to leverage or attract private capital, blended finance has contributed large resources for investments in developing countries. It is an approach for combining financial instruments in ways that allow participants in the blending to respect their respective mandates and risk-return preferences within an agreed-upon contractual structure.

Approaches like blended finance, which initially emerged as an innovative tool in the development community to crowd in private financing for sustainability projects in developing countries (Samans, 2016[47]), are gaining traction in the STI policy field as a way to combine public and private finance across the innovation chain (Miedzinski et al., 2020[48]; OECD, 2022[49]).

Research on blended finance has shown that different settings or investment purposes typically lead to different combinations of instruments, such as grants, debt, guarantees, funds or facilities and others, to be structured according to the investment situation to achieve a best possible fit of partners and instruments (Kwon et al., 2021[50]). Table 1.7 outlines some of the main instruments and their definitions, which can be clustered into four groups:

1.Grants and technical assistance originate from either public funding or philanthropic capital without any expectation of positive returns. Actors leveraging on these instruments are typically the ones initiating the transactions, and grants blended in the mix will play the catalytic role.

2.Outcome funding, impact bonds and impact-linked finance stand out from the rest in that they connect impact with financial rewards.

3.Various debt and equity instruments, like market-rate, concessional or subordinated debt, normally take higher risk expecting higher returns than debt finance.32 Hence, debt may be used in the later stages of a project’s development.33

4.First loss and guarantees do not normally seek returns but are deployed to provide derisking and attract additional capital. They are typically used in later stages of a development when scaling is needed and when there is a track-record of performance at hand.

These clusters of instruments often co-exist in single overarching policy initiatives, some of which are briefly outlined in Table 1.8.

Table 1.7. Definitions of selected blended finance concepts and instruments

At the same time, the ability of public finance to crowd-in private finance for sustainability transitions should not be overestimated. Many regulations and incentives that are hard-wired in global capital and financial markets continue to direct private finance towards profitable ventures that may not always promote sustainability. The financing of STI activities targeting the SDGs faces familiar obstacles, such as market failures in the private financing of RDI, as well as economic and technology risks.34 A particular challenge to financing the SDGs is that several of them involve mobilising STI for the preservation and production of “common pool resources”, such as biodiversity, global health and sustainable oceans. Private firms have fewer incentives to provide such public goods; they also have an incentive to maximise the use and exploitation of common goods.

New funding models for STI involving public and private actors offer a mechanism to increase STI funding to support the provision of global public goods. These funding models include blended finance involving multilateral international development banks, philanthropies and institutional investors such as pension funds. Sustainability bonds issued by governments and corporations can potentially scale-up private financing if issues of transparency, monitoring and accountability can be effectively addressed. STI-for-debt swaps, which function like climate-for-debt swaps, could also be used to encourage developing countries to invest in STI capacity building.

Action 5: Promote transformative change rather than “business-as-usual” outcomes

In promoting transformative change, STI policy measures should be directed at specific actions that help achieve transformations rather than “business-as-usual” outcomes. Many of the necessary reforms are familiar to the STI policy community, and promoting transformative change often coincides with achieving reforms to address long-standing challenges in STI systems.35 Barriers remain, however, for example in bridging aspirational strategy with the development and implementation of concrete policy interventions and in scaling-up and institutionalising corresponding policy innovations.

Change versus stability

A starting question is the concrete differences between incremental (adaptation) and transformative change, as well as how or if these two processes are related. Transformative change refers to “a radical permanent qualitative change in the subject being transformed, so that the subject when transformed has very different properties and behaves or operates in a different way” (HM Treasury, 2022, p. 122[51]); and “a major change in the structure of the economy brought about by deliberate policy efforts aimed at supporting specific long-term environmental, social, economic or other goals, or in response to climate change and other relevant long-term trends” (New Zealand Ministry of Business, Innovation and Employment, 2023, p. 1[52]). By contrast, incremental change is predominantly focused on preserving integrity and stability. Small-scale, localised and short-term adjustments are made to cope with change or challenges. These changes enable an evolution that allows maintaining fundamental structures and functions of the system (Schumer et al., 2022[53]).36

Are the two processes related? It has been postulated that the aggregation of many incremental interventions can push a system towards a threshold or tipping point that, when breached, triggers a self‑perpetuating process of deep and rapid change (Lenton et al., 2022[54]; Mey, Mangalagiu and Lilliestam, 2024[55]). This non-linear process can be attributed to various feedback dynamics common to innovation processes (Allen and Malekpour, 2023[56]; OECD, 2025[57]) (Box 1.3). The resulting transformation is generally faster, more intense or extensive than expected and can result in “tipping cascades” that impact other systems (Milkoreit, 2022[58]; Spaiser et al., 2024[59]). It proceeds through a combination of complex, dynamic and non-linear pathways, often following S-shaped curve dynamics where the pace of change ramps up and tapers off depending on the phase (Loorbach, Frantzeskaki and Avelino, 2017[60]; Meadowcroft, 2021[61]; Victor, Geels and Sharpe, 2019[62]). The result can be the reconfiguration of component parts of the system, including the pattern of interactions between them, and resulting outcomes (HM Treasury, 2022[51]).37

Box 1.3. Selected feedback dynamics common to innovation processes

• Economies of scale: Supply-side cost reductions occur as production increases and becomes more efficient and fixed costs are spread over larger volumes. This can yield a virtuous cycle where lower costs encourage adoption, leading to the scale-up of production and further cost reductions.

• Learning-by-doing: As experience accumulates, improvements in performance and cost occur, often in parallel to economies of scale. This can yield a virtuous cycle where greater adoption provides more opportunities for learning, improving quality and competitiveness, and spurring further adoption.

• Network and co-ordination effects: As adoption grows, utility may also increase for particular types of innovation. For example, this type of feedback dynamic is typically observed for platform technologies (e.g. blockchain, artificial intelligence), interoperable systems (e.g. electric vehicle charging networks, Internet of Things devices) and knowledge communities (e.g. open-source software). When a critical mass of adoption is reached, a bandwagon effect can occur and tip the market in favour of the emerging technology. Indirect network effects can also occur as complementary goods increase in quality or become more abundant (see system build-out).

• Adaptive expectations: Technological feedback is reinforced via political, institutional and cultural dynamics. Increased uptake reduces uncertainty and strengthens the confidence of users, investors and other actors. Legitimacy and credibility spur expectations, norms and new institutions like industry associations, standards bodies, educational curricula and user communities, which enable further adoption and investment and may motivate divestment from or discontinuation of the status quo (see destabilisation and phase-out).

• System build-out: Complementary innovations (e.g. products, infrastructures, business models, etc.) can help to address technical challenges and enhance the utility of the core technology. However, established technologies also often have deeply embedded complements that can become liabilities or stranded assets as transitions progress.

• Destabilisation and phase-out: Once a tipping point is reached, reinforcing feedback dynamics can accelerate the phase-out of established systems through declining sales and economies of scale, erosion of network advantages, increasing costs and reduced competitiveness. When incumbent firms anticipate declining profitability or stricter regulation, they may divest or innovate, which contributes to the cycle.

Along similar lines, the OECD Agenda for Transformative Science, Technology and Innovation Policy suggests that a progressive series of incremental changes in the STI policy mix could also potentially combine into a deeper intervention that disrupts the status quo and creates system-wide change.38 In this way, promoting transformative change may coincide with achieving multiple reforms that address long-standing challenges distributed across STI systems. Its non‑linearity makes transformative change messy and unsuited to “command-and-control” notions of policy intervention. Instead, STI policy should identify “leverage points” for interventions that acknowledge positive and negative feedback dynamics, the distribution of power within systems,39 and the necessity to sequence change to unlock potential pathways. This calls for a reappraisal and recalibration of the frameworks, tools and mechanisms currently used to develop and deploy STI policy, to embrace more reflexive, systemic and responsive processes (OECD, 2024[1]). Table 1.9 outlines policy implications and case study examples related to feedback dynamics like these. Box 1.4 details two examples where governments have taken a more systemic approach targeting multiple, interdependent feedback cycles.

Table 1.9. How science, technology and innovation policies can harness the dynamics of innovation processes

Sweden’s School Food Mission is a Vinnova pilot project intended to transform Sweden’s school food system and contribute to systemic change across the broader food system. The programme has made use of various transformative policymaking approaches. System maps have been co-created with stakeholders to identify leverage points to trigger system-wide effects. Design thinking and prototyping were used to develop pilot activities in partnership with students, municipalities and food producers, among others. The initiative’s governance mechanisms also enable co-ordination across various policy sectors.

The programme targets feedback dynamics in various ways. The mission has a strong culture of prototyping and iteration where over 1 500 students and 140 partners have engaged to test and evaluate open-source solutions. The programme has evolved through multiple phases involving the adjustment of activities and targets (learning-by-doing). It also exhibits a strong degree of system co‑ordination and stakeholder co-ownership where roadmaps and platforms have been designed to bring actors together across silos and foster legitimacy (network effects). Societal expectations have been shaped by the mission’s shared vision and co-created plans and initiatives (adaptive expectations). Finally, the mission’s focus extends beyond food to develop complementary systems needed to facilitate wider transformation. This includes infrastructure (e.g. redesigned food halls), data systems, procurement tools and curriculum reform (system build-out).

Germany’s National Hydrogen Strategy

Germany’s National Hydrogen Strategy1, launched in June 2020 and updated in July 2023, focuses on advancing and scaling hydrogen technologies, expanding infrastructure, and fostering partnerships to secure clean hydrogen supply. The update introduces new measures to accelerate production, import and use; facilitate integrated solutions and broader sector deployment; and support the development and integration of hydrogen ecosystems at the national and international levels.

The strategy targets feedback dynamics in various ways. Large-scale production projects, including international lighthouse initiatives, help to spread fixed costs while demand-side mechanisms like conditional contracts and procurement auctions ensure uptake (economies of scale). Several regulatory sandboxes and living labs provide real-world testing grounds to identify technical, logistic, legal and business model challenges, facilitating learning and regulatory adaptation (learning-by-doing). Linking to the European Hydrogen Backbone in addition to building up 1 800 km of domestic infrastructure connects domestic and regional production hubs, storage centres, and import terminals. The Southern Hydrogen Corridor initiative between Germany, Austria and Italy also facilitates strategic infrastructure development along transport and industrial corridors (Landini, Amante and Wacket, 2024[63]) (network effects).

In addition, the Hydrogen Acceleration Law (May 2024) helps to boost investor confidence and streamline approvals (German National Hydrogen Council, 2024[64]) and was further endorsed by the federal cabinet with over EUR 3 billion in procurement financing (2027-2036) (HyResource, n.d.[65]) (adaptive expectation). In addition to advancing electrolyser technologies and hydrogen infrastructure, supports also target the development of hydrogen refuelling infrastructure for buses, trucks and trains and pilot projects for fuel cell technologies (system build-out). Finally, construction of the Hydrogen Core Network aims to incorporate repurposed natural gas pipeline, with 60% of the network already available from repurposed natural gas pipeline (phase-out).

1. See www.bundeswirtschaftsministerium.de/Redaktion/EN/Dossier/hydrogen.html.

Sources: EC-OECD (2025), STIP Compass: International Database on Science, Technology and Innovation Policy (STIP), edition October 7, 2025, https://stip.oecd.org/moip/case-studies/43?utm= ; German Federal Ministry for Economic Affairs and Energy (2025[66]).

Conclusions

This chapter proposed five policy “actions” that governments should consider when reforming their STI policy mix to better contribute to transformative change agendas, focusing, for the most part, on funding and financing arrangements for STI. In the first action – promoting a policy agenda that contributes to broad transformative change – the chapter highlighted several STI policy options that can leverage synergies between different priorities. In this way, support to national competitiveness can also contribute to resilience and security, as well as sustainability transitions, if designed appropriately.

The chapter’s second policy action – balancing direct and indirect R&D funding for transformative change – introduced a simple schema that governments can use to map their policy interventions along two axes: 1) the innovation chain, i.e. from (breakthrough) R&D to demonstration, deployment and diffusion; and 2) the extent to which they are directed “top-down” by government and STI funding bodies. All instruments can play important roles in promoting transformative change, and a key challenge for policymakers is to strike an appropriate balance between them. There are no one-size-fits-all portfolios and an appropriate balance will depend on a country’s assets and priorities.

The chapter’s third policy action – strengthening co-ordination between STI and non-STI policy areas – aims to bridge policy silos to advance transformative change. Among popular approaches are MOIPs, which nevertheless remain constrained by a narrow focus on technological innovation and reliance on STI leadership and funding. The chapter highlighted how governments have begun to experiment with various governance solutions to overcome these limitations, including the ownership of missions by centre-of-government bodies and dedicated mission agencies.

Looking beyond co-operation across the public sector, in its fourth policy action, the chapter outlined how governments can mobilise public funding to crowd-in private finance for transformative change. Among these approaches is so-called “blended finance”, which combines a range of financial measures – including grants and various debt and equity instruments – that can accommodate the risk-return preferences of different actors within an agreed contractual framework. These approaches have their limits, however, especially when the goal is to use STI to preserve or produce common pool goods. Governments should continue to experiment with instruments like sustainability bonds and STI-for-debt swaps, which have the potential to direct STI finance and help scale-up private investments in RDI to promote transformative change.

Finally, in its fifth policy action, the chapter called for greater appreciation of the nature of transformative change and how it differs from and relates to incremental change. This is an important consideration with a view to achieving more than “business-as-usual” outcomes. Transformative change is non-linear and marked by various feedback dynamics that can ramp up the pace of change and reconfigure whole systems. The chapter argued that STI policymakers should identify “leverage points” for interventions that can trigger and accelerate the sorts of system-wide changes needed for transformations.

These five policy actions clearly overlap and should be viewed systemically when formulating and implementing STI policies. The fifth policy action – appreciating the non-linear dynamics of transformative change – underpins the other policy actions proposed in the chapter and should be an essential consideration when balancing the STI policy mix. Co-ordination across government on priority agendas is also essential insofar as support to research and innovation and their diffusion is widely distributed across various ministries and agencies. Co-ordination must also extend to the private sector, given its dominant role in RDI and their commercialisation. Finally, as societies and economies face multiple challenges – and opportunities – governments must balance their STI policy support to a range of priorities, including economic competitiveness, resilience and security, and sustainability transitions. This is far from easy, but the chapter has highlighted several policy options for intentionally leveraging synergies and mitigating trade-offs between them.

References

[56] Allen, C. and S. Malekpour (2023), “Unlocking and accelerating transformations to the SDGs: A review of existing knowledge”, Sustainability Science, Vol. 18/4, pp. 1939-1960, https://doi.org/10.1007/s11625-023-01342-z.

[5] Aristodemou, L. et al. (forthcoming), Assessing the directionality of R&D funding towards societal goals through new data sources and AI-assisted methods, OECD Publishing, Paris.

[85] Aristodemou, L. et al. (2023), “Measuring governments’ R&D funding response to COVID-19: An application of the OECD Fundstat infrastructure to the analysis of R&D directionality”, OECD Science, Technology and Industry Working Papers, No. 2023/06, OECD Publishing, Paris, https://doi.org/10.1787/4889f5f2-en.

[79] Åström, T. (2021), Meta-evaluation of the third round of strategic innovation programmes, https://www.technopolis-group.com/wp-content/uploads/2021/12/metautvardering-av-tredje-omgangen-strategiska-innovationsprogram-efter-sex-ar-min.pdf.

[29] Australian Government (2024), Australia’s National Science and Research Priorities, Commonwealth of Australia, https://www.industry.gov.au/sites/default/files/2024-08/australias-national-science-and-research-priorities-2024.pdf.

[12] Australian Government (2024), Future Gas Strategy, Commonwealth of Australia, https://www.industry.gov.au/sites/default/files/2024-05/future-gas-strategy.pdf.

[68] Béné, C. et al. (2012), “Resilience: New utopia or new tyranny? Reflection about the potentials and limits of the concept of resilience in relation to vulnerability reduction programmes”, IDS Working Papers, Vol. 2012/405, pp. 1-61, https://doi.org/10.1111/j.2040-0209.2012.00405.x.

[81] Berger, M., C. Criscuolo and A. Dechezleprêtre (2025), “Understanding Government Venture Capital: A primer and a taxonomy”, OECD Science, Technology and Industry Working Papers, No. 2025/06, OECD Publishing, Paris, https://doi.org/10.1787/2157a5b3-en.

[44] Bhattacharya, A. et al. (2024), Raising Ambition and Accelerating Delivery of Climate Finance, Grantham Research Institute on Climate Change and the Environment, London School of Economics and Political Science, London, https://www.lse.ac.uk/granthaminstitute/publication/raising-ambition-and-accelerating-delivery-of-climate-finance.

[27] BioRural (2024), “The EU Bioeconomy Strategy Action Plan and Progress”, web page, https://biorural.eu/the-eu-bioeconomy-strategy-action-plan-and-progress (accessed on 14 March 2025).

[69] Blühdorn, I. (2019), “The legitimation crisis of democracy: Emancipatory politics, the environmental state and the glass ceiling to socio-ecological transformation”, Environmental Politics, Vol. 29/1, pp. 38-57, https://doi.org/10.1080/09644016.2019.1681867.

[76] Causa, O. et al. (2024), “Labour markets transitions in the greening economy: Structural drivers and the role of policies”, OECD Economics Department Working Papers, No. 1803, OECD Publishing, Paris, https://doi.org/10.1787/d8007e8f-en.

[20] Chataway, J., R. Hanlin and R. Kaplinsky (2014), “Inclusive innovation: An architecture for policy development”, Innovation and Development, Vol. 4/1, pp. 33-54, https://doi.org/10.1080/2157930x.2013.876800.

[35] Ciaffi, G., M. Deleidi and S. Di Bucchianico (2024), “Stagnation despite ongoing innovation: Is R&D expenditure composition a missing link? An empirical analysis for the US (1948-2019)”, Technological Forecasting & Social Change, Vol. 206, https://doi.org/10.1016/j.techfore.2024.123575.

[7] Commission, E. (ed.) (2024), The future of European competitiveness, https://commission.europa.eu/topics/eu-competitiveness/draghi-report_en.

[16] Commonwealth of Australia (2024), Budget 2024-25: Cost of Living Help and a Future Made in Australia.

[45] CORDAID (2024), “Climate Finance”, news, https://www.cordaid.org/en/news/cop29-too-little-too-vague-on-climate-finance (accessed on 21 March 2025).

[25] DSIT (2024), The UK Science and Technology Framework, UK Department for Science, Innovation & Technology, https://www.gov.uk/government/publications/uk-science-and-technology-framework/the-uk-science-and-technology-framework (accessed on 12 March 2025).

[38] EC-OECD (2023), STIP Compass, https://stip.oecd.org/stip (accessed on 8 January 2024).

[83] Edmondson, D., F. Kern and K. Rogge (2019), “The co-evolution of policy mixes and socio-technical systems: Towards a conceptual framework of policy mix feedback in sustainability transitions”, Research Policy, Vol. 48/10, p. 103555, https://doi.org/10.1016/j.respol.2018.03.010.

[67] Eker, S. et al. (2024), “Cross-system interactions for positive tipping cascades”, Earth System Dynamics, Vol. 15/3, pp. 789-800, https://doi.org/10.5194/esd-15-789-2024.

[6] European Commission (2025), A Competitiveness Compass for the EU, European Commission, Brussels, https://commission.europa.eu/document/download/10017eb1-4722-4333-add2-e0ed18105a34_en (accessed on 7 March 2025).

[28] European Commission (2022), European Bioeconomy Policy: Stocktaking and Future Developments, Publications Office of the European Union, https://data.europa.eu/doi/10.2777/997651 (accessed on 14 March 2025).

[9] Garsous, G., D. Bourny and D. Smith (2023), “The climate implications of government support in aluminium smelting and steelmaking: An empirical analysis”, OECD Trade Policy Papers, No. 276, OECD Publishing, Paris, https://doi.org/10.1787/178ed034-en.

[66] German Federal Ministry for Economic Affairs and Energy (2025), “Hydrogen: A key element of the energy transition”, web page, https://www.bundeswirtschaftsministerium.de/Redaktion/EN/Dossier/hydrogen.html.

[64] German National Hydrogen Council (2024), Statement on the German Hydrogen Acceleration Act: Further Improving the Important Course for the Hydrogen Ramp-up, German National Hydrogen Council, https://www.wasserstoffrat.de/fileadmin/wasserstoffrat/media/Dokumente/EN/2024/2024-06-21_NWR-Statement_Hydrogen_Acceleration_Act.pdf.

[17] Ghosh, B. et al. (2020), “Transformative outcomes: Assessing and reorienting experimentation with transformative innovation policy”, TIPC Working Paper, TIPCWP 2020-02, https://doi.org/10.2139/ssrn.3686077.

[32] González Cabral, A., S. Appelt and T. Hanappi (2021), “Corporate effective tax rates for R&D: The case of expenditure-based R&D tax incentives”, OECD Taxation Working Papers, No. 54, OECD Publishing, Paris, https://doi.org/10.1787/ff9a104f-en.

[13] Government of Canada (2024), Budget 2024: Fairness for Every Generation, Government of Canada, Ottawa, Ontario, https://www.budget.canada.ca/2024/home-accueil-en.html (accessed on 13 March 2025).

[14] Government of Korea (2024), 5th Environmental Technology Environment Industry Environmental Technology Workforce Development Plan: 2024 Action Plan (Draft), Government of Korea.

[37] Halme, K. et al. (2019), “Case study on the policy mix for science-industry knowledge transfer in Finland: Contribution to the OECD TIP Knowledge Transfer and Policies Project”, https://stip.oecd.org/assets/TKKT/CaseStudies/2.pdf.

[15] HM House of Commons (2024), Autumn Budget 2024, UK Government, London, https://www.gov.uk/government/publications/autumn-budget-2024/autumn-budget-2024-html (accessed on 13 March 2025).

[51] HM Treasury (2022), The Green Book, HM Treasury, https://www.gov.uk/government/publications/the-green-book-appraisal-and-evaluation-in-central-government/the-green-book-2020#a7-transformation-systems-and-dynamic-change (accessed on 19 February 2024).

[65] HyResource (n.d.), “Germany”, web page, https://research.csiro.au/hyresource/policy/international/germany.

[75] IEA (ed.) (2023), Net Zero Roadmap: A Global Pathway to Keep the 1.5 °C Goal in Reach, https://www.iea.org/reports/net-zero-roadmap-a-global-pathway-to-keep-the-15-c-goal-in-reach.

[74] IEA (ed.) (2022), Unlocking the Potential of Distributed Energy Resources, https://www.iea.org/reports/unlocking-the-potential-of-distributed-energy-resources.

[43] IEA (2019), World Energy Investment 2019, Licence: CC BY 4.0, https://www.iea.org/reports/world-energy-investment-2019.

[24] ISED (2025), Innovation, Science and Economic Development Canada’s 2024-2025 Departmental Plan, Innovation, Science and Economic Development Canada, https://ised-isde.canada.ca/site/planning-performance-reporting/en/departmental-plans/innovation-science-and-economic-development-canadas-2024-2025-departmental-plan (accessed on 12 March 2025).

[18] Kanger, L., B. Sovacool and M. Noorkõiv (2020), “Six policy intervention points for sustainability transitions: A conceptual framework and a systematic literature review”, Research Policy, Vol. 49/7, p. 104072, https://doi.org/10.1016/j.respol.2020.104072.

[19] Kivimaa, P. and K. Rogge (2024), “Pursuing sustainability transitions and open strategic autonomy: A policy mix perspective on synergies and trade offs”, JRC Working Paper Series for a Fair, Innovative and Sustainable Economy, Vol. 07/2024, European Commission, Brussels, https://publications.jrc.ec.europa.eu/repository/handle/JRC139504.

[50] Kwon, T. et al. (2021), “Blended finance: When to use which instrument?”, Powerpoint presentation.

[63] Landini, F., A. Amante and M. Wacket (2024), “Italy, Germany, Austria sign cooperation deal on southern hydrogen link”, Reuters, https://www.reuters.com/sustainability/climate-energy/italy-germany-austria-sign-cooperation-deal-southern-hydrogen-link-2024-05-30.

[54] Lenton, T. et al. (2022), “Operationalising positive tipping points towards global sustainability”, Global Sustainability, Vol. 5, https://doi.org/10.1017/sus.2021.30.

[78] Litvinova, V. et al. (2023), “mRNA Vaccine Platform: mRNA Production and Delivery”, Russian Journal of Bioorganic Chemistry, Vol. 49/2, pp. 220-235, https://doi.org/10.1134/s1068162023020152.

[60] Loorbach, D., N. Frantzeskaki and F. Avelino (2017), “Sustainability transitions research: Transforming science and practice for societal change”, Annual Review of Environment and Resources, Vol. 42/1, pp. 599-626, https://doi.org/10.1146/annurev-environ-102014-021340.

[22] Mazarr, M. (2022), The Societal Foundations of National Competitiveness, National Defense Research Institute, RAND Corporation Santa Monica, CA, https://www.rand.org/pubs/research_reports/RRA499-1.html.

[33] McIvor, C. (forthcoming), New financing models for research and innovation, OECD Publishing, Paris, forthcoming.

[61] Meadowcroft, J. (2021), “Pathways to net zero: A decision support tool”, Transition Accelerator Reports, Vol. 3/1, pp. 1-108, https://transitionaccelerator.ca/reports/pathways-to-net-zero-a-decision-support-tool.

