This chapter outlines a blueprint for developing measurement agendas to strengthen the evidence base on how science and innovation contribute to sustainable growth. Rather than prescribing a single framework, it offers a flexible structure grounded in three strategic pillars. The first pillar focuses on principles for designing measurement agendas, including directionality, data integration, and shared ownership. The second identifies key gaps in existing metrics - such as the uptake of green technologies and the role of public sector innovation - while highlighting opportunities for expanding coverage. The third pillar advances the evaluation of policy impacts by promoting systems of indicators and embedding assessment in policy design. Together, these pillars provide a foundation for more coherent and actionable measurement strategies across countries and institutions.
Measuring Science and Innovation for Sustainable Growth
Abstract
Improving the evidence base to monitor the capacity of science and innovation policies to contribute to sustainable growth requires purposeful and dedicated measurement agendas. A measurement agendas in this area can provide a structured framework for collecting and analysing data, help inform decisions and policies, and prioritise steps towards addressing key gaps and opportunities. It can also facilitate better communication among stakeholders by developing a shared language and metrics to discuss performance and outcomes. Most importantly, an actionable measurement agenda must assign responsibilities for its design, implementation and subsequent review to ensure it fulfils its intended purpose.
Rather than attempt to produce a definitive measurement agenda for the OECD and/or its member countries and partners, this chapter provides a blueprint for future agendas. This blueprint proposes recommendations, set under broader lines of action, with the potential to advance the capacity of countries, ministries, departments and other relevant actors to monitor the contributions of science and innovation to sustainable growth and the role of policies.
The proposals are built upon three main pillars:
Pillar 1 sets out principles for developing and implementing measurement agendas at different levels. This includes recommendations for expanding the possibilities and methods of measurement. Specifically:
Build awareness and plan for multiple purposes and uses of data and statistics.
Instil directionality in measurement to drive action and results.
Make the most of the opportunities and synergies provided by different data sources and methods.
Establish shared ownership and partnerships.
Pillar 2 lays out key gaps in the measurement of science and innovation for sustainability in the existing offering of indicators, stemming from the analysis in this publication, and also offering great promise for providing much-needed insights. Specifically:
Recognise and measure the multifaceted contribution of upstream scientific research.
Measure actual adoption and use of green technologies and practices.
Monitor government support for environmental innovation beyond R&D.
Monitor environmental innovation beyond business.
Develop measurement and monitoring frameworks that capture systemic features of science and innovation’s role in the economy and the natural environment.
Monitor the direct resource and environmental footprint of STI activities.
Pillar 3 goes beyond single indicators as the sole output of measurement and provides recommendations for measuring impacts and evaluating policies that may shape the contribution of science and innovation to sustainable growth. Specifically:
Develop systems of indicators rather than aim for one-size-fits-all approaches.
Develop policy research goals and embed evaluation into policy.
Measuring the contribution of science and innovation to sustainable growth, particularly to energy and environmental transition goals, is key for designing and implementing evidence-based policies today, not only by those notionally in charge of formal science and research policies, but also by those working in different policy domains. However, measuring this contribution is complex, partly because measurement frameworks for science, technology and innovation (STI), environment and economy often treat these as completely separate domains, and links across them are uncertain and rarely apparent or comprehensive in available data sources. The measurement challenge is thus threefold:
Assert that science, technology, and innovation are being measured.
Assert that the measured objects have reasonably well-defined implications for natural resources and the environment.
Undertake both assessments jointly for the same activities.
This makes the measurement challenge a rather distinct one from, say, assessing the contribution of research and innovation to productivity, where already available and coexisting measures of investment and economic performance at the level of firms or industries can be used to make direct links.
By showing that there are several indicators and communities of practice helping track the role of a plural set of actors and activities in contributing to better environmental outcomes on a par with economic growth, this publication confirms there is considerable potential for improving our shared understanding of how science and innovation contribute to sustainable growth. However, it also shows that relatively few indicators are widely available internationally for reliable comparisons, while several impact channels are far from being characterised through reliable, quantifiable and well understood indicators.
This situation and the global nature of the natural resource and environmental science and innovation challenges under study make a strong case for collective and focused measurement action to address these evidence gaps. Measurement is a priority not only for those interested in international statistical benchmarking but also for a wide range of stakeholders operating in different contexts, within and outside governments, and it concerns data users and producers. The “evidential” interests of statisticians, programme administrators, regulators, business intelligence providers, academics, civil society organisations and policymakers converge in many instances. Still, there may be trade-offs between serving different goals. It is therefore possible to envisage multiple measurement agendas working in parallel with both common and distinctive elements that can reinforce each other.
