Science has a critical role to play in supporting socio-economic transformation and sustainable prosperity. Achieving this at the necessary scale and speed will require changes to key aspects of the way that science systems and academic research currently operate. This chapter explores how policies relating to research careers, research infrastructures and science’s engagement with society can act as important levers to promote research that addresses the big questions around socio-economic transformation. Structural changes are required in each of these areas and the chapter discusses the critical role of research performance assessment in incentivising the necessary changes.
OECD Science, Technology and Innovation Outlook 2025
Driving Change in a Shifting Landscape
Abstract
Research careers in academia should be attractive and accessible to the best and brightest young scientists from all walks of life, who can bring a variety of different perspectives into research. Academia needs to embrace and value research on solutions to complex socio‑economic challenges that cut across disciplines and sectors. A variety of transparent career paths should be developed that recognise and enable inter-sectoral mobility and value the essential contribution of professional research support staff.
Research infrastructures (RIs) play an important role as catalysts that bring together key resources – hardware, software, data, methods and expertise – across different fields and countries to promote transformative change. This catalytic role can be amplified further when RIs operate together in ecosystems to address shared goals. Flexible support and governance mechanisms for such RI ecosystems need to be developed further, particularly at the international level.
Engagement with society, including responsible science communication and co-production of scientific knowledge are essential for delivering transformative change. Science communication, including its translation into policy, is not always valued within science systems. While there is an increasing number of funding initiatives to support transdisciplinary research and citizen science that underpin practical solutions, academia as a whole has yet to fully embrace these approaches.
Performance assessment and incentive structures do not fully recognise or value the variety of contributions to, and outputs from, science that are necessary to support socio-economic transformations. The concept of research excellence needs to be broadened and a range of (quantitative and qualitative) measures and indicators deployed to assess different aspects of scientific performance at the individual, institutional and national levels. This needs to be accompanied by adjustments to research funding review and award mechanisms.
Maintaining the integrity and credibility of science is a must. There is a danger in implementing the necessary changes that academic science becomes, or is perceived as being, overly controlled by governments. This could rapidly undermine both science and public trust in science. In an increasingly polarised geopolitical environment, it is critically important to protect the freedom and autonomy of research and promote open science while ensuring the integrity and security of the global science ecosystem.
International scientific co-operation is essential to achieve the urgent socio-economic transformations necessary to ensure a healthy and prosperous future for all. Countries share common challenges in transforming their production and consumption systems and the open exchange of scientific expertise, information and data is necessary to address sustainability challenges and crises that do not recognise national borders. The integrity and security of the global research ecosystem must be strengthened, with scientific knowledge for transformations being recognised as a global public good.
There is wide recognition of the need for transformative change in economies and societies to meet a range of challenges, including competitiveness, security and sustainability. As discussed in Chapter 1, science, technology and innovation (STI) systems are expected to contribute to transformative change but need to reform to generate and deploy relevant knowledge, technologies and innovation at an unprecedented pace and scale. This was clearly acknowledged in the Declaration endorsed by OECD Science, Technology and Innovation Ministers in April 2024 (OECD, 2024[1]), when the Agenda for Transformative Science, Technology and Innovation Policies was welcomed (OECD, 2024[2]). The Agenda provides a high-level framework that can be applied to any transformative goals, though three have been highlighted that capture many contemporary STI policy concerns: 1) promoting economic competitiveness that is fair and inclusive; 2) fostering resilience and security against risks and uncertainties posed by the growing emergence of systemic threats; and 3) advancing sustainability transitions that mitigate and adapt to a legacy of unsustainable development from climate change, pollution and biodiversity loss.
While much of the emphasis – and hope – for achieving these goals is on the future roles of technology and innovation, all three have deep implications for science and for science policy. In short, they cannot be achieved without the new knowledge that science must generate. Calls for transformative change provide an urgent stimulus for systemic and structural reforms to make science systems better fit-for-purpose to support societies in addressing ongoing and future challenges and crises. This can also be seen as an opportunity to address some of the persistent problems that have accumulated in science systems as they have tried to adapt to different socio-economic, technological and policy demands over recent years.
New scientific knowledge is essential for understanding and responding to “wicked” global challenges for informing policies and decision making at multiple scales (local to global) and for developing the new technologies that are essential for effective transformations. At the same time, science systems are themselves directly affected by the transformations they help to drive. This is perhaps most evident in the technological realm; for example, the fundamental building blocks of the digital revolution emerged from public investment in research and have transformed scientific practice. The broader social transformations that digital technologies have enabled also directly impact science: witness, for example, the targeting of scientists via social media. Science is not just a passive contributor but is a major driver of transformation and highly susceptible to the impacts of transformation. As such, it needs to play an active role in shaping transformations, optimising the benefits and limiting the potential negative consequences of its outputs.
Science systems1 have been designed and have evolved with a dual focus on scientific excellence and promoting economic growth with societal benefit being implicit in both. The academic community is the main guarantor of excellence, while economic benefit provides the principal rationale for public investment in research. Innovation policy has a strong focus on translating scientific outputs into commercially viable products, while science policy has mainly focused on supporting the academic community and promoting research excellence, with more or less attention to societal and policy demand depending on the research domain. In this overall context, the academic research community has tended to resist the strong top-down direction of research and research portfolios have been largely shaped by the choices of individual researchers and review by scientific peers. In some fields, such as medical research, these portfolios are well-aligned with social, political and economic priorities. Science systems, as a whole, are certainly responsive to the socio-economic environment in which they operate. Collective priorities emerge bottom-up and often merge into or converge with top-down policy priorities, but this is usually a slow process and major shifts in the overall direction of collective research efforts are rare (aside from during acute crises). There is an embedded inertia in science systems which strengthens their resilience but limits the conditions under which they can be widely mobilised around shared priorities.
Over time, the archetypal science system has proven its capacity to generate new knowledge and the spillovers to society in terms of technologies and innovations have been enormous. Effective STI systems that are able to generate and exploit new scientific knowledge are a key feature of all leading economies. Scientific research has also shed light on the negative effects of some aspects of technological development and human behaviours, including the environmental impacts, and how these can be avoided or mitigated. Nevertheless, in recent years, there has been increasing debate over the scale of public investment and whether the productivity of science (i.e. costs vs. economic returns) is declining (OECD, 2023[3]). Driven by a combination of precarious working conditions for many and short-term incentives for all, science appears to have become risk-averse. The main focus of research efforts is on incremental research in established fields rather than exploring new ideas in new fields or addressing “big questions” for science and societies. In response, many countries have been introducing dedicated initiatives to support high-risk/high-return research with the explicit aim of promoting technological breakthroughs and/or addressing complex inter/trans-disciplinary challenges (OECD, 2021[4]). However, these initiatives are limited in scope and hindered by structural barriers, including dominant incentive and reward systems, in academia (OECD, 2020[5]).
In addition to supporting economic growth and competitiveness, publicly funded academic research has a broader public good motivation that distinguishes it from most research that is conducted in the private sector. This can be illustrated, for example, by the essential role that academic research played in responding to the COVID-19 pandemic – from understanding the biology of the virus to vaccine development or detecting and monitoring infection to the application of public health and social measures. The pandemic was one of the rare instances when science systems mobilised rapidly across a wide range of disciplines and at high scale to respond to urgent societal needs (OECD, 2023[6]). The public good function of academic research is critically important with regards to societal transformations, which cannot be driven solely by economic competitiveness considerations, albeit these are the top policy priority in many countries. Nor can they be realised at the time and scale necessary by simply focusing on scientific excellence, incremental advances and knowledge diffusion. The need for new scientific knowledge to support transformative change is urgent and research systems as they currently operate are unlikely to provide this, or to ensure its effective transfer to society, at the scale and speed needed.
Openness, academic freedom and international collaboration have long been recognised as critical foundations for scientific progress for the benefit of society, i.e. for science to serve its public good function. In an increasingly fractured geopolitical world, the freedom and autonomy of academic research need to be protected and emphasised. Science can either be instrumentalised/weaponised as a tool to support authoritarian rule and accentuate inequalities between countries and populations or it can be a driving force for inclusive socio-economic transformation. The legitimate appropriation of scientific knowledge to promote national interests, including economic growth and security, needs to be aligned with the new imperative to address shared global challenges and support all countries in making the necessary transformations. There is a need for collective mobilisation based around shared values. Openness, international co-operation and benefit-sharing are essential for maintaining the trusted global research ecosystem that is necessary to address global challenges (see Chapters 1 and 2). Ensuring the integrity and security of this global ecosystem in a geopolitically divided world and protecting research from interference, coercion and misuse by malevolent state and non-state actors is a critical challenge for science policymakers (see Chapter 2).
There are three key interrelated areas that science policymakers need to pay attention to if science is to play its part in socio-economic transformations: 1) the scientific workforce; 2) research infrastructures; and 3) the interface between science and society. There is a fourth area or meta-leverage point that impacts on all of these: research assessment and incentives, including funding. A skilled research workforce and the tools and equipment that it uses, i.e. RIs, are the bedrock of any science system. A close and trusted relationship between science and society is critical for the new scientific knowledge and technologies that are necessary to achieve socio-economic transformation are to be taken up while activities that are proven to be damaging are phased out. Research assessment and incentives, including funding, are the most effective mechanisms for shaping scientific research and promoting change in scientific institutions’ practices.
Figure 4.1 illustrates the current status quo with regard to different categories of research and research outputs. The policy leverage points can be used to shift this status quo and grow the smaller circles. It is important to note that this is not a “zero-sum” scenario – all research types and the intersections between them are important for achieving the transformation goals. Rebalancing cannot be achieved in a top-down control manner and, in this regard, the role of science policymakers is to catalyse and enable rather than closely direct and control. Policy actions in all four areas need to be carefully designed and implemented, respecting the established values of science and the autonomy of scientific institutions.