[11] Meadowcroft, J. (2011), “Engaging with the politics of sustainability transitions”, Environmental Innovation and Societal Transitions, Vol. 1/1, pp. 70-75, https://doi.org/10.1016/j.eist.2011.02.003.

[55] Mey, F., D. Mangalagiu and J. Lilliestam (2024), “Anticipating socio-technical tipping points”, Global Environmental Change, Vol. 89, p. 102911, https://doi.org/10.1016/j.gloenvcha.2024.102911.

[48] Miedzinski, M. et al. (2020), Science, Technology and Innovation for SDGs Roadmaps, https://sdgs.un.org/sites/default/files/documents/269391_BP_Roadmaps_IntlCollaberation_final_7_09_20.pdf.

[58] Milkoreit, M. (2022), “Social tipping points everywhere?: Patterns and risks of overuse”, WIREs Climate Change, Vol. 14/2, https://doi.org/10.1002/wcc.813.

[80] Montague, C., K. Raiser and M. Lee (2024), “Bridging the clean energy investment gap: Cost of capital in the transition to net-zero emissions”, OECD Environment Working Papers, No. 245, OECD Publishing, Paris, https://doi.org/10.1787/1ae47659-en.

[52] New Zealand Ministry of Business, Innovation and Employment (2023), Which Analytical Tools Are Suited to Transformative Change?, New Zealand Government, Wellington, https://www.mbie.govt.nz/assets/Analytical-tools-suited-to-transformative-change-report.pdf.

[70] Novy, A., N. Barlow and J. Frankhauser (2022), Transformative innovation, Vienna University of Economics and Business, https://www-mlgd.wu.ac.at/sre-disc/sre-disc-2022_01.pdf.

[72] OECD (2025), “Ensuring a just transition to net-zero emissions”, OECD Net Zero+ Policy Papers, No. 15, OECD Publishing, Paris, https://doi.org/10.1787/a3da7080-en.

[84] OECD (2025), INNOTAX database, https://stip.oecd.org/innotax/ (accessed on  March 2025).

[57] OECD (2025), “Leveraging positive tipping points in the race to net zero”, OECD Net Zero+ Policy Papers, No. 12, OECD Publishing, Paris, https://doi.org/10.1787/50b737f8-en.

[30] OECD (2025), Measuring Science and Innovation for Sustainable Growth, OECD Publishing, Paris, https://doi.org/10.1787/3b96cf8c-en.

[2] OECD (2025), OECD Economic Outlook, Interim Report March 2025: Steering through Uncertainty, Vol. 2024/2, OECD Publishing, Paris, https://doi.org/10.1787/89af4857-en.

[10] OECD (2025), “Packaging and sequencing policies for more effective climate action”, OECD Net Zero+ Policy Papers, No. 10, OECD Publishing, Paris, https://doi.org/10.1787/fced4e28-en.

[3] OECD (2025), R&D spending growth slows in OECD, surges in China; government support for energy and defence R&D rises sharply, https://www.oecd.org/en/data/insights/statistical-releases/2025/03/rd-spending-growth-slows-in-oecd-surges-in-china-government-support-for-energy-and-defence-rd-rises-sharply.html?utm_source=chatgpt.com (accessed on 31 March 2025).

[41] OECD (2024), “Designing Effective Governance to Enable Mission Success”, OECD Science, Technology and Industry Policy Papers, No. 168, OECD Publishing, Paris, https://doi.org/10.1787/898bca89-en.

[71] OECD (2024), “Framework for Anticipatory Governance of Emerging Technologies”, OECD Science, Technology and Industry Policy Papers, No. 165, OECD Publishing, Paris, https://doi.org/10.1787/0248ead5-en.

[31] OECD (2024), “How do governments direct support for innovation?”, Lessons from recent OECD measurement and impact analysis (MABIS) work, https://www.oecd.org/en/publications/how-do-governments-direct-support-for-innovation_c1d93d1c-en.html (accessed on 30 March 2025).

[40] OECD (2024), Mission-Oriented Innovation Policies for Net Zero: How Can Countries Implement Missions to Achieve Climate Targets?, OECD Publishing, Paris, https://doi.org/10.1787/5efdbc5c-en.

[1] OECD (2024), “OECD Agenda for Transformative Science, Technology and Innovation Policies”, OECD Science, Technology and Industry Policy Papers, No. 164, OECD Publishing, Paris, https://doi.org/10.1787/ba2aaf7b-en.

[82] OECD (2024), “OECD Agenda for Transformative Science, Technology and Innovation Policies”, OECD Science, Technology and Industry Policy Papers, No. 164, OECD Publishing, Paris, https://doi.org/10.1787/ba2aaf7b-en.

[39] OECD (2024), “OECD Mission-Oriented Innovation Policy Online Toolkit”, web page, https://stip.oecd.org/moip (accessed on 14 March 2025).

[36] OECD (2023), Artificial Intelligence in Science: Challenges, Opportunities and the Future of Research, OECD Publishing, Paris, https://doi.org/10.1787/a8d820bd-en.

[77] OECD (2023), “COVID-19, resilience and the interface between science, policy and society”, OECD Science, Technology and Industry Policy Papers, No. 155, OECD Publishing, Paris, https://doi.org/10.1787/9ab1fbb7-en.

[49] OECD (2022), Inclusive Forum on Carbon Mitigation Approaches, OECD, Paris, https://www.oecd.org/climate-change/inclusive-forum-on-carbon-mitigation-approaches/?_ga=2.237167680.1185783980.1708330587-982265485.1706514215.

[34] OECD (2021), “Effective policies to foster high-risk/high-reward research”, OECD Science, Technology and Industry Policy Papers, No. 112, OECD Publishing, Paris, https://doi.org/10.1787/06913b3b-en.

[46] OECD (2018), Making Blended Finance Work for the Sustainable Development Goals, OECD Publishing, Paris, https://doi.org/10.1787/9789264288768-en.

[73] OECD (2015), Frascati Manual 2015: Guidelines for Collecting and Reporting Data on Research and Experimental Development, The Measurement of Scientific, Technological and Innovation Activities, OECD Publishing, Paris, https://doi.org/10.1787/9789264239012-en.

[8] OECD/UNDP (2025), Investing in Climate for Growth and Development: The Case for Enhanced NDCs, OECD Publishing, Paris, https://doi.org/10.1787/16b7cbc7-en.

[23] Ostry, J., A. Berg and C. Tsangarides (2014), Redistribution, Inequality, and Growth, International Monetary Fund, Washington, DC, https://www.imf.org/external/pubs/ft/sdn/2014/sdn1402.pdf.

[21] Planes-Satorra, S. and C. Paunov (2017), “Inclusive innovation policies: Lessons from international case studies”, OECD Science, Technology and Industry Working Papers, No. 2017/2, OECD Publishing, Paris, https://doi.org/10.1787/a09a3a5d-en.

[47] Samans, R. (2016), “Blending public and private funds for sustainable development”, in Development Co-operation Report 2016: The Sustainable Development Goals as Business Opportunities, OECD Publishing, Paris, https://doi.org/10.1787/dcr-2016-en.

[53] Schumer, C. et al. (2022), Methodology Underpinning the State of Climate Action Series, Systems Change Lab, https://doi.org/10.46830/writn.22.00064.

[59] Spaiser, V. et al. (2024), “Negative social tipping dynamics resulting from and reinforcing Earth system destabilization”, Earth System Dynamics, Vol. 15/5, pp. 1179-1206, https://doi.org/10.5194/esd-15-1179-2024.

[26] UKRI (2025), “Innovation Accelerator programme”, web page, https://www.ukri.org/what-we-do/browse-our-areas-of-investment-and-support/innovation-accelerator-programme (accessed on 14 March 2025).

[42] United Nations Department of Economic and Social Affairs (2024), Report of the Inter-agency Task Force on Financing for Development 2024, United Nations, https://doi.org/10.18356/9789213588635.

[62] Victor, D., F. Geels and S. Sharpe (2019), Accelerating the Low Carbon Transition: The Case for Stronger, More Targeted and Coordinated International Action, Brookings, https://www.energy-transitions.org/publications/accelerating-the-low-carbon-transition.

[4] Weinberg, A. (1963), “Criteria for scientific choice”, Minerva, Vol. 1, pp. 159-171, https://doi.org/10.1007/BF01096248.

[86] Yamashita, I. et al. (2021), “Measuring the AI content of government-funded R&D projects: A proof of concept for the OECD Fundstat initiative”, OECD Science, Technology and Industry Working Papers, No. 2021/09, OECD Publishing, Paris, https://doi.org/10.1787/7b43b038-en.

Notes

Copy link to Notes

← 1.  The OECD Agenda for Transformative Science, Technology and Innovation Policies is a product of a meeting of the Committee for Scientific and Technological Policy at Ministerial level held in April 2024 with the theme of “Enabling sustainability transitions through science, technology and innovation: Shared challenges and transformative actions”. Ministers affirmed the need for an ambitious agenda for transformative STI policies and values-based technology governance frameworks, and provided strategic directions for the future work of the Committee for Scientific and Technological Policy. The main outcome of the meeting was the adoption of the Declaration on Transformative Science, Technology and Innovation Policies for a Sustainable and Inclusive Future (hereafter the “Declaration”). Structured around four pillars, the Declaration makes a case for transformative STI policies to accelerate sustainability transitions while embedding shared values in the governance of science and emerging technologies, reaffirming the need for international co-operation in STI to address global challenges, and focusing on making STI more inclusive and strengthening the evidence base. The Declaration is made operational by two underpinning documents welcomed by Adherents to the Declaration: the OECD Agenda for Transformative Science, Technology and Innovation Policies (OECD, 2024[1]); and the OECD Framework for the Anticipatory Governance of Emerging Technologies (OECD, 2024[71]).

← 2.  The United States remains the largest absolute spender on R&D in the world, spending USD 956 billion in 2023 (measured in current PPP USD, the standard method for international macroeconomic comparisons). The pace of growth in R&D spending picked up in the United States around a decade ago, leading to a widening gap with the EU27 and other leading industrial economies, such as Japan. Using the same measure, the second-largest spender is China (USD 917 billion in 2023), whose R&D expenditures surpassed those of the EU27 a decade ago and are now 62% greater. On this measure, the gap in the level of R&D expenditure between China and the United States narrowed significantly in 2023, with China’s R&D expenditure 96% of the United States’, up from 72% ten years before. However, when measured in USD at market exchange rates (an alternative measure to constant PPP USD), the gap remains much wider, with China’s R&D expenditures 49% of those of the United States in 2023, up from 42% in 2013 (OECD, 2025[3]).

← 3.  The OECD Fundstat database (version: March 2024) comprises data on government R&D project funding in 19 OECD Member countries (Australia, Austria, Belgium, Canada, Czechia, Estonia, Finland, France, Germany, Japan, Ireland, Latvia, Lithuania, Norway, Portugal, Sweden, Switzerland, the United Kingdom and the United States) and the European Union from 2010 onwards. Administrative data on government R&D project funding offer potential for international comparative analysis (OECD, 2015[73]). While such data provide a foundation for accountability, stemming from the government funding processes that generate them, these records are often inconsistent across countries, complicating cross-national analysis. Despite limitations, project-level data can provide insights into the directionality of government R&D funds, with varying levels of quality and completeness. The Fundstat database is an evolving database overseen by the OECD Expert Group on the Management and Analysis of R&D and Innovation Administrative Data and has been used for the analysis of directionality for artificial intelligence R&D (Yamashita et al., 2021[86]) and COVID-19 R&D (Aristodemou et al., 2023[85]).

← 4.  For example, efforts to improve the sustainability of regional or national electricity grids have been supplemented in some countries by growing engagement of public consumers in the energy system as prosumers (i.e. consumers and producers) and the evolution of community-scale integrated heat and power networks (IEA, 2022[74]).

← 5.  The EC-OECD STIP Compass database incorporates a thematic portal on indigenous knowledge and communities which includes information on scores of initiatives from a wide range of countries. See: https://stip.oecd.org/stip/indigenous-portal.

← 6.  Job losses spurred by sustainability transitions have been found to be up to 24% more costly in high-emission sectors compared to low-emission ones due to tendencies for impacted jobs to be concentrated among older workers in relatively high-paying jobs compared to educational attainment (OECD, 2025[72]). The creation of new jobs has largely benefited high-skilled workers.

← 7.  OECD research estimates that boosting the performance of the average OECD country to the level of the top three country performers in the area of adult skill outcomes would generate a productivity increase of 17%. This could be achieved through policy to support skills accumulation, lifelong learning and movement of labour, among other types of initiatives (OECD, 2025[2]).

← 8.  For example, the IEA estimates that more than a third of the emissions reductions required in 2050 to achieve net zero scenarios will come from technologies that are still in the lab (IEA, 2023[75]). In other words, unless certain pre-commercial technologies are rapidly proven and scaled up, net zero is likely out of reach. The fact that the other emissions reductions in the scenario (i.e., unrelated to those pre-commercial technologies) are very hard to achieve without substantial cost reductions and performance improvements is not deducible from this number but is a key part of achieving overall net zero. Achieving net zero, therefore, requires a mix of new R&D and demonstration activities together with the deployment and diffusion of existing technologies. Moreover, the resulting sustainability and digital transitions go beyond the adoption of new technologies and include investment in new infrastructures, the establishment of new markets, the development of new social preferences, and support for people of working age and communities in attaining new skills and opportunities as part of “just green transitions” (Causa et al., 2024[76]). Non-technological innovations, including social and process innovations, among others, will also make important contributions.

← 9.  Decades of investment into fundamental research on mRNA largely facilitated the expedited development and deployment of mRNA-based vaccines during the COVID-19 pandemic response (OECD, 2023[77]). Further, mRNA vaccine platforms build on established vaccine technologies and have the potential to accelerate development and manufacturing processes without compromising on safety (Litvinova et al., 2023[78]).

← 10.  Given the business sector’s importance for innovation, not only do governments contribute to the costs of business RDI, but also must ensure to create innovation-friendly framework conditions that incentivise companies to spend more on R&D.

← 11.  Only a few countries have also used tax measures to provide directionality for R&D in specific priority areas, such as green or energy-related R&D. This includes Italy, which has a higher tax credit rate for certain types of technological innovation support for the environment, and Korea, which has an enhanced tax credit rate for R&D aimed at so-called “new growth and basic technologies” (273 technologies in 14 areas, including future cars, next-generation electronic information devices, energy and environment) and “National Strategy technologies” (OECD, 2025[84]).

← 12.  See: https://stip.oecd.org/moip.

← 13.  See: https://fdoc.ffg.at/s/vdb/public/node/content/-5t5IDKMRcieVsqbCMVbCQ/6.0?a=true.

← 14.  See https://ised-isde.canada.ca/site/strategic-innovation-fund/en/about-program/program-guide.

← 15.  See https://stip.oecd.org/moip.

← 16.  See https://stip.oecd.org/moip.

← 17.  See https://academic.oup.com/rev/article/32/1/86/6780169?rss=1.

← 18.  See https://www.thebritishacademy.ac.uk/news/promising-results-from-first-year-of-innovative-grant-awarding-trial.

← 19.  For example, in the Japanese cross-ministerial Strategic Innovation Programme, the powerful programme directors in each programme act as chairs of their respective promotion committees and are deemed essential for the promotion and smooth operation of inter-ministerial co-ordination and science-industry co-operation. In Norway’s Pilot-E, the programme manager oversees the secretariat of the steering group that gathers the three partner agencies. Another aspect of variation in these types of programmes concerns the types of individuals who manage the research portfolio. For example, Canada’s National Research Council’s Challenge Programmes generally use their own technical experts to manage portfolios. While they have had success in steering the research towards new ambitious areas (which is where their expertise lies), commercialisation has been more of a challenge – something they are looking at incorporating earlier on in their process. In contrast, the UK Research and Innovation Challenge Fund generally employs former industry leaders who found it more difficult to steer the research, but then have played a greater role in identifying market opportunities for teams to pursue at the commercialisation stage (OECD, 2024[41]).

← 20.  See www.sprind.org/en/impulses/challenges/articles/overview.

← 21.  See www.ukri.org/what-we-do/ukri-challenge-fund.

← 22.   See https://docs.google.com/document/d/19VSOwbxvtVuc67gkb_ZWWZVwaOkT3FgZp5pkPQDucLY/edit?tab=t.0.

← 23.  For instance, many MOIPs related to achieving net zero carbon emissions are targeted directly at the 2030 and 2050 aims of the Paris Agreement for Climate Change.

← 24.  See www.enova.no/pilot-e/information-in-english.

← 25.  See www.gov.ie/en/publication/27c78-impact-2030-irelands-new-research-and-innovation-strategy.

← 26.  See www.creatingourfuture.ie/2022/07/minister-harris-publishes-more-than-18000-ideas-generated-from-national-brainstorm-creating-our-future.

← 27.  See www.info.gouv.fr/organisation/secretariat-general-pour-l-investissement-sgpi/strategies-d-acceleration-pour-l-innovation

← 28.  See www.economia.gob.cl/2023/11/22/gobierno-destinara-6-mil-millones-de-pesos-para-investigacion-de-litio-y-salares.htm.

← 29.  The example of Sweden’s Strategic Innovation Programmes (SIPs) has shown that this can result in a strong role played by powerful players in mature industries to the detriment of the transformative potential of the missions. Each SIP is steered by an industry-dominated board, drawn from the industry, academia and the public sector, with industry generally being dominant. The final evaluation concluded that the SIPs mostly resulted in incremental innovation at the individual project level (Åström, 2021[79]).

← 30.  For example, in the green transition, challenges for private investors include insufficient profitability compared to investments with similar risk profiles; difficulty assessing risks owing to information asymmetries between innovators and investors; lack of awareness and uncertainty around government priorities and regulations; and challenges in meeting “internal rate of return” requirements or “return on equity” thresholds. These imperfections in capital markets limit the amount of private capital available for low-carbon technologies (Montague, Raiser and Lee, 2024[80]).

← 31.  Supply chain finance has emerged as a new tool to increase available internal capital in small firms, thereby encouraging investment in R&D and innovation activities. By combining the supply chain to build long-term stable trade relations, supply chain finance can also provide stable capital flow and lower financing costs for small and medium-sized enterprises. The supply chain-based financial model can provide short-term financial support for enterprises and meet their long-term funding needs.

← 32.  Governments have long supported the development of venture capital and private equity investments, but this support is increasingly targeted towards green transitions using a variety of models that vary in terms of the level of public ownership and direction over investment decisions (Berger, Criscuolo and Dechezleprêtre, 2025[81]).

← 33.  Analyses of the financing initiative reported in the STIP Compass database in 2023 show that some 17% of equity financing initiatives target the sustainability transition’s transformative goal. Most of these programmes support innovative start-ups and SMEs through the provision of seed and/or growth and late‑stage venture capital. Transformative equity-based financing is often administered through direct public equity funds or co-investment funds. There are comparatively few fund-of-fund initiatives. Most equity financing is also focused on net zero priorities. Some programmes are broader than this and aim to support various technology areas, such as cleantech, med-tech, precision agriculture and the circular economy. Roughly 25% of the credit/loan and debt/risk-sharing initiatives captured in the STIP Compass database relate to sustainability transitions. Relevant credit programmes most commonly target the development of new products and processes or the upgrade of existing ones. In general, these initiatives support innovation by offering loans with subsidised rates. There is comparatively less focus on providing working capital or financing an expansion to acquire existing technologies.

← 34.  The obstacles and challenges to STI finance are not uniform across the different SDGs due to differences in financial and market structures, differences in the capital intensity of research and industrial activities, and structural differences in the share of public and private R&D funding. For example, obstacles to finance STI for clean energy will differ from those faced in the financing of research for global health challenges. Obstacles to finance STI for clean energy would involve a greater share of business financing from firms’ internal sources as well as from equity and capital markets, whereas obstacles to finance STI for research for global health challenges will rely more on public research funding by governments.

← 35.  For example, progress on a range of issues – such as strengthening various linkages in STI systems (e.g. between business and academia, between different parts of government, and between science and society), enhancing firms’ skills and organisational capabilities, and reducing precarity in research careers – will contribute to STI system reforms that hasten progress on the transformative goals. Likewise, directing STI systems towards goals like resilience and inequality can facilitate progress on these long‑standing issues if transformation-friendly values are embedded in STI policymaking. Thus, the pursuit of the transformative goals provides an opportunity to promote structural reforms that address long‑standing issues in STI systems and vice versa (OECD, 2024[82]).

← 36.  Persistence or absorption are alternative strategies focused on maintaining the system’s structure and function by mitigating risk and resisting change (Béné et al., 2012[68]).

← 37.  The ability of a handful of lead countries to accelerate the global adoption of electric vehicles is often held up as an example of policy effectively leveraging this kind of change process. In this instance, economies of scale allowed for the generation of a self-perpetuating, non-linear cycle of technological advancement, cost reduction and learning effects (Eker et al., 2024[67]).

← 38.  Public policy has its own positive and negative feedback dynamics that are also relevant for STI policymakers. According to (Edmondson, Kern and Rogge, 2019[83]), these include resource, interpretative and institutional effects, as well as socio-political, administrative and fiscal feedback.

← 39.  Transformative change is likely to face resistance from influential coalitions interested in maintaining the status quo, as well as disadvantaged groups concerned about the negative impacts of radical change (Blühdorn, 2019[69]; Novy, Barlow and Frankhauser, 2022[70]). Dominant stakeholders generally have vested interests in maintaining established industries, technologies and practices and are often able to influence structural conditions, e.g. regulation, or mask the full costs of the status quo. Accordingly, transformation is generally spurred by exogenous pressure, which can arise from a slow-moving trend like demographic change or a sudden shock.

2. Reconfiguring scientific co-operation in a changing geopolitical environment

Abstract

Growing geopolitical tensions and intense competition on emerging critical technologies are reshaping international co-operation in STI. Recent national STI policies and strategies reflect this shift with their increasing attention to security-related concerns. Focusing on public research systems, this chapter describes how governments are aiming to enhance national research and technological capabilities as they seek greater strategic autonomy that promotes both their economic and national security. This includes a growing policy emphasis on dual-use STI, as well as research security measures to protect against unauthorised knowledge leakage and foreign coercion. Governments have also become more strategic in their international STI linkages, including in their science diplomacy measures, with a view to projecting their national interests. The chapter highlights various risks and opportunities these policies pose and proposes that governments pursue balanced STI securitisation policies that are proportional, precise in their targeting, and based on committed partnerships with scientists and businesses, as well as across government.

Key messages

• Rising geopolitical tensions, accompanied by growing strategic competition on emerging critical technologies, are contributing to the growing securitisation of science, technology and innovation (STI). This includes the public research system, which is the chief focus of this chapter.

• Governments are pursuing a mix of policies that contribute to this growing securitisation. First, they are increasingly implementing promotion policies that orient research and development (R&D) funding towards enhancing national and economic security, covering, among other things, dual-use initiatives to foster mutually beneficial links between civil and defence research.

• Second, governments are using protection policies that introduce restrictions on sharing research findings with dual-use potential, as well as recent measures to strengthen research security more broadly to avoid exposing sensitive research to risks that ultimately erode safety and trust.

• Finally, they are implementing projection policies that provide strategic direction to international STI relations, including science diplomacy initiatives that support research co‑operation with like-minded countries and strategic competitors.

• These policies imply some reconfiguration of international research relations. For example, policies that aim to promote economic and national security could involve pooling research resources with like-minded countries, while research security measures could exclude or discourage collaboration with countries that are not considered safe partners for international co-operation.

• Securitisation policies that restrict international research co-operation and mobility could have negative effects on research quality, innovation performance and economic competitiveness if applied overzealously. There are also risks that securitisation policies could fragment international STI linkages to such an extent that it undermines co‑operation on tackling global challenges.

• STI securitisation measures involve different parts of government but are closely related and should be strategically oriented and co-ordinated. To improve their co-ordination and prevent over-securitisation, policymakers should adopt governing principles to design and implement balanced STI securitisation policy mixes that are proportional to the risks and opportunities at hand; formulated and implemented in partnership with scientists, businesses and across government; and precise and agile in their targeting.

Introduction

Scientific discovery and technological innovation occur in an interconnected global ecosystem that draws upon collective knowledge, talent, resources and infrastructure. Countries individually benefit from this international connectedness, which contributes to their competitiveness and societal well-being. Such connectedness is also critical for tackling challenges and managing risks at the global level, such as pandemic preparedness, environmental stewardship and food security, which require multilateral co‑operation in STI.

Rising geopolitical tensions, accompanied by growing strategic competition on emerging critical technologies, are reshaping frameworks for international STI co-operation that have emerged over the last three decades. These tensions and competition undercut opportunities for cross-border knowledge exchange, collaborative STI projects and technology transfer, while national interests are routinely framed as trade-offs with global priorities. These developments impact everything from international research collaboration to international trade and investments in high-technology products and facilities.

The 2023 edition of the OECD Science, Technology and Innovation Outlook introduced the concept of “STI securitisation”1 to discuss these trends (OECD, 2023[1]). This chapter continues along similar lines, focusing chiefly on the research aspects of STI systems and the impacts of growing securitisation on international research linkages. It consists of three main parts. The first part presents selected statistics on international scientific linkages, as measured by international research collaboration, researcher mobility and the scientific contributions of different countries to tackle global challenges, to highlight how these have evolved in recent years.

The second part of the chapter outlines the growing securitisation of STI policy, with a particular focus on newly intensive policy efforts towards achieving mastery and greater strategic autonomy in emerging science and technology in support of economic and national security objectives; the growing use of research security measures to protect against unauthorised knowledge leakage; and the increasing prominence of national interests in international science diplomacy.