Measurement agendas provide a structured framework for collecting and analysing data, which helps inform evidence-based decisions and policies, prioritising action on gaps and opportunities, taking into account what other actors are doing in this space. Working to develop and put in place a measurement agenda can facilitate better communication among stakeholders by providing a common language and a set of metrics to discuss performance and outcomes. Most importantly, it must assign responsibilities for its design, implementation and review to ensure it fulfils its intended purpose.
This monograph does not aim to produce a definitive measurement agenda for the OECD and/or its countries and partners, but aims to foster a discussion among those with the motivation and interest in developing and implementing one. This chapter thus provides a blueprint to support such development, setting out lines of action and recommendations with the potential to advance the capacity of countries, ministries, departments and other relevant actors to monitor the contributions of science and innovation to sustainable growth and its impacts. The proposals, which are informed through this publication’s analysis, are thus built upon three main pillars (Figure 6.1). They are as follows:
Pillar 1 sets out principles for developing and implementing measurement agendas at different levels, from those that define the approach followed by national or regional governments, to those potentially adopted by the OECD. This includes recommendations for expanding the possibilities and methods of measurement towards the measurement agendas’ ultimate goals.
Pillar 2 lays out key thematic evidence gaps in the existing offering of indicators, many of which have been identified throughout Chapters 2 to 5 in this publication as promising to improve understanding and provide relevant new insights.
Pillar 3 goes beyond single indicators as the usual focal output of measurement and provides recommendations for measuring impacts and evaluating policies that may shape the contribution of science and innovation and related policies to sustainable growth.
Source: Authors’ own elaboration.
Build awareness and plan for multiple purposes and uses of data and statistics
Measurement can serve multiple purposes.
Track progress: By establishing clear metrics and indicators, measurement allows organisations and countries to monitor their progress towards specific goals and objectives.
Inform and potentially guide decision making, making it possible to assess choices retrospectively and look ahead.
Ensure accountability: Measurement helps ensure that all stakeholders are accountable for their roles and responsibilities by providing transparent and measurable outcomes.
There is no single user of statistical and quantitative measurement of the contribution of science and innovation to sustainable growth. For example, within governments, potential uses range from strategy definition, legislation and budget prioritisation at the highest levels, to programme administration and regulatory enforcement activities that must assess merit and compliance in a very concrete and localised fashion. Outside government, information about the resource and environmental performance of new knowledge, processes and products is particularly relevant for business as a matter of strategy and competitiveness. At the same time, investors and consumers also rely on information to guide their choices (Box 6.1). Researchers interested in studying the drivers and dynamics of environmental innovations and their economic, ecological and social impacts look at securing data to inform their work, sometimes developing their own. The research community more broadly strives to demonstrate the relevance and impact of their work to their public and private funders.
Sustainability reporting standards are crucial in advancing environmental innovation by providing a structured framework for organisations to measure, disclose and improve their environmental performance. Standards such as the Global Reporting Initiative, the Task Force on Climate-related Financial Disclosures, and the Sustainability Accounting Standards Board ensure transparency and accountability in corporate sustainability efforts. By allowing businesses to report on resource use, emissions and sustainability initiatives, these frameworks encourage companies to adopt greener technologies, optimise processes, and invest in innovative solutions that reduce environmental impact. Transparency helps businesses track their progress and fosters competition, pushing industries toward more sustainable and innovative practices (Widyawati, 2019[1]).
Furthermore, sustainability reporting standards influence investor and consumer decisions. Investors increasingly rely on sustainability reports to assess long-term risks and opportunities, favouring companies that integrate environmental responsibility into their strategies. Similarly, consumers are more likely to support brands with strong sustainability commitments, encouraging businesses to develop eco-friendly products and services (see Chapter 5). By creating an incentive to align corporate strategies with sustainability goals, reporting standards help create market conditions where developing environmental innovations is rewarded (Christensen, Hail and Leuz, 2021[2]).
The business case for measurement can be more easily articulated if all its uses and implications can be examined holistically in conjunction with the relevant data sources available. Data uses drive measurement, e.g. administrative reporting supports compliance with regulations or allows claims to benefit from targeted support. As a result, a measurement agenda needs to heed the diverse range of purposes that better statistical and quantitative evidence can bring about, informed by full awareness of the broader context within which information about new knowledge and new applications of knowledge relating to resource and environmental outcomes exist.