Notes: The majority of publicly funded research is traditional discovery research, with a smaller fraction focused on understanding complex socio-economic and environmental challenges and a very small proportion focused on developing and implementing solutions for these “wicked” challenges (i.e. directly supporting transformations). The boxes indicate the main expected outputs that can be assessed for each of these three areas. In each area, the amount of really new, innovative high-risk/high-return research is limited.
To more effectively support socio-economic transformation, the system needs to be rebalanced so that all four types of research are more equally supported and incentivised and, most importantly, that the intersections between the different research types are expanded. Four key leverage points on which policy can act to help achieve this rebalancing are indicated in red.
N.B. The graphic is illustrative and the relative size of the bubbles is more meaningful than their absolute size.
The rest of this chapter focuses on the four key policy leverage points – the challenges and potential policy solutions. The main sections finish with overall policy recommendations, with illustrative examples of policy actions from different countries provided as endnotes. Many more examples can be found on the EC-OECD STIP Compass database (https://stip.oecd.org/stip).
Ultimately, science’s contribution to transformative change depends on the scientific workforce. It depends on the capacity of the research system to attract and maintain a diversity of talents – the brightest and the best from all walks of life – and provide these scientists and professional research support personnel with the necessary conditions to be creative and productive in addressing the “big questions” for science and society. This includes ensuring they have access to the necessary data and (digital) tools. Building and maintaining this capacity is essential for science to be able address many of the questions that need to be answered if inclusive and just transformations are to be effected. This means confronting a number of persistent human resource challenges, which have accumulated in research systems over several decades to the point that academic research is no longer considered an attractive or realistic career option for many aspiring young scientists who wish to contribute to science for society (OECD, 2023[7]).
Academic research and precarity
Academia fulfils an essential role not only in conducting a significant proportion of publicly funded research in most countries but also in training scientists and engineers that can work in the public and private sector, either within their country of training or abroad. While the majority of PhD trainees in science start out wanting to pursue an academic career, only a small minority will succeed. Some will leave research immediately after their PhD or move to research careers in the private sector. Many others continue to pursue their ambitions in academia until they either drop out frustrated after a succession of short-term positions or burn-out due to the hyper-competitive publish-or-perish culture that is prevalent in academia (OECD, 2021[8]). This is not good for them, for academic research or for science in society more broadly.
Academic research systems operate largely on the basis of multiple short-term (two to three years) research projects that are led by tenured scientists, with the hands-on research being conducted by PhD and postdoctoral researchers. Across OECD countries, most researchers under the age of 44 in academia are employed on short-term fixed-term employment contracts (Bello and Galindo-Rueda, 2020[9]). In many countries, a significant proportion of these researchers, including the majority of PhD researchers, are supported on stipends. Fixed-term employment is used as a flexible mechanism to meet short-term workforce demands but it also leads to precarity and diminishes the attractiveness of academic careers. Precarity, in turn, limits diversity both in terms of individuals and in research choices – when there is pressure to publish it is better to avoid high-risk research topics with uncertain outcomes (OECD, 2021[8]). Precarity is not simply about how funding and resources are used, although these are clearly important, it is also very much about academic career structures, workforce planning, personal and professional development processes, incentives, and rewards. Effectively addressing these issues requires a systemic approach that engages multiple actors – science policymakers, research funders, research providers/employers and early career researchers (OECD, 2021[8]; 2021[10]).
Precarity is context-specific but is an important feature of all national research systems. Precarity has increased as the emphasis on competition and (narrowly defined) scientific excellence has grown. While competition – between scientists, institutions and countries – is an important driving force for science, hyper-competition for talent and (limited) resources has serious negative effects. Shared global challenges call for collective mobilisation of the global research community and countries and research institutions should be sharing and adopting best practices to reduce academic precarity and retain diverse scientific talent, ensuring research careers remain attractive pathways for top young researchers.
Career options and competition for talent
Doctorate-level attainment in the population has grown rapidly across the OECD in the past decade, but entry rates in science, technology, engineering and mathematics (STEM) doctoral education in some countries have recently started to decline (Figure 4.2). In some countries, and in some fields, there is the perception that doctoral training is no longer attracting the best talent. In some strategically important fields such as informatics, artificial intelligence (AI) and quantum science, the decision not to pursue a PhD is simplified by attractive recruitment offers from the private sector for which a master’s degree or equivalent may suffice, at least for early career stages.
Note: Panels A-C focus on STEM disciplines (science, technology, engineering and mathematics). Panel D includes all research domains, including social sciences and humanities.
Source: Panels A-C: OECD (2025), Education Statistics database https://www.oecd.org/en/topics/education-access-participation-and-progression.html (accessed on 23 September 2025), Panel D: OECD (2025), REICO database https://www.oecd.org/en/networks/research-and-innovation-careers-observatory.html (accessed on 23 September 2025).
Despite indications of a recent slowdown in some countries, the number of doctorate holders worldwide has grown and there is intensified competition for limited tenured positions in academia and public research organisations (OECD, 2021[8]). According to recent global estimates, around 40% of postdocs leave academia (Duan et al., 2025[11]). While many highly skilled young scientists leave research altogether, others apply their research skills and expertise elsewhere, particularly in the business sector (Figure 4.3). From a societal perspective academic science and industrial science are complementary activities but, in practice, are distinct professions with very different career structures (Dasgupta and David, 1994[12]). Leaving academia to conduct research in the private sector is common, whereas movement in the opposite direction is relatively rare in most scientific domains (computer sciences being a notable exception in some countries). There is no revolving door – once one leaves academia it is highly unlikely they will be able to return, as their publication record and academic CV will inevitably depreciate over time. Lowering the structural barriers to entry into academia for highly skilled “outsiders” is a potentially important mechanism for expanding the diversity of perspectives and filling gaps in the public research workforce in rapidly expanding areas, such as AI and quantum science that are critical for socio-economic transformations.
As a percentage of national total researchers, based on full-time equivalents, 2023 or latest year available
Notes: Provisional values for Austria, Belgium, Czechia, Denmark, France, Greece, Iceland, Ireland, Italy, Lithuania, Luxembourg, the Netherlands, Norway, Portugal, Slovenia and Spain. 2023 corresponds to 2022 for Canada, Chile and Switzerland.
Source: OECD (2025), OECD Main Science and Technology Indicators Database, www.oecd.org/sti/msti.htm (accessed in March 2025).
Although the majority of PhDs and early-career researchers end up in careers outside of academia, very little attention is paid to preparing them for this or assisting them to make informed career choices. Scientific research develops individual expertise and transferable skills that can be applied in many socio-economic sectors, but these are often not appreciated at their full value. People and their qualifications are the most important tool for transferring research results and scientific thinking into the business and government sector and society more widely. For science to fully contribute to a country’s socio-economic development, it is essential to institutionalise career guidance for doctoral candidates and postdoc researchers (OECD, 2023[7]; Schillebeeckx, Maricque and Lewis, 2013[13]). In an ideal world, these young researchers will be highly motivated and productive throughout their time in academia and be able to apply the precious experience and skills they have acquired in other sectors, should they choose to do so at some stage in their careers.
The intense competition for scientific talent is not only between the public and private sector, it is also global (see Chapter 2). Many countries and institutions are struggling to attract and maintain the highly skilled personnel they require to conduct high-quality research and support transformative change. The share of international students in doctoral degree programmes has increased in most OECD countries. However, geopolitical tensions and concerns about research security and strategic autonomy are making some previously attractive research destinations less attractive and strengthened visa regulations are impairing international mobility (see Chapter 2). At a time when science is increasingly being called upon to address global challenges, policy measures to promote the free flow of scientists need to be maintained and strengthened.
Inclusive excellence
Scientific progress and support for transformative change requires the integration of diverse ideas, perspectives and skills within the research workforce. Diverse talents and perspectives can influence the choice of research topics and methods, opening up new research avenues and spurring discovery (Kozlowski et al., 2022[14]). Empirical work on team science has shown that diversity in perspectives leads to greater productivity when effective team management practices are in place (Apfelbaum, Phillips and Richeson, 2014[15]). However, science and academia are, to a considerable extent, self-replicating. First‑generation scientists who succeed in academia are rare and the rate of adoption of an academic career is low for scientists from historically under-represented groups (Hofstra et al., 2020[16]). Limited diversity in perspectives and a narrower range of research topics are not only distancing science from addressing the real needs of many sectors of society but also potentially slowing overall scientific productivity.
Studies of national research systems have provided broadly applicable evidence of disparities in opportunities across various stages of research careers, highlighting gaps in career advancement, pay, support and funding. These studies have largely focused on sex (Larivière et al., 2013[17]), race/ethnicity (Ginther et al., 2011[18]), and more recently also on the intersection of sex and race (Kozlowski et al., 2022[14]). They indicate that disparity and gaps, which vary by country and scientific field (Figure 4.4), are not due to differences in the quality of candidate individuals. Neither is quantity a strong factor, even though under-representation matters; in most countries, academia is poorly representative of the population at large. These discrepancies can be explained by a set of interconnected barriers for talented individuals from under-represented groups: relational barriers between individuals; structural barriers within research organisations; and systemic barriers rooted in cultural norms, regulations and legal frameworks (OECD, 2018[19]).
Source: OECD (2025), Education statistics database https://www.oecd.org/en/topics/education-access-participation-and-progression.html (accessed on 7 March 2025).
Promoting diverse perspectives and talents for “excellence in research” (Box 4.1) will require a concerted effort by policymakers, research organisations and funding agencies. In this, public policy has a dual role which combines steering, through the setting of standards and the provision of data and funding, with capacity building to ensure that talent is nurtured, retained and empowered across all career stages. Better data are needed to identify barriers to inclusion in research, taking into account that variables reflecting differences in background and perspectives (e.g. sex and race/ethnicity) are interconnected and mutually reinforcing, and to design and monitor policy interventions that recognise and exploit the synergy between inclusion and excellence in science.
