Strengthening National Evidence‑Informed Policymaking Ecosystems

4. Policy implications and the way forward

Abstract

This chapter outlines forward-looking policy pathways to supporting an agenda for EIPM in the future at European Level. Building on the findings from the project, these include focusing on the professionalisation of knowledge broker functions, maximizing the potential of AI, better assessing the impact of EIPM and ensuring commitment to use of reliable information with integrity. It also emphasises the importance of building new communities of practice. Finally, the chapter highlights future contributions from the JRC and OECD in these areas moving forward.

The need for building new communities of practice

The EIPM project has shown that all participating countries – regardless of the development status of their EIPM ecosystem – could make substantial progress to their EIPM practices. The systemic nature of EIPM requires bringing all the relevant parties to the table, either at national or international level.

At the inception of the project, each country set up co-ordinating committees involving representatives from various beneficiary organisations, who would meet regularly to share ideas and provide feedback to the JRC and OECD. Building this common space for dialogue in a systematic way across the seven countries allowed the project teams to identify many common challenges and key EIPM ecosystem functions. These insights were also shared across countries, as reflected in this final report.

There is a clear value in continuing to invest in such communities in the future, both at national level and at the European level and beyond. These can help support EIPM ecosystems at an EU level, helping foster knowledge exchange and benefits from common resources.

Inspired by the connective strength of this project, the JRC will be working together with SG REFORM and the European Commission DG for Research and Innovation to build a European EIPM Community of Practice, which will serve as the central hub for Europe’s knowledge brokerage. The Community of Practice should serve as a platform to jointly work on the future science-for-policy agenda. It will connect knowledge brokers, policymakers, researchers and other stakeholders dedicated to improving evidence informed policymaking.

Pathways to supporting an agenda for EIPM in the future at European level

The field of EIPM is constantly evolving. While it is important to follow existing good practices in the strategic planning and development of EIPM functions, it is equally crucial to look towards the future.

A future oriented, European-level EIPM agenda should serve several goals. First, it should help to ensure that EIPM ecosystems can address Europe’s most pressing challenges. This includes a mix of economic challenges, geopolitical threats, and challenges in terms of an erosion of trust in national governments and public institutions. While trust in science remains relatively high, the way information is shared, seen and consumed is being transformed through a mix of technological and social pressures, which are likely to be compounded by artificial intelligence.

Second, the agenda should help to ensure that EIPM ecosystems are fully responsive to the changing needs of governments. At a time of significant fiscal pressures, with the need to address the green and digital transition and necessary defence investments, there is a need to close evidence gaps. This includes, for example, sharing evidence on carbon mitigation approaches and increasing the efficiency and effectiveness of public expenditure through evidence driven spending reviews.

Third, the agenda should help to further strengthen the capacity and resilience of public administrations in view of future shocks. The ecosystem will be more resilient across changing government landscapes if it is equipped with solid and recognised capacity and if it is understood as useful. Reliable supply of evidence, and capacity to address issues in an agile and citizen-oriented manner, will increase government capacities into the future.

The agenda should also help ensure the sustained engagement with academia and researchers, as well as civil society in general. As discussed at several points in this report, differences in timelines, perspectives and incentives of researchers and policymakers can make collaboration between the two difficult, which can cause the development of policy-relevant evidence to suffer. Furthermore, the involvement of civil society into decision-making processes can help reinforce the democratic legitimacy of the policy process and increase transparency and trust with the public.

The strength of EIPM ecosystems also depends on their ability to keep up to date with state-of-the art knowledge and information technology, and to be able to prepare and anticipate challenges in the context of the digital transition, including the role of artificial intelligence. The professionalisation and institutional support of the role of knowledge brokers in maintaining healthy relationships with scientific, technological and innovation institutions and groups will be crucial.

Professionalisation of the role of the knowledge brokers

One of the key findings of the EIPM project is that the emerging field of knowledge brokerage, with individuals and organisations dedicated to fulfilling this function as their main objective, plays an essential role in strengthening the interaction between evidence and policymaking. However, most knowledge brokers have developed their competences in isolation, often following a path as researchers or experts in a specific scientific or policy field, and then learning how to effectively perform knowledge brokerage through experience. While institutions may have greater opportunities for mutual learning, the project findings show that in many countries applied research institutes often work in silos to address the needs of specific ministries. While translating evidence can be learned through experience, it poses additional challenges, such as how to work with interdisciplinary evidence or address ethical conflicts. Hence, EIPM ecosystems would benefit from increasing the professionalisation of the knowledge broker function, both at the individual and the organisational level, as well as through cross-disciplinary sharing across knowledge broker organisations.