Instil directionality in measurement to drive action and results
The 2010 OECD Innovation Strategy’s Measuring Innovation (OECD, 2010[3]) set out as its Action 5 the goal to “promote the measurement of innovation social goals and the social impacts of innovation”. In order to address questions about STI’s contribution to key societal goals, a comprehensive measurement agenda requires that “directionality” be embedded in its measurement. Paraphrasing the notion of policy directionality towards transformational goals (OECD, 2024[4]), “measurement directionality” means, in this particular context, being prepared to implement measurement and analysis instruments that are uniquely tailored to capture the specificities of resource and environmentally-focused science and innovation and sustain them for a reasonable period. The converse non-directional approach is one where one expects data based on generic measurement instruments to be ready for compilation and processing to serve concrete thematic needs. In a non-directional context, data users assume that data and indicators will ultimately be available as they expect producers to anticipate user needs. However, data producers may lack the resources, mandate or feedback required to act on such potential requirements, regardless of the sheer importance of the subject matter. Implementing measurement directionality may entail new questions in surveys, data infrastructures and methodologies that are fit for the intended purpose, as well as focused resource allocation and co-ordination among the relevant actors towards set goals. If they are not sustained, it is unlikely that reporting systems will effectively adapt and embed environmental considerations to yield dependable measurements.
Measurement directionality is particularly important for natural resource and environmental objectives since goals do not necessarily equate with other frequently measured dimensions, such as the disciplinary or sectoral distribution of STI activity. These frameworks’ classification schema may also treat objectives and outcomes as mutually exclusive categories, as this publication has shown about measuring socio-economic objectives of research and development (R&D), hiding the broader relevance of science and innovation activities through their multiple impact channels. The previous chapters have also provided ample evidence for the significant gap between measuring explicit policy goals and relevance. Accounting for this makes it possible to investigate the double or multiple “dividend” of investments in environmental science and innovation.
Supposedly neutral measurement approaches, although permitting and encouraging specific thematic developments, can act as “comfort” spaces, preventing effective action. The lack of effective development of horizontal measurement frameworks into concrete domains may result in them being perceived as inappropriate lenses to look at the environmental relevance and impacts of research and innovation. Without some reliable degree of directionality, generic recommendations to capture potential or actual effects of science and innovation are unlikely to be consistently followed through by the multiple actors that underpin measurement. STI measurement directionality for sustainable growth requires developing and applying measurement strategies specifically tuned to those agendas’ purpose.
Make the most of the opportunities and synergies provided by different data sources and methods
Address ambiguity through clear and implementable measurement concepts, definitions and taxonomies
The implementation and communication of energy and environmental innovation measurement is challenging as it implies working at the interface of multiple areas that already use somewhat abstract and broad concepts that are also hard to operationalise into tangible measures. Terms like innovation and technology are often used indistinctively, and, when proxied with specific indicators, insufficient attention is paid to explaining what these cover and exclude in many cases. The same applies to the dimensions of environmental relevance, where the measurement task can be negatively impacted by widespread ambiguity and the use of normative values in the adjectives and qualifiers. For example, the indiscriminate and unqualified use of adjectives such as “green” or “clean” implies a net positive environmental effect, when in practice, only one dimension of environmental impact may apply to what is specifically measured (e.g. renewable energy or low-carbon emission). The use of specific terms relating to environmental implications of science and innovation in surveys needs to be as neutral as possible to ensure unbiased participation and reporting, bearing in mind the coexistence in the measurement population of incentives towards “greenwashing” on the one hand, and views rejecting potentially stringent environmental policies and the evidence those are based on.
Another important dimension of conceptual ambiguity that a measurement agenda in this area needs to address is communicating the distinction between environmental intentions, relevance and outcomes of scientific and innovation activities. This publication has provided examples where measures based on explicit natural resource or environmental objectives underestimate measures of effective relevance. Different measurement approaches and perspectives will shed light on specific elements. For example, budgetary processes reveal intentions and priorities, while those carrying out the scientific and innovation activities may be better positioned to judge the types of goals their research may be relevant for. Likewise, companies can characterise their innovations’ resource efficiency and environmental protection attributes, but may not be well positioned to characterise their overall impact on ecosystems.
In addition to promoting transparency, a measurement agenda in this area needs to support the formulation of measurement concepts, definitions and taxonomies as a shared language and work through agreements to ensure that these are as interoperable as possible with multiple measurement perspectives, as opposed to relying solely on source-specific concepts and definitions.