Excellence in research refers to achieving the highest standards of science – rigour, ethics, significance and originality – with the aim of making an impactful contribution to knowledge and practice. Inclusive excellence in research is a concept that affirms the necessity of public policy intervention to actively promote a diversity of perspectives and the participation of individuals from different backgrounds to accelerate the advancement of science. It recognises that the notion of “excellence” is socially constructed, and its meaning is neither fixed nor universal. What counts as “significant”, “original” or “impactful” can diverge considerably across different contexts, communities and disciplines, reflecting the priorities of those in positions of authority who define and evaluate these standards. A narrow focus on excellence in isolation can privilege dominant perspectives while devaluing diverse ways of knowing and practicing science (Kraemer-Mbula et al., 2020[20]; López Piñeiro and Hicks, 2015[21]).
Inclusive excellence builds on the highest standards of science, and embeds equity – that is, acknowledging that researchers have different starting points and circumstances, and creating working conditions that make the probability of success independent of these (i.e. ensuring equal opportunities) – into the core practices of scientific research, education and training, as well as promotion and resource allocation. This has implications for academic career paths, evaluation and assessment, and the leadership of teams and research organisations. It provides a strong rationale for policies and measures that address regional and institutional imbalances in the allocation of research resources within a research system.
The concept of inclusive excellence can be applied across the research system in different ways. For example, in research-performing organisations, it can offer a framework to incentivise and sustain a shift in research culture. In funding organisations, it can help to ensure equitable access to funding opportunities, for example by introducing “junior” and “senior grants” to provide a level playing field when applying for research funding.
Artificial intelligence, open data and digital skills
AI, large language models (LLMs) and robotics are changing the way science is conducted and the roles researchers play (OECD, 2023[3]). These technologies are becoming ubiquitous across research, increasing the speed and efficiency of data and information processing (Box 4.2). They are even impacting on hypothesis generation, which was until recently considered to be the prerogative of the human scientist and there is increasing discussion of (semi-) autonomous “AI scientists” (Castelvecchi, 2024[22]). However, at least for the foreseeable future, these analytical tools cannot replace the human brain and the technical skills on which science depends. Scientists remain essential for driving discovery and progress through creativity, intuition and collaboration (Popper, 1961[23]). In the new world of AI and LLMs, a diversity of perspectives is likely to be an even more important determinant of scientific progress for the benefit of society (OECD, 2023[3]).
Artificial intelligence (AI) tools in science increase the capabilities of researchers in data analysis, simulation and hypothesis generation. For example, in genomics, the use of AI has helped to identify genetic variants associated with diseases, predict gene functions and understand complex genomic interactions (Zou et al., 2018[24]). In climate modelling it has increased the accuracy in forecasting short‑term weather patterns and long-term climate trends (Reichstein et al., 2019[25]), and it can speed up the development of new materials and products in a wide range of fields, including medicine (Max Planck Institute, 2025[26]).
AI has the potential to amplify scientific productivity by lowering costs and increasing research efficiency. For example, AI-powered robots can increase speed, precision and consistency in conducting experiments in laboratory environments. AI models that are trained on published scientific findings are able to anticipate human scientists in discovery. By avoiding the typical approach of local exploitation of the familiar instead of exploration of the unknown, these models can lead to insights and hypotheses unlikely to be proposed by human scientists (Sourati and Evans, 2023[27]).
In 2024, research on AI was awarded two Nobel prizes for innovations that will shape the future of medicine. The prize in physics was awarded to John J. Hopfield and Geoffrey E. Hinton (formerly of Google) for “foundational discoveries and inventions that enable machine learning with artificial neural networks.” The prize in chemistry was awarded to David Baker for “computational protein design” and to Demis Hassabis and John M. Jumper (of DeepMind) for “protein structure prediction”. It is notable that all these prize winners worked in academia and industry at different stages of their careers.
AI and large language models also raise a number of challenges for established scientific practices, particularly with regards to publications and the integrity of the scientific record (Kwon, 2025[28]; The Royal Society, 2024[29]). This is particularly relevant when one considers the policy emphasis on open access and the dominant role that publications play in research assessment and evaluation processes.
Source: Zou et al. (2018[24]); Reichstein et al. (2019[25]); Max Planck Institute (2025[26]); Sourati and Evans (2023[27]); Kwon (2025[28]); The Royal Society (2024[29]).
AI is an area of intense competition as well as productive collaboration between academia and industry and an area where the public good ethos of academia is sometimes pitched against the commercial motivation of industry. There are signs that the predominant role of industry in AI research could lead to a narrowing in the focus of research. Recent empirical work finds that “private sector AI researchers tend to specialise in data-hungry and computationally intensive deep learning methods” and that this is at the expense of “research involving other AI methods, research that considers the societal and ethical implications of AI, and applications in sectors like health” (Ahmed, Wahed and Thompson, 2023[30]; OECD, 2023[31]). Achieving a balance between the distribution of research and expertise across the public and private research sectors is important to ensure that AI is optimally developed and deployed to achieve transformative change.
Great hopes are being placed in open science, including FAIR (Findable, Accessible, Interoperable and Reusable) data and digital tools, as a catalyst for accelerated scientific progress and innovation. Access to data from different research fields, combined with administrative and other sources, is essential for understanding and managing complex crises and socio-economic transformations. AI and other software tools and computing resources are essential for analysing these data and translating them into relevant knowledge. However, the primary requirement for trustworthy data-intensive science is digitally skilled scientific personnel – data scientists, data stewards and software engineers (OECD, 2020[32]). Making research data FAIR and maintaining it over time is not a trivial task and is largely an unsupported mandate for the scientific community. If open science and data are to inform societal transformations, then the professional digital research support staff who can make this happen need to be properly valued and supported and making data FAIR needs to be incentivised.
Team science, risk taking (and incentives)
Science is increasingly conducted in teams that are often interdisciplinary and international. Interdisciplinary and transdisciplinary research, of the kind that is needed to support socio-economic transformations, is often high-risk – it can take longer to produce publishable results than conventional research and the outcomes are less predictable at the outset. While funding for interdisciplinary and multidisciplinary projects is increasingly available, reward systems, including recruitment and promotion, continue to primarily focus on individual performance (see below). For a young researcher with career ambitions in academia, incremental research in a well-recognised research area is a much safer choice than launching into a complex new interdisciplinary area.
Recognition for contributions to team science, from both researchers and professional support staff, is an important area for improvement. There is a need to establish clear guidelines to ensure fair acknowledgment of team members’ contributions. Initiatives such as the Contributor Roles Taxonomy (CRediT) offer a structured framework for documenting individual contributions to research projects, ensuring both transparency and fairness in recognition and enabling more nuanced acknowledgment in publications (Lin, 2024[33]). However, while scientific publication outputs are important, their uptake and short-term impact, as measured by citations, is often limited in new areas that do not already have well‑established communities of interest. Other more valuable outcomes for team science and interdisciplinary or transdisciplinary research, such as inputs to policy or societal decision making, are difficult to quantify and are rarely given their due credit in academic evaluations (see below).
There is a lack of attractive career paths, and appropriate evaluation frameworks, for researchers who want to work across disciplines and engage with citizens and policymakers (see above) to address complex societal challenges. Addressing this and incentivising and supporting researchers to take risks and work together to address big and complex challenges will be critical if science is to fully support socio-economic transformation.
Policy actions
1. Address the underlying structural issues that underpin research precarity and limit research choices and risk-taking for early career scientists.2
2. Promote “inclusive excellence” across science systems.3
3. Promote a variety of alternative career pathways within and beyond academia that can enable the generation and use of scientific knowledge to support transformations. Stimulate two-way mobility between academic science and other sectors, including industry.4
4. Continue to support and facilitate international mobility and the exchange of researchers and professional research support staff.5
In addition to talent, progress in science depends on access to state-of-the-art technologies and data (see the previous section). These are often provided as shared services, via RIs, which exist in all scientific domains and, in many cases, operate internationally (i.e. they provide access to both national and international research communities). RIs vary in nature, scale and structure and include large single-site experimental facilities, such as synchrotrons and telescopes (OECD, 2023[34]), and distributed smaller scale facilities, such as data networks or biobanks (OECD, 2017[35]). They are the backbone of national and international research systems and, as such, play a critical role in structuring these systems. They are also meeting places where different actors from different countries and sectors come together. RIs can play a leading role in promoting transformative change but this will require reforms in the way they are currently supported and operated.
Research infrastructure ecosystems
As demonstrated in the response to COVID-19, research infrastructures can play an important catalytic role in mobilising research to respond to crises (OECD, 2023[36]). They are uniquely positioned to bring together different actors, disciplines and countries to address complex scientific and societal challenges. Depending on their remit and scale, individual RIs can have a major influence on strategic directions and research choices in their own field and, by being open and providing services to new users, can support bottom-up research across multiple domains. Both direction and diversity in research are important for promoting transformative change and RIs can support both provided that they have the long-term strategic investment that enables them to extend their principle scientific missions to fully accommodate new users from science and beyond.
Besides their primary role as knowledge producers, most RIs also play a role in technology development and innovation. This can either be as a by-product of pursuing scientific goals – for which RIs often have to develop their own unique instruments and software – or a direct result of collaboration with industry and other partners to develop and test their products. In addition to experimental equipment and resources, RIs often have unique technical expertise, skills and know-how that can be applied beyond their immediate scientific goals. Many RIs are contributing to technological development and innovation in a range of areas, from energy to materials research, that are important for sustainable transformations. This often occurs in close collaboration with industry. Some have developed substantive initiatives that are specifically focused on leveraging in-house expertise to address green transitions and/or sustainable development. However, the funding for such activities remains limited and the incentives and mechanisms for engagement are relatively weak (OECD, 2025[37]).