Balancing and combining different types of evidence

As mentioned in 1.1 in Chapter 1, a broad approach to the concept of evidence was adopted throughout the EIPM project. It can refer to data, information, and knowledge from multiple sources, including both quantitative (statistics, measurements) and qualitative data (opinions, stakeholder input, conclusions of evaluation, expert advice). What is needed or preferred depends on the policy phase or on which policy instrument is most relevant. What counts as evidence also depends on the disciplinary perspective that is needed to solve policy issues. Cross-sectoral problems often need a multi- or interdisciplinary perspective. Combining disciplinary perspectives and methods is not easy, however, and most countries expressed an ambition to strengthen their capacity for this.

There was recognition across participating countries that EIPM requires a variety of forms of evidence beyond scientific evidence. For example, when designing social or educational policies, it is often neither feasible nor ethical to test them in a controlled environment. In such cases, it is important to draw on the expertise of practitioners to ensure that policies are sound and tailored to the local context, increasing their potential for success. This is understood in some contexts as implementation science, and there is often a parallel need to mobilise administrative data to a greater extent.

This broader perspective on what types of evidence are needed does not imply that all types of evidence should have the same value. One of the tasks of knowledge brokers is to synthesise different types of evidence with respect to their level of (un)certainty. Knowledge brokers should promote the use of evidence that is rigorous, systematic and technically valid. However, a core of scientific evidence may be combined with expertise from practitioners and stakeholders, which includes a need to ensure feedback from implementation, as well as space for experimentation to allow for citizen-centred and user driven approaches, particularly in policies aimed at designing public services. There is already substantial guidance available on how to combine and weigh different types of scientific evidence. This should be expanded to incorporate other sources of evidence relevant for EIPM.

Maximising the potential of AI while mitigating its risks

While the role of AI for EIPM was generally not addressed as part of the country analysis as such, given the paucity of relevant examples and information, it was fully discussed as part of a fruitful mutual learning exercise. Generative AI offers unprecedented opportunities for EIPM. The recent widespread development of large-scale language models allows for more effective literature searches and for automating a large part of the process of evidence review and synthesis. Furthermore, it can allow for the more widespread analysis and potential uptake of evidence by reducing the skills barriers needed for in-depth analysis of complex datasets and specialist information.

However, there are several potential risks. Perhaps most significantly, generative AI can not only synthesise and analyse texts, but it can also easily create false yet convincing information. As such, excess dependence on AI tools without capacity for human led discernment carries some risks. Furthermore, the process through which generative AI produces analysis or synthesises texts is often obscure if not completely opaque, meaning the opportunities to scrutinise methodologies and promote transparency available for human-analysed work do not exist. As such, it is of paramount importance to understand when AI is appropriate, how to use it, what its risks are, and how to discern whether the information it provides is legitimate. AI also presents government with implementation challenges, in terms of improving their data governance and organising data and promoting adequate AI numeracy skills.

Assessing the impact of EIPM

The importance of evidence-informed public policies has been increasingly recognised in Europe in recent years. At the same time, a growing need to assess and evaluate impacts of EIPM has emerged. Assessing the impact of EIPM interventions is necessary to enable participating countries to evaluate their progress, compare results across EIPM interventions, and provide accountability and transparency. At the European level this has been addressed by the development of strategies to evaluate EIPM ecosystems.

Evaluating the impact of interventions and initiatives promoting EIPM is challenging, and there are both technical and political factors that need to be considered. Technical factors include the fact that evidence is only one of many factors influencing policymaking, the time lag between EIPM interventions and their policy impact, and the fact that assessing the use of EIPM interventions often relies on self-reporting. On the political side, the power dynamic between stakeholders, including government entities, complicate the policymaking process and make it hard to reconstruct and measure science’s influence. There is a need for impact assessments that take both types of factors into account. However, measuring the actual impact of EIPM on the context and the intrinsic quality of policies will remain challenging.

Ensuring commitment to use of reliable information with integrity

In the current climate of declining trust and misinformation, it is extremely important and urgent to promote and commit to integrity principles in the production and use of evidence in support of policymaking. Both at the EU and in European countries members of the EU, there is a range of instruments that address integrity principles in the production of evidence, including scientific integrity frameworks, regulations, legal declarations of independence, and codes of conduct.1 However, these existing instruments focus on the production of scientific knowledge in general, and not on the production for and use of scientific knowledge in policymaking. Existing documents also do not necessarily apply to the work of knowledge brokers and knowledge brokerage institutions.