Make focused and co-ordinated use of statistical surveys
Surveys are the cornerstone of statistical data – especially official data – on research and innovation. Compared to data arising “organically” from administrative or commercial processes, survey data are “designed”. Survey instruments need to be cautiously focused and rationed in the digital world, as individuals’ and organisations’ ability and willingness to provide the requested information may be limited (OECD, 2018[5]).
The range of specific science and innovation survey data sources and their ability to collect additional information is limited. Still, it offers considerable space for action across most countries, provided that instruments strike a delicate balance between “business as usual” approaches and developing new indicators. Furthermore, there needs to be broader awareness of surveys in other statistical domains and their points of connection with science and innovation for environmental goals. An example would be measuring economic activities among companies using specific energy or environmental technologies, or expenditures by firms or households on those. These can be exploited to address shared questions of interest among the owners of different measurement vehicles, which requires more frequent dialogue among those with survey measurement responsibilities (Box 6.2).
Another important lesson of the experience on surveys focused on environmental innovations or the adoption of relevant advanced technologies is that the informational value of official statistics produced in this fashion is limited if countries do not minimally co-ordinate their approaches. The benchmarking information value of data often stems from being able to compare with similar industries in other countries, not across industries within a country. A measurement needs to incorporate plans for making better use of scarce statistical survey resources, focusing on the elements of environmental research and innovation that cannot be measured otherwise, and indeed be more internationally co-ordinated to avoid missing out on the value of international comparisons.
The National Council for Statistical Information (CNIS) is a French organisation that facilitates dialogue between producers and users of public statistics. The CNIS helps ensure that statistical data in France is accurate, relevant and meets the needs of various stakeholders. Its main functions include formulating user needs for statistical data, ensuring collaboration, evaluating surveys and advising on statistical methodologies and practices.
In October 2024, the CNIS held a discussion on the measurement of R&D and innovation in support of the green transition (transition écologique). This meeting addressed the state of play of measuring efforts of private actors through available sources, bringing together presentations on multiple initiatives across government led by different departments and agencies and the practical and methodological challenges encountered. While the meeting did not lead to formal recommendations, the meeting’s conclusions laid out some main lines of action to be pursued by different actors, highlighting the importance of record linking and the distinction between activities for descriptive measurement and the assessment of impacts.
Source: Authors, based on CNIS (2024[6]), Entreprises et stratégies de marché 2024 – 2e réunion – R&D et innovation en faveur de la transition écologique, https://www.cnis.fr/evenements/entreprises-et-strategies-de-marche-2024-2e-reunion-rd-et-innovation-en-faveur-de-la-transition-ecologique/ (accessed March 2025).
Make effective and transparent use of big data
Digitalisation has proven to be a major force for change in the generation and use of data and statistics. Science and innovation systems have become remarkably data-rich compared to the past, as information on innovation inputs and outputs that was only recorded in highly scattered paper-based sources is now much easier to retrieve, process and analyse, particularly through the Internet. Data analysis methods have similarly developed, blurring the boundaries between qualitative and quantitative data, working on rich and often publicly available data (OECD, 2018[5]).
This publication illustrates some of the applications of this new wave of indicators and discusses their advantages and limitations. Natural language processing techniques have helped this publication draw several relevant linkages between research activities and the Sustainable Development Goals (SDGs). Still, several others can be further explored. The possibilities of generative artificial intelligence (AI) make it tempting to prompt applications to provide rankings of leading companies and countries in terms of environmental technologies, for example, but how could those be considered reliable?
A measurement agenda in this area needs to embrace new measurement possibilities responsibly. The appropriate quality frameworks for the use of publicly disclosed data at the interface of STI and environmental activities, accounting for incentives to report or conceal information, need to be applied according to the purpose of the analysis. More transparent and coherent use of these increasingly popular techniques and alternative forms of data must allow users to properly understand their strengths and limitations.
Leverage opportunities for combining different data sources
A measurement agenda needs to recognise that the usefulness of data increases with the capacity to combine with other data to draw additional inferences, for example, when investigating which factors influence the adoption of low-carbon technologies. New questions do not need to be “expensively” added to surveys if data from a particular survey can be connected with other sources that contain the necessary information (OECD, 2018[5]). Connecting data also makes it possible to assess potential biases incurred when relying solely on one.
The benefits are substantive when combining data, making it possible to realise the complementary strengths of different sources, overcoming the limitations that impact each source separately. For example, Chapters 2 and 4 draw on analysis that combines survey and administrative data, helping identify the relevance of scientific publications and R&D projects to specific SDGs. In this case, the survey data helped train a classification model applicable to administrative data research.