RIs are increasingly co-operating with each other, exchanging know-how, and developing common protocols and practices to increase their efficiency and effectiveness. RI ecosystems, which support or co‑ordinate services and/or activities across different facilities, are also emerging. These are dynamic and evolving entities that develop around shared strategic objectives, which are often related to socio-economic transformations and/or crisis preparedness and response (Table 4.1). They operate at different geographic scales from local to international. By combining resources and expertise they strengthen research systems’ capacity to address shared challenges. Although there are mechanisms to support such ecosystems at a national or regional scale, there is an absence of effective tools or incentives to sustain effective global RI ecosystems, which are often dependent on short-term funding or subsidised by other member activities (OECD, 2025[37]).
Given that many RIs are inherently international and most provide open international access to data, they have an important role to play in promoting inclusion in research and access to scientific knowledge. International RIs, i.e. those with funding from multiple countries and international governance arrangements, can play a central role in socio-economic transformation in their host region or country. For example, the establishment of the Square Kilometre Array telescope in South Africa is premised on plans that this will stimulate the digital economy in South Africa and the region more broadly (Adams, Tiplady and Sgard, 2023[38]). However, the scientific requirements and political factors that inform the location of international RIs are complex and opportunities to exploit this transformative potential are rare. Distributed RIs, networks and the development of RI ecosystems provide more accessible opportunities for all countries to participate in research that can support socio-economic transformations. As yet, the global support and flexible governance mechanisms for such global infrastructure ecosystems are lacking but the need for a socio-economic transformation agenda can provide a stimulus for their development and vice versa.
Research infrastructures and trusted data
RIs are at the centre of the digital transition. They collect, produce and manage massive amounts of scientific data and information and have the professional expertise required to enable its robust analysis by a variety of users. These data have great societal value: they provide the basis for innovation and technological development and are also critical for informing sound evidence-based policies and decision making. Access to trustworthy data and information is crucial for industry, academia, policymakers and the public at large. Many RIs are at the forefront of AI and LLM development and deployment, mainly driven by their own research data management needs but with major spillovers for society (see the previous discussion on AI and Box 4.1).
Dedicated cyber-infrastructure, such as high-performance computers or GRID computing networks, support the entire public research enterprise (as well as much of private sector R&D). Research data are deposited in networks of data repositories that ensure their future stewardship and safe and secure access, in line with the FAIR principles (see Chapter 2). Many of these repositories also develop and provide access to a variety of data services and tools for different user communities (OECD, 2017[39]). In many cases they are leading efforts to develop the standards, protocols and tools that are necessary to ensure the interoperability of data from different fields, which is essential for generating the new knowledge needed to support societal transformations. For example, the social science and humanities RIs ODISSEI (Open Data Infrastructure for Social Science and Economic Innovations) and CLARIAH (Common Lab Research Infrastructure for the Arts and Humanities), together with SSHOC-NL (the Social Science and Humanities Open Cloud for the Netherlands, http://sshoc.nl), have developed a common initiative to create a secure digital infrastructure for linking and analysing diverse administrative and research data sets. This is enabling interdisciplinary research on “big” socio-economic issues, such as polarisation, social inequalities and the societal implications of environmental change.
At the international scale, the European Open Science Cloud – and similar data commons initiatives in other regions – aim to take advantage of the digitalisation of research by linking together trusted data sources and analytical tools and services, including AI algorithms and high-performance computing. These will be accessible, under safe and secure conditions, to both public and private sector researchers and promise to massively accelerate the interdisciplinary big data analysis necessary to inform socio‑economic transformations. Supporting and maintaining the cyber-infrastructure and developing the open technical standards and protocols that enable the interoperability of data from different sources are significant challenges but rapidly building the human capacity and skills for rigorous data-intensive research is perhaps the toughest challenge of all (OECD, 2020[32]).
Research infrastructures, skills and training
The construction and operation of RIs depends on highly skilled and specialised human resources – scientists, engineers and professional support staff. Hence, RIs have the capacity to contribute significantly to the development of skills and capabilities needed in many areas of society. This can include many specialised skill sets, such as digital expertise or technology-intensive engineering, as well as broader skills required for managing the complex operations and collaborations of RIs, such as transversal or systemic thinking and complex project management. RIs devote considerable time and effort to meeting their own training needs, which can be facilitated by working together in networks or ecosystems. They are also increasingly being used by a broader range of non-expert user communities, including citizen science practitioners and transdisciplinary research teams, who require dedicated training. However, while RIs provide both formal and informal training for scientists, technical support staff and users, they are rarely integrated into or well-connected to national education systems. For example, RIs handle massive amounts of complex data and increasingly use AI for data mining and analysis (see above); however, the training they provide in these areas is typically restricted to their own personnel and primary scientific user communities. There are opportunities for RIs to work more closely with higher education institutes and other training service providers to build digital, and other, capacities that address their own needs and those of society more broadly.
The challenges for developing and retaining RI personnel are similar to those in academia more broadly, including precarity, uncompetitive salaries and unclear career paths. In addition, career options within the public research system are often severely limited for those with very specialised skills and the draw of the private sector can be strong in some areas, such as in AI or quantum sciences. Specialised research support and technical personnel are critical for the effective operation of RIs They play an essential role maintaining and adapting the functioning of RIs. This ensures their daily operations but also enables these facilities to be rapidly mobilised during crises, such as the COVID-19 pandemic, and potentially also to play a major role in catalysing solutions for socio-economic transformation.
While networking and RI ecosystems can help promote mutual learning and staff exchange, RIs need to be supported to work with universities and relevant private sector actors to improve career paths and intersectoral mobility. This can also ensure that the invaluable tacit knowledge embedded in professional research support staff is shared and disseminated across different socio-economic sectors.
Policy actions
1. Recognise the essential role that RIs can play in supporting crisis response and socio-economic transformation and adopt strategic funding approaches that enhance their sustainability and enable flexibility.6
2. Support the co-ordinated and collaborative development, operation and use of RIs – including the promotion of global RI ecosystems – to tackle complex and interconnected global challenges (see Table 4.1 for examples).
3. Mandate and support RIs as sites for the generation and secure stewardship of high-quality FAIR data for responding to societal challenges.7
4. Leverage RIs for training and education to help address skills scarcities and mismatch associated with transformations.8
As witnessed during the COVID-19 pandemic, relevant scientific knowledge and social and technological innovations are essential for responding effectively to complex socio-economic challenges and crises. These, in turn, depend on having productive research environments and the right incentives. However, in a crisis situation, translating scientific knowledge into effective actions is absolutely dependent on the three-way relationship between science, policymakers and society (OECD, 2023[40]). Where trusted relationships exist, evidence-based policymaking can lead to effective crisis mitigation and adaptation strategies. Where there is a lack of trust between science and the public, crisis response is seriously compromised. Likewise, when relevant authorities are dismissive of science, their decisions and actions are likely to be ineffective. There are three critical areas for urgent attention if science is to effectively engage, and be trusted by, the public and policymakers: 1) integrity and science communication; 2) citizen engagement in research; and 3) the interface between science and policymaking. These are all closely linked to the implementation of open science in the broadest sense, i.e. embedding science in society (Dai, Shin and Smith, 2018[41]; UNESCO, 2021[42]; Wehn and Hepburn, 2022[43]).
Research integrity and science communication
Scientific integrity, good ethical practice and responsible scientific communication are essential if science is to be trusted. Following a number of high-profile scandals related to fraudulent and/or unethical research practices in the early 2000s, strong measures have been taken to protect the integrity of scientific research – including the development of guidelines, training, and review and reporting procedures. Scientific publishers have been key players in this and the formal publication of scientific results in professional journals is subject to strict review procedures and “self-policing” by peers. Nevertheless, following a number of highly publicised breaches of research integrity and misleading publications during COVID-19, scientific authorities in some countries have recently declared that science needs to refocus on its core values.
Ensuring the rigour of science and the scientific record is a long-term challenge that also relates to incentive and reward structures (see below). Whatever mitigation measures are implemented, they will not always be 100% effective and honest human (and machine) errors will occur. Despite the attention devoted to a small number of fraudulent and erroneous publications, the self-correcting mechanisms of science continue to operate effectively (albeit, in some cases, slowly). Processes for the correction and/or retraction of articles, such as such as Retraction Watch, are increasingly transparent and routine and overall the integrity of the formal scientific record is high. At the same time, as discussed in Box 4.2, the rapid development of AI and LLMs is introducing new challenges and opportunities for producing and detecting fraudulent research publications.
The public communication of science has, until recently, received less attention than the formal publication process for scientific articles. Public communication has tended to focus on exciting scientific discoveries and delivering facts in a one-way process rather than engaging with the public and addressing their interests and concerns. The limitations of this approach were clearly illustrated during the COVID pandemic, when a lack of transparency about the gaps and uncertainties in scientific data and information contributed to distrust in science-based interventions for many population groups (OECD, 2023[40]; 2023[44]). Effective two-way dialogue between citizens and scientists will be important if science is to effectively inform socio-economic transformation.
Notes: This figure is a simplified representation of the complex relationships between critical issues of societal concern and the challenges for science communication. These are not necessarily wholly new challenges, but they can manifest differently and are often magnified by social media.
Source: Based on OECD (2023[44]).
The science communication landscape is evolving, paving the way for a set of new actors (multiple publics, social media influencers, digital platforms, algorithms, etc.) who can create and share scientific content. There is a shift from traditional communication intermediaries (scientific journals, mainstream media) to online and social media, largely driven by the digital transformation. While this change provides a welcome opportunity to move beyond one-way communication, it can also enable misinformation and disinformation. In this new context, communicating science as it relates to critical issues of societal concern (public health, climate change, emerging tech) faces a number of significant challenges (Figure 4.5).