These three sets of STI securitisation measures are closely related and imply an emerging reconfiguration of international research linkages. Accordingly, the third part of the chapter proposes a set of governing principles policymakers could adopt to design and implement balanced STI securitisation policy mixes that are proportional to the risks and opportunities at hand; formulated and implemented in partnership with scientists, businesses and across government; and precise and agile in their targeting.

Recent trends in international STI co-operation2

International STI linkages have grown strongly since the 1990s to the benefit of research, innovation and economic development. Among these linkages is international research co-operation, which benefits the quality of research which, in turn, contributes to economic competitiveness through new knowledge generation and enhanced skills development. International research co-operation also broadens the dissemination of research results, helps tackle global challenges and can contribute to intercultural understanding.

In times of heightened geopolitical tensions, it is important to understand these implications for international research linkages. This section presents selected statistics on international research collaboration, researcher mobility and different countries’ scientific contributions to tackle global challenges to highlight how international research linkages have evolved in recent years.

Growth in international research collaboration has recently stalled

Collaborative research is at the core of an interconnected global research community. Data on co‑authorship of scientific publications involving authors with institutional affiliations in different countries provide an indication of international scientific collaboration.3 While only 2% of scientific papers had authors from more than one country in 1970 (Olechnicka, Ploszaj and Celinska-Janowicz, 2019[2]), the proportion was 27% of all publications in OECD countries in 2023, up from 22% in 2013 (Figure 2.1). The United Kingdom has the highest collaboration intensity within the top 15 science publishing economies, followed by Australia and France. The leading Asian economies exhibit lower than average international collaboration. Australia, Brazil, India and the United Kingdom experienced the largest proportional increase in collaboration intensity over the period 2013-2023.

Figure 2.1. International scientific collaboration intensity, selected countries, 2013 and 2023

As a percentage of domestically authored publications, based on fractional counts

Source: OECD calculations based on Scopus Custom Data, Elsevier, Version 1.2025, April 2025.

Statlink

More recent data, however, suggest that the trend towards increasing international collaboration has lost momentum and might be partly breaking down (Figure 2.2). The external collaboration rate for the United States and the EU27 area has remained virtually unchanged since 2018, while the People’s Republic of China’s (hereafter “China”) international collaboration intensity declined significantly between 2020 and 2023, with India surpassing it in 2021. The growing scale and advancement of China’s research system mean there are more opportunities than ever to collaborate domestically with leading research groups, which could reduce incentives for international collaboration.4 However, as Figure 2.3 shows, this decline is largely driven by a sharp fall in collaboration with the United States. It applies across most research fields and is particularly pronounced in the natural sciences and engineering, as shown in Figure 2.4. Similar data covering China’s collaboration with other countries show some decline in a few fields with Japan and the United Kingdom, but continuing strengthened ties with the EU27. Despite these declines, the intensity of China’s research collaboration with the United States remained considerably higher in 2023 than with these other countries.

Figure 2.2. Trends in international scientific collaboration, selected countries, 2013-2023

Percentage of scientific publications involving international collaboration, based on fractional counts

Source: OECD calculations based on Scopus Custom Data, Elsevier, Version 1.2025, April 2025.

Statlink

Figure 2.3. China’s bilateral collaboration intensity trends in scientific publications, 1996-2023

Normalised collaboration based on whole counts

Notes: The bilateral collaboration intensity between two countries is calculated by dividing the number of scientific publications by authors with affiliations in both countries (whole counts) by the square root of the product of the publications for each of the two countries (whole counts). This indicator is, therefore, normalised for publication output. Publications refer to all citable publications, namely, articles, reviews and conference proceedings.

Source: Calculations based on Scopus Custom Data, Elsevier, Version 1.2025, April 2025.

Statlink

Figure 2.4. Changes in collaboration between China and the United States, 2019-2023

Percentage change in year relative to 2019 baseline

Notes: Collaboration between China and the United States is defined by the number of co-authored publications between both countries (whole counts). Publications refer to all citable publications, that is articles, reviews and conference proceedings. The graph shows the changes in collaborations for each year versus the previous year, as a percentage of 2019 collaborations.

Source: Calculations based on Scopus Custom Data, Elsevier, Version 1.2025, April 2025.

Statlink

OECD science systems depend on a ready supply of internationally mobile researchers

International scientific mobility has also grown in recent decades and the research workforce of several major research performers in OECD countries is heavily dependent on foreign-born doctoral and postdoctoral researchers. In the United States, for example, some 45% of workers in science and engineering occupations at the doctorate level in 2021 were foreign-born, with the highest shares among computer and mathematical scientists. More than half of foreign-born workers in the United States in 2021 whose highest degree is in a science and engineering field were from Asia. The leading birthplaces were India (29%) and China (13%) (US National Science Foundation, 2024[3]).

Early-career researchers conduct much of the research carried out in OECD Member countries’ laboratories. Although internationally comparative statistics are difficult to come by,5 many of these researchers are internationally mobile. They go abroad to enhance their qualifications, access world-class research facilities and improve their career prospects.6 The OECD’s education statistics show that Australia, Austria, Belgium, Luxembourg, New Zealand, Switzerland and the United Kingdom have particularly high shares of international doctoral student graduates – at least 40% of their total number – as their universities attract global talent through scholarships, research opportunities and strong academic networks (Figure 2.5). Moreover, these proportions grew markedly between 2015 and 2022, with the exception of Luxembourg, the United Kingdom and the United States, where they have remained the same. In some countries, the high proportion of international doctoral students also reflects declining interest in pursuing a PhD among domestic students (OECD, 2025[4]). In France, for instance, factors such as long periods of study, uncertain career prospects and more attractive opportunities in the private sector are reported to have made doctoral studies less appealing for national candidates.7

Figure 2.5. Share of mobile PhD graduates, selected countries, 2015 and 2022

As a percentage of total PhD graduates

Note: Mobile doctoral students correspond to students in PhD programmes (ISCED level 8) enrolled in a country different from the one where they obtained their previous qualification, including homecoming nationals. Internationally harmonised data for the United States are unavailable.

Source: OECD Education Statistics Database Education access, participation, and progression | OECD (accessed on 18 July 2025).

Statlink

As Figure 2.5 shows, significant proportions of these mobile doctoral graduates are from China, particularly in English-speaking countries (internationally harmonised data for the United States are unavailable), and in neighbouring Japan and Korea. Restrictions on international mobility as part of growing securitisation measures could weaken this important source of researchers and oblige countries to look elsewhere to attract global talent.

OECD countries need to tap into widely distributed knowledge to tackle global challenges

International scientific collaboration is particularly important in research relevant to energy security and environmental sustainability. Compared to all other areas of science, sustainability and energy-relevant research is more collaborative. Furthermore, this international collaboration has increased over time in virtually all countries (Figure 2.6). At the same time, there have been major changes in the contribution of the largest global economies to energy- and environment-relevant research output. The United States and the European Union have seen large declines in the share of relevant publications while China’s share has increased rapidly and India has also seen a steady increase (Figure 2.7). This implies a reduction in the overall relative contribution of OECD countries to scientific output in this area, over and above the general scientific publication shift that has been taking place (OECD, 2025[5]).8 It also highlights the importance of international openness and exchanges that allow OECD countries to tap into this more widely distributed knowledge.9.

Figure 2.6. International collaboration intensity in energy and environment SDG-relevant scientific output, select countries, 2012 and 2022

As a percentage of domestically authored documents, fractional counts

Notes: SDG: Sustainable Development Goal. International collaboration refers to publications co-authored among institutions in different countries. Estimates are computed for each country by counting documents for which the set of listed affiliations includes at least one address within the country and one outside. Single-authored documents with multiple affiliations in different countries count as institutional international collaboration. A publication is tagged as relevant to environmental sustainability and energy if it has the highest aggregated probability for the SDGs under the “Planet” umbrella (6, 12, 13, 15 and 7).

Source: OECD calculations based on Scopus Custom Data, Elsevier, Version 1.2024.

Statlink

Figure 2.7. Trends for main contributors to energy and environmental SDG scientific publications, 2008‑2022

As a percentage of world total energy and environmental SDG documents

Notes: SDG: Sustainable Development Goal. A publication is tagged as relevant to environmental sustainability and energy if it has the highest aggregated probability for the SDGs under the “Planet” umbrella (6, 12, 13, 15 and 7).

Source: OECD calculations based on OECD SDG classifier (OECD, 2025[6]) and Scopus Custom Data, Elsevier, Version 1.2024.

Statlink

The growing securitisation of science, technology and innovation

Concepts such as “strategic autonomy” and “technology sovereignty” have emerged as increasingly prominent frames for STI policy (Edler et al., 2023[7]; OECD, 2023[1]).10 This orientation extends beyond technology to cover research as well: for example, growing concerns over safeguarding national and economic security and protecting freedom of enquiry have led many OECD countries to develop guidelines and checklists to increase awareness of and provide guidance to the academic community on research security and integrity. Individual countries are also moving towards more selective international knowledge sharing, enhancing co-operation with countries that have similar values and political interests, particularly in STI areas with national security implications.

The 2023 OECD Science, Technology and Innovation Outlook discussed at some length the growing securitisation of STI policy.11 It introduced a three-part framework – promotion, protection and projection policies – to map the policy responses of China, the European Union and the United States to growing geopolitical tensions and increasingly intense technological competition (OECD, 2023[1]). This framework has its origins in the policy analysis literature (see, for example Helwig, Sinkkonen and Sinkkonen (2021[8]); March and Schieferdecker (2021[9]); Goodman and Robert (2021[10])) and has recently been adopted by policymakers in the European Union, which used it to structure the European Union’s Economic Security Strategy (European Commission, 2023[11]), and by the Japanese government, which used it to articulate its economic security policies (METI, 2024[12]). Both policies are further described below. The framework’s advantage lies in the comprehensive picture it provides of the securitisation landscape that policymakers can use to design and deliver more joined-up and aligned policies across a range of areas.

This chapter uses this three-part framework to consider STI policy developments that focus predominantly on the research system (Figure2.8):12

1.Promotion policies, including directed R&D funding and broader industrial policies to enhance national and economic security, covering, among other things, dual-use initiatives to foster mutually beneficial links between civil and defence research.

2.Protection policies, including restrictions on sharing research findings with dual-use potential and recent measures to strengthen research security more broadly.

3.Projection policies, including international strategic research co-operation that further advances protection and promotion, as well as science diplomacy initiatives that support research co‑operation with a wider range of countries, including strategic competitors.

Each of these is discussed in more detail below.

Figure 2.8. Three-part science, technology and innovation securitisation policy framework

Note: The chapter’s focus is mostly “upstream” on research

Leveraging the civil research system to enhance economic and national security

Technological leadership has long underpinned the economic prosperity and security of OECD countries, and with geopolitical tensions on the rise, governments are prioritising technological mastery and strategic autonomy as part of their broader national and economic security policies (OECD, 2023[1]). The first type of STI securitisation policy intervention therefore concerns the promotion of critical research and technology capabilities, for example through directed R&D funding that serves economic and national security needs.

Along these lines, recent years have seen a proliferation of national strategies targeting the development of a few critical technologies, where countries primarily aspire to capture their economic benefits. For example, quantum science and technology promises to reshape computing, communication and problem-solving in fundamentally new ways (OECD, 2025[13]), and around the world, governments, leading research institutes and some of the best-known technology companies are investing billions of dollars in quantum research (Box 2.1).

Box 2.1. National quantum policies and strategies

Quantum technologies have become a matter of geopolitical contention, echoing dynamics seen in other dual-use emerging technologies such as artificial intelligence and semiconductors. Their civilian and defence uses make quantum technologies dual-use. Over 30 countries have formulated substantial policies in support of developing quantum technologies, and among them, 14 OECD countries have adopted comprehensive strategies and committed substantial public investments to quantum research and development.

Defence establishments have been pivotal in developing quantum technology policy. The “quantum race”, therefore, is not merely a pursuit of technological supremacy but a crucial dimension on global security and power. At the same time, economic imperatives have complemented security concerns in driving quantum strategy development, particularly with reference to classical computing’s physical limitations.

Apprehensions about dependency on external actors for critical components have led to explicit mandates to develop domestic supply chains, protect intellectual property and cultivate in-house expertise. The aspiration to capture the transformational potential of quantum technologies underpins strategic efforts to foster diverse innovation ecosystems. These efforts blend top-down leadership (such as national-level funding initiatives and roadmaps) with bottom-up ecosystem-building (through incubators, consortia and test beds). Strategies are often accompanied by substantial investments in research institutes, infrastructure and interdisciplinary hubs.

Source: OECD (Forthcoming[14]).

Enhancing economic and national security is increasingly a main objective of science, technology and innovation policy

Governments are also embarking on more ambitious forms of holistic industrial policy (Dechezlepretre, Diaz and Lalanne, 2025[15])13 in which STI policy plays a prominent part. These policies increasingly target ecosystems that transcend traditional industrial sectors and knowledge domains (see Chapter 6).14 However, what clearly distinguishes the most recent initiatives from those of just a few years ago is their securitisation perspective and inclusion of strategic autonomy as a key consideration. While different aspects of security – including energy, health and food security – have received growing attention, enhancing national security is increasingly entwined with economic security as a main STI policy objective. Box 2.2 outlines the European Union’s recent related measures, which are largely framed by the 2023 European Economic Security Strategy.15

Box 2.2. The European Economic Security Strategy

Published in 2023, the European Economic Security Strategy (European Commission, 2023[11]) aims to provide a framework for robust assessment and management of risks to economic security at the European Union (EU), national and business level at a time when these risks are both evolving rapidly and merging with national security concerns. It cites as a prime example of risk the speed with which critical new technologies are emerging and blurring the boundaries between civil and defence applications.

The European Economic Security Strategy uses the “3Ps” framework to propose a raft of policies that include promotion measures that foster the research and industrial base in strategic areas such as advanced semiconductors, quantum computing, biotechnology, net zero industries, clean energy or critical raw materials; protection measures to prevent the leakage of sensitive emerging technologies, as well as other dual-use items, to destinations of concern that operate civil-military fusion strategies; and projection measures (referred to as “partnering”) with countries who share common concerns on economic security and by investing in sustainable development and secure links throughout the world through the European Union’s Global Gateway initiative.

The risks related to technology security and technology leakage are being assessed based on a list of strategic technologies critical for economic security. At the same time – and reflecting the duality of perspectives on dual-use as an issue of concern for protection and promotion – a recent white paper outlined options to enhance targeted support for research and development (R&D) of dual-use technologies (European Commission, 2024[16]). This acknowledges that technologies used in the context of security and defence capabilities increasingly originate in the civilian domain, where private sector investments are higher, indirect costs lower and R&D cycles faster.

Taking up this theme, the Draghi Report on the future of European competitiveness asserts that dual‑use programmes are insufficiently developed in the European Union,¹ despite their potential to enhance collaboration between civilian and defence sectors, drive deep technical innovation that also addresses military needs, and mitigate risk by leveraging common technologies across different end uses (European Commission, 2024[17]). The Niinistö Report on strengthening Europe’s civilian and military preparedness and readiness, also published in 2024, highlights the need for Europe to better harness the much faster civil innovation cycles for technologies with dual-use potential to maintain its competitiveness and enhance military capabilities (European Commission, 2024[18]). Both reports compare EU expenditures on defence R&D unfavourably to similar investments by the United States.

In March 2025, the European Commission published a new white paper, European Defence Readiness 2030, which declares that technology will be the main feature of competition in the new geopolitical environment. It proposes new ecosystems and value chains for cutting-edge technologies, like artificial intelligence and quantum computing, to feed into both civilian and military applications. It highlights the blurred distinction between civil and defence research, particularly in the realm of “deep tech”, and affirms that relevant civil research results should play a crucial role in developing cutting-edge military capabilities (European Commission, 2025[19]).²

In July 2025, the European Commission published initial proposals for its tenth Multiannual Framework Programme (2028-2034), which for the first time embraces a “dual-use-by-design model” across much of its entirety (European Commission, 2025[20]).³ One of its priorities is the reduction of critical dependencies on non-EU technologies and a focus on dual-use technologies that can benefit both civil and defence applications. The European Commission proposes that large parts of the tenth Multiannual Framework Programme will be structured around the four policy “windows” of the European Competitiveness Fund, including one dedicated to “resilience and security, the defence industry and space”.

Notes: 1. This is not a new theme for the European Commission, which considers investment in innovation and better use of civilian technology in defence as key to enhancing Europe’s technological sovereignty and reducing its strategic dependencies. The European Commission published the “Action Plan on Synergies between Civil, Defence and Space Industries” in 2021, which aims to enhance complementarities between EU civil and defence programmes and instruments, promote “spin-offs” from defence and space R&D for civil applications, and facilitate “spin-ins” of civil-driven innovation into European defence co-operation projects. It followed this up in 2022 with its “Roadmap on Critical Technologies for Security and Defence”. Established in 2021 with a budget of EUR 8 billion over 7 years, the European Defence Fund promotes R&D co-operation between public research (typically research and technology organisations rather than universities) and firms. It supports competitive and collaborative projects throughout the entire R&D cycle, including design, prototyping and testing. 2. This white paper also integrates a series of actions to take immediately, such as activation of the National Escape Clause by member states, adoption of the Regulation on Security and Action for Europe, launch of the Strategic Dialogue with the defence industry, and a European Armament Technological Roadmap on investment into dual-use advanced technological capabilities in 2025. 3. The European Commission also published two independent expert reports on dual-use research and innovation (R&I) in June 2025 to inform decision making for the next generation of EU funding programmes. The first is a policy brief prepared by the Expert Group on Economic and Societal Impact of Research and Innovation that highlights the strategic role of dual-use R&I funding to foster security, strategic autonomy, competitiveness and sustainability, and recommends the adoption of a dual-use-by-design approach in future EU funding programmes (European Commission, 2025[21]). The second report, prepared by a small group of experts from both the civil and defence R&I communities, offers concrete examples and case studies on how dual-use R&I can work in practice (European Commission, 2025[22]).

In another example, the promotion of specific critical technologies in Japan’s Economic Security Promotion Act (2022) points in similar directions (Box 2.3).

Box 2.3. Japan’s economic security policy

Japan has been a pioneer in economic security policy with a view to minimising its vulnerability to economic coercion. Its main policy is the 2022 Economic Security Promotion Act, which sets out to enhance Japan’s strategic autonomy and pursue “strategic indispensability” through superior, made-in-Japan technologies on which the rest of the world depends. The act defines four pillars of action: 1) supply chain resilience; 2) securing essential infrastructure; 3) developing cutting-edge critical technologies; and 4) non-disclosure of patents. The National Security Secretariat acts as a co‑ordinating body for these efforts and an Economic Security Promotion Office has been established in the Cabinet Office to help unify economic security policies and ensure coherency across all government ministries and agencies.

In addition to these government-wide initiatives, individual ministries are also undertaking their own efforts to enhance economic security. For example, in 2023, the Ministry of Economy, Trade and Industry (METI) published its Economic Security Action Plan, which is centred on the “3Ps”, defined in Japan (as in the European Union) as promote, protect and partner. METI has also established a Trade and Economic Security Bureau to formulate and promote policy related to economic security within its jurisdiction. The Action Plan was updated in 2024 and again in May 2025 (METI, 2025[25]), with the most recent version highlighting the need for Japan to adapt to changing conditions in the international system since the previous edition. The 2025 edition outlines four sets of measures:

1.Further integration and updating of the 3Ps to strengthen Japan’s industrial and technological bases. Among the announced measures are an expanded list of promoted critical technologies that cover various new materials and technologies critical for economic and national security; a new industrial value chain approach that further strengthens measures at each layer of the entire value chain, including research and development, procurement, production, and sales; and a more strategic approach to attracting and retaining highly skilled foreign talent to promote Japan’s strategic autonomy and indispensability and prevent the unintended leakage of technology.

2.Toward the reconstruction of a rules-based international economic order as a global public good. This includes promoting the “Run Faster Partnership” scheme, which integrates industrial promotion and industrial protection measures aimed at co-creating industrial and technological bases with like-minded countries (with an initial focus on the Indo-Pacific region); and a more proactive role in international strategic rule-making and standardisation.

3.Promotion of public-private dialogue. This includes establishing a multi-layered public-private dialogue mechanism; developing codes of conduct and guidelines as references for firms contributing to Japan’s economic security; and supporting firms in establishing information security systems.

4.Strengthening economic intelligence. In addition to existing scenario analysis, tabletop exercises, supply chain analysis and technology analysis, METI will recruit external experts with high expertise, e.g. in specific technology and industrial areas; establish a strategic dialogue platform between private sector think tanks and the government; and, together with the National Security Secretariat and Cabinet Office, establish an economic security centre to enhance the government’s economic intelligence capabilities.

Notes: The updated 2025 Action Plan highlights four “tectonic shifts”: 1) the erosion of the liberal international economic order; 2) the intensification of competition for technological hegemony (centred on artificial intelligence) between China and the United States; 3) the growing importance of energy security; and 4) intensifying competition in next-generation strategic fields (e.g. space, oceans and unmanned aerial systems) essential to national security among major but also emerging powers.

Sources: NSS (2022[23]); METI (2024[12]; 2025[25]); Suzuki (2023[24]).

Among the main elements the act identified are Japan’s technological capabilities framed as economic measures related to ensuring national security. What marks a sharp departure from the past, though, is a new R&D initiative based on the act called the “K Program” (the Program for the Development of Key Technologies for Economic Security). This is a joint initiative of the Cabinet Office; the Ministry of Education, Culture, Sports, Science and Technology; and the Ministry of Economy, Trade and Industry (METI) that focuses on technologies that contribute to securing national economic security in domains such as maritime, aerospace and cyberspace. The K Program currently has a budget of up to JPY 500 billion (EUR 3 billion) over ten years (NSS, 2022[23]; Suzuki, 2023[24]).16

There is renewed interest in promoting dual-use research as economic and national security agendas become increasingly entwined17

These short accounts of emerging European and Japanese economic security policies highlight governments’ expectations of the far-reaching consequences of emerging technologies like artificial intelligence (AI) and quantum computing, including for national security. They also point to governments’ renewed interest in promoting dual-use R&I – involving both the civil and defence sectors – to foster economic and national security. As outlined in Chapter 1, governments are looking to actively exploit synergies between policy goals, including security and economic competitiveness, to ensure maximum return on and efficiency of STI investments. With defence budgets growing considerably in many OECD countries, including in R&D (see Chapter 1), governments are keen to leverage these new expenditures for civil purposes, too. It also seems likely that some civil R&D will be partly labelled as contributing to defence and security as countries aim to meet ambitious defence spending targets over the coming decade. Both phenomena contribute to the emergence of more explicit dual-use agendas.

While dual-use ambitions can be realised through multiple channels, two points of policy focus are emerging. The first focuses on ways to better anticipate both civil and defence needs when conducting low technology-readiness level (TRL) research related to general purpose technologies, such as AI and quantum computing, even when the field of application is not yet known. Many general-purpose technologies are inherently dual-use, and an approach that embodies a “dual-use-by-design model”, as discussed in the context of the European Union’s next framework programme (see Box 2.2), would aim to raise the awareness and reflection of researchers, administrators and funders on the potential end uses of their research. Such increased awareness and reflection would seek to promote early detection of the dual‑use potential of scientific output, with a view to enhancing understanding of both the risks and opportunities (European Commission, 2025[21]). It would also aim to promote a simultaneous alignment with civil and defence requirements, thereby minimising the modifications required to align a given technology with civil or defence standards when targeting respective markets (European Commission, 2025[22]).

The second point of policy focus concerns strengthening technology transfer between civil and defence applications at higher TRLs. While many governments have long supported two-way linkages between the civil and defence R&I systems, rapid and disruptive technological developments in the civil sector have seen governments pay growing attention to their dual-use potential in the defence sector. Accordingly, defence research funding programmes are increasingly opening and commissioning research from the civil research system.

Links between civil and defence research have been historically stronger in some countries than in others. For example, the United States has had a long-standing relationship between civil and defence research as a core feature of national security and science policy. The Department of Defense is an important funder of basic research in universities and support for doctoral programmes in a range of fields. Organisations such as the Defense Advanced Research Projects Agency have sponsored path-breaking research and facilitated new scientific networks, drawing on leading university scientists as programme managers and researchers (see Chapter 1).18 By contrast, Germany and Japan have historically maintained strict separation between civil and defence research. For example, Germany’s “civil clause” excludes most public universities from defence-related research. This is currently under review, with the Federal Ministry of Research, Technology and Space discussing with other research funders the extent to which funding incentives can be used to increase co-operation between civilian and defence research in appropriate areas.19 In other countries, such as Australia, France and the United Kingdom, while some universities and civil public research institutes have a long history of working with the defence sector, linkages are less developed and systemic than in the United States.

Despite this history of linkages, the civil and defence research systems remain relatively distinct and somewhat independent of one another, having, for instance, their own lead ministries, funding bodies and programmes, research centres and infrastructures, and rules and regulations on what knowledge can and cannot be openly shared. Defence R&I ecosystems remain relatively closed compared to their civil counterparts and are still highly nationally organised. But as economic and national security agendas become more closely entwined, there is growing convergence in the design of funding and other public policy interventions that support civil and defence research and technology development. This could signal the emergence of a more integrated R&I system that sees civil research organisations and researchers further contribute to and exploit defence and security research.