As linking different data sources can provide insights that could not be derived from working separately with the different components, a measurement agenda must also prepare the infrastructure and governance arrangements that enable data integration and its effective use. Ensuring data compatibility is potentially beneficial to policymakers and other stakeholders managing national research and innovation systems and can yield considerable benefits for individuals and organisations doing (or reporting on) research. If an individual data item is made interoperable, it can be reused across multiple systems, allowing it to be provided to authorities only once. An integrated and interoperable system considerably reduces reporting and compliance burden, freeing up more time and money for research.
Effective data stewardship is another key element. National statistical organisations are natural candidates for assuming such responsibility. However, they need to be adequately resourced, staffed, and empowered to fulfil such a goal, distinct from the goal of publishing indicators.
Establish shared ownership and partnerships
The environmental challenges addressed in this publication straddle multiple disciplines and actors. The information “marketplace” on environmental innovation is a complex and evolving one, where the roles of data producers and users often overlap, with several intermediaries offering products to meet the needs of different actors, from advice and accreditation through to rankings, to name a few. This is therefore a shared agenda with several other policy and statistical domains that requires busting silo mentalities while preserving some degree of specialisation, given the complexities within each domain.
Define and distribute ownership over elements of the agenda
Driving advances through a measurement agenda requires establishing ownership over its different components among potential contributors. From the perspective of any given country, there will be different ministries and departments taking the policy lead, and roles may vary depending on the type of scientific and innovation activity considered. The experience summed up in Chapter 4 about measuring support for business research and innovation highlighted that environmental innovation support initiatives are often found within ministries responsible for industry, natural resources, and the environment. Different stakeholders have different possible contributions regarding data production and analysis to make in isolation or in collaboration with others. These should be clearly spelt out in an agenda.
The measurement agenda also needs to recognise the impact of policies and policy practices for and around data on the data available to inform policy. A “policies for evidence” approach in the measurement roadmap will promote shared responsibility among policy and administrative decision makers over data, statistics and evidence on STI and concerted action among them to guide and sustain evidence-building and evidence-using efforts across a wide range of intertwined resources and purposes.
Integrate the measurement of relevant STI activities in existing data strategies for sustainable growth
Rather than work in isolation, measurement agendas need to capitalise on complementary measurement efforts and strategies to promote the energy and green transition. Some countries and organisations already have data strategies for sustainable growth in place (Box 6.3). These often consider technology and innovation as key elements, and include several of the indicators contained in this publication. However, they may not include other important elements or help articulate demand for other indicators in gap areas. Recognition and support from these broader strategies can be elicited to advance the measurement of the contribution of science and innovation to sustainable growth.
Recognising their importance and the difficulty of measuring the economic and environmental impacts of clean technologies as an activity that cuts across all major economic sectors, Canada’s Clean Technology Data Strategy (CTDS) aims to ensure that data are readily available to understand the economic and environmental contribution of clean tech. Established in 2017, the CTDS is a joint initiative led by Natural Resources Canada, Innovation, Science and Economic Development Canada, and the Clean Growth Hub, which supports the collection of data and regular reporting on clean tech activity. The CTDS argues that better data strengthen the evidence base for decisions, improve understanding of the emerging clean tech landscape and ensure the creation of impactful policies and programmes to support the production and adoption of clean technologies.
There are three main components in the strategy and ongoing engagement with key partners:
Development and dissemination of authoritative clean technology statistics, led by Statistics Canada. Expand the collection and production of statistics and macroeconomic indicators on the clean technology economy. This component comprises five main data products, but these do not include the surveys of Advanced Technology Adoption introduced in Chapter 3.
Leveraging industry data, led by Natural Resources Canada. Leverage public information to gather company-level data on the clean tech industry in Canada and conduct surveys to better understand the challenges and opportunities clean tech companies face.
Leveraging administrative data, led by the Clean Growth Hub. Use existing administrative data to track the impact of government programmes that support clean technology.
Source: Authors, based on Government of Canada (2025[7]), Clean Technology Data Strategy, https://ised-isde.canada.ca/site/clean-growth-hub/en/clean-technology-data-strategy.