As this project has underlined, EIPM is not an individual skill – on the contrary, it requires a whole ecosystem to function. EIPM requires good working relationships between knowledge producers, brokers and users. That means that simply mirroring regulations that safeguard scientific research from attempts by the private sector to influence methods and outcomes will not be sufficient. EIPM requires guidelines on how to combine the task of building good working relations, while creating an environment that ensures integrity both in the production and the use of scientific evidence. Too often the default for researchers is to insist that their integrity and independence can only be guaranteed by an arm’s length relationship with government and funding independent of wider societal goals.

However, this may not always be sufficient, given the risk of capture by large private interests, and also given the new risks associated with mis and disinformation. In terms of integrity, while there is a vital place for formalised advice, informal advice from researchers and experts embedded inside government is essential to effective EIPM. There is therefore a need to develop frameworks that enable researchers to work in proximity and trust, while protecting the integrity and independence of their advice. Enabling EIPM approaches also require to be aware of the new risks of mis and disinformation, which requires securing an environment that safeguards information integrity. This imply for EIPM to take place within an information environment conducive to the availability of accurate and plural information sources, enabling individuals to make informed choices. This can be done through developing strategic documents, making efforts to expand information and research, co-operating with partner countries and the private sector, and capacity building (OECD, 2024[1])

Future contributions to this agenda

The JRC and the OECD remain committed to contribute to the future agenda of EIPM, drawing on their respective strength and expertise.

The JRC Agenda

To this end, in the coming years, the JRC will work on several concrete outputs:

Knowledge Broker Manifesto

The JRC will publish a knowledge broker manifesto aimed at establishing the profession of knowledge brokers. The manifesto will serve as a guidebook for practicing knowledge brokerage ethically and effectively, compiling use cases and best practices from practitioners and researchers across various fields of scientific advice. It will also incorporate insights from behavioural sciences on how to deliver evidence that can change minds.

European Framework for EIPM

To help strengthen citizens' trust in the ability of governments to deliver effective policies based on sound evidence, the JRC will explore ways to foster commitment to integrity in both the production and use of EIPM. Existing instruments for scientific integrity, along with national legal provisions that guarantee the independence of publicly funded research organisations and advisory bodies, provide valuable inspiration. However, the goal is to design an instrument that expresses a positive commitment to EIPM integrity from researchers and from (political) policymakers as well. Such an instrument should not add to the administrative burdens on researchers and policymakers.

Update of the competence framework for policymakers and researchers

The JRC competence frameworks for policymakers and researchers have served as an important tool in this project, both for the capacity building workshops and as a framework to analyse needs and gaps in existing training and educational programmes in the seven participating countries. In turn the project has identified important new developments, such as the use of AI for EIPM, that challenge the national science-for-policy ecosystems and the individuals that are part of it. Therefore, the JRC will work on an update of the competence framework, that will also more explicitly acknowledge the competences of knowledge brokers. The update will not only cover the framework itself, but also related products such as the brochure ‘10 tips for researchers: how to achieve input on policy’.

Collaborative policymaking

The JRC will bring science advice to collaborative policymaking. Specifically, it will explore how cognitive, behavioural, and organisational sciences can help us rethink the way public administrations and governing institutions work together and inject collective intelligence in the policymaking process. It will focus on how to strengthen co-ordination across governmental departments and agencies to make them more efficient in mobilising their wealth of knowledge and resources. The underlying ambition is to both systematise empirical knowledge and develop evidence-informed interventions to transform and innovate the way public services work.

AI to support science-for-policy 7

The JRC is currently exploring possibilities to bring scientific knowledge closer to policy development, implementation, monitoring, and evaluation with the help of AI. The objective of this work is (1) to investigate if and how specific activities and tasks at the interface between Science and Policy could benefit from AI capabilities and Human-AI collaboration, (2) experiment with promising use cases, and (3) develop appropriate impact indicators to evaluate both benefits and costs of using AI in such Science-for-Policy activities.

Disinformation

The JRC contributes to the EU’s long-standing work on countering disinformation and foreign information manipulation and interference. A JRC-led study recently published in Nature confirmed that both pre-bunking and debunking can be effective when countering fallacious statements about climate change and COVID-19 vaccines. Both can reduce agreement with false claims, their assessment as credible, and the likelihood of sharing misinformation. JRC is soon publishing a study on disinformation trends analysing 20 million news articles covering Ukraine and Russia from 2013 until 2024. It will also publish an update of its report on Technology and Democracy.