Dual-use research raises several practical and more fundamental questions for civil research systems and policies

Dual-use R&I can be subject to extensive and complex export control compliance measures, including dual-use export control regulations, that introduce additional administrative overheads, significantly prolong development cycles and impose higher costs. Secure development environments with high‑security zones and restricted access may also need to be established, implying changes in the organisation of the campus, research teams, data management and IT systems, among other things. However, academic basic research has traditionally been exempt from many of these restrictions. For example, in the United States, the Department of Defense funds considerable research at universities that is inherently dual-use but also unclassified. It tends to be at the application level that distinctions are made between civil and defence uses, and separations are put in place to protect secrecy on the defence side. This distinction could become blurred if R&D funding programmes that are notionally civil become “dual-use-by-design” and target low TRL research that must already consider a range of uses, including for defence.

Talent constraints are another important challenge due to the scarcity in many OECD countries of professionals with both technical expertise and the required security clearances. Leading universities are international in their staffing, and in many systems, foreign doctoral students and postdocs play key roles. Obtaining security clearances can be a long and cumbersome process when hiring foreign researchers and doctoral students, and certain nationalities are likely to be excluded in some contexts. The classification of knowledge as sensitive or classified will also restrict its open dissemination, which could discourage early-career researchers who depend on open publication for their career progression (see Chapter 4). Finally, ethical considerations might also limit scientists’ availability and acceptance to engage in research with potential military applications (European Commission, 2025[22]).

At the same time, international competition for leading scientists has become more fierce

Despite possible future restrictions on hiring certain foreign researchers in certain fields, attracting international talent, including leading scientists, remains an important approach for countries to bolster research and technical capabilities that underpin their economic and national security. Those that do not join the global competition for highly skilled migrants risk falling behind (OECD, 2023[26]). OECD countries have for some time offered different types of incentives to attract leading scientists from abroad, including fellowships, grants, tax breaks, special visas and pension portability.20 Among these measures are talent programmes, which target leading overseas scientists with financial incentives and entry and settlement support. These programmes have become increasingly popular in recent years and often target specific areas of science and technology where countries are seeking to deepen their capabilities. Box 2.4 provides selected examples of recent initiatives.

Box 2.4. Examples of recent talent programmes to attract overseas scientists

The European Union and its member states offer a variety of funding opportunities open to researchers outside Europe. In May 2025, the European Commission announced a EUR 500 million package for 2025-2027 to attract and retain researchers based outside the European Union (EU). It includes the “Choose Europe for Science” initiative, which was launched in 2025 to attract and retain top research talent globally. It also includes the European Research Council Advanced Grant, which provides additional support to researchers moving from non‑EU countries, who can apply for an additional EUR 2 million to cover eligible start-up costs (European Commission, 2025[27]).

At the EU member state level, in 2023 the Spanish Agencia Estatal de Investigación established the ATRAE programme, which awards grants to recruit established, internationally recognised research talent (among the top 10% of global researchers in their field) who have recently spent a significant period of their professional activity abroad. The 2025 call is worth EUR 40 million, with individual grants of up to EUR 1.2 million each. Their purpose is to promote progress toward a more competitive science, technology and innovation system at both the national and international levels.1

Germany’s Federal Ministry of Research, Technology and Space launched the “Global Minds Initiative” in 2025 directed at excellent international researchers. The initiative builds on programmes of the Alexander von Humboldt Foundation and the German Research Foundation and aims to signal a culture of welcome in Germany and to offer a safe haven for scientific freedom. Funding is based on scientific excellence and is open to all themes.²

France launched the “Choose France for Science” platform in 2025 as part of its commitment to welcome international researchers who wish to work in an environment conducive to academic freedom. Operated by the French National Research Agency, it enables universities, schools and research organisations to apply for up to 50% co-funding from the government to host researchers.³

Also in 2025, the Swedish Research Council issued a call for grants of SEK 2 million to enable Swedish higher education institutions and other research organisations to recruit prominent researchers who are active outside Europe. The grant is to cover expenses for the recruitment and salary for employment in Sweden during a limited period.⁴

“Science Hub Denmark” is a nationally co-ordinated initiative aimed at enhancing the global visibility of Danish research and career opportunities in natural sciences, engineering and life sciences. It promotes Denmark as an attractive destination for top-tier international researchers, with a strong focus on excellence, societal impact and quality of life.⁵

Beyond Europe, Korea launched its “K-Tech Pass” in 2025 to attract global talent in advanced industries, including semiconductors, secondary batteries, displays, biotechnology, robotics and the defence sector. The scheme offers both entry and settlement support to foreigners with expertise in advanced technologies who have signed an employment contract with Korean firms in high-tech industries.⁶

China has scores of talent programmes operating at the national, regional and city levels. They mostly focus on attracting students and professionals from the Chinese diaspora to return to China. Perhaps the best known and one of the largest was the “Thousand Talents Programme”, which operated between 2008 and 2023 and is estimated to have attracted 7 000-8 000 participants. The part-time version of the programme attracted close scrutiny from several OECD Member countries, since programme participants moved back and forth and often set up a laboratory in China that mirrored their research lab in the OECD, thereby promoting knowledge and technology transfer on a regular basis (Barteau and Rovito, 2024[28]). The programme contributed to growing research security concerns in OECD countries,⁷ particularly around conflicts of interest and conflicts of commitment among participating scientists.

Notes: 1. For further information, see: https://www.aei.gob.es/en/node/5066. 2. For further information, see: https://www.bmftr.bund.de/EN/Research/ScienceSystem/global-minds-initiative-germany/global-minds-initiative-germany.html?nn=1102680. 3. For further information, see: https://france2030.agencerecherche.fr/ChooseFranceForScience-2025/accueil.php?lang=EN. 4. For further information, see: https://www.vr.se/english/just-now/news/news-archive/2025-04-02-new-grant-for-recruiting-researchers-active-outside-europe.html. 5. For further information, see: https://research.state-of-denmark.com/about.

1. For further information, see: https://www.kotra.or.kr/gtc_eng/subList/41000060003. 7. For example, the 2022 CHIPS and Science Act prohibits US-based researchers with federal funding from participating in foreign talent recruitment programmes sponsored by China or the Russian Federation.

Protection through research security measures

There is a growing concern about hostile actors that exploit international research collaboration to acquire research and expertise to accelerate their technological capabilities in areas critical to national and economic security. Without attention and effective management, there is anxiety that such actions may have implications for national security, economic competitiveness and the integrity of research collaboration (James et al., 2025[29]). Many OECD countries now consider unauthorised information transfer and foreign interference in public research a serious national and economic security risk, and research security, including preventing undesirable foreign state or non-state interference in fundamental and applied scientific research, has become a high priority in STI policy (OECD, 2022[30]).

While countries have well-regulated export control systems for research on chemical, biological, radiological, nuclear and explosive technologies, it is less easy to control the intangible transfer of data, information and know-how from scientific research carried out without a specific practical application in mind. This is the case for basic research, which has traditionally been exempt from export controls. At the same time, it is recognised that knowledge from many areas of basic research could be considered as potentially dual-use and, as highlighted above, policymakers are now considering ways to raise awareness among researchers and funding agencies to take this perspective into account in low TRL research. This has led to closer scrutiny of international scientific collaborations and publication practices that were previously liberal, with entire scientific fields, such as AI and quantum computing, increasingly classified as “critical”, “sensitive” or “security-relevant” to provide them with protection against espionage and foreign influence and to secure competitive advantages (German National Academy of Sciences Leopoldina and German Research Foundation, 2024[31]).

Protecting data, information and know-how are not easy in the Internet era and restrictions on access may conflict with research integrity principles and open science (OECD, 2022[30]). Scientific research operates within a global research ecosystem that relies on autonomy, openness and free exchange to function effectively.21 A blanket application of strict research security measures would pose a direct or indirect risk to the quality, productivity, integrity and, therefore, the societal and economic value of the national research system (James et al., 2025[29]). Developing the capacity to identify and manage genuine security risks while preserving the integrity of the global research ecosystem has, therefore, become a priority for many governments.

Research security threats may result from the hostile activities of threat actors or the poor risk management practices of research-performing organisations or individual researchers. In 2022, the OECD released a policy paper entitled “Integrity and security in the global research system” which made recommendations on how various actors – including national governments, research-funding agencies, research institutions, universities, academic associations and intergovernmental organisations – should approach research security and outlined efforts already under way. Recommendations included integrating research security considerations into national and institutional frameworks for research integrity; promoting a proportionate and systematic approach to risk management in research; and working across sectors and institutions to develop more integrated and effective policy (OECD, 2022[30]).

These themes are addressed below, but since the report’s publication, research security measures have continued to expand globally, driven by heightened awareness and the evolving nature of security threats. There has been a sharp rise in the number of policy initiatives focused on research security and the number of countries deploying them. Only 27 national policy initiatives were reported in 2018 in the STIP Compass database.22 By 2025, that number had grown almost tenfold to more than 250. The interest in research security has expanded worldwide, with the number of countries with research security policies more than trebling over the same period, from 12 in 2018 to 41 in 2025.

Coupling of research security and integrity

While governments are putting measures in place to improve research security, they are at the same time emphasising research integrity, which refers specifically to certain values, norms and principles that constitute good scientific practice (freedom of scientific research, openness, honesty, ethics, integrity, accountability, etc.) and regulate international research collaboration (reciprocity, equity, non‑discrimination, etc.). Research integrity applies to the behaviour of individual scientists, but also to research ecosystems, with a particular focus on mitigating national and economic security threats and foreign interference. As international collaboration becomes more widespread and the geographic distribution of scientific production changes, mitigating unauthorised information transfer and foreign interference needs to be included in considerations of research integrity. Increasing transparency, disclosing potential conflicts of interest and conflicts of commitment, and managing risks are aimed at strengthening both research integrity and security and are considered essential for promoting trust in science (OECD, 2022[30]).

Policy support is increasingly focused on research security implementation measures

While many earlier policy initiatives focused on raising awareness of research security as an issue and developing policy intelligence, such as evaluations of country-specific risks, the more recent focus has been on developing strategies, agendas and plans and providing support for their implementation. There has also been growing use of regulation, soft law and oversight since 2020. This shift indicates that countries are tightening up their research security efforts, transitioning from simply raising awareness and gathering intelligence to more concrete planning and implementation. These efforts primarily target research-performing organisations and funding agencies. The extent of initiatives focused on implementation reflects the extent to which these actors need support operating in a changing environment as well as the extent to which this environment is disrupting established practices. The most common categories of implementation support among recent policy initiatives are the development of guidance, self-service tools and advisory services (Box 2.5). Governments, funding agencies and research‑performing organisations are also establishing dedicated organisational structures to promote research security.23

Box 2.5. Emerging types of support for research security implementation

Guidance development

Recent guidance documents provide frameworks for implementing newly issued strategies or directives or add detail to previously issued guidance. These are designed not only to help institutions navigate the evolving landscape of research security but also to help create a culture of accountability. For instance, the Norwegian Directorate for Higher Education and Skills, in collaboration with the Research Council of Norway, has issued Guidelines and Tools for Responsible International Knowledge Cooperation (Norwegian Directorate for Higher Education and Skills, 2023[32]). In the United States, the White House Office of Science and Technology Policy has published Guidelines for Research Security Programs at Covered Institutions. These outline expectations for research security programmes in relation to cybersecurity, foreign travel security reporting, research security training and export control training (White House Office of Science and Technology Policy, 2024[33]). In Austria, the Ministry of Innovation, Mobility and Infrastructure (Bundesministerium für Innovation, Mobilität und Infrastruktur) is supporting research security in the applied research sector. Measures include guidance to help funding applicants self-assess the risks of possible joint projects and partners by asking the right questions, giving advice on how to identify red flags and general information on research security.

Self-service tools

With the increasing complexity of research security regulations and guidance, recent initiatives have included practical tools to help universities and research institutions apply this guidance in their day‑to‑day work. For example, the UK government’s National Protective Security Authority and National Cyber Security Centre have released a research security maturity self-assessment tool, the Trusted Research Evaluation Framework. Complementing existing Trusted Research guidance, the framework is aimed at helping academic institutions in different stages of their research security journey understand what “good” looks like across seven areas of activity. It defines what constitutes foundation, intermediate and developed capacity for multiple evaluation dimensions (National Protective Service Authority, United Kingdom, 2024[34]).

Advisory services

The establishment of advisory bodies has become a cornerstone of recent research security efforts, reflecting research institutions’ need for expertise navigating research security requirements, assessing risk in projects and evaluating potential collaborations. For example, Denmark established the Centre for Innovation and Knowledge Security within the Danish Security and Intelligence Service in 2023 to proactively advise its research institutions on how to handle threats from foreign states. The centre was created to address the growing risks of espionage, intellectual property theft and foreign interference in research.¹ The United Kingdom’s partnership approach to research security engages universities through its Research Collaboration Advice Team and aims to support the research sector to take informed decisions on research collaborations.²

Notes: 1. For further information, see: https://pet.dk/en/our-tasks/security-advisory-services/the-objective. 2. For further information, see: https://www.gov.uk/government/groups/research-collaboration-advice-team-rcat.

Vetting international research collaborations through project- versus list-based approaches

Authorities are placing restrictions of varying degrees on collaboration with certain research organisations or countries. In some cases these are linked to identified fields of research that reflect geopolitical and economic security considerations. For example, the Government of Canada’s Policy on Sensitive Technology Research and Affiliations of Concern entered into force in 2024 and stipulates that any research grant or funding application in listed sensitive technology research areas will not be funded if researchers involved in the application activities are in receipt of funding or in-kind support from listed research organisations connected to military, national defense or state security entities that could pose a risk to Canada’s national security (Innovation, Science and Economic Development Canada, 2024[35]).24

Other countries are establishing project-based approaches to identify sensitive research collaborations rather than relying on defined field- or affiliation-specific restrictions. The German federal government published its Strategy on China in 2023, setting a framework for secure co-operation with China amidst systemic rivalry (Federal Foreign Office, Germany, 2023[36]). The German Research Foundation, the German Science and Humanities Council, the German Academic Exchange Service, and the Max Plank Society subsequently published recommendations to support scientists, research institutes and universities in navigating this new context. They deliberately refrain from drawing red lines around specific countries, partner institutions or research topics and instead endorse case-by-case assessments.25

Similarly, a 2024 JASON study Safeguarding the Research Enterprise, contracted by the US National Science Foundation (NSF), recommended identifying sensitive research and risks to collaboration at the project level during research proposal evaluation research. It provides an alternative to sweeping restrictions on all collaborations in listed high-risk areas. This process-based approach has been newly adopted in the NSF’s Trusted Research Using Safeguards and Transparency (TRUST) framework.26 Inspired by this example, the Japan Science and Technology Agency (JST) has recently introduced a pilot scheme, JST-TRUST, that it applies to its calls for proposals on quantum science and semiconductor research. The scheme involves screening experts’ proposals, asking principal investigators how they do due diligence on their projects. Based on this, they consider risk-mitigation measures, to be set by the JST, if necessary. The JST-TRUST also assists with monitoring and guidance on research outcomes and publishing.27

Reframing science diplomacy to further both national and multilateral goals

The third type of STI securitisation policy intervention is rooted in the projection of national interests in international regulations, norms, standards and alliances. In this regard, science diplomacy, defined as the use of science for foreign policy purposes, has become an increasingly prominent instrument to pursue not only multilateral goals, but also national interests. While some supporters of science diplomacy still predominantly highlight its global public goods aspects,28 today it is widely recognised that science is increasingly used as a strategic tool to secure national interests and power, and for leverage in interstate rivalry. This duality is hardly new, but as strategic competition has become more prominent in the current geopolitical environment, perceptions of science diplomacy have shifted, and it has become more institutionalised as an external foreign policy tool.

Along these lines, the 2025 revised science diplomacy framework of the American Association for the Advancement of Science and the United Kingdom’s Royal Society acknowledges the new era of disruption and is “darker, more realistic, and hard-edged, than its predecessor” from 2010 (The Royal Society and AAAS, 2025[37]). The European Commission’s European Framework for Science Diplomacy, also published in 2025, makes similar observations, acknowledging that “science and technology are pieces on the global geopolitical chessboard” (Gjedssø Bertelsen et al., 2025[38]). Box 2.6 briefly outlines both reports, which are expected to influence science diplomacy policies.

Box 2.6. New landmark frameworks for science diplomacy in 2025

Science diplomacy in an era of disruption (American Association for the Advancement of Science and The Royal Society)

This report updates the 2010 framework for science diplomacy issued by the American Association for the Advancement of Science and The Royal Society. It argues that a more fragmented and dangerous world, impacted by global challenges and technological disruption, necessitates a revised approach to how science and diplomacy interact. The document proposes a simplified two-dimensional framework: science impacting diplomacy (the different ways that science interacts with diplomatic objectives) and diplomacy impacting science (the ways that diplomacy interacts and engages with the scientific enterprise).

The consultations that informed this report highlighted several key messages. For example, while science advisory mechanisms are increasingly integrated into national and multilateral institutions – reflecting the fact science is now ever more central to foreign policy – scientific and diplomatic interests may not always coincide. Treaties governing global commons sometimes conflict with sovereign national interests, prompting a re-examination of scientific values once thought universal and their implications for international scientific collaboration. The report also noted the increasingly influential roles of non-state actors, such as major companies and philanthropic organisations in the changing landscape of science diplomacy.

A European Framework for Science Diplomacy (European Commission)

With science and technology increasingly becoming a geopolitical currency, the European Union has concluded that science diplomacy can help it to project soft power and pursue its economic interests and fundamental values more effectively. Accordingly, the European Union launched the report, a European Framework for Science Diplomacy, in 2025, which is expected to be followed up with a science diplomacy strategy later in the year. The report recognises paradigm shifts in science diplomacy, driven by geopolitical and technological changes. It proposes a European-wide approach to science diplomacy that preserves spaces for exchange, fosters a shared responsibility for addressing common challenges and protecting global public goods, and defends Europe’s strategic interests. In particular, the framework highlights the need for a strategic use of science diplomacy in the current geopolitical context, involving enhanced strategic intelligence capacity (e.g. using foresight) and strengthened science diplomacy in delegations and embassies. The report provides concrete recommendations and actions on how European leadership in science diplomacy can be achieved through strategic, operational and enabling instruments.

Note: To elaborate: Strategic instruments for European science diplomacy focus on setting clear priorities and making science diplomacy visible, identifying the appropriate balance between openness and restrictedness in international science co-operation, and leveraging science diplomacy to address global challenges and sustainably manage global public goods and commons, including through partnerships with countries in the Global South. Operational instruments aim to establish the necessary structures for EU leadership in science diplomacy, foster science for policy and foresight ecosystems to enhance foreign and security policies, and strengthen the role of science and technology within diplomatic representations. Enabling instruments seek to create and connect science diplomacy communities, train and empower Europe’s current and future science diplomacy professionals, and advance the frontiers of science diplomacy through research and innovative approaches to strengthen the overall capacity and expertise in the field.

Sources: Gjedssø Bertelsen et al. (2025[38]); The Royal Society and AAAS (2025[37]).

Scholars and practitioners have debated ways to define, categorise and frame the different shapes and forms science diplomacy can take (Turekian, 2018[39]).29 This chapter does not seek to create an additional framework, but rather identifies three key aspects to consider:

1.The co-operative nature of science diplomacy: Science diplomacy for global public goods and development.

2.The competitive nature of science diplomacy: Science diplomacy in pursuit of national interests.

3.The hybrid nature of science diplomacy: Non-state actors and Track 2 diplomacy.

Each of these is further elaborated below.

Co-operative science diplomacy for global public goods and development

This form of science diplomacy strongly emerged in the 1990s and involves a mix of foreign policy and scientific personnel, often meeting in multilateral fora, to address global challenges like climate change, biodiversity loss, health security issues, etc. Examples include the Paris Agreement on Climate Change. In some fields, science diplomacy communities have emerged according to the nature of the scientific domain or natural resource in question, such as water diplomacy, health diplomacy, cyber diplomacy, etc. It is the sort of science diplomacy that some think is most at risk from rising geopolitical tensions and national security policies. By way of example, Box 2.7 describes ongoing ocean science diplomacy and its recent roles in the third United Nations Ocean Conference in Nice (France).

Box 2.7. From data to diplomacy: How ocean science shapes policy and trust

The ocean plays a vital role in the economies and livelihoods of hundreds of millions of people. If treated as a single country, the ocean economy would have ranked as the world’s fifth-largest economy in 2019, contributing 3-4% of global gross value added between 1995 and 2020 and supporting up to 133 million full-time jobs (OECD, 2025[40]). But the ocean is under many pressures, threatening not only ocean health but the future of the ocean economy as well.

Peer-reviewed science provides objective criteria that help reconcile economic ambitions with global environmental imperatives in ocean governance and management. Grounded in data, peer engagement and shared objectives, scientists from different countries can build co-operation and trust where traditional diplomacy sometimes falters, resulting in tangible policy outcomes and informed decision making. The importance of ocean science diplomacy was clearly apparent during the recent third United Nations Ocean Conference (9-13 June 2025), where the integration of up-to-date scientific findings on the state of the ocean shaped commitments anchored in shared evidence.

For instance, responding to the need for science-based marine protected areas – with the objective of conserving 30% of the ocean by 2030 – countries committed to joint ocean exploration missions and enhanced transparency in areas beyond national jurisdiction in the high seas. Notably, over 20 countries ratified the “High Seas Treaty”, a major diplomatic outcome of the conference. Once in effect, potentially as early as 2026, the agreement will establish a new legal framework for governing the high seas, contributing to the conservation and sustainable development of marine biological diversity. Science indicators and sustainable ocean data observations – spanning weather patterns, biodiversity, carbon cycles and fisheries – will provide a shared evidence base to guide policy and enable scientifically informed negotiations.

Source: OECD (2025[40]).

A related consideration is the participation of low- and middle-income countries (LMICs) in these international co-ordination efforts. For example, since LMICs are expected to account for much of the growth in global carbon emissions until 2050, it will be important for the global community to support multilateral and club-based STI collaborations that include or are driven by representatives of the Global South.30 International STI co-operation can help strengthen the national STI capabilities of LMICs, allowing them to better engage in global STI collaboration and decision making and contributing to their overall economic development.

Competitive science diplomacy in pursuit of national interests

The view of science as a purely collaborative, objective and unifying force capable of overcoming deep political divides is challenged by the reality that science can be a geopolitical asset, blurring the lines between its perceived non-political nature and its role in power dynamics (Runguis and Flink, 2020[41]). Governments are strategically harnessing scientific expertise and international collaboration to advance their country’s geopolitical influence, economic competitiveness and security objectives, often involving a deliberate balance between openness and restrictive measures to safeguard sovereign interests.

Along these lines, many countries are bolstering the science capacity of their foreign ministries and missions. For example, some countries have a large representation of science and technology diplomacy counsellors or attachés in missions abroad. The United Kingdom, for instance, has a well-established network of approximately 130 staff in over 65 locations across the world, building collaborations that aim to maintain the country’s scientific base, support the competitive advantage of the United Kingdom’s innovative businesses, and address shared opportunities and threats. These work with local science and innovation organisations to project UK STI excellence and leadership globally, build and facilitate STI of value to the United Kingdom, and provide insights and intelligence. While the thematic focus is different for each country, priorities include opportunities and risks from critical and emerging technologies, addressing climate change and biodiversity loss, and health security.31 Other G7 countries and China have similar operations, but some smaller countries are also active. For example, Hungary maintains an international network of science and technology attachés stationed at 15 key locations in major STI partner countries and centres of competitiveness and innovation (Asia-Europe Foundation, 2025[42]).

Hybrid science diplomacy involving non-state actors and Track 2 diplomacy

Science diplomacy involves an increasingly hybrid approach, combining and intertwining Track 1 (formal diplomacy primarily led by diplomats and other state actors) and Track 2 diplomacy (informal diplomacy, involving non-governmental participants and informal dialogue) (Ruffini, 2020[43]; Turekian and Gluckman, 2024[44]). While Track 1 diplomacy involves the direct pursuit of state interests through official channels and supports the negotiation of international treaties and formal agreements, Track 2 diplomacy is considered a means by which non-state actors, particularly academics and scientific organisations, can contribute new ideas and relationships to the official diplomatic process by incorporating leading thinkers from outside governmental structures. Through the soft power of science, they can establish personal scientific networks to foster trust where official diplomatic links are otherwise weak or non-existent. Box 2.8 outlines some widely cited examples.

Box 2.8. Examples of Track 2 science diplomacy

While Track 2 diplomacy might be officially sanctioned by governments, it can also be driven by the professional interests of scientists. Various Cold War era links between the West and the Soviet Union are often cited as examples, such as the Pugwash conferences, which brought together scientists from both sides of the Iron Curtain and played a significant, behind-the-scenes role in informing and laying the groundwork for major arms control.¹ More recent examples of links include those between American and Cuban scientists over sharing weather data, leading to the development of a formal agreement to install shared GPS monitoring equipment in Cuba;² the Iranian public health experts who worked together with US counterparts to replicate the Iranian primary healthcare system in the Mississippi Delta;^3,4 and in the Middle East, collaboration between the Arava Institute for Environmental Studies and the Damour for Community Development, which have been organising since 2016 the Track II Forum for Environmental Diplomacy to enable key civil society organisations and individuals representing both state and non-state actors to discuss and develop cross-border strategies to facilitate formal and informal environmental agreements between Israel, the Palestinian Authority and Jordan.⁵

Notes: 1. For further information, see: https://www.nobelprize.org/prizes/peace/1995/pugwash/speedread. 2. For further information, see American Meteorological Society (2015[45]). 3. For further information, see: https://era.ideasoneurope.eu/2022/07/13/learning-from-rivals-the-role-of-science-diplomacy. 4. For further information, see: https://www.stimson.org/2025/health-and-science-diplomacy-could-pave-the-way-to-new-us-iran-relations. 5. For further information, see: https://arava.org/initiatives-working-groups.