Build domestic and international partnerships
Within countries, the STI evidence community needs to address the persistent and significant disconnect between different users and producers of STI data, statistics and analysis (OECD, 2018[5]). Building capabilities and encouraging co-ordination among different actors will be necessary to allow new data infrastructures to emerge. Most solutions aiming to build infrastructures that transform evidence capabilities rely on social change rather than technology, underpinned by community engagement and trust building. Driving progress requires identifying the major obstacles to developing evidence infrastructures – often starting within public administrations, where data are fragmented and synergies foregone. Lack of policy awareness can block improvements to the legal framework for data exchange and reuse. It can also hinder the implementation of sustainable “business models” for data, which consider the intended statistical use by policymakers. Advancing towards meeting the ambitions of the agenda requires strong domestic partnerships and collaboration, both within government and with other stakeholders participating at the interface of the information “marketplaces” for innovation, natural resources and sustainability.
International collaboration around priority setting and implementation of measurement is particularly important for reasons, such as the transboundary nature of the innovation and environmental phenomena that need to be measured, and policy demand for international benchmarking. OECD and other international organisations can play an important role in this space.
This publication’s systematic presentation of available indicators and cases has exposed significant gaps and opportunities for improved coverage and understanding of key aspects of science and innovation’s contributions. While there has been continuous progress in terms of expanding understanding of the strengths and weaknesses of the science and innovation supply chain, including with respect to natural resource environmental issues, important blind spots remain that need to be tackled by measurement agendas. The following elements do not represent an exhaustive list and would need to be adjusted for any given measurement agenda to reflect the priorities of the relevant stakeholders.
Recognise and measure the multifaceted contribution of upstream scientific research
Measurement of research contributions to energy and environmental activities, to those with an apparent connection, provided an overly narrow view of the impact pathways for research. A measurement roadmap needs to account for the role played by scientific research not only as a provider of critical information about the natural environment and as an enabler of more radical, transformational technological innovations, but also as a vital multi- and inter-disciplinary input to policy making and citizen engagement.
This publication has demonstrated several approaches to enable consistent measurement of such contributions, providing a more comprehensive view of the role of research. This line of work can be enhanced in several directions, including through a more detailed appreciation of knowledge and service flows.
Measure actual adoption and use of green technologies and practices
This publication has also shown in Chapter 3 that a significant amount of innovation activity and outcomes may be overlooked when relying solely on input-based measures or proxies based on recorded intellectual property rights. This is likely because many environmental innovations involve operational and supply chain modifications aimed at improving resource and energy efficiency, rather than new products, and because imitation in adoption is a common way for firms to innovate.
Several efforts have been made to track progress on environmental innovation using very different approaches. However, many lack coverage, granularity, timeliness and may involve high data collection costs, especially when conducted on a large scale. While there is no international consensus on what should be the precise recurring measurement and reporting of innovation according to its environmental or societal impacts, several countries have collected at some point data that helps differentiate companies that have introduced innovation with environmental benefits across a spectrum of potential impacts. These efforts can and should be standardised to ensure coverage of a minimum set of attributes of common interest across countries. It may also be timely to consider a more active use of object-based measurement approaches so as to have more concrete elements to assess the environmental relevance of changes within companies and to be able to connect surveys and other data sources more effectively.
Monitor government support for environmental innovation beyond R&D and innovation in the public sector
Countries worldwide strive to find the appropriate level and balance of public financial support for innovation to address several competing and pressing policy objectives, particularly those at the interface between industrial and environmental policy goals. To ensure effective policy learning, impact and value for money in their investments for innovation, governments need to adopt more effective, transparent and comprehensive national monitoring mechanisms. These should go beyond the assessment of traditional R&D support instruments like grants or tax incentives, putting in place effective monitoring of support provided under instruments that are currently undermeasured from an innovation policy perspective. This type of evidence is key for developing effective support portfolios that strike an optimal balance between upstream and downstream innovation activities.
For example, instruments such as public procurement are regularly alluded to as vital for driving demand for innovative solutions to environmental challenges. Yet, the available data are very limited in terms of coverage and quality, preventing effective examination of the degree of novelty and its contribution to environmental objectives. Monitoring government support for environmental innovation also requires effective work across different policy domains, as innovation is often an enabling step towards energy and environmental transformation goals rather than the ultimate objective for its own sake.
Monitor environmental innovation beyond business
Although innovation is not the sole prerogative of business organisations, measurement of environmental innovation, like innovation more broadly, is primarily concerned with innovations introduced by business enterprises. A comprehensive assessment of the contribution of innovation to sustainable growth also needs to consider the role played directly by government and non-profit institutions in adopting new, environmentally superior solutions. The same applies to households as “institutional units” in the economy, as adopters of green technology. This publication shows in Chapter 5 some indicators about consumer preferences and behaviours, but statistics on those are not as widely available as they are for the business sector.