Poor or irresponsible science communication can seed fake news, which can be rapidly propagated via social media and facilitated by the use of AI. Fake news stories often promote alternative/non-credible scientific perspectives and are a major threat to science and to democratic processes. Ineffective science communication can undermine the credibility of scientific experts, scientific institutions and policymakers. During crises, confusing, contradictory and untargeted scientific messaging can lead to poor compliance with policy advice, putting individuals and communities at risk (OECD, 2023[40]). Responding effectively to misinformation and disinformation requires a multi-faceted approach in which the science community needs to play a leading role. This is partly about responsible science communication (Box 4.3) but also about promoting scientific and digital literacy so that people are able to distinguish between rigorous scientific information and opinion.
1. Transparency
Providing access to data on which scientific conclusions depend.
Clearly describing the methods and data used to reach a conclusion.
Communicating uncertainties.
2. Inclusivity
Reaching out to diverse groups in society.
Making science communications accessible (e.g. taking into account digital infrastructure and language barriers).
3. Integrity
Adhering to ethical and professional standards in scientific research and communication.
Being intellectually honest (e.g. not hyping scientific results) and ensuring the rigour of the research that is being communicated.
4. Accountability
Making clear who is responsible for a scientific communication and in what capacity they are communicating (e.g. personal or institutional; subject expert or scientific commentator).
Being clear on the sources that underpin a scientific communication.
Openly declaring any potential conflicts of interest or commitment for those communicating or providing the content of scientific communications.
5. Freedom and autonomy
Communicating scientific research without being constrained by external interference (e.g. political, legal, religious).
Respecting the self-governance of academic research and the right of scientists to freely communicate (in accordance with Principles 1-4).
6. Timeliness. This is particularly relevant in emergency situations and includes:
Delivering relevant and up-to-date information to citizens in a timely manner, with appropriate caveats where necessary.
Not withholding or delaying the communication of relevant scientific information while ensuring that essential quality controls have been performed prior to its release.
Source: OECD (2023[44]).
Citizen engagement
While responsible and effective science communication is important, public engagement with science needs to go beyond this. Co-design and co-production of research, including the engagement of local and indigenous communities, is essential to achieve the Agenda for Transformative Science, Technology and Innovation Policies. Citizens are important contributors to environmental and health monitoring, e.g. for biodiversity assessments and pandemic alert mechanisms, and also have critical knowledge and perspectives that need to be integrated into transformative research agendas and practice. Citizen science – the active engagement of citizens in the production of scientific knowledge – needs to be widely embraced and incentivised (OECD, 2025[45]). Beyond this, there is a need to support transdisciplinary research that combines and integrates knowledge from different disciplines as well as different public and private sector stakeholders (OECD, 2020[5]). These, and other modes of research that bring together different stakeholders to co-produce scientific knowledge and/or innovations, are increasingly being deployed across many research fields, although they still represent a very small fraction of total research activity. Increasing this requires broad acceptance of the value of citizen engagement in research within both the academic and policy communities and tailored support and review mechanisms (Box 4.4). It also requires new incentive and reward structures to promote a shift in academic culture (see below).
Why and when to promote citizen science?
1. Policy and decision makers across government should recognise the value of citizen science for science and society and embed citizen science into their considerations when formulating policies. From the research policy perspective, there are three main rationales for promoting citizen science:
increasing the scope of data collection and/or analysis and accelerating scientific discovery
addressing societal needs and challenges more effectively
promoting the democratisation, legitimacy and uptake of policies informed by scientific knowledge.
How to support citizen science?
2. Senior-level commitment (in ministries, research agencies and institutions) is critical to drive wider acceptance of the value of actively engaging citizens in research.
3. Top-down and bottom-up approaches need to be combined and supported effectively to promote citizen science, as it requires engagement from multiple actors and good interactions between them.
4. A variety of citizen science community groups and entities, including public and private intermediary agents, networks and associations, play important roles and should be supported accordingly.
Overcoming obstacles and challenges
5. The rigour and quality of citizen science, as well as the management of potential bias, must be ensured for citizen science to be widely accepted.
6. Throughout policy planning and implementation, recognition that citizens are a very heterogeneous group with different motivations, interests and barriers for getting involved with research is important.
7. Citizen science collaborations across countries can make a significant contribution to tackling global challenges.
Systemic change and assessing impact
8. The science community and its institutions need to embrace citizen science as a valuable mode of research that can complement and improve traditional research activities. This requires a shift in academic research culture that can be supported by providing appropriate career pathways and reward systems for citizen science practitioners.
9. Monitoring, evaluation and assessment of the impacts of citizen science should reflect the different rationales for its deployment and, in many cases, the most important impacts will not be fully reflected in traditional bibliometric performance measures.
Source: OECD (2025[45]).
Open science – including public access to scientific information and data – can be an important driver to promote responsible science communication and citizen engagement in research. Indeed, citizen science is increasingly viewed as the third pillar of open science (UNESCO, 2022[46]).9 While opening up scientific data and information and the processes of science to the public at large could potentially encourage misuse, the potential benefits far outweigh the risks provided that appropriate safeguards can be put in place to limit misuse. This is the case, for example, with sensitive personal data for which safe and secure access protocols are being developed so that only legitimate parties can access and analyse the data. Integration of data from multiple sources, including administrative and research data and data collected by citizens, will be important to inform and monitor the impact of socio-economic transformations.
Science, decision making and policymaking
As discussed above, the main focus of STI policy in most countries has been on translating scientific knowledge into commercial products and growth. Until recently, much less attention has been devoted to translating scientific knowledge into effective policy and decision making. Yet in the context of transformations, this is a critical area for attention. Complex transformations are characterised by considerable scientific uncertainty and relevant scientific advice needs to be timely, yet rigorous and transparent (OECD, 2015[47]). Policies and decision making need to be informed by best available scientific evidence that draws on all relevant disciplines and explicitly acknowledges gaps in existing knowledge (Box 4.5). At the same time, the independence and autonomy of science must be protected. The frameworks, processes and incentives to enable well-functioning science advisory systems and the uptake of scientific knowledge in policy making are lacking in many countries and internationally (OECD, 2018[48]). The COVID-19 pandemic was revelatory in this regard and the lessons learnt from it are directly applicable to future crises and ongoing policy development to address complex socio-economic issues (OECD, 2023[36]; 2023[40]; 2023[6]).
The following principles are important in building an effective and trustworthy science advisory process:
1. Have a clear remit, with defined roles and responsibilities for its various actors. This includes:
a clear definition and demarcation of advisory versus decision-making functions and roles
defined roles and responsibilities, and the necessary expertise for communication
an ex ante definition of the legal role and potential liability for all individuals and institutions involved
the necessary institutional, logistical and personnel support relative to its remit.
2. Involve the relevant actors, including scientists, policymakers and other stakeholders, as necessary. This involves:
using a transparent process for participation and following strict procedures for declaring, verifying and addressing conflicts of interest
engaging all the necessary scientific expertise across disciplines to address the issue at hand
giving explicit consideration to whether and how to engage non-scientific experts and civil society stakeholders in framing and/or developing advice
implementing effective procedures for the timely exchange of information and co-ordination with different national and international counterparts.
3. Produce advice that is sound, unbiased and legitimate. Such advice should:
be based on the best available scientific evidence
explicitly assess and communicate scientific uncertainties
be preserved from political (and other vested-interest group) interference
be generated and used in a transparent and accountable manner.
Source: OECD (2020[49]).
One of the principal lessons from the COVID-19 crisis is that no matter how good science advisory processes are, the effective uptake of rigorous scientific evidence by governments depends on the political willingness to consider this evidence. Policymaking is rarely determined by science alone and there are multiple sectoral perspectives that need to be considered and weighed according to different value judgements by policymakers (OECD, 2023[40]). However, consideration of best available scientific evidence (and the associated gaps and uncertainties) is critical for policies to effectively support inclusive socio‑economic transformations. The absence of political support or acceptance can be a major challenge for science. In this regard, the three-way relationship between science, society and policymaking is critically important. In well-functioning democracies, governments are accountable to citizens, and public trust in science can reinforce both evidence-based policymaking and the implementation of these policies (OECD, 2024[50]). In the ideal situation, a virtuous triangle of trust can be established between governments, science and the public.
A cultural change within science is necessary to make it more open, inclusive and responsive to societal needs. In the face of urgent global challenges, scientific knowledge needs to inform policy development and decision making at different scales, from the international to the local level.
Policy actions
1. Strengthen scientific integrity and good research conduct by shifting the focus from the quantity of research outputs to the quality, transparency and rigour (see next section).
2. Prioritise and reward responsible science communication and societal engagement, including transdisciplinary and citizen science activities (see Boxes 4.3 and 4.4).10
3. Promote open science and public access to scientific data and information while ensuring the safety and security of sensitive information.11
4. Develop effective science advisory systems that can integrate insights from different disciplines and respond in a timely manner to policymakers’ and citizens’ needs (see Box 4.5).12
5. Promote scientific and digital literacy across society and reward scientists for contributing to related education and training activities.13
For science systems to significantly change the way that they operate, one or both of the following conditions are required: 1) there is a major crisis, such as a pandemic or a war, that science clearly has a role in addressing;14 and 2) the incentive and reward structures for scientists and scientific institutions are shifted. Considering the latter, as described earlier, the incentive and reward structures in public research are heavily focused on a narrow definition of scientific excellence. Despite growing criticism and calls for change (see Box 4.6 for examples), bibliometric measures, such as citations, continue to be heavily used – often in isolation – to evaluate and reward scientific performance at the institutional and individual level and to benchmark national performance. This is what drives “the publish or perish culture”, with its adverse effects on early career scientists, diversity in science and research choices. It discourages researchers from taking the risk of working across disciplines or sectors and addressing the “big issues” that underpin inclusive socio-economic transformations.