Among non-state actors, the private sector plays an increasingly crucial and complex role in science diplomacy, wielding significant scientific, economic and political influence that can, in some cases, rival that of individual countries. Many large firms, especially global technology businesses, are major R&D funders, with their annual expenditures often comparable to or exceeding national public research programmes. Some engage directly in diplomatic efforts, cultivating relationships with foreign governments and international bodies like the United Nations and the European Union,32 and engaging directly with them on topics like emerging technologies, often bypassing national diplomats from their countries of origin. They are also critical partners in public-private partnerships for developing large research infrastructures, such as SESAME and CERN Open Lab, and demonstrated their pivotal capacity during the COVID-19 pandemic in vaccine development and global distribution. Furthermore, the private sector is central to setting international technical standards for global trade and knowledge exchange.

Developments like these have led to the emergence of technology and innovation diplomacy (Leijten, 2017[46]), which involves combining expertise from the three traditionally separate fields of technology, business and foreign policy with a view to advancing national interests. Some countries have established a diplomatic presence near innovation hubs like Silicon Valley in recent years. Denmark led this trend in 2017 with its tech embassy in Palo Alto,33 a move since emulated by the EU Office in Silicon Valley.34 In another example, Switzerland has established its Swissnex global network to strengthen its profile as a world-leading hotspot of innovation. The network has offices in 6 regions renowned for innovation, backed by around 20 counsellors based in Swiss embassies worldwide. A notable feature is the engagement of public and private stakeholders from the Swiss and local education, R&I landscape, who cover at least two-thirds of the costs of Swissnex’s activities.35

Principles for governing science, technology and innovation securitisation

These three strands of securitisation policy – critical technology promotion, research security and science diplomacy – are closely related and present policymakers with several governance challenges. Three aspects of STI policy governance particularly stand out, namely formulating the scope and focus of securitisation policies, mobilising key stakeholders to co-design and implement them, and building a knowledge and evidence base to inform policy choices and strategy:

• First are the scope and focus of STI securitisation measures. A key consideration here is their proportionality with the level of expected risk and opportunities. Governments need to strike several balances along different axes and at different levels in their policies, in particular with regards to international openness.

• Second, the R&I activities these policies seek to influence are performed by semi-autonomous researchers and private businesses. Governments must mobilise and partner with these groups for securitisation policies to succeed. They must also co-ordinate across different parts of government given the cross-cutting nature of securitisation policies.

• Finally, security-related STI policy measures should be precise and agile when targeting research, technology and industrial areas for promotion, protection and projection. This points to the need for policy risk assessment and uncertainty analysis capabilities, underpinned by useable strategic intelligence. Securitisation policies should also be monitored and evaluated to enable course corrections and promote accountability.

Proportionality, partnerships and precision are, then, a further set of “3Ps” that overlay the original security 3Ps of promotion, protection and projection (Figure 2.9).36 They amount to principles for governing securitisation policies to mitigate risks and promote strategic co-ordination. The remainder of this section briefly discusses each in turn.

Figure 2.9. Principles for governing science, technology and innovation securitisation that promote proportionality, partnership and precision

Proportionality: Scoping STI securitisation policies that balance different values, goals and interests

STI securitisation policies are inherently concerned with balancing different values, goals and interests along different axes and at different levels. The securitisation measures outlined in this chapter all pull in the direction of enhancing national interests, primarily economic competitiveness and national security, with each challenged by the need to retain some measure of international openness and research autonomy, both of which contribute to the value of R&I activities. The success of individual measures is tied to others, and they need to work together to ensure a balanced approach to securitisation. The formulation and implementation of STI securitisation policies should, therefore, be considered together as part of a broader and balanced STI securitisation strategy (also keeping in mind that this chapter has primarily explored the public research system, and that there are several other policies relevant to STI securitisation extending across the innovation chain that should also be considered).

The focus of this section is primarily on balancing research security and international openness, but there are also other important dilemmas that policymakers need to consider. For example, technology races should incorporate safeguards to manage downside risks and bridge global technology divides. In this regard, principles and guidelines can be an attractive modality for international, transnational and/or global actors to make moral and political commitments with some flexibility and accommodation for differences and changing circumstances (OECD, 2024[47]). Relatedly, clear ethical guidelines should be established for research and technologies with dual-use potential to ensure they do not undermine human rights or societal well-being (European Commission, 2025[21]).

Balancing research security and openness and the implications for international collaboration

Research security measures raise significant questions about international research collaboration, which is an important aspect of scientific openness. Countries are striving to strike the right balance between safeguarding their national and economic security while upholding academic freedom, promoting international research co-operation, and ensuring openness and non-discrimination. Implementing overly broad or extreme security practices can stifle academic freedom, hinder innovation and disrupt valuable global partnerships. On the other hand, too little security can expose sensitive research or academic collaborations to risks, ultimately eroding safety and trust.

There is wide recognition among policymakers that research security and open science need not be cast as oppositional and can, in fact, be complementary: for instance, research security measures can enable open research practices by protecting academic freedom from abuse by malicious state actors; they also often entail greater transparency on researchers’ affiliations and funding sources. In this way, they contribute to good scientific practice, but they can also be applied overzealously. The key is to find a middle ground that protects valuable work without undermining the very principles of academic freedom and the social and economic benefits of participating in open international scientific collaboration (OECD, 2022[30]; Shih, 2025[48]). The overarching principle guiding this complex equilibrium is to keep scientific engagement “as open as possible and as secure as necessary”.37 A related concept is the “small yard, high fence”, where strict, robust controls are put in place to protect narrow and specific areas of science and technology considered critical to national and economic security. However, the growing emphasis on research with a dual-use character could shift countries’ calculations and may introduce additional restrictions on international research collaboration (European Commission, 2025[21]).

STI securitisation measures also run the risk of creating a more fragmented global R&I landscape that is ill-equipped to tackle global challenges. Measures in one country can easily trigger unwelcome countermeasures in others and have a chilling effect on international collaboration to address shared global challenges. For instance, health-related research fields – such as pandemic preparedness and antimicrobial resistance – face an openness-versus-security dilemma. They address global challenges that depend on open scientific collaboration and data sharing to drive preparedness, recovery and resilience. At the same time, heightened openness in such sensitive health domains can increase the risk of misuse or misconduct, underscoring the need for vigilance. Acknowledging and managing this tension responsibly is essential to safeguard research integrity and ensure that international collaboration in health research can continue with confidence and maintain its positive impact.

The broader emphasis on research security has inadvertently led to a chilling effect on international research collaboration and academic mobility more broadly. Research-performing organisations are increasingly cautious about entering international research collaborations where security risks have been identified. This is in part a function of asymmetric knowledge, with research-performing organisations often complaining of being given insufficient information from security services to make informed judgements (James et al., 2025[29]). Research-performing organisations also complain of being faced with a range of ambiguities and sometimes contradictory signals. There are also risks that researchers feel pressured to self-censor or avoid high-risk but important research areas, adversely affecting R&I (Shih, 2025[48]). Furthermore, there are risks of prejudice, cultural bias and inadvertent discrimination against certain population groups in both list-based and process-based approaches to risk assessment. This is a major concern for the academic community and an area that needs to be carefully monitored as policies for restrictions on collaboration become more widespread.

While research-performing organisations have a responsibility to act responsibly in their international activities, neither individual researchers nor individual universities should be left alone in making assessments of difficult goal conflicts, and governments and funding agencies have a responsibility to provide national guidelines (Swedish Council for Higher Education, 2024[49]). It is possible to define international co-operation as fully compatible with national security rather than as something external and threatening to it.38 To achieve this, an “intentionality” approach is crucial, requiring a deep understanding of collaborating partners’ motivations, networks and their ultimate intentions for research outputs. Similarly, emphasising reciprocity in collaborations is vital to ensure mutual benefits and prevent non-reciprocal exchanges that can intensify securitisation concerns (Dawes, Salt and Smith, 2024[50]).

At the same time, research-performing organisations need to develop their own internal security capacity, which includes creating institutional policies, establishing risk management and due diligence processes, and hiring dedicated research security officers. They also need to continue raising awareness about research security among researchers and administrative staff. Building this capacity is a challenge, as institutions must do so with limited funding and in a competitive job market where these specialised skills are scarce. Similarly, governments are also struggling with these capacity constraints as the growing demands of research security put a strain on ministries and security agencies (James et al., 2025[29]).

A strategic dual-track approach is emerging, focusing on intensive collaboration with “like-minded” countries for cutting-edge technologies while maintaining broader co-operation with diverse countries for shared global challenges (Asano and Arimoto, 2024[51]; Turekian and Gluckman, 2024[44]). Policy frameworks must clearly define “red lines” for collaborations that flagrantly violate established norms, such as serious ethics dumping, direct military use by military institutions, illicit technology transfer or grave human rights violations. At the same time, they need to actively manage “grey areas” where different national and institutional contexts create challenges, to prevent inappropriate transgressions through adherence to principles of research integrity, ethics and “responsible internationalisation” (Shih, 2024[52]). To take effective decisions, a wide range of issues must be considered, such as openness, scientific advancement, global challenges, national security, economic security, ethics, human rights and democracy. Combining these diverse concerns into a single, cohesive approach is difficult but essential for achieving proportionality in STI securitisation measures (Schwaag Serger and Shih, 2024[53]).

Partnerships: Co-operating with scientists and businesses and across government

A comprehensive STI securitisation policy mix must find ways to bring a broad range of stakeholders, including governments, business and academia, into the discussion while at the same time building robust governance mechanisms essential to integrating a range of priorities and values. This is in a context where businesses and public research-performing organisations enjoy considerable autonomy, which presents co-ordination and mobilisation challenges, particularly where values and interests may be misaligned. Promotion, protection and projection policies also call for cross-government co-ordination, but this is notoriously challenging, with different ministries and agencies having their own specific operating procedures, mental models and frameworks, and community interests to serve.

Co-operation with businesses

Most R&D in technology-intensive economies is conducted in firms, where trade and investment restrictions, as well as new industrial policy measures, are felt most keenly. Involving firms in formulating and implementing STI securitisation policies is, therefore, crucial. This is perhaps most obvious in promotion policies, where, for instance, the new wave of industrial policies builds largely on public-private partnerships. Priority-setting and policy formulation in these contexts typically involve firms, which are often engaged in strategic foresight and technology assessment processes, policy formulation and design, and collaborative R&D with public sector research-performing organisations. Firms also benefit from policy incentives that seek to attract international talent and are typically engaged in their design.

This chapter has focused on research security policies affecting public research-performing organisations, but firms are also subject to restrictions, for example in the form of export controls and investment screening as part of economic and national security measures. They are also targets of cyberattacks and industrial espionage, as well as vectors. Some governments provide guidelines on countermeasures against technology leakage in the context of overseas expansion of production facilities. Both the European Economic Security Strategy (see Box 2.2) and Japan’s Action Plan for Economic Security (see Box 2.3), for example, include provisions on the security risks from outbound investments.

This chapter has also highlighted the growing prominence of large leading technology firms in technology diplomacy as they seek to exert influence over international norms and political agendas. These firms wield significant control over critical technologies, raising key questions around accountability, equity and governance, particularly as development of these technologies resides largely outside the oversight of the state, and corporate interests may diverge from national interests (Geneva Science and Diplomacy Anticipator, 2025[54]). Science and technology diplomacy frameworks, involving firms and state authorities, have been updated to explicitly recognise non-state actors as integral participants, shifting from a state‑only focus.

Co-operation with scientists and research-performing organisations

Directed research agendas that are oriented towards strategic goals like economic and national security must mobilise scientists and research-performing organisations if they are to succeed. Governments traditionally use managed funding programmes and other incentives for this purpose (see Chapter 1), but they must also incentivise the strengthening of linkages with other innovation system actors, notably firms, to promote innovation and national competitiveness. As already highlighted, a growing policy focus on dual‑use research and technology development could have implications for civil research, in terms of its physical environment, e.g. with high-security zones and restricted access, but also in the ways research is conducted and disseminated. Targeted education and support will be needed to help researchers better understand the complexities of dual-use research, including its risks and its opportunities (European Commission, 2025[21]). But there will also be a need for scientists and universities to be routinely involved in co-designing any new arrangements.

This is already happening around research security. The primary responsibility for implementing research security belongs to research-performing organisations and especially universities, given their autonomy is secured in many countries.39 At the time of the OECD-GSF report (OECD, 2022[30]), government research security measures were regularly criticised by the research community for being opaque or disconnected from the operational realities of research institutes. More recent initiatives show a marked improvement in how governments engage with research institutes in the development and implementation of such policies. For example, the 2024 JASON study’s recommendation to adopt a process-based, rather than a list-based, approach to identifying sensitive research was developed after discussions with a range of government agencies, university administrators and experts on issues of research security (JASON, 2024[55]). Across various levels of government, universities and research institutes appear to be routinely involved as active partners in the research security policymaking process, with nearly all new policy initiatives making efforts to collect input from research stakeholders.40

At the same time, a range of researcher-driven initiatives has emerged to promote interactions between the scientific and diplomatic communities, many working across national borders. For example, the American Association for the Advancement of Science established in 2008 the Center for Science Diplomacy, which aims to strengthen interactions and partnerships between the two communities, as well as to develop the intellectual framework and training to support the practice of science diplomacy.41 In Europe, the EU Science Diplomacy Alliance was launched in 2021 to facilitate interactions and dialogue, training, institutional capacity building and co-ordination of grant-seeking or use of joint funding.42 Similarly, DiploCientífica has developed a collaborative network that brings together scientists, policymakers and the diplomatic community in Latin America and the Caribbean to build capacity and produce constructive knowledge.43 Finally, South Africa hosts the Science Diplomacy Capital for Africa initiative that aims to facilitate cross-border collaboration between African science institutions and global partners, particularly diplomatic communities and regional bodies.44

Promoting cross-government policy coherence

Ministries with responsibilities for R&I, as well as funding agencies, are active in the growing securitisation of STI, although it has most often been led by ministries in other policy domains such as trade, foreign affairs, defence and industry. Existing links between STI policy and other policy domains remain weak in most countries and need strengthening to orchestrate government action on protection, promotion and projection policies (OECD, 2023[1]).

Strategically oriented research – for example, as part of new industrial policies – necessarily involves cross-government co-operation, particularly to help orchestrate supportive actions across the innovation chain, from basic research to technology commercialisation and diffusion. This is perhaps best illustrated by the recent popularity of mission-oriented innovation policy approaches, which typically bring together several ministries and agencies to co-ordinate actions to meet specified and time-limited shared goals (see Chapter 1). The promotion of dual-use research and technologies will also call for increased co-operation between STI ministries and agencies and their defence sector counterparts to accelerate innovation and support responsible and secure technology development (European Commission, 2025[21]).

An integrated approach to research security also calls for strengthening cross-government collaborations between science and security agencies. Such collaborations are necessary to build mutual understanding of the benefits and risks of international research collaborations and to help build risk-appropriate mitigation strategies. One purpose of such collaborations has been to build a shared understanding between the scientific and security agencies on the risks facing the research sector and thus increase buy-in for research security policy actions.45 More broadly, as research security policies proliferate, measures to streamline and harmonise overlapping guidelines, standards and organisational responsibilities are likely to be required. The establishment of new structures and new requirements for research security can come with a significant overhead in terms of costs and effort. Already, national governments and funding institutions face the need to clarify organisational roles and responsibilities within their own level of governance.46 A growing interest in information clearinghouses and learning and discussion forums may signal a demand for improved policy coherence. This would not only facilitate more consistent implementation of research security policies worldwide but also reduce burden on researchers. The Dutch National Contact Point for Knowledge Security is often held up as a good example and is a collaboration between different government ministries to support anyone connected to a knowledge institute who has questions about opportunities, risks or practical matters concerning international research co-operation.47

Cross-government collaborations also aim to support faster and more effective risk identification and mitigation of potential threats. For example, the Korean Ministry of Science and ICT is developing a new security classification for national research and development projects to enable better monitoring of these projects according to their level of risk. The new classification system defines a new category of “sensitive research” that lies between the traditional categories of “classified” and “unclassified” research. This initiative is part of a comprehensive national Plan for Strengthening the Research Security System for the Establishment of a Trusted Research Ecosystem, which is the collaborative work of nine ministries and agencies. The partnership is also conducting consultations on research asset leakage and developing a research security field guide to support research institutions.

Effective science diplomacy also needs enhanced cross-government co-ordination, and the interface between diplomatic services, science ministries and research communities is increasingly important. Policy concerns revolve around strengthening institutional capacities and personnel skills while fostering a more strategic outlook in pursuing a range of means and ends.

Precision: Building strategic intelligence and risk assessment capabilities

For securitisation measures to be proportionate, they should be based on sound risk and opportunity identification and assessments that draw on knowledge and evidence on current and future developments of new STI and their potential impacts on the economy and society. This strategic intelligence draws on a broad range of methods, such as statistical benchmarking, forecasting and modelling, foresight, technology assessment, systems and pathway mapping, and technology monitoring and evaluation. Chapter 7 outlines several types of strategic intelligence practices that would likely prove useful, including horizon scanning and technology monitoring; situation analysis; forward-looking technology assessment; adaptive foresight; multistakeholder participation; and formative (real-time) evaluation. Such efforts should also combine and integrate different disciplines, for example expertise on research, science and higher education systems and dynamics with expertise on relevant countries and national and economic security. Such a multi-disciplinary approach is important for avoiding over-securitisation (Schwaag Serger and Shih, 2024[53]).

Different technology supply chains have different vulnerability risks, and the same applies to international science collaboration: different critical technologies have varying dual-use potential, for example, and countries differ in their capacities to exploit them. This variation points to the need for a targeted policy approach, underpinned by risk management assessments that draw on the best available evidence, as well as forward-looking analysis where uncertainties preclude traditional risk-based analysis (OECD, 2023[1]). Several initiatives are now under way to develop this knowledge foundation, but more is needed. Earlier descriptions of EU and Japanese economic security policies highlighted these sorts of activities (see Box 2.2 and Box 2.3).48 Box 2.9 describes how Finland is similarly building capacity in cross‑government technology assessment to inform its economic and research and technology security policies.

Box 2.9. Strategic intelligence for economic and research security in Finland

As a small, technologically advanced economy, Finland has benefited from open international research and innovation co-operation, which has guided its science, technology and innovation (STI) policy thinking throughout the post-Cold War era. With growing awareness of research security, Finland needs to reconsider its technological and business strengths and purposefully develop growth opportunities and international high value-added businesses. Alongside strong EU STI collaboration, membership to the North Atlantic Treaty Organization also influences Finland’s STI collaboration with partners and opens export opportunities, including for the defence industry.

High-tech industries account for an important part of Finland’s exports, including products and services, and depend on significant amounts of raw materials and intermediate products sourced from abroad. In late 2024, the Ministry of Economic Affairs and Employment established a Technology Policy Unit to enhance policy co-ordination, identify growth opportunities from technologies and deepen Finland’s analytical capacities in this area as a part of an overall technology roadmap activity. This includes developing capabilities to generate strategic intelligence that aims to provide a better understanding of the challenges and opportunities from research and new technologies.

To improve cross-government co-ordination, the Ministry of Economic Affairs and Employment has also convened a working group on technology policy for regular exchange of views and co-ordination with relevant ministries and agencies. Together, this group will identify and discuss policy questions that require national co-operation and co-ordination (including research and technology security issues), strengthen goal-oriented technology anticipation and analytical capacity, and produce knowledge that supports STI and industrial policy steering.

Source: Based on correspondence between the OECD Secretariat and the Finnish Ministry of Economic Affairs and Employment and the Ministry of Education, Science and Culture.

In the research security area, several countries and institutions have developed guidance for risk assessment and avoidance of risk,49 but there is less guidance on what proportionate risk mitigation and management means in different contexts. Not all risks can be identified and there are systemic vulnerabilities that need to be considered, such as in IT systems or peer review processes. Proportionality depends on priorities, resources and context. For example, some countries have developed blacklists of areas in which all scientific collaboration with specific countries or institutions is prohibited. Others are linking risk identification and management to TRLs. The advantages and disadvantages of different approaches in different contexts is an area where different countries and institutions could learn from each other.

As the field of research security continues to develop, there is recognition of the need for continuous learning to stay ahead of emerging risks and challenges. This includes understanding the latest mechanisms of foreign interference and effective strategies for mitigating risk. To address this need, various actors across the research landscape are moving to formalise continuous learning processes, in the form of evaluating policy and practice. This evaluative process is essential for refining existing policies and ensuring they remain aligned with broader R&I objectives.50 Furthermore, to facilitate the sharing of best practices in research security, organisations at multiple levels are increasingly focused on ways to foster peer learning, including through discussion-based forums and central clearinghouses of vetted, up‑to-date information and resources on threats and mitigation strategies.51

Science diplomacy measures would also benefit from greater use of strategic intelligence. Along these lines, the Geneva Science and Diplomacy Anticipator has proposed a Framework on Anticipatory Science Diplomacy to proactively govern and deploy scientific advances before they cause disruption or inequality, ensuring science serves humanity while navigating geopolitical competition. It does this by providing relevant actors with early insights into frontier science – by identifying and scoping out the major scientific advances with the highest potential to reshape humanity and the planet – thereby allowing sufficient time to assess and debate their long-term global implications, and avoid missed opportunities by proactively shaping innovation trajectories before crises emerge (Geneva Science and Diplomacy Anticipator, 2025[54]). The OECD’s 2024 Framework for Anticipatory Governance of Emerging Technologies also provides structured guidance on how governments can embed anticipation into policy cycles, stakeholder engagement and innovation strategies, including at the international level (OECD, 2024[47]).

Conclusions

Among a wide range of types of international STI linkages, this chapter has focused on international research linkages and, specifically, the emerging securitisation of STI policy that now shapes them. Post‑Cold War international STI co-operation arrangements are being reconfigured as they transition to a new era marked by growing geopolitical rivalry and intensified inter-state competition on emerging technologies. Signals of a less open international research system are already emerging: for instance, growing international co-authorship in scientific publications has been at the core of a more interconnected global research community over the last 30 years, but is now stagnating or even in decline.

While these developments present new challenges and considerable uncertainty, STI policymakers can influence the contours of an emerging landscape of international STI linkages. Using the “3Ps” framework of promotion, protection and projection introduced in the 2023 edition of the OECD Science, Technology and Innovation Outlook, this chapter has reported on how governments increasingly target critical technologies to promote both economic and national security; implement research security measures to protect against unauthorised knowledge leakage and coercion; and use science diplomacy policies to further their national interests and accordingly more strategically manage the international openness of their research systems.

These policies carry various risks and opportunities, and policymakers should pursue balanced STI securitisation policies that are proportional to the risks at hand, precise in their targeting, and based on partnerships with scientists and businesses, as well as across government. For example, securitisation policies for STI should weigh any restrictions against the benefits of open science and innovation; they should be evidence-based, drawing on risk assessments, future-oriented analysis such as foresight and technology assessment (see Chapter 7), and evaluation insights; and they should mobilise a diverse set of stakeholders – including scientists and innovative firms which increasingly accept the necessity of securitisation measures – to increase their chances of success.

Many of the skills and organisational capabilities needed for governments to pursue balanced securitisation policies for STI remain underdeveloped. New institutions, policy frameworks and governance arrangements will also be required but will take time to develop, sometimes through trial and error. Policies that promote dual-use research, research security and science diplomacy are often managed by a range of ministries and agencies yet are closely entwined. Governments need to develop policy tools and assessment frameworks that offer a systemic view and understanding of their portfolio of securitisation policies in STI and beyond to appreciate their synergies and dissonance and promote joined-up interventions. Despite the sensitivities of the policy area, governments should also engage in international mutual learning and benchmarking of emerging good practices among like-minded countries to co-ordinate and accelerate their national development plans and implementation progress. There is still much to learn and continued sharing of policy and practice will be needed, as will policy refinement in a fast-moving space.

References

[45] American Meteorological Society (2015), “How meteorological partnership between US, Cuba was created over 20 years”, ScienceDaily, http://www.sciencedaily.com/releases/2015/05/150522131609.htm.

[51] Asano, K. and T. Arimoto (2024), “Research security for actionable science diplomacy”, Science and Diplomacy, https://doi.org/10.1126/scidip.adt9870.

[42] Asia-Europe Foundation (2025), Asia-Europe Science & Technology Diplomacy Report: Mapping Science and Technology Diplomacy in the Two Regions, Asia-Europe Foundation, https://asef.org/publications/asia-europe-science-technology-diplomacy-report.

[28] Barteau, M. and S. Rovito (eds.) (2024), International Talent Programs in the Changing Global Environment, National Academies of Sciences, Engineering, and Medicine, https://doi.org/10.17226/27787.

[50] Dawes, J., K. Salt and C. Smith (2024), “Achieving research security while pursuing science diplomacy: Considering intentionality”, Science and Diplomacy, https://doi.org/10.1126/scidip.adt9878.