The fact that actors in these other sectors draw upon solutions that business enterprises commercialise does not deny the case for monitoring the key role these play, both from the demand (driving demand) and supply (furnishing ideas that may have commercial merit) sides. Since measurement of innovation beyond the business sector is not yet common or stable, and requires further testing and experimentation, having a thematic focus on environmental innovation in exploring new approaches could potentially facilitate faster progress than has been accomplished so far.
Develop measurement and monitoring frameworks that capture systemic features of science and innovation’s role in the economy and the natural environment
The energy and green transition challenges require measurement and analysis to carefully account for the emergent properties of innovation at the level of local, regional, national and global systems. These properties do not necessarily result from the averaging or adding up of indicators at the level of individual units, such as firms or individuals, but result from interactions and change among them.
Information on capabilities and resilience in key nodes of supply chain networks can be critical for innovation with energy and environmental objectives. Systemic changes can be the outcome and the channel by which new technologies are adopted, for example, the application of enabling low-carbon technology, such as advanced batteries, across a broad range of uses (mobility, home energy storage, etc). Measuring those in conventional ways is fraught with difficulties because official statistical systems and their guarantees of confidentiality are, as currently set up, not well suited for measuring and reporting links across companies and other actors reporting data.
This publication has barely scratched the surface of such a perspective by providing “indicators” on the interconnected structure of scientific activity and R&D funding across portfolios and the relationships between domains and sustainability goals. Several more key attributes, such as pathways of diffusion and measures of network centrality, contribute to depicting the system-level features and transformations, and these should be the focus of attention and targeted measurement.
Furthermore, the assessment of systemic implications of science and innovation also needs to take into account the transition pathways and the structural adjustments faced by regions and industries. Energy and green transitions may involve paradoxical situations where the outputs of highly polluting industries play a key part in producing intermediate goods and services necessary for new, less polluting technologies. Structural adjustments may also disproportionately impact communities and lead to the accelerated obsolescence of capital and skills.
Monitor the direct resource and environmental footprint of STI activities
When looking at the contribution of science and innovation to sustainable growth, one gap in this publication is the measurement of its own direct and extended footprint. As an economic activity that draws on resources and produces different types of outputs, it has a resource and environmental footprint that needs to be accounted for. Relevant questions this line of work would need to tackle from the start include assessing the feasibility of an operationalisable definition and framework for different types of STI activities, and implementing it. Its results would be used to assess how generic and STI domain-specific policy actions contribute to reducing the environmental footprint of science and innovation, their likely impacts, and potential synergies and trade-offs with other policy objectives.
The resource and environmental footprint of downstream innovation and diffusion activities is perhaps the most challenging to undertake in a comprehensive fashion, since it is difficult to differentiate from the footprint of overall economic activity. Potential priorities for measurement and analysis concern the initial adoption and diffusion of emerging technologies. There is considerable interest, for example, and a growing literature on the impacts of the accelerating adoption of forms of AI with sizeable energy demands in the context of the “twin transitions”. Other key areas of interest include the material resource use requirements and potential pollution associated with introducing new low-carbon technologies.
The measurement of science and innovation’s contribution to sustainable growth should not operate under the assumption that all scientific and innovation contributions are necessarily positive or measurable with certainty at any given time. Therefore, potential environmental uncertainty should also be conveyed with existing or new tools that may need developing.
Discussions at the 2016 OECD Blue Sky conference on the next generation of data and indicators pointed out it was time to stop focusing solely on indicators and consider instead the entire data value chain and data cycle, considering in particular how data can be used to extract insights about the drivers and outcomes of science and innovation, and the policies that might influence their contribution to sustainable growth. A rich data and indicator ecosystem can and should be leveraged to do more than just present one-dimensional views of specific aspects of the state of science and innovation systems.
Develop systems of indicators rather than aim for one-size-fits-all approaches
Capturing the multifaceted contributions of science and innovation to sustainable growth requires developing systems of indicators that describe a broad range of intertwined social, economic and physical phenomena. Once available, effective processes need to be implemented to make sense of the data in combination. There is extremely rich information at the level of individual industries and ecosystems that does not necessarily find measurement counterparts in others, e.g. measures of resource or energy performance associated with specific products or processes. In today’s age, with the available means for large-scale data processing and analysis, it makes sense to take stock of subsystem-specific measures of knowledge generation, innovation and their outcomes using those subsystems’ own performance metrics. Rather than simply juxtaposing those into dashboards, these may be combined and analysed using meaningful economic or physical aggregation and modelling methods, based on explainable and transparent approaches.