While formal research publications continue to be an important output from research, there are other activities and outputs that society expects and needs and which must be valued and incentivised if science is to support sustainable transformations. These include: public engagement, policy support, provision of trusted FAIR data and green innovations (Figure 4.6). Unlike publication outputs, none of these, with the possible exception of technological innovation, are easy to measure or assess in an objective or quantitative manner. While qualitative assessments and peer review can give some indication of performance in these areas, such approaches are resource-intensive, are not always feasible and have their own limitations (Wilsdon, 2015[51]). Nevertheless, it is important that science policymakers and academic leaders give clear signals in terms of incentives and funding that business as usual is not sufficient for science to support socio-economic transformation. Talented young scientists and scientific institutions must be encouraged and supported in pursuing inclusive excellence in research (see Box 4.1) and to support transformative change.
There is growing recognition of the need for change in research assessment processes (Box 4.6). While this has mainly been driven by concerns about the perverse effects that current processes have on individual and institutional behaviour, it also recognises the need for a broader framing of scientific excellence, or inclusive excellence (see Box 4.1), that values different scientific contributions, activities and outputs. A number of countries and institutions are responding to these calls for change and implementing reforms to their assessment systems. Many of these put a greater emphasis on qualitative assessment and peer review, although these approaches have their own drawbacks in terms of resource requirements and potential biases. AI and LLMs are also opening up new avenues, e.g. analysis of publications to identify high-risk innovative research outputs (Machado, 2021[52]). It will be important to harmonise these various initiatives across different scales – individual, institutional, national – and countries so that they do not inadvertently introduce barriers to mobility. Research assessment reforms are a critical tool for incentivising and monitoring different aspects of scientific performance in relation to socio-economic transformations.
DORA (the San Francisco Declaration on Research Assessment) (2012): calls on research actors to avoid using journal-based metrics as a surrogate measure for the quality of scientists or their work, and considers a broad range of impact measures. Provides positive recommendations.
Science in Transition (2013): a movement started in the Netherlands to develop principles for assessing scientists for hiring/promotion/tenure.
Leiden Manifesto (2015): sets out principles and best practices for the use of quantitative indicators in research assessment.
INORMS (International Network of Research Management Societies) (2018): brings together research managers to share good practices. Key outputs of the INORMS Research Evaluation Group – including the SCOPE Framework for Research Evaluation and More than Our Rank initiative – are aimed at all stakeholders in higher education and research.
FOLEC-CLACSO (Latin American Forum on Research Assessment) (2019): a forum for exchange on research evaluation practices in Latin America, initiated by the Latin American Council for Social Sciences (CLACSO). The forum provides regional-specific guidelines for research assessment. CLACSO-FOLEC’s Research Assessment Academy trains reviewers and assessment specialists to support fairer and more situated evaluation processes.
Room for Everyone’s Talent (2019): an initiative by Dutch public knowledge institutions and funders of research. Advocating for change in the recognition and reward system, and more inclusive research culture.
Honk Kong Principles for Assessing Researchers (2019): principles to help research institutions to minimise perverse incentives, recognise and reward trustworthy research, support the inclusion of behaviours that strengthen research integrity in frameworks for career appraisal and advancement.
Science Europe (2020): produced a position statement and Recommendations on Research Assessment Processes.
TARA (Tools to Advance Research Assessment) (2021): growing out of the DoRA community, this “bottom-up” initiative aims at developing practical tools to promote responsible research assessment in research-performing organisations.
CoARA (Coalition for Advancing Research Assessment) (2022): an international initiative aimed at reforming research evaluation systems to emphasise quality, inclusivity and a diversity of contributions – moving beyond traditional metrics such as publication counts and journal impact factors.
AGORRA (A Global Observatory for Responsible Research Assessment) (2023): part of the Research on Research Institute, this observatory generates evidence and analysis to support and accelerate responsible research assessment across 14 countries.
Barcelona Declaration (2024): advocating for open research information.
Source: Adapted and updated from Curry et al. (2020[53]).
In addition to changes in performance assessment systems, many countries are implementing new research-funding initiatives to support the type of research and innovation required to address the “big questions” and support sustainable and inclusive socio-economic transformation. A number of new funding schemes support mission- and challenge-based research, high‑risk/high return research (OECD, 2021[4]), citizen science (OECD, 2025[45]), transdisciplinary research (Kaiser and Gluckman, 2025[54]; OECD, 2020[5]), and other modes of participatory research. At the institutional level, interdisciplinary centres of excellence, some with a specific remit to inform policy, are relatively commonplace and some universities are restructuring their research around transversal missions and/or local societal needs rather than traditional academic disciplines (OECD, 2020[5]). New disciplines, such as sustainability research, that directly address key aspects of the Agenda for Transformative Science, Technology and Innovation Policies are emerging. However, all these initiatives and developments represent a very small proportion of the total public research effort. The large majority of academic research continues to be conducted in traditional disciplines and public and policy engagement activities are sideline activities rather than mainstream outputs. The challenge for science policymakers is to provide the right incentives – including funding – to mobilise a significant fraction of academic science to address the big questions and support society in making urgent socio-economic transformations.
Policy actions
1. Review research assessment processes at all levels to promote inclusive excellence and take full account of the variety of scientific activities and outputs required to support transformative change (see Box 4.6).
2. Recognise the value of citizen engagement, policy advice and FAIR data provision (even in the absence of rigorous quantitative measures) in individual recruitment and career progression.15
3. Recognise the value of teamwork, interdisciplinary and-transdisciplinary research, and citizen science approaches, which are all required to address sustainability transformations, and embed this into research assessments 16
4. Provide the necessary funding support, via suitably adapted mechanisms, to scale-up the type of research and activities, including public and policy engagement, that can support transformations.17
Governments and research funders have an important role to play in investing in research that supports socio-economic transformation and supporting the communication and engagement activities necessary to translate research outcomes into effective action. However, investment alone is not sufficient, and the existing structures, operating frameworks and incentives that shape research choices, careers and practices need to be adjusted.
Many positive developments are already underway, but these tend to be small scale and peripheral to the mainstream of scientific activity. Traditional disciplinary research conducted by clever individuals must continue to be supported and serendipitous discoveries will surely contribute to positive transformations, but this alone will be insufficient. Significant structural change and new incentives are required to ensure that a diversity of bright minds are empowered and supported with the necessary infrastructure to work together and engage other stakeholders in producing and applying the scientific knowledge required to promote transformative change.
In the face of urgent global challenges, science should be a source of hope and optimism rather than scepticism and mistrust. Science has been a critical factor underpinning many countries’ socio-economic development and prosperity. In fields like medical research or agriculture, it has made a huge contribution to improving human well-being. The understanding of the universe that comes from basic research in physics and astronomy has had an incalculable impact on human culture, beliefs and values. However, an important lesson from history is that scientific knowledge and the technologies that arise from it can be instrumentalised for both good and bad purposes.
Another lesson from history is that science advances best and benefits the most people when an appropriate balance between national competition and international co-operation is achieved. No one country has all the expertise necessary and infrastructure to address the complex global challenges we are already faced with, let alone the new crises that will surely emerge over the coming decades. As the COVID-19 pandemic illustrated, we live in a globally connected world, where crises can rapidly propagate with no respect for national borders. Achieving socio-economic transformations and sustainable prosperity for all requires both the adaptation of national science systems and a strengthening of the global research ecosystem.
The effectiveness of science in promoting transformative change is dependent on science being trustworthy and being trusted by policymakers and the public at large. Building and maintaining this trust, in an increasingly polarised geopolitical environment means upholding academic freedom and research integrity. It also means supporting open science and international co-operation while protecting scientists and scientific institutions from interference by state and non-state actors.
References
[38] Adams, D., A. Tiplady and F. Sgard (2023), “Integration of socio-economic impact into the development of the Square Kilometre Array (SKA) in South Africa”, OECD Science, Technology and Industry Working Papers, No. 2023/07, OECD Publishing, Paris, https://doi.org/10.1787/04438223-en.
[30] Ahmed, N., M. Wahed and N. Thompson (2023), “The growing influence of industry in AI research”, Science, Vol. 379/6635, pp. 884-886, https://doi.org/10.1126/science.ade2420.
[15] Apfelbaum, E., K. Phillips and J. Richeson (2014), “Rethinking the baseline in diversity research”, Perspectives on Psychological Science, Vol. 9/3, pp. 235-244, https://doi.org/10.1177/1745691614527466.
[9] Bello, M. and F. Galindo-Rueda (2020), “The 2018 OECD International Survey of Scientific Authors”, OECD Science, Technology and Industry Working Papers, No. 2020/04, OECD Publishing, Paris, https://doi.org/10.1787/18d3bf19-en.
[22] Castelvecchi, D. (2024), “Researchers built an “AI scientist” – what can it do?”, Nature, Vol. 633/8029, pp. 266-266, https://doi.org/10.1038/d41586-024-02842-3.
[53] Curry, S. et al. (2020), “The changing role of funders in responsible research assessment: progress, obstacles and the way ahead”, RoRI Working Paper, No. 3, Research on Research Institute, https://doi.org/10.6084/m9.figshare.13227914.v2.
[41] Dai, Q., E. Shin and C. Smith (2018), “Open and inclusive collaboration in science: A framework”, OECD Science, Technology and Industry Working Papers, No. 2018/07, OECD Publishing, Paris, https://doi.org/10.1787/2dbff737-en.
[12] Dasgupta, P. and P. David (1994), “Toward a new economics of science”, Research Policy, Vol. 23/5, pp. 487-521, https://doi.org/10.1016/0048-7333(94)01002-1.
[11] Duan, Y. et al. (2025), “Postdoc publications and citations link to academic retention and faculty success”, Proceedings of the National Academy of Sciences, Vol. 122/4, https://doi.org/10.1073/pnas.2402053122.
[18] Ginther, D. et al. (2011), “Race, ethnicity, and NIH research awards”, Science, Vol. 333/6045, pp. 1015-1019, https://doi.org/10.1126/science.1196783.
[16] Hofstra, B. et al. (2020), “The diversity-innovation paradox in science”, Proceedings of the National Academy of Sciences, Vol. 117/17, pp. 9284-9291, https://doi.org/10.1073/pnas.1915378117.