[15] Dechezlepretre, A., L. Diaz and G. Lalanne (2025), “What are the latest trends in industrial policy? Three key findings from OECD data”, OECD Blog, https://www.oecd.org/en/blogs/2025/09/what-are-the-latest-trends-in-industrial-policy-three-key-findings-from-oecd-data.html.

[7] Edler, J. et al. (2023), “Technology sovereignty as an emerging frame for innovation policy”, Research Policy, Vol. 52/6, https://doi.org/10.1016/j.respol.2023.104765.

[27] European Commission (2025), Choose Europe for Science: A Snapshot of Initiatives in the European Union to Support Researchers Worldwide, Publications Office of the European Union, https://data.europa.eu/doi/10.2777/8296541.

[19] European Commission (2025), Joint White Paper for European Defence Readiness 2030, European Commission, Brussels, https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex:52025JC0120.

[21] European Commission (2025), Making the Most of EU Research and Innovation Investments: Rethinking Dual Use, Publications Office of the European Union, Luxembourg, https://op.europa.eu/en/web/eu-law-and-publications/publication-detail/-/publication/9d84f730-516c-11f0-a9d0-01aa75ed71a1/language-en.

[20] European Commission (2025), Proposal for a Regulation of the European Parliament and of the Council on Establishing the European Competitiveness Fund (“ECF”), Including the Specific Programme for Defence Research and Innovation Activities, European Commission, Brussels, https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52025PC0555.

[67] European Commission (2025), Technology Monitoring and Assessment: Comparing EU, US and Chinese approaches, Publications Office of the European Union, https://data.europa.eu/doi/10.2777/0772224.

[22] European Commission (2025), Unlocking the Potential of Dual-use Research and Innovation, Publications Office of the European Union, Luxembourg, https://op.europa.eu/en/web/eu-law-and-publications/publication-detail/-/publication/e11c91be-5173-11f0-a9d0-01aa75ed71a1/language-en.

[18] European Commission (2024), Safer Together: Strengthening Europe’s Civilian and Military Preparedness and Readiness, European Commission, https://commission.europa.eu/topics/defence/safer-together-path-towards-fully-prepared-union_en.

[17] European Commission (2024), The Future of European Competitiveness, Publications Office of the European Union, Luxembourg, https://commission.europa.eu/topics/eu-competitiveness/draghi-report_en#paragraph_47059.

[16] European Commission (2024), White Paper on Options for Enhancing Support for Research and Development Involving Technologies With Dual-use Potential, European Commission, Brussels, https://research-and-innovation.ec.europa.eu/document/download/7ae11ca9-9ff5-4d0f-a097-86a719ed6892_en.

[64] European Commission (2023), Commission Recommendation of 3 October 2023 on Critical Technology Areas for the EU’s Economic Security for Further Risk Assessment with Member States, European Commission, Strasbourg, France, https://defence-industry-space.ec.europa.eu/commission-recommendation-03-october-2023-critical-technology-areas-eus-economic-security-further_en.

[11] European Commission (2023), European Economic Security Strategy, European Commission, Brussels, https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:52023JC0020.

[36] Federal Foreign Office, Germany (2023), “Germany adopts its first comprehensive Strategy on China”, https://www.auswaertiges-amt.de/en/aussenpolitik/regionaleschwerpunkte/asien/strategy-on-china/2608618.

[54] Geneva Science and Diplomacy Anticipator (2025), Anticipatory Science Diplomacy: A Framework for Global Action, Geneva Science and Diplomacy Anticipator Foundation, Geneva, https://www.gesda.global/wp-content/uploads/2025/08/GESDA_ASD_Paper_2025_28July__FINAL.pdf.

[61] German Academic Exchange Service (2024), Academic Cooperation with China: A Realistic Approach: DAAD Recommendations to German Higher Education Institutions, German Academic Exchange Service, Bonn, Germany, https://static.daad.de/media/daad_de/pdfs_nicht_barrierefrei/der-daad/240307_daad_perspektive_china_en.pdf.

[31] German National Academy of Sciences Leopoldina and German Research Foundation (2024), Scientific Freedom and Security Interests in Times of Geopolitical Polarisation: Fifth Report of the Joint Committee of DFG and Leopoldina on November 1, 2024, Joint Committee of DFG and Leopoldina on the Handling of Security-Relevant Research, Halle, https://www.security-relevant-research.org/publication-progressreport2024.

[59] German Research Foundation (2023), Recommendations from the DFG: Dealing with Risks in International Research Cooperation, German Research Foundation, https://www.security-relevant-research.org/umgang-mit-risiken-2.

[65] German Science and Humanities Council (2025), Science and Security in Times of Global Political Upheaval, German Science and Humanities Council, Cologne, Germany, https://doi.org/10.57674/v8dp-8269.

[38] Gjedssø Bertelsen, R. et al. (eds.) (2025), A European Framework for Science Diplomacy: Recommendations of the EU Science Diplomacy Working Groups, Publications Office of the European Union, https://doi.org/10.2777/9235330.

[10] Goodman, M. and B. Robert (2021), “Toward a T12: Putting allied technology cooperation into practice”, CSIS Briefs, Center for Strategic & International Studies, https://csis-website-prod.s3.amazonaws.com/s3fs-public/publication/211013_Goodman_Toward_T12.pdf?gPeOoAp9ER7mdhXQW7OWYxqga_s133l (accessed on 6 March 2023).

[8] Helwig, N., E. Sinkkonen and V. Sinkkonen (2021), Strategic Autonomy and the Transformation of the EU: New Agendas for Security, Diplomacy, Trade and Technology, FIIA Report No. 67, Finnish Institute of International Affairs, Helsinki, https://www.fiia.fi/wp-content/uploads/2021/04/fiia-report-67_niklas-helwig-et-al_strategic-autonomy-and-the-transformation-of-the-eu.pdf.

[35] Innovation, Science and Economic Development Canada (2024), Policy on Sensitive Technology Research and Affiliations of Concern, Government of Canada, Ottawa, Ontario, https://science.gc.ca/site/science/en/safeguarding-your-research/guidelines-and-tools-implement-research-security/sensitive-technology-research-and-affiliations-concern/policy-sensitive-technology-research-and-affiliations-concern.

[29] James, A. et al. (2025), European Research Security: Threat Perspectives and the Responses of Policy Makers and Research Performing Organisations, Manchester Institute of Innovation Research, https://doi.org/10.71535/d5993826-83d6-4aa2-8e23-4195ad92b629.

[55] JASON (2024), Safeguarding the Research Enterprise, JASON, McLean, VA.

[56] Larrue, P., P. Tõnurist and D. Jonason (2024), “Monitoring and evaluation of mission-oriented innovation policies: From theory to practice”, OECD Science, Technology and Industry Working Papers, No. 2024/09, OECD Publishing, Paris, https://doi.org/10.1787/5e4c3204-en.

[46] Leijten, J. (2017), “Exploring the future of innovation diplomacy”, European Journal of Futures Research, Vol. 5, https://doi.org/10.1007/s40309-017-0122-8.

[9] March, C. and I. Schieferdecker (2021), “Technological sovereignty as ability, not autarky”, CESifo Working Papers, Ludwigs-Maximilians University’s Center for Economic Studies and the ifo Institute, Munich, Germany, https://www.cesifo.org/DocDL/cesifo1_wp9139.pdf.

[60] Max Planck Society (2023), Action Recommendations for Cooperation with China, Max Planck Society for the Advancement of Science, Munich, Germany, https://www.mpg.de/21535920/mpg-china-handlungsempfehlung.pdf.

[57] Mérat, V. (2022), “Précarité, harcèlement, manque d’attractivité: À l’université, la baisse du nombre de doctorants inquiète”, Le Figaro étudiant, https://etudiant.lefigaro.fr/article/precarite-harcelement-manque-d-attractivite-a-l-universite-la-baisse-du-nombre-de-doctorants-inquiete_55a41c50-37fa-11ed-9733-eb8453767bff.

[25] METI (2025), Action Plan to Enhance Industrial and Technological Infrastructures for Economic Security: Second Revised Edition, Ministry of Economy, Trade and Industry, Tokyo, https://www.meti.go.jp/policy/economy/economic_security/250530actionplanr2.pdf.

[12] METI (2024), Economic Security Action Plan, Ministry of Economy, Trade and Industry, Tokyo.

[34] National Protective Service Authority, United Kingdom (2024), “Trusted Research Evaluation Framework”, news, https://www.npsa.gov.uk/blog/news/trusted-research-evaluation-framework.

[63] National Science Foundation (2024), “NSF-backed SECURE Center will support research security, international collaboration”, news, https://new.nsf.gov/news/nsf-backed-secure-center-will-support-research.

[32] Norwegian Directorate for Higher Education and Skills (2023), Guidelines and Tools for Responsible International Knowledge Cooperation, Norwegian Directorate for Higher Education and Skills, Bergen, Norway, https://hkdir.no/en/rapporter-undersokelser-og-statistikk/guidelines-and-tools-for-responsible-international-knowledge-cooperation-report.

[23] NSS (2022), Economic Security Promotion Act, National Security Secretariat.

[13] OECD (2025), “A quantum technologies policy primer”, OECD Digital Economy Papers, No. 371, OECD Publishing, Paris, https://doi.org/10.1787/fd1153c3-en.

[66] OECD (2025), Economic Security in a Changing World, OECD Publishing, Paris, https://doi.org/10.1787/4eac89c7-en.

[6] OECD (2025), Measuring Science and Innovation for Sustainable Growth, OECD Publishing, Paris, https://doi.org/10.1787/3b96cf8c-en.

[5] OECD (2025), Measuring Science and Innovation for Sustainable Growth, OECD Publishing, Paris, forthcoming, https://doi.org/10.1787/3b96cf8c-en.

[40] OECD (2025), The Ocean Economy to 2050, OECD Publishing, Paris, https://doi.org/10.1787/a9096fb1-en.

[4] OECD (2025), “What are the key trends in international student mobility?”, Education Indicators in Focus, No. 88, OECD, Paris, https://www.oecd.org/en/publications/what-are-the-key-trends-in-international-student-mobility_2a423a76-en.html.

[47] OECD (2024), “Framework for Anticipatory Governance of Emerging Technologies”, OECD Science, Technology and Industry Policy Papers, No. 165, OECD Publishing, Paris, https://doi.org/10.1787/0248ead5-en.

[68] OECD (2023), OECD Reviews of Innovation Policy: Korea 2023, OECD Reviews of Innovation Policy, OECD Publishing, Paris, https://doi.org/10.1787/bdcf9685-en.

[1] OECD (2023), OECD Science, Technology and Innovation Outlook 2023: Enabling Transitions in Times of Disruption, OECD Publishing, Paris, https://doi.org/10.1787/0b55736e-en.

[26] OECD (2023), “What is the best country for global talents in the OECD?”, Migration Policy Debates, No. 29, OECD, Paris, https://www.oecd.org/content/dam/oecd/en/publications/reports/2023/03/what-is-the-best-country-for-global-talents-in-the-oecd_3496c15f/5186ab2d-en.pdf.

[30] OECD (2022), “Integrity and security in the global research system”, OECD Science, Technology and Industry Policy Papers, No. 130, OECD Publishing, Paris, https://doi.org/10.1787/1c416f43-en.

[14] OECD (Forthcoming), National quantum policies and strategies.

[2] Olechnicka, A., A. Ploszaj and D. Celinska-Janowicz (2019), The Geography of Scientific Collaboration, Routledge.

[62] Presidential Advisory Council on Science and Technology, Korea (2023), Research Security System Improvement Plan, https://www.pacst.go.kr/jsp/council/councilPostView.jsp?post_id=2472&etc_cd1=COUN01&cpage=2&board_id=11#this.

[43] Ruffini, P. (2020), “Collaboration and competition: The twofold logic of science diplomacy”, The Hague Journal of Diplomacy, https:// brill.com/view/journals/hjd/15/3/article-p371_8.xml.

[41] Runguis, C. and T. Flink (2020), “Romancing science for global solutions: On narratives and interpretative schemas of science diplomacy”, Humanities & Social Sciences Communications, Vol. 7/102, https://doi.org/10.1057/s41599-020-00585-w.

[53] Schwaag Serger, S. and T. Shih (2024), Finding Effective and Acceptable Approaches to Managing International Collaborations in a Geopolitically Turbulent World, Vinnova, https://www.vinnova.se/globalassets/publikationer/2024/ru-rapport-dec-2024.pdf?cb=20250107110054.

[48] Shih, T. (2025), “Challenges to research security”, pp. 1-8, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5172561.

[52] Shih, T. (2024), “The role of research funders in providing directions for managing responsible internationalization and research security”, Technological Forecasting and Social Change, Vol. 201, pp. 1-10, https://doi.org/10.1016/j.techfore.2024.123253.

[24] Suzuki, K. (2023), “How will the Economic Security Law change Japan’s sci-tech policy?”, https://www.tokyofoundation.org/research/detail.php?id=943#:~:text=Sci%2DTech%20Policy%20under%20the%20New%20Law&text=First%20is%20that%20the%20process,made%20through%20private%2Dpublic%20partnerships.

[49] Swedish Council for Higher Education (2024), Responsible Internationalisation: Interim Report on a Government Assignment, Swedish Council for Higher Education, https://www.uhr.se/globalassets/_uhr.se/publikationer/2024/responsible-internationalisation---interim-report-on-a-government-assignment-2024.pdf.

[37] The Royal Society and AAAS (2025), Science Diplomacy in an Era of Disruption, The Royal Society, https://royalsociety.org/-/media/about-us/international/science-diplomacy/science-diplomacy-in-an-era-of-disruption.pdf.

[39] Turekian, V. (2018), “The evolution of science diplomacy”, Global Policy, Vol. 9/53, pp. 5-7, https://doi.org/10.1111/1758-5899.12622.

[44] Turekian, V. and P. Gluckman (2024), “Science diplomacy and the rise of technopoles”, Issues in Science and Technology, Vol. XLI/1, https://issues.org/science-diplomacy-technopoles-turekian-gluckman.

[69] UNESCO (2025), Science diplomacy in a rapidly changing world: Building peace in the minds of men and women., UNESCO, https://doi.org/10.54677/DTIP7777.

[3] US National Science Foundation (2024), The STEM Labor Force: Scientists, Engineers, and Technical Workers, US National Science Foundation, Alexandria, VA, https://ncses.nsf.gov/pubs/nsb20245.

[58] US Senate Committee on Commerce, Science, & Transportation (2022), View the CHIPS+ Legislation, https://www.commerce.senate.gov/2022/8/view-the-chips-legislation.

[33] White House Office of Science and Technology Policy (2024), White House Office of Science and Technology Policy Releases Guidelines for Research Security Programs at Covered Institutions, White House Office of Science and Technology Policy, Washington, DC, https://bidenwhitehouse.archives.gov/wp-content/uploads/2024/07/OSTP-RSP-Guidelines-Memo.pdf.

Notes

Copy link to Notes

← 1.  The 2023 edition of the OECD Science, Technology and Innovation Outlook defined “securitisation” as the reframing of regular policy issues, such as climate change, migration and emerging technologies, into matters of “security”. The term has a more common but unrelated use in finance.

← 2.  International STI linkages are wide-ranging. For example, academic researchers routinely co-operate and exchange across borders to advance shared scientific interests. Many researchers are also internationally mobile. In the private sector, innovative firms trade and invest internationally in the production of high-technology products and services. While international STI linkages can be led by states, they are more often built from the bottom-up, via individual researchers, research organisations and firms. The focus of this chapter is chiefly on research collaboration and international researcher mobility.

← 3.  International collaboration intensity can be measured as the number of a reference territory’s publications where the set of listed affiliations includes at least one affiliation abroad, as a percentage of the total number of publications attributed to that territory.

← 4.  China’s growing research capabilities have transformed the geography of international scientific collaboration over the last couple of decades. China’s spending on R&D was second only to the United States in 2023, it has the largest number of R&D personnel globally, and it is at the forefront in many areas of science and technology.

← 5.  Early-career researchers – doctoral and postdoctoral – are often internationally mobile, but tracking their career paths can be difficult, particularly if they stop publishing and are no longer visible through bibliometric data. The absence of comprehensive data on career paths is not only a challenge for policymakers but also a problem for early-career researchers who want to make informed career choices. The Research and Innovation Careers Observatory (ReICO) is a joint initiative of the OECD and the European Commission. It aims to be the premier source for reliable data and information on careers in research and innovation (R&I). The project’s goal is to create a dynamic information hub that tracks and analyses trends in R&I talent, career paths, and mobility in OECD and EU countries, as well as other economies.

One of ReICO’s core pillars is talent circulation. This theme focuses on the movement of R&I talent across institutions, sectors and borders. It explores career transitions and how mobility contributes to dynamic and interconnected innovation systems. ReICO seeks to improve the international comparability of data on researcher mobility by working closely with national contact points through co-ordinated annual data collections. These draw on national administrative records and survey data, complemented by international sources such as the OECD Database on Immigrants. In addition, ReICO will launch the ReICO Survey of the R&I Workforce, targeting individuals, to generate new insights. It will thus shed light on the patterns and dynamics of talent circulation and provide evidence to distinguish between brain drain, brain gain and brain circulation. See https://www.oecd.org/en/networks/research-and-innovation-careers-observatory.html for more information.

← 6.  Across OECD countries, the proportion of international students increases with the level of higher education. On average, they account for 5% of bachelor’s students, rising to 15% of master’s students and 25% of those in doctoral programmes (OECD, 2025[4]).

← 7.  See Mérat (2022[57]).

← 8.  Combined, the energy and environment SDGs account for approximately 28% of scientific production. The share of scientific publications that are deemed most likely to contribute to the energy and environment remained stable in the period 2008-2018. While the total number of indexed scientific publications grew steadily throughout the observed period, the share of publications relevant to the green transition stagnated through 2018 and moderately declined thereafter (OECD, 2025[6]).

← 9.  This section is based largely on OECD (OECD, 2025[6]).

← 10.  Citing the academic literature, the 2023 edition of the OECD Science, Technology and Innovation Outlook (OECD, 2023[1]) provided the following definitions: “‘technology sovereignty’ refers to a polity’s capacity to act strategically and autonomously in an era of intensifying global technology-based competition. A related concept, ‘strategic autonomy’, is broader and refers to a polity’s capacity to act independently in strategically important policy areas. It does not imply isolation or decoupling from the rest of the world, but rather describes a polity’s capacity to develop and manage international relations independently. It is tied to technology sovereignty, insofar as the latter creates opportunities to compete at technological frontiers, with positive impacts on the polity’s ability to influence global affairs. Countries’ capacity to successfully develop, integrate and use emerging and disruptive technologies in military applications is a traditional measure of their strategic autonomy, but this capacity also applies to many commercial technologies, particularly those with dual-use potential.”

← 11.  This chapter draws attention to four types of security and their implications for international STI linkages. The widest is global security, which includes food security, health security and environmental security, all areas where STI plays a central role. These cover many well-known global challenges, including pandemics, growing antimicrobial resistance, biodiversity loss, soil erosion and climate change. National security is traditionally associated with the military and other security forces, though broad definitions can also refer to various types of global security issues, such as pandemic preparedness. Defence-related national security strongly depends on advanced technologies, many of which are increasingly developed in the civil sector. Economic security refers to risks related to the resilience of supply chains, physical and cyber security of critical infrastructure, and the weaponisation of economic dependencies or economic coercion (European Commission, 2023[11]; OECD, 2025[66]). It is also concerned with technology leakage. Finally, the chapter also refers to research security, which is concerned with preventing undesirable foreign state or non-state interference with research (OECD, 2022[30]). These four types of security are often complementary, but they can also be in tension and involve trade-offs, particularly with respect to international STI linkages.

← 12.  This chapter’s focus on the research system is in contrast to the broader perspective of the 2023 edition (OECD, 2023[1]), which also covered “downstream” policy concerns, such as industry subsidies (promotion), supply chain vulnerabilities (protection), and strategic alliances and technical standards (projection). The aim in 2023 was to provide a high-level and broad overview of the growing securitisation of STI, whereas in 2025 the aim is to more closely explore the links between different policies, primarily in one part of the innovation chain (the research system); their implications for international research linkages; and the measures governments might take to ensure policies are proportional to risks and opportunities, designed in partnership with the main stakeholders, and drawing on a mix of knowledge and evidence.

← 13.  The OECD gathers publicly available data and measures industrial strategies across OECD countries through harmonised data on industrial policy expenditures, their composition, their mode of delivery and the characteristics of their beneficiaries. For further information, see: https://www.oecd.org/en/topics/sub-issues/quantifying-industrial-strategies.html.

← 14.  Mission-oriented innovation policies incorporate a similar ecosystem perspective, but with a narrower focus on fulfilling a specific mission, including technological missions (Larrue, Tõnurist and Jonason, 2024[56]).

← 15.  Measures like these are not just confined to OECD countries. In China, for instance, the 14th Five-Year Plan for National and Economic Social Development (2021-2025) and its underpinning Dual Circulation Strategy aim to achieve self-sufficiency in core technologies and reduce the country’s reliance on foreign technologies, such as advanced semiconductors, where it has critical dependencies (OECD, 2023[1]).

← 16.  Funds are dispersed through the Japan Science and Technology Agency (JST) and the New Energy and Industrial Technology Development Organization (NEDO). The act also has introduced a public-private co-operation council that actively supports R&D under the K Program by sharing valuable and sometimes sensitive information related to public and private needs and technological solutions. This includes information on security incidents involving private enterprises that are held by relevant administrative organisations, which had normally not been shared with researchers due to confidentiality obligations under the National Public Service Act.

← 17.  This section draws widely on a keynote presentation given by Professor Andrew James (University of Manchester) to the Committee for Scientific and Technological Policy’s 125th meeting on 6 November 2024.

← 18.  Inspired in part by the United States’ success in developing productive linkages between its civil and defence technology ecosystems, China has been pursuing a Military-Civil Fusion initiative for several years. It aims to create and exploit synergies between economic development and military modernisation, and encourages defence and commercial firms to collaborate and synchronise their efforts by sharing talent, resources and innovations (OECD, 2023[1]).

← 19.  While the German federal government is in favour of leveraging synergies between military and civilian research, it recognises the need for a holistic approach to security that sees both promoted in a more complementary way.

← 20.  The EC-OECD STIP Compass database has outlined information on almost 400 policy initiatives from 60 countries related to the international mobility of human resources. See: https://stip.oecd.org/stip/interactive-dashboards/themes/TH55.

← 21.  Furthermore, several countries have constitutional or other legal provisions regarding academic freedom and the institutional autonomy of universities.

← 22.  The thematic portal on research security in the EC-OECD STIP Compass provides a unique window into research security policy initiatives worldwide. The portal enables mutual learning across countries, focusing on the types of policy instruments countries are deploying and the specific policy concerns they seek to address. As of 2025, the portal contains information on 261 research security policy initiatives from 41 countries. Following the STIP Compass policy taxonomy, the top three policy instruments reportedly used are public awareness campaigns and other outreach activities; strategies, agendas and plans; and science and technology regulation and soft law.

← 23.  These initiatives often involve the creation of dedicated offices or units with mandates to oversee research security policy. They centralise responsibility and expertise for research security, potentially allowing more consistent policy development and enforcement. For example, the Office of the Director of the National Science Foundation (NSF) established in 2023 the new Office of the Chief of Research Security Strategy and Policy, which is responsible for co-ordinating all research security policies across the NSF. Its responsibilities include: identifying and addressing potential risks to the research enterprise; developing policy and best practices; conducting outreach and education; communicating reporting and disclosure requirements; establishing policies to ensure compliance; and, importantly, conducting due diligence on applications for NSF awards (US Senate Committee on Commerce, Science, & Transportation, 2022[58]). Organisations with similar remits are also being created at universities and research institutions, in the United States and elsewhere. For example, in a research security survey conducted in 2023 by the Korean Ministry of Science and ICT, nearly half of the more than 90 Korean research institutes that responded had a dedicated body for research security (Presidential Advisory Council on Science and Technology, Korea, 2023[62]).

← 24.  This list-based approach is complemented by a broader risk assessment approach requiring general due diligence on research projects (including beyond the scope of the list), as per the complementary policy entitled the National Security Guidelines for Research Partnerships. It is also notable that the Canadian policy does not identify specific countries, but rather focuses on the risk profile of individual research organisations, only prohibiting affiliations with those that are assessed to pose the highest risk to Canada's national security (Innovation, Science and Economic Development Canada, 2024[35]).

← 25.  The German Research Foundation underlines that individual and institutional applicants wishing to co‑operate with international partners need to explain the potential risks and benefits of the collaboration, with more detailed justifications typically required for projects with greater risks or appearance of risks (German Research Foundation, 2023[59]). The German Science and Humanities Council published a position paper on “Science and security in times of global political upheaval” (German Science and Humanities Council, 2025[65]), which gives recommendations for dealing with knowledge risks in order to protect and build a resilient society. While the recommendations by the German Science and Humanities Council and the German Research Foundation are country agnostic, the Max Plank Society and the German Academic Exchange Service have published specific papers regarding interaction with China. The Max Plank Society emphasises the need for mutual understanding and a culturally sensitive approach for sound decisions and balanced co-operation with China (Max Planck Society, 2023[60]). According to the German Academic Exchange Service’s recommendations, interaction with Chinese partners should be interest-oriented, risk-aware and competence-based (German Academic Exchange Service, 2024[61]).

← 26.  TRUST applies a decision tree for assessing research proposals and ongoing projects regarding personnel appointments and research support, non-compliance with disclosure and other requirements, and potential risks to national security. See National Science Foundation (National Science Foundation, 2024[63]).