In this context, composite indices appear to provide a number of advantages over simple collections of indicators. Reducing the number of indicators on display in a dashboard offers the appearance of simplicity and facilitates initial communication with a wider user base in search of a reference “number” (i.e. policymakers, media and citizens). However, in an emerging area of measurement in need of concerted action at all levels, environmental innovation composite indexes may inadvertently lead data users to believe that the index on offer and its underlying data are a sufficient guide for serving user needs. As with any measurement, the measure should serve the goal of measurement rather than the other way around.
Develop policy research goals and embed evaluation into policy
Articulate policy research goals
There are many unanswered questions when it comes to science and innovation for sustainable growth. Relatively little is known about the interrelationships that characterise the trajectories of environmental vs other innovations and the trade-offs faced by policies. For instance, what is the right balance, from an innovation perspective, between subsidising widespread diffusion of existing low-carbon technologies and supporting investment in the upstream development of radically new technologies? Questions such as this reveal uncertainty about the impact of market scale as a driver for enhanced production efficiency and how long and risky the lag is between experimentation and commercialisation. These are key parameters for modelling and anticipating policy impacts.
Such questions, which are often not answerable with observational measurement, help formulate the necessary data measurements and the additional research efforts that are desirable to invest in. Potential research goals inform and help foster action on the key data gaps presented under Pillar 2, but may also highlight others.
Articulating these research goals publicly helps engage the research community, for example, through research funding calls and sharing key policy questions with those prepared to test the underlying hypotheses. Making relevant data interoperable and accessible under relevant stewardship arrangements is a key enabler of researcher engagement. Researchers are also actively creating data, using different approaches and gaining more recognition for their efforts.
“Design-in” data and monitoring, assessment and evaluation requirements into policies and administration
Policy evaluation is a major driver for measurement and should be part of measurement roadmaps. Evaluation cannot be left to indicators alone since the relevant policy questions often incorporate a counterfactual “what if” scenario. Still, the concept of evaluation can have different meanings in different contexts and at different moments of the policy cycle, from ex ante policy assessment to ex post policy evaluation, passing through the production of key performance indicators for the STI system or many of its components.
To ensure they deliver value for money, it is essential to institutionalise monitoring and evaluation of science and innovation policies and funding instruments with a bearing on environmental outcomes throughout the full policy cycle. Whenever possible and sensible to introduce, policy experimentation can be a way to ensure more accurate learning about policy effectiveness by providing suitable counterfactual scenarios. Cross-country impact assessment can naturally exploit the diversity of policy initiatives.
While different countries have different policy preapproval systems, it is important to ensure that past evaluations, indicators and research comprehensively inform ex ante assessment. The research policy goals highlighted in the previous subsection should therefore anticipate the policy options under consideration as much as possible. Data needs and methods for ex post or impact evaluation need to be part of policy design from inception and be considered and planned before implementation. For example, how will policy inputs, innovation, and environmental outcomes be empirically connected? It is generally advisable that monitoring accompanies policy implementation both throughout and after the implementation, and, in that evaluation phase, incorporates independent contributions.
References
[2] Christensen, H., L. Hail and C. Leuz (2021), “Mandatory CSR and sustainability reporting: economic analysis and literature review”, Review of Accounting Studies, Vol. 26/3, pp. 1176-1248, https://doi.org/10.1007/s11142-021-09609-5.
[6] CNIS (2024), Entreprises et stratégies de marché 2024 – 2e réunion – R&D et innovation en faveur de la transition écologique, https://www.cnis.fr/evenements/entreprises-et-strategies-de-marche-2024-2e-reunion-rd-et-innovation-en-faveur-de-la-transition-ecologique/.
[7] Government of Canada (2025), Clean Technology Data Strategy, https://ised-isde.canada.ca/site/clean-growth-hub/en/clean-technology-data-strategy.
[4] OECD (2024), “OECD Agenda for Transformative Science, Technology and Innovation Policies”, OECD Science, Technology and Industry Policy Papers, No. 164, OECD Publishing, Paris, https://doi.org/10.1787/ba2aaf7b-en.
[5] OECD (2018), “Blue Sky perspectives towards the next generation of data and indicators on science and innovation”, 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.
[3] OECD (2010), Measuring Innovation: A New Perspective, OECD Publishing, Paris, https://doi.org/10.1787/9789264059474-en.
[1] Widyawati, L. (2019), “A systematic literature review of socially responsible investment and environmental social governance metrics”, Business Strategy and the Environment, Vol. 29/2, pp. 619-637, https://doi.org/10.1002/bse.2393.