[54] Kaiser, M. and P. Gluckman (2025), “Are scientific assessments and academic culture impeding transformative science?”, Sustainability Science, Vol. 20/3, pp. 1109-1116, https://doi.org/10.1007/s11625-025-01631-9.
[14] Kozlowski, D. et al. (2022), “Intersectional inequalities in science”, Proceedings of the National Academy of Sciences, Vol. 119/2, https://doi.org/10.1073/pnas.2113067119.
[20] Kraemer-Mbula, E. et al. (eds.) (2020), Transforming Research Excellence: New Ideas from the Global South, African Minds, Cape Town, https://www.scienceopen.com/document_file/b6cefefe-f3ca-4522-b9a8-d45c26c01869/ScienceOpen/Transforming%20Research%20Excellence_FINAL-WEB-02012020.pdf.
[28] Kwon, D. (2025), “Science sleuths flag hundreds of papers that use AI without disclosing it”, Nature, Vol. 641/8062, pp. 290-291, https://doi.org/10.1038/d41586-025-01180-2.
[17] Larivière, V. et al. (2013), “Bibliometrics: Global gender disparities in science”, Nature, Vol. 504/7479, pp. 211-213, https://doi.org/10.1038/504211a.
[33] Lin, Z. (2024), “Modernizing authorship criteria and transparency practices to facilitate open and equitable team science”, Accountability in Research, pp. 1-24, https://doi.org/10.1080/08989621.2024.2405041.
[21] López Piñeiro, C. and D. Hicks (2015), “Reception of Spanish sociology by domestic and foreign audiences differs and has consequences for evaluation”, Research Evaluation, Vol. 24/1, pp. 78-89, https://doi.org/10.1093/reseval/rvu030.
[52] Machado, D. (2021), “Quantitative indicators for high-risk/high-reward research”, OECD Science, Technology and Industry Working Papers, No. 2021/07, OECD Publishing, Paris, https://doi.org/10.1787/675cbef6-en.
[26] Max Planck Institute (2025), “Artificial Intelligence for Materials Science”, web page, https://www.mpie.de/4867299/artificial_intelligence (accessed on 21 February 2025).
[45] 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.
[37] OECD (2025), “Fostering research infrastructure ecosystems for addressing complex scientific and societal challenges”, OECD Science, Technology and Industry Policy Papers, No. 183, OECD Publishing, Paris, https://doi.org/10.1787/34797721-en.
[1] OECD (2024), Declaration on Transformative Science, Technology and Innovation Policies for a Sustainable and Inclusive Future, OECD, Paris, https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0501.
[2] 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.
[50] OECD (2024), OECD Survey on Drivers of Trust in Public Institutions – 2024 Results: Building Trust in a Complex Policy Environment, OECD Publishing, Paris, https://doi.org/10.1787/9a20554b-en.
[3] OECD (2023), Artificial Intelligence in Science: Challenges, Opportunities and the Future of Research, OECD Publishing, Paris, https://doi.org/10.1787/a8d820bd-en.
[31] OECD (2023), Artificial Intelligence in Science: Challenges, Opportunities and the Future of Research, OECD Publishing, Paris, https://doi.org/10.1787/a8d820bd-en.
[44] OECD (2023), “Communicating science responsibly”, OECD Policy Briefs, No. 2, OECD Publishing, Paris, https://doi.org/10.1787/5c3be7ce-en.
[36] OECD (2023), “COVID-19 and policy for science”, OECD Science, Technology and Industry Policy Papers, No. 152, OECD Publishing, Paris, https://doi.org/10.1787/8f86e60b-en.
[40] OECD (2023), “COVID-19 and science for policy and society”, OECD Science, Technology and Industry Policy Papers, No. 154, OECD Publishing, Paris, https://doi.org/10.1787/0afa04e2-en.
[6] 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.
[7] OECD (2023), “Promoting diverse career pathways for doctoral and postdoctoral researchers”, OECD Science, Technology and Industry Policy Papers, No. 158, OECD Publishing, Paris, https://doi.org/10.1787/dc21227a-en.
[34] OECD (2023), “Very large research infrastructures: Policy issues and options”, OECD Science, Technology and Industry Policy Papers, No. 153, OECD Publishing, Paris, https://doi.org/10.1787/2b93187f-en.
[10] OECD (2021), “Challenges and new demands on the academic research workforce”, in OECD Science, Technology and Innovation Outlook 2021: Times of Crisis and Opportunity, OECD Publishing, Paris, https://doi.org/10.1787/75f79015-en.
[4] 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.
[8] OECD (2021), “Reducing the precarity of academic research careers”, OECD Science, Technology and Industry Policy Papers, No. 113, OECD Publishing, Paris, https://doi.org/10.1787/0f8bd468-en.
[5] OECD (2020), “Addressing societal challenges using transdisciplinary research”, OECD Science, Technology and Industry Policy Papers, No. 88, OECD Publishing, Paris, https://doi.org/10.1787/0ca0ca45-en.
[32] OECD (2020), “Building digital workforce capacity and skills for data-intensive science”, OECD Science, Technology and Industry Policy Papers, No. 90, OECD Publishing, Paris, https://doi.org/10.1787/e08aa3bb-en.
[49] OECD (2020), “Providing science advice to policy makers during COVID-19”, OECD Policy Responses to Coronavirus (COVID-19), OECD Publishing, Paris, https://doi.org/10.1787/4eec08c5-en.
[19] OECD (2018), “Gender in a changing context for STI”, in 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.
[48] OECD (2018), Scientific Advice During Crises: Facilitating Transnational Co-operation and Exchange of Information, OECD Publishing, Paris, https://doi.org/10.1787/9789264304413-en.
[39] OECD (2017), “Co-ordination and support of international research data networks”, OECD Science, Technology and Industry Policy Papers, No. 51, OECD Publishing, Paris, https://doi.org/10.1787/e92fa89e-en.
[35] OECD (2017), “Digital platforms for facilitating access to research infrastructures”, OECD Science, Technology and Industry Policy Papers, No. 49, OECD Publishing, Paris, https://doi.org/10.1787/8288d208-en.
[47] OECD (2015), “Scientific advice for policy making: The role and responsibility of expert bodies and individual scientists”, OECD Science, Technology and Industry Policy Papers, No. 21, OECD Publishing, Paris, https://doi.org/10.1787/5js33l1jcpwb-en.
[23] Popper, K. (1961), The Logic of Scientific Discovery, Science Editions, New York, NY.
[25] Reichstein, M. et al. (2019), “Deep learning and process understanding for data-driven Earth system science”, Nature, Vol. 566/7743, pp. 195-204, https://doi.org/10.1038/s41586-019-0912-1.
[13] Schillebeeckx, M., B. Maricque and C. Lewis (2013), “The missing piece to changing the university culture”, Nature Biotechnology, Vol. 31/10, pp. 938-941, https://doi.org/10.1038/nbt.2706.
[27] Sourati, J. and J. Evans (2023), “Accelerating science with human-aware artificial intelligence”, Nature Human Behaviour, Vol. 7/10, pp. 1682-1696, https://doi.org/10.1038/s41562-023-01648-z.
[29] The Royal Society (2024), “Science in the age of AI: How artificial intelligence is changing the nature and method of scientific research”, news, https://royalsociety.org/news-resources/projects/science-in-the-age-of-ai.
[46] UNESCO (2022), An Introduction to the UNESCO Recommendation on Open Science, United Nations Educational, Scientific and CulturalOrganization, Paris, https://doi.org/10.54677/xoir1696.
[42] UNESCO (2021), UNESCO Recommendation on Open Science, United Nations Educational, Scientific and Cultural Organization, Paris, https://doi.org/10.54677/mnmh8546.
[43] Wehn, U. and L. Hepburn (2022), Guidance for the Implementation of the UNESCO Open Science Recommendation re. “Opening Science to Society” (FINAL), Zenodo, https://doi.org/10.5281/zenodo.7472827.
[51] Wilsdon, J. (2015), The Metric Tide: Independent Review of the Role of Metrics in Research Assessment, SAGE Publications Ltd, London, https://doi.org/10.4135/9781473978782.
[24] Zou, J. et al. (2018), “A primer on deep learning in genomics”, Nature Genetics, Vol. 51/1, pp. 12-18, https://doi.org/10.1038/s41588-018-0295-5.
← 1. This chapter focuses mainly on public science or academic research systems. In most OECD countries, the majority of academic research is conducted in universities and supported by dedicated science funding agencies. In some countries, public research institutes, such as the CNRS in France, are the main public research provider. The distinction is made between science systems and broader concepts of innovation or STI systems, which also include government research conducted in research and technology organisations and dedicated innovation agencies, whose primary mission is to support industry. There is considerable variation across countries in the weight of different components that make up science and STI systems. Science policy and innovation policy can likewise be more or less integrated depending on the scope of different ministries and agencies.
← 2. See OECD (2021[8]) for a detailed analysis on policy options and illustrative examples of what countries are doing to reduce precarity. These range from voluntary concordats and charters for research providers to the promotion of tenure track positions or legislative action on contractual status. In 2015, as part of its University Capacity Building Strategy, South Africa introduced a New Generation of Academics Programme with successful applicants being appointed into permanent posts firmly factored into long-term university staffing plans from the outset. France and Germany have introduced legislative measures to eliminate stipends and ensure that PhD and postdoctoral researchers are employed on standard employment contracts. They have also introduced junior chair and tenure track programmes to reduce uncertainty around future careers.
← 3. Many countries are implementing a diversity of policies to: build a sound evidence base on diversity in research personnel; support accessibility to research careers for under-represented groups; reduce uncertainty in research careers; and ensure equitable access to funding opportunities. Germany has integrated gender equality as a core component of its Excellence Initiative (2007-2017), in the Excellence Strategy (since 2017) and related research funding strategies, aiming to foster a more equitable and inclusive research landscape. Canada introduced its Tri-Agency EDI Action Plan in 2018 as the foundation for a range of measures to support the equitable participation of students and researchers in the research system. UK Research and Innovation introduced the Equality, diversity and inclusion strategy in 2023, which encourages the nine UK research councils to use equality impact assessments, an evidence-based approach designed to help organisations ensure that their policies, practices, events, trainings and decision-making processes are fair and do not create disadvantages for any protected groups.