The OECD Agenda

To this end the OECD will address several elements of this agenda as part of its work on governance, drawing on multiple policy workstreams as part of its programme of work and budget:

Governing with AI

The OECD is also very active in the AI area, with a forthcoming major publication exploring the potential for AI in the public sector, including how to leverage AI for evidence informed policymaking and evaluation.

Advancing agile approaches for evidence-based policymaking

With the adoption of the Recommendation on Agile Regulatory Governance to Harness Innovation (OECD, 2021[2]), the OECD is advancing strategic and country level support to advance the adoption of processes and tools that increase flexibility and data-driven decision making for regulatory policy. This includes advancing the use of digital tools to improve the design and administration of regulation, building guidance on strategic foresight and regulatory experimentation (e.g. sandboxes, prototyping) and increasing the agility of regulatory manage tools like RIA and ex post evaluations.

Measuring and assessing policy evaluation practices

Following the adoption of the Recommendation on Public Policy Evaluation, and the public release of the Policy Evaluation Toolkit in early 2025, the OECD will be undertaking further analysis to measure and assess the implementation of the Recommendation. The toolkit is designed to help operationalise the goals and principles of this Recommendation, and provide practical guidance and examples for implementation. It is intended for a wide range of users, including government officials, evaluation workers at both national and subnational levels, legislators and regulators. It can be used to reflect on and strengthen evaluation systems, assess specific components of an evaluation system, and identify areas for improvement.

The OECD remains committed to measuring and assessing progress with implementing the Recommendation, which will address a number of issues related to assessing and measuring EIPM in the specific area of policy evaluation. These are undertaken with the support of the OECD expert group on policy evaluation, which gathers experts from a large number of EU Member States (OECD, 2025[3]).

Addressing the mis and disinformation space

In the related area of mis- and disinformation, the OECD will support countries in the implementation of the Recommendation of the Council on Information Integrity adopted in December 2024. This includes strengthening societal resilience, enhancing the transparency, accountability and plurality of information sources, and upgrading institutional architectures and open government practices (OECD, 2024[4]).

Engaging with Centres of Government, national schools of government and behavioural expert communities

The OECD also maintains peer-networks that will continue to help identify and implement good practices in the public governance area, that can also support EIPM ecosystems. The OECD is currently leading a multi-country TSI project on Centres of Government, including six European Member States, with a focus on “Enhancing Centre of Government capacities to steer complex priorities”. The annual network of Centres of Government has the potential to address key topical issues for evidence informed policymaking, including trust, governing with AI and the Recommendation on Information Integrity . Schools of government can also be engaged in efforts to strengthen relevant skills. Finally, OECD work on behavioural insights can also be mobilised for mainstreaming behavioural science into policymaking in the future.

Government at a Glance

OECD regularly publishes a comprehensive overview of data on governance practices across OECD countries in the bi-annual Government at a Glance publication. Data within the publication is high-quality and internationally comparable, and is collected using OECD survey tools. The bulk of information within the report is collected directly from government officials responsible for a relevant public governance process. The report provides countries with a benchmark to understand the effectiveness of their governance processes in comparative perspective.

Besides the cross-cutting governance related areas above, the OECD will also continue to support green economic and social agendas through its multiple thematic, economic and social committees, providing best practices and cross-country knowledge and engaging both with European and non-European countries to produce global references that are useful for policy.

This agenda reaffirms the commitment of the JRC and OECD to further support a resilient and evolving European practice of Evidence-Informed Policymaking - one that is fit to face the challenges of tomorrow.

References

[3] OECD (2025), “Public Policy Evaluation Implementation Toolkit”, https://www.oecd.org/content/dam/oecd/en/publications/reports/2025/02/implementation-toolkit-for-the-oecd-recommendation-on-public-policy-evaluation_f24516be/77faa4fe-en.pdf (accessed on 16 April 2025).

[1] OECD (2024), Facts not Fakes: Tackling Disinformation, Strengthening Information Integrity, OECD Publishing, Paris, https://doi.org/10.1787/d909ff7a-en.

[4] OECD (2024), “Recommendation of the Council on Information Integrity”, OECD Legal Instruments, OECD, Paris, https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0505 (accessed on 16 April 2025).

[2] OECD (2021), “Recommendation of the Council for Agile Regulatory Governance to Harness Innovation”, OECD Legal Instruments, OECD, Paris, https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0464.

Note

← 1.  The EU has a framework in place to promote and ensure scientific integrity. The framework is based on several key documents, initiatives and regulations, such as the European Code of Conduct for Research Integrity and the EU’s Open Science Policy. Member States have similar codes of conduct, as well as legal declarations that safeguard the independence of individual advisory bodies and research organisations.