← 27.  For further details, see: https://www.jst.go.jp/osirase/research_security/index_e.html. More broadly, the policy discussion on research security in Japan surged in 2024 during economic security policy discussions (e.g. the National Security Secretariat’s expert panel on countermeasures against leakage of critical technologies). Following this, the Ministry of Education, Culture, Sports, Science and Technology issued its Report on Approach for Ensuring Research Security at Universities in December 2024 https://www.mext.go.jp/content/20250423-mxt_kagkoku-000019002_2.pdf). The CSTI Secretariat Expert Panel launched discussions to develop research security guidelines in April 2025.

← 28.  For instance, examples featured on UNESCO’s webpage on science diplomacy [https://www.unesco.org/en/scientific-research-cooperation-why-collaborate-science-benefits-and-examples] include CERN in Switzerland, the International Centre for Theoretical Physics in Italy and SESAME in Jordan. UNESCO has also published in 2025 its report, Science diplomacy in a rapidly changing world: Building peace in the minds of men and women (UNESCO, 2025[69]).

← 29.  The growing popularity of the science diplomacy concept has led to some dilution of its meaning. As noted by the Geneva Science Diplomacy Anticipator (2025[54]), its transdisciplinary nature has made science diplomacy attractive across academic, policy and diplomatic communities, but also prone to being used as a catch-all label for any initiative involving international science collaborations, even when these lack strategic intent or demonstrable impact on foreign policy or international governance. The proliferation of the concept risks obscuring this important distinction, underscoring the need for clearer frameworks to ensure science diplomacy remains purpose-driven, coherent and impactful.

← 30.  The uneven distribution of research infrastructure capacities at the global level prevented equitable access to resources and data in many parts of the world during the COVID-19 pandemic, contributing to a disconnect between needs and resources. OECD country research funders recognised the problem, allocating around USD 200 million globally for COVID-19 projects that aimed to strengthen research capacity in LMICs, most of which focused on reinforcing laboratory capacity. Such a strengthening of research capacity can be an important contribution to health-crisis preparedness, but needs to be extended to provide effective global action for other ongoing and future challenges (OECD, 2023[1]).

← 31.  For further information, see: https://www.gov.uk/world/organisations/uk-science-and-innovation-network.

← 32.  For example, more information on Microsoft’s work with the United Nations and international organisations can be found at: https://www.microsoft.com/en-us/corporate-responsibility/united-nations.

← 33.  For further information, see the Office of Denmark’s Tech Ambassador at: https://techamb.um.dk/the-techplomacy-approach.

← 34.  For further information, see: https://www.eeas.europa.eu/delegations/united-states-america/san-francisco_en?s=253.

← 35.  For further information, see: https://swissnex.org/about-us/mission-and-organization.

← 36.  The European Economic Security Strategy (see Box 2.2) also identifies proportionality and precision as fundamental principles for any measures on economic security (European Commission, 2023[11]).

← 37.  Open science is a policy priority for all OECD countries. The EC-OECD STIP Compass database provides a window into how different countries are responding at the national policy level to promote open science. Its STI policies for Open Science portal provides information on close to 700 policy initiatives from more than 60 countries and the European Union and is regularly updated. While the information is certainly not fully comprehensive for all countries, it does provide a meta-view of where countries are focusing their efforts and, at the level of individual initiatives, provides access to summary information and links that are a valuable starting point for those who want to delve further. The portal also provides ready access to published reports and articles relating to open science policy from the OECD, other international organisations and relevant public repositories.

← 38.  Emerging frameworks suggest potential paths forward to promote complementarities between research security and open science. For example, in its 2023 Research Security System Improvement Plan, Korea notes that it must continue to promote international co-operation to drive innovation, even while adopting research security management strategies. A 2024 JASON report formalises this intuition by suggesting that technology maturity – as measured through the TRL framework – can guide institutions’ decisions on when imposing additional controls, versus maintaining openness, might best support national security (broadly defined to include economic security). The authors suggest that while potential national security issues may be apparent as early as fundamental research stages (TRLs 1 and 2), it is not until technologies reach the pilot and demonstration phase (move from TRLs 5 to 6) that their actual significance to national security can be demonstrated and warrant substantive mitigation efforts (JASON, 2024[55]).

← 39.  For example, Finland’s Constitution secures university autonomy. When ensuring research security, a researcher’s right to choose their research topic and methods cannot be restricted. This means that the applicants for Research Council of Finland funding need to take research security into account as part of the application for research funding. For this reason, the Act on Research Council of Finland was amended in May 2025 to include a paragraph on research security. The aim is to encourage research-performing organisations to identify potential risks and threats related to security in advance. The Research Council’s task is to ensure that due consideration be given to research security and to the risks associated with it in research projects, research co-operation and the use of research results.

← 40.  For instance, at the multilateral level, the European Union’s Council Recommendation on Enhancing Research Security, adopted in May 2024, was developed with significant input from R&I stakeholders. The result has been a strongly positive reception from key associations such as the League of European Research Universities, an association of 24 leading research-intensive universities in Europe (see: https://www.leru.org/news/leru-welcomes-proposals-for-more-secure-research-in-the-future). At the national government level, Korea’s recent efforts to develop its comprehensive Plan for Strengthening the Research Security System for the Establishment of a Trusted Research Ecosystem not only engaged 14 universities with industry-academia co-operations (OECD, 2023[68]) but it also linked to the Research Security Advisory Committee, composed of research experts and security experts (see: https://www.msit.go.kr/bbs/view.do?sCode=user&mId=113&mPid=238&bbsSeqNo=94&nttSeqNo=3183414).

← 41.  For further information, see: https://www.aaas.org/programs/center-science-diplomacy/about.

← 42.  For further information, see: https://www.science-diplomacy.eu.

← 43.  For further information, see: https://diplomaciacientifica.org.

← 44.  For further information, see: https://www.africasciencediplomacy.org.

← 45.  For example, Germany is currently working on developing a strategic approach that connects the constitutionally protected freedom of science with German security and economic interests while maintaining a culture of enabling reciprocal international research co-operation with reliable partners. The German Federal Ministry of Education and Research (known since May 2025 as the Federal Ministry of Research, Technology and Space) has initiated a national process bringing together stakeholders from the scientific community, the federal ministries and governments of the Länder, industry, and intelligence services. The aim of the ongoing process is to develop a common understanding on research security and essential measures to improve the status quo.

← 46.  For example, the Australian Research Council’s Foreign Interference and Security Risk Internal Audit identified one of the highest priority gaps as being the absence of an overarching framework clarifying different actors’ roles and responsibilities (https://www.arc.gov.au/funding-research/research-security).

← 47.  For further information, see the National Contact Point for Knowledge Security, Ministry of Education, Culture and Science, at: https://english.loketkennisveiligheid.nl.

← 48.  For example, in 2023 the European Commission published a recommendation on critical technology areas for the European Union’s economic security (European Commission, 2023[64]). It has also set up a Critical Technologies Observatory which identifies, monitors and assesses critical technologies for the space, defence and related civil sectors and their potential application and related value and supply chains. It also monitors and analyses existing and predictable technology gaps, the root causes of strategic dependencies, and vulnerabilities. Based on these data, the European Commission prepares a classified report for member states on critical technologies and risks associated with strategic dependencies affecting security, space and defence every two years. It also prepares technology roadmaps based on these reports, which include mitigating measures to boost research and innovation and reduce strategic dependencies affecting security and defence (European Commission, 2025[67]).

← 49.  For example, the NSF Office of the Chief of Research Security and Policy released its Guidelines for Research Security Analytics in 2023 to support the implementation of its congressionally mandated role. This includes performing risk assessments of NSF proposals and awards using analytical tools to assess non-disclosures of required information. These guidelines include a breakdown of which agency personnel may conduct research security-related activities, what monitoring activities are allowed and with what resources they are conducted, how information will be validated to ensure accuracy, and how information may be shared within the NSF and externally. This level of specificity not only clarifies roles within the NSF but also sets a standard for accountability in research security.

← 50.  For example, in the United States, the NSF has launched the Research on Research Security Program to assess the methods for identifying research security risks and strategies for preventing and mitigating them. The programme seeks to better understand the nature, scope, challenges and potential of this field – including the critical areas of cybersecurity, foreign travel, research security training and export control training – with the aim of informing best practices and guidance for the research community. For further information, see: https://new.nsf.gov/news/nsf-announces-research-research-security-program.

← 51.  The NSF has also announced a 5-year USD 67 million investment to establish the Safeguarding the Entire Community of the US Research Ecosystem (SECURE) Center, an information clearinghouse. The SECURE Center will disseminate information and reports on risks of foreign interference, provide research security training to relevant communities, and serve as a bridge between the research community and government funding agencies on security concerns. For further information, see: https://new.nsf.gov/news/nsf-backed-secure-center-will-support-research.

3. Expanding the benefits of investments in science, technology and innovation

Abstract

This chapter discusses the importance of broadening the benefits from and participation in innovation across different social groups, regions and industries. It discusses how the particular challenges facing science, technology and innovation (STI) policymakers in 2025 – accelerating frontier technology development, building resilience and improving sustainability – interact with the channels through which the benefits of innovation are distributed. It concludes by identifying key implications for STI policymakers.

Key messages

• Long-standing challenges, changing contexts. Many of the participation and inclusion challenges facing science, technology and innovation (STI) policy in 2025 – such as underrepresentation, concentration of opportunity and the importance of diffusion – echo patterns observed throughout the history of technological change. While the participation imperative is not new, the evolving context of rapid technological shifts, geopolitical competition and urgent societal challenges makes it even more vital for policymakers to address barriers that have persisted across past transformations.

• Innovation is inherently concentrated, but broader benefits require deliberate diffusion efforts. Innovation activities naturally cluster among leading firms, sectors and regions due to economies of scale and knowledge spillovers. However, translating these concentrated innovations into economy-wide productivity gains and societal benefits requires dedicated policies and investments in diffusion mechanisms.

• Broadening participation is a key lever for expanding innovation benefits. Beyond the diffusion of existing technologies, widening who participates in creating and shaping innovation directions can enhance both the quality and societal relevance of technological development while ensuring benefits reach more diverse populations.

• Regional and industrial concentration shapes inclusion opportunities. Innovation activities are highly concentrated among leading firms, sectors and regions, creating uneven opportunities for participation and benefit. Workers in highly innovative environments have greater access to high-value jobs and career advancement while those in lagging areas face limited prospects.

• STI policies may face trade-offs between excellence and inclusion. Policies directly promoting participation and those supporting centres of excellence for reasons of expediency play important roles for STI policies. What is key for the STI policy mix is supporting frontier advancement while facilitating broad-based access to benefits and participation.

• Co-ordination across policy domains is essential. Persistent participation gaps despite targeted STI interventions demonstrate that achieving inclusion requires co-ordinated action across education, labour market, social and regional development policies. STI policies alone cannot address deep-rooted structural barriers and need to be co-ordinated with other policy domains.

Introduction

Technological innovation and diffusion significantly improve people’s lives across and beyond the OECD. Whether through improvements in labour productivity, the quality of and access to medical and educational services, or in ensuring stable and affordable access to energy, technological innovation and diffusion can be significant forces for positive economic, societal and environmental progress. As policymakers increasingly direct STI investments toward strategic objectives – competitiveness, resilience and sustainability – this chapter examines why and how ensuring their broadly distributed benefits has become essential for policy success.

Broadening the participation in and widening the benefits from innovation have been recurring themes in international STI policy discussions, including in the 2018 and 2021 editions of the OECD Science, Technology and Innovation Outlook (OECD, 2018[1]; 2021[2]; Paunov and Planes-Satorra, 2021[3]). These discussions underscore the point that wider participation in technological innovation and diffusion not only enhances economic outcomes but can also improve the quality and societal relevance of innovation itself. While these insights remain valid for 2025 and beyond, policymakers face new challenges arising from emerging trends in STI.

The first challenge is the post-COVID increased focus in directed STI policies addressing strategic imperatives such as technological resilience and competitiveness (OECD, 2024[4]; Paunov and McGuire, 2022[5]). While this strategic focus is essential for national competitiveness, the risk is that the urgency and scale of these investments may inadvertently reinforce existing patterns of concentration. When finite public resources are rapidly deployed to achieve technological breakthroughs, they often flow to established centres of excellence – leading research institutions, flagship companies and innovation hubs – that can deliver results quickly, potentially widening gaps with other regions and actors unless accompanied by parallel investments in diffusion and capacity building.

The second consideration is that the vast majority of employment across the OECD is in low or medium innovation-intensive sectors, including those such as healthcare and education, where technological diffusion could significantly improve productivity and social outcomes. It is therefore important to ensure that efforts to push the technological frontier do not impede efforts to accelerate the diffusion of technological innovation where uplifts from productivity could be both economically and socially significant.

The third consideration is that innovation-intensive manufacturing industries face the dual challenge of adopting digital and low-carbon technologies simultaneously, creating a temporal challenge for firms and workers in those industries to contribute to and benefit from technological innovation.

The fourth consideration is that market concentration dynamics in some innovation-intensive sectors may affect the distribution of innovation opportunities and benefits. Understanding these dynamics and their implications for STI policy design requires policy attention and potential co-ordination with other policy domains (OECD, 2024[6]; 2023[7]).

These dynamics suggest several areas where STI policymakers can take action:

• Strengthening diffusion policies to broaden participation in and benefits from technological innovation, revisiting best practices from technological, regional and sectoral diffusion initiatives for the contemporary context.

• Embedding strategic reflections in the policy design process on how the strategic agenda‑setting for STI may impact participation in and benefits from publicly funded STI, and how policymakers in STI and other policy areas could address the economic and social implications of these impacts.

• Better understanding of the relationship between competition and technological innovation and diffusion, both within and across economies.

This chapter begins with a brief overview of the differentiated and at times lag effects of technological innovation and its diffusion through history. It then proceeds with a discussion of a simplified framework to clarify how participation in innovation affects the direction of technological progress and the distribution of its outcomes. The chapter then discusses how participation in technological innovation and its diffusion interact with the imperatives facing STI policymakers today, such as competitiveness and resilience. It then outlines some of the key channels through which the framework set out could inform STI policy in 2025 and the years ahead. The chapter concludes with some overarching considerations for policymakers as they consider the importance of broadening the participation in and benefits from technological innovation as they design and implement STI policies in the years ahead.

Why should policymakers care about broadening participation in science, technology and innovation?

The transformative changes currently underway, including the digital and green transitions, share critical characteristics with transformative technological changes throughout history, particularly in how they create concentrated benefits during development phases and uneven diffusion patterns that can limit the ability of all sectors of the population to benefit. Historical analysis of major technological transitions, notably the First Industrial Revolution (1760-1840), reveals persistent patterns where the benefits of breakthrough innovations initially concentrate among those with access to capital and technological infrastructure while broader societal gains emerge only through deliberate diffusion mechanisms and often significant social and political intervention (Acemoglu and Johnson, 2024[8]; 2023[9]; Hobsbawm, 1962[10]).

The mechanisation of British textile production, for instance, demonstrates how technological advancement can simultaneously drive productivity growth while displacing entire categories of skilled workers – handloom weavers saw their real wages more than halve between 1806 and 1820 as power looms replaced traditional craftsmanship, illustrating the stark short-term distributional consequences of technological change (Feinstein, 1998[11]; Voth, 2003[12]).

What distinguishes contemporary digital and green transitions from earlier technological revolutions is not just their speed and scope, but the recognition that broad-based and widely shared benefits from innovation are neither automatic nor guaranteed. Unlike previous eras where distributional consequences were often treated as inevitable byproducts of progress, today’s transitions occur within policy frameworks that explicitly acknowledge the need for participatory approaches to technology development and diffusion.

The evidence from historical transitions suggests that without deliberate intervention, technological progress can create persistent inequalities, even after decades of economic growth. Improvements in living standards for displaced workers during the Industrial Revolution were minimal and required sustained social and political action to materialise (see Feinstein (1998[11]); Allen (2007[13]); or Acemoglu and Johnson (2024[8])). This historical perspective underscores why current policy debates around frontier technologies like artificial intelligence (AI), quantum computing and clean energy systems must simultaneously address the concentration of development capabilities and the mechanisms for ensuring broad-based access to the benefits of technological progress.

The framework that follows in this chapter builds on these historical insights to distinguish between development dynamics, which determine who participates in innovation processes and where innovation capabilities concentrate, and diffusion dynamics, which shape how innovations spread across different populations, regions and economic sectors over time. Understanding this distinction is crucial for contemporary STI policy, as the evidence suggests that addressing inclusion challenges requires different approaches at the development stage (where the focus is on broadening participation in frontier research and innovation) versus the diffusion stage (where the emphasis shifts to ensuring widespread adoption and benefit-sharing across diverse communities and regions).

How does technological innovation impact society, and how does broader participation in innovation affect the direction of technological innovation?

There are two key dimensions for understanding how technological innovation interacts with socio‑economic outcomes. The first is to consider how the development and diffusion of new technologies affects the outcomes of different groups, and the skills and capacities they may require to benefit from these technologies. The second is to consider how participation in the development of technological innovation affects the direction and quality of that innovation. These two dimensions interact in ways that affect the inclusiveness of the gains from innovation (e.g. better health outcomes, better education access, higher productivity) and the alignment of innovation with social needs.

These outcomes are how different groups in society benefit from new technologies and innovations in the following ways (Figure 3.1):

• Direct impacts [quadrant A]: Different socio-economic groups may benefit differently from new technologies and innovations. Several conditions determine access to the benefits of innovation, including price, geographic availability, required infrastructure, user capabilities (such as digital or technical skills) as well as the very purpose of the innovation, which may cater to the needs of specific groups of the population. Social innovations are a specific sub-set of innovations aimed at addressing unmet social needs – particularly those affecting marginalised or underserved groups.

• Indirect impacts [quadrant B]: Wider economic and social impacts accrue indirectly by affecting returns to labour and capital impacted by innovation and technology. Changes in production processes driven by innovation, such as automation and the wider application of AI in production, can alter demands for skills or capacities and affect demands for different types of assets (capital, land, etc.). Moreover, technological progress can change the relative returns to labour and capital (see Autor et al. (2017[14]) and Guellec and Paunov (2017[15]) for a discussion).

Participation refers to the opportunities that different groups in society have to shape technological progress in the following ways (Figure 3.1):

• Direct participation [quadrant C]: Individuals can engage directly in the design and development of technology and innovation by being part of the research and innovation workforce. Such engagement requires having specialised skills and capacities, which vary depending on the types of activities undertaken. Beyond professional research and innovation roles, citizens more generally can engage in a myriad of other ways, ranging from weak engagements, such as providing user feedback that influences product development, to more substantial involvement, like contributing to open-source and citizen science projects (OECD, 2025[16]).

• Indirect participation [quadrant D]: Individuals may also engage indirectly in shaping innovation and technology development by participating in industry or policy decision-making processes (e.g. taking key investment decisions, shaping institutional choices over adoption); engaging in public consultations, participatory technology assessment or participatory research agenda-setting exercises influencing the directions of public or private choices on innovation and technology; and, more passively, shaping demand.

The relationship between innovation and inclusion can, therefore, be explored through four dimensions, as illustrated in Figure 3.1.

Figure 3.1. Four key dimensions of participation in innovation and its outcomes

What are the implications for science, technology and innovation policy today?

Building on the historical context and framework from the previous section, this section examines how participation dynamics interact with today’s STI policy imperatives. Key priorities policymakers face in 2025 include accelerating the development of strategic technologies, navigating green and digital transitions while ensuring competitiveness, enabling broad-based diffusion to meet productivity needs, building consensus around STI directions, and operating with agility amid fiscal constraints. Where possible, evidence from gender, regional and industrial participation is provided to illustrate these dynamics.

Navigating excellence and inclusion in an era of strategic competition

This subsection addresses how the framework’s distinction between outcomes and participation – across both direct and indirect channels – plays out at the frontier of STI policy. Frontier-oriented investments and “directional” STI policies are intended to deliver significant economic and technological gains, but they risk concentrating both the direct benefits (A: who reaps the returns from new STI products) and the indirect structural effects (B: who benefits from changing skill and capital demands) among established actors. Frontier-oriented STI investments thus face a resource allocation challenge: directing finite budgets toward established centres of excellence – which may be entirely rational for achieving rapid breakthroughs – can inadvertently concentrate both innovation returns and structural benefits among already capable actors. This creates tension between the immediate imperative to develop strategic technologies and the broader goal of ensuring widespread access to innovation benefits, particularly when existing diffusion mechanisms are already strained.

The rapid pace of digital and green transitions, combined with intensifying geopolitical competition around frontier technologies, has fundamentally altered the policy landscape for STI systems. Major economies have launched significant directed investments – such as the European Union’s European Chips Act – that aim to build capabilities in AI, quantum computing, semiconductors and clean energy technologies. These policies reflect a new era where STI policy is increasingly “directional”, pursuing specific mission-oriented objectives while navigating complex trade-offs between technological leadership and inclusive participation (Arnold et al., 2023[17]; Larrue, 2021[18]; Mazzucato, 2018[19]; OECD, 2024[4]).

A variety of metrics show that technological innovation is already highly concentrated at the firm, sectoral and regional levels. For evidence at the firm level, data on top research and development (R&D) investors show that the top 100 companies (in terms of R&D investments) account for around a striking 50% of global R&D in 2023 (Figure 3.2) There is also significant concentration within that group of top R&D investors: on average, the top 10 companies invested more than double the amount than the top 50 (including the top 10), and the top 50 invested on average over 50% more than the top 100. Top R&D investors also account for an important part of national R&D.

Figure 3.2. Average R&D investments of the top global 2 000 companies, 2023

Source: European Commission: Joint Research Centre, NINDL, E., NAPOLITANO, L., CONFRARIA, H., RENTOCCHINI, F., FAKO, P., GAVIGAN, J. and TUEBKE, A., The 2024 EU Industrial R-D Investment Scoreboard, Publications Office of the European Union, Luxembourg, 2024, https://data.europa.eu/doi/10.2760/9892018, JRC140129.

Statlink

Innovation is also geographically concentrated in a limited set of “innovation leader” regions across OECD countries, which collectively generate the bulk of national R&D and scientific output. While such a concentration can encourage economies of scale and knowledge spillovers, persistent innovation leadership and a decline in entry and job reallocation rates signal rising barriers to new entrants. However, this concentration pattern may be evolving in frontier technology areas: in fields like AI and quantum computing, smaller firms such as OpenAI, Anthropic and specialised quantum start-ups are driving significant breakthroughs alongside established tech giants, suggesting that technological disruption can still create opportunities for new entrants to challenge incumbent advantages.

In the context of public support for frontier technology development, the question for policymakers is to what extent the pursuit of research excellence and support for pre-existing clusters with strong capacities reinforces these concentration dynamics, and what are the implications of that concentration? When governments allocate finite resources to leading firms, research institutions and regions that already possess established capabilities, they may achieve technological breakthroughs more efficiently but risk widening gaps with lagging actors (quadrants B and C in Figure 3.1). For instance, while progress in AI and automation may benefit those with complementary skills and capital assets, it can simultaneously reduce returns to routine labour and concentrate benefits in technologically advanced regions. The challenge for policymakers is determining when such concentration is a necessary short-term cost for long‑term competitiveness and when it creates structural barriers that ultimately undermine innovation system resilience.

Concentrating resources in frontier technology development can be a highly effective strategy, especially when leading firms, industries and research institutions – often clustered in specific regions – already possess strengths that policies can leverage. In a globally competitive environment, building on these existing capabilities may offer the best chance for success. Spreading resources too broadly can dilute their impact, resulting in well-intentioned but ultimately ineffective outcomes. Furthermore, technological diffusion may occur over time through market forces, collaboration and talent mobility. They also depend on whether the effects of concentration are temporary or long-lasting, i.e. whether frontier developments will eventually lead to improved outcomes that spread more broadly across society over time.

The downside, however, may be a widening gap with those already lagging behind, challenging also the wider diffusion of frontier technologies, which in turn can reduce possibilities for their further development as demand and user experiences play important roles in shaping innovation pathways. The allocation of limited public resources to firms and institutions that already have well-established capabilities may further reduce possibilities for others to participate in frontier technology development and benefit from its diffusion.

Recent OECD work has widely discussed tension between concentration (which may promote excellence and rapid technological development) and inclusion. The OECD’s Industrial Policy Framework demonstrated that effective industrial strategies must balance productivity growth with addressing societal challenges, including inclusion, through co-ordinated approaches that address complementarities between different policy instruments (Criscuolo, Gonne and Lalanne, 2022[20]). Recent OECD analysis shows that industrial policies increasingly target societal goals, with green transition objectives comprising 18.6% of national STI strategies, followed by social and regional inclusion at 9.9% (Paunov and Einhoff, 2025[21]). 

The potential unintended negative impact of these STI policies (as well as the above-mentioned frontier technology policies) on broader participation does not imply that they should not be pursued, as they address other important policy priorities – such as economic competitiveness, resilience and national security. Strategies that support frontier technology development are an example of the many rationales justifying STI policy. From a Schumpeterian perspective, technological change inherently involves disruption and renewal, and OECD countries have well-established social policy frameworks that can help mitigate the social and regional impacts of such transitions. The key point is to recognise that technology policy choices are not neutral, and that they have an impact on the distribution of wealth, income and opportunities across firms, industries and places (and ultimately individuals) well into the future. The other key insight is that there is a role for STI policy to complement other efforts in shaping the nature of transitions.