← 4. See OECD (2023[7]) for a detailed analysis of policy options and examples of initiatives being implemented in different countries to diversify research career pathways. These range from capacity‑building training and exchange partnerships between public research providers and other public and private sector actors with research capabilities to improving career guidance and mentorship for early career researchers. Recognition of new research roles, including “third space professionals”, who operate at the interface between research and professional services, is also increasing. Japan has recently conducted a Survey of Japan Master’s Human Resource Profiling, exploring the reasons why master’s students choose non-academic careers instead of pursuing doctoral programmes.
France has created a postdoctoral contract in both public and private law to facilitate the professional transition of doctoral graduates to permanent positions in public or private research. Norway provides mobility grants (salary funds) for doctoral and postdoctoral fellows for internships in the public and voluntary sectors for up to six months.
← 5. International mobility is a high priority for many OECD countries that wish to strengthen their research workforce. Most countries have dedicated funding schemes to attract foreign researchers, e.g. Brain Pool Korea is a specific scheme to attract outstanding overseas scientists, including Korean scientists living abroad, to work in all sectors of the economy. Japan is providing major funding to small number of selected universities to become global research leaders, with an emphasis on international recruitment at all levels, from PhD students through to principal investigators, including professional support staff. More broadly, the European Commission is supporting a number of regulatory and administrative initiatives to harmonise conditions, including pension rights, for early career researchers across Europe to facilitate mobility and exchange.
← 6. An illustrative example of this is the importance that has been attached to strengthening the RI ecosystem, to ensure health emergency readiness, and to transform Canada’s biomanufacturing and life sciences sector after the COVID-19 pandemic. As part of a comprehensive inter-ministerial strategy (Canada’s Biomanufacturing and Life Sciences Strategy), the Canada Foundation for Innovation, in co‑ordination with domestic funding counterparts, launched two sequential competitions under the Biosciences Research Infrastructure Fund. The first stand-alone competition focused on biocontainment and large-animal facilities to conduct infectious disease research safely. The second competition adopted a strategic ecosystem approach through an integrated biomedical research and RI competition, focused on accelerating the translation of promising research into commercially viable products and processes. The competitions were designed to strengthen research, RI and talent capacity in Canada for innovation-led growth; identify, leverage and boost existing networks of players and RI capabilities across sectors; and create meaningful, sustainable synergies within the pan-Canadian research ecosystem. Central to achieving this was a robust governance structure for design, decision making and oversight, which provided strategic guidance, flexibility and coherence throughout the different stages of programme development and merit review processes. This ensured that the final recommendations for the RI competitions were targeted, anchored in real needs and driven towards the core objectives of the strategy, ensuring alignment with other federal investments to achieve maximum impact.
← 7. An illustrative example of this is the Open Data Infrastructure for Social Science and Economic Innovations (ODISSEI), a collaborative consortium in the Netherlands that aims to improve user access to social sciences and humanities (SSHs) RIs. It provides analytical tools, algorithms, a secure computing environment, and specialised services and expertise. ODISSEI regroups all SSH RIs in the Netherlands in a unique consortium. SSHs are traditionally a very scattered group of disciplines with very different data access policies and data standards. Not only has ODISSEI been able to regroup all research institutions on SSHs in the country but it also includes and provides secured access to SSH data from Statistics Netherlands. Secured access to the large diversity of SSH data has allowed the conduct of complex interdisciplinary studies of societal interest.
← 8. The CERN Accelerator School (CAS) has been running for over 40 years. Initially focused on training directly related to particle accelerators, courses have progressively diversified to a wide range of domains, including data sciences and engineering, as innovation and technologies developed for scientific equipment proved to be invaluable to a vast array of potential users. CERN works closely with industry and accelerator technologies are transforming society in areas such as security scanning, cancer treatment, and food and materials sterilisation. It has been at the origin of many spin-offs and has recently developed a “CERN Entrepreneurship Student Programme” (CESP) that brings together graduate students from around the globe for practical and theoretical trainings. Under the supervision and coaching of CERN experts and knowledge transfer professionals, students explore, evaluate and exploit CERN technologies with the aim of developing concepts for new ventures. CESP builds upon basic knowledge about entrepreneurship and venture creation.
← 9. Open science is a policy priority for all OECD countries, and several are developing monitoring systems to track the openness of scientific information, data, and software. Despite these efforts, monitoring and evaluation of open science impact remains insufficient, with most current assessment models focusing on output-based measures such as publication counts, rather than real-world societal benefits, inclusivity, and sustainability. The EOSC Observatory data highlights this gap, as only 15 countries have national monitoring for open-access publications, while none have comprehensive monitoring systems for research data. In the meantime, the French government has made strides in developing a national indicator, called the “French Open Science Monitor”, which is internationally recognised (see https://frenchopensciencemonitor.esr.gouv.fr/).
← 10. A number of countries have introduced strategies for citizen engagement and communication linked with dedicated funding and reward schemes (OECD, 2025[45]; 2023[40]; 2020[5]). Several countries or funding agencies, such as The Research Council of Norway, have embedded citizen science and engagement in national open science strategies. Others, such as the German Federal Ministry of Research, Technology and Space; UK Research and Innovation; and the National Research Foundation in South Africa, have developed dedicated strategies for public engagement or participation in science. With respect to funding, an interesting example is the European Commission’s Framework Programme, which has evolved from having a dedicated funding stream for science and society (2002) to science in society (2007) to science with and for society (2014). Since 2021, Horizon Europe has included a focus on widening participation and strengthening the European research area (2021), with societal engagement being mainstreamed across the programme, including in the dedicated funding for missions-driven research. This approach has been echoed in a number of countries which have a variety of dedicated and more generic funding schemes that support citizen engagement. In terms of recognition and rewards, there are also a variety of national initiatives, for example: the Flemmish Academy of Belgium has introduced annual Science Communication Awards; Germany has a Knowledge of the Many – Research Prize for Citizen Science sponsored by the Federal Ministry for Research, Technology and Space and the Natural History Museum; and France has a Prize for Participatory Research, sponsored by the Ministry of Research and Environmental and Agricultural Research’s funding agency.
← 11. Many countries are implementing a broad range of policies and actions to promote FAIR and open data (STI policies for Open Science portal). In 2021, the OECD Council adopted a revised Recommendation of the Council concerning Access to Research Data from Public Funding and a policy toolkit is being developed to support countries in the implementation of this Recommendation.
← 12. Science advisory structures and processes differ from one jurisdiction to another (OECD, 2015[47]) and these proved to be variously effective during the COVID-19 pandemic, when a particular challenge was integrating data and knowledge across different sources (OECD, 2023[40]). The pandemic also highlighted the challenges in co-ordinating and sharing advice across federal and national boundaries (OECD, 2018[48]). Since the pandemic, a number of countries have revised their science advisory mechanisms to try to address these challenges.
← 13. While promoting digital and scientific literacy falls largely under the remit of formal education, there a number of accompanying measures or actions that scientific agencies and institutions can also take. These range from support for science centres and museums to science festivals or open laboratories. The United Kingdom’s Festival of Social Science takes place annually and includes a series of exhibitions, lectures and public debates as well as performances, guided walks and workshops. A similar Science Festival is organised annually in different cities in Poland.
← 14. While climate change and biodiversity loss are widely regarded as major global crises, they are not at the top of the list of national priorities for many countries. This reflects the fact they are also not the immediate priority or concern for the majority of citizens in most countries. In the absence of strong societal and political pressure, the public scientific enterprise as a whole has not yet fully mobilised to support sustainability transitions and inclusive socio-economic renewal. The bottom-up mobilisation of the scientific community that characterised the COVID-19 pandemic is not happening spontaneously with regard to sustainability transitions, at least not at the necessary speed and scale.
← 15. One mechanism to ensure this is to adopt standardised academic CV templates that include only a small number of key publication outputs and support narrative accounts for other activities and outputs, e.g. in relation to science and society or science and policy. Going further, the Norwegian Career Assessment Matrix (NOR-CAM) is a framework for assessing and evaluating research(ers). Central to this framework is the move away from merely favouring quantitative measurements, for instance solely focusing on the number of publications or the ranking of the journal where the research is published, to a more comprehensive and flexible framework where multiple areas of expertise can be assessed more systematically than is currently the case.
← 16. The Dutch Recognition and Rewards programme, launched in 2020, brings together an alliance of research institutes and universities with support from the Ministry of Education, Culture and Science to promote cultural change in academia via changes in research assessment and incentives. It has five areas of focus: 1) diversifying and revitalising career paths; 2) achieving balance between individuals and the collective; 3) stimulating open science; 4) focusing on quality; and 5) stimulating academic leadership.
← 17. On behalf of the German federal government, the state of Brandenburg and several members of the Alliance of Science Organizations in Germany, the Federal Ministry of Education and Research (since May 2025 renamed to the Federal Ministry of Research, Technology and Space) established a new research institute in 2009 with the aim of acting as an intermediary between science, politics and society. In January 2023, the institute joined the Helmholtz Association of German Research Centres and was institutionalised as the Research Institute for Sustainability. The institute’s goal is to act as an international platform for science and a link between research, politics and business in society and to contribute to the formation of public opinion. With its transdisciplinary research approach, the Research Institute for Sustainability brings together science, politics, economy and society in discourses on societal challenges. The Federal Ministry of Education and Research has also been funding “social-ecological junior research groups” since 2002. These groups explicitly implement interdisciplinary and transdisciplinary research approaches to tackle complex societal challenges.