From development to diffusion and adoption: The need for policy differentiation

The relationship between knowledge creation (quadrants C and D in Figure 3.1, who participates in and shapes STI development and governance) to widespread technology adoption and benefit-sharing (quadrants A and B in Figure 3.1, who directly and indirectly gains from STI products and changing economic returns) reveals distinct yet interdependent policy challenges. Effective STI systems require not only excellence at the frontier, but also robust, inclusive mechanisms for diffusion, ensuring diverse actors have the capacity and opportunity to participate in innovation and be able to enjoy its benefits.

Frontier-oriented STI policies now face a double mandate: continue pushing the technological boundary and consider how diffusion and adoption policies can be integrated into these policies. Recent OECD evidence on AI adoption shows why this matters: uptake is still twice as high in large firms and advanced regions as in smaller firms and peripheral areas, owing to scale-dependent fixed costs, data requirements and superior absorptive capacity (OECD, 2024[22]).

Effective adoption of innovation is essential for ensuring that the benefits of new technologies reach a broad swathe of society and the economy but is hindered by highly uneven spatial and sectoral innovation activity. Workers in “leader” firms, sectors and regions consistently achieve higher wages and revenues, while those in “laggard” environments capture fewer gains. For example, patenting is overwhelmingly concentrated in large urban areas, with 90% of patent applications coming from urban inventors and large urban regions having significantly higher patenting rates than medium-sized or smaller areas (OECD, 2024[23]).

This reflects both the structural advantages of co-location and the cascading disadvantages faced by less connected or less innovative regions. Moreover, increasing entrenchment among market leaders, as seen in the declining rates of firm entry and job mobility, limits the spread of new technologies to lagging places and firms. Addressing these challenges requires targeted policy interventions – such as improving infrastructure and connectivity, advancing skills development in underperforming areas, and supporting technology adoption in small and medium-sized enterprises (SMEs) – to facilitate broader and more equitable diffusion and participation throughout the innovation system.

The push to align economic competitiveness with sustainability presents its own particular diffusion challenges. The evolution of labour and skill demands in emerging sectors, such as the move to electric vehicle production, requires new competencies not always possessed by workers in sunset industries (Curtis, O’Kane and Park, 2023[24]). If skills development and retraining lag behind technological change, this transition may end up benefiting only a narrow pool of workers in specific regions or firms. STI policy must, therefore, link innovation support to strategic workforce development and ensure that knowledge, infrastructure and market opportunities reach participants beyond traditional innovation leaders, including those in less advantaged sectors and places.

Policy design has to distinguish between two separate bottlenecks. Knowledge diffusion determines who can join the inventive process while technology adoption decides who can turn new ideas into social and economic value. Because the obstacles differ, skills and research networks on the one hand, data readiness, finance and managerial know-how on the other, each stage needs its own toolkit. Capacity building must simultaneously broaden participation in technology creation and equip diverse regions, sectors and communities to adopt innovations at scale.

Diffusion policies, whether embedded in technology frontier programmes or deployed in parallel, are the main lever for widening the benefits of and participation in STI; Table 3.1 shows examples of relevant diffusion policies. They only work when actors possess sufficient absorptive capacity, i.e. the ability to spot, absorb and apply external knowledge. Investing in those capacities is thus essential to share the gains from innovation more equitably, including across borders. For developing economies, building such capabilities is a prerequisite for meaningful engagement in global STI systems.

Translating this into practice poses a governance challenge: responsibilities for funding and decision making must be split judiciously between national and subnational authorities so that national strategic aims align with regional strengths and opportunities. Regional policymakers, in particular, need diagnostic tools to see where their ecosystems can credibly participate in strategic, frontier-technology domains.

Survey data from the G7 and Brazil (OECD, 2025[25]) reveal that diffusion frictions – acute skill shortages, low data maturity and uncertainty over returns – now trump simple awareness gaps. Firms rate three policy responses the highest:

1.modernised qualification frameworks plus hands-on, sector-specific training

2.higher quality, easily accessible public data

3.streamlined collaboration with universities and dedicated diffusion agencies.

Although dedicated technology diffusion agencies are well regarded, they currently serve only a minority of firms, underscoring the need for scalable sign-posting services, SME-oriented vendor-selection guidelines and clear accountability frameworks for safe AI use. Comparable international surveys and rigorous evaluation of these agencies are critical to identify and share what works.

Table 3.1. Examples of science, technology and innovation diffusion policies

If policymakers want the benefits of the development and diffusion of technologies to be truly broad-based and shared, STI policies need to dramatically accelerate efforts to address participation imbalances. While this subsection focuses on gender as a salient lens, the structural and systemic challenges described here apply – often with added complexity – to other groups that face barriers, such as people from lower income backgrounds, minority groups and regions that have undergone significant deindustrialisation.

Over the past decade, women’s participation in STI has notably increased, although gaps persist. Globally, the share of 25-34-year-olds with tertiary education rose from 23% to 27.5% between 2013 and 2021, and in OECD countries from 45.6% to 53.7% (OECD, 2024[32]), outpacing men in both cases. Yet this progress masks persistent disparities in key fields. In 2021, just one-third (32.5%) of graduates in science, technology, engineering and mathematics (STEM) were women, up only marginally from 31% in 2013. Representation varies sharply by discipline: women make up a slight majority in natural sciences, mathematics and statistics (53.6% of graduates), but only 27.8% in engineering and 22.7% in information technology.

These educational gaps carry through into research and innovation careers. Despite a modest increase – from 34.7% to 35.6% between 2013 and 2021 – on average women still account for just over a third of R&D personnel in OECD countries. National shares vary considerably, with some countries approaching gender parity (Iceland, Latvia and Lithuania) while others continue to record comparatively low shares despite progress in recent years (Czechia, Korea and Japan at ~22%). In patenting, the share of women inventors fell from 13.4% in 2013 to 11.3% in 2019 and remains below 7% in some innovation‐intensive economies such as Austria, Germany and New Zealand.

Improving women’s participation in STI has been a policy focus for a long time, with policymakers implementing targeted financing schemes, such as scholarships and research grants, for women to engage in STI training and activities. Table 3.2 provides examples of policy initiatives in this area.

Table 3.2. Examples of policy initiatives for women’s participation in science, technology and innovation

This subsection foregrounds quadrant D of the framework illustrated in Figure 3.1 – diversity in decision making and leadership in STI – by highlighting that participatory governance is increasingly critical as technological transformations accelerate. As technological change profoundly reshapes society, the capacity for a range of groups – not just technical experts – to influence the direction, priorities and norms of innovation becomes central to the legitimacy, equity and societal alignment of STI policy. Ensuring effective and democratic participation in STI governance elevates indirect forms of participation to a policy priority and brings the interplay between who shapes innovation and who ultimately benefits from it into sharper focus across all four quadrants.

The scale and societal implications of transformative technological developments require participatory approaches to STI policy design (quadrant D of Figure 3.1). As in previous technological revolutions, today’s frontier technologies will fundamentally reshape work, governance and social relations in ways that technical experts alone cannot fully anticipate or evaluate; the direction of technology development and diffusion is also much more actively shaped by policy than in the past.

The 2024 OECD Framework for Anticipatory Governance of Emerging Technologies emphasises early societal engagement to surface concerns, inform design choices and guide innovation toward more equitable outcomes (OECD, 2024[38]). This requires investing in the capacity building, consultation processes and institutional arrangements that enable diverse groups to contribute meaningfully to shaping technological directions.

When thinking about participation in the governance and agenda-setting of STI going forward, it is important to note that policymakers do so in a context where historical participation gaps remain unresolved. Women, for example, remain under-represented across entrepreneurial and leadership roles in STI. An analysis of a dataset of start-ups listed on Crunchbase covering firms founded between 2000 and 2017 in OECD and BRICS countries found that women-only founding teams accounted for less than 6% of all start-ups, while those with at least one female co-founder made up approximately 15% (Lassébie et al., 2019[39]). Evidence on European venture capital (VC)-funded start-ups based on data from private sources (Pitchbook and the European Data Cooperative) covering 39 000 investors and 85 000 entrepreneurs that were active in Europe between 2011 and 2021 (accounting for 80% of total VC firms and 52% of start-ups) indicates that, between 2011 and 2021, women comprised only 10% of founders and chief executive officers (CEOs) in those start-ups, while start-ups led solely by women captured a mere 2% of total VC funding. Among investors, only about one in seven senior-level VC investors was female, 90% of whom worked in predominantly male teams (European Investment Fund, 2023[40]).

These participation imbalances extend into senior roles across sectors shaping the future of STI. In 2020, fewer than 5% of Silicon Valley 150 companies had female CEOs, and women held only 26-34% of senior posts at Google, Apple, Facebook, Amazon and Microsoft (European Centre for Women and Technology, 2024[41]). In pharmaceuticals, just 17% of board seats are held by women (ISPE, 2021[42]), and in deep‑tech start-ups under one-quarter of founding teams include a woman, with overall female founder share at 14% (Davila et al., 2024[43]). Academia and public administration show similar patterns – women lead 23.6% of higher education institutions and occupy 31.1% of board roles in the European Union (European Commission, 2021[44]).

Figure 3.3. Gender equality in senior management positions in national administrations, 2011 and 2021

Note: Organisations covered are central administrations, also referred to as ministries and/or departments of a national government led by a minister. Senior administrators are the sum of level 1 and level 2 administrators. Level 1 administrators include all administrative (non-political) positions from the head of the ministry down to the level of head of directorate or similar, where a directorate is a major section within the ministry. Level 2 administrators include all positions below the head of directorate down to the level of head of division/department, where a division/department is the first level of organisation below the directorate (i.e. the second level of functional organisation). Data refer to the OECD-EU countries plus Iceland, Norway the United Kingdom and Türkiye.

Source: OECD (2023[45]) based on European Institute for Gender Equality (EIGE) Gender Statistics (database); women and men in decision-making (WMID) authorities.

Statlink

Similarly, women are also under-represented in public decision-making bodies. In August 2023, they accounted for 33% of members of parliament in national parliaments across the European Union. Only six national parliaments had more than 40% women members, while seven had less than 25%. In the case of senior management positions in national administrations, gender disparities tend to be lower and decreasing over time, although significant differences remain across countries (Figure 3.3) (EIGE, 2024[46]). When looking at research-funding organisations in the European Union, the gender composition of presidents and members of the highest decision-making body tends to be predominantly male, with some notable exceptions (Figure 3.4).

Figure 3.4. Gender composition of research-funding organisations: Members of the highest decision-making body

Percentage of total

Source: EIGE, https://eige.europa.eu/gender-statistics/dgs/indicator/wmidm_educ__wmid_resfund, accessed on 18 August 2025.

Statlink

Industrial and regional innovation concentration: The competition policy dimension

This subsection discusses how the concentration of innovation activities – across firms, sectors and regions – shapes both the distribution of outcomes (quadrants A and B) and the opportunities for participation (quadrants C and D) within STI systems. A high and rising concentration in R&D investment and innovation capacity can lead to substantial gains in technological progress but also risks reinforcing disparities in who benefits (quadrants A and B) and who is able to participate meaningfully in the innovation process and decision making (quadrants C and D). The framework clarifies that such market dynamics are not neutral: they can reinforce exclusionary patterns unless actively addressed. Recognising these risks, recent OECD work highlights the critical, complementary role of competition policy in maintaining contestable markets, safeguarding access for new entrants, and ensuring that both the outcomes and opportunities created by innovation are widely shared, not just captured by a handful of leading actors or regions.

While concentration can drive technological advancement through economies of scale and scope in R&D – particularly critical for frontier technologies like AI, quantum computing and advanced semiconductors that require massive capital investments – it also creates significant barriers to participation for smaller firms, emerging regions and new entrants.

The OECD’s extensive work on competition policy in digital markets demonstrates that these concentration effects are not merely incidental but reflect underlying market power dynamics that can become self‑reinforcing, as leading platforms leverage their positions to acquire talent, emerging competitors and complementary technologies (OECD, 2025[47]; 2024[6]). Addressing these inclusion challenges effectively requires co-ordinated intervention from competition policymakers, who possess the analytical tools and enforcement mechanisms to assess market contestability, prevent anti-competitive consolidation and ensure that innovation ecosystems remain accessible to diverse participants. The OECD’s Competition and Innovation Framework emphasises that competition policy has a key role in facilitating innovation diffusion and allowing innovations to spread across markets, complementing STI policies’ efforts to broaden participation while maintaining the competitive dynamics essential for continued technological progress (OECD, 2023[7]).

International competition also plays a decisive role in shaping the opportunities – and limitations – in low‑carbon and green technologies. For example, while the People’s Republic of China’s strategic dominance in green technology manufacturing has profoundly altered the global competitive landscape, it has also led to a rapid deployment of what are critical technologies for economic decarbonisation (ITIF, 2020[48]). The consolidation of production in China degraded European manufacturing capacity but may also have curtailed potential technological pathways where European firms might have excelled, such as advanced thin-film technologies, perovskite-silicon tandems and other alternative photovoltaic approaches. The implication is, therefore, that the international dimension of technological competition can have a profound impact on the ability of local innovation and industrial ecosystems to participate in the opportunities that can emerge in transitions.

Co-ordinated policy action

This subsection draws together the full four-quadrant framework by emphasising that realising both broad participation in, and broad benefits from, innovation requires policy coherence beyond the STI domain itself. While STI policy can directly influence who participates in developing and using new technologies (quadrants C and D in Figure 3.1) as well as who reaps their benefits (quadrants A and B in Figure 3.1), persistent barriers to inclusion are often rooted outside STI’s traditional mandate.

The evidence demonstrates that STI policies alone cannot, therefore, address all dimensions relating to the participation in STI and the benefits from its outcomes. STI governance must, therefore, co-ordinate with social, education and economic policies to tackle structural barriers that limit participation and benefit; this, of course, is not new, but the speed and implications of frontier technological innovation raises the importance of action.

Conclusions

While technological innovation can bring substantial benefits for OECD Member countries and their citizens, the pursuit of technological leadership and the uneven diffusion of technological innovation can reinforce or deepen existing divides in who participates in and benefits from the STI system. The three ‑quadrant framework that was introduced – linking direct and indirect participation with direct and indirect outcomes – provides one approach for policymakers to assess and address the multifaceted impacts of technological progress. Four key takeaways emerge for STI policymakers in 2025.

First, STI policies operate across multiple dimensions of participation and inclusion simultaneously. The illustrative framework shows that policies designed to advance one quadrant – such as excellence-based initiatives targeting frontier development – can have unintended consequences across the others. For instance, while such policies may accelerate innovation, they can also concentrate benefits in already-advantaged regions and actors. Effective STI policy requires explicit consideration of cross-quadrant effects and complementary measures to address potential exclusionary impacts.

Second, addressing global challenges demands both excellence and inclusion. The scale and speed required for digital and green transitions cannot be achieved through concentration in innovation alone. Evidence from the gender participation case study illustrates that under-representation across the innovation pipeline – from education through leadership – represents a fundamental constraint on innovation capacity. Broader participation unlocks diverse perspectives, accelerates diffusion and generates the societal legitimacy essential for successful transitions.

Third, successful inclusion policies require co-ordination beyond STI. The persistent nature of participation gaps – despite decades of targeted interventions – underscores that STI policies alone are insufficient. Structural barriers require co-ordinated action across education, labour market, social and regional development policies. The framework’s emphasis on both direct and indirect pathways highlights the need for this multi-policy approach.

References

[8] Acemoglu, D. and S. Johnson (2024), “Learning from Ricardo and Thompson: Machinery and labor in the Early Industrial Revolution and in the age of artificial intelligence”, Annual Review of Economics, Vol. 16, pp. 597-621, https://doi.org/10.1146/annurev-economics-091823-025129.

[9] Acemoglu, D. and S. Johnson (2023), Power and Progress: Our Thousand-year Struggle Over Technology and Prosperity, PublicAffairs.

[13] Allen, R. (2007), “Pessimism preserved: Real wages in the British Industrial Revolution”, Oxford Economics Series Working Papers, No. 314, Oxford Univeristy, Department of Economics, https://ideas.repec.org/p/oxf/wpaper/314.html.

[27] APC (2025), “Advanced Propulsion Centre UK”, https://www.apcuk.co.uk/.

[17] Arnold, E. et al. (2023), “Navigating green and digital transitions: Five imperatives for effective STI policy”, OECD Science, Technology and Industry Policy Papers, No. 162, OECD Publishing, Paris, https://doi.org/10.1787/dffb0747-en.

[14] Autor, D. et al. (2017), “Concentrating on the fall of the labor share”, American Economic Review, Vol. 1087/5, pp. 180-185, https://doi.org/10.1257/aer.p20171102.

[20] Criscuolo, C., N. Gonne and G. Lalanne (2022), “An industrial policy framework for OECD countries: Old debates, new perspectives”, OECD Science, Technology and Industry Policy Papers, No. 127, OECD Publishing, Paris, https://doi.org/10.1787/0002217c-en.

[24] Curtis, E., L. O’Kane and R. Park (2023), Workers and the Green-Energy Transition: Evidence from 300 Million Job Transitions, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w31539.

[43] Davila, A. et al. (2024), Women Founders in European Deep Tech Startups, European Investment Bank et al., https://supernovas.eitcommunity.eu/wp-content/uploads/2024/11/Executive_Summary_Supernovas_Digital_FV.pdf?utm_source=chatgpt.com.

[31] EC-OECD (2025), “STIP Compass”, https://stip.oecd.org/stip/.

[46] EIGE (2024), Gender Equality in the European and National Parliaments Ahead of the 2024 Election Year, European Institute for Gender Equality, https://eige.europa.eu/sites/default/files/documents/gender-equality-in-the-european-parliament-and-in-national-parliaments-in-the-european-union-ahead-of-the-2024-election-year.pdf.

[41] European Centre for Women and Technology (2024), “Women in tech statistics: Gender diversity in tech [2024]”, news, https://ecwt.eu/news/gender-diversity-in-tech-2024.

[26] European Commission (2025), “About S3”, https://ec.europa.eu/regional_policy/policy/communities-and-networks/s3-community-of-practice/about_en.

[44] European Commission (2021), She Figures 2021: Gender in Research and Innovation – Statistics and Indicators, Publications Office of the European Union, https://op.europa.eu/en/web/eu-law-and-publications/publication-detail/-/publication/67d5a207-4da1-11ec-91ac-01aa75ed71a1.

[40] European Investment Fund (2023), The VC Factor: Gender Lens Edition, European Investment Fund, https://www.eif.org/news_centre/publications/vc-factor-gender-lens-edition.pdf.

[29] Federal Ministry for Economic Affairs and Energy (2025), “Clusters4Future”, https://www.clusterplattform.de/CLUSTER/Navigation/EN/Cluster-Policy/National-Level/Clusters4Future/Clusters4Future.html.

[11] Feinstein, C. (1998), “Pessimism perpetuated: Real wages and the standard of living in Britain during and after the Industrial Revolution”, The Journal of Economic History, Vol. 58/3, pp. 625-658, https://doi.org/10.1017/s0022050700021100.

[15] Guellec, D. and C. Paunov (2017), “Digital innovation and the distribution of income”, NBER Working Paper Series, No. 23987, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w23987.

[10] Hobsbawm, E. (1962), The Age of Revolution: 1789-1848, Vintage.

[36] Innovative Initiatives Digital Luxembourg (2025), “Women in tech”, web page, https://innovative-initiatives.public.lu/initiatives/women-tech.

[42] ISPE (2021), Women in Pharma: Empowering Women as Industry Leaders, International Society for Pharmaceutical Engineering, https://ispe.org/sites/default/files/attachments/public/PE_JanFeb21_CompleteIssue_v5_LR.pdf.

[48] ITIF (2020), The Impact of China’s Production Surge on Innovation in the Global Solar Photovoltaics Industry, Information Technology & Innovation Foundation, Washington, DC, https://itif.org/publications/2020/10/05/impact-chinas-production-surge-innovation-global-solar-photovoltaics.

[30] Japan Science and Technology Agency (2025), “Science and Technology Research Partnership for Sustainable Development (SATREPS)”, https://www.jst.go.jp/global/english/index.html.

[34] Korea Foundation for Women in Science, Engineering and Technology (2025), WISET, https://wiset.or.kr/eng/sub01_01_01.do.

[18] Larrue, P. (2021), “The design and implementation of mission-oriented innovation policies: A new systemic policy approach to address societal challenges”, OECD Science, Technology and Industry Policy Papers, No. 100, OECD Publishing, Paris, https://doi.org/10.1787/3f6c76a4-en.

[39] Lassébie, J. et al. (2019), “Levelling the playing field: Dissecting the gender gap in the funding of start-ups”, OECD Science, Technology and Industry Policy Papers, No. 73, OECD Publishing, Paris, https://doi.org/10.1787/7ddddd07-en.

[19] Mazzucato, M. (2018), “Mission-oriented innovation policies: Challenges and opportunities”, Industrial and Corporate Change, Vol. 27/5, pp. 803-815, https://doi.org/10.1093/icc/dty034.

[35] MINT Vernetzt (2025), MINT Vernetzt, https://www.mint-vernetzt.de/.

[28] NGen (2025), “Next Generation Manufacturing”, https://www.ngen.ca/.

[47] OECD (2025), Competition and Digital Economy, OECD Publishing, Paris, https://www.oecd.org/en/topics/sub-issues/competition-and-digital-economy.html.

[16] OECD (2025), “Embedding citizen science into research policy”, OECD Science, Technology and Industry Policy Papers, No. 175, OECD Publishing, Paris, https://doi.org/10.1787/a1cfb1a8-en.

[25] OECD (2025), The Adoption of Artificial Intelligence in Firms: New Evidence for Policymaking, OECD Publishing, Paris, https://doi.org/10.1787/f9ef33c3-en.

[6] OECD (2024), Competition Policy in Digital Markets: The Combined Effect of Ex Ante and Ex Post Instruments in G7 Jurisdictions, OECD, Paris, https://www.oecd.org/content/dam/oecd/en/publications/reports/2024/10/competition-policy-in-digital-markets_554eb7d5/80552a33-en.pdf.

[22] OECD (2024), “Fostering an inclusive digital transformation as AI spreads among firms”, OECD Policy Briefs, OECD Publishing, Paris, https://www.oecd.org/en/publications/fostering-an-inclusive-digital-transformation-as-ai-spreads-among-firms_5876200c-en.html.

[38] OECD (2024), “Framework for Anticipatory Governance of Emerging Technologies”, OECD Science, Technology and Industry Policy Papers, No. 165, OECD Publishing, Paris, https://doi.org/10.1787/0248ead5-en.

[4] OECD (2024), “OECD Agenda for Transformative Science, Technology and Innovation Policies”, OECD Science, Technology and Industry Policy Papers, No. 164, OECD Publishing, Paris, https://doi.org/10.1787/ba2aaf7b-en.

[23] OECD (2024), OECD Regions and Cities at a Glance 2024, OECD Publishing, Paris, https://doi.org/10.1787/f42db3bf-en.

[32] OECD (2024), Population with tertiary education [data set], https://www.oecd.org/en/data/indicators/population-with-tertiary-education.html.

[7] OECD (2023), Competition and Innovation: A Theoretical Perspective: OECD Competition Policy Roundtable Background Note, OECD, Paris, https://www.oecd.org/content/dam/oecd/en/publications/reports/2023/05/competition-and-innovation-a-theoretical-perspective_6e330b82/4632227c-en.pdf.

[45] OECD (2023), Government at a Glance 2023, OECD Publishing, Paris, https://doi.org/10.1787/3d5c5d31-en.

[2] OECD (2021), OECD Science, Technology and Innovation Outlook 2021: Times of Crisis and Opportunity, OECD Publishing, Paris, https://doi.org/10.1787/75f79015-en.

[1] OECD (2018), OECD Science, Technology and Innovation Outlook 2018: Adapting to Technological and Societal Disruption, OECD Publishing, Paris, https://doi.org/10.1787/sti_in_outlook-2018-en.

[21] Paunov, C. and J. Einhoff (2025), “Innovation policy transformed? Unveiling a new paradigm through natural language processing”, OECD Science, Technology and Industry Working Papers, No. 2025/21, OECD Publishing, Paris, https://doi.org/10.1787/5ee60cb5-en.

[5] Paunov, C. and H. McGuire (2022), “Towards a new vision of innovation through COVID-19? A comparative reading of 11 countries’ strategies”, OECD Science, Technology and Industry Policy Papers, No. 136, OECD Publishing, Paris, https://doi.org/10.1787/15475840-en.

[3] Paunov, C. and S. Planes-Satorra (2021), “What future for science, technology and innovation after COVID-19?”, OECD Science, Technology and Industry Policy Papers, No. 107, OECD Publishing, Paris, https://doi.org/10.1787/de9eb127-en.

[33] STEM Learning (2025), “About STEM ambassadors”, web page, https://www.stem.org.uk/stem-ambassadors.

[37] The Research Council of Norway (2019), Policy for Gender Balance and Gender Perspectives in Research and Innovation, The Research Council of Norway, Lysaker, Norway, https://www.forskningsradet.no/contentassets/19527ed7d0b149d6b9b310f8bb354ce9/nfr_gender_policy_orig-1.pdf.

[12] Voth, H. (2003), “Living standards during the Industrial Revolution: An economist’s guide”, American Economic Review, Vol. 93/2, pp. 221-226, https://doi.org/10.1257/000282803321947083.

Click to read more